Cancer Support Source-Caregiver

1. Abstract

The Cancer Support Source-Caregiver (CSS-Caregiver), developed and validated by Zaleta et al. (2023), is an 18-item multidimensional distress screening, referral, and clinical decision-support instrument specifically engineered to identify the pervasive unmet biopsychosocial needs of informal cancer caregivers. Building upon foundational conceptual and psychometric groundwork originally introduced by Shaffer et al. (2019), the CSS-Caregiver operationalizes caregiver distress across five distinct functional domains: Emotional Well-Being, Patient Well-Being, Caregiving Tasks, Finances, and Healthy Lifestyle. Additionally, the instrument incorporates embedded, empirically derived screening algorithms capable of identifying elevated risk for clinically significant major depressive disorder and generalized anxiety disorder, alongside an optional clinical assessment item tracking tobacco and substance use.

Psychometric evaluation across diverse cohorts of informal adult caregivers demonstrates robust structural and measurement properties. Exploratory and confirmatory factor analyses confirm a five-factor latent structure that accounts for 67% of total variance, displaying exceptional model fit indices (CFI = 0.94, TLI = 0.93, SRMR = 0.06, RMSEA = 0.07). The CSS-Caregiver exhibits excellent internal consistency reliability, evidenced by an overall Cronbach’s alpha of α = 0.92, with stable test-retest reproducibility across intervals (intraclass correlation coefficient [ICC] = 0.85). Convergent validity is substantiated through strong Pearson correlations (|r| ≥ 0.50) with standardized measures of depression, anxiety, fatigue, and caregiver health burden, while discriminant validity is confirmed via negligible associations with generalized self-esteem (|r| = 0.12 to 0.22). Known-groups validity demonstrates significant variations in distress across age brackets, caregiving intensity, household income tiers, and gender. Furthermore, receiver operating characteristic analyses substantiate high diagnostic utility for the depression (≥ 3, sensitivity = 94.5%–95.0%, specificity = 63.0%–66.4%) and anxiety risk algorithms (≥ 4, sensitivity = 86.7%–90.6%, specificity = 66.0%–68.0%). As a clinical triage instrument, the CSS-Caregiver effectively bridges the gap between routine distress screening and individualized supportive oncology service allocation.

2. Keywords

Cancer Caregivers, Caregiver Burden, Caregiver Distress, Distress Screening, Psycho-Oncology, Needs Assessment, Anxiety Risk, Depression Risk, Factor Analysis, Psychometrics, Supportive Oncology, Oncology Social Work, Health-Related Quality of Life, Cancer Support Source.

3. Authors

The Cancer Support Source-Caregiver was developed by an interdisciplinary research and clinical team affiliated with the Cancer Support Community Research and Training Institute:

• Alexandra K. Zaleta, Ph.D. — Cancer Support Community Research and Training Institute, Philadelphia, PA, USA. ORCID: 0000-0002-8589-4722. Corresponding Author Email: [email protected].
• Melissa F. Miller, Ph.D., MPH — Cancer Support Community Research and Training Institute, Philadelphia, PA, USA. ORCID: 0000-0001-8288-2681.
• Erica E. Fortune, Ph.D. — Cancer Support Community Research and Training Institute, Philadelphia, PA, USA. ORCID: 0000-0002-6748-8229.
• Julie S. Olson, Ph.D. — Cancer Support Community Research and Training Institute, Philadelphia, PA, USA. ORCID: 0000-0002-4669-3880.
• Kimberly Papay Rogers, Ph.D. — Cancer Support Community Research and Training Institute, Philadelphia, PA, USA. ORCID: 0000-0003-4740-1269.
• Kelly Hendershot, MSW, LCSW — Cancer Support Community Headquarters, Washington, DC, USA. ORCID: 0000-0001-5730-4935.
• Susan Ash-Lee, MSW, LCSW — Cancer Support Community Headquarters, Washington, DC, USA. ORCID: 0000-0003-4202-4249.

4. Purpose

Informal cancer caregivers—spouses, adult children, partners, siblings, and close friends—serve as the primary pillar of support for patients traversing the complex oncologic trajectory. Caregivers frequently manage complex treatment regimens, monitor debilitating toxicities, coordinate outpatient clinical visits, advocate within health systems, and handle household economics, all while navigating intense emotional distress. Extensive epidemiological investigations have repeatedly revealed that family caregivers often experience levels of psychological distress, anxious apprehension, depressive symptomatology, and chronic exhaustion that equal or surpass the distress reported by patients themselves. In response to these pervasive challenges, oncology accreditation bodies, such as the American College of Surgeons Commission on Cancer (CoC), have established mandates for routine distress screening. However, such clinical initiatives have predominantly focused exclusively on patients, leaving informal caregivers largely overlooked in ambulatory settings.

The primary clinical and scientific purpose of the Cancer Support Source-Caregiver (CSS-Caregiver) is to provide an efficient, psychometrically sound, multidimensional assessment instrument designed specifically to identify the broad spectrum of caregiver distress, screen for clinical risks of depressive and anxiety disorders, and instantly automate targeted clinical referrals. Rather than viewing distress as an abstract, unidimensional psychological construct, the CSS-Caregiver conceptualizes caregiver burden as an ecological interplay between direct caregiving responsibilities, personal physiological self-care, financial toxicity, existential uncertainty, and systemic healthcare communication barriers.

In clinical practice, the CSS-Caregiver serves as an actionable triage platform. Completed via digital tablets in clinic waiting rooms, patient portals, or community oncology support facilities, the scale pinpoints distinct functional areas where caregivers require immediate psychoeducation, psychosocial counseling, navigational support, or financial counseling. An item-level rating of ≥ 2 (indicating moderate, serious, or very serious concern) acts as an automated threshold that alerts clinical oncology social workers, navigators, and care coordinators. Concurrently, the measure calculates discrete risk indicators for clinical depression and anxiety, ensuring that individuals who meet critical psychiatric cutoffs receive rapid clinical assessment and secondary psychiatric referral. In clinical research, the CSS-Caregiver offers an intervention-sensitive, multidimensional endpoint for randomized controlled trials testing supportive care, psychosocial counseling, and educational interventions directed toward cancer dyads and family systems.

5. Psychological Construct

The CSS-Caregiver measures multidimensional caregiver distress within an oncology framework, capturing subjective emotional, physical, instrumental, and financial strain. Unlike generic caregiver burden inventories (such as the Zarit Burden Interview), the CSS-Caregiver captures cancer-specific vulnerabilities across five correlated latent dimensions alongside two algorithmic psychiatric risk screening scores:

Emotional Well-Being

This subscale captures core affective manifestations of internal distress, social isolation, and existential strain experienced by the caregiver. It evaluates feelings of profound sadness, pervasive depressive affect, acute anxiety, catastrophic apprehension, perceived helplessness or being emotionally overwhelmed, and chronic loneliness stemming from the caregiver role. In addition, this domain evaluates emotional adjustments surrounding altered physical appearance, shifting relationship dynamics, and disrupted intimacy within the caregiving dyad.

Patient Well-Being

Caregivers do not experience distress in isolation; their subjective suffering is inextricably tied to the patient’s physical and psychological status. This subscale captures vicarious distress and empathetic strain, reflecting the caregiver’s anxiety regarding the patient’s unrelieved physical pain, intractable symptoms, emotional despair, depressive moods, and existential dread regarding impending disease progression, metastatic recurrence, or end-of-life outcomes.

Caregiving Tasks

This structural dimension quantifies instrumental distress related to functional care coordination and disease management. It measures the caregiver’s perceived difficulty and burden in managing complex medication protocols, monitoring treatment-induced side effects, communicating effectively with physicians and multidisciplinary oncology teams, acquiring reliable health literacy information, and managing work-life-caregiving conflicts.

Finances

Oncologic treatments introduce severe financial toxicity that impacts family stability. This dimension evaluates the caregiver’s distress regarding escalating direct out-of-pocket cancer medical costs, prescription co-pays, loss of personal employment income or productivity, handling regular household expenses, and the administrative burden of negotiating insurance approvals, pre-authorizations, and disability benefits.

Healthy Lifestyle

The intense physical and logistical demands of caregiving often erode the caregiver’s own health-maintaining routines. This domain measures personal self-care disruption, specifically assessing the caregiver’s ability to maintain regular physical activity, adhere to balanced nutritional habits, and achieve restorative, uninterrupted sleep. Compromise in these primary lifestyle behaviors accelerates physical fatigue and immunological vulnerability.

