Endometriosis affects 6–10% of reproductive-aged women and is associated with significant morbidity.1–3 Clinical manifestations are broad and can include dysmenorrhea, nonmenstrual pelvic pain, dyspareunia, dyschezia, and infertility.4–6 The wide range of presentations hinder the ability to make a timely diagnosis; the average time to surgical diagnosis is nearly 7 years from onset of symptoms.3,4,6–8 This delay can lead to chronic pain, reduced quality of life, infertility, and significant cost burden to the individual and health system.3,4,8,9
Multiple strategies exist for the treatment of endometriosis-related dysmenorrhea, which is defined as cyclic pelvic pain occurring during the menstrual cycle.2,4 Medical therapies include nonsteroidal antiinflammatory drugs (NSAIDs) and hormonal agents such as short-acting reversible contraception, levonorgestrel intrauterine device, and gonadotropin-releasing hormone (GnRH) modulators. Surgery is frequently performed to confirm the diagnosis and treat endometriosis implants; however, there is not a standardized surgical technique.2,10 The optimal surgical strategy for effective treatment of endometriosis is unknown as there is a paucity of conflicting data.2,4,11 Surgical interventions can range from excision or ablation of endometriosis, nerve sparing peritoneal stripping, excision of deep infiltrating nodules, bowel resection, and hysterectomy with or without bilateral salpingo-oophorectomy.2,4,11
In addition to health consequences, endometriosis is associated with significant cost. In 2009, the medical costs associated with endometriosis in the United States were estimated at $69.4 billion annually.12 Despite the recognized cost burden of this disease, cost-effectiveness data on the various treatment strategies is limited. Previous studies have investigated the direct and indirect costs regarding endometriosis; however, there are no prior studies that evaluate the cost effectiveness of a stepwise regimen to guide management.3,7,12–15 We sought to determine the cost effectiveness of sequential therapy for the treatment of endometriosis-related dysmenorrhea.
METHODS
A cost-effectiveness model was constructed to compare four stepwise strategies in the management of dysmenorrhea (Fig. 1). We adopted a health care payor perspective. Each strategy trialed a series of different medical therapies before offering surgery compared with proceeding directly to surgical management. We compared the following strategies: strategy 1) NSAIDs followed by surgery if there was no improvement; strategy 2) NSAIDs, then a short-acting reversible contraceptive or a long-acting reversible contraceptive (LARC) followed by surgery if no improvement; strategy 3) NSAIDs, then a short-acting reversible contraceptive or LARC, then a LARC or a GnRH agonist or antagonist, followed by surgery if no improvement; strategy 4) proceeding directly to surgery without attempting medical management first. Surgical management included laparoscopic ablation or excision of endometriosis, resection of deep infiltrating endometriosis and resection of endometrioma by cystectomy. We assumed a 6-month trial of each medication with progression to the next treatment in the series if no symptom improvement occurred. A 6-month trial was used as the majority of the studies used to obtain probability data used a 6-month time period to determine improvement in dysmenorrhea. Outcomes were modeled over a 3-year time horizon. Our study population was reproductive-aged women with dysmenorrhea secondary to endometriosis. Our theoretical cohort size was 4,817,894 women, which was calculated using an 8% prevalence of endometriosis among women aged 18–45 in the United States.1–3,16
;)
Fig. 1.: Decision tree for comparing four stepwise strategies in the management of endometriosis-related pain. [+] indicates collapsed branches that lead to identical branches as the ones that are displayed. Medical treatment 1: nonsteroidal antiinflammatory drugs (NSAIDs). Medical treatment 2: NSAIDs, then short-acting reversible contraceptives (SARCs) or long-acting reversible contraceptives (LARCs). Medical treatment 3: NSAIDs, then a SARC or LARC, then a LARC or GnRH modulator.Bohn. Obstet Gynecol 2021.
All model inputs were derived from the literature (Table 1). Our primary outcome of interest was the incremental cost-effectiveness ratios of competing strategies. Secondary outcomes included costs and quality-adjusted life years (QALYs). Our model was built using TreeAge Pro Software. This study was exempt from Institutional Review Board approval at Oregon Health & Science University.
Table 1.: Cost-Effectiveness Model Input Parameters
We created a stepwise strategy based on how endometriosis is routinely treated clinically.2,5,15 Within each treatment group, we chose one medication that is commonly available, and had available data regarding improvement of dysmenorrhea and cost information.14,17–25 For short-acting reversible contraceptives, model inputs were based on combined hormonal contraceptives and for LARCs, model inputs were based on the 52-mg levonorgestrel intrauterine device.18,19,25 For GnRH agonists, data were based on leuprolide acetate.20,23,25 For GnRH antagonists, data were based on elagolix.21,23,25 Previously published probabilities specific to dysmenorrhea improvement in women with laparoscopically diagnosed endometriosis were used for each method (Table 1). The probability of a woman being treated with a short-acting reversible contraceptive compared with LARC, LARC compared with GnRH agonist, or GnRH agonist compared with GnRH antagonist as seen in strategy 2 and strategy 3 in the model (Fig. 1) was based on expert opinion.
Costs were derived from the literature and inflated to 2019 dollars using the medical component of the Consumer Price Index.26 Costs for all of the medical treatments were for 6 months of treatment and obtained from Red Book Online.25 The cost of new and return office visits to evaluate the efficacy of the medical treatments was included in the model.27 The cost for laparoscopic surgery included a 12-month postsurgical period that accounted for inpatient and emergency room admissions, follow-up office visits, and prescriptions.14 A 12-month time period was used after surgery as the cost studies used were based on a 12-month postindex period.
Quality-adjusted life years are a standard measurement used in health economics to capture the effect on quality of life of different health conditions. They are a standard metric by which cost effectiveness of competing health interventions is determined.28 They are the product of the health utility of a condition, and the length of time by which a person is affected by the disease. Utilities are a measure of health preferences, and capture the decrease in health experienced with an array of conditions on a scale from 1 to 0. A utility of 1 represents “perfect health,” and 0 is equivalent to death.28,29 In our model, utilities were applied to measure women's symptoms secondary to endometriosis over a 3-year time horizon and discounted at a rate of 3%.30 The utilities for mild or no endometriosis symptoms and severe endometriosis symptoms were obtained from the literature (Table 1).
We calculated the incremental cost-effectiveness ratios of each medical strategy (strategies one through three) and compared them with the strategy of proceeding directly with surgical management to compare the difference in costs and effectiveness of each pathway. The threshold for cost effectiveness was set at a standard willingness-to-pay threshold of $100,000 per QALY; a cost less than $100,000 per each additional QALY gained is cost effective.31,32
Sensitivity analysis was performed broadly to test the robustness of the model. We performed univariate sensitivity analysis on all probabilities, costs, and utilities. A tornado diagram was created to identify the most influential variables (Appendix 1, available online at https://links.lww.com/AOG/C425). For the five most influential variables, we expanded their range from one half to twice the base estimate to evaluate whether a threshold value exists. A threshold value indicates that the model's findings would change if the value is exceeded.28 Bivariate sensitivity analysis was performed on a priori selected variables of clinical significance (eg, effectiveness of surgery, costs of GnRH modulators).33
Multivariable sensitivity analysis was performed using a Monte Carlo simulation with 10,000 iterations. For each iteration of a Monte Carlo simulation, the model is run with different inputs for each variable, sampled randomly from the distribution for each value. To achieve this multivariate sensitivity analysis, the probability and cost inputs can be varied simultaneously by sampling their distributions.29,33 Because neither probabilities nor costs can be normally distributed, beta distributions were used for probabilities and gamma distributions for costs.32
RESULTS
In a theoretical model among 4,817,894 women with endometriosis-related dysmenorrhea, Strategies 1, 2, and 3 were cost effective at standard willingness-to-pay thresholds of $100,000 per QALY gained when compared with surgery alone (strategy 4) (Table 2). When we examined costs of each strategy, strategy 4 was the costliest and strategy 2 the least. Regarding QALYS, we found that Strategies 1, 2, and 3 (sequential medical then surgical management) resulted in at least one million higher QALYs than strategy 4 (immediate surgery). Therefore, strategy 4 (proceeding directly to surgery), had the highest cost with lowest QALYs. This indicates that surgery, without a trial of any medications first, is inferior to the other three strategies as a standard, first-line approach.
