- Email: [email protected]
Examining the Value of Menopausal Symptom Relief Among US Women
VALUE IN HEALTH 19 (2016) 158–166
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jval
Examining the Value of Menopausal Symptom Relief Among US Women Benjamin M. Craig, PhD1,*, Sandra A. Mitchell, PhD, CRNP2 1 Health Outcomes and Behavior, Moffitt Cancer Center and University of South Florida, Tampa, Florida, USA; 2Outcomes Research Branch, Healthcare Delivery Research Program, Division of Cancer Control & Population Sciences, National Cancer Institute, Rockville, Maryland, USA
AB STR A CT
Background: Menopausal symptoms can cause significant distress to women, yet little is known about the value women place on these symptoms. Methods: In April 2013, 3397 US women, aged 40 to 69 years, completed an online survey that included 30 paired comparisons. Specifically, respondents were shown two menopausal symptoms described using the Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events and asked, “Which do you prefer?” From their choices, we estimated a generalized linear model to assess the values women place on symptom relief in terms of quality-adjusted life-years (QALYs). Results: Approximately half the respondents (1753 of 3397 [52%]) always preferred a reduced lifespan (up to 90 days) instead of experiencing menopausal symptoms at their worst for 30 days. For most of the symptoms (248 of 263 [94%]), including low-grade events, QALYs were significantly reduced (P o 0.05). The value women placed on relief ranged widely by symptom domain: the relief from depression, problems with memory, headache, pain in abdomen, problems with anger, and vomiting were
the most valuable. Conclusions: Overall, the value women place on menopausal symptom relief is surprisingly high. As the first national study to directly ask women about their preferences and to estimate the value of menopausal symptom relief on a QALY scale, this work provides critical evidence for health outcomes research in midlife women and can be applied in the evaluation of treatments that reduce or eliminate menopausal symptoms. This work also provides proof-of-concept for an approach to value Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events responses on a QALY scale. Keywords: menopause, Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE), preference, QALY, quality-adjusted life-year, symptoms, women’s health.
Introduction
events (AEs) [13–15]. This system is intended to complement National Cancer Institute’s CTCAE, an existing lexicon of clinician-reported AE items required for use in all National Cancer Institute–sponsored trials. The PRO-CTCAE item library is composed of 126 items that evaluate the presence, frequency, severity, and interference with usual or daily activities associated with 78 symptomatic toxicities [16–18]. Given PRO-CTCAE’s focus on self-reported AEs relating to treatment, it was particularly well suited for the creation of the MP-30D, a PRO checklist for the measurement of symptoms related to menopause. The primary focus of this article was a health valuation study of the MP-30D from the perspective of US women aged 40 to 69 years. Health valuation studies are commonplace throughout the clinical literature; yet to our knowledge, this is the first study that asked women about their preferences regarding symptoms associated with menopause [19–23]. Nearly all past studies focused on gender-neutral PRO instruments (e.g., the EuroQol five-dimensional questionnaire), most taking the societal perspective [24–27]. This reliance on gender-neutral domains (i.e., those
Menopause reflects not only the cessation of reproductive capacity but also often includes a multitude of associated symptoms that can negatively affect women’s health-related quality of life [1,2]. Whether part of the natural aging process, an adverse effect of medical treatment, or damage to the ovaries, the symptoms that accompany menopause can be measured using one of the available patientreported outcome (PRO) instruments (Table 1) [3–12]. No study, however, has asked women which symptoms and symptom attributes (frequency, severity, and interference) matter most to them. Understanding what women experience and what they value in terms of symptoms is critical for women’s health care, including the evaluation of new health technologies (e.g., Duavee), the assessment of treatment regimens (e.g., chemotherapy), and other forms of comparative effectiveness research. The National Cancer Institute’s PRO version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE) measurement system allows patients to self-report symptomatic adverse
Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: There are no conflicts of interest. * Address correspondence to: Benjamin M. Craig, Moffitt Cancer Center, 12902 Magnolia Drive, MRC-CANCONT, Tampa, FL 33612. E-mail: benjamin.craig@moffitt.org. 1098-3015$36.00 – see front matter Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jval.2015.11.002
159
VALUE IN HEALTH 19 (2016) 158–166
Table 1 – MP-30D domains: Comparison to HRQOL instruments for menopause. MP-30D domains arranged by the CTCAE system organ class*
HRQOL instruments for menopause BESS [4]
Cardiac disorders 1. Pounding or racing heartbeat: F, S X Gastrointestinal disorders 2. Pain in the abdomen: F, S, I 3. Constipation: S X 4. Increased passing of gas: P 5. Nausea: F, S X 6. Vomiting: F, S X 7. Bloating of the abdomen: F, S Metabolism and nutrition disorders 8. Decreased appetite: S, I X Musculoskeletal and connective tissue disorders 9. Aching joints; aching muscles; X back pain: F, S, I Psychiatric disorders 10. Insomnia: S, I X 11. Anxiety: F, S, I 12. Depression†: F, S, I 13. Irritability: F, S, I X 14. Problems with anger: F, S, I 15. Decreased sexual interest: S Renal and urinary disorders 16. Urinary problems†: F, I X 17. Pain or burning with urination: S Nervous system disorders 18. Numbness or tingling in your X hands or feet: S, I 19. Dizziness: S, I X 20. Headache: F, S, I X 21. Problems with memory: S, I X 22. Problems with concentration: S, I X Reproductive system and breast disorders 23. Irregular menstrual periods: P X 24. Vaginal dryness: S X 25. Pain during vaginal sex: S X 26. Breast area enlargement or X tenderness: S Skin and subcutaneous tissue disorders 27. Unexpected or excessive X sweating: F, S 28. Dry skin; Itchy skin: S Vascular disorders 29. Hot flashes: F, S X General disorders 30. Fatigue, tiredness, or lack of energy: S, I
MENCAV [3]
MENQOL [8]
X
WHQ [46]
WHQ-R [7]
MRS [11]
X
X
X
UQOL [11]
MQOL [10]
Greene [47]
X
X X
X
X
X
X X X X
X X X X
X
X
X
X
X
X
X
X
X
X
X X X X
X X X
X X X X
X
X
X
X
X
X
X X
X X
X X X X X
X X X X X
X
X X X X X
X
X
X
X X X
X X X X
X X X X
X X X X
X X
X X
X
X
X
X
X
X
X
X
X
X X X X
X
X X
X
X
X
X
X
X
X
X
X
X
X
BESS, Breast Cancer Prevention Trial Eight Symptom Scale; HRQOL, health-related quality of life; MENCAV, Calidad de Vida en Menopausia (Quality of life during menopause); MENQOL, Menopause-Specific Quality of Life; MQOL, Menopausal Quality of Life Scale; MRS, Menopause Rating Scale; PRO-CTCAE, Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events; UQOL, Utian Menopause Quality-of-Life Score; WHQ, Women’s Health Questionnaire; WHQ-R, Women’s Health Questionnaire–Revised. * The MP-30D includes 30 checkboxes (1 for each domain) and 72 five-level items describing symptom frequency (F), severity (S), and interference (I) with usual activities. Among 72 F, S, and I items (listed above), 63 originated from the PRO-CTCAE v1.0; however, the F, S, and I items for back pain, irritability, and problems with anger did not because these symptoms are not in the PRO-CTCAE v1.0. † Depression includes two symptoms (Feeling that nothing could cheer you up; Sad or unhappy feelings), and Urinary problems includes three symptoms (Frequent urination; Urges to urinate all of the sudden; Loss of control of urine).
160
VALUE IN HEALTH 19 (2016) 158–166
domains relevant to both men and women) may systematically underrepresent the burden of unfavorable health among women because women experience outcomes relating to reproduction, such as menstruation, pregnancy, breast-feeding, and menopause. Gender-specific domains, however, complicate valuation from the societal perspective. For example, this would entail asking both women and men about their preferences on menopausal symptoms as if men could experience them. Because men do not have direct experience with menopausal symptoms, however, their understanding and preferences on such symptoms is highly speculative and, thus, is of questionable merit toward informing clinical or policy recommendations on women’s health. Instead of conducting a direct valuation study of menopausal symptoms, multiple studies have indirectly linked (or mapped) validated menopausal instruments (e.g., Menopause Rating Scale) to gender-neutral health-related quality-of-life instruments that have values as a way to loosely incorporate the value of menopausal symptoms from the societal perspective [28–30]. Even though mapping has the advantage of simplification to a common, gender-neutral metric for the evaluation of all conditions and treatments, linking to gender-neutral instruments may dilute the experience of gender-specific domains. The prevalence and reliance on such mapping studies in economic evaluations, in part, motivated the approach taken in this study to ask women about their preferences directly. In addition to being the first valuation study examining menopausal symptoms, this is the first study to value more than six symptom domains. Some researchers recognize the potential benefit of including more than six domains into a PRO instrument [31–33], but others argue on practical grounds that a valuation task with more than six domains is not feasible [29]. This argument is typically advanced in reference to conventional valuation tasks (e.g., time trade-off and standard gamble), in which all domains must be presented at once. This study used paired comparisons under a partial profile approach, similar to previous work [21], which allowed for separately valuing domains and permitted valuation of more than six domains, with multiple items per domain. Overall, this study included approximately half the items in the PRO-CTCAE. The aim of this study was to assess the preferences of women aged 40 to 69 years with respect to menopausal symptoms, such as hot flashes, fatigue, and sleep disturbances. We examined two questions: 1) Does the value placed on symptom relief vary across symptoms? 2) Does the value placed on symptom relief vary as a
Fig. 1
function of symptom frequency, severity, and interference with usual or daily activities? Using a nationally representative sample of women aged 40 to 69 years, we conducted an online survey that included the MP-30D and 30 paired comparisons that asked women which health they preferred (Fig. 1). This is the first study to create a preference-based measure of menopausal symptoms on a quality-adjusted life-year (QALY) scale, where a loss of 1 QALY is equivalent to a 1-year reduction in lifespan with no health problems. Ultimately, this measure may be used to advance women’s health care by enhancing comparative effectiveness research, health technology assessment, and clinical policy recommendations as related to the management of menopausal symptoms and efforts to improve women’s health across the life course.
Methods Introduction to the MP-30D Designed for this study, the MP-30D is an adaptive checklist that measures 30 domains of menopausal symptoms that were identified through a review of existing PRO instruments to assess health concerns and experiences of menopausal women (Table 1). Each of the 30 checkboxes routed respondents to additional domain-specific items. Among the 30 domains, 27 domains routed to PRO-CTCAE items (63 of the 126 PRO-CTCAE items). The PRO-CTCAE, however, did not include items for the three remaining domains: back pain, irritability, and problems with anger. Following the format of the PRO-CTCAE, these three new symptoms were created with a five-level response scale, with higher values indicating greater symptom frequency, severity, or interference. This domain-routing approach was developed to handle the multidimensionality of menopausal symptoms without imposing an excessive burden on respondents. We assumed that the checkbox captured the presence of a menopausal symptom domain, and the items located the respondent along the latent domain on the basis of item responses (e.g., frequency, severity, and interference) [34,35]. If a respondent did not check a domain box, no questions were asked for that domain under the assumption that domain measurement was not relevant to the respondent’s experience. For the paired comparisons, each MP-30D response described a health outcome using adjectival statements [22,36]. An adjectival statement is an item response transformed into a state-
VALUE IN HEALTH 19 (2016) 158–166
ment. For example, the “Mild” response to “What was the severity of your dry skin at its worst?” is transformed into “Dry skin, mild at its worst.” The MP-30D has 72 items (2 presence items [yes/no] and 70 five-level items) that describe 282 changes in level (2 1 þ 70 4). The first change in frequency (from level 1 to 2), however, is inferred from the other attributes because having any severity or interference requires nonzero frequency, reducing the available number of changes to 263 (282 – 19).
Valuation of Symptoms In this study, utility refers to a random latent trait at the individual level that governs a person’s choice (i.e., episodic random utility model) [37] and value refers to a preferencebased measure representing the choices of a group of individuals [21]. The probability of a choice between two independent episodes, A and B, depends on their attributes and may vary because of intrapersonal variability or respondent heterogeneity [38,39]. Specifically, episodes A and B have the same value (VA ¼ VB) if and only if exactly half the respondents choose A over B. For example, if half the respondents preferred 30 days with insomnia at its worst over a loss of 90 days with no health problems (i.e., 90/365 ¼ 0.25 QALY; Fig. 2), this implies that the value of 30 days with insomnia at its worst is a loss of 0.25 QALY.
161
paired comparisons, each respondent completed three examples followed by eight pairs that asked respondents to choose between one of the symptoms (e.g., insomnia) and a loss in lifespan with no health problems (i.e., lifespan pairs; Fig. 1). The remaining 22 pairs asked respondents to choose between two symptoms (i.e., health pairs). All pairs were randomly ordered to reduce possible sequence effects. The follow-up component asked about the respondent’s menstrual patterns, parity, and current use of hormone replacement therapy and offered an opportunity to leave survey feedback.
Pair Selection
Participants were women aged 40 to 69 years recruited from a preexisting national panel of US adults. To promote concordance with the 2010 US Census, we used six demographic quotas (two age groups, three race/ethnicity groups; 2 3 ¼ 6). Once a quota was filled, no additional respondents belonging to that quota were accepted. The survey was administered online from April 3 to April 21, 2013. The approach and methods, including its sampling design and survey instrument, were adapted from the Patient-Reported Outcomes Measurement Information System-29 valuation study (1R01CA160104) [40] and approved by the University of South Florida Institutional Review Board (USF IRB no. 8236) [41].
The Appendix (available as Supplemental Materials found at http://dx.doi.org/%2010.1016/j.jval.2015.11.002) provides a didactic overview of paired comparisons, adjectival statements, and pair selection as well as a few econometric concepts that may aid the reader. Originally, the sample size was 3000 respondents with 50 responses per pair, which allowed for 480 lifespan pairs (3000 8/50) and 1320 health pairs (3000 22/50). For the lifespan pairs, eight symptom domains were selected by the investigators as potentially worth a loss in lifespan. Losses in lifespan were paired with increases in these domains from level 1 (none) or 2 (mild) to level 5 (severe). Each symptom domain was experienced for three possible durations—7 days, 14 days, or 30 days—and was traded for five gains in lifespan (2 days, 7 days, 14 days, 30 days, and 90 days) with a delay that ranged from 1 year (e.g., 2 days less than 1 year) to 5 years (e.g., 2 days less than 5 years). In total, 480 lifespan pairs were created (8 2 3 5 2). The selection of the 1320 health pairs was more complicated because of the correlation structure between items. Specifically, we selected three types of health pairs: 382 interitem, 80 intersymptom, and 858 interdomain. The interitem pairs present tradeoffs between distinct aspects of the symptom experience (e.g., severity vs. interference for insomnia), the intersymptom pairs present tradeoffs of symptoms within a domain (e.g., feeling that nothing could cheer you up vs. sad and unhappy feelings for depression), and the interdomain pairs present domain tradeoffs (e.g., insomnia vs. dizziness).
Survey Instrument
Econometrics
After consenting, respondents completed a screener including demographic and socioeconomic status questions [42] and then proceeded to the survey, which began with a single item eliciting global, self-rated health and the MP-30D followed by paired comparisons and follow-up components. At the start of the
To simplify the analysis, only the 1480 pairs for which symptoms were experienced for 30 days were included in the analytical sample (see Appendix in Supplemental Material). For each of the pairs, the sample probability of choosing A over B, pk, was approximately normally distributed by the central limit theorem and served as the dependent variable of a generalized linear model. Each alternative (A and B) was represented by a linear regression, dh, that included 263 coefficients, 1 for each indicator variable of change in level (e.g., 1 to 2). The coefficients of the generalized linear model were estimated by minimizing the weighted sum of squared error,
Participants
1X 480
ðPðAk 4 Bk Þpk Þ2 =σ 2k
k¼1
where P(Ak 4 Bk) ¼ dB/(dA þ dB) and σ2k ¼ pk (1 – pk)/nk [43,44]. Confidence intervals were estimated by percentile bootstrap with pair-quota stratification and 1000 resampling iterations.
Results
Fig. 2
Of the 4211 respondents recruited for this study, 492 (12%) did not complete the survey and 322 (7.6%) were terminated (e.g., survey timed out at 2 hours). There were no associations between demographic characteristics (i.e., age, race, and ethnicity) or
162
VALUE IN HEALTH 19 (2016) 158–166
income and incomplete surveys. As indicated in Table 2, the sample characteristics were generally concordant with the 2010 US Census estimates, although the sample had slightly more Hispanic respondents and more homogeneous distributions with respect to educational attainment and household income. Among the 3397 respondents, 15% (n ¼ 514) reported that they had not experienced any of the 30 symptoms over the past 7 days. For the remaining (85%; n ¼ 2883), the median number of symptoms experienced was 3 out of 30 possible symptoms with an interquartile range of one to six symptoms. The four most prevalent symptom domains were aching joints, aching muscles, or back pain (45%); fatigue, tiredness, or lack of energy (40%); headache (34%); and insomnia (31%). For the lifespan pairs, 1753 (52%) respondents always preferred a reduced lifespan (rather than experiencing a domain of menopausal symptoms at its worst for 30 days) and 36 (1%) always preferred experiencing symptoms rather than have a reduced lifespan. This lack of variability suggests that future studies may wish to incorporate a higher range of reductions in lifespans (490 days) or milder symptoms with shorter durations. For health pairs, five (o1%) respondents always preferred the left and three (o1%) always preferred the right. Across all pairs, the sample sizes for each pair ranged from 49 to 68 respondents (available as Appendix in Supplemental Materials).
Figure 2 shows the results of all 480 lifespan pairs. In these pairs, loss in lifespan represents the “price” women would have to pay for symptom relief. The downward sloping curves (i.e., demand curves) imply that as the price of symptom relief drops the proportion of women who prefer symptom relief increases. More than 60% of the women prefer to reduce their lifespan by 90 days to avoid the eight domains of menopausal symptoms at their worst for 30 days (Fig. 2), which suggests that women place a high value on relief from symptoms that commonly occur during menopause. Based on this evidence alone, relief from insomnia and constipation were the least valuable of the eight domains and relief from headache and vomiting were the most valuable. To better understand QALY calculation using paired comparison results, consider that 37 (63%) out of 59 respondents preferred a loss of 90 days of lifespan over experiencing insomnia at its worst for 30 days (Fig. 1). With this single result and the cumulative distribution function, P(Ak 4 Bk) ¼ dB/(dA þ dB), we learned that the value of 30 days of insomnia equals a loss of 153 days of reduced lifespan (i.e., 0.42 QALY): 30 days of insomnia/(30 days of insomnia þ 90 days of reduced lifespan) ¼ 63% 30 days of insomnia ¼ 63% (30 days of insomnia þ 90 days of reduced lifespan)
Table 2 – Sample demographics by completion status. Characteristic
Age (y) 40–54 55–69 Race White Black or African American American Indian or Alaska Native Asian Native Hawaiian or other Pacific Islander Some other race Two or more races Hispanic ethnicity Hispanic or Latino Not Hispanic or Latino Educational attainment* Less than high school High school graduate Some college, no degree Associate’s degree Bachelor’s degree Graduate or professional degree Refused/don’t know Household income ($) r14,999 15,000–24,999 25,000–34,999 35,000–49,999 50,000–74,999 75,000–99,999 100,000–149,999 Z150,000 Refused/don’t know
n (%)
P value
US 2010 Census (%)
Dropout (N ¼ 492)
Terminated (N ¼ 322)
Completed (N ¼ 3397)
264 (53.66) 228 (46.34)
173 (53.73) 149 (46.27)
1846 (54.34) 1551 (45.66)
0.944
56.70 43.30
399 (81.10) 77 (15.65) 2 (0.41) 1 (0.20) 4 (0.81)
264 (81.99) 55 (17.08) – – –
2837 (83.51) 462 (13.60) 15 (0.44) 16 (0.47) 22 (0.65)
0.356
76.70 12.10 0.80 4.80 0.10
– 9 (1.83)
– 3 (0.93)
– 45 (1.32)
75 (15.24) 417 (84.76)
58 (18.01) 264 (81.99)
521 (15.34) 2876 (84.66)
0.441
11.20 88.80
22 (4.47) 225 (45.73) 96 (19.51) 44 (8.94) 95 (19.31) 9 (1.83) 1 (0.20)
4 (1.24) 88 (27.33) 68 (21.12) 54 (16.77) 100 (31.06) 8 (2.48) –
127 (3.74) 45.45 (1544) 617 (18.16) 323 (9.51) 738 (21.73) 46 (1.35) 2 (0.06)
o0.001
11.40 28.80 22.60 9.40 17.20 10.07
22 (4.47) 43 (8.74) 68 (13.82) 113 (22.97) 73 (14.84) 55 (11.18) 40 (8.13) 16 (3.25) 62 (12.60)
14 (4.35) 31 (9.63) 47 (14.60) 79 (24.53) 54 (16.77) 29 (9.01) 31 (9.63) 11 (3.42) 26 (8.07)
143 (4.21) 298 (8.77) 475 (13.98) 720 (21.20) 579 (17.04) 386 (11.36) 333 (9.80) 150 (4.42) 313 (9.21)
0.585
13.40 11.50 10.80 14.20 18.30 11.80 11.80 8.10
3.80 1.60
* Educational attainment and household income comes from the 2010 American Community Survey 1-Year Estimates [48,49].
VALUE IN HEALTH 19 (2016) 158–166
37% 30 days of insomnia ¼ 63% 90 days of reduced lifespan 30 days of insomnia ¼ 63% 90 days of reduced lifespan/37% 30 days of insomnia ¼ 153 days of reduced lifespan Still, this is just 1 of 60 lifespan pairs for the insomnia domain. In another lifespan pair, we learned that 44 (81%) out of 54 respondents preferred a loss of 30 days of lifespan over experiencing 30 days with insomnia at its worst severity and interference, which implies that 1 month of insomnia equals 128 days of reduced lifespan (81% 30/(1 – 81) ¼ 128 days or 0.35 QALY). Using the responses from 1480 pairs in a generalized linear model estimation, Table 3 describes the value of menopausal symptom relief from the perspective of women aged 40 to 69 years. For example, the value of constipation relief on a QALY scale increases on the basis of its severity: mild at its worst (0.296), moderate at its worst (0.296 þ 0.293 ¼ 0.589), severe at its worst (0.589 þ 2.942 ¼ 3.531), and very severe at its worst (3.531 þ 0.114 ¼ 3.645). Respondents were indifferent between 30 days of relief from very severe constipation and a 109-day increase in lifespan (3.645 30). Based on these results, relief from each symptom significantly increases the quality of women’s lives (P o 0.05), except a few that were insignificant (15 of 263 [6%]). Although symptom relief is clearly preferred over experiencing symptoms, the value women place on symptom relief ranged widely across the 30 domains. Figure 3 shows the ranking of the 15 “least desirable” symptom domains and depicts variation in the value of relieving each of their symptom attributes (frequency, severity, and interference).
Conclusions This valuation study focused on the MP-30D, a new adaptive checklist for the measurement of menopausal symptoms, integrating items from the PRO-CTCAE. In an effort to enhance this new measure for comparative effectiveness research, the 263 changes in symptoms captured by the MP-30D were valued on a QALY scale, so researchers can better understand the relative value women place on symptom relief. In the process of accomplishing these objectives, this is the first study to 1) examine preferences on symptoms described using the PRO-CTCAE; 2) assess the values of women nationally; and 3) distinguish the values placed on reducing symptom frequency, severity, and interference. Results show that from the perspective of women aged 40 to 69 years, the relief in the 30 symptom domains is highly valuable and that a reduction in symptom frequency— particularly for nausea, vomiting, headaches, anger, and abdominal pain—is preferred over improvements in severity and interference. On the basis of our findings, most of the women would prefer to reduce their lifespan than to experience menopausal symptoms, even at low levels. As the first preference study of the PRO-CTCAE (including 63 of the 126 items), these results provide insights into its symptom attributes (frequency, severity, and interference). First, the relief from low levels of symptoms is valuable (e.g., grade 1), particularly with respect to dizziness, headaches, problems with memory and concentration, and urinary pain and incontinence. This finding confirms the importance of aggressive symptom relief efforts, even when symptoms may not seem, at face value, to be prominent or highly distressing. Second, symptoms that worsen into grade 3 events are highly detrimental, such as moderate severity shifting to severe or very severe, and interference shifting from somewhat to quite a bit or very much. This observation may point toward clinically meaningful cutoff points
163
for the PRO-CTCAE items and may hint at potential mappings to serious AEs. Third, PRO-CTCAE attributes have different values, which suggest nonoverlapping information about their experiences. This study examines the perspectives of women, yet we recognize that some researchers strongly recommend that valuation be conducted from the societal perspective [45]. Such a study would require asking men to value symptoms relating to menopause. Many menopausal symptoms may be experienced by men (e.g., nausea), but not all (e.g., vaginal dryness). This inexperience alone might dilute the weights, and the focus of this study was on the perspective of women, specifically those with experience of one (or more) of these symptoms. Instead of imposing the values of men on women, some men may prefer to abdicate this authority to women. Others may feel obligated to express their view because of the implications for communal resources and religious considerations (e.g., reproductive rights). Out of a sense of fairness in measurement and valuation, conventional practice is to ignore evidence on gender-specific domains in QALY calculations. Some researchers believe that valuing gender-specific domains is unnecessary, impractical, and unfair. Regardless of one’s views on this practice, this article will help researchers and policymakers to better understand the value women place on symptom relief. Future analyses may explore heterogeneity in preferences among women, including differences between women of different ages, experiences, and dependents. Women may have experienced and adapted to symptoms, which may alter the value they place on symptom relief. The value a woman places on symptom relief (e.g., anger) may also incorporate the benefits for those around her as well as her capacity to fulfill commitments and roles, including responsibilities at work and at home, particularly when frequent, severe, or debilitating symptoms (e. g., headaches) can make it difficult to engage in family and social relationships or may limit vocational functioning. Subsequent studies will examine such heterogeneity as well as the values of younger and older women (and men, where appropriate). This study is the first to assess the value of relieving symptoms that are commonly experienced during menopause, but it did not value all aspects of menopause, which may include physiologic changes, such as diminished fertility. Future studies may incorporate additional symptoms as well as outcomes of shorter or longer durations. A fundamental limitation of this study is the use of the QALY scale, which assumes that the value of a symptom is strictly proportional to its duration and independent of one’s lifespan. A more subtle limitation of this article (and most health valuation studies) is an implicit assumption that the value of relieving individual symptoms is additive and that the presence of concurrent multiple symptoms has no interaction effect on the value of relieving other symptoms. We acknowledge that the experience of multiple concurrent symptoms may magnify or reduce the value of relieving one or more symptoms. Alternative model specifications that accommodate nonadditive relationships and the value of relieving concurrent symptoms can be examined in future work. To understand the impact of chemotherapy, injury, or other life-changing events that can induce premature menopause requires clinical investigations that include measures of symptomatic toxicities such as the PRO-CTCAE or symptom instruments, such as the MP-30D. Nevertheless, this study investigates the preferences of midlife women to avoid symptoms and enhances our understanding of the value of symptom relief. These strong preferences toward symptom relief underscore the importance of gender-specific valuation and motivate efforts to evaluate new treatment approaches to manage symptoms effectively in women at midlife.
164
VALUE IN HEALTH 19 (2016) 158–166
Table 3 – The values women place on menopausal symptoms on a quality-adjusted life-year (QALY) scale. Symptoms arranged by MP-30D domains*
Change in level 1 to 2
Pounding or racing heartbeat† Pain in the abdomen
Constipation‡ Increased passing of gas Nausea Vomiting Bloating of the abdomen Decreased appetite Aching joints
Aching muscles
Back pain
Insomnia Anxiety
Feel that nothing could cheer you up
Sad or unhappy feelings
Irritability
Problems with anger
Decreased sexual interest Frequent urination† Urges to urinate all of a sudden Loss of urinary control Urinary pain Numbness or tingling in hands or feet Dizziness Headaches
Problems with memory Problems with concentration
F S F S I S P F S F S F S S I F S I F S I F S I S I F S I F S I F S I F S I F S I S F I F I F I S S I S I F S I S I S I
0.541 0.291 0.324 0.296
0.416 0.825 0.234 0.017 0.030 0.030 0.170 0.018 0.099 0.033 0.159 0.209 0.139 0.229 0.439 0.273 0.340 0.107 0.406 0.143 0.238 0.413 0.689 0.205 0.160 0.192 0.701 1.144 0.218 0.225 0.485 0.429 0.087 0.265 0.654 0.547 0.270 0.341
2 to 3
3 to 4
0.351 0.247 0.233 0.372 0.373 0.293
0.747 2.365 2.829 1.719 2.424 2.942
0.380 0.156 1.237 0.231 0.173 0.192 0.014 0.023 0.056 0.052 0.012 0.030 0.054 0.033 0.070 0.090 0.048 0.151 0.131 0.265 0.312 0.273 0.392 0.169 0.177 0.254 0.297 0.04 0.183 0.105 0.203 0.797 0.674 0.677 0.017 0.054 0.040 0.125 0.033 0.427 0.058 0.444 0.187 0.134 0.541 0.246 0.120 0.402 0.164 0.626 0.506 0.328 0.241
4 to 5
3.388 1.515 5.399 1.620 1.750 0.114 0.423 2.758 2.949 1.216 0.896 9.932 8.662 2.012 1.715 0.579 0.689 0.706 0.431 0.049 0.057 0.177 0.049 0.246 0.301 0.055 0.263 0.437 0.200 0.195 0.350 0.149 0.148 0.400 0.122 0.715 0.990 0.235 0.564 1.074 0.366 0.874 0.600 1.231 0.435 1.924 2.111 0.717 1.220 1.593 1.116 0.815 1.139 0.509 0.477 0.984 0.482 1.018 0.947 0.208 0.496 1.209 0.592 1.145 1.082 0.542 0.310 1.063 0.608 5.973 4.662 1.796 1.032 3.677 1.299 0.279 0.001 0.116 0.311 0.303 0.120 0.177 0.376 0.304 0.166 1.259 0.948 0.666 0.364 4.080 0.001 0.695 0.582 1.257 0.372 3.806 0.566 3.617 0.410 2.581 5.765 1.824 1.114 1.581 1.838 3.738 2.305 3.526 1.093 1.096 1.036 1.629 0.728 continued on next page
165
VALUE IN HEALTH 19 (2016) 158–166
Table 3 – continued. Symptoms arranged by MP-30D domains*
Irregular menstrual periods Vaginal dryness Pain during vaginal sex Breast area enlargement and pain Unexpected or excessive sweating Dry skin Itchy skin Hot flashes Fatigue, tiredness, or lack of energy
Change in level
P S S S F S S S F S S I
1 to 2
2 to 3
3 to 4
4 to 5
0.145 0.857 0.335
0.058 0.180 0.059 0.104 0.082 0.023 0.033 0.082 0.053 0.262 0.133
0.458 1.582 0.734 0.247 0.278 0.045 0.447 0.348 0.267 0.684 1.584
0.175 0.012 0.277 0.166 0.803 0.288 0.001 0.001 0.765 0.294 0.316 0.268
0.237 0.018 0.167 0.114 0.203 0.310
* Presence (P) is a two-level response scale (No/Yes). The F, S, and I items for back pain, irritability, and problems with anger are not in the PROCTCAE v1.0. Frequency (F), severity (S), and interference with usual activities (I) on a five-level response scale where higher level indicated greater symptoms. The labels depend on the attribute: frequency increases from never (1) to rarely (2), to occasionally (3), to frequently (4), and to almost constantly (5); severity increases from none at its/their worst (1) to mild at its/their worst (2), to moderate at its/their worst (3), to severe at its/their worst (4), and to very severe at its/their worst (4); and interference increases from not at all (1) to a little bit (2), to somewhat (3), to quite a bit (4), and to very much (5). † The level change from 1 to 2 in frequency (i.e., none to rarely) is implied by all other level changes for the domain; therefore, it is not shown. ‡ For example, the value of constipation relief on a QALY scale increases on the basis of its severity: mild at its worst (0.296), moderate at its worst (0.296 þ 0.293 ¼ 0.589), severe at its worst (0.589 þ 2.942 ¼ 3.531), and very severe at its worst (3.531 þ 0.114 ¼ 3.645). Therefore, respondents were indifferent between 30 d of relief from very severe constipation or 109-d increase in lifespan (3.645 30).
Fig. 3
Acknowledgments We thank Michelle Owens, Janel Phetteplace, and Carol Templeton at Moffitt Cancer Center for their contributions to the research and creation of this article. Source of financial support: Funding support for this research was provided by Dr. Craig’s support account at Moffitt Cancer Center.
Supplemental Materials Supplemental material accompanying this article can be found in the online version as a hyperlink at http://dx.doi.org/10.1016/j.
jval.2015.11.002 or, if a hard copy of article, at www.valueinhealth journal.com/issues (select volume, issue, and article).
R EF E R EN C ES
[1] Avis NE, Colvin A, Bromberger JT, et al. Change in health-related quality of life over the menopausal transition in a multiethnic cohort of middle-aged women: Study of Women’s Health Across the Nation. Menopause 2009;16:860–9. [2] Bromberger JT, Kravitz HM, Chang YF, et al. Major depression during and after the menopausal transition: Study of Women’s Health Across the Nation (SWAN) (vol 41, pg 1879, 2011). Psychol Med 2011;41: 2238–8. [3] Bermejo JB, Martinez JAV, Saiz AR, et al. Validation of a menopause quality of life scale: the MENCAV scale. Maturitas 2008;59:28–37.
166
VALUE IN HEALTH 19 (2016) 158–166
[4] Cella D, Land SR, Chang CH, et al. Symptom measurement in the Breast Cancer Prevention Trial (BCPT) (P-1): psychometric properties of a new measure of symptoms for midlife women. Breast Cancer Res Treat 2008;109:515–26. [5] Chattha R, Kulkarni R, Nagarathna R, et al. Factor analysis of Greene’s Climacteric Scale for Indian women. Maturitas 2008;59:22–7. [6] Chedraui P, Aguirre W, Hidalgo L, et al. Assessing menopausal symptoms among healthy middle aged women with the Menopause Rating Scale. Maturitas 2007;57:271–8. [7] Girod I, de la Loge C, Keininger D, et al. Development of a revised version of the Women’s Health Questionnaire. Climacteric 2006;9:4–12. [8] Hilditch JR, Lewis J, Peter A, et al. A menopause-specific quality of life questionnaire: development and psychometric properties. Maturitas 1996;24:161–75. [9] Hunter M. The Womens Health Questionnaire—a measure of mid-aged womens perceptions of their emotional and physical health. Psychol Health 1992;7:45–54. [10] Jacobs PA, Hyland ME, Ley A. Self-rated menopausal status and quality of life in women aged 40-63 years. Br J Health Psychol 2000;5:395–411. [11] Utian WH, Janata JW, Kingsberg SA, et al. The Utian Quality of Life (UQOL) Scale: development and validation of an instrument to quantify quality of life through and beyond menopause. Menopause 2002;9:402–10. [12] Zollner YF, Acquadro C, Schaefer M. Literature review of instruments to assess health-related quality of life during and after menopause. Qual Life Res 2005;14:309–27. [13] Basch EM, Reeve BB, Mitchell SA, et al. Electronic toxicity monitoring and patient-reported outcomes. Cancer J 2011;17:231–4. [14] Basch EM, Reeve BR, Mitchell SA, et al. Development of the National Cancer Institute’s Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). J Natl Cancer Inst 2014;106. [15] Hays RD, Spritzer KL, Amtmann D, et al. Upper-extremity and mobility subdomains from the Patient-Reported Outcomes Measurement Information System (PROMIS) adult physical functioning item bank. Arch Phys Med Rehabil 2013;94:2291–6. [16] National Cancer Institute. Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). National Cancer Institute, Rockville, Maryland, USA, 2010. [17] National Cancer Institute. Appendix A–Patient-Reported OutcomesCommon Terminology Criteria for Adverse Events (PRO-CTCAE) Version v1.0 Items. National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Rockville, MD, USA, 2013. [18] Mendoza T, Atkinson T. PRO-CTCAE Face-To-Face Meeting #2: Development of PRO-CTCAE Items. National Cancer Institute, Rockville, MD, USA, 2010. [19] Brooks R, Rabin R, De Charro F. The Measurement and Valuation of Health Status Using EQ-5D: A European Perspective: Evidence from the EuroQol BIO MED Research Programme. Springer, Dordrecht, The Netherlands, 2003. [20] Furlong WJ, Feeny DH, Torrance GW, et al. The Health Utilities Index (HUI (R)) system for assessing health-related quality of life in clinical studies. Ann Med 2001;33:375–84. [21] Craig BM, Reeve BB, Cella D, et al. Demographic differences in health preferences in the United States. Med Care 2014;52:307–13. [22] Craig BM, Pickard AS, Stolk E, Brazier JE. US valuation of the SF-6D. Med Decis Mak 2013;33:793–803. [23] Ratcliffe J, Couzner L, Flynn T, et al. Valuing Child Health Utility 9D health states with a young adolescent sample: a feasibility study to compare best-worst scaling discrete-choice experiment, standard gamble and time trade-off methods. App Health Econ Health Pol 2011;9:15–27. [24] Erickson P. Evaluation of a population-based measure of quality of life: the Health and Activity Limitation Index (HALex). Qual Life Res 1998;7:101–14.
[25] Fryback DG, Dunham NC, Palta M, et al. US norms for six generic health-related quality-of-life indexes from the National Health Measurement Study. Med Care 2007;45:1162–70. [26] Fryback DG, Palta M, Cherepanov D, et al. Comparison of 5 healthrelated quality-of-life indexes using item response theory analysis. Med Decis Mak 2010;30:5–15. [27] Andresen EM, Rothenberg BM, Kaplan RM. Performance of a selfadministered mailed version of the Quality of Well-Being (QWB-SA) questionnaire among older adults. Med Care 1998;36:1349–60. [28] Brazier J, Roberts J, Platts M, et al. Estimating a preference-based index for a menopause specific health quality of life questionnaire. Health Qual Life Outcomes 2005;3:13. [29] Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271–92. [30] Schneider HPG, MacLennan AH, Feeny D. Assessment of health-related quality of life in menopause and aging. Climacteric 2008;11:93–107. [31] Brazier JE, Rowen D, Mavranezouli I, et al. Developing and testing methods for deriving preference-based measures of health from condition-specific measures (and other patient-based measures of outcome). Health Technol Assess (Winchester, England) 2012;16:1–114. [32] Kuspinar A, Finch L, Pickard S, et al. Using existing data to identify candidate items for a health state classification system in multiple sclerosis. Qual Life Res 2013;23:1445–57. [33] Krabbe PFM, Stouthard MEA, Essink-Bot ML, et al. The effect of adding a cognitive dimension to the EuroQol multiattribute health-status classification system. J Clin Epidemiol 1999;52:293–301. [34] Wang X, Bradlow ET, Wainer H. User’s Guide for SCORLIGHT (Version 3.0): A Computer Program for Scoring Tests Built of Testlets Including a Module for Covariate Analysis. Educational Testing Service, Princeton, New Jersey, USA, 2005. [35] Thissen D, Steinberg L, Mooney JA. Trace lines for testlets—a use of multiple-categorical-response models. J Educ Meas 1989;26:247–60. [36] Craig B, Owens MA. Methods Report on the Women’s Health Valuation (WHV): Year 1. Moffitt Cancer Center, Tampa, Florida, USA, 2013. [37] Craig BM, Busschbach JJ. The episodic random utility model unifies time trade-off and discrete choice approaches in health state valuation. Popul Health Metr 2009;7:3. [38] Bansback N, Brazier J, Tsuchiya A, et al. Using a discrete choice experiment to estimate health state utility values. J Health Econ 2012;31:306–18. [39] Hadorn DC, Hays RD, Uebersax J, et al. Improving task comprehension in the measurement of health state preferences—a trial of informational cartoon figures and a paired-comparison task. J Clin Epidemiol 1992;45:233–43. [40] Craig BM, Schell MJ, Brown PM, et al. HRQoL Values for Cancer Survivors: Enhancing PROMIS Measures for CER. H. Lee Moffitt Cancer Center, National Institutes of Health, Rockville, MD, USA, 2011. [41] Craig BM, Reeve BB, Brown PM, et al. US valuation of health outcomes measured using the PROMIS-29. Value Health 2014;17:846–53. [42] U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau. United States Census 2010. In: U.S. Department of Commerce Economics and Statistics Administration, ed., National Processing Center, Jeffersonville, Indiana, USA, 2010. [43] McFadden D. Conditional logit analysis of qualitative choice behavior. In: Zarembka P, ed. Frontiers in Econometrics. New York: Academic Press, 1974. [44] Urban FM. Urban’s solution (minimum normit X2). In: Bock RD, Jones LV, eds. The Measurement and Prediction of Judgment and Choice. San Francisco: Holden-Day, 1968. [45] Gold M, Siegel J, Russell L, et al. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press, 1996. [46] Hunter M. The Women’s Health Questionnaire (WHQ): the development, standardization and application of a measure of mid-aged women’s emotional and physical health. Qual Life Res 2000;9:733–8. [47] Greene JG. Constructing a standard climacteric scale. Maturitas 1998;29:25–31.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jval
Examining the Value of Menopausal Symptom Relief Among US Women Benjamin M. Craig, PhD1,*, Sandra A. Mitchell, PhD, CRNP2 1 Health Outcomes and Behavior, Moffitt Cancer Center and University of South Florida, Tampa, Florida, USA; 2Outcomes Research Branch, Healthcare Delivery Research Program, Division of Cancer Control & Population Sciences, National Cancer Institute, Rockville, Maryland, USA
AB STR A CT
Background: Menopausal symptoms can cause significant distress to women, yet little is known about the value women place on these symptoms. Methods: In April 2013, 3397 US women, aged 40 to 69 years, completed an online survey that included 30 paired comparisons. Specifically, respondents were shown two menopausal symptoms described using the Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events and asked, “Which do you prefer?” From their choices, we estimated a generalized linear model to assess the values women place on symptom relief in terms of quality-adjusted life-years (QALYs). Results: Approximately half the respondents (1753 of 3397 [52%]) always preferred a reduced lifespan (up to 90 days) instead of experiencing menopausal symptoms at their worst for 30 days. For most of the symptoms (248 of 263 [94%]), including low-grade events, QALYs were significantly reduced (P o 0.05). The value women placed on relief ranged widely by symptom domain: the relief from depression, problems with memory, headache, pain in abdomen, problems with anger, and vomiting were
the most valuable. Conclusions: Overall, the value women place on menopausal symptom relief is surprisingly high. As the first national study to directly ask women about their preferences and to estimate the value of menopausal symptom relief on a QALY scale, this work provides critical evidence for health outcomes research in midlife women and can be applied in the evaluation of treatments that reduce or eliminate menopausal symptoms. This work also provides proof-of-concept for an approach to value Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events responses on a QALY scale. Keywords: menopause, Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE), preference, QALY, quality-adjusted life-year, symptoms, women’s health.
