|
|
||||||||
TRACK to:
|
|
Electronic letters published:
|
|
|||
|
Jacob M Reider, Albany NY, US Family Physician
Send response to journal:
|
While the results of this study are not surprising, I would expect that if a similar study was done in 2004 the results would be somewhat different - as adoption of information technology has increased recently. Nonetheless, the study misses an important point about the use of information technology as a research tool - the presence of data in electronic form is only the first step in building a good understanding of what's going on in the clinical encounter. For example, without a well-defined clinical nomenclature, it is hard to characterize outpatient visits in a meaningful way. Billing codes are not clinician-friendly. Even with an Electronic Health record, two physicians may chose two different ICD-9 codes for the same clinical diagnosis. Would this mislead a researcher? Perhaps. The recent addition of SNOMED to UMLS is a welcome improvement. If software vendors incorporate the SNOMED terms into their products, the accuracy of clinicians' coding will improve - as will the accuracy of any research based on such information. Competing interests: None declared |
|||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |