1. Abstract
The Self-Care Inventory – Revised (SCI-R) is a widely utilized, psychometrically validated self-report instrument designed to assess adherence to contemporary diabetes self-management behaviors among individuals diagnosed with type 1 diabetes and type 2 diabetes. Originating from the work of Annette M. La Greca and colleagues and later revised by Weinger, Butler, Welch, and La Greca (2005), the SCI-R updates older self-care assessments to reflect modern clinical recommendations in endocrinology, including intensive blood glucose monitoring, flexible insulin adjustment, ketone testing, dietary management, and emergency preparedness. Comprising 15 items, the instrument asks respondents to rate their adherence over the previous 1 to 2 months using a 5-point Likert scale ranging from 1 (Never do it) to 5 (Always do this as recommended, without fail), with non-applicable options for non-insulin treated regimens.
Extensive psychometric investigations have established that the SCI-R exhibits strong internal consistency (Cronbach’s α typically ranging between .79 and .87 across diverse adult, adolescent, and cross-cultural cohorts) and robust test-retest reliability. Construct validity is supported by consistent negative correlations with glycemic control indicators, most notably glycated hemoglobin (HbA1c), as well as positive correlations with diabetes self-efficacy and psychological coping metrics. Structural evaluations using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) have demonstrated either a robust unidimensional structure or multi-factor models mapping onto specific operational dimensions such as glucose monitoring, dietary adherence, insulin administration, and emergency behaviors. The SCI-R represents an indispensable tool across behavioral medicine, pediatric and adult endocrinology, clinical trials, and epidemiological health outcome research.
2. Keywords
Self-Care Inventory – Revised, SCI-R, diabetes self-management, glycemic control, HbA1c, adherence measurement, type 1 diabetes, type 2 diabetes, behavioral medicine, psychometrics, health psychology, diabetes self-efficacy.
3. Authors
The original Self-Care Inventory (SCI) was conceived and developed by Annette M. La Greca, Ph.D., ABPP, Distinguished Professor of Psychology and Pediatrics at the University of Miami, Coral Gables, Florida, USA. Dr. La Greca is a preeminent scholar in pediatric psychology, chronic illness management, and developmental psychopathology. Her seminal contributions established foundational psychometric standards for assessing familial and social support systems as well as behavioral self-management in youth with diabetes.
The revised version (SCI-R) was adapted and psychometrically validated for both adult and adolescent populations by a team led by Katie Weinger, Ed.D., R.N. (Senior Investigator in the Research Division at the Joslin Diabetes Center and Associate Professor of Psychiatry at Harvard Medical School), in collaboration with H. A. Butler, M.S., G. W. Welch, Ph.D., and Annette M. La Greca, Ph.D.. Subsequent cross-cultural validations have expanded the instrument’s utility across international clinical contexts, such as validation in UK adult cohorts led by Jane Speight, Ph.D., Colin R. Martin, Ph.D., and colleagues.
4. Purpose
Diabetes mellitus is an arduous, multifaceted metabolic disorder that demands continuous, lifelong behavioral adherence from patients. The successful avoidance of acute complications (such as severe hypoglycemia, diabetic ketoacidosis, and hyperosmolar hyperglycemic states) and chronic microvascular and macrovascular sequelae (including retinopathy, nephropathy, peripheral neuropathy, and cardiovascular disease) depends almost entirely on effective daily self-management. The Self-Care Inventory – Revised (SCI-R) was specifically engineered to capture the broad spectrum of behaviors essential to contemporary diabetes clinical regimens.
From a clinical perspective, the primary purpose of the SCI-R is to provide healthcare practitioners, certified diabetes care and education specialists (CDCES), and clinical psychologists with a structured, quantitative portrait of a patient’s actual routine behaviors versus recommended medical guidelines. Clinicians often encounter substantial discrepancies between a patient’s theoretical knowledge of their diabetes treatment plan and their actual day-to-day execution. By isolating individual domains—such as ketone checking, dietary record keeping, carbohydrate tracking, glucose logging, clinic appointment attendance, and wearing medical alert identification—the SCI-R helps target specific behavioral deficits that contribute to suboptimal glycemic outcomes.
In clinical trials and behavioral research, the SCI-R serves as a standardized outcome measure to evaluate the efficacy of psychoeducational curricula, continuous glucose monitoring (CGM) systems, automated insulin delivery technologies, mobile health (mHealth) applications, and behavioral support interventions. Traditional biomedical outcomes, such as changes in HbA1c, reflect complex physiological processes influenced by pharmacodynamics, hormonal fluctuations, intercurrent illness, and genetics. As such, HbA1c cannot serve as an unconfounded surrogate for behavioral adherence. The SCI-R fills this critical methodological gap by offering a direct, standardized, psychometrically grounded behavioral metric that disentangles behavioral compliance from biological responsiveness.
5. Psychological Construct
The psychological construct evaluated by the SCI-R is diabetes self-management adherence (historically termed compliance or regimen adherence). In health psychology and behavioral medicine, self-management adherence is conceptualized as an active, self-directed, decision-making process through which individuals carry out a constellation of medical, lifestyle, and preventive behaviors designed to maintain metabolic homeostasis.
Unlike simple medication adherence—which typically involves taking an oral pill at a set schedule—diabetes self-care is dynamic, conditional, and demanding. The SCI-R captures this multifaceted behavioral construct through several operational sub-domains:
- Glycemic Monitoring and Data Tracking: Refers to the proactive tracking of capillary blood glucose levels (e.g., Check blood glucose with monitor), the systematic logging of trends (e.g., Record blood glucose results), and metabolic crisis prevention via biological assay (e.g., Check ketones). These behaviors demand cognitive vigilance, analytical processing of numerical trends, and proactive problem solving.
- Nutritional and Dietary Regulation: Captures the complex behavioral self-regulation required during nutritional intake. It encompasses quantity control (e.g., Eat recommended food portions), temporal discipline (e.g., Eat meals/snacks on time), behavioral tracking (e.g., Keep food records), and cognitive appraisal of nutritional content (e.g., Read food labels). In diabetes, nutritional adherence directly impacts postprandial glucose excursions.
- Medication and Insulin Administration: Evaluates adherence to pharmacological timing, dose accuracy, and algorithmic adjustments (e.g., Take insulin at the right time, Take correct dose of insulin, and Adjust insulin). This requires both declarative knowledge (knowing the prescribed dose) and procedural expertise (calculating carbohydrate-to-insulin ratios and correction factors based on ambient glucose).
- Emergency Preparedness and Safety Behaviors: Encompasses vital risk-mitigation actions to safeguard against acute medical events. Key actions include carrying rescue carbohydrates (e.g., Carry quick acting sugar for lows), adhering to formal clinical treatment protocols (e.g., Treat low blood glucose), and publicly identifying vulnerability through medical alert items (e.g., Wear medic alert).
- Health Care Engagement and Physical Activity: Captures structural engagement with clinical providers (e.g., Attend clinic appointments) and energy expenditure behaviors (e.g., Exercise regularly), which directly moderate insulin sensitivity and cardiovascular risk.
6. Theoretical Framework
The conceptual foundation of the Self-Care Inventory – Revised rests heavily on Social Cognitive Theory (SCT), originally formulated by Albert Bandura (1986). Central to SCT is the triadic reciprocal causation model, wherein personal cognitive factors, environmental influences, and behavioral patterns continuously interact. In the context of the SCI-R, self-care behaviors are not seen as reflexive habits, but as outcomes of complex self-regulatory mechanisms governed by self-efficacy expectations (an individual’s confidence in executing specific tasks), outcome expectancies, and sociostructural barriers.
Bandura’s concept of self-regulation—comprising self-monitoring, judgment of personal performance against clinical standards, and affective self-reaction—is directly measured by the SCI-R. For instance, checking blood glucose and reading food labels represent core self-monitoring functions. When an individual identifies a discrepancy between target glucose levels and current readings, cognitive appraisal mechanisms drive corrective actions, such as insulin adjustment or hypoglycemia treatment. Patients with low perceived diabetes self-efficacy are far more prone to omit complex self-care steps, leading to behavioral dropouts captured by lower SCI-R scores.
Furthermore, the SCI-R aligns with the Self-Regulation Model of Illness Cognitions developed by Howard Leventhal and colleagues. According to Leventhal’s Common Sense Model (CSM), individuals build cognitive and emotional representations of their illness based on identity, cause, timeline, consequences, and cure/controllability. These representations guide behavioral coping strategies. When an individual views diabetes as manageable through proactive intervention (high perceived controllability), they are significantly more likely to engage in the daily self-care behaviors measured on the SCI-R. Conversely, fatalistic perceptions or acute diabetes distress erode these behaviors, manifesting as lower scores on the instrument.
