Evidence Tables

Ellis P, Robinson P, Ciliska D, et al. Diffusion and Dissemination of Evidence-based Cancer Control Interventions. Rockville (MD): Agency for Healthcare Research and Quality (US); 2003 May. (Evidence Reports/Technology Assessments, No. 79.) Evidence Tables.

Evidence Table 1. Systematic reviews of the effectiveness of cancer control interventions in adult smoking cessation (Key Question 1)

Lead Author (Year) Review Purpose Quality Assessment Rating	Inclusion Criteria Dates Articles Reviewed, Number of Studies Included, Meta-analysis Performed?	Results Conclusions
Ashenden, R⁵⁸ (1997)	Inclusion Criteria:	Results:
Review Purpose: To examine how effective lifestyle advice provided by GPs is in changing patient behavior, specifically smoking, alcohol consumption, diet, and exercise behaviors. Lifestyle advice could include the provision of written information in addition to verbal advice.	Published in English	In 13 studies the provision of brief advice (provided within the confines of a single consultation) was compared with a no-advice control group; 3 compared intensive advice (more than a single consultation) with a control group; 3 compared brief advice, intensive advice, and a control; 4 studies compared brief with intensive advice. In 14 studies advice was provided by a GP; in 6 trials advice was provided by GP and another health professional, such as nurse or counselor; and in 3 studies the advice was provided by another health professional
Quality Assessment Rating: Strong	Investigated the effectiveness of lifestyle advice provided in the GP setting	Two studies were eliminated from the analysis because of significant heterogeneity. No sensitivity analysis was performed to examine the effect of including these data. Results are based on data from 21 smoking cessation trials
Refer to Adult Healthy Diet Evidence Table 2 for additional information	Comparison made between either no intervention or usual care, control group, or between advice of differing intensities	Brief advice is associated with a small increase in abstinence rates (OR 1.32, 95% CI 1.18–1.48). The odds of quitting are slightly improved with more intensive (OR 1.46) interventions compared with brief interventions (OR 1.27).
	Followup for smoking cessation trials must have been minimum 6 months after therapy	Based on data for 16,385 subjects, it would be necessary to provide advice (either brief or intensive) to 35 smokers to produce 1 quitter; 50 patients would need to receive brief advice or 25 patients would need to receive intensive advice
	Dates of Articles Reviewed: Inception of database to May 1995	Data from 7 studies showed no added benefit of intensive advice compared with brief advice (OR 1.07, 95% CI 0.88–1.29)
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 37	GP-based health programs have a modest and variable effect on health outcomes such as lifestyle change
	Total RCTs: 37	More extensive and rigorous research is required
	Studies focusing on adult smoking cessation: 23	A greater number of GPs will need to become involved in promoting behavior change than the literature is currently suggesting if GP-based interventions are to be effective in public health
	RCTs focusing on adult smoking cessation: 23
	Meta-analysis Performed? Yes
Bains, N⁵⁹ (1997)	Inclusion Criteria:	Results:
Review Purpose: To evaluate the effectiveness of financial incentives and competitions (e.g., holidays, cash prizes) on:	Published in English	All 17 studies provided information on quit rates. Quit rates among individuals begin high (mean quit rate 34% at 1 month) but decrease over time (mean quit rate 23% at 1 year)
Participation in population-based smoking cessation programs	Study was an evaluation of a financial incentive-based smoking cessation program and participation and/or quit rates were reported	Five studies did not provide information on the proportion of the eligible population participating in the contests
Quit rates in population-based smoking cessation programs	Studies of financial incentives for worksite smoking cessation programs were specifically excluded	It is estimated that financial incentives and competitions may attract 1–2% of eligible populations. There is no evidence that particular types of incentives are able to influence participation or quit rates, but the size of the incentive does appear to be important, with larger incentives viewed as more effectively motivating smokers to quit or stay smoke-free
Quality Assessment Rating: Strong	Dates of Articles Reviewed: 1975 and spring 1997	Conclusions:
	Number of Studies Included:	Incentive-based smoking cessation programs that target an entire community have the advantage of reaching a large and diverse group; however, they may only attract those individuals who are already motivated to quit
	Total Number of Studies: 17
	Total RCTs: 1
	Studies focusing on adult smoking cessation: 17
	RCTs focusing on adult smoking cessation: 1
	Meta-analysis Performed? No
Fiore, MC, et. al,⁶⁰U.S. Department of Health and Human Services Public Health Services Clinical Practice Guideline (2000)	Inclusion Criteria:	Results:
Review Purpose: To provide a clinical practice guideline summarizing effective experimentally validated tobacco use dependence treatments and practices. This guideline is an updated version of the 1996 Smoking Cessation Clinical Practice Guideline #18.	Published in English	Screening systems to increase the assessment and documentation of smoking status significantly increase the rate of clinician intervention among patients who smoke (9 studies, OR 3.1, 95% CI 2.2–4.2), but this is not associated with a significant increase in smoking abstinence (3 studies, OR 2.0, 95% CI 0.8–4.8)
Quality Assessment Rating: Strong	RCT on the patient level; of a tobacco use treatment intervention	Physician advice to quit smoking is associated with increased abstinence (7 studies, OR 1.3, 95% CI 1.1–1.6). Abstinence rates increase with increasing duration of contact (43 studies): < 3 minutes (OR 1.3, 95% CI 1.01–1.6); 3–10 minutes (OR 1.6, 95% CI 1.2–2.0); > 10 minutes (OR 2.3, 95% CI 2.0–2.7)
Related Articles:	Followup >5 months after quit date	More than 1 person-to-person treatment session is associated with increased abstinence rates (45 studies): 2–3 sessions (OR 1.4, 95% CI 1.1–1.7); 4–8 sessions (OR 1.9, 95% CI 1.6–2.2); > 8 sessions (OR 2.3, 95% CI 2.1–3.0)
Fiore, MC⁷⁵ (2000)	Dates of Articles Reviewed: 1975 - January 1, 1999	Smoking cessation interventions (not defined) delivered by more than one clinician type increased abstinence rates (37 studies): 1 clinician type (OR 1.8, 95% CI 1.5–2.2); 2 clinician types (OR 2.5, 95% CI 1.9–3.4); 3 or more clinician types (OR 2.4, 95% CI 2.1–2.9)
Fiore, MC¹⁶⁴ (1996)	Number of Studies Included:	Self-help strategies including reactive telephone counseling (OR 1.2, 95% CI 1.02–1.3), proactive telephone counseling (OR 1.2, 95% CI 1.1–1.4), group counseling (OR 1.3, 95% CI 1.1–1.6) and individual counseling formats (OR 1.7, 95% CI 1.4–2.0) are associated with increased smoking abstinence rates (58 studies)
Anonymous¹⁶⁵ (2000)	Not clearly stated	Smoking cessation interventions that are delivered in multiple formats increase abstinence rates (54 studies): 1 format (OR 1.5, 95% CI 1.2–1.8); 2 formats (OR 1.9, 95% CI 1.6–2.2); 3 or more formats (OR 2.5, 95% CI 2.1–3.0)
	Meta-analysis Performed? Yes	Several types of counseling and behavioral therapies (not fully defined) are associated with increased abstinence rates (62 studies): intra-treatment social support (OR 1.3, 95% CI 1.1–1.6); extra-treatment social support (OR 1.5, 95% CI 1.1–2.1); general problem solving (OR 1.5, 95% CI 1.3–1.8); aversive smoking (OR 1.7, 95% CI 1.04–2.8)
		Conclusions:
		The following topics regarding formats require additional research:
		Identify which combinations of formats are effective; the efficacy of innovative approaches to self-help such as individualized computerized interventions; the efficacy of reactive hotlines/help lines; the relative efficacy of different types of self-help interventions; the efficacy of self-help materials as adjuvant treatments
		Do they add significantly to the effectiveness of other proven tobacco dependence treatments such as individual counseling, group counseling, proactive telephone counseling, and pharmacotherapy?
Hopkins, DP⁶¹ (2001)	Inclusion Criteria:	* For further explanation of grading of evidence, refer to Briss et al¹⁶⁶.
Review Purpose: To review the evidence for selected population-based tobacco use cessation interventions for efficacy, applicability, and other effects to form the basis of recommendations by the Task Force on Community Preventive Services regarding their use	Published in English	Results:
Quality Assessment Rating: Strong	A primary study which took place in an industrialized country and addressed at least one area in the conceptual framework (ETS, initiation, cessation)	Strong scientific evidence (17 studies) demonstrates that increasing unit price for tobacco products is effective in increasing tobacco use cessation and reducing consumption. The median estimate of price elasticity (percent change in consumption for 1% increase in price) was 0.41, i.e., 10% increase in price would lead to 4.1% decrease in consumption
Related Article:	Met evidence review and Community Guide definition of the interventions	Strong scientific evidence (15 studies) demonstrates that mass media education campaigns combined with other interventions (excise tax, community education, distribution of self-help materials, or individual counseling) are effective in increasing tobacco use cessation and reducing consumption. The median quit rate observed in 5 studies was 2.2% (in comparison with individuals not exposed to media campaign). The median decline in state-wide tobacco sales observed in 3 studies was 15 packets per capita per year
Briss, PA¹⁶⁶ (2000)	Provided information on > 1 outcome related to the analytic framework	Insufficient evidence (9 studies) exists to assess the effectiveness of cessation series (mass media interventions that use recurring segments to recruit, inform, and motivate tobacco product users to quit) in reducing tobacco use
	Include an exposure and a comparison group	One study evaluating media cessation contests met the inclusion criteria. Self-reported cessation was increased at 6 months (3.3% increase in quit rates). This evidence was considered insufficient to asses the effectiveness of media cessation contests
	Dates of Articles Reviewed: 1980 - May 2000	Sufficient scientific evidence (7 studies) documents that provider reminder systems (chart prompts, stickers, expanded “vital signs” to include smoking status, and flow sheets) implemented alone are effective in increasing provider delivery of advice to quit to tobacco-using patients. The studies report a median increase of 32.5% in the documentation of smoking status and median increase of 13% in delivery of advice to quit smoking
	Number of Studies Included:	Insufficient evidence (16 studies) exists to assess the effectiveness of provider education interventions when implemented alone. Five studies reported on provider determination of smoking status (median 8% increase). Ten studies reported on provider delivery of advice to quit (median 2.2% increase)
	Total Number of Studies:17	Strong scientific evidence (31 studies) demonstrates multicomponent health care system interventions that include a minimum of a provider reminder system and provider education program are effective in increasing both provider delivery of advice to quit (median increase 20%) and patient tobacco use cessation (median 4.7% increase). Additional effectiveness was demonstrated by studies that also included patient education, such as self-help cessation materials (median 22% increase in provider advice to quit and 5.7% increase in cessation)
	Total RCTs: unclear	Insufficient evidence (3 studies) exists to assess the effectiveness of provider assessment and feedback interventions. Improvements in provider recognition of tobacco use were noted (median increase 21%). However, neither provider delivery of advice to quit or patient tobacco use cessation were assessed
	Studies focusing on adult smoking cessation: unclear	Sufficient evidence (5 studies) demonstrates reducing out-of-pocket costs for effective cessation therapies increases both use of the effective therapy (median 7% increase in use of NRT) and patient tobacco use cessation (median 7.8% cessation rates from 6 to 12 months)
	RCTs focusing on adult smoking cessation: not reported	Strong scientific evidence (32 studies) demonstrates telephone cessation support is effective in increasing tobacco use cessation when implemented with other interventions (patient education, provider delivered counseling, NRT, smoking cessation clinic) in both clinical and community settings. The median increase in tobacco cessation was 2.6%
	Meta-analysis Performed? No	A description of the attributes used to develop the criterion for recommendations was reported
Lancaster, T⁶⁴ (2001)	Inclusion Criteria:	Results:
Review Purpose: To assess the effectiveness of training health professionals to deliver smoking cessation interventions	RCTs in which the intervention was training health care professionals in methods to promote smoking cessation in their patients	Training health care providers increased the smoking cessation activities of health professionals. Trained professionals were 1.5 to 2.5 times more likely to counsel patients about smoking cessation and to initiate interventions such as a quit date, suggesting a followup appointment, and offering self-help material or nicotine gum
Quality Assessment Rating: Moderate	Dates of Articles Reviewed: Not clearly stated	Six of 8 studies found no effect of training on quit rates. One study reported quit rates at 1 year of 8.8% vs. 6.1% and 4.4% for 2 control groups
Related Article:	Number of Studies Included:	The use of prompts and reminders increased the frequency of health professional intervention, but only 1 of 3 studies observed an increase in abstinence rates
Silagy, C⁷³ (1994)	Total Number of Studies: 10	Conclusions:
	Total RCTs: 10	Training health professionals to provide smoking cessation interventions had a measurable effect on professional performance but inconsistent effects on abstinence rates
	Studies focusing on adult smoking cessation: 10
	RCTs focusing on adult smoking cessation: 10
	Meta-analysis Performed? No
Lancaster, T⁶² (2001)	Inclusion Criteria:	Results:
Review Purpose: To determine the effectiveness of more intensive counseling delivered by a smoking cessation counselor to a patient on a one-to-one basis	RCT or quasi-RCT, minimum followup of 6 months, at least 1 treatment arm consisted of unconfounded intervention from a counselor trained in smoking cessation	Individual counseling by a smoking cessation counselor was more effective than control (OR 1.55, 95% CI 1.27–1.90). This result was robust to a sensitivity analysis. There was no evidence that more intensive counseling was more effective than brief counseling (OR 1.17, 95% CI 0.59–2.34). There was no evidence of a difference in effect between individual counseling and group therapy in addition to NRT (OR 1.33, 95% CI 0.83–2.13)
Quality Assessment Rating: Strong	Individual counseling of any smokers except pregnant women, defined as face-to-face encounter with counselor trained in assisting smoking cessation Sustained abstinence, or two-point prevalence used where available	Conclusions:
	Dates of Articles Reviewed: Updated to October 2000	There is consistent evidence that individual counseling from a smoking cessation specialist increases the likelihood of cessation compared with less intensive support. While most of the trials were undertaken in hospitalized smokers, counseling was also shown to be effective in a workplace setting and amongst community volunteers
	Number of Studies Included:
	Total Number of Studies: 11
	Total RCTs: unclear
	Studies focusing on adult smoking cessation: 11
	RCTs focusing on adult smoking cessation: unclear
	Meta-analysis Performed? Yes
Lancaster, T⁶³ (2001)	Inclusion Criteria:	Results:
Review Purpose: To determine the effectiveness of different forms of self-help materials (written materials, video, audiotape) compared with no treatment and with other minimal contact strategies; the effectiveness of adjuncts to self-help; and the effectiveness of tailored approaches compared with non-tailored approaches	RCT, minimum followup of 6 months, for any smokers except pregnant smokers	Nine trials comparing self-help materials provided by post with no intervention reported a significant increase in abstinence rates (OR 1.23, 95% CI 1.02–1.49)
Quality Assessment Rating: Strong	At least 1 arm consisted of a self-help intervention without repeated face-to-face therapist contact	Five trials of self-help materials and brief contact reported no significant increase in abstinence rates (OR 1.19, 95% CI 0.96–1.49)
	Self-help interventions defined as any manual or program to be used by individuals to assist a quit attempt not aided by health professionals, counselors, or group support	Eleven trials of self-help materials and advice to quit smoking by a health care worker vs. advice alone reported no significant increase in abstinence rates for people receiving self-help materials (OR 1.15, 95% CI 0.77–1.72) Three trials of self-help materials targeting specific populations found no significant reduction in quit rates (OR 1.13, 95% CI 0.85–1.50). Eight trials of self-help materials tailored to individuals' readiness to quit found significant increases in quit rates compared with standard materials (OR 1.41, 95% CI 1.14–1.75)
	Reported outcome as sustained abstinence	The addition of followup telephone calls from counselors increased quit rates (OR 1.62, 95% CI 1.33–1.97). The addition of a telephone hotline in one study also showed a benefit
	Dates of Articles Reviewed: To October 1999, updated to October 2000	There was no evidence of any difference among different types of self-help materials
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 45	Self-help materials, which provide a structured approach to smoking cessation beyond simple information, may provide a small increase in quitting compared with no intervention. There is no evidence they have an additional benefit over other minimal interventions, such as advice from a health care professional or NRT
	Total RCTs: 45	There is evidence that materials tailored to individual smokers is more effective
	Studies focusing on adult smoking cessation: 45
	RCTs focusing on adult smoking cessation: 45
	Meta-analysis Performed? Yes
Lichtenstein, E⁶⁹ (1996)	Inclusion Criteria:	Results:
Review Purpose: To review the effectiveness of telephone counseling as a smoking cessation intervention	Studies where telephone counseling is either a major part of the intervention, or is employed such that its specific effects on treatment outcome can be assessed	Reactive telephone counseling reaches only a small proportion of the smoking population. They may have a small benefit in terms of smoking cessation. There was no attempt to quantify this
Quality Assessment Rating: Moderate	Meta-analysis of studies of proactive telephone counseling	Analyses of the results of 13 RCTs of proactive counseling are reported for short-term (3 to 8 month) and long-term (12 to 18 months) cessation rates. Two studies were eliminated from the analysis of short-term effects because of heterogeneity. A further 2 studies were eliminated from the analysis of long-term cessation rates. No sensitivity analysis was performed
	Dates of Articles Reviewed: Not clearly stated; around 1996	There was evidence of increased cessation at short-term followup among individuals receiving proactive telephone counseling (OR 1.34, 95% CI 1.19–1.51). The size of the benefit was less at long-term followup (OR 1.2, 95% CI 1.06–1.37)
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 13	Reactive smoking cessation hotlines appear to be a useful public health smoking control strategy for large populations. Proactive telephone counseling produces modest improvements in smoking cessation rates
	Total RCTs: 13
	Studies focusing on adult smoking cessation: 13
	RCTs focusing on adult smoking cessation: 13
	Meta-analysis performed? Yes
Matson, DM⁷⁰ (1993)	Inclusion Criteria:	Results:
Review Purpose: To determine whether incentive-based programs conducted at worksites increase participation and long-term smoking cessation rates	Published studies of workplace smoking cessation programs involving financial incentives and compensation to promote participation and quit rates	Only 8 studies had a comparison group in which the effects of incentives and competition were separated from the effects of other interventions
Quality Assessment Rating: Moderate	Dates of Articles Reviewed: Inception of database - winter 1992	Three of these studies showed that incentives increased participation rates, and 5 enhanced smoking reduction
	Number of Studies Included:	One study found that competitions may enhance quit rates over incentives by themselves
	Total Number of Studies: 15	No study showed that incentives and/or competition enhanced smoking cessation past 6 months
	Total RCTs: 10	Conclusions:
	Studies focusing on adult smoking cessation: 15	Incentives/competition may be useful for increasing participation and smoking reduction
	RCTs focusing on adult smoking cessation: 10	Further research is needed to determine their effects on long-term quit rates, and what types of incentive procedures are most effective
	Meta-analysis Performed? No
Mullen, PD⁶⁵ (1997)	Inclusion Criteria:	Results:
Review Purpose: To examine the overall effectiveness of patient education and counseling on preventive health behaviors and to examine the effects of various approaches for modifying specific types of behavior	Published and unpublished studies that measured the effect of any education or counseling intervention on a preventive health behavior in apparently healthy individuals seen in a clinical setting in a developed country	The weighted average effect size from a random effects model for smoking/alcohol studies was 0.61 (95% CI 0.45–0.77), indicating that the behavioral outcomes produced a statistically significant difference
Quality Assessment Rating: Strong	RCTs and non-RCTs were included	Multiple regression shows that using behavioral techniques, particularly self-monitoring and several communication channels, produces larger effects for the smoking/alcohol group
Refer to Adult Healthy Diet Evidence Table 2 for additional information	Dates of Articles Reviewed: 1971 - 1994	Conclusions:
	Number of Studies Included:	Patient education and counseling contribute to behavior change for primary prevention of disease. Some techniques are more effective than others in changing specific behaviors
	Total Number of Studies: 74
	Total RCTs: 52
	Studies focusing on adult smoking cessation: 35
	RCTs focusing on adult smoking cessation: 26
	Meta-analysis Performed? Yes
Pederson, LL⁷² (2000)	Inclusion Criteria:	Results:
Review Purpose: To examine studies of smoking cessation interventions in African Americans and make comparisons with findings from the general population	Studies targeting African Americans specifically related to smoking cessation	Two studies of church-based smoking cessation interventions were reported. In both studies the intervention group was more likely to quit, although the difference was not statistically significant in either study
Quality Assessment Rating: Weak	Dates of Articles Reviewed: 1988-1998	Six studies of community-based smoking cessation programs were examined. Community-based interventions appear to be equally effective for African Americans and Caucasian Americans
	Number of Studies Included:	Four randomized trials evaluated smoking cessation interventions in a clinical setting (brief message to quit, health education program, NRT, office system prompt). No trial showed a significant increase in abstinence rates at 6 months. However, there is no good evidence that interventions in the clinical setting should be any less effective for African American populations
	Total Number of Studies: 12	Conclusions:
	Total RCTs: 9	More research is needed on the natural history of quitting, on the social norms for smoking among African American groups, and on the conceptual dimensions of race in the context of this research
	Studies focusing on adult smoking cessation: 12
	RCTs focusing on adult smoking cessation: 9
	Meta-analysis performed? No
Rice, VH⁶⁶ (2001)	Inclusion Criteria:	Results:
Review Purpose: To determine the effectiveness of nurse-delivered interventions on smoking behavior in adults	RCTs with minimum 6 month followup, of a nursing intervention (provision of advice, other content and strategies) to help patients quit smoking.	Sixteen studies comparing nursing intervention to a control or usual care group found nursing interventions significantly increased the odds of quitting (OR 1.50, 95% CI 1.29–1.73). This result did not change significantly if 2 trials accounting for significant heterogeneity were excluded from the analysis (OR 1.57, 95% CI 1.14–2.22)
Quality Assessment Rating: Strong	Required at least 2 treatment groups, and allocation must have been stated as “random”	The magnitude of benefit from low intensity interventions (1 counseling session OR 1.67, 95% CI 1.14–2.45), was not significantly different from that of high intensity interventions (counseling session, plus phone calls +/- additional face-to-face meetings OR 1.47, 95% CI 1.26–1.72)
Related Article:	Dates of Articles Reviewed: 1983 - June 2001	There is some evidence that interventions among hospitalized patients with cardiovascular disease (OR 2.14, 95% CI 1.39–3.31) may be more effective than non-cardiac hospitalized patients (OR1.2, 95% CI 0.92–1.56)
Rice, VH⁷⁴ (2000)	Number of Studies Included:	Conclusions:
	Total Number of Studies: 22	Indicates the potential benefits of smoking cessation advice and counseling given by nurses to their patients, with reasonable evidence that interventions can be effective
	Total RCTs: 22
	Studies focusing on adult smoking cessation: 22
	RCTs focusing on adult smoking cessation: 22
	Meta-analysis Performed? Yes
Ritvo, P⁷¹ (1997)	Inclusion Criteria:	Results:
Review Purpose: To determine what smoking cessation interventions are effective in family practice and what are the barriers to smoking cessation programs in family practice.	Not Reported	The data were combined according to the length of followup, the use of biochemical validation of cessation and the use of NRT. It is therefore difficult to report on the effectiveness of different smoking cessation interventions
Quality Assessment Rating: Moderate	Dates of Articles Reviewed: 1990 - 1996	The authors conclude that there is value in combining 3 key strategies: physician advice and support, NRT and cognitive-behavioral counseling
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 20	Relative efficacy of adjunctive counseling and physician training require further evaluation
	Total RCTs: unclear	One may be better than the other for achieving intervention intensity, or, alternatively, their combination may produce a synergism that proves to be most effective
	Studies focusing on adult smoking cessation: 20
	RCTs focusing on adult smoking cessation: unclear
	Meta-analysis Performed? No
Silagy, C⁶⁷ (2001)	Inclusion Criteria:	Results:
Review Purpose: To determine the effectiveness of advice from medical practitioner in promoting smoking cessation. Additional goals were to compare minimal interventions (single consultation < 20 minutes plus 1 telephone call) with more intensive interventions (any intervention involving greater time commitment) and the addition of supplemental aids such as self material to physician advice.	RCTs evaluating smoking cessation advice from a medical practitioner in which data on abstinence at 6 months were provided	Pooled data from 16 trials of brief advice vs. no advice (or usual care) demonstrated a significant increase in quit rates (OR 1.69, 95% CI 1.45–1.98). This equates to a difference in abstinence rates of approximately 2.5%
Quality Assessment Rating: Strong	Studies in pregnant women smokers were excluded	Trials comparing more intensive interventions found a small benefit from more intensive interventions (OR 1.23, 95% CI 1.02–1.49)
	Dates of Articles Reviewed: 1972 - October 2000	Indirect comparisons of studies in which followup visits were scheduled had a higher success rate (OR 2.66, 95%CI=2.06–3.45) than when followup was not scheduled (OR 1.59, 95%CI=1.33–1.90). A direct comparison found similar results (OR 1.60, 95%CI=1.1–2.33)
	Number of Studies Included:	One study reported 20 year followup data. Total mortality was 7% lower, fatal coronary disease was 13% lower and lung cancer was 11% lower in the intervention group, although these differences were not significant
	Total Number of Studies: 34	Conclusions:
	Total RCTs: 34	Simple advice has a small effect on cessation rates. Additional maneuvers appear to have only a small effect, though more intensive interventions are marginally more effective than minimal interventions
	Studies focusing on adult smoking cessation: 34-
	RCTs focusing on adult smoking cessation: 34
	Meta-analysis Performed? Yes
Stead, LF⁶⁸ (2001)	Inclusion Criteria:	Results:
Review Purpose: To evaluate the effect of proactive and reactive telephone counseling on abstinence rates	RCT or quasi-RCT, assessing proactive or reactive telephone counseling to assist smoking cessation to any population	Ten trials evaluated proactive counseling and minimal intervention with minimum intervention. Data were not combined because of significant heterogeneity. Two trials showed the addition of telephone counseling over self-help or individually tailored materials. Three trials showed a significant increased quit rate (range 2 to 8%). Three trials showed non-significant increased quit rates between 1–2% and a further two trials had an odds ratio close to 1
Quality Assessment Rating: Strong	Subjects were smokers or recent quitters as defined by trial recruitment protocols	Four studies compared telephone counseling plus a face-to-face intervention with a face-to-face intervention. There was no evidence of any benefit (OR 1.08, 95% CI 0.87–1.34)
	Minimum followup for cessation of 6 months	There was no evidence of any benefit of the addition of telephone counseling to NRT in 4 studies (OR 1.08, 95% CI 0.82–1.43)
	Dates of Articles Reviewed: Up to August 2000	One trial examined the provision of a help line to self-help materials. Quit rates were increased from 4.0 – 6.6% (p<0.05). Two additional trials showed no added benefit
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 23	Proactive telephone counseling appears to increase quit rates if it is used as the sole intervention
	Total RCTs: Unclear	Given the heterogeneity in these results, there is some uncertainty
	Studies focusing on adult smoking cessation: 23-	The addition of telephone counseling to other interventions does not appear to significantly increase abstinence rates
	RCTs focusing on adult smoking cessation: 23
	Meta-analysis performed? Yes

