Benjamin Matthew Craig

Objectives: Introduced in 2024, the PROMIS®-16 is a health-related quality of life (HRQoL) instrument that measures eight dimensions: physical function, ability to participate in social roles and activities, anxiety, depression, sleep disturbance, pain interference, cognitive function, and fatigue. Given the potential overlap in dimensions, this study compares the measurement and psychometric properties of the PROMIS-16 and EQ-5D-5L. Methods: In 2024, an online cross-sectional survey was conducted with a nationally representative sample of U.S. adults aged 18 years or older (N = 2577), including the EQ-5D-5L and PROMIS-16 as stand-alone instruments. We compared the measurement properties of the PROMIS-16 and the EQ-5D-5L, namely ceiling and floor percentages, informativity (Shannon’s indices), and convergent/divergent validity. Furthermore, we conducted an exploratory factor analysis to identify their combined structure. Results: Within construct-overlapping items, the PROMIS-16 items exhibited greater dispersion, including lower ceiling percentages and higher informativity, compared to their EQ-5D-5L counterparts. Good convergent validity between PROMIS-16 and EQ-5D-5L items was observed for physical functioning (mobility; r = 0.85) and anxiety/depression (r = 0.76–0.79). EFA captured six distinct constructs: Physical Function, Social Roles & Activities, Anxiety / Depression, Pain Interference, Cognitive Function, and Sleep / Fatigue. Possibly due to reverse coding, the two Cognitive Function items of PROMIS-16 showed inconsistent performance, with weak correlations and limited alignment with other constructs. Conclusions: The PROMIS-16 items demonstrated greater conceptual coverage and informativity than the EQ-5D-5L, but issues with inconsistent framing and potential response biases remained. On the other hand, the EQ-5D-5L demonstrated stronger concordance with respondents’ global health assessments compared to the PROMIS-16.

Section Objectives EQ-5D-5L items, self-care (SC), pain/discomfort (PD), and anxiety/depression (AD), are composites of two components. While PD (comprised of pain [PN] and discomfort [DI]) and AD (comprised of anxiety [AN] and depression [DE]) have been extensively studied, SC (comprised of washing self [WA] and dressing self [DR]) remains underexplored. Additionally, to our knowledge, the psychometric performance of composites versus components has never been investigated across age groups. This study examines the three composites and six components overall and by age group. AbstractSection Methods In 2024, a representative sample of 6,018 US adults participated in an online survey that included the EQ-5D-5L and its six components in fixed order. In the overall sample and separately across age groups (18–29, 30–39, 40–49, 50–59, 60–69, 70+), we evaluated underreporting, ceilings and floors, agreement (weighted Kappa), convergent validity, informativity, explanatory power for EQ VAS (R²), and structural validity (exploratory factor analysis [EFA]). AbstractSection Results The ceiling was lower across all three composites compared to their component items. Strong agreement was found between composites and their corresponding component items (SC: 0.78, PD: 0.91, AD: 0.92). Using component items instead of composites increased the informativity and adjusted R² values. The first-listed components showed stronger convergent validity with their composites.EFA identified three factors: physical functioning (including MO, UA, SC), PD, and AD. Splitting the AD for younger and PD for older populations showed the greatest added value while there were minimal differences between SC and the module of washing and dressing self. AbstractSection Conclusions The EQ-5D-5L composites effectively captured the information provided by their individual components, but replacing PD and AD with their components may improve instrument performance and its adaptability across health concerns relevant to different age groups.

Introduction The experimental EuroQol Toddler and Infant Populations (EQ-TIPS) instrument is currently under development as a health-related quality of life (HRQoL) measure for toddlers and infants aged 0–36 months. Using this protocol, researchers can conduct surveys with discrete choice experiments (DCEs) that examine the key properties of HRQoL instruments, specifically whether severity aligns with preferences across each attribute and the extent to which attributes influence choices. To demonstrate this protocol, we will conduct two waves of DCE surveys using different choice tasks and a common scenario (a 1-month health episode for a 1-year-old child). Methods and analysis In the first wave (a general population sample of 400 Australian adults; 14 kaizen tasks each), respondents will view a single EQ-TIPS-five-level (v3.0) profile for each task and be asked to make a series of choices that sequentially alleviate the child’s health problems. Using this exploratory evidence, we will assess whether EQ-TIPS severity aligns with preferences and the extent to which each attribute level influences choices (ie, main effects). In the second wave (1000 Australian adults; 28 paired comparisons), respondents will see two EQ-TIPS profiles for each task and be asked to choose between them. Using this confirmatory evidence, we will compare the main effects and their uncertainty by wave. For each DCE, we will estimate a main-effects conditional logit model and test for differences in effects using cluster bootstrap techniques. As sensitivity analyses, we will evaluate the effects of task sequence, attribute order and sample size on uncertainty in each wave. Ethics and dissemination The independent review board at Includovate evaluated the application for ethical clearance and approved the study on 19 February 2025. To disseminate our findings, we will prepare multiple manuscripts for publication in peer-reviewed journals and present highlights at scientific meetings, such as the EuroQol Plenary Meeting and the International Academy of Health Preference Research.