Table 2.: Outcomes, Costs, Cost per Quality-Adjusted Life Year, and Incremental Cost-Effectiveness Ratio in a Hypothetical Cohort of 4,817,894 Women With Endometriosis-Related Dysmenorrhea
We examined all possible combinations of comparison between the different strategies, with the following notable outcome between Strategies 1 and 4. Despite NSAIDs being the least expensive medical management, strategy 1 (trial of one medication, NSAIDs) was associated with the second highest cost of all the strategies due to the large percentage of women that fail therapy with NSAIDs and proceed to surgery. Next, we compared outcomes between Strategies 2 and 3, as they were the most similar strategies. Strategy 2 (trial of two medical therapies before surgery) was more cost effective than strategy 3, which included a trial of a third medication. Strategy 3 (trialing three medications), had a modest increase in cost, with a limited gain in QALYs. Although the incremental cost-effectiveness ratio was below the cost-effectiveness threshold, it did not yield much of a comparative advantage. If all women in our theoretical cohort who failed a second medication received a third medication (comparing Strategies 2 and 3) before moving on to surgery, this would result in a cost of $257 million dollars.
Sensitivity analysis demonstrated our findings are robust. Tornado diagrams were made to compare two different strategies at a time. We found that the model was most sensitive to changes in the cost and probability of dysmenorrhea improvement by surgery and the cost of GnRH agonists (Appendix 2, available online at https://links.lww.com/AOG/C425). We then further examined these variables using multi-variable sensitivity analysis. Regarding the probability of improvement in dysmenorrhea with surgery, each strategy was compared with strategy 4 (Appendix 3, available online at https://links.lww.com/AOG/C425). For surgical management to be the preferred first-line approach, the probability of improvement after surgery would need to exceed 83%.
As the cost of GnRH agonists also affected the model, we examined Strategies 2 and 3 with sensitivity analyses as the difference in these strategies is the addition of a third medication, often a GnRH modulator. With regard to GnRH modulators, the addition of these medications was cost effective and resulted in 190,000 more QALYs than strategy 2. The model was sensitive to slight changes in cost, however. In univariate sensitivity analysis, the addition of a GnRH agonist (the most expensive medical management) became the dominant strategy at an input value of $7,408 or less, at which point strategy 3 not only resulted in more QALYs but was cheaper (Fig. 2). Therefore, if the cost of GnRH agonist decreases, strategy 3 would become more cost effective than strategy 2.
Fig. 2.: Univariate sensitivity analysis regarding the cost of gonadotropin-releasing hormone (GnRH) agonist between strategy 2 and strategy 3. The vertical line displays the incremental cost-effectiveness ratio. The horizontal axis shows the cost of the GnRH agonist (leuprolide acetate). This figure demonstrates that strategy 3 goes from being the dominant strategy to being cost effective only when the cost of the GnRH agonist exceeds $7,400. Medical treatment 2: nonsteroidal antiinflammatory drugs (NSAIDs), then short-acting reversible contraceptives (SARCs) or long-acting reversible contraceptives (LARCs). Medical treatment 3: NSAIDs, then a SARC or LARC, then a LARC or GnRH modulator.Bohn. Managing Endometriosis-Related Dysmenorrhea. Obstet Gynecol 2021.
Monte Carlo analysis demonstrated our findings were robust. We examined combinations of comparison between the different strategies, focusing on strategy 2 compared with strategy 3 as they are the closest in cost and QALYs. Figure 3 demonstrates the Monte Carlo simulation results for strategy 2 (NSAID, short-acting reversible contraceptive or LARC, surgery) compared with strategy 3 (NSAID, short-acting reversible contraceptive or LARC, LARC or trial of GnRH modulators, surgery). Regardless of model variation, both strategies are cost effective.
Fig. 3.: Multivariate sensitivity analysis of strategy 3 vs strategy 2. This figure displays the outcomes of the 10,000 trials from the Monte Carlo simulation. The willingness-to-pay (WTP) line indicates a willingness-to-pay threshold set to $100,000. Each dot is the result of a single trial.Bohn. Managing Endometriosis-Related Dysmenorrhea. Obstet Gynecol 2021.