Evidence Table 2. Systematic reviews of the effectiveness of cancer control interventions in adult healthy diet (Key Question 2)

Lead Author (Year) Review Purpose Quality Assessment Rating	Inclusion Criteria Dates Articles Reviewed, Number of Studies Included, Meta-analysis Performed?	Results Conclusions
Ammerman, A⁸⁵ (2001)	Inclusion Criteria:	Results:
Review Purpose: To clarify what is known about the relative efficacy and effectiveness of behavioral interventions in promoting dietary change. Population subgroups are considered, particularly by ethnicity and gender, as well as cost-effectiveness of interventions.	Published in English	Few studies were appropriately designed or reported their findings to permit interpretation of the evidence for the efficacy of interventions by subgroup, particularly low-income or ethnic subgroups
Quality Assessment Rating: Strong	Research done in North America, Europe, or Australia, on non-institutionalized populations of adults, adolescents or children (excluding infants)	No studies that met the authors' review criteria provided data on the cost-effectiveness of dietary interventions. Interventions in high-risk groups had a greater impact than on the general public
	All intervention types considered, with at least 40 subjects per group at followup	Dietary interventions were positively associated with changes in fruit and vegetable intake (average increase of 0.6 servings/day) and a median reduction of 15.7% change in fat as a percentage of energy intake When fruit and vegetable intake were measured individually, changes in fruit intake were larger
	Dates of Articles Reviewed: 1975 - 1999	Conclusions:
	Number of Studies Included:	Dietary intervention components that included social support, small groups, goal setting, family activities, and “interactions with food,” such as cooking or taste testing, and were culturally or ethnically specific, seemed promising in increasing fruit and vegetable intake and reducing fat intake
	Total number of studies: 92	To gain the most from intervention research, the authors recommend that future studies assess dietary intake at the individual level and should collect detailed process and psychosocial data to help identify determinants of dietary change
	Total RCTs: Not clearly stated	Comparing the cost-effectiveness of these different intervention approaches will be critical to assessing their broader applicability
	Studies focusing on adult healthy diet: 71	More research is needed to determine the longer-term effectiveness of dietary interventions and to evaluate programs specifically designed to encourage the maintenance of change, preventing relapse over time
	RCTs focusing on adult healthy diet: Not clear
	Meta-analysis Performed? Yes
Ashenden, R⁵⁸ (1997)	Inclusion Criteria:	Results:
Purpose: To examine how effective lifestyle advice provided by GPs is in changing patient behavior, specifically smoking, alcohol consumption, diet, and exercise behaviors.	Published in English	Many general practice-based lifestyle interventions show promise in effecting small changes in behavior; none appears to produce substantial changes
Quality Assessment Rating: Strong	Investigated the effectiveness of lifestyle advice provided in the GP setting	Four trials were found related to GP advice regarding fat and fiber intake; 1 was positive, 1 negative, and 2 mixed
Refer to Adult Smoking Cessation	Comparison made between either no intervention or usual care, control group, or between advice of differing intensities	Advice was usually given in general lifestyle area, including diet, making interpretation difficult
Evidence Table 1 for additional information	Followup for smoking cessation trials must have been minimum 6 months after therapy	Conclusions:
	Dates of Articles Reviewed: Inception of database - 1995	GP-based health programs have a modest and variable effect on health outcomes, such as lifestyle change
	Number of Studies Included:	More extensive and rigorous research is required
	Total number of studies: 37	A greater number of GPs will need to become involved in promoting behavior change than the literature is currently suggesting if GP-based interventions are to be effective in public health
	Total RCTs: 37
	Studies focusing on adult healthy diet: 10
	RCTs focusing on adult healthy diet: 10
	Meta-analysis Performed? Only for smoking cessation data
Brunner, E⁸⁶ (1997)	Inclusion Criteria:	Results:
Purpose: To evaluate the effectiveness of dietary advice in primary prevention of chronic disease.	RCTs	After 3 to 6 months, mean net changes in each of the 5 outcomes favored intervention
Quality Assessment Rating: Moderate	Primary prevention for free-living adults, with intervention groups encouraged to consume a diet aimed at changing patterns of fat, sodium, or fiber consumption	Studies achieved a 2.5% reduction in percentage of calories from fat (95% CI -3.9,-1.1)
	Trial lasted minimum 3 months	Similar reductions in serum cholesterol, urinary sodium, and systolic and diastolic blood pressure. Relative effectiveness of different types of interventions was not assessed
	Dates of Articles Reviewed: Published materials up to July 1993	Conclusions:
	Number of Studies Included:	Individual dietary interventions in primary prevention achieved modest improvements in diet and cardivoascular disease risk status when maintained for 9–18 months
	Total number of studies: 17
	Total RCTs: 17
	Studies focusing on adult healthy diet: 17
	RCTs focusing on adult healthy diet: 17
	Meta-analysis Performed? Yes
Ciliska, D⁸⁷ (2000)	Inclusion Criteria:	Results:
Purpose: To determine the effectiveness of community-based interventions to increase fruit and vegetable consumption in people 4 years of age and older, specifically looking for outcome differences by socioeconomic status or age of the target group; location, intensity, and theoretical basis of the intervention; and level of training, education or professional status of the intervener.	Article involved an intervention intended to alter fruit and vegetable consumption	It appears easier to increase fruit intake than vegetables.
Quality Assessment Rating: Strong	Within the scope of public health practice in Ontario Participants were 4 years or older	Interventions by peer educators trained by nutritionists resulted in significant improvements in food intake.
	Study was prospective, had a comparison group, and provided information of process outcome or evaluation	Passive dissemination strategies of mailed messages and broad community interventions had no impact on intake.
	Dates of Articles Reviewed: First year of the respective database - August 1998; hand-searching went back to first issue of 1993; reference lists checked and articles back to 1988 were highlighted	Worksite multi-pronged interventions produced a statistically significant improvement in intake, but of questionable clinical significance (6.8 servings/month).
	Number of Studies Included:	No conclusions are possible regarding the relative impact of interventions in different target groups, site of the intervention, or level of preparation of the intervener.
	Total number of studies: 15	Conclusions:
	Total RCTs: 1	The most effective interventions gave clear messages about increasing fruit and vegetable consumption; incorporated multiple strategies that reinforced the messages; involved the family; were more intensive; were provided over a longer period of time, rather than 1 or 2 contacts; and were based on a theoretical framework.
	Studies focusing on adult healthy diet: 5
	RCTs focusing on adult healthy diet: 0
	Meta-analysis Performed? No
Contento, I⁸⁸ (1995)	Inclusion Criteria:	Results:
Purpose: To determine the effectiveness of nutrition education, and to identify any successful elements across interventions and the implications of the findings for program implementation, research, and policy in nutrition education.	For nutrition education for adult studies, studies examining direct nutrition education for free- living adults 18–65 years old were included	These strategies were identified as important in improving dietary behavior: attention to individual motivators and re-enforcers; personalized self-assessment; active participation, use of multiple channels; personalized education and counseling; family and peer involvement; use of behavioral techniques (in particular, goal setting monitoring, incentives and reinforcements); and environmental supports in food and point-of-purchase environments
Quality Assessment Rating: Moderate	For nutrition education for older adults, studies examining direct nutrition education for independently living adults over 65 years without a medical condition requiring ongoing care were included	Conclusions:
	For studies examining in-service preparation in nutrition education for professionals and paraprofessionals, studies that placed emphasis on continuing education or in-service training activities for elementary and high school teachers, school food service personnel, and nutrition and health professionals and paraprofessionals were included	Nutrition education is a significant factor in improving dietary practices when behavioral change is set as the goal and the educational strategies employed are designed with that as a purpose
	Dates of Articles Reviewed: 1980 - Not clearly stated	The studies having greatest impact on behavior consciously and systematically used motivational communication strategies in combination with behavioral change strategies and environmental interventions
	Number of Studies Included:
	Total number of studies: 217
	Total RCTs: Unclear
	Studies focusing on adult healthy diet: 117
	RCTs focusing on adult healthy diet: Unclear
	Meta-analysis Performed? No
Glanz, K⁸⁹ (1992)	Inclusion Criteria:	Results:
Purpose: Review a core of 25 articles (dating from 1982 through 1991) regarding medical school curricula and physician's knowledge, attitudes, and practices related to nutritional care, with a focus on prevention of coronary heart disease through cholesterol control.	Peer-reviewed journals or documents released by authoritative government and scientific organizations	While there appear to be modest increases in attention to nutrition at various levels of medical training and some improvement in physician attitudes about dietary intervention, the authors conclude that both educational opportunities and physicians' practices warrant increased and more effective attention to nutrition
Quality Assessment Rating: Moderate	Only those in which nutrition and/or cholesterol management was either the sole focus or a clearly identified focus	The review suggests multiple determinants of physicians' behaviours that presently are less than optimal for clinical care and public health. Thus, efforts to improve physicians' involvement in nutritional care should address the issue from both educational and health policy perspectives
	Control or minimal intervention group included	Studies on the determinants of physicians' nutritional care practices suggest 2 types of foci: (1) the knowledge, attitudes, and skills of physicians; and (2) the environment of health care delivery
	Dates of Articles Reviewed: 1982 - 1991	Because even the best training will fail to be implemented without reducing barriers such as lack of time, payment, and coordination of care, a two-pronged approach is essential
	Number of Studies Included:	Conclusions:
	Total number of studies: 25	Effective nutritional management requires both a sound knowledge base and the understanding, skills, and confidence to manage patients' education and behavioural change
	Total RCTs : Not clear	The current state of research and practice about physicians, preventive care, and nutrition should inform, direct, and help to improve both practice and research in the years ahead
	Studies focusing on adult healthy diet: 25
	RCTs focusing on adult healthy diet: Not clear
	Meta-analysis Performed? No
Glanz, K⁹⁰ (1995)	Inclusion Criteria:	Results:
Purpose: To summarize and provide a critical review of worksite health promotion program evaluations published between 1980 and 1995 that address nutrition and hypercholesterlomia.	Original data-based studies that report on measures of health status, behavior, attitudes, and knowledge as outcomes of worksite nutrition and cholesterol interventions.	No meta-analytic techniques were used to summarize data; data are presented narratively by study under each of the following categories:
The article discusses and critiques both intervention methods and research methodologies to identify the most effective strategies	Work published since 1980, to reflect programs conducted since the release of the first U.S. dietary guidelines Intervention strategies focusing on nutrition or cholesterol control were clearly described; an identifiable nutrition-related outcome measure was reported; and the study focused primarily on employees or a worksite population. Evaluations not available in peer-reviewed journals are not included.	Ten worksite nutrition programs were reviewed and categorized as group education, group education plus individual counselling/instruction, cafeteria-based programs, and group education plus cafeteria-based programs.
Related Papers:	Dates of Articles Reviewed:	Sixteen worksite cholesterol programs were reviewed, in five categories: monitoring, individual counselling, group sessions or classes, mediated methods using print, audiovisual, telephone, and self-help kits; and combination approaches.
Wilson, MG¹⁶⁷ (1996)	1968 to 1994*	Conclusions:
Quality Assessment Rating: Strong	Number of Studies Included:	Conclusions that can be drawn from this review are limited by the study designs used, which often lacked control groups, used nonrandomized designs, or relied on self-selected high-risk or volunteer participants.
	Total number of studies: 26	Rating of quality of the evidence in the literature as a whole lies between suggestive and indicative. It is clear that worksite nutrition and cholesterol programs are feasible and that participants will benefit in the short-term.
	Total RCTs: 12	Conclusive evidence about a causal relationship between worksite nutrition and cholesterol programs and improved behavior or health is not yet available, although studies currently underway hold promise for providing more solid evidence about the potential efficacy of these interventions.
	Studies focusing on adult healthy diet: 26
	RCTs focusing on adult healthy diet: 12
	Meta-analysis Performed? No
	*only work published since 1980 were included. The search was part of a larger review.
Mullen, PD⁶⁵ (1997)	Inclusion Criteria:	Results:
Purpose: To examine the overall effectiveness of patient education and counseling on preventive health behaviors and to examine the effects of various approaches for modifying specific types of behavior.	Published and unpublished studies measuring effects of any education or counseling intervention on a preventive health behavior in healthy individuals seen in clinical settings in a developed country	The weighted average effect size from a random effects model for nutrition/weight was 0.51 (95% CI 0.20–0.82) indicating that the behavioral outcomes produced a statistically significant difference
Quality Assessment Rating: Strong	RCTs and non-RCTs included	Multiple regression shows that using behavioral techniques, particularly self-monitoring and several communication channels, produces larger effects for the nutrition/weight group
Refer to Adult Smoking Cessation	Dates of Articles Reviewed: 1971 - 1994	Conclusions:
Evidence Tables for additional information.	Number of Studies Included:	The authors concluded that patient education and counseling contributed to behavior change for the primary prevention of disease
	Total number of studies: 74	Some techniques were found to be more effective than others in changing specific behaviors
	Total RCTs: 52
	Studies focusing on adult healthy diet: 12
	RCTs focusing on adult healthy diet: 10
	Meta-analysis Performed? Yes
Wilcox, S⁹¹ (2001)	Inclusion Criteria:	Results:
Purpose: To update previous narrative reviews systematically to examine the magnitude of nutrition and PA counseling effects highlighting gender and race/ethnicity issues.	Published in English	Intervention effects were modest, but significant for PA, BMI or weight, dietary fat, blood pressure, and total LDL serum cholesterol.
Quality Assessment Rating: Strong	Primary study conducted in health care settings investigating effects of dietary advice on CVD risk factors	Effects generally larger for samples with mean age >50 years and <6 months followup.
	CVD risk factor included as outcome variable Women >18 years	Type of comparison group, intervention, or use of behavior theory did not have a consistent impact on effects.
	Control or minimal intervention group included	Few studies focused on “persons of colour”.
	Dates of Articles Reviewed: 1980-2000	Conclusions:
	Number of Studies Included:	These findings are likely to be meaningful when considered by a public health perspective, even though they may be considered small by conventional statistical definitions.
	Total number of studies: 45	Clinical settings offer a way to reach a large proportion of the population, especially when low-income community health care clinics are included.
	Total RCTs: 34	Dietary assessments are subject to considerable measurement error, which attenuates true correlation coefficients.
	Studies focusing on adult healthy diet: 14	The authors felt that the true magnitude of effect was likely to be larger than reported in this study.
	RCTs focusing on adult healthy diet: 10
	Meta-analysis Performed? Yes

Evidence Table 3. Systematic reviews of the effectiveness of cancer control interventions in mammography (Key Question 3)

Lead Author (Year) Review Purpose Quality Assessment Rating	Inclusion Criteria Dates Articles Reviewed, Number of Studies Included, Meta-analysis Performed?	Results Conclusions
Balas, EA⁹⁴ (2000)	Inclusion Criteria:	Results:
Review Purpose: To assess the impact of clinician prompting on the provision of preventive care and to identify the effect of various covariates (reimbursement type, clinical characteristics, clinician specialty, and computerization).	RCT	The effect of prompting on mammography (n=14) 11.5% rate difference (95% CI 7.1 – 16.0)
Quality Assessment Rating: Strong	Physician prompt in the study group and no similar intervention in the control group	Generally, prompting can significantly increase preventive care performance by 13.1% (95% CI 10.5 – 15.6 - including cancer screening - fecal occult blood, Pap smear, and mammography)
Refer to Cervical Cancer Screening Evidence Tables for additional information.	Measurement of the effect on the number of preventive care activities	Conclusions:
	Dates of Articles Reviewed:	These conclusions are presented generally for preventive care performance:
	January 1, 1966 – December 31, 1996	Prompting physicians can lead to a significant improvement in health maintenance Observed increase in performance of preventive care efforts could reap substantial reductions in total mortality
	Number of Studies Included:	The many prompting tools offer a wide selection of options that are equally effective and easily applicable in most healthcare organizations (e.g., checklists attached to the patient chart, tagged notes, computer-generated encounter forms, prompting stickers, patient carried prompting cards)
	Total Number of Studies: 33
	Total RCTs: 33
	Studies focusing on screening mammography: 14
	RCTs focusing on screening mammography: 14
	Meta-analysis Performed? Yes
Bonfill, X⁹⁵ (2001)	Inclusion Criteria:	Results:
Review Purpose: To assess the effectiveness of different strategies for increasing the participation rate of women invited to community (population-based) breast cancer screening activities or mammography programs	RCT	The evidence favored five active strategies for inviting women into community breast cancer screening services: letters of invitation (OR 1.66, 95% CI=1.43–1.92), mailed education material (OR 2.81, 95% CI 1.96 – 4.02), letter of invitation plus phone call (OR 2.53, 95% CI 2.02 – 3.18), phone call (OR 1.94, 95% CI 1.70 – 2.23), and training activities plus direct reminders for the women (OR 2.46, 95% CI 1.72 – 3.50)
Quality Assessment Rating: Moderate	Published and unpublished trials in which women were invited to a community breast screening acitivity or program	Home visits did not prove to be effective (OR 1.06, 95% CI 0.80 – 1.40) and letters of invitation to multiple examinations plus educational material favored the control group (OR 0.62, 95% CI 0.32 – 1.20)
	Dates of Articles Reviewed: 1996 - 2000	Conclusions:
	Number of Studies Included:	Most active recruitment strategies for breast cancer screening programs examined in this review are more effective than no intervention
	Total Number of Studies: 14	Combinations of effective interventions can have an important effect
	Total RCTs: 14	Some costly strategies, such as a home visit and a letter of invitation to multiple screening examinations plus educational material, are not effective
	Studies focusing on screening mammography: 14	Further reviews comparing the effective interventions and studies that include cost-effectiveness, women's satisfaction, and equity issues are needed
	RCTs focusing on screening mammography: 14
	Meta-analysis Performed? No
Jepson, R⁹⁶ (2000)	Inclusion Criteria:	Results:
Review Purpose: To systematically review factors associated with the uptake of screening programmes and to assess the effectiveness of methods to increase uptake.	RCTs, quasi-RCTs, cohort and prospective case - control studies of any screening programmes, where the outcome was screening uptake	12 RCTs were identified which invited women by letter (vs. no letter) to attend mammograms. Three showed a significant effect of intervention and 5 showed no effect. Data could not be extracted from the remaining four studies
Quality Assessment Rating: Moderate	Must have used some form of multivariate analysis	5 studies were identified by the review comparing reminder letters versus control or another intervention and showed evidence of some effectiveness of reminders for mammograms
	Dates of Articles Reviewed: 1996 - 1998	Four studies evaluated the impact of educational sessions, printed materials or educational outreach visits targeted towards health care providers. The studies suggest a small increase in the uptake of screening tests in the intervention group when compared with the control group. RRs were not calculated due to lack of data One RCT evaluating a day long education session for 8 screening procedures (Pap, mammography, breast self exam, cholesterol screening, etc) reported that physician education intervention ultimately increased the proportion of women having a mammogram (p<0.01)
	Number of Studies Included:	For printed educational materials 2 RCTs were identified for which RRs could be calculated and found no effect of printed materials. However, 7 other studies in which RRs could not be calculated, 1 found that printed materials were more effective than control; the other 6 did not
	Total Number of Studies: 190	Four of 5 RCTs comparing educational telephone calls compared with control showed no effect of intervention in increasing mammography uptake
	Total RCTs: 130	Five studies evaluated the effectiveness of telephone counseling (by either a breast nurse or other specialist) to increase mammography uptake. Uptake increased significantly in three of the studies. The fourth study reported no difference in uptake (RR=1.07; 95% CI 1.00 – 1.15)
	Studies focusing on screening mammography: 34	Three studies (two RCTs, on quasi-RCT) evaluated reduced charges for free screening. One well-designed RCT found the provision of free screening (vouchers) was very effective in encouraging uptake among low-income women (RR = 4.28; 95%CI=1.91–9.60). The other two studies also reported a statistically significant effect of the intervention. Both studies targeted minority women (Hispanic and rural farming women)
	RCTs focusing on screening mammography: 16	Five RCTs evaluated the effectiveness of physician reminders in increasing uptake. RR's were calculated for three of the RCTs. All three reported an effect of the intervention, but one was only a small cluster RCT. One good quality RCT reported the mean mammography completion rate was 47.9 versus the control 34.6 which was statistically significant. (p value not reported)
	Meta-analysis Performed? No	Five RCTs were identified for the combination of physician reminders and individual letters or reminders to increase mammography uptake. RRs were calculated for four RCTs and all reported a statistically significant effect of the intervention. One did not present enough details of uptake
		Conclusions:
		Conclusions from the review are reported generally across all topics:
		Interventions for which there is evidence of effectiveness are invitation appointments, letters (less effective for mammography), telephone calls, telephone counseling, reduction of financial barriers and chart reminders for physicians
		Most educational materials have limited effectiveness, but educational home visits may increase uptake
		To increase informed uptake, future interventions should include information on the likely harms and risks, as well as the benefits of screening
		These studies should include a measure of the knowledge and whether this knowledge was used in the decision to undergo screening
		Furthermore, more studies are needed that target ethnic-minority groups and other groups where uptake is low
Kupets, R⁹⁷ (2001)	Inclusion Criteria:	Results:
Review Purpose: To determine the most effective strategies for the implementation of breast and cervical cancer screening delivered to women	Study conducted in North America	Physician-based strategies, especially manual and computer-generated reminders, appear to be the most effective approach in the implementation of breast cancer screening delivery to women. Computer-generated reminders improved the delivery of mammography to patients by an absolute rate of 6–30%
Quality Assessment Rating: Moderate	RCT	2 studies were identified and found that both studies indicate a significant improvement in the delivery of mammograms, with an absolute increase of 14–30% for performance rates and 15% for delivery rates. The NNI was 3–7 physicians There was a significant improvement of delivery of breast screening with an absolute increase of 35%, with an NNI of 2.5–3 physicians for the use of a manual reminder placed on the chart from studies identified (N=2)
Refer to Cervical Cancer Screening Evidence Tables for additional information.	Primary care physician (including family physician, GP, gynecologist and internist),	4 studies addressed patient reminders. The results were conflicting. 2 studies did not show a significant improvement, while 2 studies did show an improvement (10%, results not reported)
	Study included assessment of both breast and cervical screening.	Conclusions:
	Dates of Articles Reviewed: 1966-2000	Despite the availability of screening tests for the detection of breast cancer, the rates with which these are being offered are low
	Number of Studies Included:	Interventions targeting the patient alone showed an absolute increase in breast cancer screening of 10%, those targeting physician and patients for breast screening was 5–23% and those targeting the physician alone for breast screening ranged from 6–35%
	Total Number of Studies: 14	Generally, patient interventions with the highest level of accrual to screening were mailed letter invitations. The interventions with the highest success rates included physician reminder systems, both computerized and manual
	Total RCTs: 14
	Studies focusing on screening mammogrpahy: 14
	RCTs focusing on screening mammography: 14
	Meta-analysis Performed? No
Legler, J⁹⁸ (2002)	Inclusion Criteria:	Results:
Review Purpose: Examines mammography-enhancing intervention studies that focus no women in groups with historically lower rates of mammography use than the general population, and to determine which types of interventions are most effective for these diverse populations	English, US & international studies with concurrent controls	Combined intervention effects were estimated for different categories of intervention types using random effects models for sub-groups of studies
Quality Assessment Rating: Moderate	Reported actual receipt of mammogram (usually based on self-report) in groups of women with historically lower use of mammography	The most impressive effects were not for single category approaches but for combinations of interventions. The strongest combination of approaches used access-enhancing and individual-directed strategies and resulted in an estimated 27% increase in mammography use (95%CI=9.9–43.9, 9 studies)
	Dates of Articles Reviewed: 1984-1997	Additionally impressive was the access enhancing and system-directed combination (20% increase, 95%CI=8.2–30.6, 5 studies)
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 38	Access-enhancing strategies are an important complement to individual and system-directed interventions for women with historically lower rates of screening
	Total RCTs: 25
	Studies focusing on screening mammography: 38
	RCTs focusing on screening mammography: 25
	Meta-analysis Performed? Yes
Mandelblatt, J⁹⁹ (1995)	Inclusion Criteria:	Results:
Review Purpose: This study was designed to review research articles assessing the effectiveness of interventions to enhance physician breast cancer screening behaviour	Studies in the US only	The results were presented with clinical breast exam
Quality Assessment Rating: Moderate	Concurrent controls	In university settings, physician reminders and audit with feedback each significantly increased use of mammography by approximately 5% to 20% In community based settings the effects of physician education also had a positive impact on mammography rates, which ranged from 6% to 14%
	Start date chosen to coincide with publication of formal guidelines	Nine studies were identified which addressed the use of patient education and reminders on screening rates. Six studies reported increased mammography rates, although four were used in combination with other interventions
	Dates of articles reviewed:	Three studies were identified which addressed the use of audit and feedback interventions. All three reported a significant increase in mammography screening. The overall range of effect size was 15–24%
	January 1980–April 1993	Using patient education to influence physician behaviour in university settings had a modest impact in community trials
	Number of Studies Included:	Generally, reminders were most cost-efficient than audit with feedback
	Total Number of Studies: 20	Conclusions:
	Total RCTs: 11	Physician-based interventions can be effective in increasing screening use
	Studies focusing on screening mammography: 20	*Interventions should emphasize community practices and practices caring for underserved and older population.
	RCTs focusing on screening mammography: 11
	Meta-analysis Performed? No
Mandelblatt, J¹⁰⁰ (1999)	Inclusion Criteria:	Results:
Review Purpose: To determine the effectiveness of interventions targeted at providers to enhance the use of mammography	English language only	Behavioral interventions increased screening by 13.2% [95%CI=7.8–18.4] as compared with usual care and by 6.8% [95%C=4.8–8.7] as compared with active controls
Quality Assessment Rating: Moderate	Studies conducted in the US that used a RCT or concurrent non-RCT design, had defined outcomes, and presented data that could be abstracted for re-analysis Included studies that used either outcomes of ordering screening or completion rates of screening	Cognitive intervention strategies improved mammography rates by 18.6% [95%CI= 12.8–24.4]
	Exclusion criteria reported	Sociological interventions also had a similar magnitude of effect on screening rates [13.1% increase, 95%CI=6.8–19.3]
	Dates of Articles Reviewed: 1980-1998	Interventions targeting both patients and providers were not significantly better at increasing screening than those targeting providers alone, and multiple approaches (e.g., behavioral and cognitive) were generally not more effective than a single approach
	Number of Studies Included:	All interventions targeted at physicians were effective in increasing screening rates
	Total Number of Studies: 35	Conclusions:
	Total RCTs: 23	Interventions were more effective in increasing mammography use when compared with usual care than with active controls
	Studies focusing on screening mammography: 35	Strategies that targeted both patients and providers were not significantly more effective than those targeting providers alone
	RCTs focusing on screening mammography: 23	Decisions on the ultimate selection of an intervention to improve mammography receipt that targets providers should depend on feasibility, resources, expertise, and cost-effectiveness
	Meta-analysis Performed? Yes
Ratner, P¹⁰¹ (2001)	Inclusion Criteria:	Results:
Review Purpose: The purpose of this study was a meta-analysis to identify factors that influence the effectiveness of interventions in increasing women's use of mammography screening programs.	Papers in all languages published and indexed in Medline	More recent studies (1990-1996) were associated with higher mammography screening rates (OR 2.1, 95%CI=1.2–3.9)
Quality Assessment Rating: Moderate	Restricted attention to those that used experimental or quasi-experimental designs with mammography screening uptake as the dependent variable.	Studies designed to target older women (minimum age, 50–65 years) and those set in clinics exhibited smaller screening rates (OR 0.6, 95%CI=0.3–1.0, and OR 0.5, 95%CI=0.3–0.8, respectively)
	Exclusion criteria reported	Conclusions:
	Dates of Articles Reviewed: 1966-June 1997	Interventions primarily based in the community and that target relatively younger women (>35 years) are most effective
	Number of Studies Included:	The results of this meta-analysis do not provide explicit guidance for the implementation of specific interventions
	Total Number of Studies: 69
	Total RCTs: 34
	Studies focusing on screening mammography: 69
	RCTs focusing on screening mammography: 34
	Meta-analysis Performed? Yes
Shea, S¹⁰² (1996)	Inclusion Criteria:	Results:
Review Purpose: To conduct a meta-analysis of computer-based and manual reminder systems and to assess the overall effectiveness in ambulatory settings directed at preventive care.	Randomized, concurrent controls trials where the control group received no intervention	The following interventions improved preventive practices compared with the control condition for breast cancer screening - computer generated reminder (OR 1.88, 95% CI 1.44 – 2.45), manual reminder (OR 1.63, 95% CI 1.21 – 2.18) and computer plus manual reminder (OR 1.88, 95% CI 1.44 – 2.45) - unadjusted
Quality Assessment Rating: Strong	Ambulatory settings	Conclusions:
Refer to Cervical Cancer Screening Evidence Tables for additional information.	Dates of Articles Reviewed: 1966-December 1995	Computer generated reminders were effective for increasing, breast cancer screening
	Number of Studies Included:	Evidence from randomized controlled studies supports the effectiveness of data-driven computer-based reminder systems to improve prevention services in the ambulatory care setting
	Total Number of Studies: 16
	Total RCTs: 16
	Studies focusing on screening mammography: 11
	RCTs focusing on screening mammography: 11
	Meta-analysis performed? Yes
Shekelle, P¹⁰³ (1999)	Inclusion Criteria:	Results:
Review Purpose: To determine the best strategies for early detection and prevention currently covered by Medicare and to assess interventions designed to improve influenza and pneumococcal immunization rates, mammography rates, cervical smear cytology (pap test) and colon cancer screening	Had to address one or more of the five services of interest and employ one of the following study designs: RCT, controlled clinical trial, controlled before and after study, or interrupted time series.	The intervention with the greatest number of studies was patient reminders (131) followed by patient education (122 interventions). Patients were most often the target of the interventions (179) versus providers (113)
Quality Assessment Rating: Strong	Primarily searched for data relevant to the Medicare population.	The effectiveness of interventions to improve the use of clinical preventive and screening services for mammography were: patient financial incentives OR 3.57 (95% CI 2.36 – 5.40); patient reminder OR 2.57 (95% CI 2.22 – 2.98); organizational change OR 2.26 (95% CI 1.81 – 2.82); provider education OR 2.26 (95% CI 1.81 – 2.82); provider reminder OR1.59 (95% CI 1.36 – 1.86); feedback OR 1.49 (95% CI 1.24 – 1.80) and patient education OR 1.31 (95% CI 1.12 – 1.52)
	Exclusion criteria included	Results were also presented across all four regressions and results indicate that organizational change was consistently one of the most or the most effective interventions at increasing use of the clinical and preventive services. Patient financial incentives were also highly effective as was patient reminders which demonstrated relatively consistent effective results across all services
	Dates of Articles Reviewed: 1980 - 1995	Personalized reminders (one's signed by the patient's physician) are more effective than generic ones. And finally, feedback appeared to be a relatively ineffective intervention, as it was statistically beneficial only for increasing screening mammography
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 187	Conclusions are presented across all screening topics
	Total RCTs: 136	Organizational change and financial incentives were most consistent at producing the largest improvements in use of all preventive and screening services Patient reminders are also consistently effective across all topics. Patient reminders that are personalized or signed by the patient's physician are more effective than reminders that are generic
	Studies focusing on screening mammography: 76	Feedback is of limited, if any, effectiveness
	RCTs focusing on screening mammography: Not Clear	Multiple interventions are more effective than single interventions, although highly successful single interventions exist
	Meta-analysis Performed? Yes	Computer assisted provider reminders are more cost effective than patient reminders in the few studies that have addressed this issue
		There are insufficient data to draw conclusions about the effect of pre-intervention rates, intensity of interventions, or other factors in determining the success of interventions
Sin, J¹⁰⁴ (1999)	Inclusion Criteria:	Results:
Review Purpose: To evaluate the effectiveness of different interventions to increase breast screening uptake	English-language	Results are presented narratively by study
Quality Assessment Rating: Strong	Studies were included if uptake was used as a measure of success of the intervention and if relevant to the UK screening programme	Interventions were categorized as person-directed, system-directed, social network or multi-strategy
	Studies that promote breast screening uptake and acceptability	There is limited evidence available in the system-directed, multi-strategy and cost effectiveness categories
	Dates of Articles Reviewed: 1980-July 1998	Conclusions:
	Number of Studies Included:	Effectiveness in boosting uptake is greatest for the simple to administer interventions (example: invitations), rather than in depth ones, and these interventions tended to be person-directed
	Total Number of Studies: 28	Evidence of effectiveness of interventions in the other categories was limited by the number of studies and study design
	Total RCTs: 15
	Studies focusing on screening mammography: 28
	RCTs focusing on screening mammography: 15
	Meta-analysis Performed? No
Snell, J¹⁰⁵ (1996)	Inclusion Criteria:	Results:
Review Purpose: The focus of this meta-analysis was on studies that employed an intervention directed at either patients or physicians, or both, and measured its effects on screening rates for breast, cervical, and colorectal cancers. The objectives were to discern which intervention or combination of interventions was most successful and whether screening rates were enhanced Moderate more by targeting the patients, the physicians, or both patients and physicians.	Primary care setting directed at a patient, physician or both.	Results are presented as a combination of breast, cervical and colorectal cancer and could not be separated.
Quality Assessment Rating: Moderate	Addressed screening for breast, cervical or colorectal cancer.	Interventions targeting either physician or patient were equally successful (d = + 0.1894 and d = + 0.1756, respectively)
Refer to Cervical Cancer Screening Evidence Tables for additional information.	Reported results allowing a calculation of effect size.	Studies targeting both physician and patient demonstrated a smaller effect size (d = + 0.0514)
	Dates of Articles Reviewed: 1989-1994	Greater success was found for interventions targeting the physician both during and outside the patient visit (d =+ 0.1222 during visit, d = + 0.1849 outside visit, d = + 0.3375 both)
	Number of Studies Included:	Screening behaviour improved when the physicians were the targets of more than one, but not more than three, interventions (d = + 0.1360, d = + 0.2495, d = + 0.6829, d = - 0.0058)
	Total Number of Studies: 38	Since a combination of during- and outside-visit interventions showed a larger effect size than either alone, a multifaceted approach to changing physician behaviour seems to be the best
	Total RCTs: Unclear	Effect size by screening activity for breast screening (n = 41 cases) was d = + 0.2236 (95% CI 0.1960 – 0.2512)
	Studies focusing on screening mammography: Unclear	Conclusions:
	RCTs focusing on screening mammography: Unclear	Cancer screening activities increase with interventions that target either the physician or the patient and, when physicians are targeted, multiple interventions to serve as behaviour cues and increase awareness appear optimal
	Meta-analysis Performed? Yes
Wagner, T¹⁰⁶ (1998)	Inclusion Criteria:	Results:
Review Purpose: Compares the effectiveness of mailed patient reminders at increasing mammography screening.	Published RCTs that used a mailed reminder (to patients).	Among US studies in which controls did not receive any type of reminder, women who received reminders were approximately 50% more likely to get a mammogram (OR 1.48, x² _MH(1) = 38.27, p <.001)
Quality Assessment Rating: Moderate	Exclusion criteria reported	Tailored letters were found to be more effective than generic reminders (OR 1.87, x² _MH(1) = 4.70, p <.05)
	Dates of Articles Reviewed: 1985 - September 1996	Combining costs and effectiveness data allowed for effectiveness data of cost per woman screened, which ranged from $0.96 to $5.88
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 19	Patient reminders are effective at increasing mammography screening
	Total RCTs: 19	Mailed reminders are also relatively inexpensive given that most of the work can be done by personal computer
	Studies focusing on screening mammography: 19	More research is needed to assess the cost-effectiveness of patient reminders and their effectiveness across race, education, income and type of insurance
	RCTs focusing on screening mammography: 19
	Meta-analysis Performed? Yes
Yabroff, K¹⁰⁷ (1999)	Inclusion Criteria:	Results: Results presented by category (behavioural, cognitive & sociological)
Review Purpose: The objective of this study was to determine the effects of patient-based mammography screening strategies.	English language	Behavioural interventions increased screening by 13.2% (95% CI 4.7 – 21.2) compared with usual care, and by 13.0% (95% CI 8.7 – 17.4) when using multiple strategies and 5.6% (95% CI 0.6 – 10.6) when using a single intervention compared to active controls
Quality Assessment Rating: Moderate	US centred	Cognitive interventions using generic education strategies had little impact on screening, but those that used theory-based education (e.g., health belief model) increased rates by 23.6% (95% CI 16.4 – 30.1) compared with usual care
	Randomized or concurrent controlled design	Sociological interventions also increased screening rates and demonstrated improved mammography utilization by 12.6% (95% CI 7.4 – 17.9)
	Defined outcomes	The mode of intervention delivery appears to be an important component in the increasing rates of mammography utilization
	Data available for reanalysis	Interventions using a theoretical framework were the most effective in increasing screening rates
	Exclusion criteria provided.	The ability of these interventions to increase screening among subgroups and improve rates of ongoing screening, as well as the costs of these strategies, is unknown and is an important area for future research
	Dates of Articles Reviewed: 1980-1998	Conclusions:
	Number of Studies Included:	Overall, behavioural interventions, theory-based cognitive interventions, and sociological patient-targeted interventions appear to be effective in increasing mammography utilization, particularly when compared with usual care
	Total Number of Studies: 43	Multiple behavioural interventions and interactive theory-based cognitive interventions are effective when compared with active controls
	Total RCTs: 40	The long-term effectiveness of these interventions in increasing rates of regular mammography is only rarely reported and this will be an important area for further research
	Studies focusing on screening mammography: 43
	RCTs focusing on screening mammography: 40
	Meta-analysis Performed? Yes
Yabroff, K¹⁰⁸ (2001)	Inclusion Criteria:	Results:
Review Purpose: To assess the effectiveness of patient-targeted interventions in increasing mammography use when performed outside (out-reach) or inside the primary care medical setting (in-reach).	English language studies intended to increase mammography use published in the US	In-reach and out-reach interventions were equally effective in increasing mammography use
Quality Assessment Rating: Moderate	Exclusion criteria included	Compared to active controls, behavioural interventions with multiple strategies increased mammography use by 14% (95% CI 8.7 – 19.2) in in-reach and 18.7% (95% CI 4.9 – 32.4) in out-reach settings
	Dates of Articles Reviewed: 1980 - February 2001	Theory-based educational strategies delivered interactively increased mammography use by 10.7% (95%CI 6.8 – 14.7) and 19.9% (95% CI 10.6 – 29.1) in in-reach and out-reach settings, respectively
	Number of Studies Included:	Interventions that combined behavioural and theory-based educational strategies with usual care controls increased mammography use by 14% (95% CI 7.9 – 20.2) in in-reach and 27.3% (95% CI 14.7 – 40.0) in out-reach settings
	Total Number of Studies: 66	Sociological interventions increased mammography use by 10.7% (95% CI 3.4 –18.0) and 9.1% (95% CI 1.7 – 13.3) in in-reach and out-reach settings respectively
	Total RCTs: 54	Conclusions:
	Studies focusing on screening mammography: 66	In-reach and out-reach interventions to increase mammography use were similarly effective within intervention categories based on mechanism of action, mode of delivery, and type of control group
	RCTs focusing on screening mammography: 54	Ultimate decisions about intervention strategies will depend on the characteristics of the target population, practical considerations and relative cost-effectiveness
	Meta-analysis Performed? Yes

Evidence Table 4. Systematic reviews of the effectiveness of cancer control interventions in cervical cancer screening (Key Question 4)

Lead Author (Year) Review Purpose Quality Assessment Rating	Inclusion Criteria Dates Articles Reviewed, Number of Studies Included, Meta-analysis Performed?	Results Conclusions
Austin, SM¹¹³ (1994)	Inclusion Criteria:	Results:
Review Purpose: The objective of this study was to assess the clinical value of the physician reminder, an information intervention, in increasing compliance for selected preventive health care measures	Published study	Three RCTs were identified which addressed the effect of physician reminders on preventive care; more specifically cervical cancer screening
Quality Assessment Rating: Moderate	RTC	The OR from the combination of evidence from the 3 cervical cancer screening trials was significant (OR 1.180, 95% CI 1.020 – 1.339)
	A comparison of information or utilization management intervention in the study group with no similar assistance in the control group	Conclusions:
	An evaluation of the change in process and/or outcome of patient care	Based on results of this meta-analysis, further trials testing the effect of physician reminders on tetanus immunization would be unnecessary and probably unethical
	Dates of Articles Reviewed: Not clearly stated (<1994 assumed)
	Number of Studies Included:
	Total Number of Studies: 6
	Total RCTs: 6
	Studies focusing on cervical cancer screening: 3
	RCTs focusing on cervical cancer screening: 3
	Meta-analysis Performed? Yes
Balas, EA⁹⁴ (2000)	Inclusion Criteria:	Results:
Review Purpose: To quantify the impact of clinician prompting on the provision of preventive care and to identify the effect of various covariates (reimbursement type, clinical characteristics, clinician specialty, and computerization).	RTC	Results were presented by category (i.e., prevention) for most of the report.
Quality Assessment Rating: Strong	Physician prompt in the study group and no similar intervention in the control group	Very little data pertaining specifically to cervical cancer screening alone was reported
Refer to Mammography Screening Evidence Tables for additional information.	Measurement of the effect on the number of preventive care activities	Of the studies included, most addressed the clinical areas of cancer screening and prevention (20), immunization (14), and diabetes management (4)
	Dates of Articles Reviewed: January 1, 1966 – December 31, 1996	The effect of prompting on Pap smear (n=15) showed a rate difference of 5.8% (95% CI 1.5 – 10.1)
	Number of Studies Included:	Overall, prompting can significantly increase preventive care performance by 13.1% (95% CI 10.5 – 15.6). The effect of prompting for Pap smear specifically was 5.8% (95% CI 1.5 – 10.1)
	Total Number of Studies: 33	Academic affiliation, ratio of residents, and technique of delivery did not have a significant impact on the clinical effect of prompting
	Total RCTs: 33	Conclusions:
	Studies focusing on cervical cancer screening: 15	Prompting physicians can lead to a significant improvement in health maintenance
	RCTs focusing on cervical cancer screening: 15	The many prompting tools offer a wide selection of options that are equally effective and easily applicable in most health care organizations (e.g., checklists attached to the patient chart, tagged notes, computer-generated encounter forms, prompting stickers, patient-carried prompting cards)
	Meta-analysis Performed? Yes
Jepson, R⁹⁶ (2000)	Inclusion Criteria:	Results:
Review Purpose: To systematically review factors associated with the uptake of screening programs and to assess the effectiveness of methods to increase uptake	RCTs, quasi-RCTs, cohort and prospective case control studies of any screening programs where the outcome was screening uptake	Effective interventions aimed at individuals included invitation appointments, letters, and telephone calls; telephone counseling; and removal of financial barriers
Quality Assessment Rating: Strong	Must have used some form of multivariate analysis	Interventions that may be effective included educational home visits; opportunistic screening; multicomponent community interventions; and invitation followup prompts
Refer to Mammography Screening Evidence Tables for additional information	Dates of Articles Reviewed: 1996 - 1998	Limited effectiveness interventions were printed and audio-visual educational materials; educational sessions; risk-factor questionnaires; and face-to-face counseling
	Number of studies included:	Ineffective interventions were use of rewards and incentives.
	Total Number of Studies: 190	Other interventions either had no good-quality evidence or insufficient evidence for evaluation
	Total RCTs: 130	Conclusions:
	Studies focusing on cervical cancer screening: 12	Interventions for which there is evidence of effectiveness are invitation appointments, letters, telephone calls, telephone counseling, reduction of financial barriers, and chart reminders for physicians
	RCTs focusing on cervical cancer screening: 8	Most educational materials have limited effectiveness, but educational home visits may increase uptake
	Meta-analysis Performed? No	To increase informed uptake, future interventions should include information on the likely harms and risks, as well as the benefits, of screening. These studies should include a measure of the knowledge and whether this knowledge was used in the decision to undergo screening
		More studies are needed that target ethnic-minority groups and other groups where uptake is low
Kupets, R⁹⁷ (2001)	Inclusion Criteria:	Results:
Review Purpose: To determine the most effective strategies for the implementation of breast and cervical cancer screening delivered to women	Study conducted in North America	Of the 6 studies reviewed for computer-generated reminders, 3 showed significant improvements in cervical cancer screening
Quality Assessment Rating: Moderate	RTC	The delivery of cervical cancer screening improved by 9–30% with an NNI of 3–10 physicians
Refer to Mammography Screening Evidence Tables for additional information.	Primary care physician (including family physician, GP, gynecologist, and internist)	For the 2 studies identified for audit and feedback, neither study showed improvement for cervical cancer screening when comparing intervention versus control arm
	Study included assessment of both breast and cervical screening	The results for mailed patient reminders are mixed. Of the 4 studies assessed for the review, 2 studies showed improvement in cervical cancer screening (10%, the other study does not report numbers), and 2 show no significant improvement; in fact, there is a negative effect in the study arm with a decrease in screening of 10%
	Dates of Articles Reviewed: 1966 - 2000	Interventions targeting patients alone showed an absolute increase in cervical cancer screening of 10%. Interventions involving both the patient and physician resulted in an absolute increase of 10–30%, and those targeting physicians alone resulted in an increase of 9–40%
	Number of Studies Included:	Conclusions:
	Total Number of Studies: 14	Despite the availability of screening tests for the detection of breast and cervical cancer, the rates with which these are being offered are low
	Total RCTs: 14
	Studies focusing on cervical cancer screening: 14
	RCTs focusing on cervical cancer screening: 14
	Meta-analysis Performed? No
Pirkis, JE¹¹⁴ (1998)	Inclusion Criteria:	Results:
Review Purpose: To determine the effectiveness of patient-reminder systems and GP-reminder systems in promoting uptake of Pap tests. The a priori hypothesis was that both would be more effective than “normal care” in doing so	English	The women whose GPs had been prompted to remind them to have a Pap test were significantly more likely to do so than were control women (TRD -6.6%, 95% CI = 5.2 – 8.0)
Quality Assessment Rating: Strong	RTC	The corresponding estimate of the number of women needed to be involved in a GP reminder scheme in order to generate one additional screen is 15.2(95% CI = 12.6 – 19.3)
	GP or family medicine setting, examining the effectiveness of GP and patient reminder in increasing the proportion of women screened for cervical cancer	The TRD for the patient reminder studies was 4.9% (95% CI = - 2.6 – 7.2).
	Dates of Articles Reviewed: 1966 - December 1996	In both cases, sensitivity analysis revealed that one study stood out as an exceptional result. The omission of this study induced homogeneity among remaining studies
	Number of studies included:	Once this study was removed, the TRD's for the GP reminder and patient reminder studies were 7.9% (95% CI = 6.5 – 9.4) and 10.8% (95% CI = 8.1 – 13.6), respectively
	Total Number of Studies: 10	Conculsions:
	Total RCTs: 10	The results strongly suggest that GPs should make use of GP and patient reminder systems
	Studies focusing on cervical cancer screening: 10
	RCTs focusing on cervical cancer screening: 10
	Meta-analysis Performed? Yes
Shea, S¹⁰² (1996)	Inclusion Criteria:	Results:
Review Purpose: To conduct a meta-analysis of computer-based and manual reminder systems and to assess the overall effectiveness in ambulatory settings directed at preventive care.	RCT or concurrent controlled trials where the control group received no intervention	Results presented mainly by prevention category. Very limited information provided for cervical cancer screening specifically
Quality Assessment Rating: Strong	Ambulatory settings	Computer reminders improved preventive practices compared with the control condition for several other preventive care services, but not cervical cancer screening (OR 1.15, 95% CI 0.89 – 1.49)
Refer to Mammography Screening Evidence Tables for additional information.	Dates of Articles Reviewed: 1966 - December 1995	For all 6 classes of preventive practices combined, the adjusted OR was 1.77 (95% CI 1.38 – 2.27)
	Number of studies included:	Computer plus manual reminders vs. manual reminders: The adjusted OR for this comparison was 1.42 for all 6 preventive categories combined (95% CI 1.02 – 1.97; P=0.04), however, both methods had a lesser effect on cervical cancer screening than other preventive care practices (i.e., vaccinations, colorectal cancer screening)
	Total Number of Studies: 16	Conclusions:
	Total RCTs: 16	Conclusions presented generally for preventive services
	Studies focusing on cervical cancer screening: 9	Computer-generated reminders were effective for increasing vaccinations, breast cancer screening, colorectal cancer screening, and cardiovascular risk reduction, but they were not effective for increasing cervical cancer screening or the other 6 specific forms of preventive care examined
	RCTs focusing on cervical cancer screening: 9	Evidence from RCTs supports the effectiveness of data-driven computer-based reminder systems to improve prevention services in the ambulatory care setting
	Meta-analysis Performed? Yes
Shekelle, PG¹⁰³ (1999)	Inclusion Criteria:	Results:
Review Purpose: To determine the best strategies for early detection and prevention currently covered by Medicare and to assess interventions designed to improve influenza and pneumococcal immunization rates, mammography rates, cervical smear cytology (pap test) and colon cancer screening	Had to address one or more of the 5 services of interest and employ one of the following study designs: RCT, controlled clinical trial, controlled before-and-after study, or interrupted time series.	The intervention with the greatest number of studies was patient reminders (131) followed by patient education (122 interventions). Patients were most often the target of the interventions (179) versus providers (113)
Quality Assessment Rating: Strong	Primarily searched for data relevant to the Medicare population.	The effectiveness of interventions to improve the use of clinical preventive and cervical cancer screening were: patient financial incentives OR 3.12 (95% CI 2.62–3.72); patient reminder OR 1.84(95% CI 1.67–2.02); organizational change OR 2.65 (95% CI 2.26–3.12); provider education OR 1.59 (95%CI 1.29–1.97); provider reminder OR 1.40 (95%CI 1.27–1.54); feedback OR 1.12 (95% CI 0.97–1.30), and patient education OR 1.53 (95% CI 1.30–1.82)
	Exclusion criteria included	Results were also presented across all 4 regressions and results indicate that organizational change was consistently one of the most or the most effective interventions at increasing use of the clinical and preventive services. Patient financial incentives were also highly effective as were patient reminders which demonstrated relatively consistent effective results across all services
	Dates of Articles Reviewed: 1980 - 1995	Personalized reminders (signed by the patient's physician) were more effective than generic ones. And finally, feedback appeared to be a relatively ineffective intervention, as it was statistically beneficial only for increasing screening mammography
	Number of Studies Included:	Conclusions: Conclusions are presented across all screening topics Organizational change and financial incentives were most consistent at producing the largest improvements in use of all preventive and screening services
	Total Number of Studies: 187	Patient reminders are also consistently effective across all topics. Patient reminders that are personalized or signed by the patient's physician are more effective than reminders that are generic
	Total RCTs: 136	Feedback is of limited, if any, effectiveness
	Studies focusing on screening mammography: 65	Multiple interventions are more effective than single interventions, although highly successful single interventions exist
	RCTs focusing on screening mammography: Not Clear	There are insufficient data to draw conclusions about the effect of pre-intervention rates, intensity of interventions, or other factors in determining the success of interventions
	Meta-analysis Performed? Yes
Snell, JL¹⁰⁵ (1996)	Inclusion Criteria:	Results:
Review Purpose: To discern which intervention or combination of interventions was most successful for screening breast, cervical, and colorectal cancers, and whether screening rates were enhanced more by targeting the moderate patients, the physicians, or both patients and physicians.	Primary care setting directed at a patient, physician, or both	Results are presented as a combination of breast, cervical and colorectal cancer and could not be separated.
Quality Assessment Rating: Moderate	Addressed screening for breast, cervical, or colorectal cancer	Interventions targeting either physician or patient were equally successful (d +0.1894 and d +0.1756, respectively).
Refer to Mammography Screening Evidence Tables for additional information	Reported results allowing a calculation of effect size	Studies targeting both physician and patient demonstrated a smaller effect size (d +0.0514).
	Dates of Articles Reviewed: 1989 - 1994	Greater success was found for interventions targeting the physician both during and outside the patient visit (d +0.1222 during visit, d +0.1849 outside visit, d +0.3375 both).
	Number of Studies Included:	Screening behavior improved when the physicians were the targets of more than one, but not more than 3, interventions (d +0.1360, d +0.2495, d +0.6829, d -0.0058).
	Total Number of Studies: 38	Since a combination of during- and outside-visit interventions showed a larger effect size than either alone, a multi-faceted approach to changing physician behavior seems to be the best.
	Total RCTs: Unclear	Effect size by screening activity for cervical cancer screening (n = 35 cases) was d +0.0083 (95% CI = -0.0174 – 0.0340)
	Studies focusing on cervical cancer screening:	Conclusions:
	Unclear (45 cases included)	Cancer screening activities increase with interventions that target either the physician or the patient, and when physicians are targeted, multiple interventions to serve as behavior cues and increase awareness appear optimal
	Meta-analysis Performed? Yes
Tseng, DS¹¹⁵ (2001)	Inclusion Criteria:	Results:
Review Purpose: To perform a meta-analysis on existing RCTs to investigate the efficacy of patient letter reminders on increasing cervical cancer screening using Pap smears	RCT	The test for homogeneity showed evidence of heterogeneity (X² = 31, 9 df, p<.001). Division into sub-populations based on socio-economic status resolved the heterogeneity (X² = 5.2, 8 df, p = .75)
Quality Assessment Rating: Strong	Interventions studied in the form of a reminder letter	The studies evaluating those in lower socio-economic groups had a smaller response (OR 1.16, 95% CI 0.99 – 1.35) than those studies using mixed populations (OR 2.02, 95% CI 1.79 – 2.28)
	Published and unpublished studies that examine populations due for Pap smear screening	The pooled OR showed that patients who received the patient reminder letter were significantly more likely to return for screening than those who did not (OR 1.64, 95% CI 1.49 – 1.80)
	Dates of Articles Reviewed: 1966 - 2000	Conclusions:
	Number of Studies Included:	Patient reminders in the form of mailed letters increase the rate of cervical cancer screening
	Total Number of Studies: 10 – Total RCTs: 10	Patient letter reminders have less efficacy in lower socio-economic groups
	Studies focusing on cervical cancer screening: 10	Little information is available about the efficacy of reminders for Pap smear screening in Hispanics, the elderly, and the less educated
	RCTs focusing on cervical cancer screening: 10
	Meta-analysis Performed? Yes