Introduction
events (AEs) [13–15]. This system is intended to complement National Cancer Institute’s CTCAE, an existing lexicon of clinician-reported AE items required for use in all National Cancer Institute–sponsored trials. The PRO-CTCAE item library is composed of 126 items that evaluate the presence, frequency, severity, and interference with usual or daily activities associated with 78 symptomatic toxicities [16–18]. Given PRO-CTCAE’s focus on self-reported AEs relating to treatment, it was particularly well suited for the creation of the MP-30D, a PRO checklist for the measurement of symptoms related to menopause. The primary focus of this article was a health valuation study of the MP-30D from the perspective of US women aged 40 to 69 years. Health valuation studies are commonplace throughout the clinical literature; yet to our knowledge, this is the first study that asked women about their preferences regarding symptoms associated with menopause [19–23]. Nearly all past studies focused on gender-neutral PRO instruments (e.g., the EuroQol five-dimensional questionnaire), most taking the societal perspective [24–27]. This reliance on gender-neutral domains (i.e., those
Menopause reflects not only the cessation of reproductive capacity but also often includes a multitude of associated symptoms that can negatively affect women’s health-related quality of life [1,2]. Whether part of the natural aging process, an adverse effect of medical treatment, or damage to the ovaries, the symptoms that accompany menopause can be measured using one of the available patientreported outcome (PRO) instruments (Table 1) [3–12]. No study, however, has asked women which symptoms and symptom attributes (frequency, severity, and interference) matter most to them. Understanding what women experience and what they value in terms of symptoms is critical for women’s health care, including the evaluation of new health technologies (e.g., Duavee), the assessment of treatment regimens (e.g., chemotherapy), and other forms of comparative effectiveness research. The National Cancer Institute’s PRO version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE) measurement system allows patients to self-report symptomatic adverse
Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: There are no conflicts of interest. * Address correspondence to: Benjamin M. Craig, Moffitt Cancer Center, 12902 Magnolia Drive, MRC-CANCONT, Tampa, FL 33612. E-mail: benjamin.craig@moffitt.org. 1098-3015$36.00 – see front matter Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jval.2015.11.002
159
VALUE IN HEALTH 19 (2016) 158–166
Table 1 – MP-30D domains: Comparison to HRQOL instruments for menopause. MP-30D domains arranged by the CTCAE system organ class*
HRQOL instruments for menopause BESS [4]
Cardiac disorders 1. Pounding or racing heartbeat: F, S X Gastrointestinal disorders 2. Pain in the abdomen: F, S, I 3. Constipation: S X 4. Increased passing of gas: P 5. Nausea: F, S X 6. Vomiting: F, S X 7. Bloating of the abdomen: F, S Metabolism and nutrition disorders 8. Decreased appetite: S, I X Musculoskeletal and connective tissue disorders 9. Aching joints; aching muscles; X back pain: F, S, I Psychiatric disorders 10. Insomnia: S, I X 11. Anxiety: F, S, I 12. Depression†: F, S, I 13. Irritability: F, S, I X 14. Problems with anger: F, S, I 15. Decreased sexual interest: S Renal and urinary disorders 16. Urinary problems†: F, I X 17. Pain or burning with urination: S Nervous system disorders 18. Numbness or tingling in your X hands or feet: S, I 19. Dizziness: S, I X 20. Headache: F, S, I X 21. Problems with memory: S, I X 22. Problems with concentration: S, I X Reproductive system and breast disorders 23. Irregular menstrual periods: P X 24. Vaginal dryness: S X 25. Pain during vaginal sex: S X 26. Breast area enlargement or X tenderness: S Skin and subcutaneous tissue disorders 27. Unexpected or excessive X sweating: F, S 28. Dry skin; Itchy skin: S Vascular disorders 29. Hot flashes: F, S X General disorders 30. Fatigue, tiredness, or lack of energy: S, I
MENCAV [3]
MENQOL [8]
X
WHQ [46]
WHQ-R [7]
MRS [11]
X
X
X
UQOL [11]
MQOL [10]
Greene [47]
X
X X
X
X
X
X X X X
X X X X
X
X
X
X
X
X
X
X
X
X
X X X X
X X X
X X X X
X
X
X
X
X
X
X X
X X
X X X X X
X X X X X
X
X X X X X
X
X
X
X X X
X X X X
X X X X
X X X X
X X
X X
X
X
X
X
X
X
X
X
X
X X X X
X
X X
X
X
X
X
X
X
X
X
X
X
X
BESS, Breast Cancer Prevention Trial Eight Symptom Scale; HRQOL, health-related quality of life; MENCAV, Calidad de Vida en Menopausia (Quality of life during menopause); MENQOL, Menopause-Specific Quality of Life; MQOL, Menopausal Quality of Life Scale; MRS, Menopause Rating Scale; PRO-CTCAE, Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events; UQOL, Utian Menopause Quality-of-Life Score; WHQ, Women’s Health Questionnaire; WHQ-R, Women’s Health Questionnaire–Revised. * The MP-30D includes 30 checkboxes (1 for each domain) and 72 five-level items describing symptom frequency (F), severity (S), and interference (I) with usual activities. Among 72 F, S, and I items (listed above), 63 originated from the PRO-CTCAE v1.0; however, the F, S, and I items for back pain, irritability, and problems with anger did not because these symptoms are not in the PRO-CTCAE v1.0. † Depression includes two symptoms (Feeling that nothing could cheer you up; Sad or unhappy feelings), and Urinary problems includes three symptoms (Frequent urination; Urges to urinate all of the sudden; Loss of control of urine).
160
VALUE IN HEALTH 19 (2016) 158–166
domains relevant to both men and women) may systematically underrepresent the burden of unfavorable health among women because women experience outcomes relating to reproduction, such as menstruation, pregnancy, breast-feeding, and menopause. Gender-specific domains, however, complicate valuation from the societal perspective. For example, this would entail asking both women and men about their preferences on menopausal symptoms as if men could experience them. Because men do not have direct experience with menopausal symptoms, however, their understanding and preferences on such symptoms is highly speculative and, thus, is of questionable merit toward informing clinical or policy recommendations on women’s health. Instead of conducting a direct valuation study of menopausal symptoms, multiple studies have indirectly linked (or mapped) validated menopausal instruments (e.g., Menopause Rating Scale) to gender-neutral health-related quality-of-life instruments that have values as a way to loosely incorporate the value of menopausal symptoms from the societal perspective [28–30]. Even though mapping has the advantage of simplification to a common, gender-neutral metric for the evaluation of all conditions and treatments, linking to gender-neutral instruments may dilute the experience of gender-specific domains. The prevalence and reliance on such mapping studies in economic evaluations, in part, motivated the approach taken in this study to ask women about their preferences directly. In addition to being the first valuation study examining menopausal symptoms, this is the first study to value more than six symptom domains. Some researchers recognize the potential benefit of including more than six domains into a PRO instrument [31–33], but others argue on practical grounds that a valuation task with more than six domains is not feasible [29]. This argument is typically advanced in reference to conventional valuation tasks (e.g., time trade-off and standard gamble), in which all domains must be presented at once. This study used paired comparisons under a partial profile approach, similar to previous work [21], which allowed for separately valuing domains and permitted valuation of more than six domains, with multiple items per domain. Overall, this study included approximately half the items in the PRO-CTCAE. The aim of this study was to assess the preferences of women aged 40 to 69 years with respect to menopausal symptoms, such as hot flashes, fatigue, and sleep disturbances. We examined two questions: 1) Does the value placed on symptom relief vary across symptoms? 2) Does the value placed on symptom relief vary as a
Fig. 1
function of symptom frequency, severity, and interference with usual or daily activities? Using a nationally representative sample of women aged 40 to 69 years, we conducted an online survey that included the MP-30D and 30 paired comparisons that asked women which health they preferred (Fig. 1). This is the first study to create a preference-based measure of menopausal symptoms on a quality-adjusted life-year (QALY) scale, where a loss of 1 QALY is equivalent to a 1-year reduction in lifespan with no health problems. Ultimately, this measure may be used to advance women’s health care by enhancing comparative effectiveness research, health technology assessment, and clinical policy recommendations as related to the management of menopausal symptoms and efforts to improve women’s health across the life course.
Methods Introduction to the MP-30D Designed for this study, the MP-30D is an adaptive checklist that measures 30 domains of menopausal symptoms that were identified through a review of existing PRO instruments to assess health concerns and experiences of menopausal women (Table 1). Each of the 30 checkboxes routed respondents to additional domain-specific items. Among the 30 domains, 27 domains routed to PRO-CTCAE items (63 of the 126 PRO-CTCAE items). The PRO-CTCAE, however, did not include items for the three remaining domains: back pain, irritability, and problems with anger. Following the format of the PRO-CTCAE, these three new symptoms were created with a five-level response scale, with higher values indicating greater symptom frequency, severity, or interference. This domain-routing approach was developed to handle the multidimensionality of menopausal symptoms without imposing an excessive burden on respondents. We assumed that the checkbox captured the presence of a menopausal symptom domain, and the items located the respondent along the latent domain on the basis of item responses (e.g., frequency, severity, and interference) [34,35]. If a respondent did not check a domain box, no questions were asked for that domain under the assumption that domain measurement was not relevant to the respondent’s experience. For the paired comparisons, each MP-30D response described a health outcome using adjectival statements [22,36]. An adjectival statement is an item response transformed into a state-
VALUE IN HEALTH 19 (2016) 158–166
ment. For example, the “Mild” response to “What was the severity of your dry skin at its worst?” is transformed into “Dry skin, mild at its worst.” The MP-30D has 72 items (2 presence items [yes/no] and 70 five-level items) that describe 282 changes in level (2 1 þ 70 4). The first change in frequency (from level 1 to 2), however, is inferred from the other attributes because having any severity or interference requires nonzero frequency, reducing the available number of changes to 263 (282 – 19).
Valuation of Symptoms In this study, utility refers to a random latent trait at the individual level that governs a person’s choice (i.e., episodic random utility model) [37] and value refers to a preferencebased measure representing the choices of a group of individuals [21]. The probability of a choice between two independent episodes, A and B, depends on their attributes and may vary because of intrapersonal variability or respondent heterogeneity [38,39]. Specifically, episodes A and B have the same value (VA ¼ VB) if and only if exactly half the respondents choose A over B. For example, if half the respondents preferred 30 days with insomnia at its worst over a loss of 90 days with no health problems (i.e., 90/365 ¼ 0.25 QALY; Fig. 2), this implies that the value of 30 days with insomnia at its worst is a loss of 0.25 QALY.
161
paired comparisons, each respondent completed three examples followed by eight pairs that asked respondents to choose between one of the symptoms (e.g., insomnia) and a loss in lifespan with no health problems (i.e., lifespan pairs; Fig. 1). The remaining 22 pairs asked respondents to choose between two symptoms (i.e., health pairs). All pairs were randomly ordered to reduce possible sequence effects. The follow-up component asked about the respondent’s menstrual patterns, parity, and current use of hormone replacement therapy and offered an opportunity to leave survey feedback.
Pair Selection
Participants were women aged 40 to 69 years recruited from a preexisting national panel of US adults. To promote concordance with the 2010 US Census, we used six demographic quotas (two age groups, three race/ethnicity groups; 2 3 ¼ 6). Once a quota was filled, no additional respondents belonging to that quota were accepted. The survey was administered online from April 3 to April 21, 2013. The approach and methods, including its sampling design and survey instrument, were adapted from the Patient-Reported Outcomes Measurement Information System-29 valuation study (1R01CA160104) [40] and approved by the University of South Florida Institutional Review Board (USF IRB no. 8236) [41].
The Appendix (available as Supplemental Materials found at http://dx.doi.org/%2010.1016/j.jval.2015.11.002) provides a didactic overview of paired comparisons, adjectival statements, and pair selection as well as a few econometric concepts that may aid the reader. Originally, the sample size was 3000 respondents with 50 responses per pair, which allowed for 480 lifespan pairs (3000 8/50) and 1320 health pairs (3000 22/50). For the lifespan pairs, eight symptom domains were selected by the investigators as potentially worth a loss in lifespan. Losses in lifespan were paired with increases in these domains from level 1 (none) or 2 (mild) to level 5 (severe). Each symptom domain was experienced for three possible durations—7 days, 14 days, or 30 days—and was traded for five gains in lifespan (2 days, 7 days, 14 days, 30 days, and 90 days) with a delay that ranged from 1 year (e.g., 2 days less than 1 year) to 5 years (e.g., 2 days less than 5 years). In total, 480 lifespan pairs were created (8 2 3 5 2). The selection of the 1320 health pairs was more complicated because of the correlation structure between items. Specifically, we selected three types of health pairs: 382 interitem, 80 intersymptom, and 858 interdomain. The interitem pairs present tradeoffs between distinct aspects of the symptom experience (e.g., severity vs. interference for insomnia), the intersymptom pairs present tradeoffs of symptoms within a domain (e.g., feeling that nothing could cheer you up vs. sad and unhappy feelings for depression), and the interdomain pairs present domain tradeoffs (e.g., insomnia vs. dizziness).
Survey Instrument
Econometrics
After consenting, respondents completed a screener including demographic and socioeconomic status questions [42] and then proceeded to the survey, which began with a single item eliciting global, self-rated health and the MP-30D followed by paired comparisons and follow-up components. At the start of the
To simplify the analysis, only the 1480 pairs for which symptoms were experienced for 30 days were included in the analytical sample (see Appendix in Supplemental Material). For each of the pairs, the sample probability of choosing A over B, pk, was approximately normally distributed by the central limit theorem and served as the dependent variable of a generalized linear model. Each alternative (A and B) was represented by a linear regression, dh, that included 263 coefficients, 1 for each indicator variable of change in level (e.g., 1 to 2). The coefficients of the generalized linear model were estimated by minimizing the weighted sum of squared error,
Participants
1X 480
ðPðAk 4 Bk Þpk Þ2 =σ 2k
k¼1
where P(Ak 4 Bk) ¼ dB/(dA þ dB) and σ2k ¼ pk (1 – pk)/nk [43,44]. Confidence intervals were estimated by percentile bootstrap with pair-quota stratification and 1000 resampling iterations.
Results
Fig. 2
Of the 4211 respondents recruited for this study, 492 (12%) did not complete the survey and 322 (7.6%) were terminated (e.g., survey timed out at 2 hours). There were no associations between demographic characteristics (i.e., age, race, and ethnicity) or
162
VALUE IN HEALTH 19 (2016) 158–166
income and incomplete surveys. As indicated in Table 2, the sample characteristics were generally concordant with the 2010 US Census estimates, although the sample had slightly more Hispanic respondents and more homogeneous distributions with respect to educational attainment and household income. Among the 3397 respondents, 15% (n ¼ 514) reported that they had not experienced any of the 30 symptoms over the past 7 days. For the remaining (85%; n ¼ 2883), the median number of symptoms experienced was 3 out of 30 possible symptoms with an interquartile range of one to six symptoms. The four most prevalent symptom domains were aching joints, aching muscles, or back pain (45%); fatigue, tiredness, or lack of energy (40%); headache (34%); and insomnia (31%). For the lifespan pairs, 1753 (52%) respondents always preferred a reduced lifespan (rather than experiencing a domain of menopausal symptoms at its worst for 30 days) and 36 (1%) always preferred experiencing symptoms rather than have a reduced lifespan. This lack of variability suggests that future studies may wish to incorporate a higher range of reductions in lifespans (490 days) or milder symptoms with shorter durations. For health pairs, five (o1%) respondents always preferred the left and three (o1%) always preferred the right. Across all pairs, the sample sizes for each pair ranged from 49 to 68 respondents (available as Appendix in Supplemental Materials).
Figure 2 shows the results of all 480 lifespan pairs. In these pairs, loss in lifespan represents the “price” women would have to pay for symptom relief. The downward sloping curves (i.e., demand curves) imply that as the price of symptom relief drops the proportion of women who prefer symptom relief increases. More than 60% of the women prefer to reduce their lifespan by 90 days to avoid the eight domains of menopausal symptoms at their worst for 30 days (Fig. 2), which suggests that women place a high value on relief from symptoms that commonly occur during menopause. Based on this evidence alone, relief from insomnia and constipation were the least valuable of the eight domains and relief from headache and vomiting were the most valuable. To better understand QALY calculation using paired comparison results, consider that 37 (63%) out of 59 respondents preferred a loss of 90 days of lifespan over experiencing insomnia at its worst for 30 days (Fig. 1). With this single result and the cumulative distribution function, P(Ak 4 Bk) ¼ dB/(dA þ dB), we learned that the value of 30 days of insomnia equals a loss of 153 days of reduced lifespan (i.e., 0.42 QALY): 30 days of insomnia/(30 days of insomnia þ 90 days of reduced lifespan) ¼ 63% 30 days of insomnia ¼ 63% (30 days of insomnia þ 90 days of reduced lifespan)
Table 2 – Sample demographics by completion status. Characteristic
Age (y) 40–54 55–69 Race White Black or African American American Indian or Alaska Native Asian Native Hawaiian or other Pacific Islander Some other race Two or more races Hispanic ethnicity Hispanic or Latino Not Hispanic or Latino Educational attainment* Less than high school High school graduate Some college, no degree Associate’s degree Bachelor’s degree Graduate or professional degree Refused/don’t know Household income ($) r14,999 15,000–24,999 25,000–34,999 35,000–49,999 50,000–74,999 75,000–99,999 100,000–149,999 Z150,000 Refused/don’t know
n (%)
P value
US 2010 Census (%)
Dropout (N ¼ 492)
Terminated (N ¼ 322)
Completed (N ¼ 3397)
264 (53.66) 228 (46.34)
173 (53.73) 149 (46.27)
1846 (54.34) 1551 (45.66)
0.944
56.70 43.30
399 (81.10) 77 (15.65) 2 (0.41) 1 (0.20) 4 (0.81)
264 (81.99) 55 (17.08) – – –
2837 (83.51) 462 (13.60) 15 (0.44) 16 (0.47) 22 (0.65)
0.356
76.70 12.10 0.80 4.80 0.10
– 9 (1.83)
– 3 (0.93)
– 45 (1.32)
75 (15.24) 417 (84.76)
58 (18.01) 264 (81.99)
521 (15.34) 2876 (84.66)
0.441
11.20 88.80
22 (4.47) 225 (45.73) 96 (19.51) 44 (8.94) 95 (19.31) 9 (1.83) 1 (0.20)
4 (1.24) 88 (27.33) 68 (21.12) 54 (16.77) 100 (31.06) 8 (2.48) –
127 (3.74) 45.45 (1544) 617 (18.16) 323 (9.51) 738 (21.73) 46 (1.35) 2 (0.06)
o0.001
11.40 28.80 22.60 9.40 17.20 10.07
22 (4.47) 43 (8.74) 68 (13.82) 113 (22.97) 73 (14.84) 55 (11.18) 40 (8.13) 16 (3.25) 62 (12.60)
14 (4.35) 31 (9.63) 47 (14.60) 79 (24.53) 54 (16.77) 29 (9.01) 31 (9.63) 11 (3.42) 26 (8.07)
143 (4.21) 298 (8.77) 475 (13.98) 720 (21.20) 579 (17.04) 386 (11.36) 333 (9.80) 150 (4.42) 313 (9.21)
0.585
13.40 11.50 10.80 14.20 18.30 11.80 11.80 8.10
3.80 1.60
* Educational attainment and household income comes from the 2010 American Community Survey 1-Year Estimates [48,49].
VALUE IN HEALTH 19 (2016) 158–166
37% 30 days of insomnia ¼ 63% 90 days of reduced lifespan 30 days of insomnia ¼ 63% 90 days of reduced lifespan/37% 30 days of insomnia ¼ 153 days of reduced lifespan Still, this is just 1 of 60 lifespan pairs for the insomnia domain. In another lifespan pair, we learned that 44 (81%) out of 54 respondents preferred a loss of 30 days of lifespan over experiencing 30 days with insomnia at its worst severity and interference, which implies that 1 month of insomnia equals 128 days of reduced lifespan (81% 30/(1 – 81) ¼ 128 days or 0.35 QALY). Using the responses from 1480 pairs in a generalized linear model estimation, Table 3 describes the value of menopausal symptom relief from the perspective of women aged 40 to 69 years. For example, the value of constipation relief on a QALY scale increases on the basis of its severity: mild at its worst (0.296), moderate at its worst (0.296 þ 0.293 ¼ 0.589), severe at its worst (0.589 þ 2.942 ¼ 3.531), and very severe at its worst (3.531 þ 0.114 ¼ 3.645). Respondents were indifferent between 30 days of relief from very severe constipation and a 109-day increase in lifespan (3.645 30). Based on these results, relief from each symptom significantly increases the quality of women’s lives (P o 0.05), except a few that were insignificant (15 of 263 [6%]). Although symptom relief is clearly preferred over experiencing symptoms, the value women place on symptom relief ranged widely across the 30 domains. Figure 3 shows the ranking of the 15 “least desirable” symptom domains and depicts variation in the value of relieving each of their symptom attributes (frequency, severity, and interference).
Conclusions This valuation study focused on the MP-30D, a new adaptive checklist for the measurement of menopausal symptoms, integrating items from the PRO-CTCAE. In an effort to enhance this new measure for comparative effectiveness research, the 263 changes in symptoms captured by the MP-30D were valued on a QALY scale, so researchers can better understand the relative value women place on symptom relief. In the process of accomplishing these objectives, this is the first study to 1) examine preferences on symptoms described using the PRO-CTCAE; 2) assess the values of women nationally; and 3) distinguish the values placed on reducing symptom frequency, severity, and interference. Results show that from the perspective of women aged 40 to 69 years, the relief in the 30 symptom domains is highly valuable and that a reduction in symptom frequency— particularly for nausea, vomiting, headaches, anger, and abdominal pain—is preferred over improvements in severity and interference. On the basis of our findings, most of the women would prefer to reduce their lifespan than to experience menopausal symptoms, even at low levels. As the first preference study of the PRO-CTCAE (including 63 of the 126 items), these results provide insights into its symptom attributes (frequency, severity, and interference). First, the relief from low levels of symptoms is valuable (e.g., grade 1), particularly with respect to dizziness, headaches, problems with memory and concentration, and urinary pain and incontinence. This finding confirms the importance of aggressive symptom relief efforts, even when symptoms may not seem, at face value, to be prominent or highly distressing. Second, symptoms that worsen into grade 3 events are highly detrimental, such as moderate severity shifting to severe or very severe, and interference shifting from somewhat to quite a bit or very much. This observation may point toward clinically meaningful cutoff points
163
for the PRO-CTCAE items and may hint at potential mappings to serious AEs. Third, PRO-CTCAE attributes have different values, which suggest nonoverlapping information about their experiences. This study examines the perspectives of women, yet we recognize that some researchers strongly recommend that valuation be conducted from the societal perspective [45]. Such a study would require asking men to value symptoms relating to menopause. Many menopausal symptoms may be experienced by men (e.g., nausea), but not all (e.g., vaginal dryness). This inexperience alone might dilute the weights, and the focus of this study was on the perspective of women, specifically those with experience of one (or more) of these symptoms. Instead of imposing the values of men on women, some men may prefer to abdicate this authority to women. Others may feel obligated to express their view because of the implications for communal resources and religious considerations (e.g., reproductive rights). Out of a sense of fairness in measurement and valuation, conventional practice is to ignore evidence on gender-specific domains in QALY calculations. Some researchers believe that valuing gender-specific domains is unnecessary, impractical, and unfair. Regardless of one’s views on this practice, this article will help researchers and policymakers to better understand the value women place on symptom relief. Future analyses may explore heterogeneity in preferences among women, including differences between women of different ages, experiences, and dependents. Women may have experienced and adapted to symptoms, which may alter the value they place on symptom relief. The value a woman places on symptom relief (e.g., anger) may also incorporate the benefits for those around her as well as her capacity to fulfill commitments and roles, including responsibilities at work and at home, particularly when frequent, severe, or debilitating symptoms (e. g., headaches) can make it difficult to engage in family and social relationships or may limit vocational functioning. Subsequent studies will examine such heterogeneity as well as the values of younger and older women (and men, where appropriate). This study is the first to assess the value of relieving symptoms that are commonly experienced during menopause, but it did not value all aspects of menopause, which may include physiologic changes, such as diminished fertility. Future studies may incorporate additional symptoms as well as outcomes of shorter or longer durations. A fundamental limitation of this study is the use of the QALY scale, which assumes that the value of a symptom is strictly proportional to its duration and independent of one’s lifespan. A more subtle limitation of this article (and most health valuation studies) is an implicit assumption that the value of relieving individual symptoms is additive and that the presence of concurrent multiple symptoms has no interaction effect on the value of relieving other symptoms. We acknowledge that the experience of multiple concurrent symptoms may magnify or reduce the value of relieving one or more symptoms. Alternative model specifications that accommodate nonadditive relationships and the value of relieving concurrent symptoms can be examined in future work. To understand the impact of chemotherapy, injury, or other life-changing events that can induce premature menopause requires clinical investigations that include measures of symptomatic toxicities such as the PRO-CTCAE or symptom instruments, such as the MP-30D. Nevertheless, this study investigates the preferences of midlife women to avoid symptoms and enhances our understanding of the value of symptom relief. These strong preferences toward symptom relief underscore the importance of gender-specific valuation and motivate efforts to evaluate new treatment approaches to manage symptoms effectively in women at midlife.
164
VALUE IN HEALTH 19 (2016) 158–166
Table 3 – The values women place on menopausal symptoms on a quality-adjusted life-year (QALY) scale. Symptoms arranged by MP-30D domains*
Change in level 1 to 2
Pounding or racing heartbeat† Pain in the abdomen
Constipation‡ Increased passing of gas Nausea Vomiting Bloating of the abdomen Decreased appetite Aching joints
Aching muscles
Back pain
Insomnia Anxiety
Feel that nothing could cheer you up
Sad or unhappy feelings
Irritability
Problems with anger
Decreased sexual interest Frequent urination† Urges to urinate all of a sudden Loss of urinary control Urinary pain Numbness or tingling in hands or feet Dizziness Headaches
Problems with memory Problems with concentration
F S F S I S P F S F S F S S I F S I F S I F S I S I F S I F S I F S I F S I F S I S F I F I F I S S I S I F S I S I S I
0.541 0.291 0.324 0.296
0.416 0.825 0.234 0.017 0.030 0.030 0.170 0.018 0.099 0.033 0.159 0.209 0.139 0.229 0.439 0.273 0.340 0.107 0.406 0.143 0.238 0.413 0.689 0.205 0.160 0.192 0.701 1.144 0.218 0.225 0.485 0.429 0.087 0.265 0.654 0.547 0.270 0.341
2 to 3
3 to 4
0.351 0.247 0.233 0.372 0.373 0.293
0.747 2.365 2.829 1.719 2.424 2.942
0.380 0.156 1.237 0.231 0.173 0.192 0.014 0.023 0.056 0.052 0.012 0.030 0.054 0.033 0.070 0.090 0.048 0.151 0.131 0.265 0.312 0.273 0.392 0.169 0.177 0.254 0.297 0.04 0.183 0.105 0.203 0.797 0.674 0.677 0.017 0.054 0.040 0.125 0.033 0.427 0.058 0.444 0.187 0.134 0.541 0.246 0.120 0.402 0.164 0.626 0.506 0.328 0.241
4 to 5
3.388 1.515 5.399 1.620 1.750 0.114 0.423 2.758 2.949 1.216 0.896 9.932 8.662 2.012 1.715 0.579 0.689 0.706 0.431 0.049 0.057 0.177 0.049 0.246 0.301 0.055 0.263 0.437 0.200 0.195 0.350 0.149 0.148 0.400 0.122 0.715 0.990 0.235 0.564 1.074 0.366 0.874 0.600 1.231 0.435 1.924 2.111 0.717 1.220 1.593 1.116 0.815 1.139 0.509 0.477 0.984 0.482 1.018 0.947 0.208 0.496 1.209 0.592 1.145 1.082 0.542 0.310 1.063 0.608 5.973 4.662 1.796 1.032 3.677 1.299 0.279 0.001 0.116 0.311 0.303 0.120 0.177 0.376 0.304 0.166 1.259 0.948 0.666 0.364 4.080 0.001 0.695 0.582 1.257 0.372 3.806 0.566 3.617 0.410 2.581 5.765 1.824 1.114 1.581 1.838 3.738 2.305 3.526 1.093 1.096 1.036 1.629 0.728 continued on next page
165
VALUE IN HEALTH 19 (2016) 158–166
Table 3 – continued. Symptoms arranged by MP-30D domains*
Irregular menstrual periods Vaginal dryness Pain during vaginal sex Breast area enlargement and pain Unexpected or excessive sweating Dry skin Itchy skin Hot flashes Fatigue, tiredness, or lack of energy
Change in level
P S S S F S S S F S S I
1 to 2
2 to 3
3 to 4
4 to 5
0.145 0.857 0.335
0.058 0.180 0.059 0.104 0.082 0.023 0.033 0.082 0.053 0.262 0.133
0.458 1.582 0.734 0.247 0.278 0.045 0.447 0.348 0.267 0.684 1.584
0.175 0.012 0.277 0.166 0.803 0.288 0.001 0.001 0.765 0.294 0.316 0.268
0.237 0.018 0.167 0.114 0.203 0.310
* Presence (P) is a two-level response scale (No/Yes). The F, S, and I items for back pain, irritability, and problems with anger are not in the PROCTCAE v1.0. Frequency (F), severity (S), and interference with usual activities (I) on a five-level response scale where higher level indicated greater symptoms. The labels depend on the attribute: frequency increases from never (1) to rarely (2), to occasionally (3), to frequently (4), and to almost constantly (5); severity increases from none at its/their worst (1) to mild at its/their worst (2), to moderate at its/their worst (3), to severe at its/their worst (4), and to very severe at its/their worst (4); and interference increases from not at all (1) to a little bit (2), to somewhat (3), to quite a bit (4), and to very much (5). † The level change from 1 to 2 in frequency (i.e., none to rarely) is implied by all other level changes for the domain; therefore, it is not shown. ‡ For example, the value of constipation relief on a QALY scale increases on the basis of its severity: mild at its worst (0.296), moderate at its worst (0.296 þ 0.293 ¼ 0.589), severe at its worst (0.589 þ 2.942 ¼ 3.531), and very severe at its worst (3.531 þ 0.114 ¼ 3.645). Therefore, respondents were indifferent between 30 d of relief from very severe constipation or 109-d increase in lifespan (3.645 30).
Fig. 3
Acknowledgments We thank Michelle Owens, Janel Phetteplace, and Carol Templeton at Moffitt Cancer Center for their contributions to the research and creation of this article. Source of financial support: Funding support for this research was provided by Dr. Craig’s support account at Moffitt Cancer Center.
Supplemental Materials Supplemental material accompanying this article can be found in the online version as a hyperlink at http://dx.doi.org/10.1016/j.
jval.2015.11.002 or, if a hard copy of article, at www.valueinhealth journal.com/issues (select volume, issue, and article).
R EF E R EN C ES
[1] Avis NE, Colvin A, Bromberger JT, et al. Change in health-related quality of life over the menopausal transition in a multiethnic cohort of middle-aged women: Study of Women’s Health Across the Nation. Menopause 2009;16:860–9. [2] Bromberger JT, Kravitz HM, Chang YF, et al. Major depression during and after the menopausal transition: Study of Women’s Health Across the Nation (SWAN) (vol 41, pg 1879, 2011). Psychol Med 2011;41: 2238–8. [3] Bermejo JB, Martinez JAV, Saiz AR, et al. Validation of a menopause quality of life scale: the MENCAV scale. Maturitas 2008;59:28–37.
166
VALUE IN HEALTH 19 (2016) 158–166
[4] Cella D, Land SR, Chang CH, et al. Symptom measurement in the Breast Cancer Prevention Trial (BCPT) (P-1): psychometric properties of a new measure of symptoms for midlife women. Breast Cancer Res Treat 2008;109:515–26. [5] Chattha R, Kulkarni R, Nagarathna R, et al. Factor analysis of Greene’s Climacteric Scale for Indian women. Maturitas 2008;59:22–7. [6] Chedraui P, Aguirre W, Hidalgo L, et al. Assessing menopausal symptoms among healthy middle aged women with the Menopause Rating Scale. Maturitas 2007;57:271–8. [7] Girod I, de la Loge C, Keininger D, et al. Development of a revised version of the Women’s Health Questionnaire. Climacteric 2006;9:4–12. [8] Hilditch JR, Lewis J, Peter A, et al. A menopause-specific quality of life questionnaire: development and psychometric properties. Maturitas 1996;24:161–75. [9] Hunter M. The Womens Health Questionnaire—a measure of mid-aged womens perceptions of their emotional and physical health. Psychol Health 1992;7:45–54. [10] Jacobs PA, Hyland ME, Ley A. Self-rated menopausal status and quality of life in women aged 40-63 years. Br J Health Psychol 2000;5:395–411. [11] Utian WH, Janata JW, Kingsberg SA, et al. The Utian Quality of Life (UQOL) Scale: development and validation of an instrument to quantify quality of life through and beyond menopause. Menopause 2002;9:402–10. [12] Zollner YF, Acquadro C, Schaefer M. Literature review of instruments to assess health-related quality of life during and after menopause. Qual Life Res 2005;14:309–27. [13] Basch EM, Reeve BB, Mitchell SA, et al. Electronic toxicity monitoring and patient-reported outcomes. Cancer J 2011;17:231–4. [14] Basch EM, Reeve BR, Mitchell SA, et al. Development of the National Cancer Institute’s Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). J Natl Cancer Inst 2014;106. [15] Hays RD, Spritzer KL, Amtmann D, et al. Upper-extremity and mobility subdomains from the Patient-Reported Outcomes Measurement Information System (PROMIS) adult physical functioning item bank. Arch Phys Med Rehabil 2013;94:2291–6. [16] National Cancer Institute. Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). National Cancer Institute, Rockville, Maryland, USA, 2010. [17] National Cancer Institute. Appendix A–Patient-Reported OutcomesCommon Terminology Criteria for Adverse Events (PRO-CTCAE) Version v1.0 Items. National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Rockville, MD, USA, 2013. [18] Mendoza T, Atkinson T. PRO-CTCAE Face-To-Face Meeting #2: Development of PRO-CTCAE Items. National Cancer Institute, Rockville, MD, USA, 2010. [19] Brooks R, Rabin R, De Charro F. The Measurement and Valuation of Health Status Using EQ-5D: A European Perspective: Evidence from the EuroQol BIO MED Research Programme. Springer, Dordrecht, The Netherlands, 2003. [20] Furlong WJ, Feeny DH, Torrance GW, et al. The Health Utilities Index (HUI (R)) system for assessing health-related quality of life in clinical studies. Ann Med 2001;33:375–84. [21] Craig BM, Reeve BB, Cella D, et al. Demographic differences in health preferences in the United States. Med Care 2014;52:307–13. [22] Craig BM, Pickard AS, Stolk E, Brazier JE. US valuation of the SF-6D. Med Decis Mak 2013;33:793–803. [23] Ratcliffe J, Couzner L, Flynn T, et al. Valuing Child Health Utility 9D health states with a young adolescent sample: a feasibility study to compare best-worst scaling discrete-choice experiment, standard gamble and time trade-off methods. App Health Econ Health Pol 2011;9:15–27. [24] Erickson P. Evaluation of a population-based measure of quality of life: the Health and Activity Limitation Index (HALex). Qual Life Res 1998;7:101–14.
[25] Fryback DG, Dunham NC, Palta M, et al. US norms for six generic health-related quality-of-life indexes from the National Health Measurement Study. Med Care 2007;45:1162–70. [26] Fryback DG, Palta M, Cherepanov D, et al. Comparison of 5 healthrelated quality-of-life indexes using item response theory analysis. Med Decis Mak 2010;30:5–15. [27] Andresen EM, Rothenberg BM, Kaplan RM. Performance of a selfadministered mailed version of the Quality of Well-Being (QWB-SA) questionnaire among older adults. Med Care 1998;36:1349–60. [28] Brazier J, Roberts J, Platts M, et al. Estimating a preference-based index for a menopause specific health quality of life questionnaire. Health Qual Life Outcomes 2005;3:13. [29] Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271–92. [30] Schneider HPG, MacLennan AH, Feeny D. Assessment of health-related quality of life in menopause and aging. Climacteric 2008;11:93–107. [31] Brazier JE, Rowen D, Mavranezouli I, et al. Developing and testing methods for deriving preference-based measures of health from condition-specific measures (and other patient-based measures of outcome). Health Technol Assess (Winchester, England) 2012;16:1–114. [32] Kuspinar A, Finch L, Pickard S, et al. Using existing data to identify candidate items for a health state classification system in multiple sclerosis. Qual Life Res 2013;23:1445–57. [33] Krabbe PFM, Stouthard MEA, Essink-Bot ML, et al. The effect of adding a cognitive dimension to the EuroQol multiattribute health-status classification system. J Clin Epidemiol 1999;52:293–301. [34] Wang X, Bradlow ET, Wainer H. User’s Guide for SCORLIGHT (Version 3.0): A Computer Program for Scoring Tests Built of Testlets Including a Module for Covariate Analysis. Educational Testing Service, Princeton, New Jersey, USA, 2005. [35] Thissen D, Steinberg L, Mooney JA. Trace lines for testlets—a use of multiple-categorical-response models. J Educ Meas 1989;26:247–60. [36] Craig B, Owens MA. Methods Report on the Women’s Health Valuation (WHV): Year 1. Moffitt Cancer Center, Tampa, Florida, USA, 2013. [37] Craig BM, Busschbach JJ. The episodic random utility model unifies time trade-off and discrete choice approaches in health state valuation. Popul Health Metr 2009;7:3. [38] Bansback N, Brazier J, Tsuchiya A, et al. Using a discrete choice experiment to estimate health state utility values. J Health Econ 2012;31:306–18. [39] Hadorn DC, Hays RD, Uebersax J, et al. Improving task comprehension in the measurement of health state preferences—a trial of informational cartoon figures and a paired-comparison task. J Clin Epidemiol 1992;45:233–43. [40] Craig BM, Schell MJ, Brown PM, et al. HRQoL Values for Cancer Survivors: Enhancing PROMIS Measures for CER. H. Lee Moffitt Cancer Center, National Institutes of Health, Rockville, MD, USA, 2011. [41] Craig BM, Reeve BB, Brown PM, et al. US valuation of health outcomes measured using the PROMIS-29. Value Health 2014;17:846–53. [42] U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau. United States Census 2010. In: U.S. Department of Commerce Economics and Statistics Administration, ed., National Processing Center, Jeffersonville, Indiana, USA, 2010. [43] McFadden D. Conditional logit analysis of qualitative choice behavior. In: Zarembka P, ed. Frontiers in Econometrics. New York: Academic Press, 1974. [44] Urban FM. Urban’s solution (minimum normit X2). In: Bock RD, Jones LV, eds. The Measurement and Prediction of Judgment and Choice. San Francisco: Holden-Day, 1968. [45] Gold M, Siegel J, Russell L, et al. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press, 1996. [46] Hunter M. The Women’s Health Questionnaire (WHQ): the development, standardization and application of a measure of mid-aged women’s emotional and physical health. Qual Life Res 2000;9:733–8. [47] Greene JG. Constructing a standard climacteric scale. Maturitas 1998;29:25–31.