{"paper_id":"e8b1e30e-32c2-41b3-8c75-eca51e28df3b","body_text":"ORIGINAL RESEARCH ARTICLE\nExploring the Use of Cost-Benefit Analysis to Compare\nPharmaceutical Treatments for Menorrhagia\nSabina Sanghera 1,4 • Emma Frew 1 • Janesh Kumar Gupta 2 • Joe Kai 3 •\nTracy Elizabeth Roberts 1\nPublished online: 25 April 2015\n/C211The Author(s) 2015. This article is published with open access at Springerlink.com\nAbstract\nBackground The extra-welfarist theoretical framework\ntends to focus on health-related quality of life, whilst the\nwelfarist framework captures a wider notion of well-being.\nEQ-5D and SF-6D are commonly used to value outcomes\nin chronic conditions with episodic symptoms, such as\nheavy menstrual bleeding (clinically termed menorrhagia).\nBecause of their narrow-health focus and the condition’s\nperiodic nature these measures may be unsuitable. A viable\nalternative measure is willingness to pay (WTP) from the\nwelfarist framework.\nObjective We explore the use of WTP in a preliminary\ncost-benefit analysis comparing pharmaceutical treatments\nfor menorrhagia.\nMethods A cost-benefit analysis was carried out based on\nan outcome of WTP. The analysis is based in the UK\nprimary care setting over a 24-month time period, with a\npartial societal perspective. Ninety-nine women completed\na WTP exercise from the ex-ante (pre-treatment/condition)\nperspective. Maximum average WTP values were elicited\nfor two pharmaceutical treatments, levonorgestrel-releas-\ning intrauterine system (LNG-IUS) and oral treatment.\nCost data were offset against WTP and the net present\nvalue derived for treatment. Qualitative information ex-\nplaining the WTP values was also collected.\nResults Oral treatment was indicated to be the most cost-\nbeneficial intervention costing £107 less than LNG-IUS\nand generating £7 more benefits. The mean incremental net\npresent value for oral treatment compared with LNG-IUS\nwas £113. The use of the WTP approach was acceptable as\nvery few protests and non-responses were observed.\nConclusion The preliminary cost-benefit analysis results\nrecommend oral treatment as the first-line treatment for\nmenorrhagia. The WTP approach is a feasible alternative to\nthe conventional EQ-5D/SF-6D approaches and offers ad-\nvantages by capturing benefits beyond health, which is\nparticularly relevant in menorrhagia.\nKey Points for Decision Makers\nMenorrhagia affects health and non-health aspects of\nlife\nBroader benefits of the treatment should also be\nconsidered\nWillingness to pay is feasible and acceptable for use\nin menorrhagia\nThe cost-benefit analysis suggests oral treatment as a\nfirst-line treatment for menorrhagia\nElectronic supplementary material The online version of this\narticle (doi:10.1007/s40273-015-0280-0) contains supplementary\nmaterial, which is available to authorized users.\n& Emma Frew\nE.Frew@bham.ac.uk\n1 Health Economics Unit, Public Health Building, University\nof Birmingham, Edgbaston, Birmingham B15 2TT, UK\n2 School of Clinical and Experimental Medicine, University of\nBirmingham, Birmingham, UK\n3 Division of Primary Care and National Institute for Health\nResearch, University of Nottingham, Nottingham, UK\n4 Present Address: Health Economics Research Group, Brunel\nUniversity London, Uxbridge UB8 3PH, UK\nPharmacoEconomics (2015) 33:957–965\nDOI 10.1007/s40273-015-0280-0\n\n1 Introduction\nEconomic evaluation offers a formal toolkit to assess both\nthe costs and consequences of competing services. In the\nUK, decision makers such as the National Institute for\nHealth and Care Excellence (NICE) have adopted cost-\nutility analysis as the economic evaluation method of\nchoice, which measures outcomes using quality-adjusted\nlife-years (QALYs) with a focus on health-related out-\ncomes [1]. The conventional criterion for decision making\nis based on a health-maximisation principle with the aim of\nmaximising QALYs relative to the resources available.\nThis approach to economic evaluation, with its focus on\nhealth outcomes, is described in theoretical terms as an\n‘extra-welfarist’ approach [ 2]. To construct QALYs, it is\nrecommended that either the EQ-5D or the SF-6D instru-\nment is used to measure health-related quality of life. The\nuse of cost-utility analysis offers a framework for evi-\ndence-based decision making in which the objective is to\nmaximise health, but it offers limited support for the\nevaluation of interventions for which there are gains that go\nbeyond health alone. Cost-benefit analysis is an alternative\napproach within the economic evaluation toolkit and in\ncontrast is based on the welfarist approach. Cost-benefit\nanalysis places a monetary value on outcomes using stated\npreferences methods such as contingent valuation or\n‘willingness to pay’ (WTP). A cost-benefit analysis takes a\nwider perspective compared with a cost-utility analysis and\nthus offers the potential to incorporate costs and conse-\nquences that go beyond the healthcare sector.