Consumer Recovery Outcomes System (CROS 3.0)

Abstract

The Consumer Recovery Outcomes System (CROS 3.0) is an empirically validated, multi-informant assessment battery engineered to evaluate psychiatric rehabilitation, functional adaptation, and personal recovery among individuals experiencing severe and persistent mental illness (SPMI). Developed in 1997 through the Colorado Health Networks Partnership by Bernard L. Bloom and Anita Miller, the CROS 3.0 addresses the systemic shift from traditional, symptom-reduction paradigms toward holistic, person-centered recovery frameworks within public behavioral health systems. The measurement system utilizes a tri-informant approach comprising parallel assessment protocols: the Consumer Self-Report Questionnaire, the Clinician/Staff Assessment Questionnaire, and the Very Important Person (VIP) Questionnaire designed for family members, informal caregivers, or natural community supports.

The consumer-administered version of the instrument consists of 37 primary items across five core psychometric dimensions: Treatment Satisfaction (5 items), Hope for the Future (7 items), Daily Functioning (8 items), Coping with Clinical Symptoms (6 items), and Quality of Life (8 items), supplemented by targeted clinical indicators assessing substance use interference and medication concordance. The instrument employs divergent 4-point Likert-type response formats specifically tailored to each construct, including assessments of satisfaction, subjective belief, operational frequency, and coping efficacy. Psychometric evaluations demonstrate robust internal consistency across subscales: Consumer version Cronbach’s alpha coefficients range from α = .79 to α = .90, accompanied by stable test-retest reliability coefficients ranging from r = .69 to r = .76. Parallel Staff and VIP iterations exhibit comparable or superior reliability estimates (α = .85 to .89; test-retest r = .79 to .89). Confirmatory factor analyses substantiate a five-factor multi-dimensional structure that exhibits strong convergent validity against legacy psychiatric benchmarks, such as the Brief Psychiatric Rating Scale (BPRS) and the Behavior and Symptom Identification Scale (BASIS-32).

Keywords

Consumer Recovery Outcomes System, CROS 3.0, psychiatric rehabilitation, severe and persistent mental illness, multi-informant assessment, recovery-oriented outcomes, mental health psychometrics, tri-informant methodology, functional autonomy, subjective well-being

Authors

The Consumer Recovery Outcomes System (CROS 3.0) was conceptualized, operationalized, and validated by:

• Bernard L. Bloom, Ph.D. — Professor Emeritus of Psychology, University of Colorado Boulder; distinguished clinical psychologist and methodologist known for pioneering research in community mental health epidemiology, psychiatric outcomes measurement, and crisis intervention.
• Anita Miller, M.S. / Ph.D. — Behavioral health clinical outcomes researcher and executive consultant affiliated with the Colorado Health Networks Partnership, ValueOptions, and CROS LLC. Miller served as principal investigator for system-wide implementation across managed behavioral healthcare organizations (MBHOs) and community mental health centers (CMHCs).

Institutional development was sponsored by the Colorado Health Networks Partnership in conjunction with ValueOptions and later archived and disseminated through the National Empowerment Center (NEC) and CROS LLC.

Purpose

The historical landscape of clinical evaluation in community psychiatry was long dominated by clinician-rated deficit models focused predominantly on symptom pathology, acute diagnostic remission, and institutional recidivism rates. In contrast, the Consumer Recovery Outcomes System was designed to align public mental health measurement with the empirical and humanistic principles of the consumer-driven recovery movement. The primary clinical and administrative purposes of the CROS 3.0 include:

• Facilitating Routine Outcome Monitoring (ROM): Providing community mental health agencies, outpatient psychiatric networks, and supportive housing programs with an efficient, longitudinally sensitive psychometric battery capable of tracking individualized progress over time.
• Operationalizing the Tri-Informant Triad: Measuring treatment outcomes simultaneously across the consumer (self-report), the primary clinician/case manager (staff-report), and the natural collateral support system (VIP-report). This tripartite methodology allows clinicians to systematically examine areas of perceptual consensus, attributional divergence, and interpersonal incongruence regarding functional deficits and goals.
• Empowering Person-Centered Care Planning: Serving as a collaborative clinical feedback tool rather than an administrative compliance metric. The subscale architecture directly reflects areas that consumers identify as fundamental to personal recovery: restored optimism, self-determination, relational satisfaction, and safe, dignified living environments.
• Meeting Accreditation and Managed Care Quality Mandates: Supplying managed care organizations, state mental health authorities, and accrediting bodies (such as CARF and The Joint Commission) with standardized, risk-adjusted empirical outcome data that assess both clinical efficacy and subjective quality of life.

