ABSTRACT
Objective
A global multidisciplinary consensus initiative identified eight patient‐reported outcome domains important in obesity care: self‐esteem; physical, mental, and social health; eating; stigma; body image; and excess skin. Several of these domains are not represented in generic preference‐based measures typically used in weight loss research. We developed the BODY‐Q‐7 Dimensions (BODY‐Q‐7D) to provide a valuation‐ready framework for a preference‐based instrument.
Methods
Concepts from the BODY‐Q were used to inform the development of items that covered the eight consensus initiative domains. Psychometric performance of the BODY‐Q‐7D was examined using an online international sample (i.e., Prolific).
Results
Content validity was established with 123 participants. The field test included 770 participants, of whom 269 completed a test–retest, and 599 completed a 3‐month follow‐up. Rasch Measurement Theory analysis reduced the BODY‐Q‐7D to 7 items. Items fit the Rasch model with ordered thresholds and structural independence. Reliability was > 0.80 across three reliability coefficients. Hypothesis‐based tests of construct validity and responsiveness were supported with 22/28 (79%) and 12/16 (75%) tests accepted.
Conclusions
The BODY‐Q‐7D demonstrated strong psychometric performance in a weight loss sample. The next phase of development is to elicit and model preference weights to assign utility values to each defined health state.
Keywords: BODY‐Q, health‐related quality of life, preference‐based measure, weight loss
Study Importance
- What is already known?
-
○A global multidisciplinary consensus initiative identified eight overarching outcome domains considered important by people living with obesity. Generic preference‐based measures lack content validity for obesity research.
-
○
- What does this study add?
-
○BODY‐Q‐7D was developed with extensive input from people living with obesity and experts.
-
○BODY‐Q‐7D measures outcome domains important in the multidisciplinary management of obesity.
-
○
- How might these results change the direction of research or the focus of clinical practice?
-
○BODY‐Q‐7D health states can be valued for use in economic evaluations of treatments for weight loss.
-
○
1. Introduction
Obesity is an urgent public health concern associated with high healthcare costs. In 2020, overweight and obesity affected 38% of the global population, a proportion projected to exceed 50% by 2035 without improved prevention and treatment [1]. The global economic burden for this condition is expected to increase from 2.4% to 2.9%, corresponding to an annual cost of US $4.32 trillion by 2035 [1]. These costs underscore the need for healthcare systems to prioritize cost‐effective weight loss interventions.
Treatments for obesity can have a substantial impact on health‐related quality of life (HRQL) [2]. Effective interventions include metabolic bariatric surgery (MBS), which can result in sustained weight loss and improvement or resolution of many obesity‐related comorbidities [3]. Behavioral and pharmacological interventions, including recent glucagon‐like peptide‐1 (GLP‐1) receptor agonists, also produce clinically meaningful weight loss [4, 5, 6]. In health economic evaluations of weight loss treatments, HRQL is commonly measured from the patient perspective using preference‐based measures (PBMs) to estimate quality‐adjusted life years (QALYs). Most health economic evaluations of weight loss treatments have relied on generic PBMs to estimate QALYs, including the EQ‐5D, EQ‐5D‐5L, and SF‐6D [6, 7]. However, generic instruments may lack sensitivity to detect change in obesity‐specific aspects of HRQL following multidisciplinary management of obesity as a chronic disease.
Another approach in health economic evaluations has been to use the scores from an obesity‐specific patient‐reported outcome measure (PROM) to predict utility scores on a generic PBM [8, 9, 10, 11, 12]. PROMs that have been used to map content to generic PBMs include the Moorehead‐Ardelt II (MA‐II) [8], the Impact of Weight on Quality of Life‐Lite (IWQOL‐Lite) [9, 10], and the Obesity Problems Scale (OP‐scale) [11]. The mapping approach has recognized limitations, particularly in accurately reflecting obesity‐specific concerns that are not well represented in generic PBMs. Overall, since both approaches rely on generic PBMs to generate QALYs, the scores can underestimate QALY gains, potentially biasing cost‐effectiveness estimates and influencing reimbursement and coverage decisions for weight loss interventions.
To address the need for standardized outcome measurement in obesity research, the Standardize Quality of life measurement in Obesity Treatment (S.Q.O.T.) initiative was established [12, 13, 14]. This initiative aimed to identify a set of overarching outcome domains most relevant to obesity research and clinical care. A literature review identified 25 candidate patient‐reported outcomes (PROs) [15]. These outcomes were prioritized during international multidisciplinary consensus meetings that involved people living with obesity and healthcare professionals [12, 13, 14]. Consensus was reached that the following eight overarching outcome domains are important to measure: self‐esteem, physical health/function, mental/psychological health, social health, eating, stigma, body image, and excess skin.
Since generic PBMs do not capture all eight outcomes, a weight‐loss‐specific PBM is needed to support accurate and meaningful health economic evaluations. BODY‐Q is a validated, modular obesity‐specific PROM that can be used to measure outcomes related to weight loss treatments (e.g., diet, exercise, bariatric surgery/medicine) and body contouring to remove excess skin after weight loss or for cosmetic purposes [16, 17, 18]. BODY‐Q is composed of multiple independently functioning scales and is therefore not amenable to valuation since it is not possible to combine scores for different scales into a single score.
The aim of this study was to apply Rasch Measurement Theory (RMT) to develop a short‐form BODY‐Q‐7 Dimensions (BODY‐Q‐7D) that captures the PROs identified by the S.Q.O.T. initiative and forms a robust foundation for a future PBM. The Rasch approach was used to ensure that the dimensions provide invariant, interpretable, and interval‐level measurement, key prerequisites for defining health states suitable for valuation. The specific objectives were: (1) to develop the content of the BODY‐Q‐7D and establish its content validity and (2) to evaluate its psychometric performance in preparation for subsequent health state valuation and utility modeling.
2. Methods
2.1. Ethics
Ethical approval for this study was obtained from the Hamilton Integrated Research Ethics Board in Hamilton, Ontario, Canada (Project ID #18651). The study used the Prolific online platform (prolific.com). All surveys were administered using REDCap, a secure platform for electronic data collection, hosted through the Faculty of Health Sciences at McMaster University. Electronic informed consent was obtained from participants prior to completing the study surveys. Compensation was a prorated rate of 12 pounds sterling per hour.
2.2. Data Collection
2.2.1. Content Validity
The BODY‐Q‐7D was developed and validated in accordance with established PROM development guidelines [19, 20, 21, 22, 23]. Concepts from the BODY‐Q [16, 17, 18] were used to create items to cover the eight PROs from the S.Q.O.T. [12, 13, 14], forming a draft instrument. The draft was reviewed by S.Q.O.T. initiative members who provided feedback that was used to make revisions. The revised instrument was included in a content validity survey. On May 5, 2025, the Prolific platform was used to screen for residents of Australia, Canada, Ireland, New Zealand, UK, and USA whose BMI was ≥ 20. Those who were trying to lose weight, wanting to lose weight, or had lost weight through MBS or an obesity medication were invited to complete the content validity survey. Table S1 shows the content validity questions.
2.2.2. Field Test
More Prolific participants were screened on July 10, 2025, to recruit a larger sample for the field test. For data quality, participants were asked to repeat the screening survey on July 22, 2025. Those who provided the same answers for type of MBS and/or type, mode of administration, and/or treatment status (current vs. past use) for GLP‐1 receptor agonists were invited to complete the field test survey. This survey included demographic and clinical questions, the BODY‐Q‐7D and other BODY‐Q scales (i.e., Body, Body Image, Physical Function, Psychological, Social, Eating Behavior, Sexual) [16, 17, 18], and the EQ‐5D‐5L [24]. The survey remained open until August 9, 2025.
2.2.3. Test–Retest (TRT)
TRT data were collected between days 7 and 14. Participants who completed the survey within the time frame and reported no change in their weight‐related HRQL since the initial assessment were included in the TRT analysis. We aimed to recruit ≥ 100 participants to obtain a “very good” sample size TRT rating according to COSMIN guidelines [22].
2.2.4. Responsiveness
At least 3 months after the initial survey, participants were invited to complete a follow‐up survey. For change in the BODY‐Q‐7D scale, we asked directionality (i.e., same, worse, or better) and magnitude (e.g., a little, somewhat, quite a bit, a lot). Participants were reminded of their baseline answers for constructs using the piping function in REDCap. Reminders are recommended by Devji et al. [25], who reported that people have difficulty recalling a previous health state after 4 weeks.
2.3. Analysis
Content validity results were analyzed descriptively. Table S2 shows the psychometric tests performed. For the field test data, RMT analysis was used to examine the response threshold order, item fit, local dependency, targeting, differential item functioning (DIF), and reliability [26, 27, 28, 29]. The probabilistic Rasch model compares the relationship between the ability of the people in the sample and the difficulty of the items in a scale. When data for a scale fit the Rasch model, the scale provides invariant, interpretable, and interval‐level measurement. RUMM2030 software (RUMM Laboratory) was used with the unrestricted partial credit model for polytomous data. Rasch logits for the final scale were used to convert raw scores for the BODY‐Q‐7D into scores that ranged from 0 (worst) to 100 (best). Other BODY‐Q scales and the EQ‐5D‐5L were scored according to developers' instructions.
For TRT, single and average intraclass correlation coefficients (ICC) were calculated with and without outliers identified using box plots. Analysis used a two‐way mixed effects model for consistency. ICC values ≥ 0.7 were indicative of acceptable reliability [22]. The smallest detectable change (SDC) was computed at the individual [1.96*√(2)*SEM] and group [SDCind/√n] level [30]. The standard error of the mean (SEM) for SDC calculations was determined as follows: (SDT1 + SDT2)/2*√(1 − ICC) [22].
Table 1 shows 28 and 16 predetermined hypotheses for the initial and responsiveness surveys, respectively. Variables used to examine group differences in the initial survey included: height, weight, and goal weight; location in the weight loss journey (not started, close to the start, about halfway through, near completion, finished); satisfaction with weight (extremely dissatisfied to extremely satisfied); physical and mental health (excellent, very good, good, fair, poor); social life, eating habits, and how the body looks (extremely unhappy to extremely happy); activity level (not at all to very active); how often weight interfered with ability to do usual daily activities (never, rarely, sometimes, often, always); MBS (want vs. had); GLP‐1 receptor agonist (want vs. currently having); body contouring (want vs. do not want); a computed score for amount (none, a little, a moderate amount, a lot) of excess skin across seven parts of the body; and severity of excess skin problems in the past week (none, little bad, somewhat bad, very bad). Additional variables in the follow‐up survey used to examine construct validity included: directionality of change (same, worse, better); magnitude of change (a little, somewhat, quite a bit, a lot); and satisfaction with change in weight (dissatisfied vs. satisfied).
TABLE 1.
Tests of construct validity.
| Hypotheses | Result |
|---|---|
| Cross‐sectional hypotheses | |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
| Correlations | |
|
Yes |
|
No |
|
Yes |
|
No |
|
No |
|
No |
|
Yes |
|
No |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
No |
|
Yes |
| Hypotheses ACCEPTED | 22/28 (79%) |
| Responsiveness hypotheses based on change scores | |
|
Yes |
|
Yes |
|
Yes |
|
No |
|
Yes |
|
Yes |
| Correlations | |
|
No |
|
Yes |
|
Yes |
|
Yes |
|
No |
|
Yes |
|
Yes |
|
Yes |
|
Yes |
|
No |
| Hypotheses ACCEPTED | 12/16 (75%) |
A composite Likert change score was computed by assigning 0 for same score, −1 for worse score, +1 for improved score for the following HRQL questions: happy with how body looks, social life, eating habits, activity level, physical health, mental health, and interference of weight with activities.
Hypothesized correlations between BODY‐Q‐7D scores, clinical variables, and scores for other BODY‐Q scales and the EQ‐5D‐5L were based on the relatedness of constructs using magnitudes that are commonly applied in COSMIN‐based hypothesis testing [21]. Specifically, COSMIN considers correlations ≥ 0.5 to indicate similar constructs, 0.3–0.5 to indicate related but dissimilar constructs, and < 0.30 to indicate weak or unrelated constructs. For the field test data, we hypothesized that the strength of correlations between the BODY‐Q‐7D and the BODY‐Q Psychological scale and the EQ‐5D‐5L would be ≥ 0.5 and with clinical variables (e.g., BMI), other BODY‐Q scales, and EQ‐5D‐5L subscales would be 0.30–0.50. For the change scores, we hypothesized the direction and the relative magnitude of change scores. Associations were expected to be weak (r < 0.30), moderate (0.30–0.49), or strong (≥ 0.50) depending on the conceptual similarity between constructs measured and how much each construct was anticipated to change [31]. Construct validity was considered adequate if ≥ 75% of the hypothesis tests were confirmed as per COSMIN criteria [20, 22].