Depression Risk and Anxiety Risk Algorithms

Beyond dimensional profiling, the instrument yields two clinical risk screening indices. The Depression Risk score combines core affective and anhedonic items (e.g., sadness/depression and feeling overwhelmed) to identify caregivers exhibiting symptoms suggestive of major depressive episodes. The Anxiety Risk score synthesizes manifestations of cognitive and autonomic anxiety (e.g., anxious apprehension and pervasive worry about cancer progression) to detect individuals in need of urgent psychiatric evaluation.

6. Theoretical Framework

The conceptual framework of the CSS-Caregiver is grounded in the convergence of Lazarus and Folkman’s Transactional Model of Stress and Coping (1984), Pearlin’s Caregiver Stress Process Model (1990), and modern Family Systems and Dyadic Coping Theory (Bodenmann, 2005).

The Transactional Model of Stress and Coping

According to Lazarus and Folkman, psychological stress is not an environmental stimulus or an isolated physiological response, but rather a dynamic transaction between the individual and their environment. When faced with a loved one’s cancer diagnosis, caregivers engage in primary appraisal, evaluating whether disease progression, treatment toxicities, or economic costs pose significant threats to their family’s well-being. Concurrently, secondary appraisal involves assessing personal and external resources available to handle these demands. Distress emerges when perceived caregiving demands outstrip perceived coping resources, self-efficacy, and institutional support. The CSS-Caregiver maps these cognitive appraisals across concrete physical, financial, and logistical care challenges.

Pearlin’s Caregiver Stress Process Model

Leonard Pearlin’s sociological framework posits that caregiver distress is structured hierarchically, divided into primary stressors and secondary stressors. Primary stressors stem directly from the patient’s illness, functional dependence, cognitive changes, and physical suffering (measured within the Patient Well-Being and Caregiving Tasks subscales). Over time, these primary challenges spill over into secondary role strains, leading to occupational instability, interpersonal friction, financial strain, and personal lifestyle degradation (captured by the Finances and Healthy Lifestyle subscales). Unmitigated exposure to primary and secondary stressors contributes to psychological, physical, and autonomic exhaustion, manifesting as clinical depression and generalized anxiety.

Dyadic Coping and Interdependence

Contemporary psycho-oncology models highlight the interdependent nature of the cancer experience: patient and caregiver function as a reciprocal emotional unit. Emotional distress is communicable; patient suffering induces vicarious trauma in the caregiver, while caregiver strain undermines patient recovery and psychological resilience. By evaluating both caregiver-centric emotional strain and patient-directed concerns, the CSS-Caregiver captures this interdependent dynamic.

7. Validity

The psychometric properties of the CSS-Caregiver were evaluated by Zaleta et al. (2023) through an extensive cross-sectional validation study involving adult informal caregivers of cancer patients across the United States. Validation analyses established robust convergent, discriminant, known-groups, and criterion-referenced diagnostic validity.

Convergent Validity

Convergent validity was evaluated by correlating the CSS-Caregiver total distress and subscale scores with validated reference instruments assessing affective distress and caregiver health impacts. Bivariate Pearson correlation analyses demonstrated strong, statistically significant associations (|r| ≥ 0.50):

• Depressive Symptomatology: Correlated strongly with established measures including the Patient Health Questionnaire (PHQ-9) and PROMIS Depression metrics (r ≥ 0.55).
• Anxiety Symptoms: Exhibited pronounced positive correlations with the Generalized Anxiety Disorder-7 (GAD-7) scale (r ≥ 0.58).
• Caregiver Fatigue & Physical Burden: Showed robust correlations with standardized health burden inventories, confirming that elevated CSS-Caregiver distress tracks with functional, physical, and subjective exhaustion.

Discriminant Validity

Discriminant validity was confirmed by demonstrating low correlations with constructs theoretically distinct from caregiver distress. When correlated with generalized self-esteem (e.g., Rosenberg Self-Esteem Scale), correlation coefficients remained low and non-overlapping, ranging from r = -0.12 to r = -0.22. This demonstrates that the CSS-Caregiver measures situation-specific distress rather than enduring self-evaluation or global self-worth.

Known-Groups Validity

The instrument demonstrated known-groups validity across distinct clinical and demographic strata:

• Caregiving Intensity: Caregivers dedicating > 20 hours per week to direct care activities reported significantly higher total distress and caregiving task burden compared to those with lower time commitments (p < 0.001).
• Income and Financial Strain: Individuals reporting lower household income or inadequate health insurance coverage exhibited significantly elevated scores on the Finances factor (p < 0.001).
• Age and Gender: Younger caregivers and female caregivers exhibited higher total distress, matching established epidemiological trends in informal oncology caregiving.

Criterion Validity, Sensitivity, and Specificity

Receiver operating characteristic (ROC) curves established cutoff scores for identifying clinical psychiatric risk against standard clinical thresholds (PHQ-9 ≥ 10 for depression; GAD-7 ≥ 10 for anxiety):

• Depression Risk Screening Cutoff (≥ 3 out of 8): Exhibited high sensitivity ranging between 94.5% and 95.0%, with adequate specificity of 63.0% to 66.4% (Area Under the Curve [AUC] > 0.88).
• Anxiety Risk Screening Cutoff (≥ 4 out of 8): Demonstrated sensitivity of 86.7% to 90.6% and specificity of 66.0% to 68.0% (AUC > 0.86).

8. Reliability

The CSS-Caregiver exhibits strong measurement reliability, evaluated via internal consistency across items and temporal stability across repeated administrations.

Internal Consistency Reliability

In the primary psychometric validation study by Zaleta et al. (2023), the 18-item core CSS-Caregiver scale demonstrated high overall internal consistency, yielding a Cronbach’s alpha coefficient of α = 0.92. Individual subscale alphas demonstrated solid reliability across all five domains:

• Emotional Well-Being: α ≥ 0.85
• Patient Well-Being: α ≥ 0.81
• Caregiving Tasks: α ≥ 0.79
• Finances: α ≥ 0.84
• Healthy Lifestyle: α ≥ 0.78

Test-Retest Reliability and Intraclass Correlations

Temporal stability was evaluated in a retest subgroup across a 2- to 3-week interval, during which caregiving demands remained relatively stable. The test-retest reliability coefficient for the 18-item total distress score was r = 0.85. When evaluated with the optional substance use assessment item included, the stability coefficient was r = 0.86. Intraclass Correlation Coefficients (ICCs) across all five individual subscales equaled or exceeded 0.72, demonstrating reliable psychometric stability over time while maintaining sensitivity to meaningful clinical changes.

9. Factor Analysis

The latent structural integrity of the CSS-Caregiver was established using sequential exploratory factor analysis (EFA) followed by confirmatory factor analysis (CFA) across split derivation and validation samples.

Exploratory Factor Analysis (EFA)

Initial item pools derived from clinical expert panels, focus groups, and prior pilot work (Shaffer et al., 2019) underwent principal axis factoring with promax (oblique) rotation to account for expected inter-factor correlations. Items with low factor loadings (< 0.40), high cross-loadings (> 0.30 across multiple factors), or conceptual redundancy were removed. The final EFA revealed a clear five-factor solution comprising 18 items, with every retained factor supported by at least three defining items. This five-factor model accounted for 67% of the total variance. An additional item tracking tobacco and substance use demonstrated low factor loadings across all domains (< 0.30); however, given its clinical diagnostic relevance, it was retained as a standalone risk screening item rather than being forced into a dimensional factor.

Confirmatory Factor Analysis (CFA)

Confirmatory factor analysis was conducted on an independent validation cohort to verify the five-factor structural model. The five latent factors (Emotional Well-Being, Patient Well-Being, Caregiving Tasks, Finances, Healthy Lifestyle) were modeled as intercorrelated first-order dimensions. The CFA verified acceptable-to-good fit without post-hoc error covariances, yielding the following fit metrics:

• Chi-Square: χ² = 216.12 (df = 125, p < 0.001)
• Comparative Fit Index (CFI): 0.94
• Tucker-Lewis Index (TLI): 0.93
• Root Mean Square Error of Approximation (RMSEA): 0.07 (90% Confidence Interval: [0.06, 0.09])
• Standardized Root Mean Square Residual (SRMR): 0.06

All standardized factor loadings exceeded 0.60 (p < 0.001), confirming that the 18 items reliably reflect their intended latent constructs.