Evidence Table 5. Systematic reviews of the effectiveness of cancer control interventions in the control of cancer pain (Key Question 5)

Lead Author (Year)Review Purpose	Inclusion Criteria Dates Articles Reviewed, Number of Studies Included, Meta-analysis Performed?	Results Conclusions
Allard, P¹²² (2001)	Inclusion Criteria:	Results:
Review Purpose: To review studies on cancer pain-control interventions and describe the findings with respect to participants' attitudes and knowledge, pain management, and pain levels.	English studies focusing on educational interventions designed to improve the control of cancer pain in adults	Seven studies found role-model training to be effective as health professionals became actively involved in implementing targeted cancer pain control activities in their own care setting
Quality Assessment Rating:	Search strategy not based exclusively on study design because RCTs are scarce on palliative care	Attitude and knowledge scores of physicians and nurses improved slightly in 3 intervention communities; there was no significant improvement in patients' pain as measured by the Pain Management Index
Moderate	Dates of Articles Reviewed: 1962 -1999	Five studies reported an improvement in knowledge and attitudes of nurses concerning cancer pain after the educational intervention
	Number of Studies Included in Review:	Seven of the 25 studies included an assessment of patient's pain relief as an outcome variable
	Total Number of Studies: 33	Eight studies (2 RCTs) directed interventions at patients or their family caregivers
	Total RCTs: 6	Seven of these studies targeted patients with advanced cancer and 1 study focused on family caregivers
	Studies focusing on control of cancer pain: 33	Improvement with respect to pain relief was reported in 5 studies where it was assessed
	RCTs focusing on control of cancer pain: 6	Conclusions:
	Meta-analysis performed? No	Educational interventions can successfully improve cancer pain knowledge and attitudes of health care professionals, but have little impact on patients' pain levels
		The most promising avenue for improving cancer pain control in ambulatory settings may be brief nursing interventions targeting patients in combination with a daily pain diary
		This review suggests that further progress may occur through incorporating a systematic and valid method of documenting daily fluctuation in pain levels, and insuring that documented uncontrolled pain is followed rapidly by clinical reassessment and dose adjustment

Evidence Table 6. Strategies for dissemination of cancer control interventions in adult smoking cessation (Key question 6)

Lead Author (Year) Country Study Purpose	Study Design Target Group Quality Assessment Rating	Dissemination Strategy Evaluated	Cancer Control Intervention(s) to be Disseminated	Findings
Albright, CL¹³⁴ (1992) United States	Study Design:	Train-the-Trainer (Month long training, off site for the trainers)	Clinical preventive medicine (CPM) curriculum that encompassed 6 areas: smoking cessation, clinical nutrition, weight control and exercise, hyperlipidemia and national screening guidelines	Training the Trainers:
Purpose: To test a dissemination model for providing clinical preventive medicine (CPM) training to internal medicine faculty across the US	• Cohort (one group pre and post)		The smoking cessation section included physician interventions for smoking cessation	• Fidelity ratings obtained from videotapes of the home-site seminars given by the trainers indicated that the trainers adhered closely to the CPM curriculum. (Over 90% of the slides provided were used in all 3 years of the study)
Refer to Adult Healthy Diet Evidence Table for additional information	• n = 10 internists were trained as “trainers”			Trainers teaching home-site internal medicine faculty:
	• n = 91 faculty members attended home-site seminars by the trainers			• Knowledge of smoking cessation, cholesterol and screening-specific information of faculty members significantly increased post-seminar in the last 2 years studied (p<0.02 and p<0.001, respectively)
	Target Group: General internal medicine faculty in the US			• Faculty's ratings of their self-efficacy to implement the CPM practices for smoking cessation demonstrated significant post-test increases for all 3 years (p<0.0001)
	Quality Assessment Rating: Weak			• Post-test faculty reported using significantly more specific behavioral change strategies (diaries, contracts and self-help materials) for smoking cessation in the 2 years reported (p<0.003 and p<0.0001, respectively)
				• 85% of the faculty in 1986, 96% in 1987, and 84% in 1988 reported teaching the CPM strategies to their house staff
				House staff trained by faculty that attended the CPM seminars:
				• House staff were surveyed before and after faculty participated in the CPM seminars. Significant increases were reported in the degree to which the faculty addressed CPM in their teaching interactions (1987: p<0.0001 and 1988: p<0.04, respectively)
				• House staff ratings of their self-efficacy to implement specific preventive medicine strategies increased in both years (1987: p<0.0001 and 1988: p<0.002, respectively)
Anderson, D¹⁴⁰ (1989) United States	Study Design:	Multiple media sources (television, radio, and newspapers)	CIS - a telephone-based information and education program of the NCI	Results pertaining to smoking are reported in this evidence table
Purpose: To examine inquiries received by the Cancer Information Service (CIS), a telephone hot-line, to determine:	• Descriptive study			• Television was by far the most frequently reported source of learning about the CIS by callers seeking smoking-related information (78.5%). The second most cited source was publications at 6.9% (included newspapers, magazine, pamphlets, and posters). Radio was cited by 4.6% of smoking-related callers. Health care providers were cited by 4.0% of smoking-related callers and 3.1% cited significant others; 3.0% cited telephone assistance
1) Effects of different media in stimulating calls to the CIS, and	• Retrospective analysis of 5 years of inquiries to one national and 26 local CIS offices in 4 subject areas. A standardized call record form was completed for each call.			The following findings were not reported by topic in the paper:
2) Demographic characteristics of callers in four cancer prevention and early detection subjects: smoking, nutrition, Pap smear screening and breast self-examination	• Demographic information was only collected during the last 2 years of the study for first-time, non-health professional callers, and was limited by federal stipulations to 20% of callers in 5 CIS offices			• Combined across topics, television was the most common information source reported by callers for both sexes: 72.2% male callers cited television compared with 60.7% of female callers. As the age of the callers increased, the frequency of television cited as the information source decreased. In the 19-year old or less age group, 81.7% of callers cited television compared with 39.6% of callers in the 60-year or older age group
Refer to Adult Healthy Diet and Cervical Cancer Screening Evidence Tables for additional information	• n = 57,374 nutrition related calls over the 5 years studied			• Television was the primary source reported by callers for all education levels. In general, the lower the caller's level of education, the more frequently television was cited as the information source
	Target Group: Smokers in the US, Mexico and other countries			• Television was the predominant source for 4 of the 5 ethnic groups identified across all topics - Caucasians, African-Americans, Hispanics, and Native Americans. For callers of Asian or Pacific Island heritage, the most frequently cited source was publications (46.7%) followed by television (32.1%). Further analysis suggested news publications, not health publications, were the greatest source of information for this ethnic group
	Quality Assessment Rating: Weak
Boyd, N¹²⁶ (1998) United States	Study Design:	Targeted media awareness campaign “Quit Today!”	CIS - a telephone-based information and education program of the National Cancer Institute	• The call volume from African-American smokers was significantly higher in the experimental communities compared to control communities (p<0.008)
Purpose: To evaluate whether a targeted communications campaign could lead more African- American smokers to call the Cancer Information Service (CIS) for smoking cessation information and materials	RCT, pre-post design	(Predominantly radio with some television) plus community outreach (community organizations distributed a videotape which encouraged smoking cessation and promoted use of the CIS as a free smoking cessation resource)		• African-American smoking callers in the experimental communities were more likely to be between the ages of 30-39 years than smoking callers in other ethnic groups, including Caucasians
	• N= 7 experimental communities (received the media campaign)			• Of the African-American smoking callers in the experimental communities: 45% were male (higher proportion than other ethnic group callers), and a higher proportion had completed high school or some college compared to callers from other ethnic groups
	• N=7 control communities (no media campaign)			• Of African-American smoking callers in the experimental communities: 51% cited radio as the way they had heard about the CIS, 42% cited television and less than 4% cited the videotape, poster, or family/friend/doctor
	Target Group: African-American adult smokers			• In the experimental communities, radio messages generated 8.89 calls per 10,000 African-American smokers, and television generated 6.89 calls per 10,000 African-American smokers. All other sources combined accounted for 1.38 calls per 10,000 African-American smokers
	Quality Assessment Rating: Moderate
Cockburn, J¹²⁴ (1992) Australia	Study Design: RCT	(1) Personal delivery and presentation by an educational facilitator with a follow-up visit 6 weeks later	A quit smoking kit. This kit contained multiple interventions from brief advice to intensive interventions. All of the interventions included had been proven to be efficacious in previous research studies	• General practitioners receiving the educational facilitator approach were significantly more likely than those receiving the other 2 approaches to have seen the kit (p = 0.003)
Purpose: To compare 3 approaches for marketing a quit smoking intervention kit to general practitioners.	n = 272 (total) general practitioners (GPs) selected by a random sample of GPs working in Melbourne (of these 6 GPs refused to receive the kit and 2 others refused facilitator visits)	(2) Delivery to the receptionist by a friendly volunteer courier with a follow-up phone call 6 weeks later		• The educational facilitator approach was rated as significantly more motivating to try the kit than the other 2 approaches (p =0.0005)
	Educational facilitator group:	(3) Postal delivery with a follow-up letter 6 weeks later		• There was no significant difference between the 3 approaches in use of the minimal intervention components of the kit (statistics not reported)
	n = 80 GPs			• General practitioners receiving the educational facilitator approach were significantly more likely than those receiving the other 2 approaches to have used one of the intensive intervention components (contract cards) from the kit (p = 0.02)
	Volunteer courier group:			• General practitioners in the educational facilitator approach were more likely to believe that the kit was less complicated (p = 0.01) and reported greater knowledge on how to use the kit (p = 0.02) than physicians receiving the other dissemination strategies
	n = 92 GPs			• There were no significant differences between the 3 approaches in overall acceptability of the kit or the delivery method (statistics not reported)
	Postal delivery group:			• There were no significant differences between the 3 approaches in perceptions of cultural or structural barriers to using the kit (statistics not reported)
	n = 92 GPs			• The cost of the educational facilitator approach ($Australian 142/ physician) was 24 times that of the mailed approach. The volunteer courier approach ($Australian 14/ physician) was twice the cost of the mailed approach (research costs were excluded from calculation)
	Target Group: General practitioners in Melbourne
	Quality Assessment Rating: Moderate
Cummings, K¹³¹ (1993) United States	Study Design: Post-test only, control group design	Targeted media awareness campaign Predominantly television spots. Additional supporting printed materials (posters and pamphlets) were distributed to healthcare providers and public service announcements (PSAs) were distributed to radio stations	CIS - a telephone-based information and education program of the National Cancer Institute	• The call rate per 10,000 smokers was significantly higher in experimental markets (10.74) compared with control markets (1.56), p<0.01
Purpose: To test the impact of a targeted mass-media campaign designed to motivate smokers to call the Cancer Information Service (CIS) for information on stopping smoking.	n = 7 experimental media markets (received the targeted media campaign)		Callers received verbal information about smoking cessation and were offered a free self-help smoking cessation booklet	• In experimental markets, 28.9% of calls received were from the target audience compared with 9.5% in control markets (p<0.01)
Related Papers:	n = 7 control media markets (no media campaign)			• Television was cited as the primary source of learning about the CIS by 71.5% of callers in the experimental markets and by 57.7% of callers in the control markets (no statistics reported)
Cummings, K¹³⁵ (1989)	Target Group: Women smokers with young children			• Callers from experimental markets were more likely to be female compared to control markets (p<0.01)
	Quality Assessment Rating: Weak			• A higher proportion of callers from experimental markets were between the ages of 20–29 years compared to the control markets (p<0.01)
				Six-month follow-up interviews were conducted with callers from the target audience:
				• In total, 770 callers were eligible, 719 from experimental markets and 48 from control markets. Follow-up interviews were completed with 559 women (73%)
				• 64% of women followed-up from the experimental markets attempted to quit smoking compared to 46% of women in the control markets (not significant, p=0.06)
				• 72 women reported that they had quit smoking and had not had a cigarette for at least a week at the time of the follow-up interview. There was no significant difference in quit rates between the experimental (13%) and control (15%) markets
Dietrich, A ¹⁴⁴ (1992) United States	Study Design: RCT	(1) Facilitator visited each practice 3–4 times over 3 mths; each visit lasted approximately 120 min.	Multiple office system interventions including preventive care flow sheets, chart stickers, health education posters and brochures, and patient health diaries	Note : Results pertaining to smoking cessation are reported in this evidence table.
Purpose: To test the impact of physician education and facilitator assisted office system interventions on cancer early detection and preventive services	In total, 98 of the 102 practices that agreed to participate completed the study. The unit of randomization was the medical practice as represented by one physician	Performed an initial audit of each practice to assess the status of preventive care and assisted practices in the design and implementation of office system interventions. Practices only implemented those interventions that meet their perceived needs	(None of the interventions were computer-based)	The response rate for the cross-sectional survey pre-experiment was 91% (N=2436 patients) and 93% (N=2595) at 12 months follow-up.
Refer to Adult healthy diet, Mammography, and Cervical cancer screening evidence tables for additional information.	Four groups:	(2)Facilitator+Workshop		• Significantly more eligible patients in the Facilitator Only group reported their physician had advised them to quit smoking compared to patients in the control group at 12 month follow-up (Proportion: 0.84 vs. 0.67; p<0.05, baseline results were used as covariates)
	Facilitator only: N=24 practices	Same as (1) plus physician from each practice attended a 1 day workshop led by an expert who reviewed NCI's prevention and screening recommendations and taught specific skills. Also provided a written syllabus		• There was no significant increase in the number of eligible patients in the Facilitator+Workshop group reporting their physician had advised them to quit smoking compared to patients in the control group at 12 month follow-up (Proportion: 0.80 vs. 0.67)
	Workshop + Facilitator: N=26 practices	Note: The workshop andcontrol groups did not receive information on the use of office systems interventions for cancer prevention or early detection.		The study's overall conclusion: Community practices assisted by a facilitator in the development and implementation of an office system can substantially improve provision of cancer early detection and preventive services
	Workshop only: N=24 practices
	Control: N=24 practices; no intervention; no further detail provided
	Target Group: Office based GPs and general internists in New Hampshire and Vermont
	Quality Assessment Rating: Weak
Elder, J¹²⁷ (1991) United States	Study Design: 1 group, post-test design	Mass media awareness campaign “Quit to Win”	Patient-directed incentives to quit smoking, self-help educational materials, and a listing of smoking cessation programs available in the county of San Diego	Only findings pertaining to the evaluation of media sources to attract program participants are reported
Purpose: (1) To determine whether a community-based organization with limited resources could effectively promote smoking cessation through a program previously conducted only under the auspices of large, federally-funded demonstration programs; (2) To evaluate which media sources were influential in attracting program participants; (3) To identify factors that differentiate joiners and non-joiners of this type of program; and (4) To provide an estimate of the cost-effectiveness of such a program	Target Group: Adult smokers in San Diego County	(Television, radio, newspaper and printed advertisements distributed through community organizations, healthcare facilities and worksites)		• Telephone interviews were conducted with 3 groups: (1) program joiners (N = 148, random sample), (2) non-joiners (N = 78, who had requested information, but did not join the program) and (3) comparison group of smokers who had not heard about the program and did not join (N = 127, random digit dialing)
	Quality Assessment Rating: Weak			• The most effective informing media source for joiners was television (53.9%) with its influence increasing as income level decreased (p < 0.05). Income level was not related to any media source for non-joiners
				•People with lower incomes relied more on the influence of friends and family in deciding to participate in the contest (p < 0.05)
Epps, R¹²⁹ (1998) United States	Study Design: 1 group post test	To recruit the medical organizations:	Multiple smoking cessation interventions	Recruitment of national professional medical organizations:
Purpose: To determine whether national organizations of medical professionals could be used to effectively disseminate a research-based smoking cessation program.	Target group: (1) National professional medical organizations; (2) Medical professionals	National Cancer Institute (NCI) representatives met with the organization leaders to discuss the train-the-trainer program	(In particular, NCI's manual “How to Help Your Patient Stop Smoking” that outlines office-based interventions)	• Of the 8 organizations initially approached, 6 agreed to co-sponsor Train-the-Trainer seminars and two declined. Reasons given for declining: one had its own smoking cessation program and the other was already collaborating with another government institute to develop a program tailored to its own specialty
Three objectives:	Quality Assessment Rating: Weak	To disseminate the smoking cessation interventions:		• The combined membership of 6 participating organizations exceeded 150,000
(1) To enlist national professional organizations;		Train-the-Trainer		• During the 4 years of the program, 5 smaller specialty organizations were also recruited (number approached not stated)
(2) To co-sponsor 50 Train-the-Trainer sessions; and				• In total, 11 national organizations and their affiliates co-sponsored 53 Train-the-Trainer seminars in 22 states and the District of Columbia
(3) To train 2,000 professionals throughout the US to instruct their colleagues in smoking cessation techniques				• The interval between NCI's initial contact and an organization co-sponsoring a Train-the-Trainer seminar ranged from 3 months to 1 year
				Train-the-Trainer seminars:
				• 53 Train-the-Trainer seminars were held during the 4 years of the program and 2,098 medical professionals were trained as trainers. There are “trainers” practicing in 50 states in America, the District of Columbia and Puerto Rico
				The paper did not report any results pertaining to the “trainers” disseminating the smoking cessation interventions to professionals in their local communities
