Clinical Epidemiology (May 2019)
The positive predictive value of spontaneous subarachnoid hemorrhage diagnoses in the Danish National Patient Register
Abstract
Asger Sonne,1 Jesper B Andersen,2 Lars S Rasmussen11Department of Anaesthesia, Centre of Head and Orthopaedics, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark; 2Department of Neurosurgery, The Neuroscience Centre, Rigshospitalet, University of Copenhagen, Copenhagen, DenmarkIntroduction: Spontaneous (non-traumatic) subarachnoid hemorrhage (ntSAH) is frequently suspected in the emergency department, but the incidence rate is low. Diagnosis registers can provide valuable data for research in uncommon conditions like ntSAH. Unfortunately, validity vary in the registers. We aimed to assess the validity of the ntSAH diagnosis in the Danish National Patient Register (DNPR) and secondly to describe patients misclassified as ntSAH and to calculate the incidence rate of ntSAH.Methods: From the DNPR we extracted information on patients at least 18 years of age on admission registered with a diagnosis of ntSAH and admitted to a hospital in the Capital Region of Denmark between January 1, 2008 and December 31, 2014. Two independent investigators reviewed the medical records to assess if the diagnosis could be confirmed. Those not confirmed were categorized according to a predeveloped case report form.Results: We identified 1101 patients with a diagnosis of ntSAH; medical records were accessible for 1069 (97.7%) and 842 (78.8%) met the inclusion criteria. The diagnosis was confirmed in 537 patients (63.8% (95% confidence interval: 60.5–67.0%)). Among patients where ntSAH was not confirmed, 122 (40.0%) had a traumatic subarachnoid hemorrhage (tSAH), ntSAH had been suspected in 57 (18.9%) patients but was not substantiated during admission, while parenchymal hemorrhages were identified in 45 (14.5%) patients. The incidence rate was 5.5 (95% CI: 4.9–6.1) per 100,000 persons per year.Conclusion: We found a positive predictive value (PPV) of 63.8% for the diagnosis of non-traumatic subarachnoid hemorrhage in the Danish National Patient Register. The low PPV suggests that care should be taken if unvalidated data are used for research and that results of previous studies should be interpreted with caution.Keywords: subarachnoid hemorrhage, diagnosis register, diagnosis validation, incidence rate, non-traumatic, Danish National Patient Register