Analysis took place in SPSS version 30. ANOVA and independent t‐tests were used to assess group differences. A one‐sided t‐test was used for testing differences between two groups when directionality was hypothesized. Paired t‐tests were used to examine change for the responsiveness analysis. Pearson correlations were used to assess convergent validity. Normality of data was assessed using kurtosis and skewness and visual plots and nonparametric statistics were applied if distributions were considered non‐ normal [32]. Statistical significance was set at p < 0.05. Construct validity was based on the consistency of results for the hypotheses rather than the individual p values. As such, adjustment for multiple comparisons was not considered appropriate.
3. Results
Figure 1 shows the flow of participants and reasons for exclusions at each step of the study.
FIGURE 1.
Recruitment for each component of the study. [Color figure can be viewed at wileyonlinelibrary.com]
3.1. Content Validity
The first draft of the BODY‐Q‐7D included 11 questions covering the eight S.O.Q.T. outcomes. The scale included four response options and the following instructions: “These questions ask about how OBESITY and WEIGHT LOSS have affected you. Please answer each question based on the PAST WEEK.” This version was sent by email to 16 S.Q.O.T. committee members (2 obesity doctors, 1 endocrinologist, 2 psychologists, 2 dietitians, 2 health economists, and 1 nurse practitioner) and organizers of the meetings (3 surgeons and 3 researchers). Feedback led to several changes. The instructions were revised to broaden the scope: “These questions ask about the impact of your WEIGHT on your HEALTH. Please answer each question based on the PAST WEEK.” A question deemed outside the scope of the eight outcomes was dropped (i.e., “Did you experience any unpleasant symptoms related to eating [e.g., heartburn, nausea, vomiting]?”). A question was added to provide an alternative for body image (i.e., “How do you feel about your body?”). Minor wording changes were made to better align questions with the response options.
A total of 123 participants completed the content validity survey. Table 2 shows the participant characteristics and Table 3 shows the type of obesity medication participants were currently taking or had in the past. The sample ranged in age from 20 to 75 years (mean = 41.7, SD = 12.7). When asked if they were currently trying to lose weight, 103 (83.7%) said yes, 14 (11.4%) said no but that they wanted to, and 6 (4.9%) said no. All participants indicated that the BODY‐Q‐7D instructions were easy to understand. Table 4 shows the content validity results. The response options and 11 questions were generally deemed easy to understand and reflective of participant experiences. For questions measuring similar concepts, some participants found “moderate exercise” more relevant than “walk or move around.” The question “Did you experience any unpleasant symptoms due to excess skin caused by weight loss (e.g., rash, infections, odor)?” was considered not relevant by 11 participants and comments such as the following were provided: “No excess skin, I've never been worried about this,” “Me personally I had no extreme weight loss causing excess skin,” “I've never had problems with my skin due to my weight loss in the past ever,” “I've never lost enough weight to have much in the way of excess skin.”
TABLE 2.
Sample characteristics.
| Characteristic | Subgroups | Content validity | Field test | Follow‐up | |||
|---|---|---|---|---|---|---|---|
| N = 123 | % | N = 770 | % | N = 599 | % | ||
| What country do you live in? | USA | 80 | 65.0 | 384 | 49.9 | 299 | 49.9 |
| Canada | 7 | 5.7 | 73 | 9.5 | 56 | 9.3 | |
| UK | 35 | 28.5 | 275 | 35.7 | 217 | 36.2 | |
| Australia | 0 | 0 | 23 | 3.0 | 18 | 3.0 | |
| Ireland | 0 | 0 | 11 | 1.4 | 5 | 0.8 | |
| New Zealand | 1 | 0.8 | 4 | 0.5 | 4 | 0.7 | |
| What is your age? | 18–34 | 41 | 33.3 | 206 | 26.8 | 149 | 24.9 |
| 35–44 | 41 | 33.3 | 230 | 29.9 | 169 | 28.2 | |
| 45–54 | 18 | 14.6 | 192 | 24.9 | 154 | 25.7 | |
| ≥ 55 | 23 | 18.7 | 142 | 18.4 | 127 | 21.2 | |
| What is your gender? | Man | 65 | 52.8 | 223 | 29.0 | 182 | 30.4 |
| Woman | 57 | 46.3 | 534 | 69.4 | 410 | 68.4 | |
| Nonbinary | 1 | 0.8 | 11 | 1.4 | 5 | 0.8 | |
| Other gender | 0 | 0 | 2 | 0.3 | 2 | 0.3 | |
| What racial group(s) do you belong to? | Black | 25 | 20.3 | 72 | 9.4 | 57 | 9.5 |
| East Asian | 3 | 2.4 | 2 | 0.3 | 2 | 0.3 | |
| Southeast Asian | 1 | 0.8 | 6 | 0.8 | 5 | 0.8 | |
| Latin American | 3 | 2.4 | 12 | 1.6 | 7 | 1.2 | |
| Middle Eastern | 1 | 0.8 | 2 | 0.3 | 2 | 0.3 | |
| South Asian | 2 | 1.6 | 16 | 2.1 | 12 | 2.0 | |
| White | 84 | 68.3 | 620 | 80.5 | 484 | 80.8 | |
| Multiple races | 2 | 1.6 | 33 | 4.3 | 24 | 4.0 | |
| Prefer not to answer/Missing/Other | 2 | 1.6 | 1 | 0.1 | 1 | 0.2 | |
| Indigenous | 0 | 0 | 4 | 0.5 | 3 | 0.5 | |
| Pacific Islander | 0 | 0 | 2 | 0.3 | 2 | 0.3 | |
| As of today, what is the highest level of education you have completed? | Some high school | 1 | 0.8 | 8 | 1.0 | 8 | 1.3 |
| Completed high school | 6 | 4.8 | 81 | 10.5 | 61 | 10.2 | |
| Some college or trade school or university | 18 | 14.6 | 168 | 21.8 | 130 | 21.7 | |
| Completed college or trade school or university | 66 | 53.7 | 350 | 45.5 | 274 | 45.7 | |
| Some master's or doctoral degree | 10 | 8.1 | 39 | 5.1 | 32 | 5.3 | |
| Completed master's or doctoral degree | 22 | 17.9 | 124 | 16.1 | 94 | 15.7 | |
| In the past 3 months, how difficult was it for you to cover your household expenses and pay your bills? | Not at all difficult | 34 | 27.6 | 219 | 28.4 | 177 | 29.5 |
| A little difficult | 46 | 37.4 | 245 | 31.8 | 192 | 32.1 | |
| Somewhat difficult | 34 | 27.6 | 167 | 21.7 | 124 | 20.7 | |
| Very difficult | 6 | 4.9 | 81 | 10.5 | 63 | 10.5 | |
| Extremely difficult | 3 | 2.4 | 56 | 7.3 | 41 | 6.8 | |
| Prefer to not answer | 0 | 0 | 2 | 0.3 | 2 | 0.3 | |
| What is your current marital status? | Never married | 35 | 28.5 | 245 | 31.8 | 184 | 30.7 |
| Separated | 2 | 1.6 | 21 | 2.7 | 17 | 2.8 | |
| Divorced | 5 | 4.1 | 78 | 10.1 | 65 | 10.9 | |
| Widowed | 2 | 1.6 | 11 | 1.4 | 10 | 1.7 | |
| Living common‐law | 12 | 9.8 | 62 | 8.1 | 48 | 8.0 | |
| Married | 66 | 53.7 | 346 | 44.9 | 270 | 45.1 | |
| Other | 1 | 0.8 | 7 | 0.9 | 5 | 0.8 | |
| BMI category | < 24.9 | 17 | 12.8 | 68 | 8.8 | 51 | 8.6 |
| 25–29.9 | 35 | 28.5 | 113 | 14.7 | 82 | 13.8 | |
| 30–34.9 | 31 | 25.2 | 160 | 20.8 | 131 | 22.0 | |
| 35–39.9 | 15 | 12.2 | 158 | 20.5 | 123 | 20.6 | |
| > 39.9 | 25 | 20.3 | 271 | 35.2 | 209 | 35.1 | |
| Have you ever had bariatric surgery? | No | 71 | 57.8 | 668 | 86.8 | 526 | 87.8 |
| Yes | 52 | 42.3 | 102 | 13.2 | 73 | 12.2 | |
| Ever taken a GLP‐1 medication to lose weight? | No | 40 | 32.6 | 328 | 42.6 | 255 | 42.6 |
| Yes, I am currently taking | 48 | 39.0 | 324 | 42.1 | 258 | 43.1 | |
| Yes, I have taken in the past | 35 | 28.5 | 118 | 15.3 | 86 | 14.4 | |
| Have you ever gone to a medically managed weight loss program? | No | — | — | 524 | 68.1 | 412 | 68.8 |
| Yes, I am currently going to one | — | — | 70 | 9.1 | 52 | 8.7 | |
| Yes, I have gone in the past | — | — | 176 | 22.9 | 135 | 22.5 | |
TABLE 3.
Type of weight loss medication for the participants who reported current or past use.
| Type | Content validity (N = 83) | Field test (n = 442) | Follow‐up (n = 344) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Currently having | Had in past | Currently having | Had in past | Currently having | Had in past | |||||||
| N | % | N | % | N | % | N | % | N | % | N | % | |
| Ozempic (semaglutide) | 20 | 24.1 | 26 | 31.3 | 89 | 20.1 | 97 | 21.9 | 70 | 20.3 | 74 | 21.5 |
| Mounjaro (tirzepatide) | 14 | 16.9 | 14 | 16.9 | 140 | 31.7 | 45 | 10.2 | 112 | 32.6 | 34 | 9.9 |
| Wegovy (semaglutide) | 8 | 9.6 | 12 | 14.5 | 43 | 9.7 | 57 | 12.9 | 36 | 10.5 | 38 | 11.0 |
| Zepbound (tirzepatide) | 7 | 8.4 | 10 | 12.0 | 37 | 8.4 | 7 | 1.6 | 26 | 7.6 | 5 | 1.5 |
| Saxenda (liraglutide) | 4 | 4.8 | 9 | 10.8 | 3 | 0.7 | 35 | 7.9 | 2 | 0.6 | 27 | 7.8 |
| Trullicity (dulaglutide) | 5 | 6.0 | 3 | 3.6 | 7 | 1.6 | 15 | 3.4 | 5 | 1.5 | 13 | 3.8 |
| Rybelus | 0 | 0.0 | 0 | 0.0 | 3 | 0.7 | 5 | 1.1 | 3 | 0.9 | 5 | 1.5 |
| Retatrutide | 0 | 0.0 | 0 | 0.0 | 1 | 0.2 | 0 | 0.0 | 1 | 0.3 | 0 | 0.0 |
| Compounded semiglutide | 1 | 1.2 | 0 | 0.0 | 6 | 1.4 | 5 | 1.1 | 6 | 1.7 | 5 | 1.5 |
| Other obesity medication | ||||||||||||
| Phentermine (Adipex‐P) | 6 | 7.2 | 14 | 16.9 | — | — | — | — | — | — | — | — |
| Contrave (bupropion‐naltrexone) | 3 | 3.6 | 11 | 13.3 | — | — | — | — | — | — | — | — |
| Orlistat (Xenical, Alli) | 3 | 3.6 | 17 | 20.5 | — | — | — | — | — | — | — | — |
| Setmelanotide (Imcivree) | 3 | 3.6 | 5 | 6.0 | — | — | — | — | — | — | — | — |
| Qsymia (phentermine/topiramate) | 1 | 1.2 | 7 | 8.4 | — | — | — | — | — | — | — | — |
TABLE 4.
Response option and item results from the 123 content validity participants.
| PRO | Item | Response options | Items | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Easy to understand | Fits experience | Easy to understand | Relevant | ||||||
| N | % | N | % | N | % | N | % | ||
| Physical |
|
123 | 100 | 122 | 99.2 | 123 | 100 | 116 | 94.3 |
| Physical |
|
123 | 100 | 122 | 99.2 | 123 | 100 | 122 | 99.2 |
| Body image |
|
123 | 100 | 123 | 100 | 123 | 100 | 123 | 100 |
| Psychological |
|
123 | 100 | 122 | 99.2 | 120 | 97.6 | 119 | 96.7 |
| Body image |
|
120 | 97.6 | 121 | 98.4 | 121 | 98.4 | 121 | 98.4 |
| Self esteem |
|
122 | 99.2 | 123 | 100 | 122 | 99.2 | 121 | 98.4 |
| Stigma |
|
122 | 99.2 | 121 | 98.4 | 122 | 99.2 | 117 | 95.1 |
| Social |
|
123 | 100 | 123 | 100 | 123 | 100 | 121 | 98.4 |
| Eating |
|
123 | 100 | 123 | 100 | 123 | 100 | 123 | 100 |
| Eating |
|
123 | 100 | 123 | 100 | 123 | 100 | 122 | 99.2 |
| Excess skin |
|
123 | 100 | 121 | 98.4 | 123 | 100 | 112 | 91.1 |
For recall period, participants were asked whether they could reflect on their experiences over the “past week.” Of the 123 participants, 118 (95.9%) said yes. Most participants (N = 86, 69.9%) also felt that the past week was an appropriate time frame, although 34 (27.6%) participants considered it too short and 3 (2.4%) too long. Recommendations for a longer time frame ranged from 2 weeks to more than 1 year, while shorter time frame suggestions ranged from 4 to 5 days.
For missing content, 105 (85.4%) said that no relevant aspects of weight impacting their health were missing. Missing concepts mentioned by at least three participants included energy and sleep quality. Feedback about the scale was overall positive, such as the following: “It gave me a sense of trust in the professionalism of the party posing the questions. The questions themselves were very inoffensive and comfortable to answer. I wouldn't feel awkward while answering the questions or bad after I finished.”
3.2. Field Test
All 11 questions were included in the field test, which involved 770 participants. Table 2 shows the participant characteristics and Table 3 shows the reported use of weight loss medication. Participants ranged in age from 19 to 80 years (mean = 43.2, SD = 11.9). When asked if they were currently trying to lose weight: 618 (80.3%) said yes, 142 (18.4%) said no but they wanted to, and 10 (1.3%) said no. Regarding their weight loss journey: 72 (9.4%) had not started, 331 (43%) were close to the start, 251 (32.6%) were about halfway through, 106 (13.8%) were near completion, and 10 (1.3%) were finished.
Table 5 shows the category frequency responses. For two questions, more than half the sample reported no problems: 479 (62.2%) participants reported they did not experience any unpleasant symptoms due to excess skin caused by weight loss; and 417 (54.2%) participants reported that it was not difficult to walk or move around. The excess skin question also evidenced poor fit to the Rasch model (p < 0.001). After this question was dropped, the remaining 10 questions fit the Rasch model and had ordered thresholds.
TABLE 5.
Category frequencies for the 11 tested items tested in the field test survey.
| PRO | Item | Response categories | |||
|---|---|---|---|---|---|
| 0 | 1 | 2 | 3 | ||
| Physical |
|
36 | 101 | 216 | 417 |
| Physical |
|
79 | 133 | 262 | 296 |
| Body image |
|
314 | 265 | 135 | 56 |
| Psychological |
|
94 | 253 | 266 | 157 |
| Body image |
|
406 | 277 | 76 | 11 |
| Self esteem |
|
214 | 371 | 152 | 33 |
| Stigma |
|
111 | 308 | 241 | 110 |
| Social |
|
67 | 163 | 235 | 305 |
| Eating |
|
174 | 191 | 258 | 147 |
| Eating |
|
115 | 336 | 200 | 119 |
| Excess skin |
|
17 | 94 | 180 | 479 |
Note: Response options varied according to the BODY‐Q‐7D item: 0 = extremely/severe/never; 1 = moderately/moderate/sometimes; 2 = little/mild/often; 3 = not/no/always.
For the three outcomes with two candidate questions, we picked the questions with a more balanced distribution of responses, i.e., “How difficult was it for you to do moderate exercise (e.g., go for a brisk walk)?”, “How difficult was it for you to feel in control when you ate?”, and “How self‐conscious were you about your body?” After reducing the scale, the overall model fit for the 7‐item scale was as follows: chi‐square = 81.29, df = 56, p = 0.02.
Table S3 shows the item fit statistics and DIF results. All seven questions fit the Rasch model with nonsignificant p values after Bonferroni correction and had ordered thresholds (Figure 2). On the Rasch “ruler,” the concept where participants reported most problems was self‐esteem. No DIF was detected for gender. Three questions evidenced DIF, i.e., two questions for age group and two questions for BMI group. However, DIF had minimal impact on scoring, with Pearson correlations between the original and split analyses all ≥ 0.998. No evidence of local dependency was observed (all residual correlations were ≤ 0.20). Reliability was high with PSI values of 0.81 and 0.80 and Cronbach alpha values of 0.83 and 0.82 with and without extremes, respectively. The person‐item threshold distribution (Figure 3) demonstrated good targeting; 759 of 770 (98.6%) participants (upper histogram) scored within the scale range (lower histogram), and floor and ceiling effects were below 1%.
FIGURE 2.
Threshold map.
FIGURE 3.
Person‐item threshold distribution.
Construct validity results are shown in Table 1. Overall, 22 of 28 hypotheses for the initial survey were accepted (79%). Detailed results for group difference are shown in Table S4. Higher scores were reported by participants who were further along in their weight loss journey, more satisfied with their weight (loss), social life, daily functioning, and body appearance, more active, and happier with their eating habits and who had better mental and physical health. Lower scores were observed among participants who wanted MBS, a GLP‐1 receptor agonist, and body contouring and among those who reported weight gain or had health problems related to excess skin.
Detailed results for convergent validity are shown in Table S5. As expected, moderate negative correlations were observed between the BODY‐Q‐7D score and lower BMI (r = −0.3; p < 0.001) and lower amount of weight someone wanted to lose to reach his or her goal weight (r = −0.3; p < 0.001). For convergent validity, as expected, most BODY‐Q scales correlated 0.3–0.5 with the BODY‐Q‐7D scale, indicating the new scale measures a related but dissimilar construct. The Psychological scale (r = 0.6) correlated ≥ 0.5, indicating the BODY‐Q‐7D measured a similar construct as hypothesized. The Physical Function (r = 0.5) and Social (r = 0.5) scales correlated higher than expected (0.3–0.5), indicating that the content of the BODY‐Q‐7D was more similar than hypothesized. The BODY‐Q‐7D correlated ≥ 0.5 with the summary score for the EQ‐5D‐5L (r = 0.5), indicating it measures a similar construct (p < 0.001). Finally, correlations with EQ‐5D‐5L subscales were 0.3–0.5, indicating that they measure a related but dissimilar construct (p < 0.001).