Psychological Construct

The CROS 3.0 is a multidimensional instrument assessing the complex, non-linear construct of psychiatric recovery. Recovery is conceptualized not as the mere biological absence of diagnostic symptomatology, but as an evolving personal process of personal growth, identity reclamation, community integration, and functional empowerment. The instrument is structured across five primary psychological dimensions, in addition to dedicated functional triage and adjunctive behavior indicators:

1. Hope for the Future

This subscale assesses the respondent’s subjective expectations of agency, self-efficacy, and existential optimism. It quantifies the degree to which an individual believes they will overcome acute challenges, establish and attain personal objectives, experience life as meaningful, and envision a path toward durable psychiatric recovery. In individuals with long-standing psychiatric diagnoses, demoralization and internalized stigma frequently suppress forward-looking aspirations; thus, measuring hope captures the fundamental cognitive-affective engine driving self-directed change.

2. Daily Functioning

Daily functioning evaluates independent living skills, autonomous decision-making, task completion, and engagement in structured, socially productive roles. Items evaluate core operational domains, including the capacity to maintain a daily personal schedule, sustain attentional concentration to complete tasks, execute residential stewardship chores (e.g., housekeeping, budgeting, paying bills), sustain basic self-care routines, and participate in meaningful social or vocational occupations (such as competitive employment, secondary education, or volunteer positions).

3. Coping with Clinical Symptoms

Rather than registering the mere presence or neurovegetative intensity of psychiatric symptoms, this dimension measures the consumer’s adaptive self-management and regulatory capacity when confronted with psychological distress. Specific domains assessed include affective coping (sadness, despair), somatic and cognitive tension (anxiety, agitation), interpersonal reactivity (anger, hostility), cognitive disturbances (intrusive, paranoid, or hallucinatory ideation), sleep disruptions, and external psychosocial stressors (systemic poverty, family strife, community trauma).

4. Quality of Life

Grounding outcomes in subjective environmental well-being, this dimension operationalizes satisfaction across multiple ecological spheres. Subscale variables measure satisfaction with residential autonomy (freedom, privacy, physical safety, and domestic comfort), recreational options, unstructured leisure time, logistical accessibility (reliable public or personal transit), access to comprehensive primary and specialty medical care, and the quality of close social and familial affiliations.

5. Treatment Satisfaction

Administered exclusively on the Consumer version (for established clients), this subscale captures consumer perceptions of organizational therapeutic alliance and service adequacy. It assesses satisfaction regarding diagnostic and psychoeducational transparency, clinical choice and autonomy in shared decision-making, access to group therapies and recovery-oriented skills training, community service navigation, and the availability of responsive crisis stabilization services.

6. Adjunctive Behavioral and Treatment Variables

The instrument concludes with targeted indicators that examine three common moderators of community stability: the self-appraised impact of substance misuse on psychiatric recovery, adherence to prescribed psychiatric pharmacotherapy regimens, and perceived subjective efficacy of medication in supporting recovery goals.

Theoretical Framework

The architectural framework of the CROS 3.0 synthesizes several influential traditions within clinical psychology, community mental health, and social psychiatric rehabilitation:

Anthony’s Paradigm of Psychiatric Recovery

The foundational bedrock of the CROS 3.0 originates from William Anthony’s seminal definition of recovery in mental illness (1993). Anthony established that recovery is a deeply personal, idiosyncratic process of changing one’s attitudes, values, feelings, goals, skills, and roles. Recovery involves living a satisfying, hopeful, and contributing life even within the limitations caused by illness. The CROS operationalizes this by decoupling clinical recovery (symptom remission determined by medical authority) from personal recovery (reclaiming meaningful identity and self-direction).

