Prediction of caregiver burden in informal caregivers of frail older people.

Abstract

Background The goal of the Dutch government is to ensure that frail older people can live independently as long as possible. The proportion of frail older people will increase rapidly, therefore it is expected that the number of caregivers of frail older people at risk of caregiver burden will also increase. Caregiver burden can lead to various problems. No prediction model has been developed for caregiver burden of informal caregivers of frail older people. Objectives To identify predictors and develop a prediction model to predict caregiver burden in informal caregivers of community dwelling frail older people. Method The study included 358 caregivers and frail older people. Caregivers of frail older people who participated in the U-PROFIT trial were recruited. Candidate predictors and the outcome caregiver burden were measured with a questionnaire at baseline and after one year follow-up. Associations between candidate predictors and caregiver burden were examined using univariate linear regression analysis. A backward method in logistic regression resulted in a reduced model. The AUC was examined to assess the discriminative ability of the model. Results The significant predictors in the final model were the ability to combine care tasks with the daily activities of the caregiver and having relational problems with the care receiver. The model showed an AUC of 0.68. Conclusion The ability to combine the care tasks with the daily activities of the caregiver and having relational problems were the most important predictors for caregiver burden in caregivers of frail older people at one year follow-up. A clinical prediction model was developed. Recommendations Future research should focus on the internal validation of the prediction model and transforming the model into a clinical useful and valid instrument that may prevent negative health outcomes for the informal caregivers.