Ambulatory Follow-up and Outcomes Among Medicare Beneficiaries After Emergency Department Discharge

Key Points Question How often do Medicare beneficiaries have an ambulatory follow-up visit after discharge from the emergency department, and is ambulatory follow-up associated with postdischarge outcomes? Findings In this cohort study of 9 470 626 emergency department discharges from 2011 to 2016, most patients had ambulatory follow-up within 30 days, with lower rates among Medicaid-eligible beneficiaries, Black beneficiaries, and those treated at rural emergency departments. Ambulatory follow-up was associated with a higher risk of subsequent hospitalization but a lower risk of 30-day mortality. Meaning The findings of this study suggest that access to ambulatory care may be a key driver of outcomes among Medicare beneficiaries discharged from the emergency department.

Cox proportional hazards models with the time to each post-discharge event as the outcome and ambulatory follow-up as a timevarying covariate as the primary predictor. We incorporated beneficiary age, sex, race, Medicaid eligibility, year of the visit, principal diagnosis category and beneficiary chronic conditions as covariates. We adjusted for clustering by hospital. b Hazard ratio less than one means a longer time until the outcome event. c For the outcomes of ED visits and inpatient stays, we also incorporated mortality as a competing risk. Outcomes were censored at 30 days. Cox proportional hazards models with the time to each post-discharge event as the outcome and ambulatory follow-up as a timevarying covariate as the primary predictor. We incorporated beneficiary age, sex, race, Medicaid eligibility, year of the visit, principal diagnosis category and beneficiary chronic conditions as covariates. We adjusted for clustering by hospital. b Hazard ratio less than one means a longer time until the outcome event. c For the outcomes of ED visits and inpatient stays, we also incorporated mortality as a competing risk. Outcomes were censored at 30 days. Including burns, wounds, poisonings, superficial injuries Of note, for visits using ICD-10 codes, we used the Centers for Medicare and Medicaid Services (CMS) General Equivalence Mappings to convert all ICD-10 codes to ICD-9. We got this crosswalk from the National Bureau of Economic Research (https://data.nber.org/data/icd9-icd-10-cm-and-pcs-crosswalk-general-equivalencemapping.html), but CMS is the original source and they publish a new version each year.
We found that in most cases, an ICD-10 code maps directly to one ICD-9 code. There are some ICD-10 codes that have either have multiple approximate matches or should be represented by a set of codes, so we needed to develop logic to choose a single ICD-9 code, using flag variables that CMS provided-approximate, scenario, and choice list. The logic is summarized below. 1. If there is an exact match, keep that over an approximate match (approximate=0) 2. When there should be multiple codes to represent an ICD-10 code, keep the ICD-9 code that is listed first within the scenario (choice_list=1). 3. If there are still multiple codes after the above steps, choose the most common ICD-9 code, based on the count for that code in 2014 (the most recent full year of ICD-9 data).

Hospital Characteristics
The Following Hospital Characteristics were identified from the 2014 American Hospital Association Annual Survey: Size-Hospital size was determined based on the number of hospital beds. Hospitals with 1-99 beds were considered small hospitals, medium hospitals were those with 100-399 beds and large hospitals had 400 or more beds Region-Northeast, Midwest, South and West (other regions were excluded).
Control Type-For-Profit, Not-for-Profit, Government, Non-Federal Urban/Rural location-This was defined by the Core Based Statistical Area (CBSA). CBSA type of metropolitan or micropolitan was considered urban while rural CBSA type was considered rural.
Teaching Status-Major teaching hospitals were defined as those with membership in the Council of Teaching Hospitals (COTH), minor teaching hospitals were those without COTH membership but that reported a medical school affiliation to the American Medical Association. All other hospitals were designated as non-teaching.
Hospital safety-net status was defined using the Medicare Impact File disproportionate share patient percentage. This percentage measures the degree to which hospitals serve Medicaid beneficiaries and the uninsured.