Two presentations at the workshop focused on existing measures of recovery. Alexandre Laudet (Center for the Study of Addictions and Recovery, National Development and Research Institutes, director emeritus) discussed existing measures of recovery from substance use. Mark Salzer (Temple University) discussed measures of recovery from mental disorders.
Laudet began by saying that the discussions underscored the great variety of definitions of recovery that have been developed. It is also clear that there are some common elements and overlap, both within the fields of substance use and mental illness but also across the two fields.
Close attention to the methods used to measure recovery from substance use reveals, however, that most of the time the question that is used is simply about abstinence in a yes/no format. Furthermore, the question is often about abstinence from one specific substance, despite the fact that the use of multiple substances is typical among people with an addiction. Laudet said that this approach is widespread because it is practical. In addition, she noted, self-report questions usually refer to the past 30 days. Sometimes biological samples are collected, but unless the sample is hair, the data from the samples are limited to a shorter period of time. As discussed, another approach that has been used is to ask: Did you once have a problem with drugs and alcohol, but no longer do? (see Chapter 2).
Research shows that the SAMHSA definition of recovery contains dimensions that are meaningful to people in recovery, scientists, clinicians, and other stakeholders. The definition is multidimensional and implies change, with the main elements of a reduced relationship with substance use (either abstinence or significant reduction) and improvement in a person’s quality of life.
Recent research has measured substance use and people’s changing level of functioning and showed that different dimensions of recovery do not progress at the same time, nor necessarily in the same direction.1 For example, during the initial stages of recovery from substance use, mental health gets worse. For other dimensions, such as employment, there is gradual improvement. In the long term, there is improvement overall, across dimensions.
Laudet said she expects that it will become clear that it is not possible to measure recovery in a way that yields one number, because a single score cannot capture the construct adequately. If a prevalence estimate is needed, that will have to be based on more than just one question, such as whether the person is in recovery or not.
Laudet described three measures of recovery from substance use disorder that were discussed in a prior review of measures she prepared for SAMHSA.2 She noted that research on mental health recovery has a much longer tradition and there are more measures and instruments available for that area than for recovery from substance abuse.
The first measure she discussed was the Modular Survey, an initiative with the goal to identify and develop common indicators and measures of consumer perception of care. It has a set of 21 items that covered four domains: quality of services, perceived outcome improvement, connectedness, and commitment to change. However, Laudet pointed out, the majority of the people in recovery are likely not enrolled in services, so this set of questions is likely not suitable for SAMHSA’s current purposes.
The second available measure Laudet discussed is the recovery capital measure, which is focused on the quality and quantity of recovery capital.3 This instrument contains 23 items and measures eight domains: (1) reliance on religion, (2) spirituality, (3) recent sobriety, (4) stable income, (5) alcohol/drug-free environment, (6) percentage of lifetime spent free
1Dennis, M.L., Foss, M.A., and Scott, C.K. (2007). An eight-year perspective on the relationship between the duration of abstinence and other aspects of recovery. Evaluation Review, 31(6), 585-612.
3Sterling, R., Slusher, C., and Weinstein, S. (2008). Measuring recovery capital and determining its relationship to outcome in an alcohol dependent sample. American Journal of Drug and Alcohol Abuse, 34(5), 603-610.
from the effects of substance use, (7) satisfaction with living situation, and (8) amount of education and training. Laudet noted that recovery capital is an important concept, as was discussed by Grella (see Chapter 3), but the instrument measures aspects of what is needed to achieve recovery, not recovery itself.
The Client Assessment Inventory was the third measure discussed by Laudet.4 This instrument was developed to measure clients’ self-reports and staff evaluations of clients’ progress in the therapeutic community environment. The instrument contains 14 items and measures four broad dimensions: developmental, socialization, psychological, and community membership.
Laudet said that one of the characteristics the three measures have in common is that they are all multidimensional. They are all broadly consistent with the discussions about what ought to be included in a recovery measure, whether for substance use disorder, mental disorder (or even chronic illness). However, none of these measures are truly recovery measures, and Laudet argued that dedicated measures of recovery from substance use disorder still do not exist.
She listed the following criteria as essential in her view for a recovery measure:
- able to quantify change;
- has sound psychometric properties;
- brief to be feasible for repeated administration, especially in the context of “concurrent recovery monitoring;”5 and
- applicable across populations in terms of gender, age, cultural background, recovery path, and recovery stage.
Laudet said that although none of the measures she reviewed are suitable for SAMHSA’s goals, there are several additional measures that should be mentioned in this context. One of these is the Client Outcome Measures for Discretionary Programs developed by SAMHSA as part of the Government Performance and Results Modernization Act. The goal of this measure is to assess and track client progress for the purposes of external accountability to program funders. The questions are adminis-
4Kressel, D., De Leon, G., Palij, M., and Rubin, G. (2000). Measuring client clinical progress in therapeutic community treatment. The therapeutic community Client Assessment Inventory, Client Assessment Summary, and Staff Assessment Summary. Journal of Substance Abuse Treatment, 19(3), 267-272.
5McLellan A.T., McKay J.R., Forman R., Cacciola J., and Kemp J. (2005). Reconsidering the evaluation of addiction treatment: from retrospective follow-up to concurrent recovery monitoring. Addiction, 100(4), 447-458.
tered at entry into treatment, at discharge, and 6 months after discharge. Domains include substance use; family and living conditions; education, employment, and income; crime and criminal justice involvement; mental and physical health treatment and recovery; and social connectedness. Laudet said that although the instrument is only used in the context of service delivery, it does have many of the dimensions that have been discussed as relevant to substance use recovery, including the acknowledgement of the process of change. In other words, it may be possible to build on this measure.
Another measure that is available is the Addiction Severity Index (ASI),6 which has the goal of assessing and measuring change in addiction severity. The domains covered in this scale include employment, medical, psychiatric, family and social, alcohol and drug use, and legal status. Laudet said that this is one of the most well-known addiction measurements. This index is also primarily used in the context of services, but it contains relevant dimensions, and it would be possible to capitalize on this work.
Laudet said that an existing measure that could be particularly relevant is the World Health Organization Quality of Life Instrument (WHOQOL). This instrument has been discussed at several expert meetings and by people in recovery as a measure that might represent a good starting point for measuring important aspects of recovery. The U.N. Treatnet Group, for example, recommended that in the absence of a dedicated measure of addiction recovery, both the ASI and the WHOQOL be used to measure the following domains: maintenance of abstinence or reduction in substance dependence, improvement in personal and social functioning, improvement in mental and physical health, reduction in risky behavior that could affect health, and overall improvements in increasing access to livelihoods assets and recovery capital. In 2007, the Betty Ford Expert Panel recommended the use of the WHOQOL, along with a measure of sobriety. The WHOQOL was also the measure selected in Connecticut by a group of people in recovery as the instrument most relevant to their experiences and needs. Laudet noted that Connecticut was the first state to adopt a recovery-orientation in 1999, and the WHOQOL was subsequently included in the state’s consumer survey.
Laudet noted that the original WHOQOL contains 100 items, but an abbreviated, psychometrically strong version, the WHOQOL BREF, is also available, and is typically the one that is used. The scale yields four scores in four domains, and there is an additional overall quality-of-life question. The four dimensions measured by the WHOQOL BREF are as follows:
6McLellan A.T., Kushner H., Metzger D., Peters R., Smith I., Grissom G., Pettinati H., and Argeriou M. (1992). The Fifth Edition of the Addiction Severity Index. Journal of Substance Abuse Treatment, 9(3), 199-213.
- physical (pain, energy, sleep, mobility, activities, medication, work);
- psychological (positive feelings, cognitions, self-esteem, body image, negative feelings, spirituality);
- social relationships (personal relations, social support, sex); and
- environment (safety and security, home environment, finance, health and social care, information, leisure, physical environment, transport).
Laudet summarized the advantages of the WHOQOL BREF:
- in the public domain
- capitalizes on decades of field work
- cross-culturally validated in 15 centers worldwide
- strong psychometrics
- assesses domains that are highly relevant to recovery
- assesses both positive and negative aspects of life, objective and subjective ratings
- published norms for well and ill persons
- relatively short—the 26 items require 20 minutes to administer or 15 minutes to self-administer
- developed to be broadly applicable across disorder types, varying severity of illness, and diverse socioeconomic and cultural subgroups
- serves as a precedent for methods to develop supplemental population-specific or disorder-specific modules to best capture relevant dimensions for a given group
Laudet noted that the World Health Organization defines quality of life as being close to one’s expectation of an ideal in one’s own context. In other words, the expectations may be different depending on whether the person is an 85-year-old with arthritis or a 32-year-old athlete who just had a bad accident. This is another way of looking at objective and subjective aspects of recovery. The WHOQOL BREF also recognizes that while there are general similarities across groups, people dealing with specific situations or conditions may have certain needs and preoccupations that are not captured by an overall quality-of-life instrument.
The supplemental modules are designed to address those issues. For example, there are modules for HIV-positive individuals and individuals living with chronic pain. Laudet added that there was discussion at the Betty Ford meeting about also developing a module specific to recovery from substance use. She argued that the WHOQOL BREF along with a
substance use recovery specific module might be the most suitable avenue to explore as a psychometrically strong measure of recovery.
One of the questions raised by SAMHSA was whether remission from symptoms should be included in the agency’s working definition of recovery. Laudet said that she thinks that remission should be included because from the work she has done with people in recovery it is clear that their definition of recovery does include their relationship with substance use. Abstinence is only a means to an end, but it tends to be regarded by people as a central element in their recovery experiences. For example, they will say that they have not used alcohol or a drug for a certain number of years.
A follow-up question on this point is how to operationalize remission if it is to be measured. It could be that one is using less than previously; one is being completely abstinent; or one is being completely abstinent from both drugs and alcohol. This is a complicated question and requires further thinking.
In conclusion, Laudet argued that quantifying the problem is not particularly difficult because one can produce estimates of the number of people who meet the Diagnostic and Statistical Manual of Mental Disorders criteria for substance use, which could be done through the National Survey on Drug Use and Health (NSDUH). But quantifying the solution, that is, measuring recovery from a substance use disorder, is much more difficult because recovery is a process, and it is multidimensional. Laudet argued, however, that measuring recovery is not impossible if there is the will and the funds to do it. One option might be to add the WHOQOL BREF to the NSDUH, for example, because the NSDUH already has questions on the substance use part of the equation.
Laudet said that there are estimates of the percentage of people with a drug and alcohol problem, and these percentages do not fluctuate very significantly from year to year. It is likely that the percentage of people in recovery does not fluctuate very significantly either. To produce more precise estimates it will be important to first determine what the intended primary uses of the data will be. For example, is it to inform funding for services? If yes, for which target group: for those needing treatment or for those in recovery? The answer to these questions would inform decisions about how to develop scales that are concise and contain the most relevant dimensions. The answers could also help identify the target population for the questions: for example, all lifetime users, past-year or past-month users, those identified as having a substance use disorder, or those who self-identify as being in recovery. Laudet said that there are a number of open questions at this stage, but it seems clear that to monitor, develop, and fund services there is a need for a multidimensional measure of change, not a single score.
Kenneth Wells (University of California, Los Angeles) commented that if adding a question to an existing survey is the avenue to be pursued, it should be noted that some of the domains measured by the various instruments can produce data that would be useful to have about the population as a whole, not just people with a substance use or mental disorder. Some questions on these domains may already be included in some form in existing surveys, or, if not, questions from the instruments discussed could be adapted in a way that makes them applicable to the general population. This strategy could enable SAMHSA to obtain data on a larger number of relevant questions from an existing survey. Laudet noted that there would be some downsides to a decision that involved only using parts of a scale that has not been tested to be used as Wells suggested. Wells agreed that testing would need to be done.
