GRADE profiles

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Addendum to NICE guideline CG61, Irritable bowel syndrome in adults: Diagnosis and management of irritable bowel syndrome in primary care. London: National Institute for Health and Care Excellence (NICE); 2015 Feb. (NICE guideline, No. CG61.1.)

Cover of Addendum to NICE guideline CG61, Irritable bowel syndrome in adults

Addendum to NICE guideline CG61, Irritable bowel syndrome in adults: Diagnosis and management of irritable bowel syndrome in primary care.

Show details

Contents

< Prev Next >

Appendix HGRADE profiles

H.1. Review question 1 (antidepressants vs placebo)

Table 68GRADE profile, successfully treated, abdominal pain

Number of studies	Design	Study length	Risk of bias	Inconsistency	Indirectness	Imprecision	Number of participants	Effect	Quality
	Antidepressants - TCAs vs placebo
2	RCTs	6–12 weeks	Very serious^a	Serious^c	Serious^e	Very serious^f	N=52 intervention N=52 placebo	RR 1.82 (95%CI 0.63 to 5.25)	Very low
	Antidepressants - SSRI vs placebo
4	RCTs	6–12 weeks	Very serious^b	Serious^d	Serious^e	Very serious^g	N= 96 intervention N= 101 placebo	RR2.29 (95%CI 0.79 to 6.68)	Very low

: Studies included; TCAs, Vahedi (2008), Vij (1991); SSRIs, Kuiken (2003), Tabas (2004), Tack (2006), Vahedi (2005)
(a): Unclear if questionnaire/other tools validated and no additional follow-up (Kuiekn 2003, Tabas 2004, Tack 2006,)
(b): Unclear randomisation (Tack 2006) and unclear if questionnaire/other tools validated and no additional follow-up (Vahedi 2005 and 2008, Vij 1991)
(c): I²70%
(d): I²79%
(e): Study length 6–12 weeks
(f): 95% confidence interval crosses the minimal important difference at 0.75 and 1.25, leading to very serious uncertainty. Downgraded 2 levels.
(g): 95% confidence interval crosses the minimal important difference at 1.25 and crosses line of no effect, leading to very serious uncertainty. Downgraded 2 levels

Table 69GRADE profile, scores on abdominal pain

Number of studies	Design	Study length	Risk of bias	Indirectness	Imprecision	Number of participants	Effect	Quality
Antidepressants - TCAs vs placebo
Rajagopalan 1998	RCT	8–12 weeks	Very serious^a	Serious^c	No serious	N=11 intervention, N=11 placebo	Standardised mean difference 1.49 (95%CI 0.52 to 2.45)	Very low
Antidepressant - SSRIs vs placebo
Tack 2006	RCT	8–12 weeks	Very serious^b	Serious^c	No serious	N= 11 intervention N= 12 placebo	Standardised mean difference 4.60 (95%CI 2.93 to 6.28)	Very low

: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.
(a): Unclear randomisation, unclear concealment (Rajagopalan 1998), unclear if questionnaire/other tools validated and no additional follow-up (Rajagopalan 1998) not low dose TCA, small number of study participants,
(b): Unclear if questionnaire/other tools validated and no additional follow-up (Tack 2006)
(c): Study length 8–12 weeks

Table 70GRADE profile, successfully treated, global assessment

Number of studies	Design	Study length	Risk of bias	Inconsistency	Indirectness	Imprecision	Number of participants	Effect	Quality
	Antidepressants - TCAs vs placebo
5	RCTs	4–12 weeks	Very serious^a	No serious	Serious^c	Serious^d	N=159 intervention, N=139 placebo	RR 1.43 (95%CI 1.15 to 1.79)	Very low
	Antidepressants - SSRIs vs placebo
5	RCTs	4–12 weeks	Very serious^a	Serious^b	Serious^c	Very serious^e	N=143 intervention, N=138 placebo	RR 1.51 (95%CI 0.87 to 2.61)	Very low

: Studies included; TCAs, Abdul-Baki (2009), Myren (1982), Talley (2008), Vahedi (2008), Vij (1991); SSRIs, Kuiken (2003), Ladabaum (2010), Masand (2009), Tabas (2004), Talley (2008)
(a): Unclear randomisation (Myren 1982, Masand 2009), small number of individual study participants (all included studies)
(b): I² 73%
(c): Study length 4–12 weeks
(d): 95% confidence interval crosses the minimal important difference at 1.25, leading to serious uncertainty. Downgraded 1 level.
(e): 95% confidence interval crosses the minimal important difference at 1.25 and crosses line of no effect, leading to very serious uncertainty. Downgraded 2 levels

Table 71GRADE profile, successfully treated, symptom score

Number of studies	Design	Study length	Risk of bias	Indirectness	Imprecision	Number of participants	Effect	Quality
Antidepressants - TCA vs placebo
Vahedi 2008	RCTs	8–12 weeks	Very serious^a	Serious^c	Very serious^d	N=36 intervention, N=36 placebo	RR 1.36 (95%CI 0.81 to 2.27)	Very low
Antidepressants - SSRI vs placebo
Masand 2009	RCTs	8–12 weeks	Very serious^b	Serious^c	Serious^e	N= 27 intervention N= 27 placebo	RR 2.43 (95%CI 1.21 to 4.89)	Very low

(a): Unclear if questionnaire/other tools validated (Vahedi 2008, Masand 2009) small number of individual study participants (all included studies)
(b): Unclear randomisation, unclear allocation concealment (Masand 2009), unclear if questionnaire/other tools validated (Vahedi 2008, Masand 2009), no additional follow-up and small number of study participants (Masand 2009)
(c): Study length 8–12 weeks
(d): 95% confidence interval crosses the minimal important difference at 1.25 and crosses line of no effect, leading to very serious uncertainty. Downgraded 2 levels
(e): 95% confidence interval crosses the minimal important difference at 1.25, leading to serious uncertainty. Downgraded 1 level
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 72GRADE profile, symptom scores

Number of studies	Design	Study length	Risk of bias	Indirectness	Imprecision	Number of participants	Effect	Quality
Antidepressants - TCA vs placebo
Vahedi 2008	RCTs	8–12 weeks	Very serious^a	Serious^c	Very serious^d	N=36 intervention, N=36 placebo	Standardised mean difference 0.05 (−0.41 to 2.27)	Very low
Antidepressants - SSRI vs placebo
Masand 2009	RCTs	8–12 weeks	Very serious^b	Serious^c	Serious^e	N=25 intervention, N=25 placebo	Standardised mean difference 0.75 (0.17 to 1.32)

(a): Unclear if questionnaire/other tools validated (Vahedi 2008) small number of individual study participants (all included studies)
(b): Unclear randomisation, unclear allocation concealment (Masand 2009), unclear if questionnaire/other tools validated (Masand 2009), no additional follow-up (Masand 2009) small number of individual study participants (all included studies)
(c): Study length8–12 weeks
(d): The 95% confidence interval crosses the MID of 0.5 and −0.5and crosses the line of no difference, leading to very serious imprecision in the effect size. Downgraded 2 levels.
(e): The 95% confidence interval crosses the MID of 0.5 (indicating moderate effect), leading to serious imprecision. Downgraded 1 level
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 73GRADE profiles, quality of life 1 (SF-36)

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

Number of participants

Effect

Quality

SF-36- TCAs vs placebo

Abdul-Baki 2009

RCT

12 weeks

Serious^a

None

Very serious^b

N=31 intervention,

N=25 placebo

Mean percent difference from baseline; imipramine 11.8%±13.2%, placebo 4.3%±9.0%, p=0.02

Very low

: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.
(a): High drop-out rates, unclear if questionnaire/other tools validated
(b): Small sample size, quality of life outcomes per protocol analysis

Table 74GRADE profiles, quality of life 2 (SF-36)

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

Number of participants

Effect

Quality

SF-36TCAs and SSRIs vs placebo

Talley 2008

RCT

12 weeks

Serious^a

None

Serious^b

N17 citalopram,

N=18 imipramine,

N=16 placebo

Score change; physical component; citalopram 3.5(6.1), imipramine 7.3(7.3), placebo 6.5(4.6), p=0.40 Mental component; citalopram 0(4.1), imipramine 4.8(4.5), placebo 1.9(7.2), p=0.07

Low

: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.
(a): No additional follow-up
(b): Small sample size

Table 75GRADE profiles, quality of life 3 (IBS-QOL scores)

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

Number of participants

Effect

Quality

IBS QOL scores SSRIs vs placebo,

Ladabaum 2010

RCT

8 weeks

Serious^a

None

Serious^b

N=20 intervention,

N=25 placebo

Mean score (SD) citalopram 74 (18), placebo 74 (24), p=0.85

Low

: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.
(a): Differences between groups in drop-out rates, unclear if questionnaire/other tools validated
(b): Small sample size

Table 76GRADE profiles, quality of life 4 (IBS-QOL scores)

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

Number of participants

Effect

Quality

IBS QOL SSRIs

Tabas 2004

RCT

12 weeks

Very serious^a

None

Serious^b

N=38 intervention,

N=43 placebo

% of improvement; Food avoidance; paroxetine 25.4, placebo 13.7, p=0.03 Work function score; paroxetine 25.4, placebo 12.0, p=0.08 Social function score; paroxetine 25.4, placebo 13.7, p=0.76

Very low

: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.
(a): Unclear if questionnaire/other tools validated, no additional follow-up, poor adverse effects reporting
(b): Small sample size

H.2. Review question 2 (low FODMAP diet vs Standard diet)

Table 77GRADE profile, Outcome: GI symptoms, overall response

Number of studies	Design	Study length	Risk of bias	Indirectness	Imprecision	No. of participants	Effect	Quality
FODMAP							(95%CI)
Halmos 2014	RCT, crossover	21 days (each arm)	Very serious^a	Serious^b	Very serious^c	N=30	VAS (from baseline); Low FODMAP 22.8mm (16.7 to 28.8), p<0.001 Typical diet 44.9mm (36.6 to 53.1), p<0.001	Very low
Staudacher 2012	RCT	4 weeks	Serious^d	Serious^e	Serious^f	N=41	Incidence (mean (95%CI) days/week); Low FODMAP 0.9(0.8 to 1.1), control 1.6 (1.3 to 1.9), p=0.001	Very low
Staudacher 2011	Controlled trial	Unclear	Very serious^g	Serious^h^,ⁱ	Serious^f	N=82	Improved; Low FODMAP 37/43(86%), standard diet 19/39(49%), p<0.001	Very low

(a): No allocation concealment, investigators not blinded to study group, no additional follow-up period following study completion (downgraded 2 levels)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Unclear if VAS used had been validated, minimum detectable difference crossed by 95%CI in GI symptoms (primary outcome), small participant numbers (downgraded 2 levels)
(d): No additional follow-up period (downgraded 1 level)
(e): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(f): Small participant numbers (downgraded 1 level)
(g): Non-randomised comparison, no blinding, differences in the dietary advice given in the comparison group, unclear follow-up period (downgraded 2 levels)
(h): Unclear length of study (downgraded 1 level)
(i): Convenience sample from dietetic clinic before and after implementation of low FODMAP service (downgraded 1 level)
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 78GRADE profile, Outcome: bloating