Fowler, G¹³³ (1989) United Kingdom	Study Design: 1 group, post-test	Postal delivery (Booklet was sent with the BMA News Reviews to all BMA doctors listed as GPs)	Booklet “Help Your Patient Stop” which contained the World Health Organization (WHO) and the International Agency against Cancer (UICC) joint guidelines on smoking cessation	• The response rate to the mailed questionnaire was less than 50% after the second mailing. A shortened version of the questionnaire was then mailed out (2 mailings)
Purpose: To determine whether general practitioners (GPs) received the booklet “Help Your Patient Stop” which contained the World Health Organization (WHO) and the International Agency against Cancer (UICC) joint guidelines on smoking cessation and whether the physicians found the booklet useful.	• Cross-sectional survey			• Overall response rate to the 4 mailings: 3240 GPs (65%) returned a completed questionnaire and 388 (7.8%) doctors returned uncompleted questionnaires (of these, 341 had indicated that they were not GPs)
	• Random sample of intended recipients (approx. 1 in 6)			• 50.5% of the 3240 respondents reported that they had received the “Help Your Patient Stop” booklet and 27.7% said that they had read the booklet
	n = 5000 GPs			• There were no significant differences in gender, age or smoking status between GPs who read the booklet and those that did not
	Target Group: GPs who were members of the British Medical Association (BMA)			• Doctors responding to the shorter questionnaire were more likely to have read the booklet than those replying to the longer version (p<0.001)
	Quality Assessment Rating: Weak			• There was no significant difference between respondents to the two versions in the proportion reporting they received the booklet
				• Of the 448 respondents to the longer questionnaire who reported reading the booklet, only 86 (19.2%) correctly named one or 2 of the 3 steps identified in the booklet as essential elements in helping patients stop smoking; and only 56 (12.5%) correctly reported all 3 steps
				• Correct responses to the 3 steps was associated with increasing age of the GPs (p<0.05), but not with gender or type of practice
Lemelin, J¹⁴³ (2001) Canada	Study Design: RCT	Educational facilitators	Multiple interventions including reminder systems, flow charts and patient educational materials	Random chart audit of 100 records/practice was performed a baseline and again at follow-up:
Purpose: To evaluate a multifaceted outreach intervention, delivered by nurses trained in prevention facilitation, to improve prevention in primary care	Of the 95 practices contacted, 49 chose not to participate. In total, 46 practices were randomized. One practice in the facilitator group was lost to follow-up	(Over an 18 month period each practice was visited an average of 33 times; each visit lasted approximately 1 hour)		At baseline, the preventive performance index was not significantly different between the facilitator and control groups (31.9% and 32.1%, respectively). At follow-up, the corresponding values were 43.2% and 31.9%, the absolute increase in the facilitator group was of 11.5% was statistically significant (p<0.001)
Related Papers:	Facilitator group: N=22 practices (total of 54 physicians) completed the study; received visits from educational facilitators	The facilitators performed an initial audit and feedback of each practices baseline preventive performance rates; facilitated the development of practice goals and policy for preventive care; and assisted practices in selecting and implementing interventions to improve preventive care)		Smoking cessation counseling specific findings:
Baskerville, N¹⁶⁸ (2001)	Control group: N=23 practices (total of 55 physicians) completed the study; received no visits.			On chart audit, at baseline, smoking cessation counseling was performed with 37.6% of eligible patients in the facilitator group and with 40.5% in the control group. At follow-up, the corresponding values were 41.2% and 38.7%, there was no significant difference in change between the two groups)
(Process Evaluation)	Target Group: Primary care practices that have a payment system based primarily on capitation in Ontario, Canada			Telephone survey of random sample of 25 patients per 100 audited charts from facilitator group practices indicated a higher rate of smoking cessation counseling (baseline 64.0%; follow-up 73.0%)
Refer to Mammography and Cervical Cancer Screening Evidence Tables for additional information	Quality Assessment Rating: Weak			Overall findings from the process evaluation:
				All facilitator group practices received preventive performance audit and feedback, achieved consensus on a plan for improvement, and implemented a reminder system. 90% implemented a customized flow sheet, 10% used a computerized reminder system; 95% wanted critically appraised evidence for prevention; and 100% received patient educational materials
				• Audit and feedback, consensus building, and development of reminder systems were identified as the key components by content and bivariate analysis
				• 95% of physicians were satisfied or very satisfied with the educational facilitator approach
Marin, G¹³⁰ (1995) United States	Study Design: Interrupted time series	Media awareness and education campaign (television, radio and print) plus a community outreach strategy	Multiple interventions: self-help patient educational materials in Spanish and smoking cessation programs for Latinos	Only findings pertaining to the media awareness component of the study are reported in this evidence table
Purpose: To evaluate the effectiveness of the Programa Latino Para Dejar de Fumar in disseminating smoking cessation information in San Francisco's Latino community	(7 cross-sectional telephone surveys with Hispanics 18–65 years of age; independent random samples)			• Awareness of a group-specific smoking cessation program for Latinos at baseline was: 12.6% for current smokers and 17.7% for former smokers. In the last year surveyed, 43.3% of current smokers and 42.2% of former smokers reported awareness of a Latinos smoking cessation program
Related Papers:	Baseline:			• There was a significant increase in awareness of a smoking cessation program for Latinos among less acculturated (p<0.001), those with less than 12 years of formal education (p<0.001), and among women (p<0.01). Awareness was significantly influenced by the passing of time (p<0.001) during the 7 years of surveys
(1) Marin, G¹³² (1990)	n = 1660 (Year 1986)			• Awareness of the availability of smoking-cessation printed material targeting Latinos was 41.8% at baseline and 48.7% in the last year surveyed post-implementation of the intervention
Preliminary report	n = 2053 (Year 1987)			• The awareness of the availability of printed smoking cessation material for Hispanics increased among the more acculturated (p<0.001), among those with more than high school education (p<0.001), among women (p<0.001), and awareness increased significantly with the passage of time (p<0.001)
(2) Perez-Stable, E¹³⁹ (1993)	Post-implementation of the intervention:			• The proportion of respondents who reported having a copy of the self-help smoking cessation manual produced by the project was 7.5% in 1988 and 9.2% in 1993 (high of 19.7% was in 1989)
Preliminary report	n = 1965 (Year 1988)			• There were significant increases in the rate of reported ownership of the manual among the less acculturated respondents (p<0.001) and across the years of the project (p<0.001). There was no significant effect by level of education
	n = 1989 (Year 1989)
	n = 1959 (Year 1990)
	n = 2110 (Year 1991)
	n = 1501 (Year 1993)
	Target Group: (1) Less acculturated adult Hispanic smokers in San Francisco; and (2) Non-smokers who may know relatives or friends who smoke
	Quality Assessment Rating: Weak
Mullins, R¹²⁵ (1999) Australia	Study Design: Interrupted time series	Postal delivery of an introductory letter and the patient self-help booklet (also included a stand for the books to permit display or storage)	GPs to identify patients who smoked, advise them to quit and to give them a copy of the smoking cessation self-help booklet “The Can Quit Book”	• A survey of GPs post-dissemination of the intervention found that 5% of doctors could not recall receiving the booklets. Of those who did recall receiving the booklets, 97% reported using them, and the most frequent method of use was to personally give them to patients
Purpose: To determine whether there had been a change over time in whether general practitioners (GPs) advised their patients not smoke (pre- and post-dissemination of a smoking cessation intervention).	Survey of GPs:	The booklets continued to be mailed regularly to all of the GPs (interval not specified)		• The extent to which patients who smoke remembered GPs talking to them about smoking cessation was assessed pre-dissemination of the intervention and 3 times post-dissemination. Over time, there was a significant increase in the proportion of smokers who reported that their GP had provided them with help or information to stop smoking (p<0.001). There was no significant difference over this period in smokers who reported that their doctor asked or advised to stop smoking (without offering information or help)
	n = 947 (Year: 1993)			• One year before dissemination of the intervention, 10.7% of smokers reported that their doctor provided information or help to stop smoking. Five years after dissemination, 20.6% of smokers reported that their doctor provided information or help to stop smoking
	Survey of patients who smoke:
	n = 624 (Year: 1990)
	n = 596 (Year: 1992)
	n = 609 (Year: 1994)
	n = 563 (Year: 1996)
	Target Group: All GPs in Victoria
	Quality Assessment Rating: Weak
Muramoto, M¹³⁶ (2000) United States	Study Design: 1 group, pre-post test	Train-the-Trainer and a state-wide certification program	Multiple interventions (Differed by level of certification) emphasized the Agency of Health Care Policy and Research guidelines.	• During the first year of the project, a total of 1075 participants attended certification training. 947 of the participants received basic level certification, and 82 received specialist level certification. The majority of the participants in both the basic (75.8%) and specialist (86.6%) classes were female
Purpose: To describe the development and preliminary results from a community based certification model for training in tobacco cessation skills in Arizona	Target Group: Personnel working in Arizona's tobacco control program local community-based projects and their community partners providing tobacco treatment services		Basic level - brief intervention (4As: Ask, Advise, Assist, Arrange) delivered in the context of another service (e.g., medical or dental office, worksite or faith community)	• Knowledge was assessed post-certification training only. In both the basic and specialist programs, the average score on the knowledge tests were greater the 90%
	Quality Assessment Rating: Weak		Specialist level - intensive interventions and train individuals for basic level certification	• Self-efficacy of participants was measured pre- and post-certification training, and again at 3 months follow-up. There were significant and durable increases in both basic and specialist participant's self-efficacy (basic level p<0.001; specialist level p<0.05)
				• At 6-month follow-up post-training: (1) 80.9% of participants who received basic level certification reported that they had performed at least one brief intervention and 74.8% had made a referral to intensive services; and (2) 48.8% of participants who received specialist level certification reported delivering intensive interventions and 69.5% were teaching basic level certification classes
				• At the time this report was written, there were approximately 5 basic level certification classes per week through Arizona
Pierce, J ¹⁴⁶ (1986) Australia	Study Design: Two group pre + post design	Multi-media campaign	Multiple interventions	The number of calls to the Quit Line showed a direct relationship to the campaign's television commercials: over 50,000 calls were made in the first 3 months of the campaign compared with an expected 8,600 calls predicted using previous call rates
Purpose: To evaluate whether the media-based “Quit for Life” campaign achieved its intermediate goals of: (1) making television commercials that would be remembered by >50% of the population; (2) increase use of information an treatment services; (3) increase the intention of smokers to quit; and (4) increase the number of people trying to reduce their smoking during the campaign	The city of Sydney received the media campaign. The city of Melbourne was the control city and did not receive the campaign.	Primarily television commercials (N=389 prime-time advertising spots) supplemented with radio and newspaper advertisements	(The Quit Kits contained a self-help booklet for smoking cessation and an audiocassette tape of a structured relaxation program. The Quit Centre, located in Sydney Hospital, offered a choice of six standard anti-smoking interventions for a fee of $5)	There were 19, 196 Quit Kits were sold during the campaign and there were approximately 3,000 enrolments in Quit Centre smoking cessation classes (compared to about 500 in the previous year)
Related papers:	A cohort of 949 people in Sydney (n=576) and Melbourne (n=373) were interviewed pre-campaign and 6 months after the campaign ended.	The campaign lasted 5 months		The response rate to the follow-up cohort survey was 76% in Sidney and 73% in Melbourne: significantly more smokers from the Sydney reported quitting or cutting down on the number of cigarettes compared to smokers in Melbourne (35% vs. 18%, p<0.05). The number of smokers who reported that they had attempted to quit during the 12 month follow-up period was significantly different between the two groups (66% vs. 60%, p<0.05)
Dwyer, T¹⁴⁵ (1986)	Target group: Smokers in Sydney, Australia	All 3 of the television commercials contained a message encouraging smokers to call the “Quit Line” (a recorded telephone message service that provided callers information about the “Quit Centre” and “Quit Kits”)
(Evaluated whether the “Quit for Life” campaign was associated with a decline in smoking prevalence)	Quality assessment rating: Weak
Pierce, J¹⁴¹ (1992) United States	Study Design: Interrupted time series	Nationally televised public service announcements	CIS - a telephone-based information and education program of the NCI	In the 5-year period analyzed, the CIS received a disproportionate number of calls in 3 specific months - August 1983, January 1985, and January 1987. In each case, more than 20% of all calls in that particular year were received in that month (8% would have been expected if the calls had been distributed evenly)
Purpose: To determine whether nationally televised public service announcements were associated with increased use of the Cancer Information Service (CIS) and to assess the importance of specifically promoting the CIS telephone hot-line and to identify the characteristics of individuals who respond to such promotion.	Combined the frequency-of-call data from the CIS (5-year period) with data on the frequency and reach of television spots during the same time frame			During the same 5-year period, 12 anti-smoking television public service announcements were shown. The timing of these public service announcements did not follow the pattern of peaks and troughs seen in smoking-related calls to the CIS
	n = 279,681 smoking-related calls (from 1983 through 1987)			Three of the 12 announcements, however, did have television exposure patterns that fit with the 3 peak call months. The 3 public service announcements were: (1) “Surgeon General” in which Dr. Everett Kopp urged viewers not to smoke; (2) “Chained Smoker” which featured a man in a prison uniform chained to a giant cigarette; and (3) “Reaching Smokers” which featured situations in which a telephone gave smokers a “hand” when they felt the urge to smoke. All 3 of these public service announcements explicitly encouraged smokers to call the CIS and promoted the number to call
	Quality Assessment Rating: Weak			None of the other 9 public service announcements promoted the CIS telephone number
				Demographic data was collected from a random sub-sample of smoking-related callers. Women were 20% more likely than men to call the CIS during periods of no television promotion compared to 13% more likely in television promotion periods. Callers with more than high school education were 52% more likely to call than people with less education during periods of no television promotion. During periods of television promotion, this decreased to 35% more likely. Television promotion also increased the percentage of callers under age 40 to 36%–40% from 20%–28% in non-promotion periods
Sorensen, G¹³⁷ (1998) United States	Study Design: RCT	Given to control sites (method not specified) at the conclusion of the Working Well Trial	Tobacco cessation interventions from the Working Well Trial	Only the findings pertaining to the dissemination of the interventions to the control sites at the conclusion of the Working Well Trial are reported
Purpose: To examine durability, dissemination and institutionalization of worksite tobacco control programs.	n = 42 control worksites			• The odds ratios for changes in the control worksites from baseline or end of the intervention to the 2-year follow-up were not significant (p>0.01)
Related Papers:	Target Group: Worksite management			• The greatest period of increased smoking control activities in the control worksites occurred between baseline and the end-of-intervention survey before dissemination occurred
Patterson, R⁹² (1998)	Quality Assessment Rating: Moderate			• Dissemination of the program to control worksites had little impact on the level of smoking control in activities in control worksites
Dissemination of worksite nutrition interventions from the Working Well Trial refer to the Adult Healthy Diet chapter
Tremblay, M¹³⁸ (2001) Canada	Study Design: Interrupted times series	Multiple dissemination strategies used	Multiple interventions (Self-help materials for patients who smoke, practice guideline that was based on the Agency for Health Care Policy and Research's smoking cessation guideline, counseling aids for physicians, and lists of smoking cessation resources in Montreal)	Results from 2 cross-sectional surveys:
Purpose: To describe the theoretical model underlying the “Physicians Taking Action Against Smoking” program, a five year intervention program to improve the smoking cessation counseling practices of general practitioners (GPs) in Montreal. In addition, to report the results of 2 cross-sectional surveys of GPs which assessed their use of cessation counseling practices.	Two mailed cross-sectional surveys of random samples of GPs	(Postal delivery, workshops, and conferences)		• The response rate from the random samples of GPs for the 1998 survey was 76.6% and 69.6% for the 2000 survey. Respondents to the 1998 survey were younger (mean = 45.3, SD = 9.5) than respondents to the 2000 survey (mean = 47.8, SD = 10.4); p = 0.002). Otherwise, the respondents were similar in terms of sex ratio, language spoken and residency training
Related papers:	Survey one year post-implementation:			• For male respondents: No significant differences were found between the 1998 and 2000 surveys in reported attitudes, perceived self-ability or interest in updating smoking cessation counseling skills or reported barriers to providing counseling. For female respondents: There were significant increases from 1998 to 2000 in number reporting to have the required skills to counsel patients in smoking cessation (32.8% to 48.1%, p=0.028); and a significant decrease in respondents reporting “patients not interested” as a barrier to smoking cessation counseling (56.6% to 43.0%, p = 0.019)
O'Loughlin, J¹⁴² (2001)	n = 440 (1998)			• In both 1998 and 2000, there were significant differences between male and female respondents in the number reporting use of a system to identify smokers (M 42.8% vs. F 56.1% (1998), p < 0.05) and (M 37.6% vs. F 51.9% (2000), p < 0.05). There was no significant difference between survey years
Results of the 1998 cross-sectional survey	Survey 3 years post-implementation:			• The proportion of female GPs who reported offering patients smoking cessation self-help materials increased significantly from 1998 to 2000 (reported according to patient's stage of readiness; combined results not reported, all p < 0.05). There were no significant increases observed for male respondents
	n = 454 ( 2000)
	Target Group: All GPs in Montreal (n = 2,130)
	Quality Assessment Rating: Weak
Wilson, E¹²⁸ (1984) Canada	Study Design: 1 group, process evaluation	Postal delivery of an introductory letter and information about the “Time to Quit” program was mailed to all pharmacies in the greater Winnipeg area (A second mailing of the information was sent out after the Manitoba Pharmaceutical Association's news bulletin carried a full page discussion of the program)	A “Time to Quit” self-help smoking cessation booklet	• In Winnipeg, 82% (108) of city pharmacies received a total of 40,500 booklets (no further data provided)
Purpose: To examine whether pharmacies and pharmacists are effective at disseminating patient self-materials on smoking cessation.	• n = 132 city pharmacies		The “Time to Quit” program consists of 3 elements: (1) The self-help booklet, (2) a media campaign (predominantly television) that promoted awareness of the self-booklet and its availability at participating pharmacies), and (3) a guide to communities implementing the program	• Results of a follow-up survey indicated that 93.5% of the booklets had been picked up (no further data provided)
	Target Group: Pharmacists
	Quality Assessment Rating: Weak