Dean Kilpatrick (Medical University of South Carolina) added that if some of the measures are useful both for people in recovery and the general population, administering them to both could provide interesting information that can further inform the development of measures of recovery. Some of the studies discussed during the workshop illustrate that there might be variation in these measures among those who have not had any substance use or mental disorder.
Peter Gaumond (Office of National Drug Control Policy) commented on the question of whether to measure partial recovery or remission from a diagnosable substance use disorder, in other words, cases when one continues to use a substance but no longer meets diagnostic criteria. He argued that in these cases it would still be important to know what the relationship is between a person’s condition at that point and their quality of life. That information might also provide insight into whether it makes sense to describe people with those characteristics as being in recovery or in some other way.
Christine Grella (University of California, Los Angeles) commented that if a scale such as the WHOQOL BREF were to be added to the NSDUH or another survey, it would be very useful to also have a targeted subsample that would receive the questions more than once, in the form of follow-up surveys. A longitudinal design would be particularly useful to answer questions about the process of change. The subsample could also include oversamples of populations of specific interest, such as people who have used substances other than alcohol and marijuana. She added that one challenge in general population surveys is that such subgroups are a very small part of the population.
Salzer began by saying that he is concerned about the large number of definitions of recovery that exist, particularly in the context of mental health. The definitions that have been discussed are only a subset of the range of definitions that are in use. He argued that research on recovery is further complicated by the lack of adequate emphasis on measurement issues and how measures map onto the many definitions. However, although there are limitations in the existing measures, there is a large number of instruments, and the list includes some potentially good measures that could be used for at least some of the subdomains of interest. He added that he believes that sometimes deconstructing scales and reusing them is justified, if these decisions can be supported with evidence.
In his discussion of several measures of recovery from mental disorders, Salzer said that in the category of measures that focus on recovery-oriented attitudes, beliefs, and knowledge, the most notable one is the Recovery Attitudes Questionnaire.7 The measure has two factors: recovery is possible and needs faith, and recovery is difficult and differs among people.
Another category of recovery measures focuses on perceptions of the extent to which policies, programs, and practices create a recovery-promoting environment, and these are used in numerous states and systems. These measures include recovery-oriented systems indicators that measure the extent to which people feel supported in the areas of meaningful activities, basic material resources, peer support, choice, social relationships, formal service staff, formal services, and self or holism.8 Another measure in this category is the Recovery Self-Assessment, which has four versions; it assesses the degree to which programs implement recovery-oriented services.9 The measure has five factors: (1) life goals, (2) involvement, (3) diversity of treatment options, (4) choice, and (5) individually tailored services. A “person in recovery” version of this instrument also exists, with 36 items.
Salzer said that the measures that are most in line with the focus of the discussion so far are the clinical measures and the consumer-focused or subjective measures or recovery. This distinction was proposed by Alan
7Borkin, J.R., Steffen, J.J., Ensfield, L.B., Krzton, K., Wilder, K., and Yangarber, N. (2000). Recovery Attitudes Questionnaire: Development and evaluation. Psychiatric Rehabilitation Journal, 24, 95-102.
8Dumont, J.M., Ridgway, P., Onken, S.J., Dornan, D.H., and Ralph, R.O. (2005). Recovery Oriented Systems Indicators Measure (ROSI). Available: https://www.power2u.org/downloads/ROSI-Recovery%20Oriented%20Systems%20Indicators.pdf. [June 2016].
9O’Connell, M., Tondora, J., Croog, G., Evans, A., and Davidson, L. (2005). From rhetoric to routine: Assessing recovery-oriented practices in a state mental health and addiction system. Psychiatric Rehabilitation Journal, 28, 378-386.
