[Developing a predictive model for the caregiver strain index]

Area of Science:

• Gerontology
• Nursing
• Public Health

Background:

• Homecare for patients with multiple morbidities is increasingly prevalent.
• Caregiver burden is a significant issue for informal family caregivers.
• The Caregiver Strain Index (CSI) questionnaire measures perceived caregiver burden.

Purpose of the Study:

• To construct a diagnostic nomogram for informal caregiver burden.
• To utilize data from a predictive model for nomogram development.
• To assess the clinical utility of the nomogram in community and geriatric settings.

Main Methods:

• Binary logistic regression was employed to build the predictive model.
• Questionnaire items were used as dichotomous factors.
• The Caregiver Strain Index score was categorized as "no" (0-6) or "yes" (≥7) caregiver stress.

Main Results:

• A diagnostic nomogram was developed using four specific questionnaire items (1, 4, 5, 9).
• The nomogram demonstrated a high discriminative value with an Area Under the ROC Curve of 0.962.
• Calibration analysis indicated a high goodness of fit (Brier score = 0.686, Nagelkerke R² = 0.791).

Conclusions:

• The developed nomogram is a clinically useful tool for assessing informal caregiver burden.
• The nomogram shows high accuracy and reliability in predicting caregiver stress.
• This tool can support community nursing and geriatric care professionals in identifying and managing caregiver strain.