Introduction A typical health preference study conducts a single discrete choice experiment (DCE). For example, a health valuation study may elicit preferences on an individual’s health-related quality of life along five EQ-5D-5L attributes (Mobility, Self-care, Usual Activities, Pain/Discomfort, Anxiety/Depression). Using this protocol, researchers can conduct a dual-DCE survey (ie, with two different full-block DCEs completed sequentially). To demonstrate this protocol, we will conduct 12 dual-DCE surveys in two waves and estimate the effects of different scenarios and descriptive systems on main effects (ie, incremental differences in value between levels). Methods and analysis Each of the two DCEs in a dual-DCE survey equates to a stand-alone health valuation study. To demonstrate this protocol, each is an EQ-5D-5L valuation study, including d-efficient blocks of 15 kaizen tasks and 5 paired comparisons. In wave 1 (six surveys, 1000 US adults each), the two DCEs will differ by scenario (1-year episodes ending in recovery or death or no duration/ending described). In wave 2 (six surveys, 200 US adults each), the two DCEs will include the same 5 EQ-5D-5L attributes but differ by the number of additional attributes related to cognition: none, one composite attribute (memory/concentration) and two component attributes (memory, concentration). For each DCE, we will estimate a conditional logit model and test for differences in value using cluster bootstrap techniques. We hypothesise that the values will differ by scenarios and systems. As secondary analyses, we assess the effects of sampling, scenario/system order and DCE order. Ethics and dissemination The independent review board (IRB) at Advarra determined that this research project (Pro00080475; 11 July 2024) is exempt from IRB oversight based on the Department of Health and Human Services regulations found at 45 CFR 46.104(d)(2). Furthermore, the IRB determined that the project is not subject to requirements for continuing review. To disseminate our findings, we will prepare multiple manuscripts for publication in peer-reviewed journals and present highlights at scientific meetings, such as the EuroQol Plenary Meeting, International Academy of Health Preference Research and ISPOR.

Background A key challenge in value assessment is how to summarize effectiveness, particularly the impact of interventions on patient health-related quality of life (HRQoL). One approach is to quantify the gains in HRQoL and life expectancy together as quality-adjusted life years (QALYs); however, this approach has faced various criticisms regarding its potential discriminatory aspects toward persons with disabilities, older adults, and the most vulnerable individuals in society. Methods Instead of QALYs, we provide an alternative approach that summarizes HRQoL gains from the perspective of its stakeholders (e.g., patients, parents, and caregivers) using an “experience” scale. On an experience scale, a positive value signifies an experience better than having no experience at all, while a negative value indicates an experience worse than having no experience. To illustrate the merits of this approach, we examine US preferences on the relief of child health problems, namely a discrete choice experiment (DCE) with kaizen tasks and alternatives described using the EQ-5D-Y-3L. Results Using this approach, we demonstrate the differences in perspectives between parents (N = 179), mothers (N = 99), and fathers (N = 80) of children younger than 18 years of age, as well as the feasibility of this patient-centered approach using a brief DCE survey of less than 100 respondents each (and without QALYs). Specifically, we found that mothers place a higher value on the child’s feelings than fathers. The results also suggest other differences between the perspectives of mothers and fathers, but these differences were not statistically significant (p-values < .05). Conclusions We put forth that future value assessments may summarize gains in HRQoL on a patient experience scale (i.e., experience scale from the patient perspective) to inform decision-making.

Background Kaizen is a Japanese term for continuous improvement (kai ~ change, zen ~ good). In a kaizen task, a respondent makes sequential choices to improve an object’s profile, revealing a preference path. Including kaizen tasks in a discrete choice experiment has the advantage of collecting greater preference evidence than pick-one tasks, such as paired comparisons. Objective and Methods So far, three online discrete choice experiments have included kaizen tasks: the 2020 US COVID-19 vaccination (CVP) study, the 2021 UK Children’s Surgery Outcome Reporting (CSOR) study, and the 2023 US EQ-5D-Y-3L valuation (Y-3L) study. In this evidence synthesis, we describe the performance of the kaizen tasks in terms of response behaviors, conditional logit and Zermelo–Bradley–Terry (ZBT) estimates, and their standard errors in each of the surveys. Results Comparing the CVP and Y-3L, including hold-outs (i.e., attributes shared by all alternatives) seems to reduce positional behavior by half. The CVP tasks excluded multi-level improvements; therefore, we could not estimate logit main effects directly. In the CSOR, only 12 of the 21 logit estimates are significantly positive (p < 0.05), possibly due to the fixed attribute order. All Y-3L estimates are significantly positive, and their predictions are highly correlated (Pearson: logit 0.802, ZBT 0.882) and strongly agree (Lin: logit 0.744, ZBT 0.852) with the paired-comparison probabilities. Conclusions These discrete choice experiments offer important lessons for future studies: (1) include warm-up tasks, hold-outs, and multi-level improvements; (2) randomize the attribute order (i.e., up-down) at the respondent level; and (3) recruit smaller samples of respondents than traditional discrete choice experiments with only pick-one tasks.

