Dataset Information

Predicting health-related social needs in Medicaid and Medicare populations using machine learning.

ABSTRACT: Providers currently rely on universal screening to identify health-related social needs (HRSNs). Predicting HRSNs using EHR and community-level data could be more efficient and less resource intensive. Using machine learning models, we evaluated the predictive performance of HRSN status from EHR and community-level social determinants of health (SDOH) data for Medicare and Medicaid beneficiaries participating in the Accountable Health Communities Model. We hypothesized that Medicaid insurance coverage would predict HRSN status. All models significantly outperformed the baseline Medicaid hypothesis. AUCs ranged from 0.59 to 0.68. The top performance (AUC = 0.68 CI 0.66-0.70) was achieved by the "any HRSNs" outcome, which is the most useful for screening prioritization. Community-level SDOH features had lower predictive performance than EHR features. Machine learning models can be used to prioritize patients for screening. However, screening only patients identified by our current model(s) would miss many patients. Future studies are warranted to optimize prediction of HRSNs.

SUBMITTER: Holcomb J

PROVIDER: S-EPMC8927567 | biostudies-literature | 2022 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predicting health-related social needs in Medicaid and Medicare populations using machine learning.

Holcomb Jennifer J Oliveira Luis C LC Highfield Linda L Hwang Kevin O KO Giancardo Luca L Bernstam Elmer Victor EV

Scientific reports 20220316 1

Providers currently rely on universal screening to identify health-related social needs (HRSNs). Predicting HRSNs using EHR and community-level data could be more efficient and less resource intensive. Using machine learning models, we evaluated the predictive performance of HRSN status from EHR and community-level social determinants of health (SDOH) data for Medicare and Medicaid beneficiaries participating in the Accountable Health Communities Model. We hypothesized that Medicaid insurance co ...[more]

PMID: 35296719

Similar Datasets

Project description:BackgroundMultiple state and national health care organizations have invested in activities to screen for and address the health-related social needs (HRSNs) of their patients. However, patient perspectives concerning HRSN screening discussions and facilitated referrals to supports are largely unexplored. The main objectives of this study were to explore the ways in which Massachusetts Medicaid (MassHealth) members engage with their health care clinicians to discuss HRSNs, to identify common needs discussed, and to describe whether members feel these needs are being addressed by health care clinicians and staff.MethodsThe study team performed a cross-sectional, qualitative research study that included in-depth, open-ended interviews with 44 adult MassHealth members. Interviews were conducted between June and October 2022. Interviews were recorded, transcribed, and systematically coded for analysis, and common themes were reported. The data collected for this study were part of a larger independent evaluation of MassHealth's 2017-2022 Section 1115 Demonstration that granted authority from CMS to implement health care delivery system reforms in Massachusetts.ResultsIn this qualitative study of Medicaid members, some reportedly felt comfortable freely discussing all of their clinical and social needs with their health care clinicians, while others noted feelings of apprehension. Several members recalled being asked about their HRSNs in various clinical or community settings, while others did not. The majority of members endorsed having an unmet HRSN, including housing, nutrition, financial, or transportation issues, and many barriers to addressing these HRSNs were discussed. Finally, many members cited a preference for discussing HRSNs with community-based care coordinators and social workers at the community partner organizations rather than with their health care clinicians. Community-based care coordinators were lauded as essential facilitators in making the connection to necessary resources to help address HRSNs.ConclusionsStudy results highlight an opportunity to increase the effectiveness of HRSN screening and referral practices within the health care setting through relationship building between Medicaid members and diverse interdisciplinary care teams that include staff such as community health workers. Continued investment in cross-sector partnerships, screening workflows, and patient-clinician relationships may contribute to establishing an environment in which members can comfortably discuss HRSNs and connect with needed services to improve their health.

Project description:BackgroundWhile online health social networks (OHSNs) serve as an effective platform for patients to fulfill their various social support needs, predicting the needs of users and providing tailored information remains a challenge.ObjectiveThe objective of this study was to discriminate important features for identifying users' social support needs based on knowledge gathered from survey data. This study also provides guidelines for a technical framework, which can be used to predict users' social support needs based on raw data collected from OHSNs.MethodsWe initially conducted a Web-based survey with 184 OHSN users. From this survey data, we extracted 34 features based on 5 categories: (1) demographics, (2) reading behavior, (3) posting behavior, (4) perceived roles in OHSNs, and (5) values sought in OHSNs. Features from the first 4 categories were used as variables for binary classification. For the prediction outcomes, we used features from the last category: the needs for emotional support, experience-based information, unconventional information, and medical facts. We compared 5 binary classifier algorithms: gradient boosting tree, random forest, decision tree, support vector machines, and logistic regression. We then calculated the scores of the area under the receiver operating characteristic (ROC) curve (AUC) to understand the comparative effectiveness of the used features.ResultsThe best performance was AUC scores of 0.89 for predicting users seeking emotional support, 0.86 for experience-based information, 0.80 for unconventional information, and 0.83 for medical facts. With the gradient boosting tree as our best performing model, we analyzed the strength of individual features in predicting one's social support need. Among other discoveries, we found that users seeking emotional support tend to post more in OHSNs compared with others.ConclusionsWe developed an initial framework for automatically predicting social support needs in OHSNs using survey data. Future work should involve nonsurvey data to evaluate the feasibility of the framework. Our study contributes to providing personalized social support in OHSNs.

Project description:ImportanceRecent research highlights the association of social determinants of health with health outcomes of patients with type 2 diabetes (T2D).ObjectiveTo examine associations between health-related social needs (HRSNs) and health care quality and utilization outcomes in a Medicare Advantage population with T2D.Design, setting, and participantsThis cross-sectional study used medical and pharmacy claims data from 2019. An HRSN survey was given between October 16, 2019, and February 29, 2020, to Medicare Advantage beneficiaries. Inclusion criteria were diagnosis of T2D, age of 20 to 89 years, continuous Medicare Advantage enrollment in 2019, and response to the HRSN survey. Data were analyzed between June 2021 and January 2022.ExposuresEnrollment in Medicare Advantage, diagnosis of T2D, and completion of a survey on HRSNs.Main outcomes and measuresQuality outcomes included diabetes medication adherence, statin adherence, completion of a glycated hemoglobin (HbA1c) laboratory test in the past 12 months, and controlled HbA1c. Utilization outcomes included all-cause hospitalization, potentially avoidable hospitalization, emergency department discharge, and readmission.ResultsOf the 21 528 Medicare Advantage beneficiaries with T2D included in the study (mean [SD] age, 71.0 [8.3] years; 55.4% women), most (56.9%) had at least 1 HRSN. Among the population with T2D reporting HRSNs, the most prevalent were financial strain (73.6%), food insecurity (47.5%), and poor housing quality (39.1%). In adjusted models, loneliness (odds ratio [OR], 0.85; 95% CI, 0.73-0.99), lack of transportation (OR, 0.80; 95% CI, 0.69-0.92), utility insecurity (OR, 0.86; 95% CI, 0.76-0.98), and housing insecurity (OR, 0.78; 95% CI, 0.67-0.91) were each associated with lower diabetes medication adherence. Loneliness and lack of transportation were associated with increased emergency visits (marginal effects of 173.0 [95% CI, 74.2-271.9] and 244.6 [95% CI, 150.4-338.9] emergency visits per 1000 beneficiaries for loneliness and transportation, respectively). Food insecurity was the HRSN most consistently associated with higher acute care utilization (marginal effects of 84.6 [95% CI, 19.8-149.4] emergency visits, 30.4 [95% CI, 9.5-51.3] inpatient encounters, and 17.1 [95% CI, 4.7-29.5] avoidable hospitalizations per 1000 beneficiaries).Conclusions and relevanceIn this cross-sectional study of Medicare Advantage beneficiaries with T2D, some HRSNs were associated with care quality and utilization. The results of the study may be used to direct interventions to the social needs most associated with T2D health outcomes and inform policy decisions at the insurance plan and community level.

Dataset Information

Predicting health-related social needs in Medicaid and Medicare populations using machine learning.

Publications

Predicting health-related social needs in Medicaid and Medicare populations using machine learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets