Mai Skov Christiansen, Marie Ørskov, Ida Ehlers Albertsen, Thure Filskov Overvad
The positive predictive value of a cataract diagnosis in the Danish National Patient Register is high and suitable for use in epidemiological studies. Subtypes of cataract cannot be reliably identified in the register due to widespread use of unspecified cataract diagnoses.
PURPOSE: To assess the quality of International Classification of Diseases (ICD)-10 diagnoses for cataract in the Danish National Patient Register, which remains unexplored.
METHODS: We extracted 100 consecutive patients registered with a diagnosis of cataract (50 with H25 [senile cataract] and 50 with H26 [other cataract]). Positive predictive values (PPVs) were calculated as the proportion of diagnoses that were confirmed after manual medical record review. PPVs were calculated both overall and in prespecified subgroups.
RESULTS: The median patient age was 76 years, and most were females (61%). The overall PPV for a diagnosis of cataract (H25 or H26) in the Danish National Patient Register was 95.0% (95% confidence interval [CI] 88.8-97.8%). The PPV was 90% or above across all subgroups of the H25/H26 diagnoses, across sex, incident and recurrent diagnoses, and primary position diagnosis, but lower for secondary position diagnoses (40.0%, 95% CI 5.3-85.3%). In most cases, a more specific ICD-code was available but not used in clinical practice.
CONCLUSION: The positive predictive value of a cataract diagnosis in the Danish National Patient Register is high and suitable for use in epidemiological studies. Subtypes of cataract cannot be reliably identified in the register due to widespread use of unspecified cataract diagnoses.