\nMeasures used to capture outcomes underpinned by the\nextra-welfarism framework are commonly used across all\ntypes of clinical conditions, including those that are\nchronic but have symptoms that occur in episodes [ 3]. One\nsuch condition is heavy menstrual bleeding, which is\nclinically termed ‘menorrhagia’. Menorrhagia can be de-\nfined as ‘‘excessive menstrual blood loss which interferes\nwith the woman’s social, emotional, physical and material\nquality of life’’ [ 4]. The principal driver for treatment is\nbased on women’s experience of its interference in their\nlives [5]. An objective measure of volume of blood loss is\ntherefore no longer considered to be suitable, and it is a\nwoman’s subjective assessment of her ability to cope and\nthe perceived impact on her quality of life that is increas-\ningly used to assess treatment success [ 5]. As impact on\nquality of life is the key indicator of treatment success, it is\nimportant to ensure that the quality-of-life measure is used\naccurately to reflect women’s concerns and experiences.\nHistorically, women had surgery to treat menorrhagia;\nhowever, non-hormonal and hormonal pharmaceutical\ntreatments are now available as first-line treatment for\nwomen with menorrhagia. The first robust, UK-based\neconomic evaluation of these pharmaceutical treatments\nfor menorrhagia was conducted alongside a trial using both\nEQ-5D and SF-6D to compare levonorgestrel-releasing\nintrauterine system (LNG-IUS) with usual medical treat-\nment as first-line treatment for menorrhagia [ 6]. LNG-IUS\nis an intrauterine device that can be inserted by the general\npractitioner (GP) and also provides contraception. Usual\nmedical treatment can include one of the following:\ntranexamic acid, mefenamic acid, norethisterone, depo-\nprovera, or combined estrogen/progestogen or progesto-\ngen-only oral contraceptive pill (any formulation), which is\nprescribed by the GP (a description of each treatment is\npresented in the online resource).\nConcerns around the use of these measures, which are\nunderpinned by the extra-welfarist perspective in menor-\nrhagia, were highlighted as the treatment recommendation\nto decision makers differed depending on the measure used\nto generate the QALY [ 6]. Despite being advocated by\ndecision makers, there is evidence to suggest that these\nmeasures, which focus on health, may not be suitable for a\ncondition such as menorrhagia because women believe\nboth health and non-health aspects of life are affected by\nthe condition [ 7]. Furthermore, the standard recall periods\nof typically used measures and the episodic nature of the\ncondition also mean results could be affected by the timing\nof assessment. This combined reasoning raises questions\nabout the suitability of QALYs as an outcome measure.\nThe WTP measure, underpinned by welfarist theory,\nenables the respondent to take into consideration both\nhealth and non-health outcomes and may overcome the\nissue of timing of assessment. To demonstrate its feasi-\nbility, we explore the use of the WTP approach in a pre-\nliminary cost-benefit analysis to assess the cost\neffectiveness of LNG-IUS compared with usual medical\ntreatment (also referred to as oral treatment) as the first-line\ntreatment for menorrhagia.\n2 Methods\nA cost-benefit analysis was carried out based on an out-\ncome of WTP. The analysis is related to the UK primary\ncare setting and provides an assessment of the difference in\ncosts and WTP between interventions over a 24-month\ntime horizon. The reporting of the cost-benefit analysis\nfollows the CHEERS guidelines [ 8].\n2.1 Participants and Study Design\nFor this exploratory study, a convenience sample of 110\nwomen were recruited from general gynaecology outpa-\ntient clinics based in the Birmingham Women’s Hospital\n958 S. Sanghera et al.\n\nbetween December 2012 and January 2013. Women who\nwere menstruating but did not necessarily have experience\nof menorrhagia or its treatments were sought, so all women\nattending an appointment were approached to complete a\nbooklet questionnaire, either in the clinic or at home, and\nprovided written informed consent to participate. Respon-\ndents who took the questionnaire home to complete were\ngiven a stamped addressed envelope. Women were asked\nto value the two pharmaceutical treatments of LNG-IUS\nand oral treatment.\n2.2 Outcome Measures\nWTP is elicited from the ex-ante perspective. Individuals\nare asked to express in monetary terms how much they\nvalue a good or a service that leads to a change in outcome\n[9]. In this context, maximum WTP values were derived\nprior to the change in outcome occurring, from respondents\nwho are ‘at risk’ of the condition, or ‘at risk’ of requiring\ntreatment. Given the UK is a tax-funded system that offers\nhealthcare ‘free at the point of use’, we designed the WTP\nstudy to elicit the views of the at-risk population. The ra-\ntionale being that because society is funding the healthcare\nsystem, it is the views of those at risk that should be sought.\nThe questionnaire booklet was reviewed by clinical\nexperts in menorrhagia, psychologists and health econo-\nmists for face and content validity. Maximum WTP values\nwere elicited for both LNG-IUS and oral treatment using a\nself-complete booklet questionnaire. The booklet captured\ndata on WTP and sociodemographic details.\nA description of menorrhagia (without treatment) was\nfirst presented and was based on the domains of the dis-\nease-specific quality-of-life Menorrhagia Multi-attribute\nAssessment Scale (MMAS). This measure incorporates\nboth the health and non-health outcomes associated with\nmenorrhagia and consists of six attributes, ‘practical diffi-\nculties, ‘social life’, ‘psychological health’, ‘physical\nhealth and well-being’, ‘work/daily routine’ and ‘family\nlife/ relationships’ [ 10]. We used baseline MMAS data\nfrom a recent trial (ECLIPSE, ISRCTN86566246) to gen-\nerate the description. We then presented a scenario de-\nscribing the expected average ‘outcome’ associated with\nthe two treatments, LNG-IUS (termed Mirena in the sce-\nnarios) and oral treatment, using average follow-up MMAS\ndata from the ECLIPSE trial [ 11].\nUsing the same method the scenarios for the outcomes\nassociated with LNG-IUS and oral treatments, using the\n6-month ECLIPSE MMAS data, were generated. Infor-\nmation describing the process of care was also described in\nthe treatment scenarios (see online resource for method\nused for scenario development).\nRespondents were asked for their preferred treatment,\nand their maximum monthly out-of-pocket WTP value up\nuntil menopause first for oral treatment, and then LNG-\nIUS. A payment scale, which presents respondents with a\nrange of monetary values, was used to elicit WTP values as\nit has a higher completion rate than other methods that can\nbe used in a postal questionnaire [ 12]. The payment scale\nwas derived from a previous applied WTP study [ 12], and\nused a range from £0 to £500, which was considered to be\nmost suitable, as the questionnaire asked respondents to\nprovide a monthly WTP value. An open-ended option for\nvalues greater than £500 was offered. Following the WTP\nquestion, we asked respondents to outline the reasons for\ntheir WTP values in an open-ended question to assess the\nvalidity of the WTP responses. The respondents were then\nasked to indicate whether they found the WTP question\ndifficult to answer, and to provide reasons for their re-\nsponse. The time frame of payment ‘up until menopause’\nwas explicitly stated to ensure that WTP values were not\noverestimated [13]. The questionnaire included a reminder\nto consider the amount that they can afford to pay to ensure\nthat the responses obtained were realistic and within the\nrespondent’s means [ 14]. The time period was intuitive\ngiven the nature of the condition. The monthly payment\ntime frame was used because women generally pay\nmonthly (or every 3 months) for prescriptions for menor-\nrhagia, for sanitary protection and will experience the\nbenefits of treatment on a monthly basis.\nThe booklet questionnaire is presented in the online\nresource.\n2.3 Cost and Resource Use\nGiven that an ex-ante perspective was adopted, the women\nwere not typically being treated with LNG-IUS or oral\ntreatment, and therefore primary cost data were not avail-\nable. The costs were consequently derived using the\nECLIPSE trial data as the most appropriate available\nsource and also to enable comparability between the cost-\nutility analysis alongside the ECLIPSE trial and our cost-\nbenefit analysis [6]. Briefly, the general healthcare costs for\nboth treatments included healthcare staff costs and the cost\nof the treatments. The costs of LNG-IUS and oral treatment\nwere estimated using the British National Formulary [ 15].\nStaff costs were calculated using the nationally recognised\nreference costs [16]. All costs are reported in 2011 prices in\nUK (£) sterling using the UK hospital and community\nhealth services index [16]. The overall costs for both LNG-\nIUS and oral treatment at the 2-year time point in the\nECLIPSE trial included crossover between treatment arms,\nas the analysis within the trial was ‘intention to treat’. The\naverage costs of LNG-IUS and oral treatment per person\nwere taken from the average results of a trial-based eco-\nnomic evaluation, where a decision model was used as the\nbasis for the evaluation, and were reported to be £430 and\nExploring Cost-Benefit Analysis in Menorrhagia 959\n\n£330, respectively [ 6]. All costs are from a UK National\nHealth Service perspective. A societal perspective for costs\nwas considered but was not used to enable a comparison\nbetween previous analyses using EQ-5D and SF-6D [ 6].