Mapping the EORTC QLQ-C30 onto the EQ-5D-3L: assessing the external validity of existing mapping algorithms

Purpose

To determine the external validity of existing mapping algorithms for predicting EQ-5D-3L utility values from EORTC QLQ-C30 responses and to establish their generalizability in different types of cancer.

Methods

A main analysis (pooled) sample of 3560 observations (1727 patients) and two disease severity patient samples (496 and 93 patients) with repeated observations over time from Cancer 2015 were used to validate the existing algorithms. Errors were calculated between observed and predicted EQ-5D-3L utility values using a single pooled sample and ten pooled tumour type-specific samples. Predictive accuracy was assessed using mean absolute error (MAE) and standardized root-mean-squared error (RMSE). The association between observed and predicted EQ-5D utility values and other covariates across the distribution was tested using quantile regression. Quality-adjusted life years (QALYs) were calculated using observed and predicted values to test responsiveness.

Results

Ten ‘preferred’ mapping algorithms were identified. Two algorithms estimated via response mapping and ordinary least-squares regression using dummy variables performed well on number of validation criteria, including accurate prediction of the best and worst QLQ-C30 health states, predicted values within the EQ-5D tariff range, relatively small MAEs and RMSEs, and minimal differences between estimated QALYs. Comparison of predictive accuracy across ten tumour type-specific samples highlighted that algorithms are relatively insensitive to grouping by tumour type and affected more by differences in disease severity.

Conclusions

Two of the ‘preferred’ mapping algorithms suggest more accurate predictions, but limitations exist. We recommend extensive scenario analyses if mapped utilities are used in cost-utility analyses.

     

Generic and disease-specific estimates of quality of life in macular degeneration: mapping the MacDQoL onto the EQ-5D-3L

Purpose

The macular degeneration quality of life (MacDQoL) instrument is a validated condition-specific measure of quality of life in patients with macular degeneration. This paper presents the first mapping algorithm to predict EQ-5D from responses to the MacDQoL instrument.

Methods

Responses to the MacDQoL and EQ-5D-3L instruments from 482 patients were collected from the IVAN multicentre trial of two alternative drug treatments for neovascular age-related macular degeneration. Regression specifications were estimated using OLS, censored least absolute deviation, Tobit and two-part models. Their predictive performance was assessed using mean squared error. An internal validation sample based on a random selection of 25 % of patients was used to assess the performance of the model estimated on the remaining 75 % of patients.

Results

A two-part model had the best predictive performance on the full sample. The covariates of this model include responses and weighted impact scores for all 23 condition-specific domains of the MacDQoL, and responses to a general MacDQoL quality of life question. The selected models were successful at predicting means and standard deviations of target populations, but prediction is weaker at the upper and lower extremes of the EQ-5D-3L distribution.

Conclusion

The mapping algorithms provide a means of predicting EQ-5D-3L index scores from MacDQoL scores, and could facilitate cost-effectiveness analyses when the latter but not the former are available to researchers. Further validation of the performance of the algorithms using external data would provide a means of establishing the robustness of the algorithms.

     

Testing alternative regression models to predict utilities: mapping the QLQ-C30 onto the EQ-5D-5L and the SF-6D

Purpose

The purpose of the study was to compare alternative statistical techniques to find the best approach for converting QLQ-C30 scores onto EQ-5D-5L and SF-6D utilities, and to estimate the mapping algorithms that best predict these health state utilities.

Methods

772 cancer patients described their health along the cancer-specific instrument (QLQ-C30) and two generic preference-based instruments (EQ-5D-5L and SF-6D). Seven alternative regression models were applied: ordinary least squares, generalized linear model, extended estimating equations (EEE), fractional regression model, beta binomial (BB) regression, logistic quantile regression and censored least absolute deviation. Normalized mean absolute error (NMAE), normalized root mean square error (NRMSE), r-squared (r²) and concordance correlation coefficient (CCC) were used as model performance criteria. Cross-validation was conducted by randomly splitting internal dataset into two equally sized groups to test the generalizability of each model.

Results

In predicting EQ-5D-5L utilities, the BB regression performed best. It gave better predictive accuracy in terms of all criteria in the full sample, as well as in the validation sample. In predicting SF-6D, the EEE performed best. It outperformed in all criteria: NRMSE?=?0.1004, NMAE?=?0.0798, CCC?=?0.842 and r²?=?72.7% in the full sample, and NRMSE?=?0.1037, NMAE?=?0.0821, CCC?=?0.8345 and r²?=?71.4% in cross-validation.

Conclusions

When only QLQ-C30 data are available, mapping provides an alternative approach to obtain health state utility data for use in cost-effectiveness analyses. Among seven alternative regression models, the BB and the EEE gave the most accurate predictions for EQ-5D-5L and SF-6D, respectively.

     

Do country-specific preference weights matter in the choice of mapping algorithms? The case of mapping the Diabetes-39 onto eight country-specific EQ-5D-5L value sets

Purpose

To develop mapping algorithms that transform Diabetes-39 (D-39) scores onto EQ-5D-5L utility values for each of eight recently published country-specific EQ-5D-5L value sets, and to compare mapping functions across the EQ-5D-5L value sets.

Methods

Data include 924 individuals with self-reported diabetes from six countries. The D-39 dimensions, age and gender were used as potential predictors for EQ-5D-5L utilities, which were scored using value sets from eight countries (England, Netherland, Spain, Canada, Uruguay, China, Japan and Korea). Ordinary least squares, generalised linear model, beta binomial regression, fractional regression, MM estimation and censored least absolute deviation were used to estimate the mapping algorithms. The optimal algorithm for each country-specific value set was primarily selected based on normalised root mean square error (NRMSE), normalised mean absolute error (NMAE) and adjusted-r². Cross-validation with fivefold approach was conducted to test the generalizability of each model.

Results

The fractional regression model with loglog as a link function consistently performed best in all country-specific value sets. For instance, the NRMSE (0.1282) and NMAE (0.0914) were the lowest, while adjusted-r² was the highest (52.5%) when the English value set was considered. Among D-39 dimensions, the energy and mobility was the only one that was consistently significant for all models.

Conclusions

The D-39 can be mapped onto the EQ-5D-5L utilities with good predictive accuracy. The fractional regression model, which is appropriate for handling bounded outcomes, outperformed other candidate methods in all country-specific value sets. However, the regression coefficients differed reflecting preference heterogeneity across countries.

     

Mapping of the DLQI scores to EQ-5D utility values using ordinal logistic regression

Purpose

The Dermatology Life Quality Index (DLQI) and the European Quality of Life-5 Dimension (EQ-5D) are separate measures that may be used to gather health-related quality of life (HRQoL) information from patients. The EQ-5D is a generic measure from which health utility estimates can be derived, whereas the DLQI is a specialty-specific measure to assess HRQoL. To reduce the burden of multiple measures being administered and to enable a more disease-specific calculation of health utility estimates, we explored an established mathematical technique known as ordinal logistic regression (OLR) to develop an appropriate model to map DLQI data to EQ-5D-based health utility estimates.

Methods

Retrospective data from 4010 patients were randomly divided five times into two groups for the derivation and testing of the mapping model. Split-half cross-validation was utilized resulting in a total of ten ordinal logistic regression models for each of the five EQ-5D dimensions against age, sex, and all ten items of the DLQI. Using Monte Carlo simulation, predicted health utility estimates were derived and compared against those observed. This method was repeated for both OLR and a previously tested mapping methodology based on linear regression.

Results

The model was shown to be highly predictive and its repeated fitting demonstrated a stable model using OLR as well as linear regression. The mean differences between OLR-predicted health utility estimates and observed health utility estimates ranged from 0.0024 to 0.0239 across the ten modeling exercises, with an average overall difference of 0.0120 (a 1.6% underestimate, not of clinical importance).

Conclusions

This modeling framework developed in this study will enable researchers to calculate EQ-5D health utility estimates from a specialty-specific study population, reducing patient and economic burden.

     

A comparison of the responsiveness of EQ-5D-5L and the QOLIE-31P and mapping of QOLIE-31P to EQ-5D-5L in epilepsy

Objective

To investigate the responsiveness of and correlation between the EQ-5D-5L and the QOLIE-31P in patients with epilepsy, and develop a mapping function to predict EQ-5D-5L values based on the QOLIE-31P for use in economic evaluations.

Methods

The dataset was derived from two clinical trials, the ZMILE study in the Netherlands and the SMILE study in the UK. In both studies, patients’ quality of life using the EQ-5D-5L and QOLIE-31P was measured at baseline and 12 months follow-up. Spearman’s correlations, effect sizes (EF) and standardized response means (SRM) were calculated for both the EQ-5D-5L and QOLIE-31P domains and sub scores. Mapping functions were derived using ordinary least square (OLS) and censored least absolute deviations models.

Results

A total of 509 patients were included in this study. Low to moderately strong significant correlations were found between both instruments. The EQ-5D-5L showed high ceiling effects and small EFs and SRMs, whereas the QOLIE-31P did not show ceiling effects and also showed small to moderate EFs and SRMs. Results of the different mapping functions indicate that the highest adjusted R ² we were able to regress was 0.265 using an OLS model with squared terms, leading to a mean absolute error of 0.103.

Conclusions

Results presented in this study emphasize the shortcomings of the EQ-5D-5L in epilepsy and the importance of the development of condition-specific preference-based instruments which can be used within the QALY framework. In addition, the usefulness of the constructed mapping function in economic evaluations is questionable.

     

Exploring the relationship between EQ-5D,DLQI and PASI,and mapping EQ-5D utilities: a cross-sectional study in psoriasis from Hungary

Background

There is a growing interest in policy making for using utility measures and identifying algorithms to convert disease-specific measures into utilities.

Objectives

To analyse the relationship between EQ-5D, Dermatology Life Quality Index (DLQI) and Psoriasis Area and Severity Index (PASI) in psoriasis. To transform DLQI scores, and key clinical, demographic and health service utilisation variables into utilities.

Methods

A cross-sectional questionnaire survey of 200 consecutive adult patients with moderate to severe psoriasis was carried out in two Hungarian university clinics. The relationship between the outcome measures were analysed with correlations and with the known-groups method. Bivariate and multivariate regression algorithms on EQ-5D scores were formulated.

Results

The mean age of respondents was 51 years (SD = 12.9), 68.5 % were male, and 51.5 % received biological therapy. Median EQ-5D, DLQI, and PASI scores were 0.73, 3.0, and 3.45, respectively. EQ-5D showed a moderate correlation with the DLQI and with the PASI (r _s = ?0.48 and ?0.43, p < 0.05). Strong correlation was found between DLQI and PASI (r _s = 0.81, p < 0.05). DLQI and PASI discriminated better among groups categorised by the localisation of the lesions than EQ-5D. Presence of psoriasis on the neck and/or décolletage was associated with the greatest health related quality of life (HRQOL) impairment. Ten variables were incorporated in a multivariate algorithm that accounted for 48.8 % of EQ-5D variance (ANOVA p < 0.001).

Conclusions

This study provided the first evidence that patients with visible psoriatic lesions have significantly worse HRQOL compared to those with non-visible lesions, measured not only with DLQI but also with EQ-5D. In addition to demographic and clinical variables, our model included health service utilisation variables related to psoriasis, and explained higher proportion of EQ-5D variance than any previous findings in the literature.

     

Modelling a preference-based index for EQ-5D using a non-parametric Bayesian method

Background

Conventionally, models used for health state valuation data have been parametric. Recently, a number of researchers have investigated the use of non-parametric Bayesian methods in this area.

Objectives

In this paper, we present a non-parametric Bayesian model to estimate a preference-based index for a five-dimensional health state classification, namely EQ-5D.

Methods

A sample of 2997 members of the UK general population valued 43 health states selected from a total of 243 health states defined by the EQ-5D using time trade-off technique. Findings from non-parametric modelling are reported in this paper and compared to previously used parametric estimations. The impact of respondent characteristics on health state valuations is also reported.

Results

The non-parametric models were found to be better at predicting scores in populations with different distributions of characteristics than observed in the survey sample. Additionally, non-parametric models were found to be better at allowing for the impact of respondent characteristics to vary by health state. The results show an important age effect with sex having some effect.

Conclusion

The non-parametric Bayesian models provide more realistic and better utility estimates from the EQ-5D than previously used parametric models have done. Furthermore, the model is more flexible in estimating the impact of covariates.

     

Estimating informal care inputs associated with EQ-5D for use in economic evaluation

Objectives

This paper estimates informal care need using the health of the patient. The results can be used to predict changes in informal care associated with changes in the health of the patient measured using EQ-5D.

Methods

Data was used from a prospective survey of inpatients containing 59,512 complete responses across 44,494 individuals. The number of days a friend or relative has needed to provide care or help with normal activities in the last 6 weeks was estimated using the health of the patient measured by EQ-5D, ICD chapter and other health and sociodemographic data. A variety of different regression models were estimated that are appropriate for the distribution of the informal care dependent variable, which has large spikes at 0 (zero informal care) and 42 days (informal care every day).

Results

The preferred model that most accurately predicts the distribution of the data is the zero-inflated negative binomial with variable inflation. The results indicate that improving the health of the patient reduces informal care need. The relationship between ICD chapter and informal care need is not as clear.

Conclusions

The preferred zero-inflated negative binomial with variable inflation model can be used to predict changes in informal care associated with changes in the health of the patient measured using EQ-5D and these results can be applied to existing datasets to inform economic evaluation. Limitations include recall bias and response bias of the informal care data, and restrictions of the dataset to exclude some patient groups.

     

Determinants of time trade-off valuations for EQ-5D-5L health states: data from the Canadian EQ-5D-5L valuation study

Background

Previous studies suggest that population subgroups have different perceptions of health, as well as different preferences for hypothetical health states.

Objective

To identify determinants of health states preferences elicited using time trade-off (TTO) for the 5-level EQ-5D questionnaire (EQ-5D-5L) in Canada.

Methods

Data were from the Canadian EQ-5D-5L Valuation Study, which took place in Edmonton, Hamilton, Montreal, and Vancouver. Each respondent valued 10 of 86 hypothetical health states during an in-person interview using a computer-based TTO exercise. The TTO scores were the dependent variable and explanatory variables including age, sex, marital status, education, employment, annual household income, ethnicity, country of birth, dwelling, study site, health literacy, number of chronic conditions, previous experience with illness, and self-rated health.

Results

Average [standard deviation (SD)] age of respondents (N = 1209) was 48 (17) years, and 45 % were male. In multivariable linear regression models with random effects, adjusted for severity of health states and inconsistencies in valuations, older age [unstandardized regression coefficient (β) = ?0.077], male sex (β = 0.042), being married (β = 0.069), and urban dwelling (β = ?0.055) were significantly associated with health states scores. Additionally, participants from Edmonton (β = ?0.124) and Vancouver (β = ?0.156), but not those from Hamilton, had significantly lower TTO scores than those from Montreal.

Conclusions

Socio-demographic characteristics were the main determinants of preferences for EQ-5D-5L health states in this study. Interestingly, preferences were significantly lower in western Canadian cities compared to eastern ones, bringing into question whether a single preference algorithm is suitable for use in all parts of Canada.

     

One question might be capable of replacing the Shoulder Pain and Disability Index (SPADI) when measuring disability: a prospective cohort study

Questions

Is it possible to replace the Shoulder Pain and Disability Index (SPADI) with a single substitute question for people with shoulder pain, when measuring disability and how well does this substitute question perform as a predictor for recovery.

Design

A prospective cohort study.

Participants

A total of 356 patients with shoulder pain in primary care.

