Dataset Information

A Machine Learning Approach to Predicting Higher COVID-19 Care Burden in the Primary Care Safety Net: Hispanic Patient Population Size a Key Factor.

ABSTRACT:

Introduction

The federal government legislated supplemental funding to support community health centers (CHCs) in response to the COVID-19 pandemic. Supplemental funding included standard base payments and adjustments for the number of total and uninsured patients served before the pandemic. However, not all CHCs share similar patient population characteristics and health risks.

Objective

To use machine learning to identify the most important factors for predicting whether CHCs had a high burden of patients diagnosed with COVID-19 during the first year of the pandemic.

Methods

Our analytic sample included data from 1342 CHCs across the 50 states and D.C. in 2020. We trained a random forest (RF) classifier model, incorporating 5-fold cross-validation to validate the RF model while optimizing the model's hyperparameters. Final performance metrics were calculated following the application of the model that had the best fit to the held-out test set.

Results

CHCs with a high burden of COVID-19 had an average of 65.3 patients diagnosed with COVID-19 per 1000 patients in 2020. Our RF model had 80.9% accuracy, 80.1% precision, 25.0% sensitivity, and 98.1% specificity. The percentage of Hispanic patients served in 2020 was the most important feature for predicting whether CHCs had high COVID-19 burden.

Conclusions

Findings from our RF model suggest patient population race and ethnicity characteristics were most important for predicting whether CHCs had a high burden of patients diagnosed with COVID-19 in 2020, though sensitivity was low. Enhanced support for CHCs serving large Hispanic patient populations may have an impact on addressing future COVID-19 waves.

SUBMITTER: Goldstein EV

PROVIDER: S-EPMC9354120 | biostudies-literature | 2022 Jan-Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A Machine Learning Approach to Predicting Higher COVID-19 Care Burden in the Primary Care Safety Net: Hispanic Patient Population Size a Key Factor.

Goldstein Evan V EV Wilson Fernando A FA

Health services research and managerial epidemiology 20220101

<h4>Introduction</h4>The federal government legislated supplemental funding to support community health centers (CHCs) in response to the COVID-19 pandemic. Supplemental funding included standard base payments and adjustments for the number of total and uninsured patients served before the pandemic. However, not all CHCs share similar patient population characteristics and health risks.<h4>Objective</h4>To use machine learning to identify the most important factors for predicting whether CHCs ha ...[more]

PMID: 35937952

Similar Datasets

Project description:BackgroundQuality metrics, pay for performance (P4P), and value-based payments are prominent aspects of the current and future American healthcare system. However, linking clinic payment to clinic quality measures may financially disadvantage safety-net clinics and their patient population because safety-net clinics often have worse quality metric scores than non-safety net clinics. The Minnesota Safety Net Coalition's Quality Measurement Enhancement Project sought to collect data from primary care providers' (PCPs) experiences, which could assist Minnesota policymakers and state agencies as they create a new P4P system. Our research study aims are to identify PCPs' perspectives about 1) quality metrics at safety net clinics and non-safety net clinics, 2) how clinic quality measures affect patients and patient care, and 3) how payment for quality measures may influence healthcare.MethodsQualitative interviews with 14 PCPs (4 individual interviews and 3 focus groups) who had worked at both safety net and non-safety net primary care clinics in Minneapolis-St Paul Minnesota USA metropolitan area. Qualitative analyses identified major themes.ResultsThree themes with sub-themes emerged. Theme #1: Minnesota's current clinic quality scores are influenced more by patients and clinic systems than by clinicians. Theme #2: Collecting data for a set of specific quality measures is not the same as measuring quality healthcare. Subtheme #2.1: Current quality measures are not aligned with how patients and clinicians define quality healthcare. Theme #3: Current quality measures are a product of and embedded in social and structural inequities in the American health care system. Subtheme #3.1: The current inequitable healthcare system should not be reinforced with financial payments. Subtheme #3.2: Health equity requires new metrics and a new healthcare system. Overall, PCPs felt that the current inequitable quality metrics should be replaced by different metrics along with major changes to the healthcare system that could produce greater health equity.ConclusionAligning payment with the current quality metrics could perpetuate and exacerbate social inequities and health disparities. Policymakers should consider PCPs' perspectives and create a quality-payment framework that does not disadvantage patients who are affected by social and structural inequities as well as the clinics and providers who serve them.

Project description:Massachusetts' health care reform substantially decreased the percentage of uninsured residents. However, less is known about how reform affected access to care, especially according to insurance type.To assess access to care in Massachusetts after implementation of health care reform, based on insurance status and type.We surveyed a convenience sample of 431 patients presenting to the Emergency Department of Massachusetts' second largest safety net hospital between July 25, 2009 and March 20, 2010.Demographic and clinical characteristics, insurance coverage, measures of access to care and cost-related barriers to care.Patients with Commonwealth Care and Medicaid, the two forms of insurance most often newly-acquired under the reform, reported similar or higher utilization of and access to outpatient visits and rates of having a usual source of care, compared with the privately insured. Compared with the privately insured, a significantly higher proportion of patients with Medicaid or Commonwealth Care Type 1 (minimal cost sharing) reported delaying or not getting dental care (42.2 % vs. 27.1 %) or medication (30.0 % vs. 7.0 %) due to cost; those with Medicaid also experienced cost-related barriers to seeing a specialist (14.6 % vs. 3.5 %) or getting recommended tests (15.6 % vs. 5.9 %). Those with Commonwealth Care Types 2 and 3 (greater cost sharing) reported significantly more cost-related barriers to obtaining care than the privately insured (45.0 % vs. 16.0 %), to seeing a primary care doctor (25.0 % vs. 6.0 %) or dental provider (58.3 % vs. 27.1 %), and to obtaining medication (20.8 % vs. 7.0 %). No differences in cost-related barriers to preventive care were found between the privately and publicly insured.Access to care improved less than access to insurance following Massachusetts' health care reform. Many newly insured residents obtained Medicaid or state subsidized private insurance; cost-related barriers to access were worse for these patients than for the privately insured.

Dataset Information

A Machine Learning Approach to Predicting Higher COVID-19 Care Burden in the Primary Care Safety Net: Hispanic Patient Population Size a Key Factor.

Introduction

Objective

Methods

Results

Conclusions

Publications

A Machine Learning Approach to Predicting Higher COVID-19 Care Burden in the Primary Care Safety Net: Hispanic Patient Population Size a Key Factor.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets