Discussion

Nancy D Berkman; Stacey L Sheridan; Katrina E Donahue; David J Halpern; Anthony Viera; Karen Crotty; Audrey Holland; Michelle Brasure; Kathleen N Lohr; Elizabeth Harden; Elizabeth Tant; Ina Wallace; Meera Viswanathan

This publication is provided for historical reference only and the information may be out of date.

Discussion

Overview

During this systematic review update, the RTI International-University of North Carolina Evidence-based Practice Center (RTI-UNC EPC) identified a moderately large body of literature addressing the relationship between health literacy (including numeracy) and health outcomes. Our two key questions (KQ s) and subquestions were as follows.

Outcomes: Are health literacy skills related to (a) use of health care services, (b) health outcomes, (c) costs of health care, and (d) disparities in health outcomes or health care service use?
Interventions: For individuals with low health literacy skills, what are effective interventions to (a) improve use of health care services, (b) improve health outcomes, (c) affect the costs of care, and (d) improve health care service use and/or health outcomes among different racial, ethnic, cultural, or age groups?

These issues parallel the questions addressed in the initial review, published in 2004.¹^,⁵⁰^,⁵¹

The amount of research being published in the field has expanded substantially. The initial review was limited to the relationship between literacy and health outcomes (or interventions); it included a total of 73 articles, 44 addressing outcomes, and 29 addressing interventions. The updated review expanded the scope of studies; it included 103 new good- or fair-quality studies reported in a total of 132 unduplicated articles. Of these, 86 articles addressed the relationship between health literacy and outcomes and 16 examined the relationship between numeracy and outcomes. In addition, 45 articles reported on interventions for individuals with low health literacy, split between those testing a single intervention strategy and those testing a mix (combination) of intervention strategies.

In this chapter, we recap the principal findings for KQ 1 and KQ 2 and comment on the applicability of the available bodies of evidence. We then discuss the limitations of both the literature reviewed and our own update. Finally, we present recommendations for future research.

Principal Findings

KQ 1. Health Literacy and Outcomes

Literacy Studies

For examining the association between health literacy and health outcomes (KQ 1), we included 86 fair- or good-quality articles (72 studies) in this update. Of these, 24 articles addressed the effect of health literacy on health care service use, 72 on health outcomes, 9 on disparities, and 2 on costs. Overall, the majority of studies were assessed as being of fair quality.

Differences in health literacy level were associated with use of health care services. Specifically, lower literacy was associated with increased emergency department and hospital use, and breast cancer (mammography), and lower influenza immunization, based on moderate strength of evidence. Evidence for other health care service use was low or insufficient because of inconsistent or limited findings and outcomes.

The relationship between health literacy and health outcomes was variable. The risk of mortality for seniors was clearly higher with lower health literacy. There was also moderate evidence to support a relationship between lower health literacy and poorer ability to take medications appropriately or interpret labels and health messages and poorer overall health status among seniors. In these studies, the evidence consists of all observational studies generally having a medium risk of bias and results generally in a consistent direction. The evidence for all other outcomes was either low or insufficient because the literature consisted of a small number of studies, poorly designed studies, and/or inconsistent results. These evaluations focused on the relationship between the lowest and highest health literacy groups. The evidence was sparse for evaluating differences between those with marginal (a middle category) health literacy and adequate (the highest category) health literacy.

The evidence concerning differences by health literacy level in costs of health care (KQ 1c) was low. The two relevant studies examined different payment sources (Medicaid and Medicare), found inconsistent results, and included different patient populations. No studies examined differences in costs among those with private health insurance coverage or no coverage.

Health literacy was found to mediate the relationship between race and health for a variety of outcomes. Outcomes studied included a condition that keeps respondents from working or having a long-term illness; misinterpretation of medication labels; prostate-specific antigen levels among newly diagnosed prostate cancer patients; nonadherence to HIV medications; children having health insurance; and, among seniors, self-reported health status, physical and mental health-related quality of life, and receipt of an influenza vaccine. We cannot know whether health literacy level would also be a mediator of the relationship between race and other health outcomes that have not been tested. Only one study examined whether health literacy level mediated the relationship between Hispanic ethnicity and health outcomes and no relationship was found. In contrast, one study found that health literacy level mediated the relationship between gender and misinterpretation of medication labels. We found no studies that evaluated the relationship between age, cultural group, or other sociodemographic characteristics and health outcomes.