Evidence Table 7. Strategy for dissemination of cancer control interventions in adult healthy diet (Key question 7)

Author (Year) Country Study purpose	Study design Target group Quality Assessment	Dissemination strategy evaluated	Intervention(s)	Findings
Albright, CL¹³⁴(1992) United States	Study Design: Cohort, one group, pre/post n = 10 internists were trained as “facilitators” at Stanford, then they delivered seminars to 91 faculty members home-site seminars	Train-the-Trainer (Month-long training, off-site)	CPM curriculum that encompassed in 6 subject areas: smoking cessation, clinical nutrition, weight control and exercise, hyperlipidemia and national screening guidelines. The curriculum also provided content on clinical teaching and medical decision-making The clinical nutrition section included interventions that promoted adult healthy diets	Training the Trainers:
Purpose: To test a dissemination model for providing clinical preventive medicine (CPM) training to internal medicine faculty across the U.S.	Target Group: General Internal Medicine faculty in the US			Fidelity ratings obtained from videotapes of the home-site seminars given by the trainers indicated that the trainers adhered closely to the CPM curriculum (For example, slides were provided for teaching purposes. Over 90% of the slides provided were used in all 3 years of the study)
Refer to Adult Smoking	Quality Assessment			Trainer teaching home-site internal medicine faculty:
Cessation Evidence	Rating: Weak			Knowledge of smoking cessation, chotesterol and screening specific information of faculty members significantly increased post-seminar in the last 2 of the 3 years studied (P<0.02 and P<0.001)
Table for additional information.				Faculty's rating of their self-efficacy to implement the CPM practices for clinical nutrition demonstrated significant post-test increases for all 3 years (P<0.0001)
				Faculty use of specific behaviour-charge interventions to promote healthy diet was increased (diaries, self-help materials, and social support) (P<0.05)
				85% of the faculty in 1986, 96% in 1987, and 84% in 1988 reported teaching the CPM strategies to their house staff
				House staff trained by faculty that attended the CPM seminars:
				House staff were surveyed before and after faculty participated in the CPM seminars. Significant increases were reported in the degree to which the faculty addressed CPM in their teaching interactions (1987:p<0.0001 and 1988:p<0.04)
				House staff ratings of their self-efficacy to implement specific prevention medicine strategies increased in both years (1987:p<0.0001 and 1988:p<0.002)
Anderson, DM¹⁴⁰(1989) United States	Study Design: Descriptive study	Multiple media sources (television, radio, and newspapers)	CIS - a telephone based information and education program of the National Cancer Institute.	Results pertaining to nutrition are reported in this evidence table.
Purpose: To examine inquiries received by the CIS, a telephone hot-line, to determine:	Retrospective analysis of 5 years of inquiries to one national and 26 local CIS offices in 4 subject areas. A standardized call record form was completed for each call.			From the most frequently reported to the least frequently reported, source of callers' learning about the CIS hot-line were: television (33.9%), radio (28.2%), publications (included newspapers, magazine, pamphlets and posters) (26.9%), health care providers (4.3%), significant others (4.6%); and telephone assistance (2.2%)
1) effects of different media in stimulating calls to the CIS, and	Demographic information was only collected during the last 2 years of the study for first-time, non-health professional callers and was limited by federal stipulations to 20% of callers in 5 CIS offices n = 57,374 nutrition- related calls over the five years studied			The following findings were not reported by topic in the paper: Combined across topics, television was the most common information source reported by callers for both sexes (72.2% male callers and 60.7% of female callers)
2) demographic characteristics of callers in 4 cancer prevention and early detection subjects: smoking, nutrition, Pap smear screening and breast self-examination	Target Group:			There was an inverse relationship between frequency of television cited as an information source and callers age and education. In the 19-year old or less age group 81.7% of callers cited television compared to 39.6% of callers in the 60-year or older age group
Refer to Adult Smoking	Smokers in the US, Mexico and other countries			Television was the predominant source for 4 of the 5 ethnic groups identified across all topics - Caucasians, African-Americans, Hispanics, and Native Americans. For callers of Asian or Pacific Island heritage, the most frequently cited source was publications (46.7%), followed by television (32.1%). Further analysis suggested news publications, not health publications, were the greatest source of information for this ethnic group
Cessation and Cervical	Quality Assessment
Cancer Screening	Rating: Weak
Evidence Tables for additional information
Buller, DB¹⁴⁷(1999) United States	Study Design: RCT	Peer-educators (n=42) (Chosen for ‘centrality’, rated highest by peers in communication ties and flow)	5-a-Day message (to increase fruit and vegetable intake) plus accompanying 5-a-Day printed materials	Immediate changes post-intervention in awareness, attitudes, and dietary behavior:
Purpose: To test a peer-education strategy to promote the 5-a-Day message.	Experimental group:	Peer - educators attended a 16-hour training program which included training in persuasive communication techniques		Employees receiving peer- education increased their awareness of the 5-a-Day program (p<0.001), knowledge of the 5-a-Day concept, attitudes toward fruit and vegetable intake (p=0.024 - <0.001)), and number of daily servings of fruits and vegetables consumed (Increase of 0.77 on 24-hour intake recall p<0.001 and 0.46 on food frequency questionnaire items p=0.002)
Related Papers:	n = 505 employees (in 46 cliques) assigned to receive 5-a-Day peer-education plus the general 5-a-Day program	The expectation was that they would spend 2 hours/week discussing fruit and vegetable intake with coworkers		Persistence of changes in awareness, attitudes, and dietary behavior (6-month followup):
(1) Buller, DB¹⁴⁸(2000) (Same project; see additional findings below)	Control group:			There was a general persistence of the statistically significant increases in the peer-education group, but of reduced magnitude for knowledge of the 5-a-Day program and diet-related attitudes
(2) Larkey, LK¹⁴⁹(1999) (Same project, see additional findings below)	n = 492 employees (in 46 cliques) assigned to receive the general 5-a-Day program (cafeteria promotions and workplace mail) only			Statistically significant increases total number of daily servings persisted when measured by 24-hour intake recall (0.41, p=0.034), but not by food frequency questionnaire
	Target Group: Anglo and Hispanic worksite population in the US (lower socioeconomic labor and trade employees)
	Quality Assessment
	Rating: Weak
Buller, DB¹⁴⁸(2000) United States	Study Design: RCT	Peer-educators (n=42) (Chosen for ‘centrality’, rated highest by peers in communication ties and flow)(Peer- educators attended a 16-hour training program which included training in persuasive communication techniques The expectation was that they would spend 2 hours/week discussing fruit and vegetable intake with coworkers)	5-a-Day message (to increase fruit and vegetable intake) plus accompanying 5-a-Day printed materials	Findings reported in this evidence table pertain only to the peer-education (experimental group)
Purpose: To assess the implementation and reach (extent of implementation) of the peer-education component of the randomized trial described above in Buller, D (1999) ID#3018.	Experimental group:			57% of employees reported printed materials stimulated discussion of fruits and vegetables with co-workers during program, 31% still discussing 6 months later, 69% discussed printed material with a family member during intervention Greater contact with peer- educators was related to larger immediate increases in total consumption of fruits and vegetables (p=0.003) as measured by food frequency items When food types were examined separately, peer-educator contact was positively related to immediate increased vegetable intake (p=0.002), but not to fruit or juice intake
Related Papers:	n = 505 employees (in 46 cliques) assigned to receive 5-a-Day peer-education plus the general 5-a-Day program			The more employees reported reading the printed material, the smaller the observed immediate increase in fruit consumption (p=0.002)
(1) Buller, DB¹⁴⁷(1999) (Same project, see additional findings above)	Control group:			There was no significant association between peer contact and changes in total intake of these foods at the 6-month followup
(2) Larkey, LK¹⁴⁹(1999) (Same project, see additional findings below)	n = 492 employees (in 46 cliques) assigned to receive the general 5-a-Day program (cafeteria promotions and workplace mail) only
	Target Group: Anglo and Hispanic worksite population in the US (lower socio-economic labor and trade employees)
	Quality Assessment
	Rating: Weak
Dietrich, AJ ¹⁴⁴(1992) United States	Study Design: RCT	(1) Facilitators: Visited each practice 3–4 times over 3 months; each visit lasted approximately 120 minutes. Performed an initial audit of each practice to assess the status of preventive care and assisted practices in the design and implementation of office system interventions. Practices only implemented those interventions that meet their perceived needs	Multiple office-system interventions including preventive care flow sheets, chart stickers, health education posters and brochures, and patient health diaries (None of the interventions were computer-based)	Results pertaining to adult healthy diet are reported in this evidence table.
Purpose: To test the impact of physician education and facilitator assisted office-system interventions on cancer early detection and preventive services	In total, 98 of the 102 practices that agreed to participate completed the study. The unit of randomization was the medical practice as represented by one physician.	(2) Facilitator + workshop Same as (1) plus physician from each practice attended a 1-day workshop led by an expert who reviewed NCI's prevention and screening recommendations and taught specific skills. Also provided a written syllabus		The response rate for the cross-sectional survey pre-experiment was 91% (n = 2436 patients) and 93% (n = 2595) at 12-month followup.
Refer to Adult smoking cessation,	Four groups:	Note: The workshop only and the control groups did not receive information on the use of office-systems interventions for cancer prevention or early detection		Significantly more eligible patients in the Facilitator Only group reported their physician had advised them to reduce fat intake compared to patients in the control group at 12-month followup (Proportion: 0.56 vs. 0.47, p<0.05, baseline results were used as covariates).
Mammography, and Cervical cancer screening Evidence Tables for additional information	Facilitator only: n = 24 practices Workshop + Facilitator: n = 26 practices			There was nosignificant increase in the number of eligible patients in the facilitator + workshop group reporting their physician had advised them to reduce fat intake compared with patients in the control group at 12-month followup (Proportion 0.51 vs. 0.47)
	Workshop only: n = 24 practices			There was nosignificant increase in the number of eligible patients in the facilitator only or facilitator + workshop groups reporting their physician had advised them to increase fibre consumption compared to patients in the control group at 12-month followup (Proportion 0.48 (facilitator only) vs. 0.38 (control); Proportion 0.41 (facilitator + workshop) vs. 0.38 (control))
	Control: n = 24 practices; no intervention; no further detail provided			Report's overall conclusion: Community practices assisted by a facilitator in the development and implementation of an office system can substantially improve provision of cancer early detection and preventive services
	Target Group: Office-based GPs and general internists in New Hampshire and Vermont.
	Quality Assessment
	Rating: Weak
Larkey, LK¹⁴⁹(1999) United States	Study Design:	Peer-educators (n = 42) (Chosen for ‘centrality’, rated highest by peers in communication ties and flow)	5-a-Day message (to increase fruit and vegetable intake) plus accompanying 5-a-Day printed materials	Peer-health educators were more likely to use “role modeling”(p=0.0004) and “creating context” (p<0.0001) as collective (group) change strategies, while “encouragement” (p=0.0009) and “responding to employee needs” (p=0.0001) were more likely to be used as individual change strategies
Purpose: To assess which persuasive strategies are used by peer-health educators, which strategies are used more in one-on-one vs. group contexts, and which strategies are most likely to be used by males and by females	Qualitative, descriptive study of peer-educator within original randomized trial			Strategy use differed by gender. “Mock competition,” “giving materials,” and “encouragement” were used by men significantly more than “creating context” and “keeping 5-a-Day visible”. Women used “creating context” and “keeping 5-a-Day visible” significantly more than “mock competition”, “giving materials” and “encouragement” (p<0.0001 for all contrasts)
Related Papers:	Target Group: Anglo and Hispanic worksite population in the US			Hispanic peer health educators were more likely to use individual change strategies than their non-Hispanic counterparts (p=0.0128)
(1) Buller, DB¹⁴⁸(2000) (Same project, see additional findings above)	Quality Assessment
(2) Buller, DB¹⁴⁷(1999) (Same project, see additional findings above)	Rating: Weak
Patterson, RE⁹²(1998) United States	Study Design: RCT	Intervention materials were given to control sites (method not specified) at the conclusion of the Working Well Trial	Nutrition interventions from the Working Well Trial	Only the findings pertaining to the control sites at the conclusion of the Working Well Trial are reported in this table.
Purpose: To test whether the Working Well Trial nutrition intervention activities were maintained after the research program or were adopted by control sites.	n = 54 control sites			In the control worksites, there was a significant increase in nutrition activity score (composite of nutrition classes or weight loss programs; self-help nutrition manuals and guides; videotapes, posters or brochures related to nutrition) from baseline to the end of the Working Well trial (2-year interval between start and completion of the trial) (p=0.0012)
Related Papers:	Targe Group: Worksite management			There was no significant increase in the control sites, between the end of the Working Well trial (point of dissemination of the nutrition interventions to control sites) and at the followup survey conducted 2 years later on nutrition activity score
Sorensen, G¹³⁷ (1998) Dissemination of worksite smoking cessation interventions from the Working Well Trial to control cited at the end of the trial.	Quality Assessment			At followup, there was no significant difference between nutrition activity scores in the intervention compared with the control worksites
Refer to Adult smoking cessation Evidence Tables for additional information	Rating: Moderate
Samuels, SE¹⁵⁰(1993) United States	Study Design: One group, post-intervention, process evaluation	Media awareness campaign (television and print media)	Telephone hot-line which provided advice and offered an information booklet to callers	Only findings pertaining to the media awareness campaign effect on calls to the telephone hot-line are reported in this table.
Purpose: To report on the first 3 years of the Project LEAN campaign, a national social marketing intervention, designed to promote dietary change.	Target Group: General adult population			As a result of the ads and campaign publicity, the hot-line received nearly 300,000 consumer calls
	Quality Assessment			During the first 12 months of the campaign, calls peaked at 25,000 to 28,000 a month
	Rating: Weak			As publicity declined, so did calls to the hot-line
				Hot-line terminated at end of 18 months due to expense (over $300,000 per year)
Tziraki, C (2000)¹⁵¹United States	Study Design: RCT	(1) Workshop (One staff member from each practice was invited to attend a 3 hour training session. Training was provided in the 4 major components of the manual; how to organize the office environment; how to screen patient adherence; how to provide dietary advice; and how to implement a patient followup system)	NCI sponsor Nutrition Manual for cancer prevention (contains multiple interventions)	Followup interviews and observational assessments were conducted at 4–6 months post-dissemination with a physician and staff member from each practice. Adherence scores were calculated for 4 areas: office organization, nutrition screening, nutrition advice or referral, and patient followup
Purpose: To determine the effectiveness of two strategies for promoting the use of an NCI nutrition manual by primary care physicians and their office staff	810 practices were randomized; 55 practices had a change in status and became ineligible after randomization	(2) Postal delivery of the manual only (no training)	- Modeled after the NCI publication: “How to help your patients stop smoking.” The manual addressed brief counseling techniques, office system organization, material resources, staff training, and patient educational materials	< 50% of practices assigned to the workshop group sent representatives to the training workshop (120 of 244)
	Workshop group: n = 244 practices received the manual and were invited to a training workshop			The workshop group was significantly more adherent to the manual's recommendations for office organization at followup than the Postal-Delivery group (28.5% vs. 24.7%, p<0.005) and Control group (28.5% vs. 23.0%, p<0.001); these analyses included all practices in the Workshop group regardless of attendance at the training session. Of those practices who attended the workshop, 30.6% were adherent to the recommendations for office organization
	Postal delivery group: n = 256 practices were mailed the manual			The Workshop group was significantly more adherent to the manual's recommendations for nutrition screening at followup than the Postal Delivery group (23.5% vs. 21%, p<0.05) and Control group (23.5% vs. 20.5%, p<0.05). Of those practices attending the workshop, 25% were adherent to the recommendations for nutrition screening
	Control group: n = 255 practices; did not receive the manual			There was no significant difference between the Postal delivery and control groups for office organization (24.7% vs. 23.0%) or nutrition screening (21% vs. 20.5%)
	Target Group: Free-standing primary care practices in Pennsylvania and New Jersey			There was no statistically significant difference between the 3 groups for nutrition advice (Workshop 54.9%; Post 53%; Control 52.3%) nor for patient followup (Workshop 14.6%; Postal Delivery 13.6%; Control 13.6%). The attending workshop practices were significantly more likely than either Postal Delivery (57% vs. 53%, p<0.05) or Control groups (57% vs. 52.3%, p<0.05) to provide nutrition screening
	Quality Assessment
	Rating: Weak

Evidence Table 8. Strategies for dissemination of cancer control interventions in mammography (Key question 8)

Author (Year) Country Study Purpose	Study Design Target Group Quality Assessment	Dissemination Strategy Evaluated	Interventions	Findings
Dietrich, AJ¹⁴⁴ (1992) United States	Study Design: RCT	(1) Facilitator Visited each practice 3 to 4 times over 3 months; each visit lasted approximately 120 minutes. Performed an initial audit of each practice to assess the status of preventive care and assisted practices in the design and implementation of office system interventions. Practices only implemented those interventions that meet their perceived needs	Multiple office system interventions including preventive care flow sheets, chart stickers, health education posters and brochures, and patient health diaries (None of the interventions were computer-based)	Results pertaining to mammography are reported in this evidence table.
Purpose: To test the impact of physician education and facilitator assisted office system interventions on cancer early detection and preventive services	In total, 98 of the 102 practices that agreed to participate completed the study. The unit of randomization was the medical practice as represented by one physician.	(2) Facilitator + workshop Same as (1) plus physician from each practice attended a 1 day workshop led by an expert who reviewed NCI's prevention and screening recommendations and taught specific skills. Also provided a written syllabus		The response rate for the cross-sectional survey pre-experiment was 91% (n = 2,436 patients) and 93% (n = 2,595) at 12 months followup:
Findings specific to smoking cessation, healthy diet and cervical cancer screening are reported in their respective topic sections of this results chapter	Four groups:	Note: The workshop only and the control groups did not receive information on the use of office systems interventions for cancer prevention or early detection		• More patients in each of the 3 experimental groups reported having a mammogram than patients in the control group at 12 month followup (facilitator + workshop vs. controls proportion: 0.78 vs. 0.57, p<0.01; facilitator only vs. controls 0.77 vs. 0.57, p<0.01; and workshop only vs. controls 0.71 vs. 0.57, p<0.01; baseline proportions were used as covariates)
	Facilitator only: n = 24 practices			• There was no significant difference between the facilitator + workshop, facilitator only or workshop only groups in proportion of patients reporting having had a mammogram at 12 month followup
	Workshop + Facilitator: n = 26 practices			A chart review of 2,032 patient records was also performed:
	Workshop only :n = 24 practices			• Rate ratio for performance of mammography was 1.54 (95% CI 1.09 to 2.17) for Facilitator Only group compared to control group
	Control: n = 24 practices; no intervention; no further detail provided			• Rate ratio for performance of mammography was 1.60 (95% CI 1.19 to 2.15) for Facilitator + Workshop group compared with control group
	Target Group: Office based GPs and general internists in New Hampshire and Vermont.			Report's overall conclusion: Community practices assisted by a facilitator in the development and implementation of an office system can substantially improve provision of cancer early detection and preventive services
	Quality Assessment Rating: Weak
Kinsinger, LS¹⁵³ (1998) United States	Study Design: RCT	Academic detailing (Facilitators met with practice physicians and staff in the intervention group an average of 3 times with additional telephone calls or drop-in visits over a period of 12–18 months to assist in developing office systems tailored to increase breast cancer screening)	Office systems (Defined as an organized approach within a medical practice for routinely providing a given service, such as breast cancer screening, to patients for whom this service is indicated. These systems involve teamwork among a number of office staff, not just physicians. Tools such as flow sheets, chart prompts and patient educational materials can all be part of an office system, but most important is how these materials are integrated within the usual procedures of the practice)	Significant increases in 3 of 5 indicators in intervention compared to control practices from baseline to followup:
Purpose: To evaluate an outreach intervention designed to improve performance rates of breast cancer screening through implementation of office systems in community primary care practices	Experimental group: Academic detailing on how to develop office systems for breast cancer screening			Indicator 1: Practices with ≥ 50% of records having an entry on a flow sheet increased from 10–29% in intervention practices compared to a decrease from 19-7% in control practices (p = 0.02); Indicator 2: Practices in which ≥ 50% of physicians report having written preventive care policy increased from 16–57% in intervention practices compared to a decrease from 13–7% in control practices (p = 0.01); and Indicator 3: Practices in which ≥ 50% of physicians report that nurses frequently or sometimes recommend mammograms to patients increased in intervention practices from 41–58% compared to a decrease from 48-33% in controls (p = 0.04)
	n = 32 practices			No significant differences were found for the 2 other indicators in intervention compared to control practices from baseline to followup. Indicator 4: The percentage of practices in which ≥ 50% of physicians report that nurses identify patients due for mammograms (intervention 37–65%; control 39–44%). Indicator 5: The percentage of practices where ≥ 50% of physicians report frequent use flow sheets or computerized reminders to identify patients due for mammograms (intervention 35–65%; control 29–44%). Mean number of indicators increased significantly in intervention practices (1.3–2.8) compared with control practices (decrease 1.5-1.4)(p = 0.0003). However, at followup, only 23% of intervention practices reported a complete office system for breast cancer screening The proportion of records with “mention” of mammogram in the last year increased significantly more in intervention practices than in controls (12.7% vs. 3.5%, p = 0.014). However, there was no difference between intervention and control practices in the change in proportion of women's records with a mammogram report in the last year (4.7% vs. 3.4%).
	n = 68 physicians
	n = 1,444 patient records
	Control group:
	No information on use office systems provided.
	n = 30 practices
	n = 66 physicians
	n = 1,523 patient records
	Target Group: Physicians and staff of family and general internal medical practices in predominantly rural counties in North Carolina (to be eligible the physicians had to provide at least 20 hours of primary care per week)
	Quality Assessment Rating: Strong
Lemelin, J¹⁴³ (2001) Canada	Study Design: RCT	Educational facilitators Over an 18 month period each practice was visited an average of 33 times; each visit lasted approximately 1 hour	Multiple interventions including reminder systems, flow charts and patient educational materials	Random chart audit of 100 records/practice was performed a baseline and again at followup:
Purpose: To evaluate a multifaceted outreach intervention, delivered by nurses trained in prevention facilitation, to improve prevention in primary care	Of the 95 practices contacted, 49 chose not to participate. In total, 46 practices were randomized. One practice in the facilitator group was lost to followup	The facilitators performed an initial audit and feedback of each practices baseline preventive performance rates; facilitated the development of practice goals and policy for preventive care; and assisted practices in selecting and implementing interventions to improve preventive care)		• At baseline, the preventive performance index was not significantly different between the facilitator and control groups (31.9% and 32.1%, respectively). At followup, the corresponding values were 43.2% and 31.9%, the absolute increase in the facilitator group was of 11.5% was statistically significant (p < 0.001)
Related Papers:	Facilitator group: n = 22 practices (total of 54 physicians) completed the study; received visits from educational facilitators			Mammography specific findings:
Baskerville, N ¹⁶⁸ (2001)	Control group: n = 23 practices (total of 55 physicians) completed the study; received no visits.			• On chart audit, at baseline, 53.6% of eligible patients had mammograms in the facilitator group and 53.4% in the control group. At followup, the corresponding values were 67.5% and 58.7%; there was no significant difference in change between the two groups
(Process Evaluation)	Target Group: Primary care practices that have a payment system based primarily on capitation in Ontario, Canada			Overall findings from the process evaluation:
Refer to Adult Smoking Cessation Evidence Table 6 and Cervical Cancer Screening Evidence Table 9 for additional information	Quality Assessment Rating: Weak			• All facilitator group practices received preventive performance audit and feedback, achieved consensus on a plan for improvement, and implemented a reminder system. 90% implemented a customized flow sheet, 10% used a computerized reminder system, 95% wanted critically appraised evidence for prevention, and 100% received patient educational materials
				• Audit and feedback, consensus building, and development of reminder systems were identified as the key components by content and bivariate analysis
				• 95% of physicians were satisfied or very satisfied with the educational facilitator approach
Paskett, ED¹⁵⁴ (1999) United States	Study Design: Four groups, post-test only	Mail delivery of introductory letters followed by telephone recruitment	Choice of 3 increasingly intensive interventions:	• Of the 97 worksites that completed the baseline survey, 63 (65%) accepted and offered a program to their employees: 14 worksites chose the intensive nurse training, 14 sponsored worksite classes, and 35 chose the educational display of brochures
Purpose: To assess how worksites in a selected community would respond to sponsoring a breast cancer education program if the program could be varied to match the interest level and degree of involvement the worksite wanted in such a program	n = 102 worksites approached		(1) Educational display with brochures, (2) nurse-led educational sessions or (3) training worksite nurses to provide educational classes and one-on-one counseling	• Worksites with a greater percentage of female employees over age 40 were more likely to sponsor a program (p < 0.05)
	n = 97 completed baseline survey			• Worksites that chose to sponsor a program were more likely to have already sponsored breast cancer education programs at their worksites (p = 0.027) or to have a medical department (p = 0.006)
	Target Group: Senior management			• Type of component chosen was significantly associated with a history of sponsoring other health education programs (p < 0.001). Worksites that had sponsored a similar program in the past were more likely to send a nurse to be trained
	Quality Assessment Rating: Weak			• Of worksites that had never sponsored a breast cancer program (n = 73), 43 (59%) were receptive to this program and the majority of these 29 (67%) chose the educational display with brochures
Scott, TL¹⁵⁵ (1999) United States	Study Design: Descriptive study	Compared 2 strategies:	Intervention manual:	• There was little difference in interventions implemented between plans that attended the workshop and those that did not (no statistical analysis reported)
Purpose: To assess the dissemination of a mammography intervention manual in a managed care setting and to measure the effect of the manual on the managed care organization's choice of intervention strategies.	n = 8 Managed care organizations; selected based on low mammography rates and geographical diversity (2 managed care organizations were selected in each of 4 regions) One managed care organization in each of the 4 geographical regions was randomly assigned to the intensive workshop group	(1) One-day workshop and user guide to accompany the intervention manual	Summarized research findings, highlighted the most effective intervention strategies and provided practical material (such as template letters and telephone scripts)	• Seven of the managed care organizations implemented more interventions in the year after receiving the manual than in the year prior to dissemination of the intervention manual and there was an improvement in the type of interventions implemented (i.e., evidence-based interventions). The 7 managed care organizations that used the manual all implemented an intervention directed to physicians. Some managed care organizations also implemented interventions directed towards patients (e.g., reminder letters)
	Target Group: Managed care organizations	(2) Passive dissemination (Delivery of the intervention manual to the managed care organizations)		• Mammography rates in the year prior to dissemination of the manual were compared to the rates in the year after dissemination. In all 7 of the managed care organizations that used the intervention manual, mammography-screening rates increased (range: 0.22–4.0%). In the 1 managed care organization that did not use the intervention manual, its mammography-screening rate decreased 2.67%
	Quality Assessment Rating: Weak			• A key factor for intervention implementation appeared to be the length of employment of the point person. The 2 plans that implemented the least intensive interventions had point people who had only been in their positions for short periods of time
				• Factors facilitating use of the manual and implementation of interventions were: (1) motivation of the point person to improve mammography rates, (2) support of senior management, (3) adequate resources (time, personnel, and funds), and (4) organization and content of the intervention manual
				• Barriers to use and implementation were (finances, time, and programming) and data limitations (identifying population to be targeted and getting correct contact information)
Williams, PT¹⁵⁶ (1994) United States	Study Design: One group, pre-post test	Academic detailing (by either a study nurse or physician)	Multiple interventions: medical record prompts, recall systems and patient educational materials	Only topic-specific findings pertaining to breast and cervical cancer screening are reported.
Purpose: To test the feasibility of “academic detailers” calling on GPs in their offices and to determine if they: (1) facilitate the office management of cancer prevention activities, and (2) increase doctors' knowledge and use of educational and patient service resources of the American Cancer Society (ACS). The study also sought to determine what barriers prevent performance of cancer prevention and screening activities in GPs' offices	n = 10 practices			Practices were assessed at baseline and at followup. The followup time frame was not reported.
	n = 22 physicians			• Baseline: Only one of the practices used the ASC patient information. Followup: All 10 practices used the ACS patient information and 9 displayed the information in the wall racks provided
	n = 85 staff members			• Baseline: Two practices used some form of prompt on the medical record (both indicated the date of the last Pap test). In 2 other practices, nurses were responsible for determining what preventive procedures were due (but no chart summary or prompt existed). Followup: There were only minor changes to medical records. Practices that had not previously used chart summaries or prompts did not add them. However, practices that previously used chart summaries or prompts added items, typically Pap test and mammography notations
	Target Group: GPs and their support staff			• Baseline: One practice had a recall system for scheduling mammography and 5 had a recall system for Pap tests Followup: One practice with a Pap test recall system at baseline added mammography recalls, and one practice with no recall system at baseline added both mammography and Pap recalls
	Quality Assessment Rating: Weak			• The total cost of the 17 office visits by the academic detailers was $US 913
				• Barriers to delivering preventive care: time, administrative process and lack of third party reimbursement

