OBJECTIVES

The extent that universal social determinants of health (SDH) screening in clinical encounters, as recommended by the American Academy of Pediatrics, has been implemented in inpatient pediatric settings is unknown. We aimed to determine the national prevalence and predictors of standardized SDH screening in US level 2 to 4 neonatal care units (NICUs), describe characteristics of SDH screening programs, and ascertain beliefs of clinical leaders about this practice in the NICU setting.

METHODS

We randomly selected 100 hospitals with level 2 to 4 NICUs among each of 5 US regions (n = 500) and surveyed clinical leaders from January to November 2021 regarding standardized SDH screening. Responses were weighted for number of level 2 to 4 NICUs in each region and nonresponse.

RESULTS

Overall response rate was 34% (28%–40% by region). Twenty-three percent of US level 2 to 4 NICUs reported standardized SDH screening. We found no associations of hospital characteristics, such as region, size, or safety-net status, with implementation of this practice. Existing programs conducted systematic screening early in the hospitalization (84%), primarily led by social workers (92%). We identified practice variation regarding the type of screening tool, but there was substantial overlap among domains incorporated in the screening. Reported barriers to implementation included perceived lack of resources, inadequate referrals, and lack of an inpatient screening tool.

CONCLUSIONS

The prolonged neonatal hospitalization provides opportunities to systematically address SDH. Yet, only 23% of US level 2 to 4 NICUs have implemented this practice. To scale-up implementation, quality improvement may support adaptation of screening and referral processes to the NICU context.

Social determinants of health (SDH) are the conditions in which people are born, grow, work, live and age, and contribute to health outcomes across the lifespan.1  Adverse SDH, such as housing instability and food insecurity, occur more often among families living in poverty or near poverty and are associated with worse child health and health care utilization.2,3  In 2016, the American Academy of Pediatrics (AAP) recommended universal screening for adverse SDH and subsequent provision of referrals to community resources when needed at pediatric clinical care encounters.4  Since that time, implementation of SDH screening and referral has rapidly increased in outpatient pediatric settings,5  and has emerged more recently within acute care settings, such as the pediatric emergency department.6,7 

Data examining the US preterm population has shown that adverse SDH disproportionally affect families with preterm versus term infants.8  Despite the high prevalence of SDH among families with preterm infants, a recent mixed methods study of 2 US level 3 neonatal care units (NICUs) showed that systematic assessment of SDH was low and variable (0.2%–38%).9  The low rate of screening occurred despite the fact that neonatal providers working in these facilities believed that systematic SDH screening was within the scope of their practice and could be beneficial for the families they serve.9  The national prevalence of standardized SDH screening and referral in US inpatient pediatric settings like NICUs is currently unknown. Understanding the extent that this practice occurs, as well as hospital characteristics that may be associated with standardized SDH screening, will address important research and clinical gaps. Furthermore, the characteristics of SDH screening and referral systems and neonatal providers’ beliefs about this practice have not been investigated. Understanding these issues will help clinical teams working in NICUs to develop or augment their own processes for standardized SDH screening and referral to scale-up this practice.

In light of these gaps, the aims of our study were to (1) determine the national prevalence of standardized SDH screening in level 2 to 4 US NICUs; (2) examine the extent that hospital characteristics were associated with standardized SDH screening; (3) describe characteristics of current standardized SDH screening and referral processes; and (4) ascertain beliefs of clinical leaders on standardized SDH screening in the NICU setting.

We electronically surveyed clinical leaders (division chiefs or medical or clinical directors) of level 2 to 4 NICUs that were randomly selected from hospitals across 5 US regions. We first obtained the 2018 American Hospital Association list of US hospitals and restricted our sample to hospitals that provided “neonatal intensive” and/or “intermediate neonatal care,” representing level 2 to 4 NICUs. A total of 1391 hospitals were eligible for recruitment (205 in the Northeast, 356 in the Midwest, 331 in the West, 328 in the Southeast, and 171 in the Southwest). Our goal was to randomly select 100 hospitals from each of the aforementioned 5 US regions for a total of 500 hospitals. This sample size estimation assumed a 15% prevalence (based on a pilot assessment), a conservative 40% response rate, and a 5% margin of error. Second, we randomly sorted eligible hospitals from the American Hospital Association list within each US region and ascertained E-mail addresses of clinical leaders sequentially through personal contacts, internet searches, and telephone calls until we reached our goal sample size. To reach 100 hospitals in the Northeast, we moved to the 112th hospital randomly sorted on our list; and for the Midwest, West, Southeast, and Southwest, we moved to the 113th, 112th, 117th, and 110th hospital, respectively. Reasons for moving down the randomly sorted list included: unable to obtain clinical leader’s contact information (n = 19), nonfunctional E-mail address (n = 36), level 1 nursery as highest level of care (n = 3), or clinical leader declined to participate (n = 6).