\n2.4 Analysis\nAverage maximum WTP values are compared to the cost\nof providing the service to generate the net present value\n(NPV) for each treatment option. If the present value of\nbenefits (expressed through WTP) outweighs the present\nvalue of costs (present value of benefits - present value of\ncosts), then the net benefits are said to be positive\n(NPV [ 0) and it is in society’s interest to recommend the\ntreatment choice. The treatment choice that yields the\nmaximum NPV is the most efficient.\nThe incremental net benefit that shows the difference\nbetween the net benefits across the treatments (NPV oral\ntreatment - NPV LNG-IUS) is also presented. To adjust to\nthe present value, the recommended discount rate of 3.5 %\nwas applied to both the costs and outcomes [ 3]. The WTP\nvalues derived for both LNG-IUS and oral treatment were\nbased on a monthly amount, to obtain the present value the\nWTP value was discounted for every month up to and\nincluding 24 months. WTP data were found to be non-\nnormal and were therefore log transformed [ 17]. A paired\nt-test was then applied to the log transformed data to ex-\nplore the difference between the WTP values for each\ntreatment. Protest answers and non-response were removed\nfrom the analysis.\nThe base-case analysis is presented using the cost data\ndescribed above, which relates to the outcome of the eco-\nnomic evaluation alongside the ECLIPSE trial [ 6] and is\nbased on an ‘intention-to-treat’ analysis. The base-case\nanalysis was carried out in addition to two sensitivity\nanalyses to assess uncertainty in the results.\nSensitivity analysis:\n1. An assessment of uncertainty in the mean NPV is\ncarried out by bootstrapping. Bootstrapping involves\nrandomly sampling values, with replacement from the\nobserved values. Multiple samples are drawn, as 1000\nbootstrap datasets are generated using STATA (Ver-\nsion 11.0), and each dataset is considered to be a\nreiteration of the trial [ 18]. Bootstrapped 95 % confi-\ndence intervals around the mean values are presented\nand the distributions of the bootstrapped values are\nthen presented graphically.\n2. A second sensitivity analysis using alternative cost data,\nwhich were not derived from the ECLIPSE model, was\napplied to identify the impact of the source of cost data.\nIn this sensitivity analysis, the resource use and cost data\nrelated to the exclusive use of either LNG-IUS or oral\ntreatment are applied, treatment cross-over is not\nconsidered. Table 1 outlines the cost data used in the\nsensitivity analysis. As oral treatment comprises a range\nof pharmaceutical treatments, the average cost of oral\ntreatment was weighted according to the frequency with\nwhich each treatment is prescribed [6].\n3 Results\n3.1 Base-Case Results\nThe maximum average WTP for LNG-IUS was £365 and\nfor oral treatment was £372. This difference was not sta-\ntistically significant ( p = 0.1247; p \\ 0.05). The max-\nimum average WTP for oral treatment was 13 % higher\nthan the cost of the intervention, and the maximum average\nWTP for LNG-IUS was 15 % lower than the cost of\ntreatment (Table 2).\nThe base-case results indicate that oral treatment pro-\nvides a positive NPV of £45, resulting in a welfare gain,\nand LNG-IUS produces a negative NPV of £ -68, leading\nto a welfare loss (Table 2). When comparing the two\ntreatments, the incremental net benefit exceeds zero sug-\ngesting that oral treatment is cost beneficial compared with\nLNG-IUS. Based on the mean values, oral treatment could\nbe considered the most cost-beneficial intervention.\n3.2 Sensitivity Analysis\nIn sensitivity analysis 1, the confidence intervals associated\nwith the NPVs for both treatments overlap. The NPV is\n£-68 [95 % CI £ -186 to £50] for LNG-IUS and £45\n[95 % CI £ -55 to £146] for oral treatment. This suggests\nthere is some uncertainty between which treatment is most\ncost beneficial. However, when presented using graphed\nplots the bootstrapped NPV for LNG-IUS and oral treat-\nments show a clearer picture with respect to the welfare\ngains and losses (Figs. 1, 2). In most cases, oral treatment\nproduces a positive NPV, as a greater proportion of the\nbootstrapped NPV values lie above £0 (Fig. 1). In contrast,\nthe plots for LNG-IUS are the inverse of those for oral\ntreatment, as in most cases LNG-IUS produces a negative\nNPV. These bootstrapped plots suggest that oral treatment\nis more likely to be cost beneficial relative to LNG-IUS and\nreinforce the base-case result.\nIn sensitivity analysis 2, the mean WTP for LNG-IUS is\n41 % greater than the cost of LNG-IUS. The mean WTP for\noral treatment is 280 % greater than the cost of oral treatment\n(Table 3). Both treatments generate a positive NPV.\nThe results still indicate that oral treatment remains the\nmost cost-beneficial treatment as it generates a greater\n960 S. Sanghera et al.