Internal consistency, sensitivity, specificity and predictive value of the short version of the Zarit Burden Interview-7 scores: systematic review and generalization meta-analysis.

OBJECTIVES

The Zarit Burden Interview (ZBI) is one of the most widely used instruments for assessing caregiver burden. One of the most used short scales of the ZBI is the 7-item version. The aim of the study was to conduct a systematic review and generalization meta-analysis of the reliability and validity of the scores of the short version ZBI-7 items to determine whether they provide average values that justify their application and, secondly, to identify the characteristics of the studies that could explain the heterogeneity among the estimates.

METHODS

Forty-eight articles were identified that applied the short version of the ZBI-7, of which only 14 reported a reliability coefficient and 7 reported sensitivity, specificity and predictive values. A random effects model was applied in the statistical analyses.

RESULTS

The average Cronbach's alpha coefficient was 0.866 (95 % CI [0.835, 0.891], k = 15), with high heterogeneity (I2 = 86.4 %). Mean sensitivity was 98.6 % (95 % CI [70.0 %-95.3 %], k = 3), mean specificity was 87.4 % (95 % CI [93.6 %-99.7 %], k = 3), mean positive predictive value was 93.7 % (95 % CI [81.3 %-98.1 %], k = 3), mean negative predictive value was 97.3 % (95 % CI [87.5, 99.4], k = 3).

CONCLUSIONS

According to psychometric theory, the ZBI-7 can be considered reliable and valid for exploratory purposes in both research and clinical practice.