Number of studies	Design	Study length	Risk of bias	Indirectness	Imprecision	No. of participants	Effect	Quality
FODMAP							(95%CI)
Halmos 2014	RCT, crossover	21 days (each arm)	Very serious^a	Serious^b	Very serious^c	N=30	VAS (from baseline); Low FODMAP 45.1mm (35.1 to 55.0), p<0.001 Typical diet 24.2mm (17.1 to 31.2)	Very low
Staudacher 2012	RCT	4 weeks	Serious^d	Serious^e	Serious^f	N=41	Incidence (mean (95%CI) days/week); Low FODMAP 3.8(3.0 to 4.6), control 5.7 (4.9 to 6.4), p=0.002	Very low
Staudacher 2011	Controlled trial	Unclear	Very serious^g	Serious^h^,ⁱ	Serious^f	N=82	Improved; Low FODMAP 32/39(82%), standard diet 17/35(49%), p=0.002	Very low

(a): No allocation concealment, investigators not blinded to study group, no additional follow-up period following study completion (downgraded 2 levels)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Unclear if VAS used had been validated, secondary outcome (downgraded 2 levels)
(d): No additional follow-up period (downgraded 1 level)
(e): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(f): Small participant numbers (downgraded 1 level)
(g): Non-randomised comparison, no blinding, differences in the dietary advice given in the comparison group, unclear follow-up period (downgraded 2 levels)
(h): Unclear length of study (downgraded 1 level)
(i): Convenience sample from dietetic clinic before and after implementation of low FODMAP service (downgraded 1 level)
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 79GRADE profile, Outcome: abdominal pain

Number of studies	Design	Study length	Risk of bias	Indirectness	Imprecision	No. of participants	Effect	Quality
FODMAP							(95%CI)
Halmos 2014	RCT, crossover	21 days (each arm)	Very serious^a	Serious^b	Very serious^c	N=30	VAS (from baseline); Low FODMAP 43.8mm (35.0 to 52.5), p<0.001 Typical diet 22.5mm (16.3 to 28.6)	Very low
Staudacher 2012	RCT	4 weeks	Serious^d	Serious^e	Serious^f	N=41	Incidence (mean (95%CI) days/week); Low FODMAP 3.6(2.8 to 4.4), control 4.8 (4.1 to 5.5), p=0.02	Very low
Staudacher 2011	Controlled trial	Unclear	Very serious^g	Serious^h^,ⁱ	Serious^f	N=82	Improved; Low FODMAP 29/34(85%), standard diet 20/33(61%), p=0.023	Very low

(a): No allocation concealment, investigators not blinded to study group, no additional follow-up period following study completion (downgraded 2 levels)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Unclear if VAS used had been validated, secondary outcome (downgraded 2 levels)
(d): No additional follow-up period (downgraded 1 level)
(e): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(f): Small participant numbers (downgraded 1 level)
(g): Non-randomised comparison, no blinding, differences in the dietary advice given in the comparison group, unclear follow-up period (downgraded 2 levels)
(h): Unclear length of study (downgraded 1 level)
(i): Convenience sample from dietetic clinic before and after implementation of low FODMAP service (downgraded 1 level)
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 80GRADE profile, Outcome: dissatisfaction with stool consistency

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

No. of participants

Effect

Quality

FODMAP

(95%CI)

Halmos 2014

RCT, crossover

21 days (each arm)

Very serious^a

Serious^b

Very serious^c

N=30

VAS (from baseline);

Low FODMAP 47.8mm (37.6 to 57.9), p<0.001 Typical diet 25.9mm (18.9 to 32.9)

Very low

(a): No allocation concealment, investigators not blinded to study group, no additional follow-up period following study completion (downgraded 2 levels)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Unclear if VAS used had been validated, secondary outcome (downgraded 2 levels)

Table 81GRADE profile, Outcome: flatuence/wind

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

No. of participants

Effect

Quality

FODMAP

(95%CI)

Staudacher 2012

RCT

4 weeks

Serious^d

Serious^e

Serious^f

N=41

Incidence (mean (95%CI) days/week);

Low FODMAP 4.3(3.3 to 5.3), control 5.6 (4.6 to 6.5), p=0.07

Very low

Staudacher 2011

Controlled trial

Unclear

Very serious^g

Serious^h^,ⁱ

Serious^f

N=82

Improved;

Low FODMAP 33/38(87%), standard diet 14/28(50%), p=0.001

Very low

(a): No additional follow-up period (downgraded 1 level)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Small participant numbers (downgraded 1 level)
(d): Non-randomised comparison, no blinding, differences in the dietary advice given in the comparison group, unclear follow-up period (downgraded 2 levels)
(e): Unclear length of study (downgraded 1 level)
(f): Convenience sample from dietetic clinic before and after implementation of low FODMAP service (downgraded 1 level)
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 82GRADE profile, Outcome: diarrhoea

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

No. of participants

Effect

Quality

FODMAP

(95%CI)

Staudacher 2012

RCT

4 weeks

Serious^d

Serious^e

Serious^f

N=41

Incidence (mean (95%CI) days/week);

Low FODMAP 1.4(0.4 to 2.4), control 2.2 (1.3 to 3.1), p=0.24

Very low

Staudacher 2011

Controlled trial

Unclear

Very serious^g

Serious^h^,ⁱ

Serious^f

N=82

Improved;

Low FODMAP 30/36(87%), standard diet 18/29(62%), p=0.052

Very low

(a): No additional follow-up period (downgraded 1 level)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Small participant numbers (downgraded 1 level)
(d): Non-randomised comparison, no blinding, differences in the dietary advice given in the comparison group, unclear follow-up period (downgraded 2 levels)
(e): Unclear length of study (downgraded 1 level)
(f): Convenience sample from dietetic clinic before and after implementation of low FODMAP service (downgraded 1 level)
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

Table 83GRADE profile, Outcome: constipation

Number of studies

Design

Study length

Risk of bias

Indirectness

Imprecision

No. of participants

Effect

Quality

FODMAP

(95%CI)

Staudacher 2012

RCT

4 weeks

Serious^d

Serious^e

Serious^f

N=41

Incidence (mean (95%CI) days/week);

Low FODMAP 0.8(0.3 to 1.3), control 1.0 (0.5 to 1.5), p=0.56

Very low

Staudacher 2011

Controlled trial

Unclear

Very serious^g

Serious^h^,ⁱ

Serious^f

N=82

Improved;

Low FODMAP 10/21(67%), standard diet 10/22(45%), p=0.161

Very low

(a): No additional follow-up period (downgraded 1 level)
(b): FODMAP diet usually advised for 8 weeks (downgraded 1 level)
(c): Small participant numbers (downgraded 1 level)
(d): Non-randomised comparison, no blinding, differences in the dietary advice given in the comparison group, unclear follow-up period (downgraded 2 levels)
(e): Unclear length of study (downgraded 1 level)
(f): Convenience sample from dietetic clinic before and after implementation of low FODMAP service (downgraded 1 level)
: Note: Inconsistency could not be assessed as these are individual studies only without meta-analysis.

H.3. Review question 3 (linaclotide)

Table 84GRADE Profile, Outcome: Quality of Life (QOL)

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg (%)	C = Placebo (%)	Effect		Quality
IBS-QOL Responder (>14 point improvement)								Relative (95% CI)	Absolute (95% CI)
Johnston^a 2010	RCT	Serious^b	n/a	No serious	Very serious^c	31/84 (26)	31/85 (26)	1.01 [0.68, 1.50]	0 more per 100 (12 fewer, 18 more)	Very low
IBS-QOL Scale (34 items each rated on 5 point Likert scale, low score = worse QOL)								Mean difference (improvement) (95% CI)
Johnston^a 2010	RCT	Serious^b	n/a	No serious	Very serious^d	84	85	I=14, C=14.5 No p value reported		Very low
Quigley^e 2012	RCT	Very serious^f	n/a	No serious	No Serious	405	395	I=18.4, C=15.2 LS mean difference 3.3% (1.0, 5.5) p=0.004		Low
Quigley^g 2012	RCT	Serious^h	n/a	No serious	No Serious	401	403	I=16.6, C=11.1 LS mean difference 5.5% (3.4, 7.6) p<0.0001		Moderate

(a): Only the comparable dose arm (290µg) from this study is reported
(b): Per protocol analysis used for mean change endpoints but numbers per arm does not reflect drop-outs. Use of rescue medication (laxatives) was permitted but not reported by study arm. Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(c): Point estimate not reaching MID (GRADE default suggestion), 95% CIs incorporate both deterioration and improvement in QOL score
(d): No p value, SD / CIs reported.
(e): First of two studies reported in this further analysis of two RCTs (Rao et al 2012)
(f): Rescue medication (laxatives) and bulk laxatives and stool softeners were permitted throughout but not reported by study arm. Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(g): Second of two RCTs reported (Chey et al 2012)
(h): Rescue medication (laxatives) was permitted but not reported by study arm. Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.

Table 85GRADE Profile, Outcome: symptoms

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg (%)	C = Placebo (%)	Effect		Quality
2.1 FDA Pain responder (≥30% improvement in pain, for ≥half the study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a (12 weeks)	RCT	Serious^b	No Serious	No serious	Serious^c	399/806 (50)	287/798 (36)	1.38 [1.23, 1.54]	14 more per 100 (8 more, 19 more)	Low
Chey 2012 (26 weeks)	RCT	Serious^b	n/a	No serious	No Serious	197/401 (49)	126/403 (31)	1.57 [1.32, 1.87]	18 more per 100 (10 more, 27 more)	Moderate
2.2 FDA Stool frequency responder (increase of ≥1 CSBM per week, for ≥half of study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a (12 weeks)	RCT	Serious^b	Serious^d	No serious	No Serious	388/806 (48)	208/798 (26)	1.85 [1.45, 2.37]	22 more per 100 (12 more, 26 more)	Low
Chey 2012 (26 weeks)	RCT	Serious^b	n/a	No serious	No Serious	175/401 (44)	75/403 (19)	2.34 (1.85, 2.96)	25 more per 100 (16 more, 36 more)	Moderate
2.3 FDA Combined responder for Pain and Stool Frequency for ≥half of study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a (12 weeks)	RCT	Serious^b	Serious^e	No serious	No Serious	271/806 (34)	139/798 (17)	1.95 [1.30, 2.94]	17 more per 100 (5 more, 34 more)	Low
Chey 2012 (26 weeks)	RCT	Serious^b	n/a	No serious	No serious	130/401 (32)	53/403 (13)	2.47 (1.85, 3.29)	19 more per 100 (11 more, 30 more)	Moderate
2.4 FDA Pain responder (≥30% improvement in pain, for ≥two thirds of study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a (12 weeks)	RCT	Serious^b	Serious^f	No serious	Serious^g	295/806 (37)	186/798 (23)	1.58 [1.02, 2.46]	14 more per 100 (0 more, 100 more)	Very low
Chey 2012 (26 weeks)	RCT	Serious^b	n/a	No serious	No Serious	148/401 (37)	70/403 (17)	2.12 (1.66, 2.72)	19 more per 100 (11 more, 30 more)	Moderate
2.5 FDA Combined responder for Pain and Stool Frequency for ≥two thirds of study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a (12 weeks)	RCT	Serious^b	No Serious	No serious	No Serious	100/806 (12)	32/798 (4)	3.09 [2.10, 4.51]	8 more per 100 (4 more, 14 more)	Moderate
Chey 2012 (26 weeks)	RCT	Serious^b	n/a	No serious	No serious	10/401 (3)	48/403 (12)	4.82 (2.48, 9.40)	45 more per 100 (18 more, 100 more)	Moderate
2.6 EMA abdominal pain/discomfort responders (≥30% improvement with neither worsening from baseline, for ≥half of study weeks)								Relative (95% CI)	Absolute (95% CI)
2^h	RCT	Serious^b	No Serious	No serious	Seriousⁱ	439/806 (54)	320/798 (40)	1.36 [1.22, 1.51]	14 more per 100 (9 more, 20 more)	Low
2.7 EMA Global Relief responders								Relative (95% CI)	Absolute (95% CI)
3^j	RCT	Very serious^b^,^k	Serious^l	No serious	No serious	312/890 (35)	165/883 (19)	1.87 [1.33, 2.63]	16 more per 100 (6 more, 30 more)	Very low

(a): Chey 2012 (at 12 weeks only), Rao 2012
(b): Use of rescue medication (laxatives) was permitted but not reported by study arm. Bulk laxatives and stool softeners were also permitted by one study (Rao et al 2012). Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(c): Lower end of 95% CI below the threshold for MID (GRADE default suggestion)
(d): Random effects analysis identifies significant heterogeneity - I2 68% Chi² 3.17, P=0.08. Partial CI overlap
(e): Random effects analysis identifies significant heterogeneity - I2 80% Chi²5.02, P=0.03. CI overlap
(f): Random effects analysis identifies significant heterogeneity - I2 87%, Chi² 7.88, P=0.005. CIs do not overlap
(g): 95%CIs cross threshold for MID (GRADE default suggestion)
(h): Rao and Chey via Quigley 2012
(i): 95%CIs cross threshold for MID (GRADE default suggestion)
(j): Rao and Chey via Quigley 2012, Johnston 2010
(k): Downgraded due to risk of recall bias, degree of relief was measured weekly, rating symptoms retrospectively vs. symptoms prior to trial inauguration (n=2 studies)
(l): Random effects analysis identifies significant heterogeneity - I² 75%,Chi²=8.08, P=0.02. No overlap between 2/3 study CIs

Table 86GRADE Profile, Outcome: stool score/general changes in bowel habit

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg (%)	C = Placebo (%)	Effect		Quality
6.1 Consitpation Severity (% with decrease of ≥1 point on BSFS^* for ≥half study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a (12 weeks)	RCTs	Very serious^b^,^c	No Serious	No serious	No serious	485/806 (60)	327/798 (41)	1.47 [1.33, 1.62]	19 more per 100 (14 more, 25 more)	Low
Chey 2012 (26 weeks)	RCT	Very serious^b^,^c	N/A	No serious	No serious	221/401 (55)	139/403 (35)	1.60 (1.36, 1.88)	21 more per 100 (12 more, 30 more)	Low
6.2 Constipation severity (5 point scale) (12 weeks) Higher score = worse								Mean difference
Chey 2012	RCT	Serious^b	n/a	No serious	Very serious^d	401	403	Least Sq mean change I = −1.2 C = −0.6 (no SD) p<0.0001		Very low
Rao 2012	RCT	Very serious^b	n/a	No serious	Very serious^d	405	395	Least Sq mean change I = −1.2 C = −0.6 (no SD) p<0.0001		Very low
Johnston 2010	RCT	Serious^b	n/a	No serious	Very Serious^d	84	85	I= 1.35 C= 0.75 (no SD, 95% CIs and no p-value reported)		Very low
6.2 Constipation severity (5 point scale) (26 weeks) Higher score = worse								Mean difference
Chey 2012	RCT	Serious^b	n/a	No serious	Very serious^d	401	403	Least Sq mean change I = −1.2 C = −0.6 (no SD) p<0.0001		Very low

(a): Chey 2012, Rao 2012
(b): Use of rescue medication (laxatives) not reported by study arm has major potential for confounding on constipation severity. Bulk laxatives and stool softeners were also permitted by one study (Rao et al 2012). Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(c): 30% improvement (EMA recs for continuous outcomes) on 7 point BSFS = decrease of 2.1 points thus derivation of responder status using 1 point change for ≥half study weeks was not deemed to be clinically relevant.
(d): Does not meet MID (30% improvement (EMA recs for continuous outcomes) on 5 point scale =1.5 points), no 95% CIs.
*: BSFS = Bristol Stool Form Scale

Table 87GRADE Profile, Outcome: relapse or flatulence or bloating

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg (%)	C = Placebo (%)	Effect		Quality
7.1 Bloating (% with improvement of ≥30% for ≥half of the study weeks)								Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No Serious	No serious	No serious	348/806 (43)	214/798 (27)	1.61 [1.40, 1.85]	16 more per 100 (11 more, 23 more)	Moderate

(a): Chey 2012, Rao 2012
(b): Use of rescue medication (laxatives) not reported by study arm has potential for confounding on bloating. Bulk laxatives and stool softeners were also permitted by one study (Rao et al 2012). Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.

Discontinuation, Safety and Adverse Events

Table 88GRADE profile - discontinuation (all reasons)

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linclotide 290µg (%)	C = Placebo (%)	Effect		Quality
								Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No Serious	No serious	Serious^c	202/808 (25)	160/800 (20)	1.25 [1.04, 1.50]	5 more per 100 (3 more, 12 more)	Low

(a): Chey 2012, Rao 2012
(b): Use of rescue medication (laxatives) not reported by study arm has potential for confounding on perceived efficacy. Bulk laxatives and stool softeners were also permitted by one study (Rao et al 2012).
(c): CIs cross line of MID (GRADE default suggestion)

Table 89GRADE profile - reason for discountinuation

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg (%)	C = Placebo (%)	Effect		Quality
Adverse Event								Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No Serious	No serious	No serious	73/808 (9)	20/800 (3)	3.62 [2.23, 5.87]	7 more per 100 (3 more, 12 more)	Moderate
Adverse Event = Diarrhoea								Relative (95% CI)	Absolute (95% CI)
3^c	RCTs	Serious^b	No Serious	No Serious	No serious	55/893 (6)	3/885 (0.3)	18.19 [5.72, 57.88]	6 more per 100 (2 more, 19 more)	Moderate
Withdrew Consent								Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No Serious	No Serious	Serious^d	49/808 (6)	51/800 (6)	0.95 [0.65, 1.39]	0 fewer per 100 (2 fewer, 2 more)	Low
Insufficient Therapeutic Response								Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No serious	No serious	No serious	20/808 (2)	37/800 (5)	0.54 [0.32, 0.92]	2 fewer per 100 (0 fewer, 3 fewer)	Moderate

(a): Chey 2012, Rao 2012
(b): Use of rescue medication (laxatives) not reported by study arm has potential for confounding on perceived efficacy. Bulk laxatives and stool softeners were also permitted by one study (Rao et al 2012).
(c): Chey 2012, Rao 2014, Johnston 2013
(d): CIs cross line of effect

Table 90GRADE profile - adverse events

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg (%)	C = Placebo (%)	Effect		Quality
At least one Adverse Event								Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No Serious	No serious	Serious^c	491/808 (60)	438/799 (55)	1.11 [1.02, 1.21]	6 more per 100 (1 more, 12 more)	Low
Diarrhoea								Relative (95% CI)	Absolute (95% CI)
3^d	RCTs	Serious^b	No Serious	No serious	No serious	172/893 (19)	25/884 (3)	6.80 [4.52, 10.23]	16 more per 100 (10 more, 26 more)	Moderate
Abdominal Pain								Relative (95% CI)	Absolute (95% CI)
3^d	RCTs	Serious^b	No Serious	No serious	Serious^e	44/893 (5)	29/884 (3)	1.50 [0.95, 2.38]	2 more per 100 (0 fewer, 5 more)	Low
Flatulence								Relative (95% CI)	Absolute (95% CI)
2^a	RCT	Serious^b	No Serious	No serious	No serious	35/808 (4)	15/799 (2)	2.31 (1.27, 4.20)	2 more per 100 (1 more, 6 more)	Moderate
Abdominal Distension								Relative (95% CI)	Absolute (95% CI)
2^a	RCT	Serious^b	No Serious	No serious	Serious^e	18/808 (2)	9/799 (1)	1.98 (0.90, 4.39)	1 more per 100 (0 fewer, 4 more)	Low
Nausea								Relative (95% CI)	Absolute (95% CI)
Johnston 2010	RCT	Serious^b	n/a	No serious	Serious^e	1/85 (1)	5/85 (6)	0.2 (0.02, 1.68)	5 fewer per 100 (6 fewer, 27 more)	Low
UTI								Relative (95% CI)	Absolute (95% CI)
Johnston 2010	RCT	No Serious	n/a	No serious	Serious^e	5/85 (6)	2/85 (2)	2.5 (0.50, 12.5)	4 more per 100 (1 fewer, 27 more)	Moderate

(a): Chey 2012, Rao 2012
(b): Use of rescue medication (laxatives) not reported by study arm has potential for confounding on adverse events. Bulk laxatives and stool softeners were also permitted by one study (Rao et al 2012). Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(c): Point estimate does not reach MID (GRADE default suggestion) with CIs also below threshold.
(d): Chey 2012, Rao 2012, Johnston 2010
(e): CIs cross line of effect

Table 91GRADE profile - serious adverse events

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Linaclotide 290µg	C = Placebo	Effect		Quality
Serious Adverse Events^a								Relative (95% CI)	Absolute (95% CI)
3^b	RCTs	No serious	No Serious	No serious	Serious^c	7/893 (1)	9/884 (1)	0.79 [0.30, 2.04]	2 fewer per 1000 (7 fewer, 11 more)	Moderate

(a): Chey 2012, n=4 (intervention arm) cuff syndrome (1), appendicitis (1), cystopexy (1) and Hodgkin’s disease (1). Rao 2012, n=2 (intervention arm) asthma (1) and pericardial effusion and pericarditits leading to withdrawal from the study (1). Johnston 2010, n=1 (intervention arm) faecal impaction requiring hospitalisation (1).
(b): Chey 2012, Rao 2012, Johnston 2010
(c): CIs cross line of effect

H.4. Review question 4 (lubiprostone)

Table 92GRADE profile, Outcome: Quality of Life (QOL)

Quality assessment						Number of patients		Effect	Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Lubiprostone	C = Placebo	Mean Difference	Quality
1.1 IBS QOL (34 Questions) Higher score = worse
Drossman 2009 ^a	RCT	Serious^b	Cannot be assessed.	No serious	Very serious^c	769	385	Reported as Non-Significant only (no p value)	Very Low
Johanson 2008	RCT (phase 2)	Serious^b	N/A	No serious	Very Serious^c	145	48	Reported as Non-Significant only (no p value)	Very Low
1.2 Life interference (11 point scale, sub scale of IBS-SS) Higher score = worse
Whitehead 2011	RCT 14 day crossover (14 day washout)	Serious^d	N/A	No serious	Serious^e	60^f	60^f	0.23 [−0.48, 0.94]	Low

(a): Drossman reported data from 2 previously unpublished RCTs (no references therefore available)
(b): Use of rescue medications (laxatives) was permitted with no reporting of use by study arm. Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(c): No effect size or confidence intervals
(d): Unclear if ITT analysis performed. Unclear if sub analysis of scale is validated in isolation as surrogate measure of QOL.
(e): Point estimate is below MID (on 11 point scale, 30% improvement (EMA recs) = 3.3 points). CIs cross the line of effect.
(f): Total sample was 60 but was crossover study hence 60 in each arm

Table 93GRADE Profile, Outcome: symptom severity

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Lubiprostone	C = Placebo	Mean difference		Quality
2.1 IBS SS (Symptom Severity, /500) Higher score = worse
Whitehead 2011	RCT 14 day crossover (14 day washout)	No Serious	N/A	No serious	Very serious^a	62^b	62^b	7.68 [−34.89, 50.25]		Low
2.2 Overall Responder Status (degree of relief over time)^c						I = Lubiprostone (%)	C = Placebo (%)	Relative (95% CI)	Absolute (95% CI)
Drossman^d 2009	RCT	Very serious^e	Unable to assess	No serious	No serious	138/769 (18)	39/385 (10)	1.77 [1.27, 2.47]	8 more per 100 (3 more, 15 more)	Low

(a)

Point estimate below MID (on a 500 point scale, 30% improvement (EMA recs) = 150 points,) CIs cross line of effect.

