Discussion
The present study was designed to compare, using a cost-effectiveness analysis, the cost and the effectiveness of four surgical strategies used for the treatment of AUB-O,E,N in France. It showed that mini-invasive 2G and 1G procedures, which are recommended as preferred solutions for the surgical treatment can also be considered as cost-effective procedures, as they are associated with lower severe complication rates and lower costs, when compared to curettage and hysterectomy. Conversely, hysterectomy is the most expensive procedure, because of the high cost of the initial surgery and the higher rate of hospitalization for complications over time; curettage is the least effective one with the highest rate of failures over time.
Our results are in accordance with previous cost-effectiveness studies, confirming that 2G and 1G are cost-effective. Our study is the first and the only one for which cost and effectiveness data are entirely extracted from a unique source of data, i.e. the PMSI database. Roberts15 and Garside16 both point out that theirs models included hypothetical patients and they built them with multiple sources and limited available data. Miller18 also regrets the incorporation in his model of data from older studies, some of which predate the advent 2G endometrial ablation technologies. Only Fernandez13, Bischoff-Everding17 and Cooper14performed their cost analysis using economic data collected on individual patients. These studies are however based on small sample of patients (147, 88 and 660 patients, respectively), which may not be representative of AUB surgical management. Our study has the advantage of being based on observed cost and efficiency data from patients in real-life conditions, in the largest and exhaustive cohort of French patients, over a long observation period. The analysis with KM curves allows considering the time of occurrence of events since the initial surgery in patients with heterogeneous follow-up time. Indeed, the sensitivity analysis based on patients with 60 months of follow-up confirmed the conclusions at 18 months, demonstrating the robustness of our results.
The use of the PMSI database for performing a cost-effectiveness analysis presents other advantages. Since it is the basis of prospective hospital funding, hospitals have to produce standardized reports for each stay performed. Data extracted from the PMSI then presented a high exhaustiveness (all public and private hospitals are included and no sampling is done) and a high level of quality, with limited coding errors. The large number of cases documented allowed the study outcomes to be determined with high precision. Bias due to sampling errors and to loss to follow-up should be minimal. The economic burden of AUB is also expected to be well documented, as the management of the initial surgery and of severe complications occurred in hospital setting. At the time of study completion, hospital external activity was not available in the PMSI database: activity such as consultations or some procedures could not be tracked. Therefore, failure and complications that would have occurred out of hospital and that would not have led to hospitalization shall not have been tracked as they are not reported in the PMSI database. We might have an overestimation of failure/complications avoided. A linkage is currently available between the PMSI and SNIIRAM (French sick funds comprehensive reimbursement database). However, SNIIRAM could not be used in this study, as its access was not possible within the study delay. Other costs could therefore not be tracked, such as sick leaves, leading to potential underestimation of associated costs. However, the overestimation of effectiveness as well as the underestimation of costs might not be different between the 4 groups.
1G and 2G appear to be the preferred techniques. The cost of the device could be offset by its greater use with a logical reduction in the purchase cost and by the economy of occupying operating theaters because performing 2G surgery is twice as fast as for 1G.
In conclusion, this study based on the largest cohort of patients surgically treated for AUB-O,E,N confirms that 1G and 2G techniques are the most efficient strategies in real-life conditions. Despite the fact that hysterectomy and curettage are not recommended as first line strategies, that curettage is the least effective strategy and hysterectomy the most expensive one, both strategies are still used in more than half patients as first line surgical treatment. Switching to 2G techniques would not require a specific training for surgeons, when compared to curettage or hysterectomy, and it would decrease the annual AUB budget impact of about 65 millions \euro (as the extra-cost of hysterectomy compared to 2G is 24,008 \euro and 3000 annual hysterectomies are done in France with 90% of which are not justified for this indication). Improving the 2G French tariffs could foster this switch, allowing a more adapted treatment to these patients, with a limited expected increase of costs for the National Health Insurance. Moreover, most second-generation endometrial ablation are still performed under general or epidural anaesthetic in an operating theatre. However, the new small diameter devices should encourage movement of minimal invasive surgery out of the traditional operating theatre and this next development will accentuate the cost-effectiveness of the treatment of heavy menstrual bleedings by 2G techniques.
Finally, we are at a turning point where we have to balance the superior efficiency of radical surgeries (total or subtotal hysterectomy) and the lesser efficiency of 2G or 1G techniques, but despite all, that being significant with less than 20% reoperations, and their economic impact for national health insurance.
Disclosure of interests : GD, JF, DT, PD, HF and IB received personal fees from Hologic for their participation to Steering Committee during the work. PD, HF and DT reports personal fees from HOLOGIC France, outside the submitted work. HEVA works with all pharmaceutical and medical devices industries like JANSSEN, BMS, PFIZER, NOVARTIS, BSCI. JF reports personal fees from HEVA during the conduct of the study and personal fees from Astellas Pharma outside the submitted work. IB received personnel fees from Novartis, Allergan, CSL-Berhing, Merck outside the submitted work.
Contributions of authorship : This study was initiated by LdL, LL, IB and supported by Hologic. IB, LDL and LL elaborated the protocol. TL and GC performed the PMSI data extraction, their data-management and their statistical analysis under the control of LL and LDL. IB performed the cost-effectiveness analysis. GD, HF, DT, PD and JF were members of the Steering Committee which oversaw the implementation of the study and contributed to the interpretation of the study. The authors were fully responsible for all content and editorial decisions. All the authors made substantial contributions during all the stages of the work, including the conception and design of the work, acquisition, analysis and interpretation of data, manuscript development, revision of the manuscript and have approved its final version.
Details of ethics approval : The study was conducted in accordance with relevant international and French regulatory requirements. Since this was a retrospective study of an anonymized database and had no influence on patient care, ethics committee approval was not required. Use of the PMSI-MCO database for this type of study has been approved by the French national data protection agency (CNIL; annual authorization #1419102 v7 – 2015-111111-56-18 / order M14N056 and M14L056).
Funding : The study was funded by Hologic (no grant number), a company that markets NovaSure®, a radiofrequency endometrial ablation device. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.