In a recent study, it was demonstrated that the maximum simulated likelihood (MSL) estimator produces significant biases when applied to the bivariate normal and bivariate Poisson-lognormal models. The study’s conclusion suggests that similar biases could be present in other models generated by correlated bivariate normal structures, which include several commonly used specifications of the mixed logit (MIXL) models. This paper conducts a simulation study analyzing the MSL estimation of the error components (EC) MIXL. We find that the MSL estimator produces significant biases in the estimated parameters. The problem becomes worse when the true value of the variance parameter is small and the correlation parameter is large in magnitude. In some cases, the biases in the estimated marginal effects are as large as 12% of the true values. These biases are largely invariant to increases in the number of Halton draws.

Objectives Develop a score summarising how successfully a child with any surgical condition has been treated, and test the clinical validity of the score. Design Discrete choice experiment (DCE), and secondary analysis of data from six UK-wide prospective cohort studies. Participants 253 people with lived experience of childhood surgical conditions, 114 health professionals caring for children with surgical conditions and 753 members of the general population completed the DCE. Data from 1383 children with surgical conditions were used in the secondary analysis. Main outcome measures Normalised importance value of attribute (NIVA) for number/type of operations, hospital-treated infections, quality of life and duration of survival (reference attribute). Results Quality of life and duration of survival were the most important attributes in deciding whether a child had been successfully treated. Parents, carers and previously treated adults placed equal weight on both attributes (NIVA=0.996; 0.798 to 1.194). Healthcare professionals placed more weight on quality of life (NIVA=1.469; 0.950 to 1.987). The general population placed more weight on survival (NIVA=0.823; 95% CI 0.708 to 0.938). The resulting score (the Children’s Surgery Outcome Reporting (CSOR) Treatment Success Score (TSS)) has the best possible value of 1, a value of 0 describes palliation and values less than 0 describe outcomes worse than palliation. CSOR TSSs varied clinically appropriately for infants whose data were included in the UK-wide cohort studies. Conclusions The CSOR TSS summarises how successfully children with surgical conditions have been treated, and can therefore be used to compare hospitals’ observed and expected outcomes.

Introduction A decade ago, the first national valuation study of the EQ-5D-Y-3L (Y-3L) involved a discrete choice experiment (DCE) that asked 4155 US adult respondents to complete 40 paired comparisons, choosing between two dying children. Instead of choosing between dying children, the respondents in this novel protocol are asked whether ‘being in a coma’ is better or worse than experiencing ‘health problems’ (ie, experience scale) and how they would relieve health problems (ie, kaizen tasks). Our aims are to compare the preference evidence of the paired comparison and kaizen tasks and to conduct a DCE for the valuation of Y-3L profiles on an experience scale. Methods and analysis Under this protocol, we will conduct an online survey that collects preference evidence from 600 US adult respondents on the health of a 10-year-old child for a week. Across all scenarios, each child will be described as either being ‘in a coma’ or having ‘health problems’, namely five three-level attributes (Y-3L). In this DCE, each respondent will be randomly assigned to one of four D-efficient blocks, including five coma comparisons (ie, Y-3L vs coma), 10 paired comparisons (Y-3L vs Y-3L) and 10 kaizen tasks (preference paths). In addition to comparing evidence by task (aim 2), the analysis plan includes the estimation of main-effects conditional logit models to create a Y-3L value set on an ‘experience scale’ where positive (negative) experiences have positive (negative) values (0 is ‘being in a coma’ and 1 is full health). Ethics and dissemination The institutional review board (IRB) (Advarra) determined that this project (Pro00072276) is exempt from IRB oversight based on DHHS 45 CFR 46.104(d)(2) and is not subject to requirements for continuing review. The results will be prepared for publication in peer-reviewed journals and presented at scientific meetings. The data and code will be made available on reasonable request.

Health preference research (HPR) is the subfield of health economics dedicated to understanding the value of health and health-related objects using observational or experimental methods. In a discrete choice experiment (DCE), the utility of objects in a choice set may differ systematically between persons due to interpersonal heterogeneity (e.g., brand-name medication, generic medication, no medication). To allow for interpersonal heterogeneity, choice probabilities may be described using logit functions with fixed individual-specific parameters. However, in practice, a study team may ignore heterogeneity in health preferences and estimate a conditional logit (CL) model. In this simulation study, we examine the effects of omitted variance and correlations (i.e., omitted heterogeneity) in logit parameters on the estimation of the coefficients, willingness to pay (WTP), and choice predictions. The simulated DCE results show that CL estimates may have been biased depending on the structure of the heterogeneity that we used in the data generation process. We also found that these biases in the coefficients led to a substantial difference in the true and estimated WTP (i.e., up to 20%). We further found that CL and true choice probabilities were similar to each other (i.e., difference was less than 0.08) regardless of the underlying structure. The results imply that, under preference heterogeneity, CL estimates may differ from their true means, and these differences can have substantive effects on the WTP estimates. More specifically, CL WTP estimates may be underestimated due to interpersonal heterogeneity, and a failure to recognize this bias in HPR indirectly underestimates the value of treatment, substantially reducing quality of care. These findings have important implications in health economics because CL remains widely used in practice.