Bellack10 and is similar, although not identical, to the perspective presented by Kim Mueser (see Chapter 4). Most of what is known about mental health recovery comes from studies that have used a clinical model. Some of these studies measure time out of the hospital or since last hospitalization. One study with a multidimensional approach looked at a 2-year period of functioning within specified normal limits in the domains of symptomatology, participating in work or school, living independently, and maintaining social relationships.11 Salzer noted that these approaches tend to be focused on fairly broad behaviors.
In terms of measures focused on the subjective experience of recovery, Salzer noted that there have been several reviews published. A review by Burgess and colleagues looked at how measures fit nine criteria:
- explicitly measures domains related to personal recovery,
- brief and easy to use (50 or fewer items),
- consumer perspective,
- scientifically scrutinized,
- sound psychometric properties,
- applicable to Australian context,
- acceptable to consumers, and
- promotes dialogue between consumers and providers.12
Salzer noted that although the review used applicability to the Australian context as a criterion, all of the measures are also applicable in the United States. Burgess and colleagues identified four measures that met all the criteria: the Recovery Assessment Scale (RAS), the Illness Management and Recovery Scales (IMR), the Stages of Recovery Instrument (STORI), and the Recovery Process Inventory (RPI).
Another recent review of measures of recovery from mental disorder was conducted by Shanks and colleagues.13 The focus of this review was on the fit with the time framework of recovery, and the authors defined recovery as connectedness, hope, identity, meaning, and empowerment.
10Bellack, A.S. (2006). Scientific and consumer models of recovery in schizophrenia: Concordance, contrasts, and implications. Schizophrenia Bulletin, 32, 432-442.
11Liberman, R.P., Kopelowicz, A., Ventura, J., and Gutkind, D. (2002). Operational criteria and factors related to recovery from Schizophrenia. International Review of Psychiatry, 14, 256-272.
12Burgess, P., Pirkis, J., Coombs, T., and Rosen, A. (2011). Assessing the value of existing recovery measures for routine use in Australian mental health services. Australian and New Zealand Journal of Psychiatry, 45, 267-280.
13Shanks, V., Williams, J., Leamy, M., Bird, V.J., Le Boutillier, C., and Slade, M. (2013). Measures of personal recovery: A systematic review. Psychiatric Services, 64(10), 974-980.
The measures identified in this review as best fitting the criteria specified by the authors included the four on the Burgess list, and an additional two: the Maryland Assessment of Recovery (MARS) and the Questionnaire About the Process of Recovery (QPR).
Salzer provided a brief overview of each of the measures that were highlighted in the Shanks review. The IMR has 15 items that assess personal goals, knowledge of mental illness, involvement of significant others, impaired functioning, symptoms, stress, coping, relapse prevention, hospitalization, medication, and use of drugs and alcohol.14 The scale has no specific subscales. The MARS has 25 items, and a 12-item version is also available.15 The scale measures six domains: self-direction or empowerment, holistic, nonlinear, strengths-based, responsibility, and hope. The QPR has 22 items with two subscales, for intrapersonal and interpersonal tasks associated with recovery.16 The RPI has 22 items that measure six factors: anguish, connected to others, confidence and purpose, others’ care or help, living situation, and hopeful or cares for self.17
The RAS has three versions, with 41, 24, and 20 items.18 The factors measured are personal confidence and hope, willingness to ask for help, goal and success orientation, reliance on others, and not being dominated by symptoms. Salzer noted that he particularly liked that the scale measures whether one’s life continues to be dominated by symptoms because it reflects a view of recovery that focuses on living a satisfying and fulfilling life with or without the presence of mental illness symptoms.
Finally, the STORI has 50 items, with 30-item versions available: it measures five stages and four recovery processes.19 The five stages are moratorium (a stage of hopelessness and self-protective withdrawal), awareness (the realization that recovery and a fulfilling life is possible, preparation (the search for personal resources and external sources of
14Salyers, M.P., Godfrey, J.L., Mueser, K.T., and Labriola, S. (2007). Measuring illness management outcomes: A psychometric study of clinician and consumer rating scales for illness self-management and recovery. Community Mental Health Journal, 43, 459-480.
15Drapalski, A.L., Medoff, D., Unick, G.J., Velligan, D.L., Dixon, L.B., and Bellack, A.S. (2012). Assessing recovery of people with serious mental illness: Development of a new scale. Psychiatric Services, 63, 48-53.
16Neil, S., Kilbride, M., and Pitt, L, (2009). The Questionnaire about the Process of Recovery (QPR): A measurement tool developed in collaboration with service users. Psychosis, 1, 145-155.
17Jerrell, J.M., Cousins, V.C., and Roberts, K.M. (2006). Psychometrics of the Recovery Process Inventory. Journal of Behavioral Health Services and Research, 33, 464-473.
18Corrigan, P.W., Giffort, D., Rashid, F., Leary, M., and Okeke, I. (1999) Recovery as a psychological construct. Community Mental Health Journal, 35, 231-239.
19Andresen, R., Oades, L., and Caputi, P. (2003). The experience of recovery from schizophrenia: Toward an empirically validated stage model. Australian and New Zealand Journal of Psychiatry, 37, 586-594.
help), rebuilding (taking positive steps toward meaningful goals), and growth (a sense of control over one’s life and looking forward to the future). The four processes are hope, responsibility for wellness and control of life generally, establishing a positive identity, and finding meaning and purpose in life.
Salzer said that there are several psychometrically sound measures with some evidence of validity, and many of them have been informed by consumer perspectives on recovery. He pointed out, however, that the measures do not necessarily map onto the existing definitions of recovery. In addition, the construct of recovery that is being measured sometimes appears to be an amalgamation of other constructs, including constructs measured with already existing scales, such as quality-of-life measures and self-efficacy measures. He added that sensitivity to change also remains a concern for these measures. It is important to have a measure that is sensitive to change based on intervention and that can capture the process aspect of recovery. However, he noted, the existing measures do not typically have evidence of sensitivity to change.
In terms of SAMHSA’s goals to measure recovery, one specific challenge raised by Salzer is that the agency has a definition with several associated components, as well as principles of recovery. These elements complicate the question of what is being measured. He argued that several of the principles could be measured with existing measures, or subscales, if measuring them is the goal. Those principles include hope, person driven, peer support, relational, strengths/responsibility, and respect.
Salzer said that some of the components of health, home, purpose and community in SAMHSA’s definition can also be measured, although there are some operationalization challenges. For example, health could be operationalized as overcoming or managing one’s disorders or symptoms. To measure this, it would be possible to use a symptom measure or to use the subscale of not being dominated by symptoms from the RAS. It would also be possible to look at whether a person is making informed, healthy choices that support physical and emotional well-being. Salzer said that these items could be measured by using the knowledge, relapse prevention planning, and medication taking items from the IMR, although he noted that some experts might disagree about whether the medication item belongs here. Other measures could include the personal responsibility items from the MARS and STORI scales (such items as “I work hard to find ways to cope with problems in my life”). Another option would be to use a healthy behavior checklist and just ask people how much are they smoking, are they taking care of their health, eating right, are they exercising, are they being active, and so on.
There are fewer options for measuring the home component of SAMHSA’s definition of recovery, which includes having a stable place
to live and a safe place to live. Salzer said that one possibility to measure the safe home dimension could be the Kloos and Shah neighborhood safety measure.20
Another component in SAMHSA’s definition is purpose, which Salzer said could be measured in several different ways. One option would be to use item #5 from the IMR: “How much time do you spend working, volunteering, being a student, being a parent, taking care of someone else or someone else’s house or apartment? That is, how much time do you spend doing activities that are expected of you for or with another person? (This would not include self-care or personal home maintenance).” Another option would be to use the Temple University Community Participation Measure that asks about participation in 22 areas, importance of participation, and sufficiency.21 The Community Participation Indicators is another measure that could be suitable.22 This scale also measures various aspects of participation. Salzer noted that he likes the purpose component of SAMHSA’s definition and particularly likes the idea of a community participation approach to measuring it because it reflects a focus on what happens to a person who is in recovery.
