Diagnosis coding in the Danish National Patient Registry for respiratory syncytial virus infections

This national registry-based epidemiological study aimed to evaluate the usefulness of the Danish National Patient Registry (DNPR) as a tool for epidemiological studies of respiratory syncytial virus (RSV) hospitalizations. Information on RSV diagnoses in records of hospitalizations among inpatients tested for RSV in Denmark from January 1996 to May 2003 in the DNPR was validated against the RSV test result in records from the 18 laboratories testing for RSV among hospitalized patients in Denmark. Of 16,733 RSV-positive samples representing a total of 14,898 hospitalizations in the DNPR, 68% (10,111) had been registered in the DNPR with 1 or more RSV diagnoses. Age influenced the odds ratio (OR) of being coded with a RSV diagnosis among patients with a RSV positive test, hence 73% of infants compared to 57% of patients beyond infancy had been registered in the DNPR with 1 or more RSV diagnoses. The OR of being coded with a RSV diagnosis among patients with an RSV positive test was increased over the study period and was also increased by increasing length of hospitalization and by the presence of secondary diagnoses. The OR was decreased by the presence of asthma and other chronic conditions in the patients. The OR was furthermore influenced by RSV seasonality and by hospital. Taking into account a loss of 27% of RSV hospitalizations, DNPR can be used for epidemiological studies of RSV among infants in Denmark.