Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Addiction. Author manuscript; available in PMC 2017 April 1.
Published in final edited form as:
Published online 2015 December 17. doi:  10.1111/add.13219
PMCID: PMC5034732

Measuring benefits of opioid misuse treatment for economic evaluation: health related quality of life of opioid dependent individuals and their spouses as assessed by a sample of the US population



To understand how the general public views the quality of life effects of opioid misuse and opioid use disorder on an individual and his/her spouse, measured in terms used in economic evaluations.


Cross-sectional internet survey of a US-population-representative respondent panel conducted December 2013-January 2014.




2,054 randomly-selected adults; 51% male (before weighting).


Mean (95% CI) and median health “utility” for 6 opioid misuse and treatment outcomes: active injection misuse; active prescription misuse; methadone maintenance therapy at initiation, and when stabilized in treatment; and buprenorphine therapy at initiation, and when stabilized. Utility is a numerical representation of health-related quality of life used in economic evaluations to “adjust” estimated survival to include peoples' preferences for health states. Utilities are determined by surveying the general population to estimate the value they assign to particular health states—on a scale where 0=the value of being dead, and 1.0=the value of being in perfect health. Spouse spillover utility is assigned to a spouse of an individual who is in a particular health state.


Mean individual utility ranged from 0.574 (95%CI: 0.538, 0.611) for active injection opioid misuse to 0.766 for stabilized buprenorphine therapy (95%CI: 0.738, 0.795), with other states in between. Female respondents assigned higher utility to the active prescription misuse and buprenorphine therapy at initiation states than did males (p<0.05); all other states did not differ by respondent gender. Mean spousal utilities were significantly lower than 1.0 but mostly higher than individual utility, and were similar between male and female respondents.


In the opinion of the US public, injection opioid misuse results in worse health-related quality of life than prescription misuse, and methadone therapy results in worse health-related quality of life than buprenorphine therapy. Spouses are negatively affected by their partner's opioid misuse and early treatment.


Opioid misuse is a pervasive public health problem worldwide: there were an estimated 28.6-38.0 million heroin or prescription opioid past-year users globally in 20121, and approximately 69,000 deaths from opioid-related overdose.2 In 2013 in the U.S. alone, there were an estimated 4.5 million past-month non-medical users of prescription opioids, and 289,000 heroin users3, with 16,235 and 8,257 deaths, respectively.4

Economic evaluation can determine costs and benefits to guide prioritization—i.e., what are the best values for our dollars to reduce disease burden.5 Economic evaluation measures benefits through Quality Adjusted Life Years (“QALYs”), an outcome measure that encapsulates longevity and quality of life.6 QALYs are the outcome of choice for health economic evaluations conducted in the US, the UK, Canada, and Australia, among other countries.7-10 The inclusion of substance use disorders in benefits mandated by the Affordable Care Act brings substance misuse treatment on par with other medical conditions—and requires that outcomes be evaluated similarly.11

To date, QALYs have focused primarily on the patient, and excluded caregivers and family members.12 There is a growing call to include family effects in economic evaluation, to more accurately capture the entirety of benefits.13,14 Both individuals' and others' outcomes have historically been considered in substance misuse.15 In medicine, however, concern has generally been focused on the patient alone (despite evidence of the collateral effects of health16,17). In this study we assessed the health-related quality of life of opioid misuse and treatment on individuals to enable the calculation of QALYs for opioid misuse, and on spouses to inform knowledge of effects beyond the individual.

QALYs represent the value people place on life in a certain health state—whether that is life in methadone maintenance therapy or life with diabetes. For resource allocation decisions, we need to know how society as a whole values life in a health state, meaning what the general population feels about a particular state of health, and this is usually accomplished through a general population survey.6 Health economists use these values to assign a weight to each year of life that is gained by a health intervention, to compare among gains afforded by different interventions for different conditions. Health state utility is the metric that represents this weight, and was the focus of this research. Our aims were as follows:

  1. To estimate mean population utilities for opioid misuse and opioid use disorder treatment;
  2. To estimate mean population utilities for spouses of opioid dependent/in treatment individuals;
  3. To explore the scope of effects that should be included in these measurements.



We measured utilities for 6 of opioid misuse and treatment states, selected to represent a spectrum of states that would be useful for future cost-effectiveness analyses: active injection opioid misuse (“active injection misuse”); active prescription opioid misuse (“active prescription misuse”); methadone maintenance therapy at the initiation stage (“initiation methadone”), and when the user is considered stabilized in treatment (“stabilized methadone”); and buprenorphine therapy at the initiation stage (“initiation buprenorphine”), and when the user is considered stabilized in treatment (“stabilized buprenorphine”). Utilities are defined on a scale of 0-1.0, where zero=the value of being dead and 1.0=the value of being in perfect health, and all health states are valued relative to these endpoints. Utility is thus a universal metric that allows comparisons across diseases.

Utilities can be measured in a variety of ways: the “direct” method is to describe a health condition to someone in the general population – with or without personal experience with it--and have him/her assign it a number between 0 and 1.0. The “indirect” method is to have a person who has a condition report on his/her symptoms, health status, and functioning using a structured questionnaire (a “generic” utility instrument), and then to assign a utility value to this report using a pre-existing set of values collected from the general population.18 Broadly speaking, direct methods are considered theoretically superior but more administratively burdensome to implement than indirect methods; indirect methods are more convenient and can easily be integrated into clinical trials, but have practical limitations in their sensitivity to effects.19,20 Both methods produce theoretically equivalent results, although in practice differences across methods have been observed.21

We used a direct method in this study, the “standard gamble” (SG).18 The SG asks a respondent to consider the risk of death s/he would be willing to take in order to avoid living in a particular health state, and assigns value to health on the 0-1.0 scale described above.18 For example, if someone would take 10% risk of death to avoid living with diabetes, the SG for diabetes would be 0.90, or 90% of the value of perfect health. The SG is considered the theoretically optimal method for estimating utilities for societal-perspective economic evaluation.6

Study design

We conducted a cross-sectional, on-line survey of a US-population-representative respondent panel. We administered two survey versions (Figure 1): an “individual-focused” version, in which respondents assessed the utility of living in each of 6 opioid misuse or treatment health states, and a “spouse-spillover” version, in which respondents assessed the utility of the spouse of an individual in these states. We developed a hypothetical descriptive vignette for each state following established practice22, incorporating the literature, input from clinical experts, and qualitative data from people experienced with the condition (i.e., interviews with persons currently in methadone maintenance or buprenorphine therapy; n= 10). Both versions employed conventional practice for the SG technique18: for the individual-focused version, each respondent was asked to imagine him/herself living as an individual described in a vignette; for the spouse-spillover version, each respondent was asked to imagine that s/he was the spouse of a person living as described in a vignette, but was otherwise healthy.23

Figure 1
Survey schema: two survey versions, some data combined for analysis.

We developed two sets of vignettes to explore the scope of effects that should optimally be included in descriptions. Health state vignettes traditionally include descriptions of physical and emotional health, based on the premise that health-related quality of life is a function of health-related domains.6 We developed one set of vignettes in this traditional manner (“physical/emotional health descriptors”), and another set with additional descriptors commonly included in substance use disorder outcomes (employment, family relationships, and criminal justice involvement; “expanded descriptors”).24 We intentionally excluded co-morbid conditions from our vignettes because our goal was to assess the independent utility of opioid misuse.25 A “remission” health state was included in the expanded descriptors set to assess the utility of prior misuse. We used a randomized, split sample design to compare utilities for each set of vignettes: each respondent was assigned vignettes from one set only to avoid contamination across assessments. A sample vignette is presented in Figure 2; all vignettes are in the appendix.

Figure 2
Sample vignettes for initiation stage of methadone therapy, using physical/emotional health descriptors and expanded descriptors.

The spouse-spillover version assessed the active misuse and initiation treatment vignettes, described with the physical/emotional descriptors (to maintain comparability with utilities as they are commonly measured6,23). Spousal utility was estimated as a paired, “couple” utility, wherein the respondent provided a single utility value for the couple as a unit, and as a spouse-only utility, wherein the respondent provided a utility solely for the effect on an otherwise healthy individual of having a spouse living as described in the vignette. This spouse-only viewpoint represents the actual spillover utility of the opioid misuse or treatment state on a spouse; we included the couple viewpoint as a thought exercise and confirmation measure because of the potential conceptual difficulty of valuing the spouse-only perspective.23 Respondents in the spouse-spillover survey version also assessed the utility of each vignette from the perspective of an individual living as in the vignette--“self” viewpoint, as in the individual-focused version--to provide both a thought exercise before the spouse-spillover viewpoint questions and to supplement the data collected in individual-focused version.

Survey procedure

Respondents completed practice 0-100 point rating scale (RS) questions and SG questions assessing their own current health before the hypothetical vignettes. We randomly assigned vignettes to respondents to achieve a balanced sample across vignettes and a balanced respondent burden. Each respondent received 3-6 vignettes in the individual-focused version and 1-2 in the spouse-spillover version (in which each vignette was described using three viewpoints --self, couple, and spouse-only). Each vignette was assessed by RS prior to the SG evaluations, as a warm-up exercise.18

The standard gamble questions used the “magic pill” convention to present the evaluation scenario23 (see Appendix for sample questions), and visual aids developed by Prosser et al.23 Error screens were used to alert respondents to illogical answers and offer reconsideration.26 Demographic characteristics were collected at the end of the survey or provided from previously-collected panel data. The survey was pre-tested on a convenience sample of respondents in paper form (n=8), and with members of the on-line panel in programmed form (n=54).


We used the GfK “KnowledgePanel”27 for our sample. Panel members are recruited using address-based sampling via postal mail. Participants are provided with internet access and/or computer equipment or vouchers redeemable for cash in exchange for survey participation. GfK maintains sampling weights based on the US census demographic data to correct for sampling biases and to approximate population responses; the final sample is effectively considered random.

The sample size for the survey was designed to detect meaningful differences in mean values between vignettes based on existing estimates of values for similar health states using similar measures.28 Minimally important differences in utilities across measurement techniques and conditions range from 0.03-0.07.29-33 We sought a sample of approximately 425-475 responses per vignette to detect these differences based on conservative assumptions about variation in our observed means.


We excluded responses that failed invariance criteria--defined as those in which all SG responses from a respondent including the practice question were the same and equal to 0 (the minimum), 0.5 (the starting point for the exercise) or 1.0 (the maximum).26 We calculated means, confidence intervals, medians, percentiles, minimum values, and maximum values for all vignettes; sample responses were weighted to reflect the US population using weights provided by GfK. We compared mean utilities across vignettes with simple linear regression models of utilities on health states to identify significant differences overall and in pairwise comparisons (at a significance level of p<0.05). Analyses were conducted in Stata version 12 (Stata Corporation, College Park, TX). The study was approved by the Harvard T.H. Chan School of Public Health and Weill Cornell Medical College institutional review boards.


Data were collected from 2,054 respondents from December, 2013-January, 2014: 1,178 in the individual-focused version and 876 in the spouse-spillover version (57% overall participation rate; Figure 1). The total number of completed SG responses varied from 370-555 across vignettes due to the randomization pattern and missing data (Table 1). After weighting, the sample reflected the US population (Table 2). Invariant responses totaled 1,383 utilities (16.8% of total sample) and were excluded from the analysis (Figure 1)26; there were no significant demographic differences between respondents who provided invariant and non-invariant responses (results not shown).

Table 1
Number of responses by survey version, type of descriptors used in vignette, health state, and frame of reference (i.e., viewpoint), before exclusion of invariant responses*
Table 2
Sample characteristics: full sample and analytic sample, number of respondents and percentages unweighted and weighted to the US population

Utilities for all vignettes are presented in Table 3. Mean utilities were significantly lower than 1.0 (i.e., perfect health) for all of the vignettes for an individual and for a spouse, for both sets of vignette descriptors. For the vignettes described with physical/emotional health descriptors, mean individual and spousal utility was statistically significantly lowest (i.e., worst) for active injection misuse compared with other states (p<0.001, except for spousal initiation buprenorphine: p=0.053; Table 3). Mean individual utility for methadone therapy was significantly lower than buprenorphine therapy at both initiation and stabilized stages (p<0.01 for both), but did not significantly differ between states for the same treatment: p=0.13 for initiation vs. stabilized methadone, p=0.76 for initiation vs. stabilized buprenorphine. There was no statistically significant difference in mean spousal utility among the active prescription misuse, initiation methadone, and buprenorphine stages (active prescription vs. initiation methadone: p=0.62, vs. initiation buprenorphine: p=0.81; initiation methadone vs. buprenorphine: p=0.51). Mean utilities were significantly lower for vignettes described using the expanded descriptors for both active misuse and both initiation treatment stages, but differences were non-significant for the stabilized treatment stages.

Table 3
Standard gamble utilities for stages of opioid misuse and treatment for individuals and spouses, by type of descriptors used in hypothetical vignettes: mean, 95% CI, and 25th percentile, median, and 75th percentile*; after exclusion of invariant responses. ...


The US population values misusing opioids and being in opioid use disorder treatment substantially worse than being in perfect health, for both the individual user and for his/her spouse. These values quantify the health-related quality of life of opioid misuse and treatment, and allow them to be compared on equal footing to other health conditions and treatments. They can be used to calculate QALYs, the accepted metric of benefit in economic analysis in many countries, and as such can be incorporated into cost-effectiveness analyses-- of different opioid use disorder treatments, and comparison with other health interventions. Emerging techniques allow effects on spouses to also be included in cost-effectiveness analyses, as an additional loss in QALYs.14 The population also places value on effects of opioid misuse that extend beyond what is usually included in economic assessments of benefits, suggesting that the benefits of opioid use disorder treatment may be underestimated with traditional economic methods.

Our results indicate a general population preference for prescription over injection opioid misuse, and for treatment with buprenorphine over methadone. These results are not surprising because of the stigma attached to injection drug use and methadone34: prescription opioids are more socially acceptable than heroin35, and office-based buprenorphine therapy in the U.S is closer to other chronic disease management practices than methadone maintenance treatment, especially in the early stages of treatment.36 The implications of these preferences creates a paradox in that, all else equal, conditions with lower burden offer commensurately less benefit from cure—so treatment of one prescription opioid misuser for one year provides less social benefit than treatment of one injection opioid misuser for one year, just as treatment for one case of the flu, for example, would provide less social benefit than treatment for one case of meningitis. What we choose to spend resources on is a function of both the social benefit derived and many other factors, such as risk, cost, and longevity, but societal preferences indicate one piece of this equation. Our results also suggest population preferences for treatments: initial treatment on buprenorphine therapy is preferred to initial treatment on methadone, so each year lived in the early stages of buprenorphine therapy provides more social benefit than a similar year on methadone, separate of differences in cost or effectiveness or both. These differences disappear in the later stages of treatment when benefits are equal. Finally, we found that the early stage of methadone therapy was valued worse than active prescription opioid misuse (although any treatment is better than injection misuse). This finding is somewhat disturbing but not implausible—early methadone therapy is marked by users' general displeasure with programs and high levels of drop-outs.37 Until a user transitions to a more stable state in treatment the negatives may outweigh the positives. It is important, therefore, for evaluations of opioid use disorder treatments to consider and capture differences in outcomes across the stages of treatment, and the impact of keeping users in treatment until the stable-stage benefits are realized. When additional aspects of life are considered in preferences, stabilized treatment is preferred even more—suggesting that benefits of treatment are greater than economic valuation using current standard techniques would suggest. There is a need for additional research into the role of health-related quality of life and societal preferences in retaining individuals in treatment.

Our results also speak to the effects of opioid misuse on spouses. A compelling case has been made in the literature for the inclusion of family spillover effects in economic evaluation6,12 and our findings offer support for these considerations in the opioid use context. Our results showed that utility losses for spouses of opioid users are roughly similar to those for other chronic conditions.23 Interestingly, we found that active injection misuse was somewhat worse on spouses than other stages, but the other stages were valued equally-- there may be a threshold of effect for a spouse once treatment is initiated. Our spousal utility data are very limited however and point to the need for additional research in this area.

Our utilities for opioid misuse are similar to the few that have been previously reported in the literature but offer some methodological advantages. Cost-effectiveness analysis of opioid treatments require utilities elicited from a representative sample -- of the general population when direct methods are used, and of the opioid misusing population when an indirect method is used. Existing utility estimates for opioid treatment have relied on indirect methods, using clinical trial samples that run the risk of selection bias. Reported mean utilities in the literature include 0.67 for opioid dependent individuals in HIV treatment38, 0.67-0.70 for injection opioid users in treatment39, and 0.58-0.68 for lifetime and current substance misuse of any kind.40,41 These utilities represent very specific health states and select samples of individuals—those participating in clinical trials with specific eligibility criteria. Our values are based on vignettes that describe a more universal experience of opioid misuse and treatment. Nevertheless, the similarity in values between these studies and ours confirms the general magnitude of utility for opioid misuse and treatment. Our values are useful in circumstances when generality is sought, while utilities from clinical trial populations offer specificity in population and experience that may be useful in other circumstances. All meet criteria of population values and all can be used in cost-effectiveness analysis.

As with all utility surveys, there are limitations to our results. Our primary caveat concerns our vignettes. Although carefully constructed using primary data and expert input, our vignettes are by definition limited to a subset of all possible experiences with opioid misuse and treatment. We attempted to include a set of health states sufficiently generic and encompassing to be useful for economic evaluation, but acknowledge the inherent simplification required. Moreover, our expanded descriptors vignettes contain only some of the many potential dimensions of life affected by opioid misuse, and were intended as a test of methodological design, not for use in cost effectiveness analysis—for that we recommend the physical/emotional descriptor utilities, to ensure comparability to cost-effectiveness analyses conducted for other health conditions.6 Similarly, we did not include co-occurring conditions or poly-substance use in our vignettes, despite their prevalence with opioid misuse.42-45 When conducting a cost-effectiveness analysis of treating co-occurring conditions, each individual condition has an associated utility, and any two in tandem have a “joint” utility that is a distinct value. Joint utilities can be assessed as a distinct health state, or as a combination in their components, using recently-developed methods.25 Our utility estimates apply only to opioid misuse states on their own, and hence are a first set of estimates for use in cost-effectiveness analyses. Finally, the SG is a cognitively complex task and while it was preferred for our purposes, we did find some evidence of “protest” (i.e., invariant) responses in our data--respondents who refuse to engage in the valuation task and provide the same answer across questions.26 Our rate of invariant responses was commensurate with those observed in other utility surveys so do not raise particular concerns for this study.26 Measurement of family effects is particularly challenging and methods are still under development12,46; we used current and tested methods of elicitation23 yet acknowledge the potential for measurement error in our data.