7. Validity
The psychometric validity of the SCI-R has been extensively confirmed across diverse clinical and demographic cohorts, establishing exceptional construct, convergent, discriminant, and predictive validity.
Criterion and Predictive Validity
In the seminal adult validation study by Weinger et al. (2005) involving both type 1 (n = 138) and type 2 (n = 252) diabetes patients, the SCI-R total score demonstrated statistically significant inverse correlations with objective glycemic control measured via glycated hemoglobin (HbA1c). Higher SCI-R self-care adherence scores were robustly associated with lower HbA1c levels in both type 1 (r = -.36, p < .001) and type 2 diabetes cohorts (r = -.28, p < .001). Longitudinal studies have similarly verified that baseline SCI-R scores predict subsequent metabolic control, glycemic variability, and the incidence of acute diabetic ketoacidosis over 6- to 12-month follow-up intervals.
Convergent Validity
Convergent validity is well documented through strong, theoretically consistent associations with related validated psychological instruments. Weinger et al. (2005) reported robust positive correlations between SCI-R total scores and diabetes self-efficacy, as assessed by the Confidence in Diabetes Self-Care Scale (CIDS) (r = .57, p < .001 for type 1; r = .52, p < .001 for type 2). The scale also correlates positively with the Problem Areas in Diabetes (PAID) coping indices and the Summary of Diabetes Self-Care Activities (SDSCA) measure (Pearson’s r typically ranging from .50 to .68 across matching behavioral subdomains).
Discriminant Validity
The SCI-R demonstrates distinct divergence from unrelated psychological constructs. Correlations between SCI-R total scores and general trait anxiety or non-health-related self-esteem scales are consistently near zero (r values typically < .12, non-significant). Furthermore, the SCI-R demonstrates modest, statistically coherent negative correlations with depressive symptoms (evaluated via the CES-D) and diabetes-related emotional distress (PAID total score; r = -.25 to -.34), confirming that while psychological distress impedes adherence, self-care remains a distinct behavioral construct.
8. Reliability
The reliability of the SCI-R has been confirmed through repeated assessments of internal consistency, item-total correlations, and temporal stability across clinical trials and observational studies.
Internal Consistency
In the initial adult standardization study by Weinger, Butler, Welch, and La Greca (2005), the overall internal consistency was high, yielding a Cronbach’s alpha of α = .87 across the total validation sample. When stratified by diabetes etiology, the coefficient was α = .87 for individuals with type 1 diabetes and α = .86 for individuals with type 2 diabetes. Subsequent international validations have reported comparable results: Khagram, Martin, Davies, and Speight (2013) found α = .79 to .82 among UK adults with type 2 diabetes participating in the AT.LANTUS Follow-on study, while adolescent adaptations have consistently yielded alpha coefficients between α = .81 and α = .85.
Test-Retest Reliability and Stability
Temporal stability evaluated over 1- to 4-week test-retest intervals in metabolically stable patients demonstrates robust intra-class correlation coefficients (ICC). Weinger et al. observed test-retest correlations exceeding r = .78 (p < .001), indicating that individual reporting of self-care remains stable in the absence of therapeutic or medical intervention. Corrected item-total correlations for the 15 items generally exceed the standard psychometric threshold of .30 (typically ranging from .32 to .66), demonstrating that all items contribute meaningfully to the overarching construct.
9. Factor Analysis
The latent structure of the SCI-R has been evaluated through both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA), uncovering both hierarchical multidimensionality and a reliable overarching general factor.
Exploratory Factor Analysis (EFA)
During initial structural development, principal axis factoring with promax and varimax rotations yielded multi-factor configurations representing distinct clinical behavioral demands:
- Factor 1: Blood Glucose Monitoring and Logging (Item loadings: Check blood glucose [.74], Record blood glucose results [.81], Check ketones [.52]).
- Factor 2: Dietary Regulation and Food Tracking (Item loadings: Eat recommended food portions [.76], Keep food records [.65], Read food labels [.68], Eat meals/snacks on time [.58]).
- Factor 3: Medication and Insulin Adjustment (Item loadings: Take insulin at right time [.82], Take correct dose [.86], Adjust insulin [.59]).
- Factor 4: Preventive and Emergency Behaviors (Item loadings: Carry quick acting sugar [.64], Treat low blood glucose [.71], Wear medic alert [.42], Attend clinic appointments [.45], Exercise regularly [.38]).
Confirmatory Factor Analysis (CFA) and Model Fit
In structural equation modeling analyses, such as those performed by Khagram et al. (2013), a unidimensional higher-order model and a correlated four-factor model were evaluated using maximum likelihood estimation. The correlated four-factor model demonstrated good fit to the empirical data: Comparative Fit Index (CFI) = .935 to .958, Tucker-Lewis Index (TLI) = .921 to .946, and Root Mean Square Error of Approximation (RMSEA) = .048 to .058 (90% CI [.041, .065]), with standardized root mean square residual (SRMR) ≤ .052. Because the latent factors are substantially inter-correlated (inter-factor correlations ranging from .42 to .71), psychometricians strongly endorse the clinical and research use of the overall composite score as a parsimonious index of global diabetes self-care.
10. Instrument / Measurement Tool
- Instrument Name: Self-Care Inventory – Revised (SCI-R)
- Target Population: Adolescents (ages 12+) and adults diagnosed with Type 1 or Type 2 diabetes mellitus
- Administration Method: Self-administered (paper-and-pencil, computerized, or web-based survey)
- Completion Time: Approximately 3 to 5 minutes
- Total Number of Items: 15 items
- Response Scale: 5-point Likert scale:
- 1 = Never do it
- 2 = Almost never do it
- 3 = Sometimes do it
- 4 = Almost always do it
- 5 = Always do this as recommended, without fail
- (A specific "Not applicable" / "Not on insulin" option is provided for medical-regimen-specific questions such as items 4, 12, 14, and 15)
- Recall Period: Behavior during the past 1 to 2 months
- Scoring Methodology:
- Item scores range from 1 to 5. Items marked as non-applicable are omitted from calculation.
- Mean Scoring: Compute the sum of valid, completed items divided by the total number of valid items answered (scale 1.0 to 5.0).
- Standardized Rescaling (0 to 100): Frequently used in research:
Standardized Score = [(Mean Item Score - 1) / 4] × 100, yielding a range from 0 (complete non-adherence) to 100 (perfect adherence). - Subscale Scores: Calculated similarly as mean item scores within respective behavioral factors (Glucose Monitoring, Dietary Behaviors, Insulin/Medication Administration, Safety/Emergency Behaviors).
- Interpretation: Higher scores consistently reflect superior behavioral adherence to recommended clinical self-care guidelines.
11. Permissions & Fee and Test Year
The original Self-Care Inventory was introduced in 1995 by Dr. Annette M. La Greca and colleagues, with the comprehensive manual developed in 2004 (La Greca, 2004). The modernized adult validation was published by Dr. Katie Weinger and colleagues in 2005.
Copyright and Usage Policy: The SCI-R is protected by academic copyright. However, it is widely made available for academic, non-commercial research and clinical practice. Researchers and healthcare institutions seeking to administer the SCI-R must obtain permission from the original developers (Dr. Annette M. La Greca, University of Miami, or Dr. Katie Weinger, Joslin Diabetes Center) or access the manual and survey via standard institutional channels. Commercial use, pharmaceutical drug trials, or integration into proprietary software applications may require explicit licensing agreements and associated administrative fees.
12. References
Khagram, L., Martin, C. R., Davies, M. J., & Speight, J. (2013). Psychometric validation of the Self-Care Inventory-Revised (SCI-R) in UK adults with type 2 diabetes using data from the AT.LANTUS Follow-on study. Health and Quality of Life Outcomes, 11, Article 24. https://doi.org/10.1186/1477-7525-11-24
La Greca, A. M. (2004). Manual for the Self-Care Inventory. Unpublished manuscript, University of Miami, Coral Gables, FL.
La Greca, A. M., Auslander, W. F., Greco, P., Spetter, D., Fisher, E. B., Jr., & Santiago, J. V. (1995). I get by with a little help from my family and friends: Adolescents’ support for diabetes care. Journal of Pediatric Psychology, 20(4), 449–476. https://doi.org/10.1093/jpepsy/20.4.449
Weinger, K., Butler, H. A., Welch, G. W., & La Greca, A. M. (2005). Measuring diabetes self-care: A psychometric analysis of the Self-Care Inventory-Revised with adults. Diabetes Care, 28(6), 1346–1352. https://doi.org/10.2337/diacare.28.6.1346