| Outcomes | Illustrative comparative risks (95% CI) | Relative effect (95 % CI) | No of Participants (studies) | Quality of the evidence (GRADE) | |
|---|---|---|---|---|---|
| Assumed risk | Corresponding risk | ||||
| Control | Post-surgical pharmacological therapy | ||||
| Pain recurrence (VAS) cm - Pelvic pain Follow-up: 12 months | Control group mean 6.2 (SD 0.9) | The mean pain recurrence (VAS) - pelvic pain in the intervention groups was 1.2 lower (1.47 to 0.93 lower) | MD −1.2 (1.47 to −0.93) | 187 (1 study) | ⊕⊕⊕⊝ Moderate1 |
| Pain recurrence (VAS) cm-Dysmenorrhoea Follow-up: 12 months | Control group mean 6.4 (SD 1.3) | The mean pain recurrence (VAS) - dysmenorrhoea in the intervention groups was 0.7 lower (1.04 to 0.36 lower) | MD −0.7 (1.04 to −0.36) | 187 (1 study) | ⊕⊕⊝⊝ Low1,2 |
| Pain recurrence (VAS) cm - Deep dyspareunia Follow-up: 12 months | Control group mean 4.8 (SD 1.2) | The mean pain recurrence (VAS) - deep dyspareunia in the intervention groups was 0.4 lower (0.76 to 0.04 lower) | MD −0.4 (0.76 to −0.04) | 187 (1 study) | ⊕⊝⊝⊝ Very low1,3 |
| Pain recurrence (questionnaire based) - Abdominal pain at 12 months post treatment completion | 569 per 1,000 | 404 per 1,000 (279 to 586) | RR 0.71 (0.49 to 1.03) | 120 (1 study) | ⊕⊕⊝⊝ Low2,4 |
| Pain recurrence (questionnaire based) - Dysmenorrhoea at 12 months post treatment completion | 346 per 1,000 | 301 per 1,000 (190 to 471) | RR 0.87 (0.55 to 1.36) | 158 (1 study) | ⊕⊝⊝⊝ Very low3,4 |
| Pain recurrence (questionnaire based) - Dyspareunia at 12 months post treatment completion | 304 per 1,000 | 161 per 1,000 (85 to 301) | RR 0.53 (0.28 to 0.99) | 144 (1 study) | ⊕⊕⊝⊝ Low2,4 |
| Pain recurrence (Andersch and Milsom) - Pelvic pain Follow-up: 12 months | Control group mean 4 (SD 3.6) | The mean pain recurrence (Andersch and Milsom) - pelvic pain in the intervention groups was 0.4 lower (2.15 lower to 1.35 higher) | MD −0.4 (2.15 to 1.35) | 53 (1 study) | ⊕⊕⊝⊝ Low3 |
| Pain recurrence (dichotomous) Follow-up: 12 months | 216 per 1,000 | 168 per 1,000 (119 to 241) | RR 0.78 (0.55 to 1.12) | 476 (4 studies) | ⊕⊝⊝⊝ Very low2,5 |
| Pain recurrence (dichotomous) Follow-up: 13–24 months | 286 per 1,000 | 200 per 1,000 (134 to 294) | RR 0.7 (0.47 to 1.03) | 312 (3 studies) | ⊕⊝⊝⊝ Very low2,6 |
| Pain recurrence (dichotomous) Follow-up: 60 months | 480 per 1,000 | 446 per 1,000 (254 to 797) | RR 0.93 (0.53 to 1.66) | 54 (1 study) | ⊕⊝⊝⊝ Very low3,7 |
| Dysmenorrhoea Follow-up: 12 months | 383 per 1,000 | 84 per 1,000 (31 to 230) | RR 0.22 (0.08 to 0.6) | 95 (2 studies) | ⊕⊕⊕⊝ Moderate8 |
| Reoperation (women with endometriosis) | 30 per 1,000 | 35 per 1,000 (12 to 101) | RR 1.17 (0.4 to 3.4) | 327 (3 studies) | ⊕⊝⊝⊝ Very low3,9 |
| Endometriosis recurrence (dichotomous) - Disease recurrence at 5–6 months Follow-up: 5–6 months | 401 per 1,000 | 397 per 1,000 (301 to 530) | RR 0.99 (0.75 to 1.32) | 285 (1 study) | ⊕⊝⊝⊝ Very low3,4 |
| Endometriosis recurrence (dichotomous) Follow-up: 12 months | 70 per 1,000 | 101 per 1,000 (20 to 515) | RR 1.44 (0.28 to 7.36) | 310 (3 studies) | ⊕⊝⊝⊝ Very low3,10,11 |
| Endometriosis recurrence (dichotomous) Follow-up: 24 months | 133 per 1,000 | 29 per 1,000 (1 to 500) | RR 0.22 (0.01 to 3.75) | 45 (1 study) | ⊕⊝⊝⊝ Very low3,12 |
| Endometrioma recurrence (dichotomous) - Recurrence at 13–36 months | 189 per 1,000 | 104 per 1,000 (68 to 163) | RR 0.55 (0.36 to 0.86) | 463 (3 studies) | ⊕⊕⊝⊝ Low2,13,14 |
| Endometrioma recurrence (dichotomous) Follow-up: 60 months | 125 per 1,000 | 210 per 1,000 (44 to 1,000) | RR 1.68 (0.35 to 8.03) | 35 (1 study) | ⊕⊕⊝⊝ Low3,7 |
| Patient Satisfaction | Not estimable | Not estimable | RR 1.21 (0.80 to 1.82) | 95 (2 studies) | See comment |
Blinding: unclear risk. Placebo is not described and seems unlikely that blinding could be maintained when the interventions are depot and oral hormonal treatments
95% Confidence Interval crosses 1imprecision threshold
95% Confidence Interval crosses 2 imprecision thresholds
Randomisation, Allocation concealment: unclear risk. No information provided. Blinding: High risk. No placebo used
Allocation concealment: unclear risk. Not mentioned in Alborzi 2011, Loverro 2001 or Bianchi 1999. Blinding: high risk. No placebo used in Alborzi 2011, Bianchi 1999 or Vercellini 1999. Incomplete data reporting: unclear risk. 22% withdrawal overall in Vercellini 1999 due to reasons other than symptom recurrence or major protocol violations (similar in each group). 18% withdrawal overall in Alborzi 2011 after randomisation due to “poor patients follow up” with reasons not reported and unequal loss across groups (11/58 letrozole group, 18/58 dipherelin group and 1/59 no treatment group)
Allocation concealment: unclear risk. Not mentioned in Busacca 2001 or Muzii 2000. Blinding: high risk. No placebo use in Busacca 2001, Muzii 2000 or Vercellini1999. Incomplete data reporting: unclear risk. 22% withdrawal overall in Vercellini 1999 due to reasons other than symptom recurrence or major protocol violations (similar in each group). Other bias: unclear risk. No baseline characteristics reported in Muzii 2000
Allocation concealment: unclear risk. Not mentioned.
Blinding: unclear/high risk of performance bias. Unclear how patients were blinded to IUD presence in Tanmahasamut 2012 and Vercellini 2003 reported as an open label study with outcome assessors not blinded to treatment group (high risk of detection bias).
Allocation concealment: unclear risk. Not mentioned in Bianchi 1999, Busacca 2001 or Sesti 2009. Blinding: high risk. No placebo use in Bianchi 1999 or Busacca 2001.
Allocation concealment: unclear risk. Not mentioned in Alborzi 2011, Bianchi 1999 or Busacca 2001. Blinding: high risk. No placebo used in Alborzi 2011, Bianchi 1999 or Busacca 2001. Incomplete data reporting: unclear risk. 18% withdrawal overall in Alborzi 2011 after randomisation due to “poor patients follow up” with reasons not reported and unequal loss across groups (11/58 letrozole group, 18/58 dipherelin group and 1/59 no treatment group)
Using random effects model. Heterogeneity: Chi² = 5.72, df = 2 (P = 0.06); I² = 65%. Removal of Alborzi 2011 (RR = 16.48 95%CI 0.99 - 272.92) from the pooled analysis removes inconsistency (Heterogeneity: Chi² = 0.38, df = 1 (P = 0.54); I² = 0%) and the pooled fixed effects result for Bianchi 1999 and Busacca 2001 becomes RR = 0.76 (95%CI 0.30 - 1.90)
Blinding: high risk. No placebo used. Incomplete data reporting: high risk. 4/15 (27%) loss to follow up in treatment group in Tsai 2004.
Allocation concealment: unclear risk. Not mentioned in Muzii 2000 or Sesti 2009. Blinding: unclear risk - no placebo use in Muzii 2000 or in Seracchioli 2010 (although outcome assessors were blinded to treatment group. Incomplete data reporting: unclear risk. 8% withdrawal overall in relevant treatment arms in Sesti 2009. Other bias: unclear risk. No baseline characteristics reported in Muzii 2000
Using fixed effects model Heterogeneity: Chi² = 3.25, df = 2 (P = 0.20); I² = 39%
From: 11, Management strategies
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.