2023 HSR&D/QUERI National Conference

1181 — Impacts of the ICD-9 to ICD-10 Transition on Diagnostic Coding of Post-Traumatic Headache in the Veterans Health Administration

Lead/Presenter: Jason Sico ,  VA Headache Centers of Excellence, Research and Evaluation Center
All Authors: Kimber AL (VA Headache Centers of Excellence, Research and Evaluation Center, West Haven, CT; Yale University), Fenton BT (VA Headache Centers of Excellence, Research and Evaluation Center) Phadke M (VA Headache Centers of Excellence, Research and Evaluation Center, Yale University SOM) Wang K (VA Headache Centers of Excellence, Research and Evaluation Center, Yale University SOM) Ney JP (VA Headache Centers of Excellence, Research and Evaluation Center) Seng EK (VA Headache Centers of Excellence, Research and Evaluation Center, Yeshiva University, Albert Einstein SOM) Lorenze NP (VA Headache Centers of Excellence, Research and Evaluation Center, Yale University SOM) Grinberg AS (VA Headache Centers of Excellence, Research and Evaluation Center, Yale University SOM) Sico JJ (VA Headache Centers of Excellence, Research and Evaluation Center, Yale University SOM )

Objectives:
This study aims to understand how transition from the 9th version of the International Classification of Diseases (ICD-9) to the 10th version (ICD-10) affected coding of PTH in VHA. We examined both patient and clinic level factors influencing PTH coding, and the agreement between clinical impression and ICD-codes.

Methods:
This retrospective chart review, combined with administrative data, identified 2,195 patients with a PTH diagnosis in the last two years of ICD-9 (10/01/13-09/30/15). Each veteran had 2 outpatient visits from the last 2 fiscal years of ICD-9 and another 2 from the first 2 years of ICD-10 (10/01/13-09/30/17). Medicare Generalized Equivalency Maps were used to determine equivalent ICD-9 and ICD-10 codes. The abstracted clinical impression was compared with VHA electronic health record data diagnoses for one headache visit in ICD-10. Agreement was measured with Cohen’s kappa. Patients with PTH diagnosis in ICD-9 were classified based on whether they had 1.) a PTH code in both ICD-9 and 10 (retained diagnosis), 2.) a headache diagnosis other than PTH in ICD-10 (changed diagnosis), or 3.) no headache diagnosis coded in ICD-10 (lost diagnosis). Descriptive statistics and chi-squared tests were completed in SAS 9.4.

Results:
Only 18.3% retained their PTH diagnosis in ICD-10 (N = 401), while 45.4% of patients had no headache diagnosis coded in the ICD-10 period. Among the 36.3% with a changed diagnosis in ICD-10 (N = 797), the majority were coded to either headache not-otherwise specified (NOS) (55.6%) and/or migraine (50.7%). In the patients with a reviewed clinical impression, there was moderate agreement with the ICD-10 code (PTH ? = .53, 95% CI [0.48,0.59], Migraine ? = .54 CI [0.48,0.60], NOS ? = .46 CI [0.40,0.52]).13.6% of patients with a clinical impression of PTH in ICD-10 did not have the ICD-10 diagnosis. In both time periods and for all 3 ICD-10 classifications, primary care was seen most frequently. Women were more likely to have a PTH diagnosis change to another headache type in ICD-10, while men were more likely to retain PTH diagnosis or have no headache coded in ICD-10. Older patients were significantly more likely to have no headache diagnosis coded in ICD-10. Patients seen by primary care, neurology, and physical medicine and rehabilitation were most likely to have their PTH diagnosis changed in ICD-10 (44.5%, 50.5%, and 37.6%). Patients with PTH in ICD-9 that were seen in ICD-10 by psychiatry, the emergency room, or pain clinic were more likely to have no headache diagnoses coded (92.1%, 78.1%, and 53.9%).

Implications:
With the ICD-9 to ICD-10 transition, PTH coding is inconsistent with variation by gender, age, and clinic type. In the selected short time window, changes are unlikely due to resolved PTH. Coding issues are likely such as 1.) diagnosing headache phenotype rather than assessing secondary headache, 2.) assigning unspecific headache diagnoses rather than reviewing patient notes, and/or 3.) recoding reimbursable claims due to insufficient documentation.

Impacts:
Evidence from this project demonstrate that efforts to improve coding and documentation are warranted to ensure accuracy of PTH and access to specialized care which fall under the mandate of the Headache Centers of Excellence.