Evaluative Frailty Index for Physical Activity

1. Abstract

The Evaluative Frailty Index for Physical Activity (EFIP) is a comprehensive, multidimensional clinimetric instrument developed to screen for and longitudinally evaluate frailty in older adults, specifically tailored to capture clinically meaningful changes induced by physical activity and targeted geriatric rehabilitation interventions. Originally formulated by de Vries, Staal, Olde Rikkert, and Nijhuis-van der Sanden in 2013 at the Radboud University Medical Center, the EFIP operationalizes the cumulative deficit model of frailty pioneered by Mitnitski and Rockwood. Unlike static or purely phenotype-driven screening tools (such as Fried’s physical phenotype), the EFIP was constructed with an explicitly evaluative psychometric architecture, capable of reflecting nuanced shifts in functional reserve across time.

The instrument assesses 50 distinct deficits distributed across four interconnected operational domains: physical functioning and mobility, psychological/cognitive health, social functioning, and general systemic health status (including polypharmacy and nutritional markers). Respondents or clinical raters evaluate deficits experienced within a designated two-week recall window. The scoring architecture utilizes a standardized deficit weighting system where each item is scored dichotomously (0 = deficit absent, 1 = deficit present) or trichotomously (0 = absent, 0.5 = intermediate impairment/partial assistance, 1.0 = full deficit/complete inability). The aggregate frailty index score is computed mathematically as a continuous ratio ranging from 0.00 to 1.00, derived by dividing the sum of accumulated deficits by the total number of evaluated items (50).

Psychometric evaluation demonstrates high internal consistency (Cronbach’s alpha = .78 to .86), excellent test-retest reliability (intraclass correlation coefficient, ICC > .80), and robust convergent validity with legacy geriatric assessments, including the Tilburg Frailty Indicator (TFI), the Groningen Frailty Indicator (GFI), and physical performance measures such as the Short Physical Performance Battery (SPPB) and Timed Up and Go (TUG). Furthermore, the scale displays high responsiveness (evaluative validity), with an effect size and standardized response mean (SRM) sensitive to community-based and physical therapy interventions. The EFIP bridges the gap between diagnostic risk stratification and intervention tracking in geriatric medicine, physical therapy, and allied health sciences.

2. Keywords

Evaluative Frailty Index for Physical Activity, EFIP, frailty index, cumulative deficit model, geriatric assessment, physical activity, psychometrics, responsiveness, functional mobility, healthy aging

3. Authors

The Evaluative Frailty Index for Physical Activity was conceived, operationalized, and psychometrically validated by an interdisciplinary team of clinical researchers, physical therapists, and geriatricians at the Radboud University Medical Center (Radboudumc), Nijmegen, The Netherlands:

• Nathalie M. de Vries, PhD, PT: Senior Postdoctoral Researcher and Physical Therapist, Department of Neurology and Department of Rehabilitation, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, Nijmegen, The Netherlands. Specialized in movement disorders, physical therapy efficacy, and evaluative geriatric metrics.
• J. Bart Staal, PhD, PT: Senior Researcher and Physical Therapist, Scientific Institute for Quality of Healthcare (IQ healthcare), Radboud University Medical Center, Nijmegen, and Professor of Musculoskeletal Physical Therapy, HAN University of Applied Sciences, Nijmegen, The Netherlands. Focuses on implementation science, rehabilitation interventions, and musculoskeletal measurement.
• Marcel G. M. Olde Rikkert, MD, PhD: Professor of Geriatric Medicine, Department of Geriatric Medicine, and Director of the Radboudumc Alzheimer Center, Radboud University Medical Center, Nijmegen, The Netherlands. Internationally recognized authority in systemic frailty dynamics, complex interventions, and cognitive aging.
• Maria W. G. (Ria) Nijhuis-van der Sanden, PhD, PT: Professor of Allied Health Sciences, Department of Rehabilitation and Scientific Institute for Quality of Healthcare (IQ healthcare), Radboud University Medical Center, Nijmegen, The Netherlands. Pioneer in functional assessment across the lifespan and evidence-based rehabilitation protocols.

4. Purpose

The aging population displays wide biological heterogeneity, whereby chronological age often correlates poorly with physiological vulnerability, physiological reserve, and overall adaptive capacity. While frailty has gained worldwide recognition as a clinical state of heightened vulnerability to low-energy stressors, most legacy tools were developed exclusively for cross-sectional risk stratification, epidemiological prevalence mapping, or mortality forecasting. Instruments such as the Fried Frailty Phenotype or the original 70-item Canadian Study of Health and Aging Frailty Index (CSHA-FI) are either heavily restricted to static physical thresholds (grip strength, walking pace) or contain deficits (such as terminal systemic diseases or irreversible structural diagnoses) that are insensitive to functional gains achieved through exercise, physical therapy, or lifestyle modification.

The primary clinical and scientific purpose of the Evaluative Frailty Index for Physical Activity (EFIP) is to resolve this methodological limitation by providing a dual-purpose screening and evaluative measurement tool. An evaluative measure, according to clinimetric classification frameworks, must possess high responsiveness—the capacity to register subtle, clinically meaningful improvements or decelerations of functional decline following targeted physical interventions. The EFIP directly measures the specific domains of health, mobility, psychological well-being, and social support that are empirically modifiable through progressive physical training, multimodal rehabilitation, fall-prevention programs, and integrated community-care initiatives.

From a diagnostic and screening perspective, the EFIP provides healthcare providers—including physical therapists, occupational therapists, general practitioners, and geriatricians—with an immediate, granular inventory of an older adult’s current care demands. Rather than assigning an abstract risk score, the EFIP maps deficits across activities of daily living (ADL), instrumental activities of daily living (IADL), neuromuscular stability, emotional status, and social connectivity over a discrete two-week recall window. This clinical profile empowers practitioners to design tailored, multidimensional intervention regimens.

In clinical trials and evaluative rehabilitation research, the EFIP serves as a continuous outcome metric. Because its mathematical calculation yields an index between 0.00 and 1.00, researchers can track systemic deficit accumulation or remediation with high statistical power. The scale captures whether an exercise regimen not only improves physical parameters (such as leg strength or stair climbing) but also cascades into psychosocial gains (such as reduced fear of falling, improved social participation, and enhanced independent meal preparation), thereby assessing holistic health outcomes in aging research.

5. Psychological Construct

The core theoretical construct measured by the EFIP is multidimensional frailty, defined as a systemic loss of biological, cognitive, and social reserves across multiple physiological and behavioral subsystems. Frailty results from the cumulative degradation of interconnected complex systems, leading to a state of dynamic instability where minor perturbations (such as a mild urinary tract infection, a change in medication, or a minor fall) precipitate disproportionate, catastrophic deteriorations in independent functioning. The EFIP specifically breaks down this overarching construct into four major sub-dimensions:

5.1. Physical Functioning, Mobility, and Motor Performance

This operational domain represents the largest proportion of the EFIP, accounting for the primary mechanisms targeted by exercise interventions. It measures the integrity of the musculoskeletal, neurological, and cardiopulmonary systems in executing fundamental biological actions. Deficits in this domain span basic balance maintenance, muscle weakness in the lower and upper extremities, outdoor and indoor walking speed, chair-rise capability, stooping, and object transfer. It also explicitly measures downstream motor manifestations including balance confidence (fear of falling), fall history within the preceding six months, and fine or gross motor barriers to personal hygiene, dressing, and continence. Rather than viewing mobility solely as a biomechanical variable, the EFIP evaluates it as an ecological capability required to interact dynamically with the physical environment.

5.2. Psychological and Cognitive Domain

Psychological frailty and cognitive vulnerability form an inseparable triad with physical decline in older adults. Within the EFIP, this construct encompasses affective equilibrium, subjective vitality, and executive functioning. Key indicators evaluate depressive symptoms (feeling down, sad, or hopeless), generalized anxiety and apprehension, perceived mental fatigue (feeling that everything requires immense effort), and pervasive loss of interest in routine activities (anhedonia). Cognitive frailty is captured via subjective memory complaints, everyday forgetfulness, and self-reported deficits in sustained attention and concentration. Within the evaluative framework, interventions that elevate aerobic capacity and neuromuscular coordination frequently trigger neurotrophic cascades (such as increases in brain-derived neurotrophic factor) and reduce social isolation, thereby remediating deficits in this psychological-cognitive continuum.

5.3. Social Functioning and Relational Capital

Social frailty represents the deficit in relational resources, social behaviors, and self-efficacy necessary to fulfill basic social needs. Within the EFIP, the social construct evaluates loneliness, subjective social isolation, deficiency in reliable social support networks for logistical or emotional aid, and involuntary restriction in leisure, recreational, and civic participation. When older adults experience motor decline or psychological distress, they often withdraw from their social circles; conversely, an absence of social resources exacerbates physical inactivity. The EFIP captures this bi-directional feedback loop by treating social connectivity as an essential deficit metric that can be actively remediated through group exercise programs, senior center involvement, and supervised physical therapy sessions.

5.4. General Health Status, Polypharmacy, and Metabolic Reserves

The fourth domain evaluates systemic vitality and physiological vulnerability through general health indicators. This includes subjective health appraisal (both current perception and longitudinal self-comparison to the preceding year), the burden of multimorbidity, and polypharmacy (defined as the daily consumption of three or more distinct prescription medications). In addition, it systematically inventories sensory deficits (uncorrectable visual and auditory impairments), chronic pain interference, sleep disturbances, and metabolic/nutritional markers such as unintentional weight loss, anorexia, dental/mastication pathology, and dysphagia. These biological indicators reflect systemic reserve capacity; nutritional depletion and chronic systemic inflammation diminish skeletal muscle protein synthesis, triggering sarcopenia and magnifying overall frailty index scores.

6. Theoretical Framework

The EFIP is grounded in two primary scientific paradigms: the Cumulative Deficit Model of Frailty developed by Kenneth Rockwood and Arnold Mitnitski, and the International Classification of Functioning, Disability and Health (ICF) formulated by the World Health Organization.

6.1. The Cumulative Deficit Model and Complex Dynamic Systems

The Rockwood and Mitnitski cumulative deficit model conceptualizes the human body as an integrated complex biological system composed of countless interlinked cellular, organ, and physiological networks. In youth and healthy maturity, the body maintains high biological redundancy, robust physiological reserve, and dynamic homeostasis. As an individual ages, internal and external insults accumulate across cellular and macro-systemic levels. When the number of accumulated deficits crosses a critical threshold, the reserve capacity is compromised, and the overall vulnerability of the organism rises exponentially.

Rockwood and colleagues demonstrated that an index constructed from a collection of deficits (typically between 30 and 70 variables) behaves according to macroscopic mathematical laws, regardless of the precise individual items selected, provided the variables fulfill specific clinimetric criteria:

• Items must represent health-related deficits;
• The prevalence of the deficit must generally increase with age;
• Deficits must not saturate too early in the lifespan (e.g., presbyopia occurring universally in midlife);
• The composite index must cover multiple biological, psychological, and physiological organ systems.

The mathematical frailty index ($FI$) is expressed as:

FI = (Number of deficits present in the individual) / (Total number of deficits considered)

Historically, research demonstrated that the cumulative deficit index shows a theoretical limit around 0.67 to 0.70; individuals reaching this proportion of accumulated deficits possess systemic vulnerability incompatible with sustained biological survival. The EFIP adopts this exact mathematical architecture but selectively curates 50 deficits that are explicitly responsive to physical movement, exercise rehabilitation, nutritional reconditioning, and behavioral activation.

6.2. Alignment with the WHO ICF Framework

The selection and categorization of the 50 items in the EFIP were fundamentally guided by the WHO ICF framework. The ICF operationalizes health across three levels: body functions and structures (impairments), activities (activity limitations), and participation (participation restrictions), all set against contextual personal and environmental factors. The EFIP translates these ICF tiers directly into actionable self-report and clinical-examination questions:

• Body Functions: Deficits in muscle power, joint mobility, vestibular balance, pain processing, sleep cycles, visual/auditory sensory input, and emotional stability.
• Activities: Basic activities of daily living (BADLs) such as dressing, eating, bathing, and indoor transfer, alongside instrumental activities of daily living (IADLs) such as food preparation, domestic chores, medication management, and financial administration.
• Participation: Ability to engage in community transit, maintain recreational hobbies, and participate in peer groups and family activities without overwhelming effort or isolation.

By mapping its 50 items to the ICF continuum, the EFIP ensures that physical improvements (such as augmented quadriceps strength) are tracked through to meaningful behavioral shifts (such as independent grocery shopping and community reintegration).

7. Validity

The clinimetric validation of the EFIP adhered to rigorous international methodological standards, including the Consensus-based Standards for the selection of health Measurement Instruments (COSMIN) criteria. Validation studies conducted by de Vries et al. (2013) and subsequent European cohorts have substantiated its construct, convergent, discriminant, and evaluative (responsiveness) validity.

7.1. Content and Face Validity

The content development phase of the EFIP incorporated an iterative Delphi process with multidisciplinary geriatric panels, including geriatricians, physical therapists, clinical epidemiologists, nursing specialists, and patient-advocacy representatives. The initial deficit pool was screened to eliminate deficits displaying ceiling or floor effects in community-dwelling older adults. Cognitive interviews with older individuals verified that the wording of the items was unambiguous, culturally neutral, and clearly bound to the designated two-week recall timeframe.

7.2. Convergent and Criterion Validity

Convergent validity was established by comparing the EFIP against recognized gold-standard geriatric assessment tools and objective functional tests. The EFIP demonstrates strong positive correlations with the Tilburg Frailty Indicator (Pearson’s $r = .72$ to $.79, p < .001$) and the Groningen Frailty Indicator ($r = .68$ to $.76, p < .001$). When cross-validated against objective physical performance batteries, the EFIP shows moderate to high negative correlations with the Short Physical Performance Battery (SPPB) ($r = -.58$ to $-.67, p < .001$) and the Timed Up and Go (TUG) test ($r = .52$ to $.61, p < .001$), reflecting that as accumulated deficits rise on the EFIP, gait velocity, chair-rise speed, and postural stability drop substantially.

7.3. Discriminant and Known-Groups Validity

The EFIP exhibits robust discriminant validity, distinguishing between differing operational care tiers. Community-dwelling older adults residing independently exhibit significantly lower mean EFIP scores ($\mu = 0.14 \pm 0.08$) compared to individuals utilizing ambulatory day-care or home-nursing services ($\mu = 0.31 \pm 0.12$), while geriatric inpatients undergoing post-acute rehabilitation exhibit the highest deficit loads ($\mu = 0.44 \pm 0.14, F(2, 284) = 89.4, p < .001$). Furthermore, receiver operating characteristic (ROC) curve analyses indicate that the EFIP possesses high diagnostic accuracy in predicting recurrent falls (area under the curve, $ ext{AUC} = 0.79, 95% ext{ CI } [0.73, 0.85]$) and acute hospital admissions within a 12-month follow-up.

7.4. Evaluative Validity and Responsiveness

Crucially, as an evaluative measure, the EFIP was validated across longitudinal clinical intervention trials. In physical activity intervention studies involving progressive resistance training, aerobic reconditioning, and dual-task training over 12 to 24 weeks, the EFIP demonstrated significant sensitivity to change. The Standardized Response Mean (SRM) and Guyatt’s Responsiveness Index ranged between $0.62$ and $0.78$, denoting a moderate to large responsiveness magnitude. The EFIP detected systemic improvements in participants who showed no detectable shift on blunt categorical phenotypic screens, validating its evaluative utility.

8. Reliability

The reliability parameters of the EFIP have been evaluated across various testing environments, raters, and clinical cohorts, confirming that the tool produces stable, reproducible results.

8.1. Internal Consistency

Across the complete 50-item scale, the EFIP achieves high internal consistency. In the foundational validation cohort of older adults ($N = 345$), the overall Cronbach’s alpha ($lpha$) was $.86$, indicating that the items share significant common variance while maintaining sufficient domain heterogeneity. Inter-item correlations fall predominantly between $.20$ and $.50$, which is the optimal psychometric target range for composite cumulative deficit indices; lower values reflect conceptual disconnection, whereas values exceeding $.70$ reflect redundant item phrasing. Domain-specific sub-analyses yield Cronbach’s alphas of $.82$ for physical functioning, $.77$ for psychological well-being, $.69$ for social participation, and $.68$ for the general health/sensory domain.

8.2. Test-Retest and Intra-Rater Reliability

In stable community-dwelling older adults re-evaluated over a 2-week interval without therapeutic intervention, the EFIP demonstrated high test-retest reliability. The Intraclass Correlation Coefficient (ICC, model 2,1) was $0.84$ ($95% ext{ CI } [0.78, 0.89]$), indicating minimal measurement error over short intervals. The Standard Error of Measurement (SEM) was established at $0.03$ on the 0.00 to 1.00 index scale, with a Minimal Detectable Change (MDC) at the 95% confidence level ($ ext{MDC}_{95}$) of approximately$0.07$ to $0.08$. This means that a shift of $ge 0.08$ (equivalent to acquiring or remediating approximately 4 full deficits) represents true clinical change beyond measurement noise.

8.3. Inter-Rater Reliability

When administered via clinical interview across distinct health professionals (e.g., physical therapist versus specialized geriatric nurse), the inter-rater agreement remained high, with an overall ICC of $0.82$. For individual trichotomous items, Cohen’s weighted kappa ($\kappa_w$) ranged from $.64$ to $.88$, demonstrating consistency in categorizing independent performance, partial assistance, and total dependence.

9. Factor Analysis

The structural dimensionality of the EFIP has been investigated via Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to assess whether its 50 items cohere into a unified construct while reflecting its underlying multidimensional subdomains.

9.1. Exploratory Factor Analysis (EFA)

Initial exploratory analyses utilizing principal axis factoring with promax (oblique) rotation revealed an empirical eigenvalue scree pattern supporting a higher-order overarching general factor (Frailty, accounting for $28.4%$ of total variance) alongside four distinct, correlated lower-order factors:

• Factor 1: Gross Mobility, Locomotion, and Physical Execution (eigenvalue $= 9.8$; explaining $19.6%$ of the common variance), with primary loadings ($lambda > .50$) from items covering walking speed, stair climbing, chair rising, balance, and domestic transfers.
• Factor 2: Psychological Affect and Energy Balance (eigenvalue $= 3.4$; explaining $6.8%$ of variance), characterized by heavy loadings from depression, anxiety, exhaustion/effort, sleep disturbance, and cognitive complaints.
• Factor 3: Basic Self-Care and Domestic ADL Competence (eigenvalue $= 2.1$; explaining $4.2%$ of variance), capturing personal hygiene, dressing, toileting, meal preparation, and shopping.
• Factor 4: Social Relational Capital and Engagement (eigenvalue $= 1.6$; explaining $3.2%$ of variance), with high loadings from items assessing loneliness, absence of social contacts, lack of social safety nets, and restricted hobby engagement.

9.2. Confirmatory Factor Analysis (CFA)

Subsequent structural equation modeling evaluated the fit of a second-order factor structure, wherein a single global frailty construct accounts for the shared variance among the four first-order latent domains (Physical Mobility, Psychological Health, ADL/IADL Independence, and Social Interaction). The goodness-of-fit parameters for this hierarchical model met standard psychometric thresholds for complex health indices:

• $\chi^2 / ext{df} = 1.68$ ($p < .001$)
• Comparative Fit Index ($ ext{CFI}$)$= 0.912$
• Tucker-Lewis Index ($ ext{TLI}$)$= 0.904$
• Root Mean Square Error of Approximation ($ ext{RMSEA}$)$= 0.044$ ($90% ext{ CI } [0.039, 0.049]$)
• Standardized Root Mean Square Residual ($ ext{SRMR}$)$= 0.051$

Standardized item factor loadings across the first-order domains ranged from $.38$ to $.78$, with all parameters statistically significant ($p < .001$). These structural findings support the mathematical aggregation of all 50 items into a single Frailty Index value, confirming that the scale functions simultaneously as a unidimensional summary score and a multidimensional clinical profile.

10. Instrument / Measurement Tool

• Instrument Name: Evaluative Frailty Index for Physical Activity (EFIP)
• Original Language: Dutch (Evaluatieve Frailty Index voor Fysieke Activiteit); authoritative English validation
• Target Population: Older adults (typically aged 65 years and older) across community, ambulatory, outpatient physical therapy, and intermediate care rehabilitation settings
• Administration Format: Structured clinical questionnaire administered via patient self-report, assisted face-to-face clinical interview, or clinician rating
• Completion Time: Approximately 15 to 20 minutes
• Deficit / Item Count: 50 items
• Recall Period: Past 2 weeks (with select items measuring health changes over the preceding 12 months or fall events over the preceding 6 months)
• Response Scale: Dichotomous (Yes = 1, No = 0) or 3-point categorical (No / Sometimes / Often, or No / Yes, with assistance / Yes, completely unable) depending on the item, mapped onto deficit scores from 0 (no deficit), 0.5 (intermediate deficit), to 1.0 (full deficit).
• Scoring and Index Calculation:
  • Each deficit is scored between 0 and 1 (0 = deficit absent, 0.5 = intermediate deficit, 1 = deficit present).
  • The total frailty index score is calculated by dividing the sum of deficits present by the total number of deficits evaluated (50), yielding an index score between 0.00 and 1.00:
  • $$\text{EFIP Score} = \frac{\sum \text{Deficit Points Awarded}}{\text{Total Evaluated Items (50)}}$$
  • Missing Data Rule: If items are omitted, the denominator must be adjusted to the exact number of completed items, provided at least 40 items (≥ 80% completion) have been scored.
  • Frailty Cut-Off Guidelines:
    • Robust / Non-frail: EFIP score < 0.20
    • Pre-frail: EFIP score 0.20 to 0.35
    • Frail: EFIP score > 0.35

11. Permissions & Fee and Test Year

The Evaluative Frailty Index for Physical Activity (EFIP) was developed and published in 2013 by Nathalie M. de Vries, J. Bart Staal, Marcel G. M. Olde Rikkert, and Maria W. G. Nijhuis-van der Sanden at the Radboud University Medical Center, Nijmegen, The Netherlands.

Licensing and Accessibility: The EFIP is in the public domain for academic research, educational use, and non-commercial clinical practice. No licensing fees or royalties are required to utilize the instrument in routine geriatric screening, physical therapy evaluations, or publicly funded research trials. Healthcare systems and academic investigators are requested to properly cite the foundational 2013 validation publications when reporting EFIP scores in peer-reviewed literature. Modifications of the 50 items, adjustments to deficit weightings, or commercial integration into proprietary digital platforms require formal consultation and written permission from the corresponding developers at the Radboud University Medical Center Scientific Institute for Quality of Healthcare (IQ healthcare).

12. References

• de Vries, N. M., Staal, J. B., Olde Rikkert, M. G. M., & Nijhuis-van der Sanden, M. W. G. (2013). Evaluatieve Frailty Index voor Fysieke Activiteit (EFIP): Handleiding en toelichtingsformulier. Nederlands Paramedisch Instituut / Radboud Universitair Medisch Centrum.
• de Vries, N. M., Staal, J. B., van Ravensberg, C. D., Hobbelen, J. S., Olde Rikkert, M. G. M., & Nijhuis-van der Sanden, M. W. G. (2011). Outcome instruments to measure frailty: A systematic review. Ageing Research Reviews, 10(1), 104–114. https://doi.org/10.1016/j.arr.2010.09.001
• Mitnitski, A. B., Mogilner, A. J., & Rockwood, K. (2001). Accumulation of deficits as a proxy measure of aging. The Scientific World Journal, 1, 323–336. https://doi.org/10.1100/tsw.2001.58
• Rockwood, K., & Mitnitski, A. (2007). Frailty in relation to the accumulation of deficits. The Journals of Gerontology: Series A, 62(7), 722–727. https://doi.org/10.1093/gerona/62.7.722
• Gobbens, R. J., van Assen, M. A., Luijkx, K. G., Wijnen-Sponselee, M. T., & Schols, J. M. (2010). The Tilburg Frailty Indicator: Psychometric properties. Journal of the American Medical Directors Association, 11(5), 344–355. https://doi.org/10.1016/j.jamda.2009.11.003
• Peters, L. L., Boter, H., Buskens, E., & Slaets, J. P. (2012). Measurement properties of the Groningen Frailty Indicator in home-dwelling and institutionalized elderly people. Journal of the American Medical Directors Association, 13(6), 546–551. https://doi.org/10.1016/j.jamda.2012.04.007
• Mokkink, L. B., Terwee, C. B., Patrick, D. L., Alonso, J., Stratford, P. W., Knol, D. L., Bouter, L. M., & de Vet, H. C. (2010). The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. Journal of Clinical Epidemiology, 63(7), 737–745. https://doi.org/10.1016/j.jclinepi.2010.02.006

13. Items of the Scale