Analyses

Convergent, divergent, and “known” groups validity were assessed by using hypotheses testing. Responsiveness was assessed using the Receiver Operating Curve and hypothesis testing. In addition, we performed multivariate regression to assess if the substitute question showed similar properties as the SPADI and if it affected the model itself, using recovery as an outcome.

Results

The Spearman correlation coefficient between the total SPADI score and the substitute question was high, and moderate with the Shoulder Disability Questionnaire. The correlation between the substitute question and the EQ-5D-3L was low and the responsiveness was acceptable. The substitute question did not significantly contribute to both prognostic prediction models as opposed to the SPADI. Regardless all models showed poor to fair discrimination.

Conclusion

The single question is a reasonable substitute for the SPADI and can be used as a screening instrument for shoulder disability in primary clinical practice. It has slightly poorer predictive power and should therefore not be used for prognosis.

     

An empirical comparison of the measurement properties of the EQ-5D-5L,DEMQOL-U and DEMQOL-Proxy-U for older people in residential care

Purpose

This study aimed to empirically compare the measurement properties of self-reported and proxy-reported (in cases of severe cognitive impairment) generic (EQ-5D-5L) and condition-specific (DEMQOL-U and DEMQOL-Proxy-U) preference-based HRQoL instruments in residential care, where the population is characterised by older people with high rates of cognitive impairment, dementia and disability.

Methods

Participants were recruited from seventeen residential care facilities across four Australian states. One hundred and forty-three participants self-completed the EQ-5D-5L and the DEMQOL-U while three hundred and eight-seven proxy completed (due to the presence of severe dementia) the EQ-5D-5L and DEMQOL-Proxy-U. The convergent validity of the outcome measures and known group validity relative to a series of clinical outcome measures were assessed.

Results

Results satisfy convergent validity among the outcome measures. EQ-5D-5L and DEMQOL-U utilities were found to be significantly correlated with each other (p?<?0.01) as were EQ-5D-5L and DEMQOL-Proxy-U utilities (p?<?0.01). Both self-reported and proxy-reported EQ-5D-5L utilities demonstrated strong known group validity in relation to clinically recognised thresholds of cognition and physical functioning, while in contrast neither DEMQOL-U nor DEMQOL-Proxy-U demonstrated this association.

Conclusions

The findings suggest that the EQ-5D-5L, DEMQOL-U and DEMQOL-Proxy-U capture distinct aspects of HRQoL for this population. The measurement and valuation of HRQoL form an essential component of economic evaluation in residential care. However, high levels of cognitive impairment may preclude self-completion for a majority. Further research is needed to determine cognition thresholds beyond which an individual is unable to reliably self-report their own health-related quality of life.

     

Validation and comparison of the psychometric properties of the EQ-5D-3L and EQ-5D-5L instruments in Greece

Purpose

To validate and compare the psychometric properties of the EQ-5D-3L with the EQ-5D-5L classification systems in Greece.

Methods

Participants (n = 2279) over 40 years old, sampled from the greater area of Athens using a multistage stratified quota sampling method, completed both EQ-5D versions, while information was also collected on socio-demographics and health-related characteristics. The EQ-5D-5L and EQ-5D-3L were evaluated in terms of agreement, ceiling effects, redistribution and inconsistency, informativity, and convergent and known-groups validity.

Results

The agreement between the EQ-5D-3L and EQ-5D-5L was high (ICC = 0.85). Ceiling effects decreased significantly in the EQ-5D-5L in all domains (P < 0.001), with “usual activities” (?21.4 %) and “self-care” (?20.1 %) showing the highest absolute and “anxiety/depression” the highest relative reduction (?32.46 %). Inconsistency was low (5.7 %). The increase in prevalence of problems was larger than the decrease in their severity, resulting in a lower mean health utility for the EQ-5D-5L. Overall absolute and relative informativity improved by 70.5 % and 16.4 %, respectively, in the EQ-5D-5L. Both instruments exhibited good convergent and known-groups validity, with evidence of a considerably better convergent performance and discriminatory ability of the EQ-5D-5L.