Evidence Table 9. Strategies for dissemination of cancer control interventions in cervical cancer screening (Key question 9)

Author (Year) Country Study purpose	Study Design Target Group Quality Assessment	Dissemination Strategy Evaluated	Intervention(s)	Findings
Anderson, DM¹⁴⁰ (1989) United States	Study Design:	Multiple media sources (television, radio, and newspapers)	CIS - a telephone based information and education program of the NCI	Results pertaining to cervical cancer screening are reported in this evidence table.
Purpose: To examine inquiries received by the Cancer Information Service (CIS), a telephone hot-line, to determine:	Descriptive study			Telephone assistance (phone book and directory assistance) was the most frequently reported source of learning about the CIS by callers seeking Pap smear screening information (27.7%). The second most cited source was health care providers at 22.7%.Publications (included newspapers, magazine, pamphlets and posters) were cited by 19.7% of Pap screening-related callers. Television was cited by 17.7% of Pap screening-related callers, 10.3% cited significant others and 2% cited radio
1) Effects of different media in stimulating calls to the CIS, and	Retrospective analysis of 5 years of inquiries to one national and 26 local CIS offices in 4 subject areas. A standardized call record form was completed for each call.			The following findings were not reported by topic in the paper: Combined across topics, television was the most common information source reported by callers for both sexes: 72.2% male callers cited television compared to 60.7% of female callers. As the age of the callers increased, the frequency of television cited as the information source decreased. In the 19 year old or less age group, 81.7% of callers cited television compared to 39.6% of callers in the 60 year or older age group
2) Demographic characteristics of callers in 4 cancer prevention and early detection subjects: smoking, nutrition, Pap smear screening and breast self-examination	Demographic information was only collected during the last 2 years of the study for first-time, non-health professional callers, and was limited by federal stipulations to 20% of callers in 5 CIS offices			Television was the primary source reported by callers for all education levels. In general, the lower the caller's level of education, the more frequently television was cited as the information source Television was the predominant source for 4 of the 5 ethnic groups identified across all topics - Caucasians, African-Americans, Hispanics, and Native Americans. For callers of Asian or Pacific Island heritage, the most frequently cited source was publications (46.7%), followed by television (32.1%). Further analysis suggested news publications, not health publications, were the greatest source of information for this ethnic group
Refer to Adult Smoking Cessation and Adult Healthy Diet Evidence Tables for additional information	n = 57,374 nutrition related calls over the 5 years studied
	Target Group: Smokers in the US, Mexico and other countries
	Quality Assessment Rating: Weak
Dietrich, AJ¹⁴⁴ (1992) United States	Study Design: RCT	(1) Facilitator visited each practice 3 to 4 times over 3 months; each visit lasted approximately 120 minutes. Performed an initial audit of each practice to assess the status of preventive care and assisted practices in the design and implementation of office system interventions. Practices only implemented those interventions that meet their perceived needs	Multiple office system interventions including preventive care flow sheets, chart stickers, health education posters and brochures, and patient health diaries(None of the interventions were computer-based)	Results pertaining to cervical cancer screening are reported in this evidence table.
Purpose: To test the impact of physician education and facilitator assisted office system interventions on cancer early detection and preventive services	In total, 98 of the 102 practices that agreed to participate completed the study. The unit of randomization was the medical practice as represented by one physician.	(2) Facilitator + Workshop: Same as (1) plus physician from each practice attended a 1 day workshop led by an expert who reviewed NCI's prevention and screening recommendations and taught specific skills. Also provided a written syllabus		The response rate for the cross-sectional survey pre-experiment was 91% (n = 2436 patients) and 93% (n = 2595) at 12 months follow-up There was no significant increase in the number of eligible patients in the facilitator only group reporting having a Pap test compared with patients in the control group at 12 month follow-up (Proportion 0.71 vs. 0.61)
Refer to Adult Smoking Cessation, Adult Healthy Diet, and Mammography Evidence Tables for additional information	Four groups:	Note: The workshop only and the control groups did not receive information on the use of office systems interventions for cancer prevention or early detection.		There was no significant increase in the number of eligible patients in the facilitator + Workshop group reporting having a Pap test compared to patients in the control group at 12 month follow-up (Proportion 0.65 vs. 0.61)
	Facilitator only: n = 24 practices			Report's overall conclusion: Community practices assisted by a facilitator in the development and implementation of an office system can substantially improve provision of cancer early detection and preventive services
	Workshop + Facilitator: n = 26 practices
	Workshop only: n = 24 practices
	Control: n = 24 practices; no intervention; no further detail provided
	Target Group: Office based GPs and general internists in New Hampshire and Vermont.
	Quality Assessment Rating: Weak
Lemelin, J¹⁴³ (2001) Canada	Study Design: RCTI	Educational facilitators (Over an 18 month period each practice was visited an average of 33 times; each visit lasted approximately 1 hour)	Multiple interventions including reminder systems, flow charts and patient educational materials	Results pertaining to cervical cancer screening are reported in this evidence table.
Purpose: To evaluate a multifaceted outreach intervention, delivered by nurses trained in prevention facilitation, to improve prevention in primary care	Of the 95 practices contacted, 49 chose not to participate. In total, 46 practices were randomized. One practice in the facilitator group was lost to followup	The facilitators performed an initial audit and feedback of each practices baseline preventive performance rates; facilitated the development of practice goals and policy for preventive care; and assisted practices in selecting and implementing interventions to improve preventive care)		Random chart audit of 100 records/practice was performed a baseline and again at followup:
Related Papers: Baskerville, N¹⁶⁸ (2001)	Facilitator group: n = 22 practices (total of 54 physicians) completed the study; received visits from educational facilitators			At baseline, the preventive performance index was not significantly different between the facilitator and control groups (31.9% and 32.1%, respectively). At followup, the corresponding values were 43.2% and 31.9%, the absolute increase in the facilitator group of 11.5% was statistically significant (p < 0.001)
Refer to Adult Smoking Cessation, and Mammography Evidence Tables for additional information	Control group: n = 23 practices (total of 55 physicians) completed the study; received no visits			Cervical cancer screening specific findings:
	Target Group: Primary care practices that have a payment system based primarily on capitation in Ontario, Canada			On chart audit, at baseline, Pap testing was performed with 60.8% of eligible patients in the facilitator group and with 57.9% in the control group. At followup, the corresponding values were 66.2% and 59.1%, there was no significant difference in change between the two groups
	Quality Assessment Rating: Weak			Overall findings from the process evaluation:
				All facilitator group practices received preventive performance audit and feedback, achieved consensus on a plan for improvement, and implemented a reminder system. 90% implemented a customized flow sheet, 10% used a computerized reminder system, 95% wanted critically appraised evidence for prevention, and 100% received patient educational materials
				Audit and feedback, consensus building, and development of reminder systems were identified as the key components by content and bivariate analysis
				95% of physicians were satisfied or very satisfied with the educational facilitator approach
Williams, P¹⁵⁶ (1994) United States	Study Design:	Academic detailing (By either a study nurse or physician)	Multiple interventions: medical record prompts, recall systems and patient educational materials	Only topic-specific findings pertaining to breast and cervical cancer screening are reported.
Purpose: To test the feasibility of “academic detailers” calling on GPs in their offices and to determine if they:	One group, pre-post intervention			Pre-intervention: Only one of the practices used the ASC patient information. Post intervention: All 10 practices used the ACS patient information and 9 displayed the information in the wall racks provided
(1) Facilitate the office management of cancer prevention activities; and	n = 10 practices			Pre-intervention: Two practices used some form of prompt on the medical record (both indicated the date of the last Pap test). In 2 other practices, nurses were responsible for determining what preventive procedures were due (but no chart summary or prompt existed). Post-intervention: There were only minor changes to medical records. Practices that had not previously used chart summaries or prompts did not add them. However, practices that previously used chart summaries or prompts, added items, typically Pap test and mammography notations
(2 Increase doctors' knowledge and use of educational and patient service resources of the American Cancer Society (ACS). The study also sought to determine what barriers prevent performance of cancer prevention and screening activities in GPs' offices	n = 22 physicians			Pre-intervention: One practice had a recall system for scheduling mammography and 5 had a recall system for Pap tests. Post-intervention: One practice with a Pap test recall system at baseline added mammography recalls, and one practice with no recall system pre-intervention added both mammography and Pap recalls The total cost of the 17 office visits by the academic detailers was $US 913
Refer to Mammography Evidence Tables for additional information	n = 85 staff members			Barriers to delivering preventive care: time, administrative process and lack of third party reimbursement
	Target Group: GPs and their support staff
	Quality Assessment Rating: Weak

Evidence Table 10. Strategies for the dissemination of cancer control interventions in control of cancer pain (Key question 10)

Author (Year) Country Study Purpose	Study Design Target Group Quality Assessment Rating	Dissemination Strategy Evaluated	Interventions	Findings
Breitbart, W¹⁵⁷ (1998)	Study Design: One group pre—post-test design	Train-the-Trainer	Multidisciplinary approach to pain management	Impact of the Observership program on participants' local training activities at 1-year followup:
United States	The study included all of the health care providers who participated in the Observership program between February 1, 1993 and June 15, 1996	(2-week Observership program off-site at Memorial Sloan-Kettering Cancer Center. Observers were selected based on their potential to promote pain management education and training at their own institution)		The followup survey was mailed to the 88 participants who had completed the program in 1993 or 1994. Only 34 (37%) participants returned completed questionnaires. The responding observers reported a significant increase in their local educational and training activities in the year following participation in the Network Project (p<0.01)
Purpose: To evaluate the efficacy of the Network Project Observership program at the Memorial Sloan-Kettering Cancer Center. The program consists of a two-week observership in cancer pain management, psychosocial oncology and cancer rehabilitation.	n=152 observers			There was no significant difference between physician and non-physician observers in the increase in number of educational presentations made in the year following participation in the Observership program
	(Within this group there were 69 physicians, 45 nurses, 16 psychologists, 15 social workers and 7 from other disciplines (e.g. , pharmacists)			Change in knowledge and attitudes of participants toward cancer pain management:
	Target Group: Health care providers			There was a significant improvement in participants' knowledge of cancer pain and cancer pain management after observership training compared with baseline (p<0.05)
	Quality Assessment Rating: Weak
DuPen, A¹⁵⁸ (2000)	Study Design: RCT	Compared 2 dissemination strategies:	Cancer pain algorithm for ambulatory care	Impact of training session on physician adherence to guidelines in prescribing pain medications:
United States	Randomization was by institution (n=9)	(1) 1-day training session with opinion leaders	The algorithm is a decision-tree model based on the Agency for Healthcare Policy and Research (AHCPR) guideline for cancer pain management	Overall, there was no significant change in opioid prescribing practice of physicians in the training group (baseline mean 2.3, SD 0.8; followup mean 2.6, SD 0.6) compared with physicians in the control group (baseline mean 2.4, SD 0.8; followup mean 2.4, SD 0.8) from baseline to followup post-training
Purpose: To evaluate the efficacy of an educational intervention to transfer knowledge on implementing a previously tested algorithm for cancer pain management.	Experimental group:	Used an expert role-model approach with physician-nurse team presenters and provided participants with reference materials including a pain flow sheet that could be incorporated into their routine practice		Prescribing of co-analgesics was low in both groups at baseline and at followup. Training group baseline mean 0.7, SD 1.25 and followup mean 1.3, SD 1.5. Control group baseline mean 0.6, SD 1.3 and followup mean 0.7, SD 1.3 (no statistical analysis reported)
Related Paper:	The number of institutions randomized to the experimental group was not clear (10 oncologists and 20 oncology nurses attended a training session on use of the algorithm)	(2) Passive dissemination Provided the algorithm manual to institutions in the control group with no training		Impact on pain outcomes in patients with locally advanced or metastatic cancer (assessed over a 4-month period):
DuPen, S¹⁶⁹ (1999)	Control group:			There was a significant decrease in the mean level of usual pain reported by patients of physicians/nurses who were in the training group compared with patients of physicians/nurses in the control group (training group baseline mean 3.6, SD 1.9 to followup mean 2.8, SD 1.9 vs. control group baseline mean 3.0, SD 2.0 to followup mean 3.0, SD 2.0) (p = 0.05)
Evaluated the efficacy of the cancer pain algorithm	The number of institutions in the control group was not clear. The control group included 10 oncologists and 18 nurses. No training sessions were provided			Post-hoc analyses found that patients seen in the first 140 days after their physicians attended the training session achieved significantly better outcomes for the worst pain (p = 0.04) and neuropathic pain (p = 0.03) than patients seen more than 140 days after training
	Cancer patients of providers participants in this study were assessed over a 4-month period			Potential confounders: Nurses in the training session group were significantly younger (mean 38 years) compared with nurses in the control group (mean 43 years) (p = 0.02). Study patients in the training group had a significantly higher female-to-male ratio (p = 0.02). There were significant gender differences at baseline with women reporting significantly higher levels of usual (p = 0.02) and worst (p = 0.02) pain than male patients. Patient adherence to pain medication was a confounder in both groups
	n = 54 patients of providers in experimental group
	n = 51 patients of providers in control group
	Target Group: Oncologists and oncology nurses
	Quality Assessment Rating: Weak
Weissman, DE¹⁵⁹ (1995)	Study Design: 1 group pre/post test	Train-the-trainer	Multidisciplinary approach to pain management	Change in cancer pain knowledge:
United States	The study assessed participants of the role model conferences held in 1992-1993	(One-day off-site role model conference to train participants to be role models in their local communities for appropriate pain management)		Participants demonstrated significant improvement in cancer pain knowledge as a result of the 1-day conference (p<0.001 for both of the conference cohorts reported)
Purpose: To report the results of the first year the expanded Wisconsin Cancer Pain Initiative Role Model program. The goal of this program was to train physicians or nurse educators and their clinical partners (nurse, physician or pharmacist) to become role models for better pain management in their clinical practices	n = 196 participants	(There were 3 key components to the conference: workshops, lectures, and development of an Action Plan by each team. The Action Plan was a detailed list of strategies the team would undertake to help change pain management practices)		Impact on subsequent training and educational activities: Within 12 months of attending one of the role-model conferences, 56 (64%) of the 87 team reported that they had either completely (n=37) or partially (n=19) met their Action Plan goals
Related Paper:	In total 87 teams of physicians or nurse educators with one of their clinical partners attended a role model conference during this time period. Participants included 56 physicians, 128 nurses and 12 pharmacists			A total of 227 projects were completed by the 56 teams in the 12 months following the conference (mean = 4.2 projects/team); the most common project was a cancer pain lecture or in-service program for physicians, nurses, or pharmacists; 85 projects with long-term impact were completed (e.g., integrating pain assessment into clinical practice, ongoing health professional education programs, and developing a “pain team”)
Weissman, DE¹⁷⁰ (1993)	Target Group:			Acceptability of the conference format to train role models:
Preliminary report on the Wisconsin Cancer Pain Initiative Role Model program	Physicians and nurses involved in some aspect of education (e.g. medical school faculty, hospital-based nurse supervisors, or physicians who supervise medical students or residents)			Greater than 94% of conference participants responded that the content of the conference was relevant to their practice and at an appropriate educational level. More than 90% of participants reported that they felt the 4 main objectives of the conference were accomplished. The objectives were: (1) disseminating up-to-date cancer pain treatment information; (2) encouraging appropriate attitudes; (3) training physician/nurse educators and partners in pain managements and (4) helping to promote information transfer
	Quality Assessment Rating: Weak