Informative Presence Bias in Comorbidity Data of Medicare Advantage-Enrolled Beneficiaries

Background: Analyses of Medicare administrative claims data are faced with methodological challenges, including accounting for the potential effect of insurance status on documented comorbidities. We present an example of how failing to account for informative presence bias related to beneficiary enrollment status in such analyses may lead to flawed results. Methods: In this retrospective observational study of Medicare beneficiaries undergoing isolated coronary artery bypass grafting (CABG) from 1999-2019, we compare the distribution of documented comorbidities between beneficiaries with Medicare Advantage (MA) and Traditional Medicare (TM) plans. Long-term survival was then compared in both unweighted and overlap weighted analyses with and without the inclusion of documented comorbidities. Results: Among 3,015,066 Medicare beneficiaries undergoing CABG from 1999-2019, 2,345,476 underwent isolated CABG and had suitable data for analysis. The annual proportion of MA-enrolled beneficiaries undergoing CABG remained stable from 1999-2007 (1.1-4.5%) and then progressively increased annually, reaching 38.2% in 2019. The incidences of documented comorbidities were substantially lower among MA-enrolled versus TM-enrolled beneficiaries. Among MA-enrolled and TM-enrolled beneficiaries, respectively, the unweighted median survival difference was only 8 [-12,28] days (10.02 [9.96,10.07] vs 10.00 [9.98,10.01] years); the weighted (adjusted for demographics and procedural characteristics, but not beneficiary comorbidities) median survival difference was also minimal at -2 [-28,24] days (10.00 [9.95,10.06] vs 10.01 [9.98,10.04] years). However, the weighted (with adjustments including beneficiary comorbidities) median survival difference demonstrated a substantial survival disadvantage for MA-enrolled beneficiaries compared to their TM-enrolled counterparts: -604 [-626,-575] days (9.78 [9.73,9.83] vs 11.44 [11.41,11.47] years), respectively. Conclusions: Comorbidities in MA-enrolled beneficiaries may be severely under-reported in Medicare data. Studies failing to account for this are susceptible to informative presence bias with a significant treatment effect. In the absence of policy changes, increasing enrollment in MA plans will continue to decrease the population of Medicare beneficiaries with suitable data for study in comparative analyses. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement Data acquisition and effort of AK and JKB was supported by a philanthropic gift of Satish and Yasmin Gupta to Baylor Scott & White The Heart Hospital, Plano, TX ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The Baylor Scott & White Research Institute institutional review board provided approval (IRB#: 019-270, dated 8-23-2019). I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Due to a data use agreement with CMS, we are unable to provide the data referenced in our study.