Comparison of Active and Passive Surveillance for Cerebrovascular Disease: The Brain Attack Surveillance in Corpus Christi (BASIC) Project

Abstract

To provide a scientific rationale for choosing an optimal stroke surveillance method, the authors compared active surveillance with passive surveillance. The methods involved ascertaining cerebrovascular events that occurred in Nueces County, Texas, during calendar year 2000. Active methods utilized screening of hospital and emergency department logs and routine visiting of hospital wards and out-of-hospital sources. Passive means relied on International Classification of Diseases, Ninth Revision (ICD-9), discharge codes for case ascertainment. Cases were validated by fellowship-trained stroke neurologists on the basis of published criteria. The results showed that, of the 6,236 events identified through both active and passive surveillance, 802 were validated to be cerebrovascular events. When passive surveillance alone was used, 209 (26.1%) cases were missed, including 73 (9.1%) cases involving hospital admission and 136 (17.0%) out-of-hospital strokes. Through active surveillance alone, 57 (7.1%) cases were missed. The positive predictive value of active surveillance was 12.2%. Among the 2,099 patients admitted to a hospital, passive surveillance using ICD-9 codes missed 73 cases of cerebrovascular disease and mistakenly included 222 noncases. There were 57 admitted hospital cases missed by active surveillance, including 13 not recognized because of human error. This study provided a quantitative means of assessing the utility of active and passive surveillance for cerebrovascular disease. More uniform surveillance methods would allow comparisons across studies and communities.

Use and Utility of Administrative Health Data for Stroke Research and Surveillance

“…In the Brain Attack Surveillance in Corpus Christi project, the authors compared active and passive surveillance of cerebrovascular events in Nueces County, Texas, during a full calendar year. 27 The active surveillance arm consisted of screening, in real-time, admission, and emergency department visit logs in the 6 hospitals of the region, using 8 cerebrovascular screening diagnostic terms as well as routine visits to medical offices, nursing homes, radiology centers, and the office of the medical examiner. Passive surveillance used ICD-9 discharge codes.…”

Accuracy of ICD‐9‐CM Codes by Hospital Characteristics and Stroke Severity: Paul Coverdell National Acute Stroke Program

“…Disease surveillance and quality‐of‐care studies frequently use information from administrative claims databases, such as Medicare, by identifying patients with specific conditions based on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) codes. Studies comparing the accuracy of ICD‐9‐CM codes with medical record review find variations in agreement depending on code selection,2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 as well as patient characteristics, such as age, race/ethnicity,10, 11 sex,11 and length of hospitalization 3. Aside from teaching hospital status10 and hospital departments,16 potential variations in accuracy based on other hospital characteristics are largely unknown.…”

Emerging temporal trends in tissue plasminogen activator use

“…The study procedures were previously described in full, and are briefly described here. 12,16,17 In the active component, trained abstractors review hospital admission logs on a daily basis using validated screening terms. 18 In addition, the hospital wards and intensive care units are reviewed to identify in-house strokes and those missed upon admission.…”