Numeracy Studies

In this update we reviewed 16 fair-quality studies that examined the relationship between numeracy and various outcomes, including use of health care services, health outcomes, costs, and disparities. Most studies examining the relationship of numeracy to health outcomes were cross-sectional in design. Four studies were randomized controlled trials that analyzed their data in a cross-sectional manner for this analysis; one used a prospective cohort design.

In general, the strength of evidence for the relationship between numeracy and outcomes was insufficient or low given the small number of studies, which often had a high risk of bias or collectively gave mixed results. Only one study addressed the relationship between numeracy and use of health care services; this study reported no effect of numeracy on up-to-date screening for breast and colon cancer, but appears to be limited by inadequate power. Similarly, several studies demonstrated that the relationships between numeracy level and accuracy of risk perception (five studies), knowledge (four studies), skill in taking medication (six studies), and disease prevalence and severity (three studies) are mixed. The evidence for the relationship between numeracy and other health outcomes (e.g., self-efficacy, behavior) was insufficient to draw conclusions. No studies addressed the costs associated with differences in numeracy level. However, two studies examined whether numeracy level mediates health disparities and found that numeracy appeared to mediate the relationship between race and hemoglobin A1c and between gender and HIV medication management capacity.

Health Literacy and Numeracy Studies

Seven studies addressed the effects of both health literacy and numeracy on various outcomes. ⁹^,¹⁰^,⁴⁷^,⁹⁸^,¹²⁵^,¹²⁶^,¹⁷¹ Of these seven studies, six performed adjusted analyses on the same outcomes, thereby allowing assessment of whether these exposures affect health outcomes differently.⁹^,⁴⁷^,⁹⁸^,¹²⁵^,¹²⁶^,¹⁷¹ All of these studies must be interpreted with caution, however, because the proportion of individuals with low health literacy was small, raising the possibility of ceiling effects, which could obscure effects in the health literacy analyses. One study showed that ability to read nutrition labels was lower in both those with low health literacy skills (less than ninth grade) measured by the Rapid Estimate of Adult Literacy in Medicine (REALM) and low numeracy skills (less than ninth grade) measured by the Wide Range Achievement Test for mathematics (WRAT-math)⁹. However, it noted that the outcome was more highly correlated with numeracy (ρ 0.67) than health literacy (ρ 0.52). Similarly, another study showed that both health literacy skills (percent correct on the Short Test of Functional Health Literacy in Adults [S-TOFHLA]) and numeracy (percent correct on the Applied Problems Subtest of the Woodcock-Johnson Test) were related to HIV medication management capacity,⁴⁷ although the beta-coefficient was higher for numeracy in a regression model including both literacy and numeracy skill. A third study¹²⁶ showed that both health literacy skills (measured by the REALM) and numeracy (measured by a 6-item hybrid test including 3-items from Schwarz and Woloshin and 3 additional items from investigators) were related to the proportion of INR tests within range, although the correlation was higher for numeracy (r 0.12) than for health literacy (r 0.02). In contrast, two other studies found relationships between numeracy and health outcomes, but not between literacy and health outcomes. One of these studies found a relationship between numeracy (measured by the WRAT-math) and body mass index (BMI), but no relationship between literacy (measured by the REALM) and BMI.¹⁰ The other found a relationship between diabetes-specific numeracy (measured by the Diabetes Numeracy Test) and HgbA1c, but no relationship between literacy and HgbA1c.¹⁷¹ Only a single study¹²⁵ suggested a stronger relationship between literacy and health outcomes than numeracy and health outcomes. This study showed a greater likelihood of parent's using nonstandard dosing instruments to dose children's medicines related to their TOFHLA reading comprehension score (split at the median; adjusted OR, 2.4; 95% CI, 1.3-4.7) compared with their TOFHLA numeracy score (split at the median; OR, 1.4; 95% CI, 0.8 to 2.7).

KQ 2. Interventions To Improve Health Literacy

In this update we identified 42 new fair- or good-quality studies addressing the effect of interventions designed to mitigate the effects of low health literacy. Twenty-one used one specific strategy to mitigate the effects of low health literacy, and21 used a mixture of strategies combined into one intervention.

Interventions With Single Design Features

In general, the strength of evidence regarding the effect of specific design features of interventions for low-health-literacy populations is low or insufficient. This is attributable, in large part, to differences in the interventions (and subsequently results) for studies broadly grouped in the following design feature categories: alternative document design, alternative numerical presentation, additive and alternative pictorial representation, and improved readability and alternative document design.

Looking closely within categories, however, we noted that several specific design features resulted in improvements in comprehension for low-health-literacy populations in one or a few studies. These features, which bear further study in broader populations, include: presenting essential information by itself (i.e., information on hospital death rates without other distracting information, such as information on consumer satisfaction);¹⁸⁸ presenting essential information first (i.e., information on hospital death rates before information about consumer satisfaction);¹⁸⁸ presenting quality information with the higher number (rather than the lower number) indicating better quality;¹⁸⁸ using the same denominators to present the baseline risk of disease and treatment benefit;²¹⁹ adding icon arrays to numerical presentations of treatment benefit;²¹⁶^,²¹⁹ and adding video to verbal narratives.¹⁸⁴ Additionally, reexamining data from our 2004 review within these categories further suggests potential benefit from using reduced reading level and/or illustrated narratives.²³²^,²³⁶ In contrast, one study raised questions about whether certain design features, such as colored traffic symbols to denote death rates in hospitals of varying quality or symbols accompanying nonessential quality information, may actually worsen health choices among those with low health literacy.¹⁸⁸

Interventions With a Combination of Features

The strength of evidence for studies combining multiple strategies to mitigate the effects of low health literacy on outcomes was more variable that it was for single-feature interventions. We found consistent moderate strength of evidence that studied interventions change health care service use. Specifically, intensive self-management and adherence interventions appear to be effective in reducing emergency department visits and hospitalizations. Additionally, educational interventions and/or cues for screening increased colorectal cancer and prostate cancer screening. We note, however, that the health benefits of additional prostate cancer screening are questionable²⁵¹^,²⁵² and that increased screening rates could be a marker for poor decisionmaking.

We additionally found consistent evidence of moderate strength that some interventions change health outcomes. For instance, intensive disease-management programs appear to be effective at reducing disease prevalence. Furthermore, self-management interventions increased self-management behavior; however, in the only study that stratified its analysis by health literacy level, improvements were sometimes greater for those who had adequate health literacy and at other times greater for those with inadequate health literacy in adjusted analyses. The effects of other interventions on other health outcomes, including knowledge, self-efficacy, adherence, health-related skills, quality of life, and cost were mixed; thus, the strength of evidence was insufficient.

Components of effective interventions were their high intensity, theory basis, pilottesting before full implementation, emphasis on skill building, and delivery of the intervention by a health professional. Interventions that changed distal outcomes appeared to work by intermediately increasing knowledge or self-efficacy or by changing behavior.

Too few studies addressed the effects of literacy interventions on the outcomes of behavioral intent, or disparities to draw any meaningful conclusions; the strength of evidence is insufficient.

What This Update Adds to the Literature Included in the 2004 Review

Our results expand findings from our 2004 review in several ways. The size of the literature in the 2010 update review, examining the relationship between health literature and health outcomes (KQ 1) is larger than was available for the earlier review and encompasses a larger variety of outcomes (Table 62). In the 2004 review, we found that lower health literacy level was related to poorer knowledge of matters related to health outcomes and use of health services. Therefore, we did not reexamine this relationship during the update. In the earlier review, we recommended that future research examining the relationship between health literacy and health outcomes consistently control for potential confounding variables to more accurately measure the strength of the relationship between health literacy and the outcome. Unlike the earlier review, in the update, primary study outcomes are generally evaluated using multivariate analysis and control for potential confounding variables, providing a better and less biased estimate of the direction and magnitude of effect for our findings. Based on these more rigorous studies, we identified a relationship between health literacy level and additional health related outcomes. In 2004, we also recommended that studies more closely examine the factors that mediate the relationship between health literacy and health outcomes. In 2004, we had found only one study that directly examined racial disparities.¹⁵⁸ For the update, we found a limited body of research that begins to provide evidence of variables that may be on the pathway of effect between health literacy and health outcomes; these include factors such as knowledge, self-efficacy, and beliefs such as stigma related to their disease. New studies suggest that health literacy could be a mediator of racial disparities in health outcomes.

In 2004, we also recommended that studies stratify outcomes by numeracy level to gain a greater understanding of how these skills may uniquely affect health outcomes and under what conditions numeracy would be a useful indicator for targeting individuals for interventions. For the update, we found a small body of evidence concerning the relationship between numeracy level and health outcomes (Table 63). This is not only useful in and of itself, but it also is the next step in expanding our understanding of the skills that are needed to be health literate.

For KQ 2, our findings also expand findings from the 2004 review in several ways. In the 2004 review, we recommended that additional and more varied studies of interventions be pursued and that all studies measure the interventions' effects in a broader range of outcomes and by literacy subgroup. Studies in the current report have largely addressed these recommendations (see Table 64 and Table 65).

First, they address more varied interventions and provide insights into the utility of particular intervention design features. In our 2004 report, there were relatively few interventions of any type. Thus, we focused on how interventions affected outcomes rather than attempting to parse interventions into specific elements. In the current report, we reviewed studies by the specific intervention design features studied (see Table 64); only when that was not possible (i.e., because interventions used multiple design features) did we review studies by the outcomes involved (see Table 65). Using this new organizational structure, we identified several intervention design features that bear further study, including some identified through our 2004 review; these include presenting essential information by itself (i.e., information on hospital death rates without other distracting information, such as information on consumer satisfaction);¹⁸⁸ presenting essential information first (i.e., information on hospital death rates before information about consumer satisfaction);¹⁸⁸ presenting quality information with the higher number (rather than the lower number) indicating better quality;¹⁸⁸ adding icon arrays to numerical presentations of treatment benefit;²¹⁶^,²¹⁹ adding video to verbal narratives;¹⁸⁴ and using reduced reading level and/or illustrated narratives.²³²^,²³⁶ We also were able to illuminate what factors may be key in making the mixed interventions effective. Common features across nearly all of the mixed interventions that improved distal outcomes (e.g., self-management, hospitalizations, mortality) were their high intensity, theory basis, pilottesting before full implementation, emphasis on skill building, and delivery of the intervention by a health professional (e.g., pharmacist, diabetes educator; see intervention studies evidence tables in Appendix D).¹⁸²^,¹⁸³^,²⁰²^,²⁰⁷

Second, studies in the current report provide insight into the impact of interventions on a broader spectrum of outcomes. In our 2004 review, the majority of studies focused only on the outcome of knowledge (see Table 64 and Table 65). In the current review, studies focused on a broader range of outcomes, including disease self-efficacy, behavior, adherence, disease prevalence and severity, quality of life, preventive services use, emergency department visits, hospitalizations, and costs. Additionally, six studies in our update examined the impact of interventions on three or more outcomes⁷⁹^,¹⁸²^,¹⁸⁷^,¹⁹⁴^,¹⁹⁷^,²⁰² (see intervention studies evidence tables in Appendix D); they preliminarily suggest that effective interventions to mitigate the effects of low health literacy may work by increasing knowledge,¹⁹⁷^,²⁰² increasing self-efficacy,¹⁸⁷ or changing behavior.¹⁸²^,¹⁸⁷^,¹⁹⁷^,²⁰²

Third, a little over half the studies examined the effect of interventions by health literacy subgroup. This allows investigators to determine whether the intervention is more or less effective among those with low health literacy and whether interventions might ameliorate health disparities.

