Illness Management and Recovery (IMR) Scale

Abstract

The Illness Management and Recovery (IMR) Scale is a standardized, multidimensional psychometric instrument developed by Kim T. Mueser and colleagues in 2004 at the New Hampshire-Dartmouth Psychiatric Research Center. Created to evaluate outcomes within the evidence-based Illness Management and Recovery program, the scale measures an individual’s self-management of severe mental illness (such as schizophrenia, schizoaffective disorder, and bipolar disorder) alongside their personal trajectory of psychiatric recovery. The instrument exists in two parallel versions: a 15-item Client Self-Rating version and a corresponding 15-item Clinician Rating version. Each item targets a specific behavioral, cognitive, or functional facet of recovery and illness management—including personal goal progression, psychiatric knowledge, social network involvement, structured role functioning, symptom distress, impairment, relapse prevention, coping efficacy, self-help engagement, medication adherence, and substance-related interference. Every item utilizes a distinct, clinically anchored 5-point Likert-type scale (ranging from 1 to 5), yielding a total composite score (ranging from 15 to 75) where higher values signify superior self-management and advanced recovery. Confirmatory and exploratory factor analyses across diverse international samples have consistently revealed a robust multidimensional architecture, typically comprising three primary factors: Recovery, Illness Management, and Substance Use. Psychometric investigations document adequate to high internal consistency (Cronbach’s alpha ranging from .72 to .87 for total scores), strong test-retest stability (intraclass correlation coefficients between .70 and .89), solid convergent validity with validated recovery and quality-of-life instruments, and significant sensitivity to clinical change following structured psychosocial interventions.

Keywords

Illness Management and Recovery Scale, IMR, psychiatric rehabilitation, severe mental illness, recovery-oriented care, psychometrics, self-management, psychiatric outcome measurement, schizophrenia, relapse prevention

Authors

The Illness Management and Recovery (IMR) Scales were developed in 2004 by a multidisciplinary team of psychiatric rehabilitation researchers and clinicians affiliated with the New Hampshire-Dartmouth Psychiatric Research Center (Concord, NH) and the Substance Abuse and Mental Health Services Administration (SAMHSA):

• Kim T. Mueser, Ph.D. — Center for Psychiatric Rehabilitation and Department of Occupational Therapy, Boston University; formerly of Dartmouth Medical School and the New Hampshire-Dartmouth Psychiatric Research Center. (Contact: [email protected] / [email protected]).
• Susan Gingerich, M.S.W. — Independent Consultant and Trainer, co-developer of the IMR Curriculum and Social Skills Training programs.
• Michelle P. Salyers, Ph.D. — Department of Psychology, Indiana University-Purdue University Indianapolis (IUPUI) and ACT Center of Indiana.
• Angela B. McGuire, Ph.D., M.S.W. — Health Services Research and Development Service, Department of Veterans Affairs, and Indiana University.
• R. U. Reyes, M.A. — New Hampshire-Dartmouth Psychiatric Research Center.
• H. Cunningham, B.A. — New Hampshire-Dartmouth Psychiatric Research Center.

Purpose

The Illness Management and Recovery (IMR) Scale was specifically designed to bridge a critical methodological gap in psychiatric rehabilitation: the absence of a brief, standardized, dual-perspective outcome instrument capable of capturing both objective behavioral indicators of illness self-regulation and subjective dimensions of personal recovery. During the early 2000s, mental health policy and service delivery models shifted dramatically from symptom suppression and maintenance toward a recovery-oriented paradigm. Concurrently, the Substance Abuse and Mental Health Services Administration (SAMHSA) commissioned the National Evidence-Based Practices Project, which identified illness self-management as an essential psychosocial intervention for individuals diagnosed with severe mental illnesses (SMI). The IMR curriculum was packaged as a comprehensive, structured psychosocial intervention consisting of psychoeducation, behavioral tailoring for medication adherence, relapse prevention planning, social skills training, and cognitive-behavioral strategies for persistent symptoms.

To assess fidelity, monitor longitudinal progress, and evaluate therapeutic efficacy, Mueser and colleagues devised the IMR Scales. The primary clinical and research objectives of the instrument encompass:

• Standardized Outcome Evaluation: Providing an empirically validated benchmark for measuring gains achieved by participants enrolled in the 10-to-12-month IMR curriculum or comparable psychiatric rehabilitation programs.
• Multiperspective Concordance Monitoring: Utilizing parallel client self-report and clinician-rated forms to evaluate alignment, therapeutic alliance, and perceptual discrepancies between consumers and mental health practitioners regarding illness impact, functional capacity, and symptom distress.
• Individualized Treatment Planning: Pinpointing specific deficits across self-care, social network engagement, relapse prevention awareness, or vocational participation, thereby guiding individualized behavioral interventions and goal setting.
• Cross-System Quality Improvement: Serving as a robust metric for community mental health agencies, outpatient psychiatric centers, and clinical trials investigating recovery-focused interventions across varied international public healthcare systems.

Psychological Construct

The psychological construct evaluated by the IMR Scale is illness self-management and personal recovery in severe mental illness. Rather than framing recovery purely as clinical remission or the absence of psychopathology, the IMR construct conceptualizes recovery as a multidimensional, ongoing personal process of living a meaningful, self-directed life while actively managing psychiatric symptoms and mitigating the disruptive consequences of emotional disorders.

The construct integrates cognitive, behavioral, affective, and systemic-social domains, operationalized across several interrelated facets:

• Goal Formulation and Agency (Item 1): Evaluates an individual’s ability to identify personally meaningful life aspirations (e.g., employment, independent housing, creative endeavors) and mobilize active, goal-directed behavior to accomplish them, reflecting self-efficacy and agency.
• Psychoeducational Mastery and Knowledge (Item 2): Quantifies the depth of declarative and procedural understanding a client possesses regarding symptom profiles, pharmacological actions, psychological treatments, and empirically supported coping techniques.
• Social Support Mobilization and Social Capital (Items 3 & 4): Assesses the structural and functional dimensions of the individual’s interpersonal ecosystem. Item 3 evaluates natural support systems (family, friends, intimate partners) actively involved in collaborative psychiatric care, while Item 4 measures weekly contact with wider community networks outside the immediate family (co-workers, peers, neighbors).
• Engagement in Meaningful, Structured Social Roles (Item 5): Measures time invested in socially recognized, productive roles—including competitive employment, formal education, volunteer positions, and parenting—which directly counter the social marginalization and occupational deprivation frequently accompanying chronic psychiatric diagnoses.
• Symptom Impact and Distress Tolerance (Items 6 & 7): Separates the subjective affective burden of psychiatric symptoms (distress) from the objective behavioral interference those symptoms exert on everyday instrumental and social functioning.
• Active Relapse Prevention and Crisis Planning (Items 8, 9, & 10): Assesses proactive self-monitoring, knowledge of idiosyncratic early warning signs, formalization of a written psychiatric advance directive or crisis plan, alongside tracking the actual frequency of clinical relapses and psychiatric hospitalizations over longitudinal intervals.
• Everyday Coping Competence (Item 11): Evaluates the client’s perceived self-efficacy and daily behavioral execution of stress-reduction strategies and symptom-management techniques.
• Peer Support and Consumer-Run Service Utilization (Item 12): Assesses integration into independent mutual-aid networks, peer-led support groups, drop-in centers, and standardized wellness tools such as the Wellness Recovery Action Plan (WRAP).
• Pharmacological Self-Management (Item 13): Quantifies treatment adherence and collaborative medication management, evaluating the consistency with which prescribed psychiatric regimens are followed.
• Substance-Related Interference (Items 14 & 15): Explicitly measures the degree to which alcohol and drug use disrupt social relations, financial security, housing stability, clinical appointment adherence, and psychiatric symptom stability.

Theoretical Framework

The theoretical framework underpinning the Illness Management and Recovery Scale synthesizes four foundational paradigms within behavioral science and contemporary psychiatric rehabilitation: the Stress-Vulnerability-Coping Model, Albert Bandura’s Social Cognitive Theory, the Transtheoretical Model of Health Behavior Change, and the Empowerment / Recovery Model.

The Stress-Vulnerability-Coping Model

Originally formulated by Joseph Zubin and Bonnie Spring (1977) and later elaborated by Robert Liberman and colleagues (1984), this model posits that the clinical course of severe psychiatric disorders is governed by a dynamic interaction among biological vulnerability (genetic, neurodevelopmental factors), environmental stressors (life events, socio-economic challenges), and protective factors (coping competencies, social support, medication). The IMR Scale operationalizes these protective factors directly: enhancing psychiatric knowledge, acquiring structured relapse prevention plans, engaging social supports, and maintaining treatment adherence systematically attenuate the psychophysiological impact of ambient stress, thereby minimizing clinical exacerbations.

Bandura’s Social Cognitive Theory and Self-Efficacy

Bandura (1986, 1997) demonstrated that human agency and behavioral execution are fundamentally mediated by perceived self-efficacy—an individual’s confidence in their ability to execute the courses of action required to manage prospective situations. The IMR Scale treats illness self-management not as a passive compliance construct, but as an active acquisition of self-regulatory skills. When individuals acquire knowledge, experience mastery in small goal-oriented increments (Item 1), practice cognitive-behavioral coping (Item 11), and observe peer modeling in mutual-help organizations (Item 12), their self-efficacy beliefs increase, leading to superior emotional regulation and sustained behavioral tenacity.

The Transtheoretical Model (Stages of Change)

Developed by Prochaska and DiClemente (1983), the Transtheoretical Model conceptualizes health-related behavior change as progression across distinct cognitive and motivational phases: precontemplation, contemplation, preparation, action, and maintenance. The behavioral response anchors across the IMR items reflect this progression. For instance, Item 8 (Relapse Prevention Planning) grades an individual from precontemplative non-awareness (“I don’t know how to prevent relapses”), through contemplation and preparation (“I know 1 or 2 things I can do, but I don’t have a written plan”), to active maintenance and collaborative action (“I have a written plan that I have shared with others”).

The Consumer Empowerment and Recovery Paradigm

Departing from the historical deficit-oriented medical model, the psychiatric recovery paradigm (Anthony, 1993; Deegan, 1988) emphasizes subjective purpose, personal meaning, civic participation, and autonomy regardless of ongoing psychiatric symptoms. The IMR Scale mirrors this paradigm by measuring engagement in normal societal roles (employment, education, parenting) and self-determined personal goals, affirming that recovery is not identical to biological cure.

Validity

The psychometric validity of both the Client Self-Rating and Clinician Rating versions of the IMR Scale has been rigorously evaluated in numerous clinical trials and observational studies across diverse cultural and national cohorts.

Convergent and Concurrent Validity

Convergent validity has been established by correlating IMR scores with established psychometric instruments measuring overlapping psychological constructs:

• Recovery and Empowerment: Statistically significant positive correlations have been reported between the IMR Client version and the Recovery Assessment Scale (RAS) (Pearson’s $r = .50$ to $.68, p < .001$), demonstrating that the scale captures core consumer-defined recovery processes (Salyers et al., 2007; Hasson-Ohayon et al., 2007).
• Quality of Life: Strong positive correlations appear with subjective quality-of-life indices, such as the Brief Quality of Life Interview (QOLI) ($r = .42$ to $.56, p < .001$), indicating that higher illness management competence directly reflects greater life satisfaction.
• Psychopathology and Distress: Negative correlations have been confirmed between the IMR total score and standardized symptom assessment scales, including the Brief Psychiatric Rating Scale (BPRS) ($r = -.34$ to $-.48, p < .01$) and the Positive and Negative Syndrome Scale (PANSS) ($r = -.31$ to $-.45, p < .01$). The distress and functioning items specifically demonstrate moderate-to-high correlations with symptom severity indices.
• Hope and Optimism: Significant associations emerge between IMR scores and validated hope scales, such as the Herth Hope Index ($r = .44, p < .001$), supporting the construct validity of the goal-progress and coping items.

Discriminant Validity

The scale discriminates effectively between distinct clinical subgroups. In validation studies, patients in acute inpatient psychiatric crisis demonstrated significantly lower IMR composite scores ($M \approx 42.1, SD = 8.4$) compared to stabilized community mental health outpatients participating in maintenance recovery programs ($M approx 56.8, SD = 7.9; t(182) = 11.4, p < .001$). Furthermore, the substance-specific items (Items 14 and 15) reliably distinguish between individuals with and without verified dual diagnoses (co-occurring substance use disorders) on standardized diagnostic schedules.

Predictive and Longitudinal Criterion Validity

Longitudinal randomized controlled trials (RCTs) assessing the IMR intervention (e.g., Salyers et al., 2010; Färdig et al., 2011; Levitt et al., 2009) have documented that baseline and interval IMR scores prospectively predict psychiatric rehospitalization rates. Patients scoring higher on the total IMR scale and specifically on Item 8 (relapse prevention) and Item 13 (medication adherence) demonstrated significantly fewer emergency room visits and hospital readmissions over 12-month follow-up horizons ($Odds Ratio = 0.74, p < .05$).

Client-Clinician Concordance

Studies examining concordance between the client self-rating and clinician rating (e.g., Hasson-Ohayon et al., 2007; Salyers et al., 2007) yield moderate, statistically significant bivariate correlations ($r = .35$ to $.60, p < .001$). Concordance is typically highest on objective, observable behaviors—such as time in structured roles, psychiatric hospitalizations, and medication compliance—and slightly lower on internal, subjective experiences, such as symptom distress and perceived coping efficacy, confirming the additive clinical utility of administering both versions.

Reliability

The IMR Scale has demonstrated robust reliability parameters across varied clinical settings, languages, and study designs.

Internal Consistency

Estimates of internal consistency via Cronbach’s alpha for the overall 15-item scale indicate acceptable to good reliability:

• Client Self-Rating Version: Cronbach’s $\alpha$ coefficients typically range between $.72$ and $.82$ across diverse validation cohorts (e.g., $\alpha = .72$ in Mueser et al., 2004; $\alpha = .79$ in Salyers et al., 2007; $\alpha = .82$ in the Swedish validation by Färdig et al., 2011).
• Clinician Rating Version: Clinician versions consistently demonstrate slightly higher internal consistency, typically yielding $\alpha$ coefficients between $.79$ and $.87$ (e.g., $\alpha = .83$ in Hasson-Ohayon et al., 2007; $\alpha = .85$ in McGuire et al., 2012), attributable to standardized clinical training and lower variability in interpretation across items.

Test-Retest Stability

Test-retest reliability has been evaluated over intervals ranging from one to four weeks among clinically stable outpatients:

• The Client Self-Rating scale exhibits intraclass correlation coefficients (ICC) ranging from $.70$ to $.86$, reflecting temporal stability in the absence of active clinical interventions.
• The Clinician Rating scale demonstrates similarly strong test-retest reproducibility, with ICC values typically ranging between $.78$ and $.89$.

Inter-Rater Reliability

When evaluated among independent mental health clinicians assessing the same client using clinical records and structured interview data, the Clinician Rating version exhibits substantial inter-rater agreement, with reported ICCs exceeding $.80$ across subscales.

Factor Analysis

The latent dimensionality of the IMR Scale has been investigated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) across several large international cohorts (e.g., United States, Israel, Sweden, the Netherlands, and Japan).

Empirical Factor Solutions

Although initially conceived as a single composite index of illness management and recovery, empirical factor analyses frequently reject a strictly unidimensional model in favor of three-factor or four-factor oblique structures:

The Canonical Three-Factor Model (Salyers et al., 2007; Hasson-Ohayon et al., 2007)

The most widely replicated structure across both client and clinician versions decomposes the 15 items into three correlated domains:

1. Recovery Factor: Comprises items reflecting subjective progress and social engagement, typically including Item 1 (Progress towards personal goals), Item 3 (Involvement of family/friends), Item 4 (Contact outside family), Item 5 (Time in structured roles), and Item 11 (Coping). Factor loadings for these items generally range from $.48$ to $.78$.
2. Illness Management Factor: Encompasses items measuring clinical symptom control, knowledge, and treatment engagement, typically loading Item 2 (Knowledge), Item 6 (Symptom distress), Item 7 (Impairment of functioning), Item 8 (Relapse prevention planning), Item 9 (Relapse of symptoms), Item 10 (Psychiatric hospitalizations), Item 12 (Self-help activities), and Item 13 (Medication adherence). Loadings generally range from $.42$ to $.75$.
3. Substance Use Factor: Forms a distinct, robust two-item factor consisting of Item 14 (Functioning affected by alcohol use) and Item 15 (Functioning affected by drug use), with item-factor loadings consistently exceeding $.80$.

Confirmatory Model Fit Indices

In CFA studies evaluating the three-factor oblique model (e.g., Salyers et al., 2007; Färdig et al., 2011; Sklar et al., 2012), goodness-of-fit statistics consistently surpass conventional psychometric thresholds:

• Root Mean Square Error of Approximation (RMSEA): Values typically range from $.042$ to $.061$ (indicating close approximate model fit).
• Comparative Fit Index (CFI): Values consistently fall between $.92$ and $.96$.
• Tucker-Lewis Index (TLI): Coefficients regularly range from $.90$ to $.95$.
• Standardized Root Mean Square Residual (SRMR): Values span $.045$ to $.058$.

Comparative model testing has repeatedly demonstrated that this three-factor configuration provides a substantially superior fit compared to a single-factor unidimensional model (where CFI often falls below $.80$ and RMSEA exceeds $.09$), highlighting that recovery and illness self-management constitute related but distinct dimensions of psychiatric outcomes.

Instrument / Measurement Tool

The structural characteristics, administrative guidelines, and scoring mechanisms of the instrument are detailed below:

• Standard Designation: Illness Management and Recovery (IMR) Scale.
• Instrument Variants:
  • Client Self-Rating Version (completed independently by the consumer or administered via interview).
  • Clinician Rating Version (completed by a case manager, psychiatrist, primary clinician, or psychiatric nurse).
• Number of Items: 15 items per version.
• Administration Time: Approximately 10 to 15 minutes for the client self-report; 5 to 10 minutes for the clinician-rated version.
• Target Population: Adolescents and adults diagnosed with severe mental illnesses (e.g., schizophrenia-spectrum disorders, bipolar affective disorder, major depressive disorder with psychotic features, severe personality pathology).
• Reading Level / Format: Written at approximately a 5th-grade reading level to accommodate individuals with neurocognitive deficits or literacy challenges.
• Response Anchors: Every single item possesses its own customized 5-point Likert scale (graded from 1 to 5), anchored with explicit behavioral descriptors rather than generic agree/disagree labels. Across all items, a rating of 1 represents the least adaptive state (e.g., absent goals, zero coping, severe distress, frequent relapses) while a rating of 5 represents optimal self-management and recovery (e.g., achieved goals, excellent coping, zero distress, active written crisis plan).
• Scoring Procedures:
  • Composite Total Score: Derived by summing all 15 items (possible range: 15 to 75). Alternatively, a mean score is computed (sum divided by 15, yielding a score between 1.00 and 5.00), which preserves the intuitive 1-to-5 metric.
  • Subscale Scores: When utilizing the empirical 3-factor structure, mean scores are computed separately for the Recovery Subscale (Items 1, 3, 4, 5, 11), the Illness Management Subscale (Items 2, 6, 7, 8, 9, 10, 12, 13), and the Substance Use Subscale (Items 14, 15).
  • Handling Missing Data / Medication Non-prescription: Item 13 contains a skip rule for individuals not prescribed psychiatric pharmacotherapy. When Item 13 is skipped, the composite mean score is computed across the 14 completed items (sum divided by 14). If more than three items are missing, total score calculation is conventionally invalidated.

Permissions & Fee and Test Year

The Illness Management and Recovery (IMR) Scales were published in 2004 by Kim T. Mueser, Susan Gingerich, Michelle P. Salyers, Angela B. McGuire, R. U. Reyes, and H. Cunningham at the New Hampshire-Dartmouth Psychiatric Research Center in collaboration with SAMHSA.

Licensing and Accessibility: The scale is in the public domain as part of the federally sponsored Substance Abuse and Mental Health Services Administration (SAMHSA) Evidence-Based Practices (EBP) KIT series. It may be utilized, reproduced, translated, and administered for clinical, educational, non-profit, and research purposes free of charge without formal royalty fees. Researchers and clinical directors intending to employ the instrument in programmatic evaluations or academic studies are encouraged to cite the original development team and consult the publicly accessible toolkits distributed via SAMHSA and affiliated academic centers.

References

• Anthony, W. A. (1993). Recovery from mental illness: The guiding vision of the mental health service system in the 1990s. Psychosocial Rehabilitation Journal, 16(4), 11–23. https://doi.org/10.1037/h0095655
• Bandura, A. (1997). Self-efficacy: The exercise of control. W. H. Freeman and Company.
• Bullock, W. A., O’Rourke, M., Farrer, E., Breedlove, A., Smith, M. K., & Claggett, A. (2005, August). Evaluation of the Illness Management and Recovery (IMR) Program. Paper presented at the 113th Annual Meeting of the American Psychological Association, Washington, D.C.
• Deegan, P. E. (1988). Recovery: The lived experience of rehabilitation. Psychosocial Rehabilitation Journal, 11(4), 11–19. https://doi.org/10.1037/h0099565
• Färdig, R., Lewander, T., Melin, L., Kobbeba, F., & Fredriksson, M. (2011). Psychometric properties of the Swedish version of the Illness Management and Recovery Scale. Nordic Journal of Psychiatry, 65(3), 173–181. https://doi.org/10.3109/08039488.2010.514943
• Hasson-Ohayon, I., Roe, D., & Kravetz, S. (2007). The psychometric properties of the Illness Management and Recovery Scale: Client and clinician versions. Psychiatry Research, 160(2), 228–235. https://doi.org/10.1016/j.psychres.2007.05.011
• Levitt, A. J., Mueser, K. T., DeGenova, J., Heinz, M., Clark, D., Sloma, H., & Zou, H. (2009). Randomized controlled trial of Illness Management and Recovery in multiple-unit supportive housing. Psychiatric Services, 60(12), 1629–1636. https://doi.org/10.1176/ps.2009.60.12.1629
• Liberman, R. P., Falloon, I. R., & Aitchison, R. A. (1984). Behavioral social skills training for schizophrenia: A review and blueprint. Schizophrenia Bulletin, 10(3), 415–439. https://doi.org/10.1093/schbul/10.3.415
• McGuire, A. B., Salyers, M. P., Rollins, A. L., Bond, G. R., & Mueser, K. T. (2012). The Illness Management and Recovery Scale: Psychometric properties and utility. Psychiatric Services, 63(3), 279–281. https://doi.org/10.1176/appi.ps.201100067
• Mueser, K. T., Gingerich, S., Salyers, M. P., McGuire, A. B., Reyes, R. U., & Cunningham, H. (2004). The Illness Management and Recovery (IMR) Scales (Client and Clinician Versions). New Hampshire-Dartmouth Psychiatric Research Center.
• Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51(3), 390–395. https://doi.org/10.1037/0022-006X.51.3.390
• Salyers, M. P., Godfrey, J. L., Mueser, K. T., & Labriola, S. (2007). Measuring illness management and recovery: Validation of clinician and consumer-rated scales. Psychiatric Services, 58(9), 1215–1220. https://doi.org/10.1176/ps.2007.58.9.1215
• Salyers, M. P., McGuire, A. B., Rollins, A. L., Bond, G. R., Mueser, K. T., & Macy, V. R. (2010). Integrating assertive community treatment and illness management and recovery for consumers with severe mental illness. Community Mental Health Journal, 46(3), 219–229. https://doi.org/10.1007/s10597-009-9238-6
• Sklar, M., Groessl, E. J., O’Connell, M., Davidson, L., & Aarons, G. A. (2012). Factor structure of the Illness Management and Recovery Scale with consumers with severe mental illness. Psychiatric Services, 63(12), 1251–1254. https://doi.org/10.1176/appi.ps.201200155
• Zubin, J., & Spring, B. (1977). Vulnerability: A new view of schizophrenia. Journal of Abnormal Psychology, 86(2), 103–126. https://doi.org/10.1037/0021-843X.86.2.103

13. Items of the Scale (Questionnaire)