3.3. TRT
Of 365 participants invited, 303 completed the TRT survey. We excluded 27 participants who reported change, 2 who provided an invalid Prolific ID, and 5 who completed the scale outside of 7–14 days. ICC values with and without outliers for the 269 participants were 0.72 (n = 269) and 0.81 (n = 258), respectively. The SDC was 20.3 and 17.8 at the individual level and 1.2 and 1.1 at the group level with and without outliers, respectively.
3.4. Responsiveness
Responsiveness results are shown in Table 1 with detailed results in Tables S6 and S7. A total of 12 of 16 hypotheses (75%) were accepted. Change scores were positive for those who reported improved weight‐related HRQL (mean = 14.7, SD = 14.2; p < 0.001). Participants who reported no change had a mean change score that was significantly different from zero (p < 0.001), with lower confidence bounds slightly above zero (LB = 1.9). Due to kurtosis of three, small sample size (n = 49), and non‐ normal visual plot, nonparametric statistics were applied to examine the group that reported worsening HRQL. A Wilcoxon signed rank test showed that follow‐up scores (median = 39) were lower than baseline scores (median = 42), representing a negative change score (W = −2.5, p = 0.013). As expected, greater improvement (p < 0.001) was observed for participants who lost weight, were satisfied with weight change, and experienced a larger BMI reduction. There was a negative correlation between the change score and change in BMI, which was weaker than expected (r = −0.2; p < 0.001). The composite Likert change score for weight‐related HRQL had a moderate positive Spearman correlation with the BODY‐Q‐7D change score (r = 0.3; p < 0.001).
For convergent validity the correlations with the BODY‐Q‐7D change score and change scores for other BODY‐Q scales were as expected (r ≥ 0.3, p < 0.001) with the exception of Physical Function, which was weaker than expected (r = 0.2, p < 0.001). BODY‐Q Social, Sexual, and Eating Behavior change scores, as predicted, had weaker correlations with the BODY‐Q‐7D change score (≤ 0.001). The correlation for the BODY‐Q‐7D and overall EQ‐5D‐5L change scores was lower than predicted (r = 0.2, p < 0.001).
4. Discussion
Building on prior BODY‐Q research [16, 17, 18] and the S.Q.O.T. initiative to standardize outcomes in multidisciplinary obesity care [12, 13, 14], we developed the BODY‐Q‐7D as a concise, rigorously constructed measure designed to support a future PBM. The instrument captures seven key PRO domains identified as important to people living with obesity and healthcare providers, ensuring that the resulting health states reflect outcomes that matter. Content validity was established in an international sample of 123 individuals living with obesity—more than twice the COSMIN‐recommended minimum for survey‐based content validity studies [20, 22]. Field test data further demonstrated reliability, construct validity, structural integrity consistent with RMT, and evidence of responsiveness. This staged development approach, consistent with other condition‐specific PBM programs [33, 34, 35, 36], prioritizes the establishment of a psychometrically robust descriptive system before undertaking valuation. By doing so, it ensures that the defined health states are measurement‐sound and suitable for subsequent utility modeling and QALY estimation.
A question that we tested but found that it did not perform adequately within the BODY‐Q‐7D scale was dropped (i.e., “Did you experience any unpleasant symptoms due to excess skin caused by weight loss [e.g., rash, infections, odor]?”). This question evidenced poor fit to the Rasch model and had a high ceiling effect. Although it was an important concept according to the S.Q.O.T. initiative, having excess skin and experiencing skin‐related symptoms are not universally applicable to people who lose weight. Without a specific question about skin, the 7‐item scale should indirectly measure concerns for people with excess skin. Participants in the study who wanted body contouring to remove excess skin reported lower scores on the BODY‐Q‐7D than those who do not want body contouring. Furthermore, BODY‐Q‐7D scores were lower for participants who reported their excess skin problem to be somewhat or very bad compared with those who reported their skin problem as a little bad or not at all. In addition, research has shown that scores on the BODY‐Q Excess Skin scale correlate moderately or strongly with other BODY‐Q scales, including Body Image (r = 0.58), Body Appearance (r = 0.50), Sexual (r = 0.48), Psychological (r = 0.47), and Social (r = 0.40) [17].
An important finding is that self‐esteem was the concept with the highest item location (difficulty) in the RMT analysis. This finding aligns with the S.Q.O.T. initiative finding that self‐esteem was the most important PRO for people living with obesity [12]. We also found that the 3 items that ranked lowest in the Rasch hierarchy for the BODY‐Q‐7D scale (i.e., social, physical, psychological) are concepts included in generic PBMs, such as the EQ‐5D‐5L [24] and the SF‐6D [37]. These findings reinforce the need for an obesity‐specific PBM that asks about important concepts to patients, including self‐esteem, stigma, eating concerns, and body image.
The next step in the development of the BODY‐Q‐7D will be to design and implement a valuation study to create preference‐based scoring for its health states. Evidence is required to demonstrate that the BODY‐Q‐7D performs coherently in applied economic evaluations. This research will include the collection of evidence of discriminatory power across clinically meaningful groups, sensitivity to change at the utility level, and plausible behavior of utility decrements across the full severity (e.g., mild, moderate, severe) spectrum. Together, these properties are essential for ensuring that the BODY‐Q‐7D can support valid estimation of QALYs and inform cost‐effectiveness analyses. Given the time and cost associated with valuation studies [38], it was important that we conduct the groundwork to develop a psychometrically robust scale using a modern psychometric method before undertaking valuation. The BODY‐Q‐7D was shown to perform well across COSMIN criteria [20, 21, 22], as it demonstrated content validity, internal consistency, reliability, construct validity, and responsiveness.
Our study has some limitations. We used an online platform to recruit a sample of people living with obesity. We double screened to exclude anyone whose answers changed for data quality purposes, but clinical and demographic data could not be independently verified. While our sample is heterogeneous, the sample was limited to English speaking countries and included more people who identify as White, female, and in obesity Class 3. We had over 100 people who did not respond to our invitation to complete the 3‐month follow‐up survey, and the reason for nonresponse could not be determined.