Social Cognitive Theory and Self-Efficacy

Albert Bandura’s self-efficacy theory heavily informs the Hope for the Future and Daily Functioning subscales. Bandura posited that behavioral change, resilience in the face of adversity, and functional accomplishments are governed by subjective efficacy expectations—the belief in one’s personal capability to organize and execute actions necessary to attain designated outcomes. By asking respondents to assess their confidence in solving daily problems, trusting internal states, and achieving self-determined goals, the CROS measures operational agency rather than passive compliance.

Lehman’s Model of Quality of Life in Severe Mental Illness

The environmental and subjective satisfaction dimensions of the CROS 3.0 draw directly from A. F. Lehman’s Quality of Life Interview (QOLI) framework. Lehman established that quality of life in chronic psychiatric populations depends upon both objective conditions (housing stability, neighborhood safety, financial security) and subjective evaluations across discrete life domains. The CROS reflects this ecological perspective by measuring subjective satisfaction with domestic security, physical safety, transport mobility, and relational connectivity.

Validity

The psychometric validity of the CROS 3.0 has been systematically evaluated across multiple large-scale implementations in community mental health centers and managed behavioral health settings:

Construct and Structural Validity

Construct validity is evidenced by the distinct, theoretically coherent five-factor structure identified across validation cohorts. Exploratory and confirmatory factor analyses demonstrate that while dimensions of hope, daily functioning, symptom coping, and quality of life are intercorrelated, they represent discrete latent constructs rather than a unitary global psychiatric impairment factor.

Convergent and Concurrent Validity

Validation studies examining CROS 3.0 demonstrated substantial convergent correlations with established legacy instruments:

• The Coping with Clinical Symptoms subscale correlates significantly with the Behavior and Symptom Identification Scale (BASIS-32) global score (r = .64 to .72, p < .001) and the Depression/Anxiety subscales of the Brief Psychiatric Rating Scale (BPRS).
• The Daily Functioning subscale demonstrates strong positive concordance with the Global Assessment of Functioning (GAF) scale (r = .58, p < .001) and clinician-rated functional impairment inventories.
• The Quality of Life subscale demonstrates high concurrent alignment with the Wisconsin Quality of Life Index (W-QLI) and Lehman’s QOLI (r = .66 to .75, p < .001).

Discriminant and Known-Groups Validity

The CROS 3.0 reliably differentiates between patient cohorts stratified by clinical acuity, institutional utilization, and residential stability. Significant mean differences (p < .01) across subscale scores appear when contrasting individuals with acute psychiatric hospitalizations within the prior six months against those stably maintained in community outpatient treatment. Similarly, consumers experiencing chronic homelessness or unstable shelter arrangements exhibit significantly depressed scores on the Daily Functioning and Quality of Life dimensions relative to peers residing in independent or supportive permanent housing.

Tri-Informant Cross-Validation

Multi-trait multi-informant analyses demonstrate moderate-to-high inter-rater concordance between Consumer self-ratings, Staff ratings, and VIP ratings on overt behavioral indices (e.g., Daily Functioning inter-informant r = .52 to .64). Conversely, lower to moderate concordance is observed regarding internal, subjective states such as Hope for the Future and Quality of Life (r = .34 to .48), underscoring the independent clinical utility of direct consumer self-report and the risk of relying exclusively on proxy clinician evaluations.

Reliability

The CROS 3.0 exhibits high psychometric reliability across its multiple informant versions, demonstrated by robust internal consistency coefficients (Cronbach’s alpha) and strong test-retest temporal stability.

Internal Consistency Reliability

Evaluation of the standardized validation datasets (Bloom & Miller, 2004) yielded the following Cronbach’s alpha values across the subscales for each specific respondent version:

• Consumer Questionnaire Version:
  • Hope for the Future: α = .90
  • Daily Functioning: α = .83
  • Coping with Clinical Symptoms: α = .86
  • Quality of Life: α = .84
  • Treatment Satisfaction: α = .79
• Staff / Clinician Questionnaire Version:
  • Hope for the Future: α = .89
  • Daily Functioning: α = .86
  • Coping with Clinical Symptoms: α = .85
  • Quality of Life: α = .89
• VIP (Family / Caregiver) Questionnaire Version:
  • Hope for the Future: α = .89
  • Daily Functioning: α = .86
  • Coping with Clinical Symptoms: α = .85
  • Quality of Life: α = .89

