Background/Purpose: Electronic Medical Records (EMR) offer great opportunities for pharmacoepidemiologic, health outcome and health services research. However, the critical limiting factor in the widespread use of these data is the accuracy, precision and validity of diagnoses. EpicCare is a large and growing proprietary EMR software system that has been adopted by several large tertiary care facilities.The goal of this FDA funded project was to assess the validity of an ICD code of 274.* to identify patients meeting a clinical diagnosis of gout from among those who have ever been prescribed colchicine.
Methods: We identified 143 patients in the Stanford EpicCare system in the past 6 years that met our case definition for gout -at least one prescription of colchicine and one instance of ICD-9 code of 274.*. The records of these patients were individually reviewed and data on the following aspects were abstracted: physician diagnosis (notes), performance of arthrocentesis, use of urate lowering therapy, and documentation of each of the American College of Rheumatology (ACR) criteria for diagnosis/classification of gout. Data were analyzed quantitatively and qualitatively.
Results: Overall 143 case records were reviewed, of which 3 did not have any physician authored clinical documents. A physician diagnosis of gout was documented in 114 (80%). Three charts revealed a diagnosis of pseudogout and 31 did not have any physician documentation of gout. Among those records with a physician diagnosis of gout, 35 records documented intra-articular urate crystals, 36 records had documentation that met the ACR criteria, and 19 records had documentation of ACR criteria and urate crystals. The median number of ACR criteria met in the 114 charts was 4, with an interquartile range of 2 to 6. Among those records with a positive crystal identification, only 48% had documentation that met the ACR criteria. Among those records where arthrocentesis was performed and found to be negative for crystals, physician diagnosis was documented in 90%, and ACR criteria were met in 40%. In the case records that did not show documentation of arthrocentesis or any relevant laboratory evaluation, 75% had physician diagnosis and about 13% met the American College of Rheumatology criteria.
Conclusion: This study showed that in the setting of a tertiary medical center, Electronic Medical Records are an excellent resource for gout research. Documentation of the individual ACR criteria and performance of arthrocentesis is infrequent and hence these may not be useful gold standards for validation studies of gout in the EMR. From our qualitative review, documentation of physician diagnosis may be considered a useful benchmark for assessing the utility of case definitions for gout.
Disclosure:
N. Narang,
None;
E. Krishnan,
savient,
1,
URL, takeda, metbolex,ARDEA,
2,
METABOLEX TAKEDA,
5.
« Back to ACR/ARHP Annual Meeting 2012
ACR Meeting Abstracts - https://acrabstracts.org/abstract/valdiation-of-a-diagnosis-of-gout-in-the-epiccare-electronic-medical-records/