5. Conclusion
The BODY‐Q‐7D demonstrated strong psychometric properties in a large heterogeneous international sample of people living with obesity. This short, rigorously designed scale functions as a PROM and was structured to allow for preference‐based scoring. The next step for this research is to collect preference‐based weights for the health states using an established approach. Further information about the BODY‐Q PBM can be accessed as https://qportfolio.org/body‐q/.
Funding
This study was funded by internal research funding provided to Anne Klassen through the Klassen's Department at McMaster University.
Disclosure
The BODY‐Q was authored by Drs. Anne Klassen, Andrea Pusic, and Stefan Cano. The copyright to the work is owned by Memorial Sloan‐Kettering Cancer Center, Memorial Hospital for Cancer and Allied Diseases, Sloan‐Kettering Institute for Cancer Research (New York City, New York, USA), McMaster University (Hamilton, Ontario, Canada), Brigham and Women's Hospital (Boston, Massachusetts, USA), and Stefan Cano. The developers receive a share of license revenues as royalties for its use in for‐profit research based on their institution's inventor sharing policy.
Conflicts of Interest
Stefan Cano reported being president of Modus Outcomes outside the submitted work. Anne Klassen reported research consulting through EVENTUM Research outside the submitted work. The other authors declare no conflicts of interest.
Supporting information
Table S1: Questions used to determine content validity.
Table S2: Psychometric tests performed.
Table S3: Rasch Measurement Theory (RMT) analysis item fit statistics and differential item functioning (DIF).
Table S4: Construct validity results.
Table S5: Convergent validity results.
Table S6: Construct validity for change scores.
Table S7: Convergent validity results for change scores.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. World Obesity, “World Obesity Atlas 2023,” published March 2023, https://www.worldobesityday.org/assets/downloads/World_Obesity_Atlas_2023_Report.pdf.
- 2. Nygaard Flølo T., Liu H., Andersen J. R., and Kolotkin R., “Exploring the Relationship Between Obesity, Weight Loss and Health‐Related Quality of Life: An Updated Systematic Review of Reviews,” Clinical Obesity 16 (2025): e70049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Colquitt J. L., Pickett K., Loveman E., and Frampton G. K., “Surgery for Weight Loss in Adults,” Cochrane Database of Systematic Reviews 8 (2014): CD003641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Peirson L., Douketis J., Ciliska D., Fitzpatrick‐Lewis D., Ali M. U., and Raina P., “Treatment for Overweight and Obesity in Adult Populations: A Systematic Review and Meta‐Analysis,” Journal of Evidence‐Based Medicine 2, no. 4 (2014): E306–E317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Jastreboff A. M., Aronne L. J., Ahmad N. N., et al., “Tirzepatide Once Weekly for the Treatment of Obesity,” New England Journal of Medicine 387, no. 3 (2022): 205–216. [DOI] [PubMed] [Google Scholar]
- 6. Wilding J. P. H., Batterham R. L., Calanna S., et al., “Once‐Weekly Semaglutide in Adults With Overweight or Obesity,” New England Journal of Medicine 384 (2021): 989–1002. [DOI] [PubMed] [Google Scholar]
- 7. Xia Q., Campbell J. A., Ahmad H., et al., “Health State Utilities for Economic Evaluation of Bariatric Surgery: A Comprehensive Systematic Review and Meta‐Analysis,” Obesity Reviews 21, no. 8 (2020): e13028. [DOI] [PubMed] [Google Scholar]
- 8. Sauerland S., Weiner S., Dolezalova K., et al., “Mapping Utility Scores From a Disease‐Specific Quality‐Of‐Life Measure in Bariatric Surgery Patients,” Value in Health 12, no. 2 (2009): 364–370. [DOI] [PubMed] [Google Scholar]
- 9. Guo W., Xie S., Wang D., and Wu J., “Mapping the IWQOL‐Lite Onto EQ‐5D‐5L and SF‐6Dv2 Among Overweight and Obese Adults,” Quality of Life Research 33 (2024): 737–749, 10.1007/s11136-023-03598-9. [DOI] [PubMed] [Google Scholar]
- 10. Brazier J. E., Kolotkin R. L., Crosby R. D., and Williams G. R., “Estimating a Preference‐Based Single Index for the Impact of Weight on Quality of Life‐Lite (IWQOL‐Lite) Instrument From the SF‐6D,” Value in Health 7, no. 4 (2004): 490–498. [DOI] [PubMed] [Google Scholar]
- 11. Sun S., Stenberg E., Cao Y., et al., “Mapping the Obesity Problems Scale to the SF‐6D: Results Based on the Scandinavian Obesity Surgery Registry (SOReg),” European Journal of Health Economics 24, no. 2 (2023): 279–292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. de Vries C. E. E., Terwee C. B., Al Nawas M., et al., “Outcomes of the First Global Multidisciplinary Consensus Meeting Including Persons Living With Obesity to Standardize Patient‐Reported Outcome Measurement in Obesity Treatment Research,” Obesity Reviews 23, no. 8 (2022): e13452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Dijkhorst P. J., de Vries C. E. E., Terwee C. B., et al., “A Core Set of Patient‐Reported Outcome Measures to Measure Quality of Life in Obesity Treatment Research,” Obesity Reviews 26, no. 2 (2025): e13849. [DOI] [PubMed] [Google Scholar]
- 14. Dijkhorst P. J., Monpellier V. M., Terwee C. B., et al., “Core Set of Patient‐Reported Outcome Measures for Measuring Quality of Life in Clinical Obesity Care,” Obesity Surgery 34, no. 8 (2024): 2980–2990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. de Vries C. E. E., Kalff M. C., Prinsen C. A. C., et al., “Recommendations on the Most Suitable Quality‐Of‐Life Measurement Instruments for Bariatric and Body Contouring Surgery: A Systematic Review,” Obesity Reviews 19, no. 10 (2018): 1395–1411. [DOI] [PubMed] [Google Scholar]
- 16. Klassen A. F., Cano S. J., Scott A., Tsangaris E., and Pusic A. L., “Assessing Outcomes in Body Contouring,” Clinics in Plastic Surgery 41, no. 4 (2014): 645–654, 10.1016/j.cps.2014.06.004. [DOI] [PubMed] [Google Scholar]
- 17. Klassen A. F., Cano S. J., Alderman A., et al., “The BODY‐Q: A Patient‐Reported Outcome Instrument for Weight Loss and Body Contouring Treatments,” Plastic and Reconstructive Surgery. Global Open 4, no. 4 (2016): e679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. De Vries C. E. E., Mou D., Poulsen L., et al., “Development and Validation of New BODY‐Q Scales Measuring Expectations, Eating Behavior, Distress, Symptoms, and Work Life in 4004 Adults From 4 Countries,” Obesity Surgery 31, no. 8 (2021): 3637–3645. [DOI] [PubMed] [Google Scholar]
- 19. US Food and Drug Administration , Guidance for Industry: Patient‐Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims, (US Department of Health and Human Services, 2009), https://www.fda.gov/media/77832/download.