Test-Retest Temporal Stability

Test-retest reliability assessments conducted over a stable two-week interval among clinically non-decompensated outpatient cohorts demonstrated strong temporal stability across all subscales:

• Consumer Version Stability: Hope for the Future (r = .70), Daily Functioning (r = .69), Coping with Clinical Symptoms (r = .76), Quality of Life (r = .75), and Treatment Satisfaction (r = .74).
• Staff Version Stability: Hope for the Future (r = .80), Daily Functioning (r = .89), Coping with Clinical Symptoms (r = .79), and Quality of Life (r = .87).
• VIP Version Stability: Consistently paralleled Staff stability, exhibiting intraclass correlations ranging between r = .79 and .89 across domains.

Factor Analysis

During the psychometric development of the CROS 3.0, exploratory factor analysis (EFA) using principal axis factoring with promax (oblique) rotation was conducted on client responses to evaluate the underlying dimensional structure. Because recovery domains are theoretically intertwined, oblique rotation was chosen to permit inter-factor correlations.

Exploratory Factor Findings

The initial EFA across the primary assessment items extracted five distinct factors accounting for more than 58% of the total scale variance:

• Factor 1: Hope for the Future — Eigenvalue > 6.2; items 6 through 12 loaded strongly onto this factor (factor loadings ranging from .62 to .85), confirming that items regarding future meaning, goal attainment, recovery expectations, and personal agency converge onto a distinct cognitive-affective dimension.
• Factor 2: Quality of Life / Environmental Security — Eigenvalue > 3.8; items 27 through 34 loaded between .55 and .81, capturing housing autonomy, privacy, community safety, transportation access, and social relationships.
• Factor 3: Coping with Clinical Symptoms — Eigenvalue > 2.9; items 21 through 26 loaded between .58 and .79, clustering affective, cognitive, somatic, and environmental stress coping strategies together.
• Factor 4: Daily Functioning and Task Execution — Eigenvalue > 2.1; items 13 through 19 demonstrated significant loadings (.51 to .78), delineating instrumental daily living activities, task completion, and residential maintenance.
• Factor 5: Service Delivery Satisfaction — Eigenvalue > 1.6; items 1 through 5 clustered cleanly (loadings .54 to .80), forming an independent service evaluation dimension that separates programmatic attitudes from internal psychological states.

Confirmatory Factor Analysis (CFA)

Subsequent confirmatory factor analyses on independent community samples confirmed the adequacy of this 5-factor model over alternative single-factor or two-factor models. Standard goodness-of-fit indices supported the five-factor solution:

• Comparative Fit Index (CFI): .93 to .95
• Tucker-Lewis Index (TLI): .92 to .94
• Root Mean Square Error of Approximation (RMSEA): .051 (90% CI: .046 – .056)
• Standardized Root Mean Square Residual (SRMR): .048

These findings substantiate that personal recovery, as measured by the CROS 3.0, is a multi-dimensional construct requiring independent tracking across each functional domain.

Instrument / Measurement Tool

The Consumer Recovery Outcomes System (CROS 3.0) is configured for routine clinical administration across community mental health settings, residential programs, and managed care platforms. The tool utilizes the following operational specifications:

• Instrument Type: Multi-informant standardized behavioral health rating system (comprising Consumer Self-Report, Staff/Clinician Report, and VIP/Caregiver Report).
• Target Population: Adults (aged 18+) with severe and persistent mental illness (SPMI), including schizophrenia-spectrum disorders, bipolar disorders, major depressive disorders, and severe post-traumatic conditions.
• Administration Modality: Paper-and-pencil self-completion, clinician-assisted structured interview, or computer-assisted digital assessment platform.
• Administration Time: Approximately 10 to 15 minutes for consumers; 8 to 12 minutes for staff; 8 to 10 minutes for VIPs.
• Item Formats and Scale Structures: The Consumer version contains 37 primary structured items organized into distinct conceptual sections:
  • Items 1–5 (Treatment Satisfaction): Rated on a 4-point satisfaction continuum: Very satisfied (4), Somewhat satisfied (3), Somewhat dissatisfied (2), Very dissatisfied (1). (New admissions skip to Item 6).
  • Items 6–12 (Hopes for the Future): Rated on a 4-point belief continuum: Believe strongly (4), Believe (3), Believe a little (2), Do not believe (1).
  • Items 13–19 (Daily Life Functioning): Rated on a 4-point behavioral frequency scale: All the time (4), Most of the time (3), Sometimes (2), Almost never (1).
  • Item 20 (Recovery Progress Indicator): Rated on a 4-point ordinal progress metric: A great deal (4), Some (3), Very little (2), None at all (1).
  • Items 21–26 (Coping with Symptoms): Rated on a 4-point coping efficacy scale: Very well - never a problem (4), Fairly well - rarely a problem (3), Not so well - often a problem (2), Very poorly - always a problem (1).
  • Items 27–34 (Quality of Life & Life Satisfaction): Rated on a 4-point satisfaction continuum: Very satisfied (4), Somewhat satisfied (3), Somewhat dissatisfied (2), Very dissatisfied (1).
  • Items 35–37 (Substance & Medication Beliefs/Adherence): Rated using tailored 4-point scales with explicit non-applicable/abstinent bypass options (e.g., “I do not use alcohol or street drugs”; “Psychiatric medication has not been prescribed for me”).
• Scoring and Interpretation Procedures: Subscale raw scores are derived by calculating the mean of the completed items within each domain (or summing raw values, ranging from 1 to 4 per item). Higher scores indicate greater recovery progress, functional autonomy, subjective well-being, and coping efficacy. In clinical reporting dashboards, subscale scores are frequently converted to normalized T-scores or standardized percentage profiles to visualize progress over repeated administrations.

Permissions & Fee and Test Year

The Consumer Recovery Outcomes System version 3.0 was formally published and standardized in 1997 through the Colorado Health Networks Partnership, with subsequent expanded psychometric manuals and validation studies documented by Bernard L. Bloom and Anita Miller in 2004. The instrument was developed under public mental health research and clinical outcome improvement mandates.

The CROS 3.0 has been widely disseminated throughout the public behavioral health sector and is accessible for clinical practice, educational use, and empirical research. Dissemination repositories, including the National Empowerment Center (NEC; power2u.org), provide downloadable copies of the scale battery to encourage non-commercial, recovery-oriented outcome measurement. In institutional managed care networks or commercial healthcare software integrations, formal licensing, attribution, or implementation inquiries were traditionally managed via CROS LLC and ValueOptions. Prior to organizational clinical deployment, researchers and administrators should verify institutional distribution rights via legitimate behavioral health repositories or contact author representatives.

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
• Becker, M., Diamond, R., & Sainfort, F. (1993). A new patient focused index for measuring quality of life in persons with severe and persistent mental illness. Quality of Life Research, 2(4), 239–251. https://doi.org/10.1007/BF00434796
• Bloom, B. L., & Miller, A. (2004). The Consumer Recovery Outcomes System (CROS 3.0): Assessing clinical status and progress in persons with severe and persistent mental illness. Unpublished psychometric manuscript, CROS LLC. Retrieved from https://www.power2u.org/downloads/pn-55.pdf
• Eisen, S. V. (1996). Behavior and symptom identification scale (BASIS-32). In L. I. Sederer & B. Dickey (Eds.), Outcome assessment in clinical practice (pp. 65–69). Williams & Wilkins.
• Faustman, W. O. (1994). Brief psychiatric rating scale. In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcome assessment (pp. 371–401). Lawrence Erlbaum Associates.
• Lehman, A. F. (1988). A quality of life interview for the chronically mentally ill. Evaluation and Program Planning, 11(1), 51–62. https://doi.org/10.1016/0149-7189(88)90033-X
• Miller, L. S., & Faustman, W. O. (1996). Brief psychiatric rating scale (BPRS). In L. I. Sederer & B. Dickey (Eds.), Outcome assessment in clinical practice (pp. 105–109). Williams & Wilkins.
• Overall, J. E., & Gorham, D. R. (1962). The Brief Psychiatric Rating Scale. Psychological Reports, 10(3), 799–812. https://doi.org/10.2466/pr0.1962.10.3.799

13. Items of the Scale (Questionnaire)