Methodologies to Measure Upcoding in Provider Settings: A Scoping Review

Format:

Book

Author/Creator:

Levin, Jonathan S.

Contributor:

Crespin, Daniel J.

Dworsky, Michael

Kim, Jin

Reid, Rachel O.

Whaley, Christopher M.

Language:

English

Other Title:

Methodologies to Measure Upcoding in Provider Settings

Place of Publication:

RAND Corporation 2024

Summary:

Increases in payment for more-complicated patients can incentivize upcoding, a practice in which hospitals code more secondary diagnoses or complications to classify admissions or visits at higher complexity levels. However, the lack of standardized methodologies to measure upcoding hinders researchers and policymakers from fully understanding its prevalence, the medical conditions especially likely to be upcoded, the types of hospitals and geographic areas with high upcoding rates, and the overall impact of upcoding on health care spending. To address this gap, the authors of this report provide a review of methodologies to measure upcoding using literature from 2000 to 2023. Using seven electronic databases, the authors identified 39 articles that measure upcoding, which they classified into the following categories: variations in payments (n = 14 studies), variations in provider and patient attributes (n = 6), prediction algorithms (n = 6), self-reports (n = 3), and validations (n = 10). In most of these studies (n = 26), researchers examine upcoding at aggregated levels (e.g., frequency of upcoding at hospitals or in geographic areas) rather than identifying individually upcoded admissions (n = 13). Only three studies measure upcoding by comparing coding intensity with a gold standard, while the rest examine upcoding indirectly via measures of coding intensity. In their conclusion, the authors recommend that researchers and policymakers designing studies on provider-based upcoding consider outcome measures based on severity levels, payment changes that incentivize upcoding, provider and patient characteristics associated with upcoding, and indicators in administrative claims or electronic health record data.