Conclusions

Both EQ-5D versions demonstrated good construct validity and had consistent redistribution. The EQ-5D-5L system may be preferable to the EQ-5D-3L, as it exhibited superior performance in terms of lower ceiling effects, higher absolute and relative informativity, and improved convergent and known-groups validity efficiency.

     

An EQ-5D-5L value set based on Uruguayan population preferences

Purpose

To derive a value set from Uruguayan general population using the EQ-5D-5L questionnaire and report population norms.

Methods

General population individuals were randomly assigned to value 10 health states using composite time trade off and 7 pairs of health states through discrete choice experiments. A stratified sampling with quotas by location, gender, age and socio-economic status was used to respect the Uruguayan population structure. Trained interviewers conducted face-to-face interviews. The EuroQol valuation technology was used to administer the protocol as well as to collect the data. OLS and maximum likelihood robust regression models with or without interactions were tested.

Results

We included 794 respondents between 20 and 83 years. Their characteristics were broadly similar to the Uruguayan population. The main effects robust model was chosen to derive social values. Values ranged from ?0.264 to 1. States with a misery index = 6 had a mean predicted value of 0.965. When comparing the Uruguayan population with the Argentinian EQ-5D-5L crosswalk value set, the prediction for states which differed from full health only in having one of the dimensions at level 2 were about 0.05 higher in Uruguay. The mean index value, using the selected Uruguayan EQ-5D-5L value set, for the general population in Uruguay was 0.895. In general, older people had worse values and males had slightly better values than females.

Conclusion

We derived the EQ-5D-5L Uruguayan value set, the first in Latin America. These results will help inform decision-making using economic evaluations for resource allocation decisions.

     

Comparative performance of the EQ-5D-5L and SF-6D index scores in adults with type 2 diabetes

Purpose

To explore the comparative performance including discriminative and longitudinal validity of EQ-5D-5L and SF-6D index scores in adults with type 2 diabetes.

Methods

Data from an on-going cohort study of adults with type 2 diabetes in Alberta, Canada, were used. Known-groups approach was used to examine discriminative validity. Correlation and agreement indices and scatter and Bland–Altman plots were used to examine the relationship between the two measures. Longitudinal validity was explored using Wilcoxon signed-rank test, effect size, and standardized response mean.

Results

In 1832 participants at baseline (age 64.3, standard deviation 10.6 years; 45% female), median EQ-5D-5L score was 0.85 [interquartile range (IQR) 0.17], and floor and ceiling effects of 0.1 and 16.1%, respectively; median SF-6D score was 0.72 (IQR 0.24), and floor and ceiling effects of 0.1 and 3.2%, respectively. EQ-5D-5L and SF-6D index scores were significantly correlated with an overall Spearman correlation coefficient of 0.73, and an ICC of 0.62 (95% CI 0.42–0.74). Both EQ-5D-5L and SF-6D scores demonstrated statistically significant differences in self-reported chronic conditions, depressive symptoms, and diabetes-related distress, and were able to detect changes in depressive symptoms and diabetes distress across all change groups.

Conclusions

Both EQ-5D-5L and SF-6D index scores provide valid measurement in this patient population. Considerable overlap between the measures means it is not necessary to include both in surveys, however, the advantages and disadvantages of each should be considered.

     

Measurement of utility in asthma: evidence indicating that generic instruments may miss clinically important changes

Background

Accurate assessment of preference-based health-related quality of life is important in determining the value of asthma interventions.

Objective

To examine the sensitivity and responsiveness of the EQ-5D and the AQL-5D to differences in asthma control measured by the Asthma Control Questionnaire (ACQ-5).

Methods

The Observational Study of Asthma Control and Outcomes was a prospective survey of persistent asthma patients ≥12 years old in Kaiser Colorado. Patients received a survey three times in 1 year, including the ACQ-5, AQL-5D and EQ-5D-3L (including VAS). Censored Least Absolute Deviations (CLAD) and logistic regression were used, controlling for sociodemographics and smoking.

