Hospital Characteristics Associated With Social Needs Activities in the US

Abstract

This study assesses what hospital characteristics, including hospital participation in payment and delivery reform, are associated with activities related to health-related social needs.

New standards have recently been announced for hospitals and health plans related to health-related social needs (HRSNs) activities.¹ Contemporary assessments of hospital progress on these standards can help identify additional policy and payment supports needed to ensure equitable and meaningful adoption. We assessed what hospital characteristics, including hospital participation in payment and delivery reform, are associated with HRSN-related activities.

Methods

We conducted a cross-sectional analysis of acute care hospitals responding to the 2021 American Hospital Association (AHA) online survey, typically completed by administrators delegated by hospital CEOs. We focused on items related to screening for and having programs/interventions to address 5 HRSNs prioritized by the Centers for Medicare & Medicaid Services (CMS): food, housing, utilities, interpersonal violence, and transportation (additional details provided in the eTable in Supplement 1).

To adjust for hospital nonresponse, we constructed weights using logistic regression to predict HRSN item response based on hospital characteristics (eMethods in Supplement 1). The prevalence of HRSN screening and intervention items was calculated overall and by hospital characteristics. Associations between hospital characteristics and screening and interventions for all 5 HRSNs were assessed using 2 separate multivariable logistic regression models adjusting for hospital size, ownership, rurality, safety-net status, critical access status, teaching status, region, and participation in value-based payment (VBP) programs (accountable care organizations, bundled payment, and patient-centered medical homes). Models included robust standard errors, and average marginal effects (AMEs) reflecting incremental differences in the predicted probabilities were reported.

Analyses were performed using Stata version 18.0 (StataCorp) and statistical significance was defined as a 95% CI that did not include 1. This study was deemed exempt from approval by the University of California San Francisco institutional review board.

Results

Of 4306 acute care hospitals, 2749 (63.8%) responded to HRSN screening and intervention items. Compared with HRSN item nonrespondents, respondents were more likely to be located in the Northeast and to be larger, non-profit, metropolitan, teaching, and noncritical access hospitals. Overall, 82.8% of hospitals reported screening for at least 1 of the 5 CMS priority HRSNs; 45.6% screened for all 5. Food insecurity screening was most common (72.2%) and screening for utility needs was least common (51.1%). More hospitals (89.7%) reported providing interventions to address any HRSN, but fewer (38.1%) provided interventions for all 5. Intervening on transportation needs was the most common (77.0%) and intervening on utilities was the least common (46.5%). The prevalence of screening for and intervening on individual, any, and all 5 HRSNs across hospital characteristics are shown in the Table.

Table. Prevalence of Activities to Screen and Address Health-Related Social Needs (HRSNs), by Hospital Characteristics.

Characteristic	Screening (n = 2519)a							Interventions (n = 2610)a
	Any HRSNb	All 5 CMS priority HRSNs	Food insecurity	Housing instability	Utility needs	Interpersonal violence	Transportation needs	Any HRSNb	All 5 CMS priority HRSNs	Food insecurity	Housing instability	Utility needs	Interpersonal violence	Transportation needs
Overall	82.8	45.6	72.2	69.4	51.1	69.0	69.6	89.7	38.1	72.8	58.8	46.5	61.4	77.0
Hospital size, No. of beds
<100	75.5	36.8	63.2	60.3	42.0	61.2	58.9	85.8	31.0	64.6	50.0	39.3	52.4	69.0
100-400	89.5	53.2	81.2	77.7	59.1	75.8	79.3	92.9	43.5	80.2	66.0	51.9	69.3	84.0
>400	95.7	62.4	85.0	85.5	68.1	83.8	88.1	97.9	54.4	88.1	77.4	63.1	78.7	91.9
Ownership
Nonfederal government	70.8	34.5	59.5	58.2	41.1	58.2	55.8	83.4	28.7	61.4	48.4	38.1	51.1	69.2
Nonprofit	89.6	53.0	80.7	76.4	57.8	75.8	77.6	93.8	44.4	81.2	65.7	52.3	67.0	81.6
For profit	70.8	29.5	53.8	55.8	36.8	55.6	55.1	81.6	25.0	53.7	44.6	34.1	52.5	68.7
Rurality
Metropolitan	89.3	52.6	79.9	77.8	58.1	75.0	78.5	92.7	44.8	79.5	66.1	53.0	68.7	82.6
Micropolitan	82.3	48.6	69.5	68.2	53.4	70.2	69.7	87.9	36.5	66.9	53.6	45.7	60.7	74.5
Rural	67.8	26.9	55.8	50.4	32.8	54.0	48.3	84.0	23.6	61.1	45.3	31.7	44.8	65.6
Hospital designation
Safety netc
Yes	88.2	50.6	71.4	77.0	57.8	74.9	75.4	93.0	38.6	73.5	65.7	47.9	67.3	82.1
No	81.7	44.6	76.3	67.9	49.8	67.8	68.4	89.0	38.0	72.6	57.4	46.2	60.2	75.9
Critical accessd
Yes	72.7	32.3	61.2	57.7	38.2	57.3	53.7	84.7	26.9	63.0	48.1	35.1	46.2	67.4
No	87.6	51.8	77.3	74.9	57.2	74.5	77.0	92.1	43.4	77.4	63.9	51.9	68.6	81.5
Teachinge
Yes	92.0	58.8	83.5	81.6	63.7	80.4	82.6	95.3	48.2	83.4	69.0	56.8	73.7	86.3
No	77.7	38.2	66.0	62.6	44.1	62.6	62.3	86.6	32.4	66.9	53.1	40.7	54.5	71.8
Region
Northeast	94.6	59.9	86.6	84.0	65.3	84.3	80.7	94.1	46.7	87.8	67.7	56.6	71.6	83.1
Midwest	83.5	47.6	74.7	69.3	52.0	69.5	69.4	90.6	40.2	73.8	58.4	47.1	60.8	77.6
South	76.9	39.5	63.5	60.5	46.2	61.9	65.6	87.1	32.5	67.0	53.2	42.3	59.1	74.6
West	86.3	45.7	76.5	78.2	50.7	72.9	70.8	90.8	40.7	73.6	65.2	47.5	60.7	77.0
Alternative payment models
Accountable care organization
Yes	90.6	53.3	81.2	78.0	58.5	76.4	78.2	93.6	44.6	78.6	66.5	52.0	68.4	82.8
No	73.3	35.8	61.0	59.1	42.3	59.2	59.1	84.3	29.3	65.5	49.1	39.3	51.6	69.4
Bundled payment
Yes	92.3	61.0	86.0	80.7	66.3	77.9	82.8	93.7	52.5	84.5	70.9	61.4	69.0	84.6
No	79.2	39.6	67.0	64.8	45.4	65.3	64.5	88.3	32.9	68.8	54.6	41.2	58.6	74.3
Patient-centered medical home
Yes	91.9	58.3	86.6	81.5	63.6	76.7	76.3	93.9	52.5	85.7	73.8	60.4	71.2	82.6
No	80.5	42.3	68.3	66.2	48.1	66.8	67.8	88.4	34.5	69.4	55.0	43.1	58.8	75.8

Abbreviation: CMS, Centers for Medicare and & Medicaid Services.

^a Numbers represent the percentage of hospitals that answered affirmatively to each category.

^b HRSNs included (1) food insecurity, (2) housing instability, (3) utility needs, (4) interpersonal violence, and (5) transportation needs.

^c Safety-net hospitals defined as those in the highest quartile of the Disproportionate Share Hospital index nationally, which measures the proportion of Medicaid and low-income Medicare patients a hospital serves.

^d Critical access hospitals are small, rural hospitals with formal designations and funding from CMS.

^e Teaching hospitals defined as those with medical school affiliations reported to the American Medical Association or members of the Council of Teaching Hospitals and Health Systems.

Teaching hospitals (AME, 0.06 [95% CI, 0.02 to 0.11]) and those that participated in accountable care organizations (AME, 0.04 [95% CI, 0.005 to 0.09), bundled payments (AME, 0.10 [95% CI, 0.06 to 0.15]), and patient-centered medical homes (AME, 0.06 [95% CI, 0.01 to 0.10]) were more likely to screen for all 5 CMS priority HRSNs compared with reference groups, while for-profit (AME, −0.10 [95% CI, −0.18 to −0.03]), rural (AME, −0.14 [95% CI, −0.20 to −0.07]), and Southern (AME, −0.07 [95% CI, −0.13 to −0.009]) hospitals were less likely (Figure). Hospitals that participated in accountable care organizations (AME, 0.05 [95% CI, 0.009 to 0.09]), bundled payments (AME, 0.11 [95% CI, 0.07 to 0.15]), and patient-centered medical homes (AME, 0.11 [95% CI, 0.06 to 0.15]) were more likely to have interventions to address all 5 CMS priority HRSNs, while rural (AME, −0.12 [95% CI, −0.18 to −0.05]), critical access (AME, −0.07 [95% CI, −0.13 to −0.003), and safety-net (AME, −0.06 [95% CI, −0.12 to −0.006) hospitals were less likely.

Figure. Hospital Characteristics Associated With Activities to Screen and Address Health-Related Social Needs (HRSNs).

Average marginal effects reflect the average difference in predicted probability of screening for or having programs/interventions to address all 5 priority Centers for Medicare and & Medicaid Services (CMS) HRSNs relative to the respective reference group. For example, teaching hospitals were 6 percentage points more likely to screen for all 5 CMS priority HRSNs compared with nonteaching hospitals. Reference groups are labeled and denoted as points without whiskers. Whiskers represent 95% CIs.

Discussion

This 2021 national survey demonstrated that nearly half of acute care hospitals reported screening for all 5 CMS priority HRSNs, substantially higher than a prior 2017-2018 estimate of 24.4%.² While fewer than half of hospitals reported having screening and interventions to address all 5 CMS priority HRSNs, findings suggest that VBP participation was associated with HRSN program adoption, potentially because these models provide the flexibility to use funds to develop, scale, and sustain such programs. This differs from a prior 2017-2018 analysis showing that VBP program participation among physician practices was not associated with HRSN screening, which may reflect the shifting national enthusiasm for HRSN integration.³ Despite caring for patients with a higher burden of social needs,⁴ rural, safety-net, and critical access hospitals provided fewer HRSN-related services. Consistent with earlier data, this suggests these hospitals will likely require additional financial resources, workforce supports, and stronger community-based partnerships, particularly for specific HRSNs such as housing that are more resource-intensive to address.^5,6

Study limitations include differences in hospital characteristics between HRSN item respondents and nonrespondents, potential bias related to self-reported survey data, and lack of information on the reach and scope of HRSN programs.

Section Editors: Jody W. Zylke, MD, Deputy Editor; Karen Lasser, MD, and Kristin Walter, MD, Senior Editors.

Supplement 1.

eTable. Detailed Description of HRSN Items and Hospital Characteristics

eMethods. Survey Weights for Survey Nonresponse

eReference

Supplement 2.

Data Sharing Statement