- 20. Terwee C. B., Prinsen C., Chiarotto A., et al., “COSMIN Methodology for Assessing the Content Validity of PROMs—User Manual,” published February 2018, https://www.cosmin.nl/wp‐content/uploads/COSMIN‐methodology‐for‐content‐validity‐user‐manual‐v1.pdf.
- 21. Gagnier J. J., Lai J., Mokkink L. B., and Terwee C. B., “COSMIN Reporting Guideline for Studies on Measurement Properties of Patient‐Reported Outcome Measures,” Quality of Life Research 30, no. 8 (2021): 2197–2218. [DOI] [PubMed] [Google Scholar]
- 22. Mokkink L. B., Prinsen C., Patrick D. L., et al., “COSMIN Study Design Checklist for Patient‐Reported Outcome Measurement Instruments,” published July 2019, https://gut.bmj.com/content/gutjnl/70/1/139/DC1/embed/inline‐supplementary‐material‐1.pdf.
- 23. Reeve B. B., Wyrwich K. W., Wu A. W., et al., “ISOQOL Recommends Minimum Standards for Patient‐Reported Outcome Measures Used in Patient‐Centered Outcomes and Comparative Effectiveness Research,” Quality of Life Research 22, no. 8 (2013): 1889–1905. [DOI] [PubMed] [Google Scholar]
- 24. Herdman M., Gudex C., Lloyd A., et al., “Development and Preliminary Testing of the New Five‐Level Version of EQ‐5D (EQ‐5D‐5L),” Quality of Life Research 20, no. 10 (2011): 1727–1736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Devji T., Carrasco‐Labra A., Qasim A., et al., “Evaluating the Credibility of Anchor Based Estimates of Minimal Important Differences for Patient Reported Outcomes: Instrument Development and Reliability Study,” BMJ 369 (2020): m1714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Rasch G., “A Mathematical Theory of Objectivity and Its Consequences for Model Construction,” European Meeting on Statistics, Econometrics and Management Sciences (Amsterdam: 1968), https://rasch.org/memo1968.pdf.
- 27. Andrich D., Rasch Models for Measurement (Sage Publications, 1988). [Google Scholar]
- 28. Hobart J. and Cano S., “Improving the Evaluation of Therapeutic Interventions in Multiple Sclerosis: The Role of New Psychometric Methods,” Health Technology Assessment 13 (2009): 1–177, 10.3310/hta13120. [DOI] [PubMed] [Google Scholar]
- 29. Cano S. J. and Hobart J. C., “The Problem With Health Measurement,” Patient Preference and Adherence 5 (2011): 279–290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Geerinck A., Alekna V., Beaudart C., et al., “Standard Error of Measurement and Smallest Detectable Change of the Sarcopenia Quality of Life (SarQoL) Questionnaire: An Analysis of Subjects From 9 Validation Studies,” PLoS One 14, no. 4 (2019): e0216065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Smit E. B., Bouwstra H., Roorda L. D., et al., “A Patient‐Reported Outcomes Measurement Information System Short Form for Measuring Physical Function During Geriatric Rehabilitation: Test‐Retest Reliability, Construct Validity, Responsiveness, and Interpretability,” Journal of the American Medical Directors Association 22, no. 8 (2021): 1627–1632. [DOI] [PubMed] [Google Scholar]
- 32. Kim H. Y., “Statistical Notes for Clinical Researchers: Assessing Normal Distribution (2) Using Skewness and Kurtosis,” Restorative Dentistry and Endodontics 38, no. 1 (2013): 52–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Young T. A., Rowen D., Norquist J., and Brazier J. E., “Developing Preference‐Based Health Measures: Using Rasch Analysis to Generate Health State Values,” Quality of Life Research 19, no. 6 (2010): 907–917. [DOI] [PubMed] [Google Scholar]
- 34. Mavranezouli I., Brazier J. E., Young T. A., and Barkham M., “Using Rasch Analysis to Form Plausible Health States Amenable to Valuation: The Development of CORE‐6D From a Measure of Common Mental Health Problems (CORE‐OM),” Quality of Life Research 20, no. 3 (2011): 321–333. [DOI] [PubMed] [Google Scholar]
- 35. Sundaram M., Smith M. J., Revicki D., Elswick B., and Miller L. A., “Rasch Analysis Informed the Development of a Classification System for a Diabetes‐Specific Preference‐Based Measure of Health,” Journal of Clinical Epidemiology 62, no. 8 (2009): 845–856. [DOI] [PubMed] [Google Scholar]
- 36. Keetharuth A. D., Rowen D., Bjorner J. B., and Brazier J. E., “Estimating a Preference‐Based Index for Mental Health From the Recovering Quality of Life Measure,” Value in Health 24, no. 8 (2021): 1160–1172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Brazier J. E., Usherwood T., Harper R., and Thomas K., “Deriving a Preference‐Based Single Index From the UK SF‐36 Health Survey,” Journal of Clinical Epidemiology 51, no. 11 (1998): 1115–1128. [DOI] [PubMed] [Google Scholar]
- 38. Brazier J. E., 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 Technology Assessment 16, no. 32 (2012): 1–114, 10.3310/hta16320. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Questions used to determine content validity.
Table S2: Psychometric tests performed.
Table S3: Rasch Measurement Theory (RMT) analysis item fit statistics and differential item functioning (DIF).
Table S4: Construct validity results.
Table S5: Convergent validity results.
Table S6: Construct validity for change scores.
Table S7: Convergent validity results for change scores.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.