Use of electronic data to identify new zoster cases for a vaccine effectiveness study
In the pilot phase of a study of the long-term effectiveness of ZostavaxTM(live zoster vaccine indicated for the prevention of herpes zoster (HZ)), we examined the performance of combinations of diagnostic codes, medications, and laboratory tests in identifying new HZ cases.
In the Kaiser Permanente Northern California Region, we used computerized data to identify first HZ diagnoses (no HZ diagnoses in the previous 12 months) among study participants 50 years of age or older from 2007 through 2012. We then used ICD9 codes, “internal” diagnostic terms (more specific than ICD9 codes), health care setting (outpatient, emergency department, hospital), coding position (primary diagnosis or other), antiviral medications, and VZV and HSV laboratory tests to create mutually exclusive categories of potential cases which were expected to have a low, medium, or high predictive value for incident HZ. Representative samples of medical charts from each category were reviewed and diagnoses were adjudicated by two physicians independently, with discordant cases adjudicated by a third clinician
We identified 39,570 cases of HZ over the pilot study period, after restricting the diagnosis to cases that did not also have a post-herpetic neuralgia (PHN) internal diagnostic term at the time of their first HZ code. 36,417 (92%) cases were classified in the high category, which consisted of any HZ diagnoses with an antiviral medication or positive VZV lab result, or a primary HZ diagnosis so long as there was no indication of HSV. Of 200 adjudicated cases from the high category, 185 (92.5%) were confirmed as having received a diagnosis for incident HZ (95% CI, 87.9%-95.7%).
It is feasible to use electronic data to identify a large subset of new HZ diagnoses that are highly predictive of the diagnosis of new-onset HZ.
GSK: Investigator, Research grant
Merck: Investigator, Research grant
Pfizer: Investigator, Research grant
Novartis: Investigator, Research grant
MedImmune: Investigator, Research grant
Sanofi Pasteur: Investigator, Research grant
J. Bartlett, None
N. Lewis, None
B. Fireman, None
L. Aukes, None
M. Marks, Merck Sharp & Dohme, Corp.: Employee and Shareholder, Salary
P. Saddier, Merck Sharp & Dohme, Corp.: Employee and Shareholder, Salary