Patient Reported Outcomes Measurement Information System

Abstract

The Patient Reported Outcomes Measurement Information System (PROMIS®) represents an internationally recognized, state-of-the-art cooperative research initiative initiated by the National Institutes of Health (NIH) to develop, validate, and standardize modern psychometric instruments for clinical practice and clinical research. Among the core instruments within this framework is the PROMIS Global Health scale (also known as the PROMIS-10 Global Short Form), an efficient 10-item instrument designed to assess generic health-related quality of life (HRQoL) across physical, mental, and social health domains. The scale captures individuals’ evaluations of their overall health status, functioning, emotional well-being, fatigue, pain, and social participation. It yields two primary composite summary scores: Global Physical Health (GPH) and Global Mental Health (GMH). The measure uses 5-point Likert scales for categorical evaluation alongside a 0 to 10 numerical rating scale for pain intensity, which is subsequently recoded into a standardized 5-point metric. In accordance with modern psychometric standards rooted in Item Response Theory (IRT) and classical test theory (CTT), raw scores are transformed into standard T-score distributions (mean = 50, standard deviation = 10) calibrated against the 2000 United States general census population. Psychometric investigations across diverse international general populations and clinical cohorts (e.g., musculoskeletal, oncological, and psychiatric populations) demonstrate robust internal consistency (Cronbach’s alpha and IRT marginal reliability estimates typically ranging from .81 to .89), well-replicated two-factor structural validity, strong convergent validity with established legacy measures (such as the SF-36, EQ-5D, and VR-12), and pronounced responsiveness to therapeutic interventions. This article delivers an exhaustive academic analysis of the PROMIS Global Health instrument, detailing its theoretical architecture, psychometric properties, factor structure, administration criteria, scoring algorithms, and standard scale items.

Keywords

Patient Reported Outcomes Measurement Information System, PROMIS, PROMIS-10, Global Physical Health, Global Mental Health, Health-Related Quality of Life, Item Response Theory, Psychometrics, Factor Analysis, Health Status Measurement, Outcome Assessment, HealthMeasures

Authors

The Patient Reported Outcomes Measurement Information System (PROMIS®) was developed by the PROMIS Cooperative Group, an extensive multi-center academic consortium established in 2004 under the auspices of the NIH Roadmap for Medical Research. Prominent foundational psychometricians and investigators involved in the development, calibration, and governance of PROMIS and the PROMIS Global Health scale include:

• David Cella, Ph.D. — Department of Medical Social Sciences, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA.
• Ron D. Hays, Ph.D. — Division of General Internal Medicine and Health Services Research, David Geffen School of Medicine, University of California, Los Angeles (UCLA), and the RAND Corporation, Santa Monica, California, USA.
• Karon F. Cook, Ph.D. — Department of Medical Social Sciences, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA.
• Darren A. DeWalt, M.D., M.P.H. — Division of General Medicine and Clinical Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA.
• James F. Fries, M.D. — Department of Medicine, Stanford University School of Medicine, Palo Alto, California, USA.
• Caroline B. Terwee, Ph.D. — Department of Epidemiology and Data Science, Amsterdam University Medical Centers, Vrije Universiteit Amsterdam, Amsterdam, Netherlands (Principal Investigator of the Dutch-Flemish PROMIS National Center).
• HealthMeasures — The official administrative and distribution portal for PROMIS, Neuro-QoL, ASCQ-Me, and NIH Toolbox, coordinated via Northwestern University, Evanston/Chicago, Illinois, USA.

