Elsevier

Drug and Alcohol Dependence

Volume 82, Issue 2, 28 April 2006, Pages 111-118
Drug and Alcohol Dependence

Validity of the EQ-5D as a generic health outcome instrument in a heroin-dependent population

https://doi.org/10.1016/j.drugalcdep.2005.08.012Get rights and content

Abstract

Objective

To evaluate the validity of the EuroQol (EQ-5D) in a population of chronic, treatment-resistant heroin-dependent patients.

Methods

The EQ-5D is studied relative to the Maudsley Addiction Profile (MAP), the Symptom Checklist (SCL-90) and the European Addiction Severity Index (EuropASI) which were used to assess the participant's physical functioning, mental health and social integration, respectively. Data were gathered from 430 patients participating in the Dutch heroin trials with an intended 12-month treatment period. The EQ-5D was used as a separate health outcome measure. Statistical analyses were conducted using Spearman's and Pearson's correlations.

Results

The EQ-5D dimensions mobility, self-care and usual activities generally showed low correlations with relevant parameters of the MAP-HSS, SCL-90 and EuropASI (r = 0.132–0.369). The EQ-5D dimension pain/discomfort showed low to moderate hypothesized correlations with all disease-specific measures (r = 0.153–0.496). The EQ-5D dimension anxiety/depression showed moderate to high correlations with the SCL-90 (including the sum score) and some of the EuropASI parameters (r = 0.133–0.615). The EQ-5D utility scores were moderately correlated with the MAP-HSS (r = −0.468) and the SCL-90 (r = −0.491) total score and with response to treatment at month 12.

Conclusion

The majority of hypothesized associations between the EQ-5D and the disease or domain-specific measures could be confirmed. The validity of the EQ-5D-based utility score appears to be suitable in the evaluation of chronic, heroin-dependent populations.

Introduction

Nowadays, many proven effective interventions are available for the treatment of heroin-dependent patients (van den Brink and van Ree, 2003). The interventions differ in their target populations, effectiveness and in the costs per treatment. It is common to evaluate new treatments not only in terms of effectiveness or direct health benefit for the patient (clinical outcome), but also in terms of efficiency integrating the costs and changes in health status and comparing it with the best available alternative(s) (Gold et al., 1996). To compare the cost-effectiveness of different interventions and to compare the results with the cost-effectiveness of interventions in other domains of medicine, general and standardized comparable indicators of effect are needed, such as measures assessing health status. One of the most frequently applied general health status measures is the EuroQol questionnaire (EQ-5D, The EuroQol Group, 1990). The EQ-5D is a brief, simple and easy-to-use self-completion questionnaire. It is often used in economic evaluations of health care as a complement to disease-specific outcome measures. The EQ-5D is available in many languages and its use and qualities are described for a growing number of different populations and settings. To our knowledge the EQ-5D, which is an instrument of choice in economic evaluations, has not been used in (randomized) studies in drug-dependent populations, hence, little is known about its validity in these populations (Dijkgraaf et al., 2005).

This paper attempts to establish the validity of the EQ-5D in a population of chronic, treatment-resistant heroin-dependent patients participating in the Dutch heroin trials. The EQ-5D is studied relative to three disease-specific instruments commonly used in addiction research: the Maudsley Addiction Profile (MAP, Marsden et al., 1998), the Symptom Checklist (SCL-90, Arrindell and Ettema, 1986) and the European Addiction Severity Index (EuropASI, Kokkevi and Hartgers, 1995).

Section snippets

Trials, treatments and patients

In two recent Dutch trials, medically co-prescribed heroin in addition to methadone maintenance treatment was compared to methadone maintenance treatment only (n = 549). These trials were conducted among inhaling (n = 375) and injecting (n = 174) heroin-dependent patients and were carried out simultaneously (van den Brink et al., 2003). Participants were chronic heroin-dependent patients who had been treated unsuccessfully in methadone maintenance programs. The trials were multi-centred with

Results

Table 2 shows socio-demographic data and baseline substance abuse characteristics of the study population. Most patients were male, of Dutch/Western-European ethnicity with a mean age of 39 years and of low education. More than two-thirds of all patients lived independently (alone or with partner). About 60% of all patients used heroin predominantly by inhalation. Mean number of years of heroin use on a regular basis was over 16 years.

Summary of the results

The EQ-5D dimensions mobility and self-care generally showed low correlations with the disease-specific items or dimensions from the MAP-HSS, SCL-90 and EuropASI, whereas usual activities showed low to moderate correlations. The pain/discomfort dimension showed low to moderate correlations with all disease-specific measures and a moderate correlation with the MAP-HSS sum score. The anxiety/depression dimension showed moderate to high correlations with the SCL-90 (including the sum score) and

Acknowledgement

This research was funded by the Central Committee on the Treatment of Heroin Addicts (CCBH), Utrecht, The Netherlands.

References (28)

  • P. Dolan

    Modeling valuations for EuroQol health states

    Med. Care

    (1997)
  • P.J. Dorman et al.

    How do scores on the EuroQol relate to scores on the SF-36 after stroke?

    Stroke

    (1999)
  • EuroQol Group, 1990. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy 16,...
  • EuroQol Group, 2005. EQ-5D: an instrument to describe and value health. www.euroqol.org (last accessed 17th August...
  • Cited by (56)

    • Hook effect detection and detection-range-controllable one-step immunosensor for inflammation monitoring

      2020, Sensors and Actuators, B: Chemical
      Citation Excerpt :

      for Passing–Bablok analysis, and the R Score, slope, and offset values were calculated. The linearity test was carried out in accordance with the CLSI guideline EP06-A [23,24]. The test samples for linearity and the hook effect were spiked to 5, 20, 34, 48, 63, 77, 90, 105, 120, 150, 200, 300, 500, 750, and 1000 mg/L of the cCRP antigen in canine serum.

    • Opioid agonist treatment reduces losses in quality of life and quality-adjusted life expectancy in heroin users: Evidence from real world data

      2019, Drug and Alcohol Dependence
      Citation Excerpt :

      Given the higher prevalence rates of chronic medical comorbidities (e.g., HIV or HCV infection) of the OAT group, the non-OAT group showed significantly lower utility values (0.63) along the 8 years of follow-up, and the OAT effect on QOL was even larger after adjustment for confounders. Our findings of QOL corroborate the results of other studies (Larson et al., 2007; Rosenblum et al., 2003; van der Zanden et al., 2006; Volkow and McLellan, 2016). Moreover, we found the difference between the loss-of-LE and loss-of-QALE from improving QOL was 1.9 [ = 9.7-(26.6-18.8)] QALY, which could be directly compared with other clinical conditions, in contrast to estimations using disability-adjusted life year (DALY) (Degenhardt et al., 2013; Peacock et al., 2018).

    • Integrated cognitive behavioral therapy for ADHD in adult substance use disorder patients: Results of a randomized clinical trial

      2019, Drug and Alcohol Dependence
      Citation Excerpt :

      Although we originally proposed to use number of days of excessive use in the past two months (Van Emmerik-van Oortmerssen et al, 2013), we later (but before closure of the dataset) decided that a shorter time interval of one week would better reflect substance use at end of treatment as substance use in the past two months reflects substance use during treatment rather than at the end of treatment. Depressive symptoms were assessed with the Beck Depression Inventory (BDI) (Beck and Steer, 1987), anxiety symptoms were assessed with the Beck Anxiety Inventory (BAI) (Beck et al., 1988), and quality of life was assessed with the 3-level EQ-5D (Van der Zanden et al., 2006). These are all self-report instruments.

    View all citing articles on Scopus
    View full text