Describing and analyzing admission decisions in skilled nursing facilities using large-scale health informatics data (Preprint)

Abstract

Occupancy rates within Skilled Nursing Facilities (SNFs) in the United States have reached record lows. An increase in occupancy is considered a principal measure of recovery within the long-term care sector. Thus, there is incentive for SNFs to increase the number of patients moving into their facility if possible, which often entails accepting more referred patients. In order to better understand drivers of referral patterns, we provide the first comprehensive analysis of financial, clinical, and operational factors that impact whether a referral is accepted or denied by a SNF using a large health informatics database.

Our key objectives are to provide: 1) A description of the distribution of referrals sent to SNFs in terms of key patient- and facility-level features; 2) An analysis of key financial, clinical, and operational variables and their impacts on referral patterns; and, 3) A discussion identifying key potential reasons behind referral decisions in the context of learning health systems.

We extract and clean referral data, including information about SNF daily operations (occupancy and nursing hours), patient-level factors (insurance type and primary diagnosis), and facility-level factors (overall five-star rating and urban versus rural status). We compute descriptive statistics and apply regression modelling to identify and describe the relationships between these factors and referral decisions, considering them individually and controlling for other factors in order to fully understand their impact within the decision-making process.

Referrals from 627 SNFs across 41 US states from January 2020 to March 2022 are obtained. In analyzing daily operation values, we do not observe SNF occupancy or nursing hours having any significance in regards to a referral being accepted or denied. In analyzing patient-level factors, we find that both primary diagnosis category and insurance type of the patient are significant in relation to the acceptance or denial of a referral (P<0.05). Finally, in analyzing facility-level factors, we again find that overall five-star rating as well as urban versus rural status of a SNF are significant in relation to the acceptance or denial of a referral (P<0.05). Our work extends, demonstrating which categories within these features relates to referrals being more commonly accepted or denied.

While many factors may influence a denial decision, care challenges associated with individual diagnosis and financial challenges associated with different remuneration types are the strongest drivers. Understanding referral patterns is an essential step in being more intentional in the process of accepting or denying referrals. We interpret our results within an adaptive leadership framework to help guide SNFs in being more purposeful with their decisions as they strive to realize appropriate occupancy; that is, to increase their occupancy in ways that help facilities meet their goals in connection with the needs of patients.