Significant changes have been made since this review was first published in 1999 by Andrew Prentice. However the main objective has remained the same: to determine the effectiveness and safety of GnRHas in the treatment of the painful symptoms associated with endometriosis.

The review published in 1999 stated under 'Type of Participants' that "the diagnosis of endometriosis was made by direct visualisation (laparoscopy). Trials where the diagnosis had been made by history alone or by some other imaging technique would have been considered...". This was modified so that "the clinical diagnosis of endometriosis had to be made by direct visualisation (laparoscopy)" only. Trials where the diagnosis was made by techniques other than direct visualisation were excluded. Trials where GnRHa is administered in post‐surgical participants as adjuvant therapy was also specifically stated to be excluded in this current review.

Numerous modifications have been made under 'Type of Interventions'. The review published in 1999 compared GnRHa, any dosage or route of administration, with no treatment, placebo, danazol, gestrinone, progestogens, combined oral contraceptive pill, surgical ablation of endometriotic deposits, surgical treatments that purport to interrupt neural pathways (e.g. LUNA), combination of GnRHas and hormone replacement therapy, and another GnRHa. Treatments designed only to achieve relief of symptoms such as treatment with non‐steroidal anti‐inflammatory drugs or other analgesics were not considered. The current review has removed GnRHas comparisons with gestrinone, progestogens (Prentice 2000), combined oral contraceptive pill (Davis 2007), and combination of GnRHas and hormone replacement therapy as they are described under separate reviews. The current review limited comparisons of GnRHas with other medical therapies only and excluded comparisons with any surgical intervention (Jacobson 2009). Since the main objective of the review was to look at the effectiveness and safety of GnRHas in treatment of endometriosis‐associated painful symptoms, trials that compared GnRHas with other analgesics would have been considered but no trials were identified. The current review also considered trials which compared GnRHas with the relatively new LNG IUS but excluded trials that compared GnRHas with GnRH antagonists as that is a registered title of a review to be conducted by the Menstrual Disorders and Subfertility Group of Cochrane Collaboration.Trials that compared one type of GnRHa with another were excluded as that would not have contributed towards the objective, instead trials which compared different dosages, length of treatment, routes of administration, and treatment regimes of GnRHas were considered.

Outcomes of pain relief, adverse effects and resolution of endometriotic implants were considered in both reviews. Quality of life and the additional use of analgesics were additional outcomes that were considered in the current review. Cost‐effectiveness was specifically stated as an outcome not considered in the current review.

Risk of bias. Funnel plot to be conducted if eight or more studies included has been altered to 10 or more studies.