Purpose

The fundamental purpose of the PROMIS Global Health scale is to provide a brief, psychometrically sound, universally applicable metric of generic health-related quality of life that transcends disease-specific symptomatology. Historically, outcome assessment in biomedical and psychiatric investigations was dominated by either lengthy generic batteries (such as the Medical Outcomes Study 36-Item Short Form Survey [SF-36] or the Sickness Impact Profile) or narrow, organ-specific functional indices (such as the Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC] or the Oswestry Disability Index [ODI]). While legacy generic tools like the SF-36 provided valuable comparative population norms, their administrative burden often rendered them impractical for high-throughput clinical registries, routine primary care screening, and adaptive clinical trials. Furthermore, legacy scales were developed almost exclusively using Classical Test Theory (CTT), which suffers from sample-dependent item parameter estimations, varying standard errors across the functional continuum, and notable floor and ceiling effects.

PROMIS was designed to overcome these systematic limitations through the development of comprehensive, mathematically calibrated item banks operating under Item Response Theory (IRT). The 10-item PROMIS Global Health short form was specifically selected from these extensive calibrated banks to reflect the conceptual domains central to the World Health Organization (WHO) definition of health: physical, mental, and social well-being. The primary research and clinical objectives of the instrument encompass:

• Comprehensive Health Status Characterization: Quantifying broad perceptions of overall health, quality of life, physical capabilities, psychological distress, fatigue, pain intensity, and social role fulfillment in ten concise items.
• Universal Comparability across Clinical Populations: Serving as a standard generic baseline that allows researchers and epidemiologists to compare disease burden, functional outcomes, and treatment efficacy across divergent clinical entities (e.g., comparing the functional impact of severe rheumatoid arthritis versus clinical depression or congestive heart failure).
• Longitudinal Monitoring and Responsiveness: Detecting meaningful changes in subjective physical and mental function over time in response to pharmacotherapy, surgical intervention, psychological psychotherapy, or physical rehabilitation.
• Minimizing Patient and Administrative Burden: Providing reliable summary scores (Global Physical Health and Global Mental Health) in under two minutes of completion time, mitigating survey fatigue and missing data in epidemiological surveillance and everyday ambulatory clinical workflows.
• Standardized Metric Integration: Facilitating the transformation of raw psychometric scores into a standardized T-score distribution with a fixed population reference (mean = 50, standard deviation = 10), enabling clinicians to immediately interpret an individual’s functional standing relative to the general population.

Psychological Construct

The PROMIS Global Health scale operationalizes generic health-related quality of life through a multidimensional conceptual structure that bifurcates into two higher-order latent constructs: Global Physical Health (GPH) and Global Mental Health (GMH). These dimensions reflect the reciprocal, biocognitive manifestations of disease, function, and wellness.

1. Global Physical Health (GPH)

The Global Physical Health construct represents the individual’s appraisal of their somatic integrity, bodily efficiency, functional independence, and physical symptom burden. In the 10-item scale, GPH is derived from four core observed indicators:

• Overall Self-Rated Physical Health (Item 3): A direct evaluative appraisal capturing the individual’s internal perception of their physiological status and constitutional resilience.
• Physical Functioning / Instrumental Mobility (Item 7): The capacity to execute activities of daily living (ADLs) that require somatic exertion, including ambulation, stair climbing, lifting, and domestic tasks (e.g., carrying groceries or moving furniture). Deficits in this domain indicate physical impairment or mobility restriction.
• Pain Intensity (Item 10): The average perceived severity of somatic nociceptive or neuropathic sensations experienced over the recall interval, assessed using an 11-point numeric rating scale (recoded into a 5-point ordinal scale). Pain serves as a direct somatic indicator of physical pathology and physiological distress.
• Fatigue (Item 9, recoded/scored within GPH in select parameterizations, or evaluated alongside Item 8): In the standard scoring algorithm formulated by Hays et al. (2009), Item 8 (frequency of emotional problems) functions as an inversely weighted indicator within physical health, or Item 9 (average fatigue) anchors physical depletion. Specifically, in the final validated scoring algorithm, GPH comprises Item 3 (physical health rating), Item 7 (everyday physical activities), Item 8 (emotional distress recoded/weighted), and recoded Item 10 (pain rating). This reflects the somatic manifestations and physical functional consequences of systemic stress.

2. Global Mental Health (GMH)

The Global Mental Health construct encapsulates cognitive clarity, emotional equilibrium, affective regulation, and perceived quality of social interactions. It represents subjective psychological well-being and social role participation through four key indicators:

• Overall Quality of Life (Item 2): An overarching, evaluative assessment of personal existence, life satisfaction, and psychological congruence between personal aspirations and lived experience.
• Mental and Cognitive Health (Item 4): A targeted self-evaluation of mental functioning, including emotional mood stability, executive processing, and cognitive coherence.
• Satisfaction with Social Activities and Relationships (Item 5): The subjective gratification, fulfillment, and emotional security derived from personal relationships, family connections, and informal social networks.
• Performance of Social Roles and Duties (Item 6): The perceived efficacy and competence with which an individual executes typical occupational, domestic, familial, and community-based roles. This links psychological efficacy to real-world behavioral functioning.

Items 1 (general self-rated health) and 9 (fatigue rating) provide valuable exploratory and clinical information regarding overall constitutional vitality. Item 1 represents an independent single-item global health benchmark frequently utilized in longitudinal public health registries (e.g., the National Health and Nutrition Examination Survey [NHANES]).

Theoretical Framework

The development of PROMIS is grounded in modern measurement theory, primarily drawing upon Item Response Theory (IRT) and the World Health Organization’s Biopsychosocial Model of Health. Historically, subjective outcome measurement relied almost entirely on Classical Test Theory, which assumes that an observed score is simply the sum of a true score and unsystematic random measurement error ($X = T + E$). However, CTT exhibits notable psychometric limitations: scale characteristics are strictly sample-dependent, standard errors of measurement are treated as uniform across the entire trait continuum, and comparison across different forms requires extensive cross-walking calibrations.

1. The WHO Biopsychosocial Model

In 1948, the World Health Organization established a benchmark definition: “Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.” PROMIS operationalizes this definition by creating a unified measurement architecture organized into three primary domains: Physical Health, Mental Health, and Social Health. Sub-domains within this architecture are structured hierarchically, ensuring that the PROMIS Global Health scale functions not as an arbitrary amalgamation of disjointed symptoms, but as an integrated representation of somatic, psychological, and social functioning.

2. Modern Item Response Theory (IRT) Architecture

Unlike traditional measurement tools, the items constituting the PROMIS Global Health scale were selected and calibrated using modern polytomous IRT models, specifically Samejima’s Graded Response Model (GRM). In the GRM, the probability $P_{jk}( h\eta)$ that an individual with a given latent health trait standing $\theta$ endorses response category $k$ or higher for item $j$ is formulated as:

$$P^*_{jk}(\theta) = \frac{\exp\left(a_j(\theta – b_{jk})\right)}{1 + \exp\left(a_j(\theta – b_{jk})\right)}$$

where $a_j$ represents the item discrimination parameter (indicating how sharply the item differentiates between individuals with varying levels of health status), and $b_{jk}$ denotes the category threshold parameter representing the trait level $\theta$ at which an individual has a 50% probability of endorsing category $k$ or higher. Through this IRT calibration, PROMIS items were evaluated for their information functions across the latent trait distribution. This ensured that the 10 selected items provide optimal measurement precision across the spectrum of general and clinical populations, avoiding ceiling effects in healthy individuals and floor effects in populations with severe chronic illnesses.

Validity

The psychometric validity of the PROMIS Global Health scale has been established across diverse populations, clinical subspecialties, and international language adaptations.

1. Construct and Structural Validity

Construct validity was initially confirmed in the foundational PROMIS validation cohort of $N = 21,133$ individuals representative of the United States general population (Hays et al., 2009). Exploratory and confirmatory factor analyses confirmed that a two-dimensional construct—Global Physical Health (GPH) and Global Mental Health (GMH)—provided an optimal fit to the data. Inter-factor correlations between GPH and GMH generally range between $r = .55$ and $r = .70$, confirming that while physical and psychological domains share substantive variance (reflecting overall somatic-affective well-being), they remain empirically distinct constructs that must not be collapsed into an undifferentiated single score.