10. Instrument / Measurement Tool

The operational specifications of the CSS-Caregiver are summarized below:

• Test Type: Original multidimensional distress, needs assessment, and clinical risk screening questionnaire.
• Administration Format: Self-administered; available via paper-and-pencil, integrated web platforms, electronic medical record (EMR) portals, or waiting room tablet computers.
• Target Population: Adult informal family and friend caregivers (aged ≥ 18 years) providing unpaid physical, practical, or emotional care to individuals diagnosed with any cancer type or stage.
• Number of Items: 18 core items, plus 1 optional standalone screening item assessing tobacco/substance use.
• Response Scale: 5-point Likert scale:
  • 0 = Not at all
  • 1 = Slightly
  • 2 = Moderately
  • 3 = Seriously
  • 4 = Very seriously
• Factor Structure / Subscales:
  • Emotional Well-Being: Items 1, 2, 3, 4, 5 (Range: 0–4 mean score).
  • Caregiving Tasks: Items 6, 7, 8, 9 (Range: 0–4 mean score).
  • Healthy Lifestyle: Items 10, 11, 12 (Range: 0–4 mean score).
  • Finances: Items 13, 14, 15 (Range: 0–4 mean score).
  • Patient Well-Being: Items 16, 17, 18 (Range: 0–4 mean score).
• Scoring and Index Calculation:
  • Total Distress Score: Calculated by summing all 18 completed items (Range: 0–72) or by computing the overall mean across items (Range: 0–4). In instances of missing data, the mean of the available completed items is used to calculate the score.
  • Subscale Domain Scores: Calculated as the mean of completed items within each domain (sum divided by the number of completed items in that factor).
  • Actionable Referral Thresholds: An item-level score of ≥ 2 (“Moderately”, “Seriously”, or “Very seriously”) indicates an actionable unmet need that prompts specific supportive oncology interventions.
  • Depression Risk Score: Derived by summing two specific depression indicator items (Item 1 and Item 3; range 0–8). A cutoff score of ≥ 3 denotes high risk for clinical depression.
  • Anxiety Risk Score: Derived by summing two specific anxiety indicator items (Item 2 and Item 18; range 0–8). A cutoff score of ≥ 4 denotes elevated risk for clinical anxiety.

11. Permissions & Fee and Test Year

• Test Publication Year: 2023 (Initial conceptual pilot published in 2019).
• Copyright Holder: Cancer Support Community (CSC) Research & Training Institute.
• Commercial Use: Prohibited without express written authorization and formal licensing agreements.
• Fee: Available free of charge for non-profit academic research, clinical quality improvement, and non-commercial supportive oncology care programs.
• Permissions and Access Inquiries: Clinicians and investigators must contact the corresponding author: Alexandra K. Zaleta, Ph.D., Cancer Support Community Research and Training Institute, 520 Walnut Street, Suite 1170, Philadelphia, PA 19106, USA; Email: [email protected].

12. References

• Bodenmann, G. (2005). Dyadic coping and its significance for marital functioning. In T. A. Revenson, K. Kayser, & G. Bodenmann (Eds.), Couples coping with stress: Emerging perspectives on dyadic coping (pp. 33–49). American Psychological Association. https://doi.org/10.1037/11031-002
• Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer Publishing Company.
• Pearlin, L. I., Mullan, J. T., Semple, S. J., & Skaff, M. M. (1990). Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist, 30(5), 583–594. https://doi.org/10.1093/geront/30.5.583
• Shaffer, K. M., Miller, M. F., Longacre, M. L., Chen, N. T., & Zaleta, A. K. (2019). Development of CancerSupportSourceTM-Caregiver: A distress and needs screener for cancer caregivers. Psycho-Oncology, 28(8), 1735–1742. https://doi.org/10.1002/pon.4925
• Zaleta, A. K., Miller, M. F., Fortune, E. E., Olson, J. S., Rogers, K. P., Hendershot, K., & Ash-Lee, S. (2023). CancerSupportSourceTM-Caregiver: Development of a distress screening measure for cancer caregivers. Psycho-Oncology, 32(3), 418–428. https://doi.org/10.1002/pon.6092

13. Items of the Scale (Questionnaire)