\n\nNPV than LNG-IUS and the incremental net benefit ex-\nceeds zero.\n3.3 Response to Outcome Measure\nOne hundred and ten women completed and returned the\nquestionnaire. Both LNG-IUS and oral treatment received\nthe same number of non-responses (four in each). Seven\nprotest answers, which relate to the individual refusing to\nprovide a WTP value, were identified from the qualitative\nexplanations offered.\nNinety-nine women with an average age of 37 years\nprovided a WTP value for LNG-IUS and oral treatment.\nLNG-IUS was the preferred treatment (47), followed by\noral treatment (39) and no preference (11). Two respon-\ndents did not answer the question (see online resource for\nTable 1 Cost data used in sensitivity analysis\nUnit cost (£) a Source\nLNG-IUS\nConsultation (GP 10 min) 26.67 Curtis 2011 (16)/expert opinion b\nInsertion\nGP (20 min) 53.33 Curtis 2011 (16)/expert opinion\nPractice nurse (20 min) 17.00 Curtis 2011 (16)/expert opinion\nDevice cost 88.00 BNF 62 (15)\nSterile pack (insertion) 21.63 NICE (4) (inflated to 2011)\nFollow-up\n6-week review: (GP 10 min) 26.67 Curtis 2011 (16)/expert opinion\n3 month review: (GP 10 min) 26.67 Curtis 2011 (16)/expert opinion\nUnit cost (£)\na Frequencyc Source\nOral treatment\nProgestogen (Cerazette) 8.68 21 BNF 62 (15)\nTranexamic acid (Cyclokapron) 14.30 19 BNF 62 (15)\nMefenamic acid (Ponstan) 15.72 8 BNF 62 (15)\nNorethisterone 2.18 2 BNF 62 (15)\nCombined oral contraceptive (Microgynon) 2.82 1 BNF 62 (15)\nMethoxyprogesterone acetate injections (Depo-provera) 6.01 6 BNF 62 (15)\nConsultation: (GP 10 min) 26.67 Curtis 2011 (16)/expert opinion\nReview of medication (GP 10 min) 26.67 Curtis 2011 (16)/expert opinion\nBNF British national formulary, GP general practitioner, LNG-IUS levonorgestrel-releasing intrauterine system, NICE National Institute for\nHealth and Care Excellence\na The cost year is 2011\nb Expert opinion refers to clinical experts in menorrhagia (JG, JK)\nc The frequency is used to calculate the weighted average cost of oral treatment. The values are derived from data in a model-based economic\nevaluation [6]\nTable 2 Base-case results:\nmean WTP and cost of\ntreatment\nIntervention WTP Cost NPV (WTP - cost) INB (NPV oral - NPV LNG-IUS)\nLNG-IUS £365 £433 £ -68\nOral treatment £372 £326 £45 £113\nMean difference £ -7 £107\nCost data are reported in UK (£) sterling and the cost year is 2011. Costs are rounded to the nearest whole\nnumber\nCost data relate to the results of the economic evaluation alongside the ECLIPSE trial [ 6], which are based\non an ‘intention-to-treat’ analysis. The initial costs used in the economic evaluation alongside the ECLIPSE\ntrial are described in Table 1\nINB incremental net benefit, LNG-IUS levonorgestrel-releasing intrauterine system, NPV net present value,\nWTP willingness to pay\nExploring Cost-Benefit Analysis in Menorrhagia 961\n\nfurther information). Eighty percent of women said they\nhad experience of heavy periods at one time in their lives,\nbut this may not necessarily mean experience of heavy\nperiods over consecutive cycles as defined by menorrhagia.\nThe two most commonly cited reasons for a WTP value\nwere related to the ‘effect of the treatment’ and ‘afford-\nability’. There were three respondents that misunderstood\nthe WTP question.\nOver 60 % of women who completed at least one WTP\nquestion said that the question was not difficult to answer.\nOf those who did find the question difficult to answer, the\nmost common reason was related to ‘not being used to\nvaluing healthcare’. For those who did not find the val-\nuation difficult, the most commonly cited reason was ‘a\nreasonable amount to pay for the expected benefits’.\n4 Discussion\nOver a 24-month time horizon, the total cost of oral\ntreatment is cheaper than LNG-IUS (£326 compared with\n£433 respectively). The NPV of oral treatment is greater\nthan LNG-IUS (£45 compared with £ -68, respectively).\nThus, oral treatment produced a positive incremental net\nFig. 1 Base-case results:\nbootstrapped net present\nvalue—oral treatment\nFig. 2 Base-case results:\nbootstrapped net present\nvalue—levonorgestrel-releasing\nintrauterine system\nTable 3 Sensitivity analysis:\nmean WTP and cost of\ntreatment\nIntervention WTP Cost NPV (WTP - cost) [95 % CI] INB (NPV oral - NPV LNG-IUS)\nLNG-IUS £365 £260 £106 [£ -10 to £221]\nOral treatment £372 £98 £274 [£168 to £380] £168\nMean difference £ -7 £162\nCI confidence interval, INB incremental net benefit, LNG-IUS levonorgestrel-releasing intrauterine system,\nNPV net present value, WTP willingness to pay\nCost data are reported in UK (£) sterling and the cost year is 2011. Costs are rounded to the nearest whole\nnumber\n962 S. Sanghera et al.\n\nbenefit (equal to £113). On the basis of these results, oral\ntreatment could be recommended as the first-line treatment\nfor menorrhagia.\nThe findings from both sensitivity analyses support the\nbase-case analysis. The bootstrapped plots in sensitivity\nanalysis 1 demonstrate that oral treatment is the most likely\ntreatment to be cost beneficial. In sensitivity analysis 2,\nwhere cost data are not taken from the ECLIPSE trial de-\ncision model, which used intention-to-treat analysis but\ninstead relate to the exclusive use of either LNG-IUS or\noral treatment, both oral treatment and LNG-IUS generated\na positive NPV of £274 and £106 respectively. However,\noral treatment yielded the maximum NPV and therefore\nwas still indicated to be the most efficient choice.\nTherefore, the base-case analysis and sensitivity ana-\nlyses suggest that oral treatment is the most cost-beneficial\ntreatment and therefore should be recommended as the\nfirst-choice treatment for menorrhagia in clinical practice.\nIn a privately financed healthcare system, the resource\nallocation decision from a cost-benefit analysis is relatively\nstraightforward as a positive NPV indicates that the inter-\nvention(s) be recommended for use in practice. In contrast,\nwhen making resource allocation decisions in a publicly\nfunded healthcare system where a budget constraint exists,\nit is unlikely to be feasible that all interventions with a\npositive NPV are recommended for clinical practice [ 18].\nUnder budget constraints, the aim is to maximise benefits\nand therefore the interventions could be ranked against one\nanother and the intervention with the greatest NPV im-\nplemented [ 9]. Whilst this issue is not resolved, in this\ncase, we adopted the decision rule that the treatment choice\nthat yields the maximum NPV is the most efficient and\nshould be implemented.\n4.1 Strengths and Limitations\nThis is the first study, to our knowledge, that applies a\ncost-benefit analysis to compare LNG-IUS against oral\ntreatment in menorrhagia. Furthermore, a cost-benefit\nanalysis is rarely conducted a nd reported in the literature\nand a strength of the current analysis, is that it is based on\nWTP values that have been elicited from the ex-ante\nperspective, which is the oretically preferred [ 9]. An ad-\nditional strength is that once the questionnaires were de-\nveloped, they were checked by clinical experts in\nmenorrhagia, by psychologists and external health econ-\nomists to assess and improve their face and content va-\nlidity. Thus, rather than basin g the ex-ante questionnaire\nscenarios, for menorrhagia and treatment effectiveness,\non expert opinion alone or expected outcomes, novel\nmethods were used to base the scenarios on observed\nevidence from the ECLIPSE trial, which increases the\nreliability of the findings.\nA limitation of the exploratory study is that we did not\ndetermine how many women from our convenience sample\nhad experience of the treatments for menorrhagia. This\ninformation would help to determine the extent to which\nour sample reflects a true ex-ante perspective. The sample\nused does to some extent reflect the ‘at-risk’ population\ngroup, which would be made up of both women who have,\nand do not have, the condition. However, where women\nhave experience of both menorrhagia and its treatments this\ndoes not strictly meet the ex-ante criterion.\nThe costs for the base-case analysis were taken from the\naverage results of a trial-based economic evaluation to\nenable comparability between that cost-utility analysis and\nour cost-benefit analysis [ 6]. A potential limitation of this\napproach is that the possibility of changing and stopping\ntreatment was not presented in the WTP scenario and\ntherefore would not have been considered when providing\na WTP value. However, when using cost data that are only\nrelated to the WTP scenario the same treatment was found\nto be superior. In this case, although the overall cost-benefit\ndecision did not differ, the extent of the welfare gain\nproduced by oral treatment compared with LNG-IUS did\nvary, and was dependent on the cost data used in the cost-\nbenefit analysis.\nIt was not possible to conduct a comprehensive cost-benefit\nanalysis because societal costs were not available. Only\nhealthcare costs were considered in this evaluation. The dif-\nfering perspectives across costs and benefits could bias the\nresults in favour of the benefits of the treatments as a broader\nperspective is used. However, when assessing incremental net\npresent values across treatments, the impact is limited as the\nsame approach is applied to both interventions assessed. In-\ncorporating societal costs, such as lost productivity and out-of-\npocket prescription fees, would not be straightforward be-\ncause of double counting. Arguably, the WTP outcome al-\nready incorporates lost productivity as the WTP scenarios\nincluded impact on work/daily routine. Therefore, if changes\nin productivity are also counted on the cost side of the equa-\ntion, it is possible that the benefits of treatment are double\ncounted [19]. The other aspect of societal cost is the cost of\nprescriptions, which would only be relevant to oral treatment\nand the exclusion of this could be considered as bias in favour\nof oral treatment. However, we estimate this cost to be small\nand unlikely to change the treatment recommendation.