In conclusion, these results represent the first directly-elicited, population-based utilities for opioid misuse, and can be used in economic evaluations of opioid use disorder treatments as well as to compare the value of these interventions to other health interventions. They assist in bringing opioid treatment into the mainstream of health economic resource allocation decision-making methods by providing a basis for the calculation of QALYs, the outcome measure most-frequently used in cost-effectiveness analysis. Numerous areas for further research are indicated by our results. First, opioid misuse rarely exists in isolation from other health conditions, and the dual nature of substance misuse with other conditions raises measurement questions for utilities and QALYs.47 Ongoing work is exploring the effect of opioid misuse in the presence of co-occurring conditions.48 Second, the inclusion of spillover effects in economic evaluation is moving toward greater acceptance and data are needed to facilitate this development. Our estimates of spousal effects are the first evidence of QALYs associated with opioid treatment beyond the individual, and go far to improving our understanding of the full value of resources allocated to this condition.15 Finally, the trajectory of individuals from misuse to remission is well-known to be non-linear and often circular. More knowledge is needed on the values assigned to different pathways. Larger population studies that can evaluate more health states, as well as longitudinal studies of individuals progressing along these pathways will be useful in this regard. Further work on the value of opioid use disorder treatment will improve our ability to critically assess the best investments of our health resources, to improve population well-being across health all conditions.

Supplementary Material

Supp AppendixS1

Appendix: Health state vignettes

Sample screen shots of standard gamble questions from survey


We gratefully acknowledge the research assistance provided by Adrianna Saada, MPH, Jared Leff, MS, and Alison Dunning, MS. We thank Ann B. Beeder, MD, David A. Fiellin, MD, and Louise F. Haynes, MSW, for providing input on the health state descriptions. Lisa A. Prosser, PhD, is jointly responsible for the development of the methodology used to study spillover health states and we acknowledge her contribution to this research. We are also grateful to the survey respondents who provided data for our study. And finally, we are appreciative of the anonymous reviewers and the journal editors who provided feedback on earlier versions of this work. Financial support for this study was provided by a grant from the National Institute on Drug Abuse, R01 DA033424. The funding agreement ensured the authors' independence in designing the study, interpreting the data, writing, and publishing the report. The results are solely those of the authors and do not represent the official views of the National Institutes of Health or the National Institute on Drug Abuse. Preliminary results were presented at the 10th World Congress of the International Health Economics Association, Dublin, Ireland, July, 2014.


Disclosures: The authors have no competing interests to declare.