(b)

Total sample was 62 but was crossover study hence 62 in each arm

(c)

Relief measured “How would you rate your relief of IBS symptoms over the past week compared to how you felt before you entered the study?” (7 point scale, 1=significantly worse, 2=moderately worse, 3=a little bit worse, 4=unchanged, 5=a little bit relieved, 6=moderately relieved, 7=significantly relieved). Classifications of responders:

Weekly – moderate or significantly relieved for that week (secondary study endpoint).
Monthly – moderately relieved or better in 4 out of 4 weeks OR significantly relieved in 2 out of 4 weeks. Could not discontinue treatment during 4 week period and % of days of rescue medication did not increase from baseline (Secondary study endpoint)
Overall – Monthly responders for at least 2 of the 3 months of the study (primary study endpoint).

(d)

Drossman study reported data from 2 previously unpublished RCTs

(e)

Use of rescue medications (laxatives) was permitted with no report of use by study arm. Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm. Outcome reporting/recall bias suspected as question asked to identify responder status was leading and retrospective, with no mention of validation.

Table 94GRADE Profile, Outcome: abdominal pain (10 point scale) higher score = worse

Quality assessment

Number of patients

Effect

Quality

Number of studies

Design

Risk of bias

Inconsistency

Indirectness

Imprecision

I = Lubiprostone

C = Placebo

Mean Difference

Whitehead 2011

RCT

Serious^a

N/A

No serious

Very Serious^b

I = −0.80

C = −0.85

Mean difference (95%CI)

0.05 [−0.74, 0.81]

Calculated by reviewer.

Very low

(a): It was not stated whether ITT analysis was performed
(b): Effect size below MID (on 10 point scale, 30% = improvement of 3 points (EMA recs)). 95% CIs cross line of effect.

Table 95GRADE Profile, Outcome: stool score/general changes in bowel habit

Quality assessment						Number of patients		Effect	Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Lubiprostone	C = Placebo	Mean Change (improvement)	Quality
6.1 Spontaneous Bowel Movements (frequency per week) greater frequency desirable
Drossman^a 2009	RCT	Serious^b	N/A	No serious	Very serious^c	769	385	Reported as non-significant only. No P value.	Very low
Johanson 2008	RCT	Serious^b	N/A	No serious	Serious^d	(all dose arms) 145	48	I : 1.9 (BL 3.6, Wk12 5.5) C : 0.5 (BL 4.3, Wk12 4.8) P=0.0296	Low
6.2 Constipation Severity (5 point scale) Higher score = worse
Johanson 2008	RCT (phase 2)	Serious^b	N/A	No serious	Very serious^e	(all dose arms) 145	48	I : −0.6 (BL 2.2, Wk12 1.6) C : −0.3 (BL 2.1, Wk12 1.8) P=0.0056	Very low
6.3 Stool Output (Days with hard/lumpy stools or no stools (%))
Whitehead 2011	RCT 14 day crossover (14 day washout)	No Serious	N/A	No serious	Very Serious^f	60	60	% days without event (difference) I : −16.7 (BL 59.4, F/UP 42.7) C : −15.9 (BL 59.4, F/UP 43.5) (no p values reported)	Low

(a): Drossman study reported data from 2 previously unpublished RCTs
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm. Fibre/diet/fluid/exercise/other relevant meds were not reported by study arm.
(c): No effect size reported
(d): MID is met in intervention arm (30% improvement (EMA recs) based on mean of baseline frequency = 1.2 movements per week). No SD or 95%CIs reported
(e): MID is not met (30% improvement (EMA recs) based on 5 point constipation severity scale =1.5 point improvement). No SD or 95% CIs reported
(f): No SD or 95%CIs and effect size does not reach MID (30% improvement in days with hard/lumpy or no stools = 5 days without event. This was not met in either arm. I = Actual days = from 16.5 to 12.2, difference 4.3 days, C = Actual days = from 16.5 to 11.9 days, difference 4.6 days)

Table 96GRADE Profile, Outcome: bloating

Quality assessment						Number of patients		Effect	Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	I = Lubiprostone	C = Placebo	Mean difference	Quality
Bloating (11 point scale) Higher score = worse
Whitehead 2011	14 day crossover RCT (+14 day washout)	No serious	N/A	No serious	Serious^a	60^b	60^b	0.04 [−0.94, 1.02]	Moderate

(a): Effect size does not reach MID (30% improvement (EMA recs) on 10 point scale = 3 points) and CIs cross line of effect
(b): Total sample was 60 but was crossover study hence 60 in each arm

Discontinuation and Adverse Events

Table 97GRADE profile, Outcome: discontinuation (all reasons)

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
3^a	RCTs	Serious^b	No serious	No serious	Serious^c	193/820 (23)	106/436 (24)	0.99 [0.81, 1.21]	0 fewer per 100 (5 fewer to 5 more)	Low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008, 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm. This could affect perceived efficacy and thus discontinuation.
(c): Point estimate does not reach MID (GRADE default suggestion), CIs cross line of effect

Table 98GRADE profile, Outcome: discontinuation due to adverse event

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
3^a	RCTs	Serious^b	Serious^c	No serious	Serious^d	44/820 (5)	25/436 (6)	1.08 [0.44, 2.67]	0 more per 100 (3 fewer, 10 more)	Very low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm
(c): Significant heterogeneity. Random Effects analysis, I²61%, Chi² = 5.17 (p=0.08)
(d): Point estimate does not reach MID (GRADE default suggestion), CIs cross line of effect

Table 99GRADE profile, Outcome: discontinuation due to lack of efficacy

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
3^a	RCTs	Serious^b	No serious	No serious	Serious^c	32/820 (4)	22/436 (5)	0.84 [0.49, 1.43]	1 fewer per 100 (3 fewer, 2 more)	Low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm.
(c): Point estimate does not reach MID (GRADE default suggestion) and CIs cross line of effect

Table 100GRADE profile, Outcome: discontinuation due to non-compliance

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
3^a	RCTs	Serious^b	No serious	No serious	Serious^c	22/820 (3)	6/436 (1)	1.83 [0.77, 4.34]	1 more per 100 (0 fewer, 5 more)	Low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm.
(c): CIs cross line of no effect.

Table 101GRADE profile, Outcome: discontinuation due to withdrawn consent

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
3^a	RCTs	Serious^b	No serious	No serious	Serious^c	66/820 (8.2)	38/436 (9)	0.89 [0.61, 1.29]	1 fewer per 100 (3 fewer to 3 more)	low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm
(c): Point estimate indicates no MID (GRADE default suggestion). CIs cross line of effect.

Table 102GRADE profile, Outcome: adverse event (at least one)

Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)
2^a	RCTs	Serious^b	No serious	No serious	Serious^c	422/825 (51)	225/435 (52)	1.00 [0.90, 1.12]	0 fewer per 100 (5 fewer, 6 more)	Low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm
(c): Point estimate indicates no difference. CIs cross line of no effect.

Table 103GRADE profile, Outcome: adverse event = nausea

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
2^a	RCTs	Serious^b	No Serious	No serious	No serious	76/825 (9)	21/435 (5)	2.14 [1.34, 3.41]	6 more per 100 (2 more, 12 more)	Moderate

(a): Drossman 2009 (data from 2 RCTs combined), Johanson 2008
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm

Table 104GRADE profile, Outcome: adverse event = diarrhoea

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
2^a	RCTs	Serious^b	Serious^c	No serious	Serious^d	58/825 (7)	17/435 (4)	2.63 [0.68, 10.23]	6 more per 100 (1 fewer, 36)	Very low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm
(c): Significant heterogeneity. I² 69% Chi² = 3.23, (p=0.07)
(d): Point estimate indicates the risk of diarrhoea is greater in the lubiprostone group but the CIs cross the threshold for MID (GRADE default suggestion).

Table 105GRADE profile, Outcome: adverse event = abdonimal distension

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
2^a	RCTs	Serious^b	No serious	No serious	Very serious^c	21/825 (2)	13/435 (3)	1.01 [0.51, 2.00]	0 more per 100 (1 fewer, 3 more)	Very low

(a): Drossman 2009 (Drossman 2009 included data from two RCTs so this study is counted as 2), Johanson 2008 48µg dose arm only.
(b): Use of rescue medications (laxatives) was permitted with no report of use by study arm
(c): Point estimate indicates the risk of abdominal distension is borderline higher in the lubiprostone group but this is below the MID (GRADE default suggestion) and the CIs cross the line of effect.

Table 106GRADE profile, Outcome: serious adverse events^a

Quality assessment						Number of patients		Effect		Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Lubiprostone (%)	Placebo (%)	Relative (95% CI)	Absolute (95% CI)	Quality
2^b	RCTs	No serious	N/Ac	No serious	Serious^d	10/831 (1)	4/435 (1)	1.35 [0.47, 3.90]	0 more per 100 (0 fewer, 3 more)	Moderate

(a): Drossman 2009 n=8, cardiac arrest on background of multiple co-morbidities leading to death (1), Non-cardiac related chest pain (1), not specified (6). Johanson 2008 n=2 (48µg arm only) (out of a total of 3 across all doses), perforated appendix (1), cholecystitis (1), ectopic pregnancy (1). (NB Whitehead 2011 made no mention of SAEs)
(b): Drossman 2009 (data from 2 RCTs combined), Johanson 2008, (48 µg dose arm only)
(c): Second study has zero events in both arms
(d): CIs cross line of effect.

H.5. Review question 5a (relaxation therapy)

Table 107GRADE profile, Relaxation vs routine clinical care/ control/ enhanced medical care (dichotomous outomes)

Quality assessment							No of patients		Effect estimate		Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	Treatment	Comparator	Relative (95% CI)	Absolute	Quality
Outcome: Adequate relief
Shinozaki 2010	RCT	Very serious^a	No indirectness	No inconsistency	Serious^b	None	9/11 (81.8%)	3/10 (30%)	RR 2.73 (1.02, 7.32)	519 more per 1000 (from 6 more to 1000 more)	VERY LOW

(a): Randomisation not described, allocation concealment not reported, blinding of investigators not reported, unclear whether patients continued to take other medication during study, attrition not reported for this small study (n=21) (Shinozaki, 2010). Unclear whether outcome was a validated tool. Downgraded 2 levels.
(b): Lower limit of 95%CI crosses MID at 1.25, leading to uncertainty in clinical effectiveness of the treatment. Downgraded 1 level.

Table 108GRADE profile, Relaxation vs routine clinical care/ control/ enhanced medical care

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	Treatment (T)	Comparator (C)	Mean difference (95% CI)	Quality
Outcome: SIBSQ (Total score = 98; ≥30% improvement = ≥29.4 points increase/decrease from baseline)
Shinozaki, 2010	RCT	Very serious^(a)	No serious	no serious	Very serious^(c)	No serious	11	10	T= 48.9; C=36.3 MD=12.60 (−1.91 to 27.11) Mean change from baseline ^* : T=−3.2; C=−19.6 Difference: 16.4	VERY LOW
Outcome: SDS (Total score = 70; ≥30% improvement = ≥21 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(a)	No serious	no serious	Very serious^(c)	No serious	11	10	T=44.6; C=45.8 MD= −1.20 (−8.48 to 6.08) Mean change from baseline ^* : T= −1.8; C= −0.01 Difference:1.79	VERY LOW
Outcome: STAI State Anxiety (Total score= 60; ≥30% improvement =≥18 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(a)	No serious	no serious	Very serious^(c)	No serious	11	10	T=47.2; C=51.4 MD= −4.20 (−12.21 to 3.81) Mean change from baseline ^* : T= −2.8; C=−3.2 Difference:0.4	VERY LOW
Outcome: STAI Trait Anxiety (Total score= 60; ≥30% improvement =≥18 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(a)	No serious	no serious	Very serious^(c)	No serious	11	10	T=54.5; C=52.8 MD= 1.70 (−8.87 to 12.27) Mean change from baseline ^* : T= −1.5; C=−4.0 Difference: 2.5	VERY LOW
Outcome: SF36 Physical function- 8 weeks follow up^(*) (Total Score 100;; ≥30% improvement =≥30 points increase?/decrease? from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	No serious	30	35	T=70.6; C=67.5 MD= 2.73 (−3.40 to 8.86) Mean change from baseline ^* : T=+7.5; C=−2.3 Difference:9.8	VERY LOW
Outcome: SF36 Physical function-52 weeks follow up ^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	no serious	Very serious^(c)	No serious	13	21	T=91.9; C=88.8 MD= 3.10 (−8.00 to 14.20) Mean change from baseline ^* : T=-+12.5; C=+2.3 Difference:10.2	VERY LOW
Outcome: SF36 Role physical- 8 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	No serious	30	35	T= 53.9; C= 46.6 MD= 6.59 (−8.01 to 21.19) Mean change from baseline ^* : T=+27.65; C=+3.3 Difference: 24.35	VERY LOW
Outcome: SF36 Role physical-52 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	no serious	Serious^(d)	No serious	13	21	T=38.1; C=64.5 MD= 10.50 (−16.89 to 37.89) Mean change from baseline^*: T=+30.7; C=+1.6 Difference: 29.1	VERY LOW
Outcome: SF36 Bodily pain- 8 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	no serious	30	35	T=54.64; C= 54.6 MD 1.29 (−6.41 to 8.99) Mean change from baseline ^* : T=+9.75; C=+5.5 Difference: 4.25	VERY LOW
Outcome: SF36 Bodily pain- 52 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	no serious	Very serious^(c)	no serious	13	21	T=64.2; C=68 MD= −3.80 (−19.18 to 11.58) Mean change from baseline ^* : T=+11.2; C=+8.7 Difference: 2.5	VERY LOW
Outcome: SF36 General health- 8 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	no serious	30	35	T=48.2; C= 49.15 MD= −1.05 (−9.40 TO 7.30) Mean change from baseline ^* : T=+3.4; C=+0.05 Difference: 3.35	VERY LOW
Outcome: SF36 General health - 52 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	no serious	Very serious^(c)	no serious			T=65.9; C=66 MD= −0.10 (−15.85 TO 15.65) Mean change from baseline ^* : T=+6.2; C=+0.6 Difference: 5.6	VERY LOW
Outcome: SF36 Vitality- - 8 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	no serious	30	35	MD= 2.38 (−4.01 TO 8.78) Mean change from baseline ^* : T=11.7; C=+7.8 Difference: 3.9	VERY LOW
Outcome: SF36 Vitality- 52 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	no serious	Very serious^(c)	no serious	13	21	T= 61.5; C=59.2 MD= 2.30 (−12.48 TO 17.08) Mean change from baseline ^* : T=+13.5; C=+9.2 Difference:4.3	VERY LOW
Outcome: SF36 Social functioning- 8 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	no serious	30	35	MD= −3.46 (−12.08 TO 5.16) Mean change from baseline ^* : T=+17.0; C=+14.7 Difference: 2.3	VERY LOW
Outcome: SF36 Social functioning- 52 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	no serious	Very serious^(c)	no serious	13	21	T= 76.9; C=80.3 MD= −3.40 (−18.97 TO 12.17) Mean change from baseline ^* : T=+10.8; C=+7.2 Difference: 3.6	VERY LOW
Outcome: SF36 Role emotional- 8 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	no serious	Very serious^(c)	no serious	30	35	MD= 6.23 (−5.19 TO 17.66) Mean change from baseline ^* : T= +23.4; C= +2.9 Difference: 20.5	VERY LOW
Outcome: SF36 Role emotional- 52 weeks follow up^(*)(Total Score 100;; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=66.7; C=75 MD= −8.30 (−37.70 to 21.10) Mean change from baseline ^* : T=+8.6; C= +4.3 Difference: 4.3	VERY LOW
Outcome: SF36 Mental health- 8 weeks follow up^(*)(Total Score 100; ≥30% improvement =≥30 points increase/decrease from baseline)
2^≠	RCT	Very serious^(a) ^,^(b)	No serious	No serious	Very serious^(c)	no serious	30	35	MD= 2.24 (−4.12 to 8.60) Mean change from baseline ^* : T=+7.3; C=+8.9 Difference:1.6	VERY LOW
Outcome: SF36 Mental health- 52 weeks follow up^(*)(Total Score 100; ≥30% improvement =≥30 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=71.4; C=77.1 MD= −5.70 (−17.06 to 5.66) Mean change from baseline ^* : T=+7.3; C=+12.4 Difference: 5.1	VERY LOW
Outcome: BSSS- Frequency- 52 weeks follow up^(*)(Total Score 48; ≥30% improvement =≥16.4 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=16.2; C= 17 MD= −0.80 (−3.61 to 2.01) Mean change from baseline ^* : T- −4.4; C=−4.0 Difference: 0.4	VERY LOW
Outcome: BSSS-Distress- 52 weeks follow up^(*)(Total Score 48; ≥30% improvement =≥16.4 points increase/decrease from baseline)
Shinozaki, 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=13.2; C=12.5 MD= 0.70 (−2.29 to 3.69) Mean change from baseline ^* : T=−4.5; C=−3.8 Difference: 0.7	VERY LOW
Outcome: BSSS- Interference- 52 weeks follow up^(*)(Total Score 48; ≥30% improvement =≥16.4 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=13.2; C=12.5 MD= 0.70 (−2.29 to 3.69) Mean change from baseline ^* : T=−4.5; C=−3.8 Difference: 0.7	VERY LOW
Outcome: Automatic thoughts questionnaire- 52 weeks follow up^(*)(Total Score 120; ≥30% improvement =≥36 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=40.31; C=19.56 MD= −0.17 (−9.47 to 9.13) Mean change from baseline ^* : T= −5.9; C=−3.16 Difference: 2.74	VERY LOW
Outcome: Locus of control of behaviours- 52 weeks follow up^(*)(Total Score 85; ≥30% improvement =≥25.5 points increase/decrease from baseline)
Shinozaki 2010	RCT	Very seriouss^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T=24.23; C=27.9 MD=−3.67 (−13.88 to 6.54) Mean change from baseline ^* : T= −5.55; C=−1.19 Difference: 4.36	VERY LOW
Outcome: Hospital anxiety and depression scale- 52 weeks follow up^(*)(Total Score 42; ≥30% improvement =≥12.6 points increase?/decrease? from baseline)
Shinozaki 2010	RCT	Very serious^(b)	No serious	No serious	Very serious^(c)	no serious	13	21	T= 10.7; C=11 MD= −0.30 (−4.68 to 4.08) Mean change from baseline ⁺ : T=−3.3; C=−3.1 Difference: 0.2	VERY LOW

(*): For the SF36 outcomes, one study reported the outcomes at 4 separate time points. As IBS is a chronic condition it was decided to assess the quality on the 8 week follow up (where 2 studies report results)and the latest follow up point from baseline (52 weeks).For outcomes that were not pooled, the latest time point (52 weeks follow up) is used.
(+): Mean change from baseline calculated by analyst
(≠): Shinozaki 2010, Boyce 2003
(a): Randomisation not described, allocation concealment not reported, blinding of investigators not reported, unclear whether patients continued to take other medication during study, attrition not reported for this small study (n=21) (Shinozaki, 2010). Downgraded 2 levels.
(b): In Boyce (2003) baseline scores for all SF36 domains (apart from vitality and mental health) were lower in the Relaxation Therapy (RT) group compared to routine clinical care (RCC). Attrition at week 8 was 26.5% in RCC and 45.7% in RT; at week 52 attrition was 38% in RTT and 64% in RT. Only per protocol data was presented in the study. Downgraded 2 levels.
(c): 95%CI for mean difference between groups post treatment included both positive and negative effects making the direction of effect and the effect size very uncertain. The mean change from baseline for both groups did not reach clinical significant difference. Downgraded 2 levels.
(d): 95%CI for mean difference between groups post treatment included both positive and negative effects making the direction of effect and the effect size very uncertain.The mean change from baseline for the relaxation group was clinically significant at >30% change from baseline.Downgraded 1 level.

Table 109GRADE profile, Relaxation vs enhanced medical care

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	Treatment (T)	Comparator (C)	Mean difference from baseline difference (95% CI)	Quality
Outcome: Impairment severity score – bodily impairment (Total Score 12; ≥30% improvement =≥3.6 points increase/decrease from baseline)
Lahman 2010	RCT	Serious^(a)	No serious	No serious	Serious^(c)	no serious	40	40	T=1.69; C=0.79 MD= −0.39 (−0.77 TO −0.01) Mean change from baseline ⁺ : T=−0.51; C=−0.04 Difference:0.47	LOW
Outcome: Impairment severity score – psychic impairment (Total Score 12; ≥30% improvement =≥3.6 points increase/decrease From baseline)
Lahman 2010	RCT	Serious^(a)	No serious	No serious	Serious^(d)	no serious	40	40	T=1.64; C= 1.88 MD= −0.24 (−0.59 to 0.11) Mean change from baseline ⁺ : T=−0.42; C=−0.09 Difference: 0.33	LOW
Outcome: Impairment severity score – social impairment (Total Score 12; ≥30% improvement =≥3.6 points increase/decrease from baseline)
Lahman 2010	RCT	Serious^(a)	No serious	No serious	Serious^(d)	no serious	40	40	T=1.01; C=1.14 MD= −0.13 (−0.53 to 0.27) Mean change from baseline ⁺ : T=0.07; C=−0.08 Difference:0.15	LOW
Outcome: Overall IBS symptoms (Total Score 40; ≥30% improvement =≥12 points increase/decrease from baseline)
Lahman 2010	RCT	Serious^(b)	No serious	No serious	Serious^(c)	no serious	40	40	T=26.2; C=30.6 MD= −4.40 (−7.23 to - 1.57) Mean change from baseline ⁺ : T=−5.6; C=−0.04 Difference:5.56	LOW
Outcome: Abdominal pain (Total Score 40; ≥30% improvement =≥12 points increase/decrease from baseline)
Lahman 2010	RCT	Serious^(b)	No serious	No serious	Very serious^(e)	no serious	40	40	T= 25.7; C=27.3 MD=−1.60 (−6.07 to 2.87) Mean change from baseline ⁺ : T=−7.3; C= −4.1 Difference: 2.2	VERY LOW
Outcome: Deterioration – diarrhoea & constipation (Total Score 40; ≥30% improvement =≥12 points increase/decrease from baseline)
Lahman 2010	RCT	Serious^(b)	No serious	No serious	Very serious^(e)	no serious	40	40	T= 29.1; C=29.2 MD= −0.10 (−3.45 TO 3.25) Mean change from baseline ⁺ : T=−4.3; C= −2.2 Difference: 2.1	VERY LOW
Outcome: Bloating (Total Score 40; ≥30% improvement =≥12 points increase/decrease from baseline)
Lahman 2010	RCT	Serious^(b)	No serious	No serious	Serious^(c)	no serious	40	40	T=28.1; C=33.2 MD= −5.10 (−8.41 to - 1.79) Mean change from baseline ⁺ : T=−7.3; C=−1.7 Difference: 5.6	LOW

: For all outcomes in Table X, outcomes reported at 5 weeks and 3 months; due to chronic nature of IBS, only the outcomes for 3 month follow up reported as most clinically relevant.
(+): Mean change from baseline calculated by analyst
(a): Unclear whether outcome measure BSS/ ISS is validated, paper states that it is widely used in Germany. Downgraded 1 level.
(b): Outcome measure is patient- reported subjective measurement on a scale of 10–50, lack of further detail in study. Downgraded 1 level.
(c): The mean change from baseline for both groups does not reach clinical significance. The confidence intervals do not cross the line of no difference indicating the estimate of the effect is precise. Downgraded 1 level.
(d): The mean change from baseline for both groups does not reach clinical significance. The confidence intervals do cross the line of no difference, but are narrow indicating a precise estimate. Downgraded 1 level.
(e): 95%CI for mean difference between groups post treatment included both positive and negative effects making the direction of effect and the effect size very uncertain. The mean change from baseline for both groups did not reach clinical significant difference.Downgraded 2 levels.

Table 110GRADE profile, Relaxation vs hypnotherapy

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	Treatment (T)	Comparator (C)	Median (Range)	Quality
Outcome: Overall symptom score (Total score 30 ≥30% improvement =≥9 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious ^(a)^,^(b)	No serious	No serious	Very serious^(c)	no serious	13	12	T = 11; C =7.5 Test (p-value): NR Median change from baseline⁺: T= 0.0; C= −7.0 Difference: 7.0	VERY LOW
Outcome: GHQ – Sum (Total score 36 ≥30% improvement =≥14.8 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T= 22 (11–35); C= 22.5 (5–64) Test (p-value): NR Median change from baseline⁺: T= +2.5; C=−4 Difference: 6.5	VERY LOW
Outcome: HADS- Anxiety (Total score 21 ≥30% improvement =≥6.3 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=8 (0–15); C=10.5 (2–15) Test (p-value): NR Median change from baseline⁺: T=+3; C=−1 Difference: 4	VERY LOW
Outcome: HADS- Depression (Total score 21 ≥30% improvement =≥6.3 points increase/decrease from baseline)
1 Forbes, 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=4 (0–15) C=4 (0–13) Test (p-value): NR Median change from baseline⁺: T=0; C=−1.5 Difference: 1.5	VERY LOW
Outcome: SF36- Physical function (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=87 (70–100) C=75 (35–100) Test (p-value): NR Median change from baseline⁺: T=−8; C=+8 Difference: 16	VERY LOW
Outcome: SF36- Physical role (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T= 25 (0–100) C= 50 (0–100) Test (p-value): NR Median change from baseline⁺: T=−50; C=+25 Difference: 75	VERY LOW
Outcome: SF36- Emotional role (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=100 (0–100) C= 67 (0–100) Test (p-value): NR Median change from baseline⁺: T=0; C=0 Difference: 0	VERY LOW
Outcome: SF36- Social function (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T= 75 (37–100) C= 44 (12–100) Test (p-value): NR Median change from baseline⁺: T=0; C=−6 Difference: 6	VERY LOW
Outcome: SF36- Pain (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=56 (12–84) C= 46 (0–100) Test (p-value): NR Median change from baseline⁺: T=+5; C=+5 Difference: 0	VERY LOW
Outcome: SF36- Mental state (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T= 62 (40–88) C= 52 (36–84) Test (p-value): NR Median change from baseline⁺: T=−10; C=0 Difference: 10	VERY LOW
Outcome: SF36- Vitality (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=50 (15–95) C= 30 (5–75) Test (p-value): NR Median change from baseline⁺: T=0; C=−3 Difference: 3	VERY LOW
Outcome: SF36- Perception of health (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T=52 (20–100) C= 53 (5–87) Test (p-value): NR Median change from baseline⁺: T=−13; C=+16 Difference: 29	VERY LOW
Outcome: SF36- Health change (Total score 100 ≥30% improvement =≥30 points increase/decrease from baseline)
Forbes 2000	RCT	Very serious^(a)	No serious	No serious	Very serious^(c)	no serious	13	12	T= 50 (25–100) C= 67 (0–100) Test (p-value): NR Median change from baseline⁺: T=0; C=+17 Difference: 17	VERY LOW

(a): Allocation concealment not reported, over 50% attrition (equal between groups), results reported on an available case basis, higher self- rating of health (SF36 and HAD) in the audiotape group compared to the hypnotherapy group (though the authors reported that this did not reach significance), the same person recorded the relaxation tape as undertook the hypnotherapy,patients were allowed to continue with pre-existing therapy for IBS, downgraded 2 levels.
(b): The primary outcome of overall symptom score is not a validated outcome measure and the overall symptom scores were not comparable at baseline. The results were calculated by the analyst and were not clearly reported in the study, downgraded 1 level.
(c): Only median values were reported in the paper, no interquartile range was reported; this meant that imprecision could not be assessed based on the specific threshold of MID. Therefore the default for continuous outcomes for GRADE was used; the optimal information size, calculated by the GRADE working group is 400; the number of events in intervention and control groups in this study did not reach this threshold, downgraded 2 levels.

H.6. Review question 5b (CCBT and Mindfulness therapy)

Table 111GRADE profile 1a, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Waitlist (online discussion forum) (W-ODF)

Quality assessment							No of patients		Effect estimate		Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M/E	W-ODF	Relative (96% CI)	Absolute	Quality
Outcome: IBS symptoms: GSRS-IBS^d Responder (≥50% reduction in total score) (10-wks)
Ljotsson 2010	RCT	Very Serious^a^,^b	Serious^(d)	Not applicable	No serious	No serious	15/42 (35.7%)	1/43 (2.3%)	RR 15.36 (2.12 to 111.13)	33 more per 100 (from 3 more to 100 more)	LOW

(a): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(b): Participants were ‘self-referred’, potential selection bias of participants who were more likely to experience an effect. Reason for withdrawal not reported.
(c): GSRS-IBS (Gastrointestinal Symptom Rating Scale for IBS) (total score: 13 to 91, with 13 = no discomfort at all)
(d): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the UK compared to Sweden.

Table 112GRADE profile 1b, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Waitlist (online discussion forum) (W-ODF)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M/E (T)	W-ODF (C)	Mean difference (95% CI)	Quality
Outcome: Quality of life: IBS-QoL^d (10-wks)
2^(a)	RCT	Very Serious^b^,^c	Serious^(h)	No serious	No serious^f	No serious	65	70	T = 72.8, C = 52.9(at 3 months follow up) MD = 17.93 (11.25 to 24.60)	LOW
Outcome: IBS symptoms: GSRS-IBS (10-wks)
2^(a)	RCT	Very Serious^b^,^c	Serious^(h)	No serious	Serious^g	No serious	65	70	T = 32.4 C = 47.3 (at 3 months follow up) MD = −13.6 (−17.23 to −8.88)	VERY LOW

(a): Ljotsson (2010, 2011b)
(b): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(c): Participants were ‘self-referred’, potential selection bias of participants who were more likely to experience an effect. Reason for withdrawal not reported.
(d): IBS-QoL (total score: 0 to 100, with 0 = minimum QoL). Drossman (2007) suggested minimum clinical important difference as ≥14 points improvement from baseline.
(e): GSRS-IBS (Gastrointestinal Symptom Rating Scale for IBS) (total score: 13 to 91, with 13 = no discomfort at all). FDA and EMA suggested MID = 30% reduction = at least 23.4 points increase from baseline.
(f): Mean difference between groups showed significant effect, treatment group reached the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +20.6; C = −0.9], no downgrade.
(g): Although mean difference between groups showed significant effect, both groups end scores did not reach the ≥30% MID [Mean change from baseline: T = −16.1; C = −2.3]
(h): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the UK compared to Sweden.

Table 113GRADE profile 1c, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Waitlist (online discussion forum) (W-ODF)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M/E (T)	W-ODF (C)	Mean difference (95% CI)	Quality
Outcome: Primary outcomes: Abdominal pain, tenderness, constipation (lower score is better): the GI symptom diary^a (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very serious^b^,^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 3.0; C = 5.2 MD = −2.20 (−3.33 to −1.07)	VERY LOW
Outcome: Total pain: the GI symptom diary^a (lower score is better): (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very serious^b^,^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 1.4; C = 1.6 MD = −1.00 (−1.66 to −0.34)	VERY LOW
Outcome: Constipation: the GI symptom diary^a(lower score is better): (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very serious^b^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 0.3; C = 0.6 MD = −0.40 (−0.62 to −0.18)	VERY LOW
Outcome: Diarrhoea: the GI symptom diary^a (lower score is better): (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very serious^b^,^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 0.4; C = 0.6 MD = −0.20 (−0.46 to 0.06)	VERY LOW
Outcome: Bloating: the GI symptom diary^a (lower score is better): (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very serious^b^,^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 0.9; C = 1.7 MD = −0.80 (−1.16 to −0.44)	VERY LOW
Outcome: Flatulence: the GI symptom diary^a (lower score is better): (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very Serious^b^,^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 0.9; C = 1.4 MD = −0.50 (−0.82 to −0.18)	VERY LOW
Outcome: Belching: the GI symptom diary^a (lower score is better): (mean diary rating) (10-wks)
Ljotsson 2010	RCT	Very serious^b^,^c	Serious^(e)	Not applicable	Serious^d	No serious	42	43	T = 0.4; C = 0.5 MD = −0.10 (−0.31 to 0.11)	VERY LOW

(a): The GI symptom diary (mean daily rating) (5-point scale: 0 = not a problem; 4 = debilitating). FDA and EMA suggested MID = 30% improvement = at least 1.5 points decrease from baseline. For the composite primary outcome, 30% improvement = at least 4.5 points decrease from baseline.
(b): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(c): Participants were ‘self-referred’, potential selection bias of participants who were more likely to experience an effect. Reason for withdrawal not reported.
(d): Both groups end scores (mean change from baseline) did not reach the ≥30% MID.
(e): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the

CCBT-Mindfulness/Exposure vs Internet delivered stress management

Table 114GRADE profile 2a, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Internet delivered stress management (ISM)

Quality assessment							No of patients		Effect estimate		Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M/E	ISM	Relative (96% CI)	Absolute	Quality
Outcome: IBS symptoms: Adequate relief^a (responder) (10-wks)
Ljotsson 2011	RCT	Very serious^b^,^c	Serious^(f)	Not applicable	Very serious^d	No serious	68/98 (69.4%)	56/97 (57.7%)	RR 1.20 (0.97 to 1.49)	12 more per 100 (from 2 fewer to 22 more)	VERY LOW
Outcome: IBS symptoms: Adequate relief^a (responder) (6-mth FU)
Ljotsson 2011	RCT	Very serious^b^,^c	Serious^(f)	Not applicable	Serious^e	No serious	64/98 (65.3%)	43/97 (44.3%)	RR 1.47 (1.13 to 1.92)	21 more per 100 (from 6 more to 41 more)	VERY LOW

(a): Adequate relief (responder) [Question: “In the past week, have you had adequate relief from IBS pain or discomfort?”]
(b): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(c): Participants were ‘self-referred’, potential selection bias of participants who were more likely to experience an effect. Reason for withdrawal not reported.
(d): The RR did not reach the MID and the 95%CI crosses over 1.25.
(e): The 95%CI crosses over 1.25.
(f): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the UK compared to Sweden

Table 115GRADE profile 2b, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Internet delivered stress management (ISM)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M/E (T)	ISM (C)	Mean difference (95% CI)	Quality
Outcome: Quality of life: IBS-QoL^a (10-wks)
Ljotsson 2011	RCT	Very serious^c^,^d	Serious⁽ⁱ⁾	Not applicable	No serious^e	No serious	97	94	T = 75.7; C = 65.7 MD = 10.00 (4.47 to 15.53)	LOW
Outcome: Quality of life: IBS-QoL^a (6-mth FU)
Ljotsson 2011	RCT	Very serious^c^,^d	Serious⁽ⁱ⁾	Not applicable	No serious^f	No serious	87	82	T = 74.9; C = 68.7 MD = 6.20 (0.20 to 12.20)	LOW
Outcome: IBS symptoms: GSRS-IBS^b (10-wks)
Ljotsson 2011	RCT	Very serious^c^,^d	Serious⁽ⁱ⁾	Not applicable	Serious^g	No serious	96	90	T = 36.3; C = 41.1 MD = −4.80 (−8.41 to −1.19)	VERY LOW
Outcome: IBS symptoms: GSRS-IBS^b (6-mth FU)
Ljotsson 2011	RCT	Very serious^c^,^d	Serious⁽ⁱ⁾	Not applicable	Serious^h	No serious	87	82	T = 33.4; C = 39.3 MD = −5.90 (−9.93 to −1.87)	VERY LOW

(a): IBS-QoL (total score: 0 to 100, with 0 = minimum QoL). Drossman (2007) suggested minimum clinical important difference as ≥14 points improvement from baseline.
(b): GSRS-IBS (Gastrointestinal Symptom Rating Scale for IBS) (total score: 13 to 91, with 13 = no discomfort at all). FDA and EMA suggested MID = 30% reduction = at least
(c): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(d): Participants were ‘self-referred’, potential selection bias of participants who were more likely to experience an effect. Reason for withdrawal not reported.
(e): Mean difference between groups showed significant effect, treatment group reached the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +18.6; C = +10.2], no downgrade.
(f): Mean difference between groups showed significant effect, treatment group reached the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +17.8; C = +13.2], no downgrade.
(g): Although mean difference between groups showed significant effect, both groups end scores did not reach the ≥30% MID [Mean change from baseline: T = −11.2; C = −6.2]
(h): Although mean difference between groups showed significant effect, both groups end scores did not reach the ≥30% MID [Mean change from baseline: T = −14.1; C = −8.0]
(i): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the UK compared to Sweden

Mindfulness group training vs Support group

Table 116GRADE profile 3a, Mindfulness group training (MG) vs Support group (SG)

Quality assessment							No of patients		Effect estimate		Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	MG	SG	Relative (96% CI)	Absolute	Quality
Outcome: IBS symptoms: IBS-SS Responder^a (at least 50 points reduction from baseline) (10-wks)
Gaylord 2011	RCT	Very serious^b^,^c	No serious	Not applicable	Serious^d	No serious	25/36 (69.4%)	18/39 (46.2%)	RR 1.50 (1.01 to 2.25)	23 more per 100 (from 0 more to 58 more)	VERY LOW
Outcome: IBS symptoms: IBS-SS Responder^b (at least 50 points reduction from baseline) (3-mth FU)
Gaylord 2011	RCT	Very serious^b^,^c	No serious	Not applicable	Serious^d	No serious	27/36 (75.0%)	21/39 (53.8%)	RR 1.39 (0.99 to 1.97)	21 more per 100 (from 1 fewer to 52 more)	VERY LOW

(a): IBS-SS (severity scale: maximum score = 500)
(b): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(c): Women only study. Reason for withdrawal not reported.
(d): The 95%CI crosses over 1.25.

Table 117GRADE profile 3b, Mindfulness group training (MG) vs Support group (SG)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	MG (T)	SG (C)	Mean difference (95% CI)	Quality
Outcome: Quality of life: IBS-QoL^a (10-wks)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^e	No serious	36	39	T = 74.99; C = 70.92 MD = 4.07 (−3.30 to 11.44)	VERY LOW
Outcome: Quality of life: IBS-QoL^a (3-mth FU)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^f	No serious	36	39	T = 76.73; C = 71.05 MD = 5.68 (−2.39 to 13.75)	VERY LOW
Outcome: IBS symptoms: IBS-SS Abdominal pain severity^b (10-wks)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	36	39	T = 35.00; C = 50.49 MD = −15.49 (−28.42 to −2.56)	VERY LOW
Outcome: IBS symptoms: IBS-SS Abdominal pain severity^b (3-mth FU)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	36	39	T = 31.11; C = 45.49 MD = −14.38 (−26.61 to −2.15)	VERY LOW
Outcome: IBS symptoms: IBS-SS Bloating severity^b (10-wks)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	36	39	T = 42.57; C = 49.22 MD = −6.65 (−19.84 to 6.54)	VERY LOW
Outcome: IBS symptoms: IBS-SS Bloating severity^b (3-mth FU)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	36	39	T = 37.46; C = 47.55 MD = −10.09 (−23.55 to 3.37)	VERY LOW
Outcome: IBS symptoms: IBS-SS Dissatisfaction with bowel habit^b (10-wks)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	36	39	T = 49.94; C = 65.15 MD = −15.21 (−28.27 to −2.15)	VERY LOW
Outcome: IBS symptoms: IBS-SS Dissatisfaction with bowel habit^b (3-mth FU)
Gaylord 2011	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	36	39	T = 45.69; C = 62.56 MD = −16.87 (−29.60 to −4.14)	VERY LOW

(a): IBS-QoL (total score: 0 to 100, with 0 = minimum QoL). Drossman (2007) suggested minimum clinical important difference as ≥14 points improvement from baseline.
(b): IBS-SS (severity scale for individual symptoms: maximum score = 100, with ≥30% MID = 30 points change from baseline)
(c): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(d): Women only study. Reason for withdrawal not reported.
(e): Mean difference between groups showed no significant effect, both groups end scores did not reach the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +10.19; C = +3.7]
(f): Mean difference between groups showed no significant effect, both groups end scores did not reach the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +11.93; C = +3.83]
(g): Mean change from baseline did not reach the MID of ≥30 points change.

Mindfulness-based stress reduction vs Treatment as usual

Table 118GRADE profile 4a, Mindfulness-based stress reduction (MBSR) vs Treatment as usual (TAU)

Quality assessment							No of patients		Effect estimate		Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	MBSR	TAU	Relative (96% CI)	Absolute	Quality
Outcome: IBS symptoms: IBS-SS Responder^a (at least 50 points reduction from baseline) (8-wks)
Zernicke 2012	RCT	Very serious^b^,^c	No serious	Not applicable	Very serious^e	No serious	10/43 (23.3%)	10/47 (21.3%)	RR 1.09 (0.50 to 2.37)	2 more per 100 (from 11 fewer to 29 more)	VERY LOW

(a): Zernicke (2012)
(b): IBS-SS (severity scale: maximum score = 500)
(c): Medication for IBS was allowed but no information was provided regarding usage between groups.
(d): No information on what consisted of the Treatment As Usual arm.
(e): The 95%CI crosses over both MIDs of 0.75 and 1.25.

Table 119GRADE profile 4b, Mindfulness-based stress reduction (MBSR) vs Treatment as usual (TAU)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	MBSR (T)	TAU (C)	Mean difference (95% CI)	Quality
Outcome: Quality of life: IBS-QoL (8-wks)
Zernicke 2012	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^e	No serious	43	47	T = 75.0; C = 63.1 MD = 11.90 (1.91 to 21.89)	VERY LOW
Outcome: Quality of life: IBS-QoL (6-mth FU)
Zernicke 2012	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^f	No serious	43	47	T = 74.3; C = 66.5 MD = 7.80 (−2.77 to 18.37)	VERY LOW
Outcome: IBS symptoms: IBS-SS (6-mth FU)
Zernicke 2012	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^g	No serious	43	47	T = 193.6; C = 213.8 MD = −20.20 (−71.57 to 31.17)	VERY LOW

(a): IBS-QoL (total score: 0 to 100, with 0 = minimum QoL). Drossman (2007) suggested minimum clinical important difference as ≥14 points improvement from baseline.
(b): IBS-SS (severity scale: maximum score = 500, with ≥30% MID = 150 points change from baseline)
(c): Medication for IBS was allowed but no information was provided regarding usage between groups.
(d): No information on what consisted of the Treatment As Usual arm.
(e): Although mean difference between groups showed significant effect, both groups end scores did not reach the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +9.7; C = +1.5]
(f): Mean difference between groups showed no significant effect, both groups end scores did not reach the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +9.0; C = +4.9]
(g): Mean change from baseline did not reach the MID of ≥150 points change (mean change from baseline: T = −55; C = −35.2)

CCBT-Exposure vs Waitlist control

Table 120GRADE profile 5, CCBT-Exposure (CCBT-E) vs Waitlist control (WC)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-E (T)	WC (C)	Mean difference (95% CI)	Quality
Outcome: Quality of life: IBS-QoL^a (6-wks)
Hunt 2009	RCT	Very serious^c^,^d	No serious	Not applicable	No serious^e	No serious	13	18	T = 84.0; C = 111.0 MD = −27.00 (−45.25 to −8.75)	LOW
Outcome: IBS symptoms: GSRS-IBS^b (6-wks)
Hunt 2009	RCT	Very serious^c^,^d	No serious	Not applicable	Serious^f	No serious	13	18	T = 35.0; C = 52.0 MD = −17.00 (−26.19 to −7.81)	VERY LOW

(a): IBS-QoL (only raw score reported, total score: 0 to 170, with 0 = minimum QoL). Drossman (2007) suggested minimum clinical important difference as ≥14 points improvement from baseline based on the 0 to 100 scale. For the raw score of 170, the calculated MID would be ≥23.8 points improvement from baseline.
(b): GSRS-IBS (Gastrointestinal Symptom Rating Scale for IBS) (total score: 13 to 91, with 13 = no discomfort at all). FDA and EMA suggested MID = 30% reduction = at least
(c): Participants were ‘self-reported’ as being diagnosed as having IBS by a medical professional. No information on exclusion criteria.
(d): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.). No information on what is the ‘waitlist control’ group.
(e): Mean difference between groups showed significant effect, treatment group end scores reached the ≥23.8 points from baseline MID [Mean change from baseline: T = −38; C = −12], no downgrade.
(f): Although mean difference between groups showed significant effect, both groups end scores did not reach the ≥23.8 points from baseline MID [Mean change from baseline: T = −22; C = −9]

CCBT-Mindfulness vs CCBT-Mindfulness/Exposure

Table 122GRADE profile 6a, CCBT-Mindfulness (CCBT-M) vs CCBT-Mindfulness/Exposure (CCBT-M/E)

Quality assessment							No of patients		Effect estimate	Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M (T)	CCBT-M/E (C)	Mean difference (95% CI)	Quality
Outcome: Quality of life: IBS-QoL^a (10-wks)
Ljotsson 2014	RCT	Serious^c	Serious^(h)	Not applicable	No serious^d	No serious	146	146	T = 73.6; C = 79.2 MD = −5.60 (−9.88 to −1.32)	MODERATE
Outcome: Quality of life: IBS-QoL^a (6-mth FU)
Ljotsson 2014	RCT	Serious^c	Serious^(h)	Not applicable	No serious^e	No serious	134	133	T = 76.5; C = 81.4 MD = −4.90 (−9.46 to −0.34)	MODERATE
Outcome: IBS symptoms: GSRS-IBS^b (10-wks)
Ljotsson 2014	RCT	Serious^c	Serious^(h)	Not applicable	Serious^f	No serious	146	146	T = 38.2; C = 31.8 MD = 6.40 (3.41 to 9.39)	LOW
Outcome: IBS symptoms: GSRS-IBS (6-mth FU)
Ljotsson 2014	RCT	Serious^c	Serious^(h)	Not applicable	Serious^g	No serious	135	134	T = 37.3; C = 32.2 MD = 5.10 (2.03 to 8.17)	LOW

(a): IBS-QoL (total score: 0 to 100, with 0 = minimum QoL). FDA and EMA suggested MID = 30% improvement = at least 30 points increase from baseline.
(b): GSRS-IBS (Gastrointestinal Symptom Rating Scale for IBS) (total score: 13 to 91, with 13 = no discomfort at all). FDA and EMA suggested MID = 30% reduction = at least
(c): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(d): Mean difference between groups showed significant effect, both groups end scores reached the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +16.1; C = +19.6], no downgrade.
(e): Mean difference between groups showed significant effect, both groups end scores reached the MID of ≥14 points improvement from baseline [Mean change from baseline: T = +19.0; C = +21.8], no downgrade.
(f): Although mean difference between groups showed significant effect, both groups end scores did not reach the ≥30% MID [Mean change from baseline: T = −9.3; C = −14.3]
(g): Although mean difference between groups showed significant effect, both groups end scores did not reach the ≥30% MID [Mean change from baseline: T = −10.2; C = − 13.9]
(h): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the UK compared to Sweden

Table 123GRADE profile 6b, CCBT-Mindfulness (CCBT-M) vs CCBT-Mindfulness/Exposure (CCBT-M/E)

Quality assessment							No of patients		Effect estimate		Quality
No of studies	Design	Risk of bias	Indirectness	Inconsistency	Imprecision	Other considerations	CCBT-M	CCBT-M/E	Relative (96% CI)	Absolute	Quality
Outcome: Adverse events (cluster)^a (10-wks)
Ljotsson 2014	RCT	Serious^b	Serious^(e)	Not applicable	Serious^c	No serious	19/145 (13.1%)	29/142 (20.4%)	RR 0.64 (0.38 to 1.09)	7 fewer per 100 (from 13 fewer to 2 more)	LOW
Outcome: Adverse events (cluster)^a (6-mth FU)
Ljotsson 2014	RCT	Serious^b	Serious^(e)	Not applicable	Serious^d	No serious	9/127 (7.1%)	3/131 (2.3%)	RR 3.09 (0.86 to 11.17)	5 more per 100 (from 0 fewer to 23 more)	LOW

(a): Adverse events (No. of participants reported) [Cluster of residual discomfort, worsening of symptoms, stress because of the study, depressed or anxious mood]
(b): No information on baseline use of other IBS treatments (e.g. pharmacological treatments, dietary interventions, etc.).
(c): The 95%CI crosses over the MID of .0.75
(d): The 95%CI crosses over the MID of 1.25.
(e): The study was undertaken in Sweden. There is uncertainty that the intervention would be applicable to the UK population due to the differences in delivery of CCBT in the UK compared to Sweden

Bookshelf ID: NBK550715

Contents

< Prev Next >

PubReader
Print View
Cite this Page
Addendum to NICE guideline CG61, Irritable bowel syndrome in adults: Diagnosis and management of irritable bowel syndrome in primary care. London: National Institute for Health and Care Excellence (NICE); 2015 Feb. (NICE guideline, No. CG61.1.) Appendix H, GRADE profiles.
PDF version of this title (3.6M)

In this Page

Review question 1 (antidepressants vs placebo)
Review question 2 (low FODMAP diet vs Standard diet)
Review question 3 (linaclotide)
Review question 4 (lubiprostone)
Review question 5a (relaxation therapy)
Review question 5b (CCBT and Mindfulness therapy)

Related NICE guidance and evidence

Recent Activity

Clear Turn Off Turn On

GRADE profiles - Addendum to NICE guideline CG61, Irritable bowel syndrome in ad...
GRADE profiles - Addendum to NICE guideline CG61, Irritable bowel syndrome in adults
Abbreviations - Mental Health Problems in People with Learning Disabilities: Pre...
Abbreviations - Mental Health Problems in People with Learning Disabilities: Prevention, Assessment and Management
cdc42 GTP-Binding Protein, Saccharomyces cerevisiae
cdc42 GTP-Binding Protein, Saccharomyces cerevisiae
A member of the Rho family of MONOMERIC GTP-BINDING PROTEINS from SACCHAROMYCES CEREVISIAE. It is involved in morphological events related to the cell cycle. This enzyme was f...<br/>Year introduced: 2002(2000)
MeSH

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

Bookshelf

Addendum to NICE guideline CG61, Irritable bowel syndrome in adults: Diagnosis and management of irritable bowel syndrome in primary care.

Appendix HGRADE profiles

H.1. Review question 1 (antidepressants vs placebo)

Table 68GRADE profile, successfully treated, abdominal pain

Table 69GRADE profile, scores on abdominal pain

Table 70GRADE profile, successfully treated, global assessment

Table 71GRADE profile, successfully treated, symptom score

Table 72GRADE profile, symptom scores

Table 73GRADE profiles, quality of life 1 (SF-36)

Table 74GRADE profiles, quality of life 2 (SF-36)

Table 75GRADE profiles, quality of life 3 (IBS-QOL scores)

Table 76GRADE profiles, quality of life 4 (IBS-QOL scores)

H.2. Review question 2 (low FODMAP diet vs Standard diet)

Table 77GRADE profile, Outcome: GI symptoms, overall response

Table 78GRADE profile, Outcome: bloating

Table 79GRADE profile, Outcome: abdominal pain

Table 80GRADE profile, Outcome: dissatisfaction with stool consistency

Table 81GRADE profile, Outcome: flatuence/wind

Table 82GRADE profile, Outcome: diarrhoea

Table 83GRADE profile, Outcome: constipation

H.3. Review question 3 (linaclotide)

Table 84GRADE Profile, Outcome: Quality of Life (QOL)

Table 85GRADE Profile, Outcome: symptoms

Table 86GRADE Profile, Outcome: stool score/general changes in bowel habit

Table 87GRADE Profile, Outcome: relapse or flatulence or bloating

Discontinuation, Safety and Adverse Events

Table 88GRADE profile - discontinuation (all reasons)

Table 89GRADE profile - reason for discountinuation

Table 90GRADE profile - adverse events

Table 91GRADE profile - serious adverse events

H.4. Review question 4 (lubiprostone)

Table 92GRADE profile, Outcome: Quality of Life (QOL)

Table 93GRADE Profile, Outcome: symptom severity

Table 94GRADE Profile, Outcome: abdominal pain (10 point scale) higher score = worse

Table 95GRADE Profile, Outcome: stool score/general changes in bowel habit

Table 96GRADE Profile, Outcome: bloating

Discontinuation and Adverse Events

Table 97GRADE profile, Outcome: discontinuation (all reasons)

Table 98GRADE profile, Outcome: discontinuation due to adverse event

Table 99GRADE profile, Outcome: discontinuation due to lack of efficacy

Table 100GRADE profile, Outcome: discontinuation due to non-compliance

Table 101GRADE profile, Outcome: discontinuation due to withdrawn consent

Table 102GRADE profile, Outcome: adverse event (at least one)

Table 103GRADE profile, Outcome: adverse event = nausea

Table 104GRADE profile, Outcome: adverse event = diarrhoea

Table 105GRADE profile, Outcome: adverse event = abdonimal distension

Table 106GRADE profile, Outcome: serious adverse eventsa

H.5. Review question 5a (relaxation therapy)

Table 107GRADE profile, Relaxation vs routine clinical care/ control/ enhanced medical care (dichotomous outomes)

Table 108GRADE profile, Relaxation vs routine clinical care/ control/ enhanced medical care

Table 109GRADE profile, Relaxation vs enhanced medical care

Table 110GRADE profile, Relaxation vs hypnotherapy

H.6. Review question 5b (CCBT and Mindfulness therapy)

Table 111GRADE profile 1a, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Waitlist (online discussion forum) (W-ODF)

Table 112GRADE profile 1b, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Waitlist (online discussion forum) (W-ODF)

Table 113GRADE profile 1c, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Waitlist (online discussion forum) (W-ODF)

CCBT-Mindfulness/Exposure vs Internet delivered stress management

Table 114GRADE profile 2a, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Internet delivered stress management (ISM)

Table 115GRADE profile 2b, CCBT-Mindfulness/Exposure (CCBT-M/E) vs Internet delivered stress management (ISM)

Mindfulness group training vs Support group

Table 116GRADE profile 3a, Mindfulness group training (MG) vs Support group (SG)

Table 117GRADE profile 3b, Mindfulness group training (MG) vs Support group (SG)

Mindfulness-based stress reduction vs Treatment as usual

Table 118GRADE profile 4a, Mindfulness-based stress reduction (MBSR) vs Treatment as usual (TAU)

Table 119GRADE profile 4b, Mindfulness-based stress reduction (MBSR) vs Treatment as usual (TAU)

CCBT-Exposure vs Waitlist control

Table 120GRADE profile 5, CCBT-Exposure (CCBT-E) vs Waitlist control (WC)

CCBT-Mindfulness vs CCBT-Mindfulness/Exposure

Table 122GRADE profile 6a, CCBT-Mindfulness (CCBT-M) vs CCBT-Mindfulness/Exposure (CCBT-M/E)

Table 123GRADE profile 6b, CCBT-Mindfulness (CCBT-M) vs CCBT-Mindfulness/Exposure (CCBT-M/E)

Views

In this Page

Related NICE guidance and evidence

Recent Activity

Table 106GRADE profile, Outcome: serious adverse events^a