JMIR Publications

ABSTRACT

Background:

Accurate identification of advanced CKD cohort in clinical databases is complicated and often unreliable. Identifying these patients in clinical databases can allow targeting this population for their specialized clinical and research needs.

Objective:

We aimed to examine the reliability of conventionally used diagnosis codes and eGFR-based methods to identify advanced CKD patietns in a regional Veteran healthcare system (VHS) database.

Methods:

Using the VA Informatics and Computing Infrastructure services, we extracted the source cohort of Veterans with advanced CKD in a regional VHS based on a combination of the latest eGFR value ≤ 30 ml/min/1.73m2 or existing ICD-10 diagnosis code for stage 4 or 5 CKD in the last 12 months. We estimated the prevalence of advanced CKD using the ICD-10 diagnosis codes and various eGFR-based definitions published in the literature. Finally, we categorized the cohort into operational definitions of high, intermediate, and low probability advanced CKD based on the combination of eGFR values and diagnosis codes and evaluated the reliability by examining the likelihood of sustained reduction of eGFR  30 over a 6-month follow-up

Results:

Of the 133,756 active Veteran enrollees at NF/SG VHS, we identified 93,216 Veterans having a measured creatinine value and a source cohort of 1,759 Veterans with advanced non-dialysis CKD with a latest (index) eGFR ≤30 ml/min/1.73m2 or a diagnosis code for stage 4 or 5 CKD. Of the 1759 Veterans in the source cohort, only 1,102(62.9%) were recognized to have advanced CKD by diagnosis codes, whereas 1,391(79.1%) had the index eGFR < 30 ml/min/1.73m2. The prevalence of advanced CKD among Veterans at NF/SG VHS varied between (1-1.5%) based on the identification method of advanced CKD. The sensitivity and positive predictive value of the diagnosis codes compared to eGFR based definitions of advanced CKD varied between 55-65% and 55-74%, respectively. Overall, 928, 480 and 315 Veterans were identified to have high, intermediate, and low probability advanced CKD, respectively, with 94.2% of high probability, 71% of intermediate, and 16.1% of low probability groups continuing to remain in advanced CKD stage (eGFR < 30) at 6- month follow up.

Conclusions:

The prevalence of advanced CKD is higher among Veterans at NF/SG VHS compared to general population. The accuracy of identifying advanced CKD cohort by ICD-based diagnosis codes is low in the VA clinical database. We report a simplified and pragmatic EHR-based model to identify advanced CKD within a regional VHS in real-time with a tiered approach that allows allocation of resources to the groups requiring immediate attention and are at risk of progression to end stage kidney disease