1. United Nations Office on Drugs and Crime. World Drug Report 2014. New York: United Nations; 2015.
2. [Accessed 3/30/15];Information sheet on opioid overdose, November 2014. 2014 2015, at
3. Substance Abuse and Mental Health Services Administration Center for Behavioral Health Statistics and Quality. The NSDUH Report: Substance Use and Mental Health Estimates from the 2013 National Survey on Drug Use and Health: Overview of Findings. Rockville, MD: Substance Abuse and Mental Health Services Administration; Sep 4, 2014.
4. National Overdose Deaths from Select Prescription and Illicit Drugs, National Center on Health Statistics, CDC WONDER. [Accessed 4-1-15];2015 2015, at
5. Gold M, Siegel J, Russell L, Weinstein M. Cost-Effectiveness in Health and Medicine. New York: Oxford University Press; 1996.
6. Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Cost Effectiveness in Health and Medicine. New York, N Y: Oxford University Press; 1996.
7. National Institute for Health and Care Excellence. Guide to the methods of technology appraisal. 2013;2103
8. Australian Government Department of Health. Guidelines for preparing submissions to the Pharmaceutical Benefits Advisory Committee. Canberra, Australia: Commonwealth of Australia; 2015.
9. Canadian Agency for Drugs and Technologies in Health. Guidelines for the economic evaluation of health technologies. Canada. Ottowa: 2006.
10. Miller W, Robinson L, Lawrence R, editors. Valuing Health for Regulatory Cost-Effectiveness Analysis. Washington, DC: The National Academies Press; 2006.
11. Barry CL, Huskamp HA. Moving beyond parity--mental health and addiction care under the ACA. N Engl J Med. 2011;365:973–5. [PMC free article] [PubMed]
12. Tilford JM, Payakachat N. Progress in measuring family spillover effects for economic evaluations. Expert Rev Pharmacoecon Outcomes Res. 2015;15:195–8. [PMC free article] [PubMed]
13. Basu A, Meltzer D. Implications of spillover effects within the family for medical cost-effectiveness analysis. J Health Econ. 2005;24:751–73. [PubMed]
14. Al-Janabi H, Van Exel J, Brouwer W, Coast J. A framework for incorporating family health spillovers in economic evaluation. Med Decis Making. epub ahead of print. [PMC free article] [PubMed]
15. Sindelar JL, Jofre-Bonet M, French MT, McLellan AT. Cost-effectiveness analysis of addiction treatment: paradoxes of multiple outcomes. Drug Alcohol Depend. 2004;73:41–50. [PubMed]
16. Christakis NA. Social networks and collateral health effects. BMJ. 2004;329:184–5. [PMC free article] [PubMed]
17. Christakis NA, Fowler JH. The spread of obesity in a large social network over 32 years. NEJM. 2007;357:370–9. [PubMed]
18. Hunink M, Weinstein MC, Wittenberg E, et al. Decision Making in Health and Medicine Integrating Evidence and Values. 2nd. Cambridge, UK: Cambridge University Press; 2014.
19. Prosser LA, Grosse SD, Wittenberg E. Health utility elicitation: is there still a role for direct methods? Pharmacoeconomics. 2012;30:83–6. [PubMed]
20. Richardson J, Khan MA, Iezzi A, Maxwell A. Comparing and Explaining Differences in the Magnitude, Content, and Sensitivity of Utilities Predicted by the EQ-5D, SF-6D, HUI 3, 15D, QWB, and AQoL-8D Multiattribute Utility Instruments. Med Decis Making. 2015;35:276–91. [PubMed]
21. Arnold D, Girling A, Stevens A, Lilford R. Comparison of direct and indirect methods of estimating health state utilities for resource allocation: review and empirical analysis. BMJ. 2009;339:b2688. [PubMed]
22. Llewellyn-Thomas HA. Health state descriptions. Purposes, issues, a proposal. Med Care. 1996;34:DS109–18. [PubMed]
23. Prosser LA, Lamarand K, Gebremariam A, Wittenberg E. Measuring family HRQoL spillover effects using direct health utility assessment. Med Decis Making. 2015;35:81–93. [PMC free article] [PubMed]
24. McLellan AT, Luborsky L, Woody GE, O'Brien CP. An improved diagnostic evaluation instrument for substance abuse patients. The Addiction Severity Index. J Nerv Ment Dis. 1980;168:26–33. [PubMed]
25. Bo H, Fu AZ. Predicting utility for joint health states: a general framework and a new nonparametric estimator. Med Decis Making. 2010;30:E29–39. [PubMed]
26. Wittenberg E, Prosser LA. Ordering errors, objections and invariance in utility survey responses: a framework for understanding who, why and what to do. Appl Health Econ Health Policy. 2011;9:225–41. [PubMed]
28. Connock M, Juarez-Garcia A, Jowett S, et al. Methadone and buprenorphine for the management of opioid dependence: a systematic review and economic evaluation. Health Technol Assess. 2007;11:1–171, iii-iv. [PubMed]
29. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10:407–15. [PubMed]
30. Luo N, Johnson J, Coons SJ. Using instrument-defined health state transitions to estimate minimally important differences for four preference-based health-related quality of life instruments. Med Care. 2010;48:365–71. [PubMed]
31. Redelmeier DA, Guyatt GH, Goldstein RS. Assessing the minimal important difference in symptoms: a comparison of two techniques. J Clin Epidemiol. 1996;49:1215–9. [PubMed]
32. Schunemann HJ, Griffith L, Jaeschke R, Goldstein R, Stubbing D, Guyatt GH. Evaluation of the minimal important difference for the feeling thermometer and the St. George's Respiratory Questionnaire in patients with chronic airflow obstruction. J Clin Epidemiol. 2003;56:1170–6. [PubMed]
33. Walters SJ, Brazier JE. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Qual Life Res. 2005;14:1523–32. [PubMed]
34. Earnshaw V, Smith L, Copenhaver M. Drug Addiction Stigma in the Context of Methadone Maintenance Therapy: An Investigation into Understudied Sources of Stigma. International journal of mental health and addiction. 2013;11:110–22. [PMC free article] [PubMed]
35. Mars SG, Bourgois P, Karandinos G, Montero F, Ciccarone D. “Every ‘never’ I ever said came true”: transitions from opioid pills to heroin injecting. The International journal on drug policy. 2014;25:257–66. [PMC free article] [PubMed]
36. Nosyk B, Anglin MD, Brissette S, et al. A call for evidence-based medical treatment of opioid dependence in the United States and Canada. Health Aff (Millwood) 2013;32:1462–9. [PMC free article] [PubMed]
37. Reisinger HS, Schwartz RP, Mitchell SG, et al. Premature discharge from methadone treatment: patient perspectives. J Psychoactive Drugs. 2009;41:285–96. [PMC free article] [PubMed]
38. Aden B, Dunning A, Nosyk B, Wittenberg E, Bray JW, Schackman BR. Impact of Illicit Drug Use on Health-Related Quality of Life in Opioid Dependent Patients Undergoing HIV Treatment. J Acquir Immune Defic Syndr. 2015 [PMC free article] [PubMed]
39. Nosyk B, Sun H, Guh DP, et al. The quality of eight health status measures were compared for chronic opioid dependence. J Clin Epidemiol. 2010;63:1132–44. [PubMed]
40. Pyne JM, Booth BM, Farahati F, Tripathi S, Smith GR, Marques PR. Preference-weighted health status associated with substance use-disorders treatment. J Stud Alcohol. 2006;67:436–44. [PubMed]
41. Pyne JM, French M, McCollister K, Tripathi S, Rapp R, Booth B. Preference-weighted health-related quality of life measures and substance use disorder severity. Addiction. 2008;103:1320–9. discussion 30-2. [PMC free article] [PubMed]
42. Strain EC. Assessment and treatment of comorbid psychiatric disorders in opioid-dependent patients. Clin J Pain. 2002;18:S14–27. [PubMed]
43. Rosenblum A, Joseph H, Fong C, Kipnis S, Cleland C, Portenoy RK. Prevalence and characteristics of chronic pain among chemically dependent patients in methadone maintenance and residential treatment facilities. JAMA. 2003;289:2370–8. [PubMed]
44. Morasco BJ, Gritzner S, Lewis L, Oldham R, Turk DC, Dobscha SK. Systematic review of prevalence, correlates, and treatment outcomes for chronic non-cancer pain in patients with comorbid substance use disorder. Pain. 2011;152:488–97. [PMC free article] [PubMed]
45. Brecht ML, Huang D, Evans E, Hser YI. Polydrug use and implications for longitudinal research: ten-year trajectories for heroin, cocaine, and methamphetamine users. Drug Alcohol Depend. 2008;96:193–201. [PMC free article] [PubMed]
46. Wittenberg E, Prosser LA. Disutility of illness for caregivers and families: a systematic review of the literature. Pharmacoeconomics. 2013;31:489–500. [PMC free article] [PubMed]
47. Fu AZ, Kattan MW. Utilities should not be multiplied: evidence from the preference-based scores in the United States. Med Care. 2008;46:984–90. [PubMed]
48. Aden B, Nosyk B, Wittenberg E, Schackman BR. Health-Related Quality of Life in HIV-Infected and At-Risk Women: The Impact of Illicit Drug Use and Hepatitis C on a Community Preference Weighted Measure. Med Decis Making. 2013;34:800–8. [PMC free article] [PubMed]