\nFinally, we recognise the limitation associated with re-\nporting costs in the 2011 price year and WTP values in\n2013. Different years of valuation are not unusual in health\neconomics as the utility values derived from other standard\nmeasures such as EQ-5D, which is based on time trade-off\npreference values from the 1990s, are similarly not derived\nduring the same year as costs but are presented in the same\neconomic evaluation. Between these two particular price\nyears, inflation has been particularly low and therefore the\nExploring Cost-Benefit Analysis in Menorrhagia 963\n\nlack of adjustment would introduce little if any bias. Fur-\nthermore, by not changing the cost year we have presented\nresults that are directly comparable to the cost utility\nanalysis results.\nGiven the limited availability of cost-benefit analyses\ncurrently reported in the literature, a strength of the current\narticle is that we have reported all relevant methods and\nresults as explicitly as possible including additional infor-\nmation to that which is required by recommended guide-\nlines, such as CHEERS, to which published economic\nevaluations are typically recommended to adhere. As far as\nwe are aware, the current study is the first cost-benefit\nanalysis to attempt to follow CHEERS guidelines [ 8] and\nthe shortcomings of those guidelines in terms of their\nrelevance to the full and clear reporting of economic\nevaluations that take the form of a cost-benefit analysis\nhave been apparent.\n4.2 Comparison with Other Studies\nVery few cost-benefit analyses have been published. To the\nbest of our knowledge, this is the only cost-benefit analysis\nfocussing on menorrhagia. A recent cost-benefit analysis\nhas been carried out but in the area of spinal surgery in\nSwitzerland, where WTP was elicited from the ex-post\nperspective using patient values [ 20]. The authors sug-\ngested that further methodological work be carried out on\nthe use of ex-ante WTP values, as this perspective is rec-\nommended for publicly funded healthcare systems. Despite\nthe ex-ante perspective WTP and cost-benefit analysis be-\ning theoretically preferred [ 9], WTP is typically elicited\nfrom the ex-post perspective [ 21].\nIn terms of comparisons with other UK studies reporting\ntreatment recommendations for menorrhagia, in contrast to\nour findings, the NICE guidelines recommend LNG-IUS as\nthe first-line treatment for menorrhagia [ 4]. Similarly, the\neconomic evaluation alongside the ECLIPSE trial using\nEQ-5D also found LNG-IUS the most cost-effective in-\ntervention [ 6]. However, the recommendation for oral\ntreatment to be first-line treatment in our cost-benefit\nanalysis does correspond with the recommendation from\nthe economic evaluation alongside the ECLIPSE trial using\nSF-6D [6]. Decision makers currently recommend EQ-5D\nfor the valuation of outcomes [ 1], therefore LNG-IUS\nwould be considered the most cost-effective treatment,\ndespite other measures demonstrating that LNG-IUS is not\nthe most cost-effective intervention.\n4.3 Implications and Further Research\nThe results of the current analysis are not attempting to\noverturn the NICE guideline recommendation but instead\npresent an exploration of the use of an alternative measure.\nThe results of this analysis present potentially important issues\nabout the use of the conventional measures from the extra-\nwelfarist perspective, EQ-5D and SF-6D, within the context of\ndecision making for certain diseases such as menorrhagia. The\ncost-benefit analysis approach showed oral treatment to be the\nmost efficient use of society’s resources. We have previously\nshown [6] that the type of measure used to value outcomes has\nimportant implications for recommendations to decision\nmakers. To improve the generalisability and robustness of the\nresults, more research needs to be conducted using the WTP\napproach on a larger sample size that more closely resembles\nthe general population.\nFurther research to explore the role of cost-benefit\nanalysis and the use of the welfarist approach for certain\nconditions that affect non-health aspects of quality of life is\nrequired both generally by methodologists and specifically\nin applied research for clinical conditions, such as\nmenorrhagia.\nEthics Ethical approval was obtained from the National Research\nEthics Service Committee South West—Exeter. The research was\ntherefore performed in accordance with the ethical standards outlined\nin the 1964 Declaration of Helsinki.Written consent was obtained\nfrom the participants prior to their inclusion in the study.\nAcknowledgments We thank the women who participated in the\nstudy and the National Institute for Health Research for funding the\nresearch (Grant No: 02/06/02). J. K. Gupta reports honoraria received\nfrom Bayer (UK), the manufacturer of LNG-IUS (Mirena). All other\nauthors report no conflicts of interest.\nAuthor Contributions SS, EF and TR conceived and designed the\nstudy. SS developed and administered the questionnaire, carried out\nthe data analysis, conducted the CBA and wrote the manuscript. EF\nand TR supported the questionnaire development and analysis. JG and\nJK facilitated data collection. All authors edited the manuscript. SS,\nEF and TR are the guarantors of this work.\nOpen Access This article is distributed under the terms of the\nCreative Commons Attribution-NonCommercial 4.0 International\nLicense (http://creativecommons.org/licenses/by-nc/4.0/), which per-\nmits any noncommercial use, distribution, and reproduction in any\nmedium, provided you give appropriate credit to the original author(s)\nand the source, provide a link to the Creative Commons license, and\nindicate if changes were made.\nReferences\n1. National Institute of Health and Clinical Excellence. Guide to the\nmethods of technology appraisal 2013. London: National Institute\nof Health and Clinical Excellence; 2013.\n2. Birch S, Donaldson C. Valuing the benefits and the costs of\nhealthcare programmes: where’s the ‘extra’ in extra-welfarism.\nSoc Sci Med. 2003;56:1121–33.\n3. National Institute of Health and Clinical Excellence. Guide to the\nmethods of technology appraisal. London: National Institute of\nHealth and Clinical Excellence; 2008.\n964 S. Sanghera et al.\n\n4. National Collaborating Centre for Women’s and Children’s\nHealth. Heavy menstrual bleeding. London: Royal College of\nObstetricians and Gynaecologists; 2007.\n5. Shapley M, Jordan K, Croft PR. Why women consult with in-\ncreased vaginal bleeding: a case–control study. Br J Gen Pract.\n2002;52:108–13.\n6. Sanghera S, Roberts T, Barton P, Daniels J, Middleton L, Gen-\nnard L, Kai J, Gupta J. LNG-IUS vs. usual medical treatment for\nmenorrhagia: an economic evaluation alongside a randomised\ncontrolled trial. PLoS One. 2014;9(3):e91891. doi: 10.1371/\njournal.pone.0091891.\n7. Sanghera S, Frew E, Kai J, Gupta J, Roberts TE. An assessment\nof economic measures used in menorrhagia: a systematic review.\nSoc Sci Med. 2013;98:149–53.\n8. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, et al.\nConsolidated health economic evaluation reporting standards\n(CHEERS) statement. BMJ. 2013;346:f1049.\n9. McIntosh E, Clarke PM, Frew EJ, Louviere JL. Applied methods\nof cost-benefit analysis in health care. Oxford: Oxford University\nPress; 2010.\n10. Shaw RW, Brickley MR, Evans L, Edwards MJ. Perceptions of\nwomen on the impact of menorrhagia on their health using multi-\nattribute utility assessment. Br J Obstet Gynaecol.\n1998;105:1155–9.\n11. Gupta J, Kai J, Middleton L, Pattison H, Gray R. Levonorgestrel\nintrauterine system vs medical therapy for menorrhagia. N Engl J\nMed. 2013;368:128–37.\n12. Whynes DK, Frew E, Wolstenholme JL. A comparison of two\nmethods for eliciting contingent valuations of colorectal cancer\nscreening. Health Econ. 2003;22(4):555–74.\n13. Smith RD. Construction of the contingent valuation market in\nhealthcare: a critical assessment. Health Econ. 2003;12:609–28.\n14. Smith RD. It’s not just what you do, it’s the way that you do it:\nthe effect of different payment card formats and survey admin-\nistration on willingness to pay for health gain. Health Econ.\n2006;15:281–93.\n15. Joint Formulary Committee. British National Formulary 62.\nLondon, UK: BMJ Group and Pharmaceutical Press; 2011.\n16. Curtis L. Unit costs of health and social care: 2011. Canterbury:\nPersonal Social Service Research Unit; 2011.\n17. Shackley P, Donaldson C. Should we use willingness to pay to\nelicit community preferences for healthcare? New evidence from\nusing a ‘marginal’ approach. J Health Econ. 2002;21:971–91.\n18. Shackley P, Donaldson C. Willingness to pay for publicly fi-\nnanced healthcare: how should we use the numbers? Appl Econ.\n2000;32:2015–21.\n19. Glick HA, Doshi JA, Sonnad SS, Polsky D. Economic evaluation\nin clinical trials. Oxford: Oxford University Press; 2007.\n20. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stod-\ndart GL. Methods for the economic evaluation of health care\nprogrammes. 3rd ed. Oxford: Oxford University Press; 2005.\n21. Haefeli M, Elfering A, Mcintosh E, Gray A, Sukthankar A, et al.\nA cost-benefit analysis using contingent valuation techniques: a\nfeasibility study in spinal surgery. Value Health. 2008;11:575–88.\nExploring Cost-Benefit Analysis in Menorrhagia 965","source_license":"CC0","license_restricted":false}