Elan D Louis1, Lakeisha M Applegate, Eileen Rios. 1. Gertrude H. Sergievsky Center, College of Physicians and Surgeons, Columbia University, New York, USA. EDL2@columbia.edu
Abstract
BACKGROUND/AIMS: Health outcomes research often uses administrative databases. Patients with the diseases of interest are identified using International Classification of Diseases (ICD-9 CM) codes. The utility of the code for essential tremor (ET), 333.1, remains untested. We determined the positive predictive value (PPV) of the code 333.1. METHODS: Patients with the ICD-9 CM code 333.1 were identified from billing records at the Neurological Institute of New York. Their medical records were reviewed to determine whether they met Consensus Criteria for ET. RESULTS: Of 964 patients who carried the code 333.1, only 472 met diagnostic criteria for ET (i.e. PPV = 49.0%). The additional use of ICD 9-CM codes for parkinsonism and dystonia (as exclusionary criteria) only marginally improved this value (57.8%). Common diagnoses among the false positives were Parkinson's disease, dystonia, enhanced physiological tremor, drug-induced tremor, orthostatic tremor, and psychogenic tremor. Patients seen by general neurologists (vs. movement disorder specialists) were half as likely to meet diagnostic criteria for ET (34.6 vs. 51.9%, OR 0.50, 95% CI 0.23-0.70, p < 0.001). CONCLUSION: The code 333.1 performed poorly when attempting to identify ET cases. Given the very high prevalence of ET, a unique diagnostic code would seem to be in order.
BACKGROUND/AIMS: Health outcomes research often uses administrative databases. Patients with the diseases of interest are identified using International Classification of Diseases (ICD-9 CM) codes. The utility of the code for essential tremor (ET), 333.1, remains untested. We determined the positive predictive value (PPV) of the code 333.1. METHODS:Patients with the ICD-9 CM code 333.1 were identified from billing records at the Neurological Institute of New York. Their medical records were reviewed to determine whether they met Consensus Criteria for ET. RESULTS: Of 964 patients who carried the code 333.1, only 472 met diagnostic criteria for ET (i.e. PPV = 49.0%). The additional use of ICD 9-CM codes for parkinsonism and dystonia (as exclusionary criteria) only marginally improved this value (57.8%). Common diagnoses among the false positives were Parkinson's disease, dystonia, enhanced physiological tremor, drug-induced tremor, orthostatic tremor, and psychogenic tremor. Patients seen by general neurologists (vs. movement disorder specialists) were half as likely to meet diagnostic criteria for ET (34.6 vs. 51.9%, OR 0.50, 95% CI 0.23-0.70, p < 0.001). CONCLUSION: The code 333.1 performed poorly when attempting to identify ET cases. Given the very high prevalence of ET, a unique diagnostic code would seem to be in order.
Authors: Karen Curtin; Annette E Fleckenstein; Reid J Robison; Michael J Crookston; Ken R Smith; Glen R Hanson Journal: Drug Alcohol Depend Date: 2014-11-16 Impact factor: 4.492
Authors: Sophia Miryam Schüssler-Fiorenza Rose; Dawei Xie; Joel E Streim; Qiang Pan; Pui L Kwong; Margaret G Stineman Journal: BMC Health Serv Res Date: 2016-10-01 Impact factor: 2.655