Results

There were 6666 completed surveys (1799 individuals completed all three survey waves). After controlling for covariates, each one-point increase in ACQ-5 was associated with a decrease of 0.066, 0.058, 0.074 and 6.12 in EQ-5D(US), EQ-5D(UK), AQL-5D and VAS scores. Uncontrolled asthma (ACQ-5 > 1.5) was associated with a decrease of 0.15, 0.17, 0.11 and 10, respectively (vs. ACQ ≤ 1.5). AQL-5D scores were statistically significantly different across categories of ACQ-5 scores of 0.5 (the minimum clinically important difference [MCID]), while EQ-5D scores were not significant across most categories. The AQL-5D appeared more robust to changes in control over time (responsiveness) compared to EQ-5D-3L.

Conclusion

The AQL-5D appears more responsive to changes in asthma control over time and more sensitive to detecting differences corresponding to the ACQ-5 MCID than the EQ-5D-3L. Using the EQ-5D-3L without an asthma-specific measure such as the AQL-5D may miss clinically important changes in asthma control.

     

Psychometric properties of the EQ-5D-5L in patients with hip or knee osteoarthritis: reliability,validity and responsiveness

Purpose

To study the psychometric properties, including reliability, validity and responsiveness, of the Spanish EQ-5D-5L questionnaire for patients with hip or knee osteoarthritis (OA).

Methods

We included 758 patients with hip or knee OA who completed the EQ-5D-5L and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) at baseline, of whom 644 also did 6 months later. The EQ-5D-5L contains five questions from which a utility index is derived. The WOMAC covers three dimensions: pain, stiffness and physical function. Floor and ceiling effects were calculated. Reliability was assessed using Cronbach’s alpha. Convergent validity was tested using the Spearman correlation coefficient between EQ-5D-5L and WOMAC. We examined known-groups validity by comparing the EQ-5D-5L between subgroups defined by WOMAC scores using ANOVA or the Kruskal–Wallis test. Effect sizes were calculated to assess responsiveness, and minimal clinically important difference (MCID) was estimated.

Results

The EQ-5D-5L showed minimal floor and ceiling effects (<?3%). Cronbach’s alpha was 0.86. The EQ-5D-5L index was strongly correlated with WOMAC pain and function scores (??0.688 and ??0.782). Patients with higher WOMAC scores had significantly (p?<?0.0001) lower EQ-5D-5L index. The 20.19% had hip or knee replacement during the follow-up. Effect sizes were small among non-surgical patients, but >?0.80 among “improved” surgical patients, being the MCID for improvement 0.32 points.

Conclusions

The results support the reliability, validity and responsiveness of the EQ-5D-5L, overcoming the limitations of the EQ-5D-3L in these patients. Therefore, the EQ-5D-5L could be very useful as an outcome measure, at least in patients with hip or knee OA.

     

Health-related quality of life among pulmonary tuberculosis patients in Pakistan

Purpose

Health-related quality of life (HRQoL) of pulmonary TB patients has not been assessed in Pakistan. We assessed self-reported HRQoL of pulmonary TB patients in Karachi, Pakistan utilizing the EQ-5D and EQ-VAS prior to, during, and after completion of TB treatment.

Methods

We enrolled 226 pulmonary TB patients in a longitudinal cohort study. Health-utility scores were estimated by the EQ-5D five dimensions and the EQ-Visual Analogue Scale (VAS) at baseline (month 0) and each monthly follow-up visit until treatment completion at month 6. Repeated-measures ANOVA was used to investigate effect of time into treatment on EQ-5D and EQ-VAS scores.

Results

EQ-5D health utility and EQ-VAS scores increase with treatment progression. For the enrolled TB patients, the mean EQ-5D utility scores more than doubled from 0.43 to 0.88, p?<?.001, effect size η²?=?0.40 from treatment initiation to treatment completion.

Conclusion

Perceived HRQoL of TB patients improves with treatment progression. This can inform targeted treatment plans as well as TB policy and funding for high-burden countries.