Billing Codes In Big Data Often Misidentify Diagnoses

Use of billing codes successful large information sets to find diagnoses tin consequence successful up to two-thirds of cases being mistakenly identified, caller UCLA-led investigation finds.

Databases often utilized for aesculapian investigation specified arsenic those for nan Centers for Medicare & Medicaid Services aliases nan National Inpatient Survey typically trust connected ambulatory billing codes to place diseases aliases aesculapian procedures, but their accuracy is seldom verified successful publications that trust connected this data, nan researchers constitute successful a study published successful nan peer-reviewed diary British Journal of Surgery.

Though nan findings successful this insubstantial focused connected hernia diagnoses, reliance connected billing codes successful investigation reports tin lead to akin discrepancies pinch different diseases aliases conditions, said Dr. Edward Livingston, wellness sciences professor of room astatine the David Geffen School of Medicine astatine UCLA and nan investigation letter's elder author.

Researchers often presume that if a codification appears for a definite test successful 1 of these large information sets that nan illness genuinely is present. Our investigation demonstrates that is not nan lawsuit successful galore instances. Research relying connected these codes to place diseases whitethorn lead to mendacious conclusions because of this problem."

Dr. Edward Livingston, wellness sciences professor of surgery, David Geffen School of Medicine, UCLA

The researchers examined records for 1.36 cardinal patients, of whom 41,700 were diagnosed pinch hernias based connected nan coding-12,800 (45%) pinch diaphragmatic hernias, 7000 (24%) pinch ventral hernias, and 8,800 (31%) pinch inguinal hernias.

But nan researchers had corresponding images for 28,600 of code-based diagnosed patients. Of those, nan images verified that 10,234 (36%) really had hernias; 4,325 (34%) diaphragmatic hernias, 3,069 (44%) were ventral hernias, and 2,840 (32%) were inguinal hernias.

The researchers fishy that nan discrepancy stems from physicians basing their coding connected nan objective problem for which they initially examined nan diligent and not connected what they yet found. For instance, a diligent sojourn for a imaginable hernia will beryllium coded arsenic a hernia successful nan grounds and remains that measurement moreover if that first test is ruled retired during consequent examination.

"These findings item a basal weakness successful utilizing administrative information for illness identification," nan researchers write. "Encounter coding occurs because a test is considered, and not needfully proven. We recovered that reliance connected billing codes for hernia recognition could consequence successful 2/3 of cases being erroneously identified. This rumor extends beyond hernia, highlighting a superior limitation successful utilizing administrative information for objective research. Validation of coding accuracy against existent illness beingness is basal earlier assuming test validity."

Hila Zilicha, Dr. Douglas Bell, and Dr. Yijun Chen co-authored nan paper.

The UCLA Department of Surgery Research Funds and nan National Center for Advancing Translational Science (NCATS) of nan National Institutes of Health nether nan UCLA Clinical and Translational Science Institute (UL1TR001881) funded nan research.