(AJ) The Use of United States Administrative Claims Data to Study Patients with Friedreich Ataxia: The Impact of Specific and Non-Specific Coding Periods

Publication Number: 1034

Background: Friedreich Ataxia (FA) is a rare mitochondrial disorder characterized by progressive loss of muscle control resulting in impaired movement and speech. When the International Classification of Diseases (ICD) codes switched to the 10th revision (in Oct 2015), the FA-specific diagnosis code was removed and not reinstated until September 30, 2020. This period of non-specific coding complicates the design and implementation of retrospective claims analyses characterizing patients with FA.

Objectives: To explore ICD-10 codes captured during the non-specific period among patients who had specific FA diagnosis prior to the ICD-10 revision in October 2015.

Methods: This study used the United States IBM MarketScan Commercial database subset from Aug 2010 to Sept 2020. Patients with ≥1 inpatient, or ≥2 outpatient visits separated by ≥30 days, with a primary or secondary diagnosis of FA (ICD-9 334.0) prior to Oct 2015 were identified. The most frequently recorded primary ICD-10 diagnosis codes during the non-specific period were summarized, overall and among the subset of visits coded by a neurologist.

Results: A total of 669 patients with FA were identified from the period of Aug 1, 2010, to Sept 30, 2015. Among those with available follow-up during the non-specific period, the code for a group of conditions classified as early-onset cerebellar ataxia (ICD-10 G11.1) was the most frequently captured diagnosis code for inpatient and outpatient visits overall, and when coded by a neurologist. For outpatient neurologist diagnoses, G11.1 accounted for 46% of encounters, followed by codes G11.8 (other hereditary ataxias, 4%), and G11.3 (cerebellar ataxia with defective DNA repair, 4%). The top three inpatient neurologist codes were G11.1 (34%), M41.45 (neuromuscular scoliosis, 18%), and I49.01 (ventricular fibrillation, 14%).

Conclusions: This study confirms that identifying FA patients from administrative claims databases is challenging during the non-specific coding period. G11.1 was the most frequently used for primary diagnoses among patients with FA, however this code is also used for various other ataxias. Further analyses should be conducted to determine the predictability of these diagnostic codes, such as the development and validation of an identification algorithm, with expert input.