Multicentre methodological study to create a publicly available score of hospital financial standing in the USA

Objectives To create a straightforward scoring procedure based on widely available, inexpensive financial data that provides an assessment of the financial health of a hospital.Design Methodological study.Setting Multicentre study.Participants All hospitals and health systems reporting the required...

Full description

Saved in:
Bibliographic Details
Main Authors: Harlan M Krumholz, Radoslav Zinoviev, Richard Ciccarone, Rick Antle, Howard P Forman
Format: Article
Language:English
Published: BMJ Publishing Group 2021-07-01
Series:BMJ Open
Online Access:https://bmjopen.bmj.com/content/11/7/e046500.full
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Objectives To create a straightforward scoring procedure based on widely available, inexpensive financial data that provides an assessment of the financial health of a hospital.Design Methodological study.Setting Multicentre study.Participants All hospitals and health systems reporting the required financial metrics in the USA in 2017 were included for a total of 1075 participants.Interventions We examined a list of 232 hospital financial indicators and used existing models and financial literature to select 30 metrics that sufficiently describe hospital operations. In a set of hospital financial data from 2017, we used principal coordinate analysis to assess collinearity among variables and eliminated redundant variables. We isolated 10 unique variables, each assigned a weight equal to the share of its coefficient in a regression onto Moody’s Credit Rating, our predefined gold standard. The sum of weighted variables is a single composite score named the Yale Hospital Financial Score (YHFS).Primary outcome measures Ability to reproduce both financial trends from a ‘gold-standard’ metric and known associations with non-fiscal data.Results The validity of the YHFS was evaluated by: (1) cross-validating it with previously excluded data; (2) comparing it to existing models and (3) replicating known associations with non-fiscal data. Ten per cent of the initial dataset had been reserved for validation and was not used in creating the model; the YHFS predicts 96.7% of the variation in this reserved sample, demonstrating reproducibility. The YHFS predicts 90.5% and 88.8% of the variation in Moody’s and Standard and Poor’s bond ratings, respectively, supporting its validity. As expected, larger hospitals had higher YHFS scores whereas a greater share of Medicare discharges correlated with lower YHFS scores.Conclusions We created a reliable and publicly available composite score of hospital financial stability.
ISSN:2044-6055