News: ICD-10-CM codes can be used to identify incident stroke events, study finds

A cohort study published in the Journal of the American Medical Association (JAMA) found that the detection of incident stroke events is moderately accurate when using ICD-10-CM codes in Medicare claims and very accurate in ruling out non-stroke cases, JustCoding reported.

The study consulted information from two other studies, the Atherosclerosis Risk in Communities (ARIC) study and the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study, which included 5,194 and 6,359 participants, respectively. The participants were aged 65 years or older, had no prior strokes, and had linked Medicare inpatient fee-for-service claims data.

The study used Medicare-linked ARIC data to develop a list of diagnosis codes for incident stroke detection and was validated using Medicare-linked REGARDS data.

Highlights from the calculations using ARIC data include:

• Codes as the primary diagnosis had a sensitivity of 81.8% and specificity of 99.1%
• Any diagnosis code reported had a sensitivity of 94.5% and specificity of 98.4%

Highlights from the calculations using REGARDS data include:

• Codes as the primary diagnosis had a sensitivity of 70.7% and specificity of 99.1%
• Any diagnosis code reported had a sensitivity of 77.9% and specificity of 98.9%

Overall, ICD-10-CM codes can be used to identify incident stroke events in Medicare claims with moderate sensitivity and high specificity; however, the effectiveness of detection varies depending on the placement of the codes within the claim (i.e., primary, secondary, etc.).

Editor’s note: This article originally appeared in JustCoding. To read the full JAMA study, click here.