Limitations

Limitations of the Literature

Readers should interpret the findings from our systematic review in the context of several limitations. As with all systematic reviews, our results and conclusions depend on the quality of the published literature. A limitation across KQ s was heterogeneity in outcomes, populations, and study designs; this level of diversity in the knowledge base precluded us from pooling results statistically.

Specific limitations of the literature for studies addressing KQ 1 (i.e., the effects of health literacy and/or numeracy on health outcomes) included the following:

Lack of specification of thresholds for distinguishing levels of health literacy that consider the relevance of those levels to (1) the outcomes and population being studied and (2) the body of similar work in the field.²⁵³
Lack of an analytic framework or logic model for determining the appropriate set of potential confounding variables that need to be included in multivariate models. While studies generally controlled for some sociodemographic variables and other factors, the choice of variables varies across studies.
The potential for over controlling. Many studies included education (which is highly correlated with health literacy) as part of their multivariate model. Additionally, some studies included mediators of the effect of health literacy in their model; this may result in underestimating the aggregate effect of health literacy.

Small sample sizes, making it impossible to determine whether null findings represented a true lack of effect or simply reflected limitations in statistical power.

Studies conducted in just one clinic or in other narrowly defined patient populations, rendering the applicability of findings to other settings or populations unknown. Only two studies were conducted within nationally representative samples: the National Assessment of Adult Literacy conducted in 2003 and the earlier National Adult Literacy Survey in 1992.

Health literacy tools that continue to focus primarily on reading ability despite the Institute of Medicine's call for skills-based health literacy tools⁵³ (i.e., tools focused on a combination of oral or verbal, navigational, computer, or other skills necessary for individuals to manage their health).At the time of this update review, we identified none in the literature. Thus, we could not determine the relationship between a wider array of skills or abilities and health outcomes. We did, however, find evidence that development of tools that can measure these additional skills has begun.²⁵⁴

A limited number of studies examining the role of health literacy on health disparities. Most research focused on whether health literacy mediated the relationship between race and health outcomes.

The limitations of the literature for studies addressing KQ 2 (i.e., the effects of interventions to mitigate low health literacy) included the following:

Lack of an adequate control or comparator group in many studies, limiting the ability to determine the true effect(s) of the intervention.
Measurement of multiple outcomes with insufficient attention to ensure that each is adequately powered to detect a difference.
Testing interventions that combined various design features to mitigate the effect of low health literacy but offering no way to determine the effectiveness of individual components.
Failure to perform adequately controlled subgroup analyses that would elucidate differential effects of interventions in low- and high-health-literacy populations. This is important to the extent that the field's overall goal is to reduce disparities related to the impact of low health literacy rather than simply to improve outcomes for individuals at all health literacy levels.
Failure to report adequately the design features that would allow future content analyses of effective interventions.

Limitations of Our Review

In addition to clarifying the limitations of the overall body of literature, we must also acknowledge the limitations of our systematic review and update of the 2004 report. First, we included only those studies in which investigators quantitatively measured the literacy of their populations. We may have missed some important studies addressing the relationship of health literacy on health outcomes or important interventions that either did not measure health literacy or measured it only by self-report. Second, we excluded studies that included only outcomes focused on communication or decisionmaking.^255-260 Our reasoning was that, in our judgment, patient-physician communication likely moderated rather than mediated the effect of intent for behavior on health outcomes. However, this may have meant we missed outcomes or interventions important to some researchers, clinicians, and policymakers. Third, we did not conduct dual independent abstraction of all information for review. Rather, a single reviewer abstracted information and a second reviewer checked it; we feel this process was sufficiently rigorous to allow accurate conclusions, and it is the basic strategy the RTI-UNC EPC has used for this step for more than a decade. We did, however, perform dual review for article inclusion and dual rating of the risk of bias of individual studies and the strength of evidence in relation to outcomes, highlighting an overall rigorous process. Fourth, we did not formally integrate the analyses from our 2004 and current reviews, although based on our review of summary materials, we suspect this would have a minimum impact on our overall conclusions.

Opportunities for Future Research

This update shows that the field of health literacy has advanced since our 2004 review. However, many opportunities remain for important future research. The need for such investigations is considerable for gaining a better understanding of the outcomes of health care, given levels of health literacy, and for expanding the knowledge base about the impact of interventions intended to improve health literacy.

Future Research Into the Relationship Between Health Literacy and Health Outcomes

Instrument Cutpoints

The field will greatly benefit from researchers prespecifying the most relevant cutpoints for distinguishing levels of health literacy within the population being studied, considering how the cutpoints selected compare to those that have been used in measuring similar populations and outcomes. Currently, investigators use cutpoints inconsistently, such that “adequate” and “inadequate” or “low” health literacy levels have different definitions across studies. This problem makes comparing results from these studies difficult. Additionally, the literature as a whole does not lend itself to explaining at what particular level lower health literacy is related to significantly poorer outcomes of health care.

Furthermore, sometimes a middle group, often referred to as having “marginal health literacy,” is identified; other times, no such group is specified. Sometimes research teams combine the middle health literacy group with the higher health literacy group; sometimes they combine it with the lower health literacy group.

In short, those conducting work in this area in the future should more rigorously defend their choice of inadequate, marginal, and adequate levels of health literacy.

Skills-Based Measures

Testing skills-based health literacy measures will be an important focus of future research. Our current review expanded the tools that measure health literacy to include those that focus on numeracy. However, we found no tools that measure oral health literacy. New instruments are likely to be available in the near future that can be used as alternative measures of health literacy that capture additional and potentially critical skills. For example, a 2009 Institute of Medicine workshop and resulting report, Measures of Health Literacy, highlight several skills-based measurement tools that are under development—one designed for use in clinics and a second for population-based surveillance.261 Future research should consider these and other measures that may explain the interplay of a wider range of health literacy skills and outcomes.

Future research should also consider capturing changing competencies over time based on greater knowledge or experience (or both), resulting in health literacy levels changing over time. For this type of measurement, prospective research designs will be critical, allowing researchers to measure health literacy at different times while in treatment or after different amounts of experience managing a chronic condition.

Links Between Low Health Literacy and Outcomes

Additional work is needed to help us understand the pathways between low health literacy and health outcomes. A few studies examined variables that may be in the analytic pathway between health literacy and health outcomes and mediate the relationship between the two— including knowledge, self-efficacy, and beliefs. More research is needed investigating these potential mediators in relation to a wider range of outcomes and populations. Other potential variables that warrant serious attention as mediators or moderators of the relationship include measures of education, social support, cultural competency, decisionmaking skills, and trust in the information source.

Population Subgroups

Additional research is needed to understand whether health literacy has a differential effect in various subgroups of the population. For example, we lack data evaluating whether the effect of low health literacy would be significantly different in different groups defined by various sociodemographic factors. Of particular interest are the following comparisons: white populations vs. various racial and/or ethnic minority populations, nonelderly vs. elderly individuals, and male vs. female patients.

Methodologic Limitations

Current work should continue to address the basic methodological deficiencies we found during this update and the problems we noted in the previous review. For instance, researchers need to determine a minimal set of confounding variables to be considered for all multivariate analyses; sample sizes need to be larger so that investigators truly have sufficient power to detect differences among the three health literacy levels.

Applicability of Research

The degree to which results from the studies done to date can be applied broadly is limited. Considering the “PICOTS” framework (patients/populations, interventions, comparators, outcomes, timeframes, and settings) for considering the generalizability of a body of research, we conclude that the ability of decisionmakers to generalize results from the current body of work is not great. Most current studies were limited to one clinic or one geographic area; thus, we lack evidence that the results would apply in more broadly defined populations or settings. The field needs to examine the relationships between health literacy and health outcomes in more diverse and representative populations.

Future Research Into Interventions to Mitigate the Effects of Low Health Literacy

Opportunities to study interventions to mitigate the effects of low health literacy are also substantial.

Effective Design of Health-Related Documents

Additional work is needed on the design features of documents. As discussed above, we identified several design features of health-related interventions that could mitigate the effects of low health literacy. However, the majority have been examined in only one or a few studies in clinical populations; thus, they warrant further investigation.

An important question to answer is, “What needs study and what does not?” Our review failed to turn up evidence regarding several document design features widely recommended by experts in the field of health literacy; these include grouping or “chunking” of ideas and teach-back.²⁶² However, whether these features require specific investigation in relation to health literacy when they have been well studied in other fields is not clear. For instance, the field of psycholinguistics has done extensive testing of simplified sentence and document structure and the cohesiveness of concepts in the text; this body of work, albeit not necessarily stemming from the health sector, may obviate the need for specific testing of these approaches in the health literacy field per se.²⁶³ Furthermore, the educational literature has tested techniques of explicit instruction that are recommended for poor readers—i.e., instruction that has a clear task and is broken into small steps with practice and feedback at every step—and determined that they are effective.²⁶³ Rather than spending time and energy on additional testing, exploring the extent to which other fields can inform the work of health literacy may be more appropriate.

Some design features, however, may warrant explicit testing. Given the evidence from multiple areas of study that motivation increases the effects of comprehension and behavior,⁹⁸^,²⁶³^,²⁶⁴ more study of the impact of illustrations, videos, fotonovelas, and other novel approaches that may increase motivation for information-processing through their visual appeal seems warranted. Researchers in health literacy should seek guidance from the health communication literature to guide these efforts.²⁶⁵

Further testing of techniques based on oral and numerical delivery of information will also be useful. Oral information receives different cognitive processing than written information and has a naturally simpler syntax that may help low-literacy individuals.²⁶³ Numbers and graphical numerical information have many alternative forms of presentation. These have been shown to affect understanding in high-literacy individuals; they should be tested for comprehension among those with lower literacy.^266-271

Finally, investigation of “work-around” interventions should be undertaken. These can include use of patient advocates, who could accompany individuals to medical appointments and facilitate subsequent care.

Effective Components of Combination Interventions

Additional work is also needed to determine the effective components of already-tested interventions that have employed a combination of features to mitigate the effects of low health literacy. While a combination of intervention features has repeatedly been shown to ensure the success of interventions, paring away ineffective features could save delivery time and result in more cost-effective delivery. Several possibilities for accomplishing this task exist. For instance, one approach is to conduct a qualitative content analysis of existing interventions. Another approach is to conduct additional trials to test components of effective interventions. A final approach is to conduct a meta-regression; in such analyses, investigators enter data about the features of existing interventions into a statistical program to determine their relative impact on relevant outcomes. While the field may be too young for this now, meta-regression could be a very useful technique as additional studies with similar intervention features and outcomes become available. To prepare for such a meta-regression, investigators in the field might agree on a useful set of intervention design features to be tested and consistently report on the incorporation of these features into multicomponent interventions.

Effective Practice and Policy Interventions

Additional work is also needed to determine the effect of practice and policy interventions. We found almost no studies that addressed such interventions.

Implications of This Report for Clinicians and Policymakers

In addition to identifying areas for future research, this report informs clinicians and policymakers. First, it continues to raise awareness that low health literacy has a substantial impact on healthcare service use, health outcomes, cost, and disparities and warrants the attention of both clinicians and policymakers. Second, it highlights effective interventions that could be implemented in clinical practice now and/or supported by policy. These interventions have been rated as having moderate strength of evidence in our review and include intensive adherence, self-management, and disease management interventions delivered by clinical practitioners. Finally, for policymakers, our update highlights the critical need for research funding to test practice and policy interventions, which to date have gone largely untested. The recent Department of Health and Human Services National Action Plan to Improve Health Literacy helps enumerate these and other critical actions for clinicians and policymakers addressing health literacy.⁵²

Publication Details

Copyright

Copyright Notice

Publisher

Agency for Healthcare Research and Quality (US), Rockville (MD)

NLM Citation

Berkman ND, Sheridan SL, Donahue KE, et al. Health Literacy Interventions and Outcomes: An Updated Systematic Review. Rockville (MD): Agency for Healthcare Research and Quality (US); 2011 Mar. (Evidence Reports/Technology Assessments, No. 199.) Discussion.