Finally, to measure community, Salzer said that a large number of measures exist that measure social network size and aspects such as perceived social support.
Salzer said that if he had to develop a brief measure of recovery based on SAMHSA’s definition, he would consider the following items:
- “Coping with my mental illness is no longer the main focus of my life.” (health, RAS item)
- “I am making informed, healthy choices that support my physical and emotional well-being.” (health)
- “I can live in my current housing as long as I would like.” (home)
- “I feel safe in my current housing situation.” (home)
- “How much time do you spend working, volunteering, being a student, being a parent, taking care of someone else or someone else’s house or apartment? That is, how much time do you spend doing activities that are expected of you for or with another
20Kloos, B., and Shah, S. (2009). A social ecological approach to investigating relationships between housing and adaptive functioning for persons with serious mental illness. American Journal of Community Psychology, 44, 316-326.
21Salzer, M.S., Brusilovskiy, E., Prvu-Bettger, J., and Kottsieper, P. (2014). Measuring community participation of adults with psychiatric disabilities: Reliability of two modes of data collection. Rehabilitation Psychology, 59(2), 211-219.
22Heinemann, A. (2007). Community Participation Indicators. Available: http://www.ric.org/app/files/public/3598/CPI-commnity-participation.pdf [February 2016].
person? (This would not include self-care or personal home maintenance.) (purpose, IMR item #5)
- “I have people in my life who provide support, friendship, love, and hope.” (community)
Salzer said that although mixing and matching items from various scales would be a controversial approach, he is pragmatic and realizes that a very lengthy interview may not be an option. Whether the use of the six items he proposed would work could be tested.
Kim Mueser (Boston University) clarified that the medication item on the IMR scale is framed in terms of whether people who have decided to take medication indeed take it as directed. If people choose not to take medication, it is not rated.
Amaro commented that the extent to which housing is under an individual’s control varies by a person’s place in the social hierarchy. Some of the characteristics of housing depend on local ordinances and policies and whether one has a criminal record. Segregation can also play a role in where a person lives. Exposure to drugs in one’s community can act as a trigger and influence relapse. Amaro argued that some of the items that are conceptualized as individual-level factors are not individually determined for many people.
Salzer replied that the personal and environmental factors certainly interact, and he noted that one of his areas of research is looking at environmental factors as they influence recovery, well-being, and participation. Although it is true that the six questions he proposed would be administered in a vacuum, the current goal is to measure recovery, not to measure the determinants of recovery. He added that in some of his research measuring recovery among people with serious mental illness, he and his colleagues found that objective environmental influences, such as the amount of crime in the neighborhood, had very little influence on recovery. The team is now in the process of conducting follow-up studies that assess the subjective experience of these environmental factors.
Amaro asked Salzer to clarify what he sees as the issues with the measure’s ability to capture change. She said that one issue she has come across in her use of the scales with people in treatment is that some of the questions are about change in circumstances that can take a long time to shift because of social barriers (e.g., housing, employment, education, involvement with the criminal justice system).
Salzer said that if sensitivity to change is interpreted as sensitivity to changes that a mental health or a substance use treatment program can generate, treatment can only have limited effects, especially for concepts for which there is a lot of environmental influence. However, if one were to look at services that address those environmental factors, such as sup-
ported housing programs, the influence of those factors might be better reflected in the measures.
Mueser said that the IMR is sensitive to change, but it is not a pure measure of recovery: it is a measure of illness self-management and recovery. Several studies have shown that the IMR program does change the movable factors that it targets.
Graham Kalton (Westat) said that he is still wrestling with the question of how to operationalize recovery. If someone who used to have no social contact moved toward slightly more social contact but still falls far short of what one would like to see, would that be considered recovery? He also noted that there was relatively little discussion about issues that may be specific to how these scales would work in surveys of the general population. Salzer agreed with Kalton that the main challenge is that recovery means different things to different people.