Variation in outcome reporting in endometriosis trials: a systematic review.

OBJECTIVE: We reviewed the outcomes and outcome measures reported in randomized controlled trials and their relationship with methodological quality, year of publication, commercial funding, and journal impact factor.

DATA SOURCES: We searched the following sources: (1) Cochrane Central Register of Controlled Trials, (2) Embase, and (3) MEDLINE from inception to November 2014.

STUDY ELIGIBILITY: We included all randomized controlled trials evaluating a surgical intervention with or without a medical adjuvant therapy for the treatment of endometriosis symptoms.

STUDY DESIGN: Two authors independently selected trials, assessed methodological quality (Jadad score; range, 1-5), outcome reporting quality (Management of Otitis Media with Effusion in Cleft Palate criteria; range, 1-6), year of publication, impact factor in the year of publication, and commercial funding (yes or no). Univariate and bivariate analyses were performed using Spearman Rh and Mann-Whitney U tests. We used a multivariate linear regression model to assess relationship associations between outcome reporting quality and other variables.

RESULTS: There were 54 randomized controlled trials (5427 participants), which reported 164 outcomes and 113 outcome measures. The 3 most commonly reported primary outcomes were dysmenorrhea (10 outcome measures; 23 trials), dyspareunia (11 outcome measures; 21 trials), and pregnancy (3 outcome measures; 26 trials). The median quality of outcome reporting was 3 (interquartile range 4-2) and methodological quality 3 (interquartile range 5-2). Multivariate linear regression demonstrated a relationship between outcome reporting quality with methodological quality (β = 0.325; P = .038) and year of publication (β = 0.067; P = .040). No relationship was demonstrated between outcome reporting quality with journal impact factor (Rho = 0.190; P = .212) or commercial funding (P = .370).

CONCLUSION: Variation in outcome reporting within published endometriosis trials prohibits comparison, combination, and synthesis of data. This limits the usefulness of research to inform clinical practice, enhance patient care, and improve patient outcomes. In the absence of a core outcome set for endometriosis we recommend the use of the 3 most common pain (dysmenorrhea, dyspareunia, and pelvic pain) and subfertility (pregnancy, miscarriage, and live birth) outcomes. International consensus among stakeholders is needed to establish a core outcome set for endometriosis trials.