2. Convergent and Discriminant Validity

Convergent validity has been established by cross-calibrating PROMIS Global Health against recognized legacy metrics:

• SF-36 / RAND-36: The PROMIS Global Physical Health score demonstrates high correlations with the SF-36 Physical Component Summary (PCS) ($r = .75$ to $.82$) and the Physical Functioning subscale ($r = .78$). Conversely, the PROMIS Global Mental Health score correlates strongly with the SF-36 Mental Component Summary (MCS) ($r = .74$ to $.80$) and the Mental Health/Vitality subscales ($r = .71$ to $.76$).
• EQ-5D: PROMIS GPH correlates robustly with the EQ-5D-3L and EQ-5D-5L index utility scores ($r = .65$ to $.75$), while GMH correlates selectively with the EQ-5D anxiety/depression dimension.
• Discriminant Evidence: Discriminant validity is evidenced by the lower correlation observed between PROMIS Global Physical Health and legacy mental health scales (e.g., correlations between GPH and the PHQ-9 depression scale typically fall around $r = -.35$ to $-.42$), contrasted with the strong inverse correlation between PROMIS GMH and PHQ-9 ($r = -.68$ to $-.74$).

3. Known-Groups and Predictive Validity

PROMIS Global Health scores differentiate between cohorts stratified by chronic disease burden, age, and socioeconomic variables. Studies show that individuals with multiple comorbid medical conditions (e.g., chronic obstructive pulmonary disease, type II diabetes, active rheumatoid arthritis) score 1.0 to 1.5 standard deviations lower on the GPH metric compared to age-matched healthy peers. Furthermore, baseline PROMIS GPH and GMH scores predict postoperative recovery trajectories, one-year mortality, surgical revision rates, readmission rates, and long-term vocational disability in orthopedic, cardiothoracic, and neurological cohorts.

Reliability

The PROMIS Global Health scale exhibits high reliability across both classical psychometric indices and IRT-derived measurement precision metrics:

1. Internal Consistency

In the primary general population calibration sample ($N = 21,133$), Classical Test Theory internal consistency evaluations revealed high reliability coefficients for both summary composite scales:

• Global Physical Health (GPH): Cronbach’s $\alpha = .81$, McDonald’s $\omega = .82$.
• Global Mental Health (GMH): Cronbach’s $\alpha = .86$, McDonald’s $\omega = .87$.

Subsequent multi-center cross-cultural validation trials—such as the Dutch-Flemish general population validation conducted by Terwee et al. (2014) ($N = 2,012$) and various German, Asian, and Latin American validation studies—have consistently replicated internal consistency values, with Cronbach’s alpha coefficients consistently exceeding $.80$ across clinical cohorts and healthy controls.

2. IRT Marginal Reliability and Information Curves

IRT information curves demonstrate that PROMIS GPH and GMH provide high measurement reliability (reliability $ge .80$, corresponding to an information value $ge 5.0$) across an expansive range of the latent health continuum (typically between $-2.5 < theta < +1.5$ standard deviations). This confirms robust precision among patients suffering from mild, moderate, and severe functional deficits.

3. Test-Retest Reliability

In longitudinal evaluations of clinically stable populations over intervals ranging from 7 to 14 days, the intraclass correlation coefficients (ICC) consistently range from $.83$ to $.89$ for Global Physical Health and from $.80$ to $.87$ for Global Mental Health. These figures confirm high test-retest reproducibility in the absence of clinical intervention.

Factor Analysis

The structural dimensionality of the PROMIS Global Health item pool has been rigorously examined through both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA), demonstrating a correlated two-factor structure.

1. Confirmatory Factor Analytic (CFA) Fit Indices

During the primary development phase, CFA models contrasting a unidimensional model against a two-factor correlated model demonstrated that the two-factor model (GPH and GMH) fit the data well. Exemplary model fit indices reported across the empirical literature (e.g., Hays et al., 2009; Terwee et al., 2014) include:

• Comparative Fit Index (CFI): $> .96$ (values above $.95$ indicate excellent fit).
• Tucker-Lewis Index (TLI): $> .95$.
• Root Mean Square Error of Approximation (RMSEA): $.048$ to $.062$ (90% Confidence Interval: $[.041, .068]$), meeting accepted criteria for good model fit ($< .06$).
• Standardized Root Mean Square Residual (SRMR): $< .035$.

2. Standardized Factor Loadings

Standardized factor loadings under a correlated two-factor structural equation model indicate that the observed indicators load substantially on their intended latent dimensions:

• Global Physical Health (GPH):
  • Item 3 (Overall Physical Health): $lambda = .84 – .89$
  • Item 7 (Everyday Physical Activities): $lambda = .77 – .82$
  • Item 10 (Recoded Pain Intensity): $lambda = .61 – .69$
  • Item 8 (Emotional distress, cross-loaded/recoded indicator): $lambda = .48 – .56$
• Global Mental Health (GMH):
  • Item 2 (Overall Quality of Life): $lambda = .79 – .85$
  • Item 4 (Overall Mental Health): $lambda = .83 – .88$
  • Item 5 (Satisfaction with Social Relationships): $lambda = .71 – .78$
  • Item 6 (Performance of Social Roles): $lambda = .74 – .81$

Item 1 (General Health rating) and Item 9 (Fatigue rating) typically demonstrate cross-domain loadings onto both physical and mental latent vectors, consistent with their systemic biological nature. In the final, formalized scoring models, Item 1 is retained primarily as a stand-alone universal health metric, and the remaining indicators are cleanly partitioned into GPH and GMH.

Instrument / Measurement Tool

• Test Type: Generic Patient-Reported Outcome Measure (PROM); Health-Related Quality of Life (HRQoL) Short Form.
• Administration Format: Self-administered paper-and-pencil questionnaire, tablet/computerized web interface, or clinician-assisted interview. Also available via Computerized Adaptive Testing (CAT) through compatible clinical platforms.
• Total Item Count: 10 items.
• Target Population: Adults (18 years and older); pediatric adaptations (PROMIS Pediatric Global Health) exist for children and adolescents aged 8–17 years.
• Completion Time: Approximately 1.5 to 3 minutes.
• Response Scales:
  • 5-point Likert scale (Items 1–3, 6, 9): 1 = Poor, 2 = Fair, 3 = Good, 4 = Very good, 5 = Excellent
  • 5-point Likert scale (Item 4): 1 = None, 2 = Mild, 3 = Moderate, 4 = Severe, 5 = Very severe
  • 5-point Likert scale (Items 5, 7, 8): 1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, 5 = Always
  • 11-point Numeric Rating Scale (Item 10): 0 = No pain to 10 = Worst pain imaginable
• Reverse Scoring and Recoding Protocols:
  • Item 8 (Emotional distress): Must be reverse-coded prior to summation so that higher scores represent superior health ($5 = \text{Never}$, $4 = \text{Rarely}$, $3 = \text{Some\times}$, $2 = \text{Often}$, $1 = \text{Always}$).
  • Item 10 (Pain rating 0–10): Must be recoded from its 11-point continuous scale into a 5-point category scale as follows:
    • Score 0 recoded as 5 (None/No pain)
    • Scores 1 to 3 recoded as 4 (Mild pain)
    • Scores 4 to 6 recoded as 3 (Moderate pain)
    • Scores 7 to 9 recoded as 2 (Severe pain)
    • Score 10 recoded as 1 (Worst pain imaginable)
• Scoring and Transformation Mechanics:
  • Global Physical Health (GPH) Raw Score: Sum of Item 3, Item 7, reverse-coded Item 8, and recoded Item 10 (Raw score range: 4 to 20).
  • Global Mental Health (GMH) Raw Score: Sum of Item 2, Item 4, Item 5, and Item 6 (Raw score range: 4 to 20).
  • Standardized Metric Conversion: Raw composite scores are converted into standardized T-scores using official PROMIS scoring conversion tables. The T-score distribution is standardized to have a mean of 50 and a standard deviation (SD) of 10 in the United States general census population. A score of 50 reflects normal general population function; scores of 40 or below (one full standard deviation lower) indicate clinically meaningful impairment.

Permissions & Fee and Test Year

The Patient Reported Outcomes Measurement Information System (PROMIS) initiative began in 2004 under the NIH Roadmap for Medical Research. The PROMIS Global Health 10-item scale was formally validated and published by Hays et al. in 2009, followed by international validation initiatives, including the Dutch-Flemish version by Terwee et al. in 2014.

Licensing and Accessibility: PROMIS instruments are public goods. They are freely available for non-commercial academic research, public health tracking, and routine clinical care via the official HealthMeasures portal. HealthMeasures is the official repository funded to maintain, host, and distribute PROMIS, Neuro-QoL, ASCQ-Me, and NIH Toolbox instruments. While paper short forms are freely accessible, commercial integration into Electronic Health Record (EHR) software platforms, proprietary clinical trial registries, or commercial digital health solutions may require formal administrative registration, licensing agreements, and associated distribution fees administered via the PROMIS Health Organization (PHO) and Northwestern University.

References

• Cella, D., Yount, S., Rothrock, N., Gershon, R., Cook, K., Reeve, B., Ader, D., Fries, J. F., Bruce, B., Rose, M., & PROMIS Cooperative Group. (2007). The Patient-Reported Outcomes Measurement Information System (PROMIS): Progress of an NIH Roadmap cooperative group during its first two years. Medical Care, 45(5 Suppl 1), S3–S11. https://doi.org/10.1097/01.mlr.0000258615.42478.55
• Hays, R. D., Bjorner, J. B., Revicki, D. A., Spritzer, K. L., & Cella, D. (2009). Development of physical and mental health summary scores from the Patient-Reported Outcomes Measurement Information System (PROMIS) global items. Quality of Life Research, 18(7), 873–880. https://doi.org/10.1007/s11136-009-9496-9
• Hays, R. D., Spritzer, K. L., Schalet, B. D., & Cella, D. (2017). PROMIS®-29 v2.0 profile physical and mental health summary scores. Quality of Life Research, 27(7), 1885–1891. https://doi.org/10.1007/s11136-018-1842-3
• Reeve, B. B., Hays, R. D., Bjorner, J. B., Cook, K. F., Crane, P. K., Teresi, J. A., Thissen, D., Revicki, D. A., Weiss, D. J., Hambleton, R. K., Liu, H., Markward, N. J., Lai, J.-S., & Cella, D. (2007). Psychometric evaluation and calibration of health-related quality of life item banks: Plans for the Patient-Reported Outcomes Measurement Information System (PROMIS). Medical Care, 45(5 Suppl 1), S22–S31. https://doi.org/10.1097/01.mlr.0000260475.24248.04
• Rothrock, N. E., Hays, R. D., Spritzer, K., Yount, S. E., Riley, W., & Cella, D. (2010). Relative reliability of short-form and computer-adaptive testing implementation of PROMIS adult self-reported outcome measures. Quality of Life Research, 19(8), 1079–1087. https://doi.org/10.1007/s11136-010-9666-0
• Terwee, C. B., Roorda, L. D., de Vet, H. C. W., Peacock, C. A., Luijten, M. A. J., & Hays, R. D. (2014). Dutch-Flemish translation and psychometric evaluation of the PROMIS Global Health scale. Quality of Life Research, 23(Suppl 1), 35–36. https://doi.org/10.1007/s11136-014-0803-2

Items of the Scale