DISCUSSION
Our study adds a different perspective to the literature with regard to clinical management of endometriosis. Endometriosis treatment remains one of the most controversial topics in gynecology, due in part to a lack of knowledge about the pathogenesis and natural history of endometriosis, as well as limited high-quality data to guide individual care.34 Medical management is often used to treat the symptom of dysmenorrhea, but surgery is necessary to definitely diagnose the disease, and can also improve symptoms.10,35 Our study found that all sequential medical then surgical treatment pathways are cost effective in the treatment of endometriosis when compared with proceeding immediately to surgery.
We found that, although cost effective, requiring trial of a third medication offered little comparative advantage before proceeding directly to surgery after the second therapy fails. Yet, for the woman who is anxious about surgical intervention, or when a prolonged wait for a surgical specialist occurs, trial of a GnRH modulator may be worthwhile. However, if all women in our cohort who failed a second medication were required to trial a third medication before being offered surgery, this would result in an increased cost of $257 million dollars.
The decision of when to surgically intervene is an important one, and should be informed by the needs of the woman as well as surgical resources. At baseline model inputs, proceeding directly to surgery is the most costly strategy, and thus would not be a preferred, routine first-line approach. It is important to note that a wide range of surgical techniques are used in the treatment of endometriosis; thus, data on the effectiveness of surgical intervention are heterogenous.22 The probability of symptom improvement may be directly correlated to surgical technique and surgeon expertise.36 We found a key threshold value for surgical effectiveness: when improvement after surgery exceeds 83%, meaning 83% of women who underwent surgery have improvement in dysmenorrhea after 6 months, proceeding directly to surgery would become the preferred, routine first-line approach. Data on the effectiveness of different surgical techniques, collected using a standardized scale, are urgently needed to better inform endometriosis management. In addition to a need for identifying effective standardized techniques, these data suggest that surgery may be the more cost effective, or even dominant, strategy among women deemed good surgical candidates, predicted to have more than 80% chance of improvement. Additionally, we recognize care may need to be individualized to the woman's unique preferences and needs. The majority of women have endured years of pain and some may not wish to delay diagnosis through obligatory trials of medications.
Our study should be interpreted with the following limitations in mind. Cost-effectiveness analyses are limited by assumptions made and estimates published in the literature. Data on endometriosis outcomes are limited and challenging to apply. Outcome data are heterogenous, in particular with regard to surgical management. To mitigate bias, we performed extensive sensitivity analysis. This model does not account for side effects of medical management that may influence the acceptability of the strategy and lead a patient to elect early discontinuation. It was assumed that all therapies would be trialed for 6 months to allow adequate time to determine treatment failure; however, many women will not tolerate side effects for that duration of time or may find it to be an unacceptable length of time to trial a medication. Furthermore, this stepwise approach does not account for complications from surgery (gynecologic surgery in women with endometriosis is associated with higher complication rates) or recurrent dysmenorrhea after treatment.36,37 Lastly, this model cannot account for individual physical exam and radiology findings that would be used to guide clinical care, nor can it be applied to women seeking fertility, as hormonal contraceptives are counterproductive to this goal.
Endometriosis is a prevalent and highly morbid condition affecting millions of women globally. Our findings have important clinical implications to help guide management. Strong consideration should be given to offering surgical management for treatment of dysmenorrhea after two to three medications are tried and failed. Delaying surgical management in a woman with pain refractory to more than three medications may decrease quality of life and further increase cost.
REFERENCES
1. Fuldeore MJ, Soliman AM. Prevalence and symptomatic burden of diagnosed endometriosis in the United States: national estimates from a cross-sectional survey of 59,411 women. Gynecol Obstet Invest 2017;82:453–61. doi: 10.1159/000452660
2. Tommaso F, Flyckt R. Clinical management of endometriosis. Obstet Gynecol 2018;131:557–71. doi: 10.1097/AOG.0000000000002469
3. Fuldeore M, Yang H, Du EX, Soliman AM, Wu EQ, Winkel C. Healthcare utilization and costs in women diagnosed with endometriosis before and after diagnosis: a longitudinal analysis of claims databases. Fertil Steril 2015;103:163–71. doi: 10.1016/j.fertnstert.2014.10.011
4. Agarwal SK, Chapron C, Giudice LC, Laufer MR, Leyland N, Missmer SA, et al. Clinical diagnosis of endometriosis: a call to action. Am J Obstet Gynecol 2019;220:354.e1–12. doi: 10.1016/j.ajog.2018.12.039
5. Vercellini P, Vigano P, Somigliana E, Fedele L. Endometriosis: pathogenesis and treatment. Nat Rev Endocrinol 2014;10:261–75. doi: 10.1038/nrendo.2013.255
6. DiVasta AD, Vitonis AF, Laufer MR, Missmer SA. Spectrum of symptoms in women diagnosed with endometriosis during adolescence vs adulthood. Am J Obstet Gynecol 2018;218:324.e1–11. doi: 10.1016/j/ajog.2017.12.007
7. Soliman AM, Yang H, Du EX, Kelley C, Winkel C. The direct and indirect costs associated with endometriosis: a systematic literature review. Hum Reprod 2016;31:712–22. doi: 10.1093/humrep/dev335
8. Nnoaham KE, Hummelshoj L, Webster P, d'Hooghe T, de Cicco Nardone F, de Cicco Nardone C, et al. Impact of endometriosis on quality of life and work productivity: a multicenter study across ten countries. Fertil Steril 2011;96:366–73.e8. doi: 10.1016/j.fertnstert.2011.05.090
9. Greene R, Stratton P, Cleary SD, Ballweg ML, Sinaii N. Diagnostic experience among 4,334 women reporting surgically diagnosed endometriosis. Fertil Steril 2009;91:32–9. doi: 10.1016/j.fertnstert.2007.11.020
10. Rolla E. Endometriosis: advances and controversies in classification, pathogenesis, diagnosis, and treatment. F1000Res 2019;8:F1000. doi: 10.12688/f1000research.14817.1
11. Giudice LC. Clinical practice. Endometriosis. N Engl J Med 2010;362:2389–98. doi: 10.1056/NEJMcp1000274
12. Simoens S, Dunselman G, Dirksen C, Hummelshoj L, Bokor A, Brandes I, et al. The burden of endometriosis: costs and quality of life of women with endometriosis and treated in referral centres. Hum Reprod 2012;27:1292–9. doi: 10.1093/humrep/des073
13. Soliman AM, Surrey E, Bonafede M, Nelson JK, Castelli-Haley J. Real-world evaluation of direct and indirect economic burden among endometriosis patients in the United States. Adv Ther 2018;35:408–23. doi: 10.1007/s12325-018-0667-3
14. Soliman AM, Taylor HS, Bonafede M, Nelson JK, Castelli-Haley J. Incremental direct and indirect cost burden attributed to endometriosis surgeries in the United States. Fertil Steril 2017;107:1181–90.e2. doi: 10.1016/j.fertnstert.2017.03.020
15. Soliman AM, Surrey ES, Bonafede M, Nelson JK, Vora JB, Agarwal SK. Health care utilization and costs associated with endometriosis among women with Medicaid insurance. J Manag Care Spec Pharm 2019;25:566–72. doi: 10.18553/jmcp.2019.25.5.566
16. United States Census Bureau. American community survey 1-year estimates 2018; 2018. Accessed May 21, 2018.
https://www.census.gov/programs-surveys/acs/news/data-releases/2018/release.html
17. Marjoribanks J, Ayeleke RO, Farquhar C, Proctor M. Nonsteroidal anti-inflammatory drugs for dysmenorrhoea. TheCochrane Database of Systematic Reviews 2015, Issue 7. Art. No.: CD001751. doi: 10.1002/14651858.CD001751.pub3
18. Parazzini F, Di Cintio E, Chatenoud L, Moroni S, Ardovino I, Struzziero E, et al. Estroprogestin vs. gonadotrophin agonists plus estroprogestin in the treatment of endometriosis-related pelvic pain: a randomized trial. Gruppo Italiano per lo Studio dell'Endometriosi. Eur J Obstet Gynecol Reprod Biol 2000;88:11–4. doi: 10.1016/s0301-2115(99)00131-1
19. Lockhat FB, Emembolu JE, Konje JC. Serum and peritoneal fluid levels of levonorgestrel in women with endometriosis who were treated with an intrauterine contraceptive device containing levonorgestrel. Fertil Steril 2005;83:398–404. doi: 10.1016/j.fertnstert.2004.07.961
20. Crosignani PG, Luciano A, Ray A, Bergqvist A. Subcutaneous depot medroxyprogesterone acetate versus leuprolide acetate in the treatment of endometriosis-associated pain. Hum Reprod 2006;21:248–56. doi: 10.1093/humrep/dei290
21. Taylor HS, Giudice LC, Lessey BA, Abrao MS, Kotarski J, Archer DF, et al. Treatment of endometriosis-associated pain with elagolix, an oral GnRH antagonist. N Engl J Med 2017;377:28–40. doi: 10.1056/NEJMoa1700089
22. Comptour A, Chauvet P, Canis M, Gremeau AS, Pouly JL, Rabischong B, et al. Patient quality of life and symptoms after surgical treatment for endometriosis. J Minim Invasive Gynecol 2019;26:717–26. doi: 10.1016/j.mig.2018.08.005
23. Wang ST, Johnson SJ, Mitchell D, Soliman AM, Vora JB, Agarwal SK. Cost-effectiveness of elagolix versus leuprolide acetate for treating moderate-to-severe endometriosis pain in the USA. J Comp Eff Res 2019;8:337–55. doi: 10.2217/cer-2018-0124
24. Bellows BK, Tak CR, Sanders JN, Turok DK, Schwarz EB. Cost-effectiveness of emergency contraception options over 1 year. Am J Obstet Gynecol 2018;218:508.e1–9. doi: 10.1016/j.ajog.2018.01.025
25. Truven Health Analytics. Micromedex solutions. Red Book Online Search; 2019.
26. U.S Bureau of Labor Statistics. Consumer price index for all urban consumers medical care [CPIMEDSL]. FRED; 2018.
27. American Medical Association. CPT/Medicare payment search. Accessed May 21, 2018.
https://www.ama-assn.org/practice-management/cpt
28. World Health Organization. Making choice in health: WHO guide to cost-effectiveness analysis 2003. Accessed May 21, 2018.
https://www.who.int/choice/publications/p_2003_generalised_cea.pdf
29. Hunink MM, Weinstein M, Wittenberg E, Drummond MF, Pliskin JS, Wong JB, et al. Decision making in health and medicine. Cambridge University Press; 2014.
30. Siegel JE, Weinstein MC, Russell LB, Gold MR. Recommendations for reporting cost-effectiveness analyses. Panel on cost-effectiveness in health and medicine. JAMA 1996;276:1339–41. doi: 10.1001/jama.276.16.1339
31. Neumann PJ, Cohen JT, Weinstein MC. Updating cost-effectiveness--the curious resilience of the $50,000-per-QALY threshold. N Engl J Med 2014;371:796–7. doi: 10.1056/NEJMp1405158
32. Bullard KA, Edelman AB, Williams SM, Rodriguez MI. Ulipristal acetate compared to levonorgestrel emergency contraception among current oral contraceptive users: a cost-effectiveness analysis. Contraception 2019;100:222–7. doi: 10.1016/j.contraception.2019.05.004
33. Bullard KA, Shaffer BL, Greiner KS, Skeith AE, Rodriguez MI, Caughey AB. Twenty-week abortion bans on pregnancies with a congenital diaphragmatic hernia: a cost-effectiveness analysis. Obstet Gynecol 2018;131:581–90. doi: 10.1097/AOG.0000000000002483
34. Capraş RD, Urda-Cîmpean AE, Bolboacă SD. Is scientific medical literature related to endometriosis treatment evidence-based? A systematic review on methodological quality of randomized clinical trials. Medicina (Kaunas) 2019;55:372. doi: 10.3390/medicina55070372
35. Rimbach S, Ulrich U, Schweppe KW. Surgical therapy of endometriosis: challenges and controversies. Geburtshilfe Frauenheilkd 2013;73:918–23. doi: 10.1055/s-0033-1350890
36. Nezhat C, Vang N, Tanaka PP, Nezhat C. Optimal management of endometriosis and pain. Obstet Gynecol 2019;134:834–9. doi: 10.1097/AOG.0000000000003461
37. Hidaka T, Nakashima A, Hashimoto Y, Saito S. Effects of laparoscopic radical surgery for deep endometriosis on endometriosis-related pelvic pain. Minim Invasive Ther Allied Technol 2012;21:355–61. doi: 10.3109/13645706.2011.617758
