Exploring the validity of identifying care home residents through a new national register

Aims:
Identifying permanent care home residents in Denmark by national registers is subject to error. The current register-based method has a sensitivity of 87% and a positive predictive value of 57%. The Danish National Health Data Authority has generated a new register named Care Home Data (in Danish: Plejehjemsdata) to increase the quality of register-based studies on care home residents. This study aimed to investigate the validity of Care Home Data.

Methods:
We generated the gold standard by retrieving information from the four municipalities of Southern Jutland on all individuals living permanently in a care home facility in 2019. Care Home Data generates information on care home residents by pairing addresses of every apartment in Danish care home facilities with the addresses of Danish citizens. The agreement between Care Home Data and the gold standard was analysed by calculating the sensitivity and positive predictive value.

Results:
According to the municipalities, a total of 2081 individuals resided permanently in care home facilities in Southern Jutland in 2019 (gold standard). Care Home Data identified 2128 permanent care home residents; of which 2019 individuals were identified by both the municipalities and Care Home Data (true positives); 62 individuals were not identified by Care Home Data (false negatives), and 109 individuals identified by Care Home Data did not appear in data from the municipalities (false positives). This gave a sensitivity of Care Home Data of 97.0% and a positive predictive value of 94.9%.

Conclusions:
Care Home Data is a much improved tool for identifying citizens permanently residing in care homes with very high sensitivity and positive predictive value.