A cost effectiveness analysis of goserelin compared with danazol as endometrial thinning agents

Objective To analyse the cost, effectiveness and cost effectiveness of two endometrial thinning agents prior to laser ablation for dysfunctional uteri ne bleeding: danazol and goserelin. Setting A district general hospital. Design A retrospective cost effectiveness analysis, from the perspective of the health service, based on data from an open, randomised, parallel group comparative study of 160 pre-menopausal women with dysfunctional uterine b leeding. Methods Within the trial, length of operation and duration of hospital stay was recorded for each woman. Resource use due to complications of surgery and adverse drug events was evaluated by one of the authors (R.G.). Additio nal surgery after completion of the study was collected using a postal ques tionnaire which was distributed to every woman who had undergone surgery. R esource use was costed using detailed unit coals from a specific NHS trust and from published sources. A cost effectiveness analysis was undertaken re lating differential cost to differential rates of amenorrhoea at women's la st point of follow up. Results Information on amenorrhoea was available from 138 women, of whom II I had completed the questionnaire to indicate longer term follow up. Women who did not complete the clinical trial were not included in this economic evaluation. On average, women randomised to goserelin spent less time in th eatre and on the ward. Based on longer term follow up, rates of retreatment were similar in the two groups. The mean (SD) health service cost of women in the goserelin group was pound 323.84 pound 309.94), compared with pound 243.45 (pound 265.23) in the danazol group; median (range) costs were poun d 220.29 (pound 191-pound 2127) and pound 159.76 (pound 140-pound 1426) in the two groups, respectively. These costs were significantly higher for gos erelin (P = 0.0001). The goserelin group also had a higher rate of amenorrh oea (38.8% vs 28.6%, P = 0.23). Based on mean differences in cost, the incr emental cost of goserelin per additional woman with amenorrhoea was pound 7 88; based on median differences in cost: the ratio was pound 590. Conclusions The shorter duration in theatre and stay in hospital provided a modest offset of the higher acquisition cost of goserelin, but the overall cost of management remained significantly higher than managing women with danazol. The rates of amenorrhoea indicated that goserelin was more effecti ve at 24 weeks and approximately two years after surgery, although statisti cal significance was only achieved at 24 weeks. The economic impact of wome n withdrawn from treatment was not considered, but sensitivity analysis ind icates that these women may have had a large effect on the overall result o f this study. Purchasers will need to decide whether the additional cost of management with goserelin is justified by the increased rates of amenorrho ea and the reduced withdrawals from treatment.