We administered a 10 to 15 minute Research Electronic Data Capture survey through an E-mail link from January to November 2021, with a maximum of 5 attempts for nonresponders. We did not offer an incentive for response. Our preliminary analysis of the results revealed that 49 of 170 (28.8%) of clinical leaders responded “don’t know” to our primary study question about presence of a standardized SDH screening and referral process. For this reason, our research team subsequently reached out, by E-mail or phone, to social workers at the hospitals in which the clinical leader did not know whether this practice occurred. The goal of this second wave of an abbreviated survey for social workers was to determine the answer to this key question and follow-up questions assessing the characteristics of the existing programs only. We reached 41 out of 49 of these hospitals. Therefore, the final analytic sample included 163 NICUs with a valid response to our primary question. This study (both the original application and the amendment to additionally approach social workers) was approved by our university’s institutional review board.

Our survey was delivered to clinical leaders who were asked to respond according to the NICU where they “primarily work.” We ascertained our main study question, presence of standardized SDH screening, through the following question: “Does your unit routinely use standardized social determinants of health screening tool(s) to elicit families’ unmet social needs?,” followed by a clarifying statement “please note we are referring to a written, standardized screening tool (eg, screening survey, such as WE CARE, PRAPARE, Accountable Health Communities Health-Related Social Needs Screening Tool, etc.)” The answer choices were “yes,” “no,” or “don’t know.” We chose to specifically define presence of standardized SDH screening per use of a tool, as specified in the aforementioned AAP policy statement.4 

The survey additionally assessed general hospital characteristics, including highest level of neonatal care provided (categories described in the AAP Practice Guideline on Levels of Neonatal Care10  were listed), number of annual admissions, number of full time equivalent social workers dedicated to the NICU, and proportion of admissions that were Medicaid, Black and Hispanic patients, and patients with limited English proficiency. In the survey, we defined “safety-net status” as caring for ≥75% uninsured or Medicaid patients. A high proportion of >2 times the US census national average (13% for Black, 18% for Hispanic, and 10% for limited English proficiency).11,12  Among clinical leaders that reported using a standardized SDH screening tool, we assessed characteristics of their units’ processes, including which provider type typically uses it, when during the hospitalization the tool was typically administered, which type of tool was used, and how referrals for unmet social needs occurred after screening. Finally, we asked participants to indicate their beliefs as clinical leaders regarding the importance and responsibility of SDH screening in the NICU setting.

Weighted analyses were used to account for the probability of hospitals being selected among all eligible hospitals in each region and for nonresponse rates across regions to provide national estimates. For instance, there were 38 respondents from the Northeast. Each of these was weighted by 5.4 to represent the 205 eligible NICUs from the Northeast. Similarly, there were 34 respondents from the Midwest, weighted by 10.5 to represent the 356 eligible Midwest NICUs. First, we examined overall prevalence of standardized SDH screening among level 2 to 4 NICUs in the United States. Next, we examined bivariate relationships among hospitals that did and did not have standardized SDH screening using χ2 tests. Among facilities that had standardized SDH screening, we examined characteristics of the programs. Finally, we examined bivariate relationships in beliefs regarding standardized SDH screening among clinical leaders with and without a standardized SDH screening process in their hospital using χ2 tests, using a level of significance of α = 0.05. All analyses were conducted by using SAS procedures for complex survey designs in SAS 9.4 (SAS Institute, Cary, NC, USA).

The overall response rate was 170 of 500 (34%); stratified by region it was 40% in the Northeast, 35% in the Midwest, 36% in the West, 28% in the Southeast, and 34% in the Southwest. Among 170 clinical leaders who responded, 75.3% were physicians. Among respondents, 163 had information on our main outcome of interest: use of standardized SDH screening. Among these, 37 hospitals reported having a standardized SDH screening process. After the weighting procedures described, this represented an overall prevalence of 23% of US level 2 to 4 facilities with a standardized SDH screening process. Presence of this practice did not vary according to hospital characteristics, such as highest level of neonatal care, US region, safety-net hospital status, or proportion of non-Hispanic Black or Hispanic patients served (Table 1).

TABLE 1

Hospital Characteristics According to Presence of a Standardized SDH Screening Process or Not

CharacteristicOverall
N = 163, n (Weighted %)
Standardized SDH Screening
n = 37, n (Weighted %)
No Standardized SDH Screening
n = 126, n (Weighted %)
Pa
Overall 163 37 (23.0) 126 (77.0)  
Type of hospital    .10 
 Nongovernment (not-for-profit) 124 (74.3) 28 (23.2) 96 (76.8)  
 Government 18 (12.3) 7 (37.9) 11 (62.1)  
 Investor-owned (for-profit) 21 (13.4) 2 (8.5) 19 (91.5)  
Teaching status    .69 
 Teaching 98 (62.4) 25 (24.1) 73 (75.9)  
 Nonteaching 65 (37.6) 12 (21.2) 53 (78.8)  
Safety-net hospital statusb    .84 
 Yes 55 (35.1) 12 (23.4) 43 (76.6)  
 No 107 (64.0) 25 (23.1) 82 (76.9)  
Rural versus urbanc    .08 
 Rural 11 (7.4) 4 (45.1) 7 (54.9)  
 Urban 152 (92.6) 33 (21.3) 119 (78.7)  
US Region    .83 
 Northeast 38 (14.7) 8 (21.1) 30 (78.9)  
 Midwest 34 (25.6) 9 (26.5) 25 (73.5)  
 West 26 (23.8) 7 (26.9) 19 (73.1)  
 Southeast 30 (23.6) 5 (16.7) 25 (83.3)  
 Southwest 35 (12.3) 8 (22.9) 27 (77.1)  
Highest level of neonatal care    .05 
 Level 2 41 (23.9) 10 (27.2) 31 (72.8)  
 Level 3 84 (51.8) 13 (15.1) 71 (84.9)  
 Level 4 38 (24.2) 14 (35.7) 24 (64.3)  
Number of neonatal admissions    .94 
 ≤200 45 (26.5) 8 (21.1) 37 (78.9)  
 201–400 38 (22.6) 10 (24.1) 28 (75.9)  
 >400 80 (50.9) 19 (23.5) 61 (76.5)  
Hispanic patient admissions    .76 
 ≤18% (≤national average) 53 (33.4) 9 (20.6) 44 (79.4)  
 ≤19% to 39% (between 1–2×national average) 82 (47.1) 21 (25.1) 61 (74.9)  
 ≥40% (>2×national average) 20 (13.2) 4 (16.9) 16 (83.1)  
 Don’t know 8 (6.2) 3 (33.1) 5 (66.9)  
Black patient admissions    .35 
 ≤12% (≤national average) 79 (41.1) 15 (19.0) 64 (81.0)  
 13% to 26% (between 1–2×national average) 45 (30.0) 9 (18.9) 36 (81.1)  
 ≥27% (>2×national average) 32 (23.6) 10 (31.8) 22 (68.2)  
 Don’t know 7 (5.3) 3 (38.8) 4 (61.2)  
Limited English proficient patient admissions    .53 
 ≤10% (≤national average) 53 (34.3) 9 (18.3) 44 (81.7)  
 11–20% (between 1–2×national average) 66 (39.0) 18 (26.8) 48 (73.2)  
 ≥21% (>2×national average) 30 (17.8) 4 (15.9) 26 (84.1)  
 Don’t know 13 (8.9) 5 (31.3) 8 (68.7)  
Social worker full-time equivalent    .46 
 <1.0 FTE 75 (44.8) 15 (21.5) 60 (78.5)  
 1.0 FTE 49 (29.9) 9 (19.1) 40 (80.9)  
 >1.0 FTE 39 (25.2) 13 (30.4) 26 (69.6)  
CharacteristicOverall
N = 163, n (Weighted %)
Standardized SDH Screening
n = 37, n (Weighted %)
No Standardized SDH Screening
n = 126, n (Weighted %)
Pa
Overall 163 37 (23.0) 126 (77.0)  
Type of hospital    .10 
 Nongovernment (not-for-profit) 124 (74.3) 28 (23.2) 96 (76.8)  
 Government 18 (12.3) 7 (37.9) 11 (62.1)  
 Investor-owned (for-profit) 21 (13.4) 2 (8.5) 19 (91.5)  
Teaching status    .69 
 Teaching 98 (62.4) 25 (24.1) 73 (75.9)  
 Nonteaching 65 (37.6) 12 (21.2) 53 (78.8)  
Safety-net hospital statusb    .84 
 Yes 55 (35.1) 12 (23.4) 43 (76.6)  
 No 107 (64.0) 25 (23.1) 82 (76.9)  
Rural versus urbanc    .08 
 Rural 11 (7.4) 4 (45.1) 7 (54.9)  
 Urban 152 (92.6) 33 (21.3) 119 (78.7)  
US Region    .83 
 Northeast 38 (14.7) 8 (21.1) 30 (78.9)  
 Midwest 34 (25.6) 9 (26.5) 25 (73.5)  
 West 26 (23.8) 7 (26.9) 19 (73.1)  
 Southeast 30 (23.6) 5 (16.7) 25 (83.3)  
 Southwest 35 (12.3) 8 (22.9) 27 (77.1)  
Highest level of neonatal care    .05 
 Level 2 41 (23.9) 10 (27.2) 31 (72.8)  
 Level 3 84 (51.8) 13 (15.1) 71 (84.9)  
 Level 4 38 (24.2) 14 (35.7) 24 (64.3)  
Number of neonatal admissions    .94 
 ≤200 45 (26.5) 8 (21.1) 37 (78.9)  
 201–400 38 (22.6) 10 (24.1) 28 (75.9)  
 >400 80 (50.9) 19 (23.5) 61 (76.5)  
Hispanic patient admissions    .76 
 ≤18% (≤national average) 53 (33.4) 9 (20.6) 44 (79.4)  
 ≤19% to 39% (between 1–2×national average) 82 (47.1) 21 (25.1) 61 (74.9)  
 ≥40% (>2×national average) 20 (13.2) 4 (16.9) 16 (83.1)  
 Don’t know 8 (6.2) 3 (33.1) 5 (66.9)  
Black patient admissions    .35 
 ≤12% (≤national average) 79 (41.1) 15 (19.0) 64 (81.0)  
 13% to 26% (between 1–2×national average) 45 (30.0) 9 (18.9) 36 (81.1)  
 ≥27% (>2×national average) 32 (23.6) 10 (31.8) 22 (68.2)  
 Don’t know 7 (5.3) 3 (38.8) 4 (61.2)  
Limited English proficient patient admissions    .53 
 ≤10% (≤national average) 53 (34.3) 9 (18.3) 44 (81.7)  
 11–20% (between 1–2×national average) 66 (39.0) 18 (26.8) 48 (73.2)  
 ≥21% (>2×national average) 30 (17.8) 4 (15.9) 26 (84.1)  
 Don’t know 13 (8.9) 5 (31.3) 8 (68.7)  
Social worker full-time equivalent    .46 
 <1.0 FTE 75 (44.8) 15 (21.5) 60 (78.5)  
 1.0 FTE 49 (29.9) 9 (19.1) 40 (80.9)  
 >1.0 FTE 39 (25.2) 13 (30.4) 26 (69.6)  

FTE, full-time equivalent.

a

χ2P value.

b

Defined in survey as ≥75% uninsured or Medicaid patients.

c

Rural counties are defined by the AHA as those with an urban core of ≥10 000 but <50 000 population (micropolitan) and those with no urban core.

Table 2 shows characteristics of standardized SDH screening processes among hospitals that routinely perform standardized SDH screening. All NICUs with standardized processes performed universal screening. Social workers administered screening tools to caregivers among 92% of NICUs and 84% administered the tool in the first week of hospitalization. There was variation in the screening tools, with iHELP,13  Accountable Health Communities,14  PRAPARE,15  and SEEK16  commonly reported among hospitals using existing tools. However, up to 35% of respondents reported using a tool developed by their own individual units. The specific SDH that were assessed were more uniform. For example, >70% of hospitals screened for food, housing, utilities, transportation, child care, and parental education and employment. With regard to process, 25% of NICUs with standardized screening reported that their screening tool was embedded in the electronic health record. After SDH screening, most NICUs reported social work consultation (94%) and referral to community resources (80%) to address unmet social needs identified. The main perceived barriers to implementing this practice among NICUs without standardized screening included perceptions of limited resources to implement a standardized tool (62%), lack of adequate referrals to connect families with services once unmet needs are identified (48%), and lack of validated screening tools for use in the NICU setting (34%). Other barriers to implement this practice reported in the “free text” section included: workflow issues related to the NICU context (eg, high census and acuity), lack of champions, lack of embeddedness of screening tool into the electronic health record, and concern that asking families to disclose unmet social needs may undermine provider-family relationship, particularly among undocumented families.

TABLE 2

Characteristics of Standardize SDH Screening Processes (N = 37)

Characteristicn (weighted %)
Provider type that administered screening tool (all that apply)  
 Social worker 33 (91.7) 
 Nurse 9 (26.1) 
 Physician or other advanced practice provider 4 (14.5) 
 Case manager 4 (13.9) 
Timing of screening tool administration (all that apply)  
 Prenatal consultation 7 (22.7) 
 Within first week of admission 31 (83.9) 
 Within 2 weeks of discharge 2 (3.2) 
 Othera 1 (1.7) 
Type of screening tool used  
 iHELP 5 (15.6) 
 ACH (Accountable Health Communities) 2 (7.4) 
 PRAPARE (Protocol for Responding to and Assessing Patient’s Assets, Risks and Experiences) 2 (7.2) 
 SEEK (Safe Environment for Every Kid) 3 (6.3) 
 Tool developed by local unit 13 (35.3) 
 Otherb 12 (28.1) 
SDH assessed on the tool (all that apply)  
 Food 31 (86.6) 
 Housing 34 (93.4) 
 Utilities (eg, heat) 30 (78.7) 
 Family income 25 (65.4) 
 Transportation 33 (89.4) 
 Child care 29 (77.1) 
 Parental education 24 (71.0) 
 Parental employment 29 (78.8) 
 Parental immigration status 13 (36.7) 
 Otherc 11 (25.1) 
Screening tool embedded in electronic health record  
 Yes 10 (24.6) 
Follow-up procedures after SDH screening (all that apply)  
 Social work consultation 35 (94.3) 
 Referral to community based resources 31 (80.4) 
 Telephone hotline that links families to programs 7 (16.8) 
 Hospital-based programs/resources 2 (5.7) 
Characteristicn (weighted %)
Provider type that administered screening tool (all that apply)  
 Social worker 33 (91.7) 
 Nurse 9 (26.1) 
 Physician or other advanced practice provider 4 (14.5) 
 Case manager 4 (13.9) 
Timing of screening tool administration (all that apply)  
 Prenatal consultation 7 (22.7) 
 Within first week of admission 31 (83.9) 
 Within 2 weeks of discharge 2 (3.2) 
 Othera 1 (1.7) 
Type of screening tool used  
 iHELP 5 (15.6) 
 ACH (Accountable Health Communities) 2 (7.4) 
 PRAPARE (Protocol for Responding to and Assessing Patient’s Assets, Risks and Experiences) 2 (7.2) 
 SEEK (Safe Environment for Every Kid) 3 (6.3) 
 Tool developed by local unit 13 (35.3) 
 Otherb 12 (28.1) 
SDH assessed on the tool (all that apply)  
 Food 31 (86.6) 
 Housing 34 (93.4) 
 Utilities (eg, heat) 30 (78.7) 
 Family income 25 (65.4) 
 Transportation 33 (89.4) 
 Child care 29 (77.1) 
 Parental education 24 (71.0) 
 Parental employment 29 (78.8) 
 Parental immigration status 13 (36.7) 
 Otherc 11 (25.1) 
Screening tool embedded in electronic health record  
 Yes 10 (24.6) 
Follow-up procedures after SDH screening (all that apply)  
 Social work consultation 35 (94.3) 
 Referral to community based resources 31 (80.4) 
 Telephone hotline that links families to programs 7 (16.8) 
 Hospital-based programs/resources 2 (5.7) 
a

Prenatally and on admission to maternity.

b

SWYC (Survey of Well-Being of Young Children) n = 1, nondescribed other n = 11.

c

Mental health (n = 5), substance use (n = 3), interpersonal violence (n = 3), insurance (n = 3), religious preference (n = 2), language preference (n = 1), social supports (n = 3), access to infant resources (n = 1).

Table 3 summarizes beliefs of clinical leaders regarding SDH screening in the NICU setting. Overall, >80% of leaders from both facilities with and without standardized SDH screening agreed or strongly agreed that it was a priority, feasible; and >90% of leaders agreed or strongly agreed that it would improve family engagement and may improve infant health and development. Both groups agreed or strongly agreed with the statement that social workers are responsible for assessing SDH more often than physicians or nurses. In posthoc analyses, among clinical leaders from hospitals without standardized SDH screening, those who considered it a priority were more likely to report that physicians are responsible for assessing SDH (52%) compared to those who did not consider it a priority (27%). We did not find differences between these hospitals’ characteristics.

TABLE 3

Beliefs of Clinical Leaders of Level 2 to 4 Neonatal Care Units According to Presence of a Standardized SDH Screening Process or Not

BeliefStrongly Disagree or Disagree, n (%)Neutral, n (%)Agree or Strongly Agree, n (%)P
Addressing SDH is a priority in my unit     
 SDH screening 0 (0) 7 (19.2) 30 (80.8) .16 
 No SDH screening 8 (5.9) 14 (10.7) 104 (83.4)  
Social workers are responsible for assessing SDH    .06 
 SDH screening 0 (0) 5 (10.2) 32 (89.8)  
 No SDH screening 4 (3.5) 5 (4.3) 117 (92.2)  
Physicians are responsible for assessing SDH    .37 
 SDH screening 7 (18.5) 11 (27.5) 19 (54.0)  
 No SDH screening 37 (30.8) 29 (21.4) 60 (47.7)  
Nurses are responsible for assessing SDH    .43 
 SDH screening 5 (13.7) 10 (26.0) 22 (60.3)  
 No SDH screening 16 (12.3) 31 (25.4) 79 (62.3)  
Standardized SDH screening is feasible in the NICU    .28 
 SDH screening 0 (0) 3 (6.5) 34 (93.5)  
 No SDH screening 4 (2.4) 19 (15.5) 103 (82.2)  
Standardized SDH screening builds trust with families    .02 
 SDH screening 0 (0) 6 (13.4) 31 (86.6)  
 No SDH screening 2 (0.9) 51 (41.0) 73 (58.1)  
Standardized SDH screening facilitates discussion of sensitive topics (ie, homelessness)    .21 
 SDH screening 0 (0) 1 (3.3) 36 (96.7)  
 No SDH screening 1 (0.5) 15 (12.3) 110 (87.2)  
Addressing SDH may improve family engagement    .57 
 SDH screening 0 (0) 1 (1.5) 36 (98.5)  
 No SDH screening 2 (2.0) 7 (4.9) 117 (93.1)  
Addressing SDH may improve infant health and development    .77 
 SDH screening 0 (0) 1 (1.5) 36 (98.5)  
 No SDH screening 0 (0) 2 (2.2) 124 (97.8)  
BeliefStrongly Disagree or Disagree, n (%)Neutral, n (%)Agree or Strongly Agree, n (%)P
Addressing SDH is a priority in my unit     
 SDH screening 0 (0) 7 (19.2) 30 (80.8) .16 
 No SDH screening 8 (5.9) 14 (10.7) 104 (83.4)  
Social workers are responsible for assessing SDH    .06 
 SDH screening 0 (0) 5 (10.2) 32 (89.8)  
 No SDH screening 4 (3.5) 5 (4.3) 117 (92.2)  
Physicians are responsible for assessing SDH    .37 
 SDH screening 7 (18.5) 11 (27.5) 19 (54.0)  
 No SDH screening 37 (30.8) 29 (21.4) 60 (47.7)  
Nurses are responsible for assessing SDH    .43 
 SDH screening 5 (13.7) 10 (26.0) 22 (60.3)  
 No SDH screening 16 (12.3) 31 (25.4) 79 (62.3)  
Standardized SDH screening is feasible in the NICU    .28 
 SDH screening 0 (0) 3 (6.5) 34 (93.5)  
 No SDH screening 4 (2.4) 19 (15.5) 103 (82.2)  
Standardized SDH screening builds trust with families    .02 
 SDH screening 0 (0) 6 (13.4) 31 (86.6)  
 No SDH screening 2 (0.9) 51 (41.0) 73 (58.1)  
Standardized SDH screening facilitates discussion of sensitive topics (ie, homelessness)    .21 
 SDH screening 0 (0) 1 (3.3) 36 (96.7)  
 No SDH screening 1 (0.5) 15 (12.3) 110 (87.2)  
Addressing SDH may improve family engagement    .57 
 SDH screening 0 (0) 1 (1.5) 36 (98.5)  
 No SDH screening 2 (2.0) 7 (4.9) 117 (93.1)  
Addressing SDH may improve infant health and development    .77 
 SDH screening 0 (0) 1 (1.5) 36 (98.5)  
 No SDH screening 0 (0) 2 (2.2) 124 (97.8)  

N and weighted percentages shown, among 37 neonatal care facilities with a standardizes screening process and 126 that did not.

To our knowledge, this is the first study to examine prevalence of standardized SDH screening and referral among a nationally representative sample of level 2 to 4 NICUs in the United States. We found that although the vast majority of clinical leaders believed that addressing SDH was feasible, beneficial for infants and families, and a priority for neonatal care, only 23% of units had implemented this practice. NICUs with standardized processes conduct systematic universal screening early in the hospitalization, primarily led by social workers, and coupled with referrals to hospital or community resources. We identified practice variation regarding the type of screening tool, but there was substantial overlap among domains incorporated in the screening.

Although standardized SDH screening and referral has rapidly increased in pediatric outpatient settings,5  it has not yet come to scale in inpatient acute care settings like the NICU. Social care interventions are particularly relevant to the NICU setting because caregivers of premature infants are disproportionally low-income and the prolonged neonatal hospitalization itself is likely to exacerbate baseline unmet social needs.17,18  Therefore, systematic SDH screening and provision of referrals for resources is endorsed by the AAP4  and proposed as a “potentially better practice” to address the high burden of adverse SDH among NICU families.19  However, we found that despite clinical leaders’ perceived high acceptability and value of addressing SDH during the NICU hospitalization, the uptake of this practice was substantially low. This finding reflects that the integration of social care interventions as part of routine clinical care in inpatient settings are likely driven by factors beyond professional recommendations or belief in the utility of the practice. Indeed, a recent qualitative study of clinical stakeholder’s perspectives on SDH screening in inpatient pediatric wards identified context-related implementation barriers including prioritization of medical needs in acute care settings, lack of clinicians’ education regarding screening and follow-up procedures, and lack of preestablished relationships between clinicians and families.20  In our study, leaders of NICUs without standardized screening identified additional logistic barriers such as lack of validated screening tools, limited resources, and perceived lack of adequate referrals to connect families with the services needed.

In contrast, our study showed that US NICUs already using standardized SDH screening have overcome some of these barriers by adapting existing screening tools or by developing their own. Implementation of SDH screening and referral in NICUs is important because of the potential long-term improvements to health and well-being after connections with community resources. Randomized controlled trials in outpatient pediatric settings examining low touch SDH screening and referral have demonstrated effectiveness of these interventions in increasing families’ receipt of resources with fewer reported adverse SDH at follow-up.21,22  Gottlieb et al also demonstrated that SDH screening and referral were associated with better caregiver reported children’s general and emotional health.22  These trials used different multicomponent screening surveys coupled with written resource sheets for families and/or patient navigation, highlighting the fact that the specifics of the tool selected may be less important than implementing a tool that is multicomponent and aligned with quality resources for families. Hospitals looking to implement SDH screening may consider any of the existing tools or even parts of tools on the basis of their available resources. Specific guidelines for inpatient SDH screening do not exist. These guidelines should address the mode and timing of screening, type of screening provider, and communication of screening results and referrals to outpatient providers as part of discharge hand-offs for continuity of care. Tools that are embedded in the electronic health record may facilitate this hand-off communication.

Medicaid managed care organization programs in over 30 states now encourage screening for SDH and referrals for social services.23  In addition, the shift toward value-based care offers significant opportunities to address SDH while improving care value and quality.24  Therefore, we hypothesized that NICUs serving a majority of publicly insured or racially minoritized families would be more likely to have standardized SDH screening and referrals processes in place. However, we found that none of the hospital characteristics examined were associated with this practice, including size, level of care, geographic region, safety-net status, or full-time equivalent of social workers. To date, many neonatal care facilities have remained siloed within larger health care systems and the typical structure of bundled-care reimbursement in NICUs has yet to catch up to account for social risk assessments. Although the funding stream from value-based payment reforms has the potential to accelerate initiatives to address SDH at the population level, this has not yet occurred in NICU settings.2527  Overall, effective interventions to address SDH in high-risk clinical settings like the NICU will involve not only hospitals but also payers and cross-sector partnerships.28 

Our results emphasize the need to build capacity through staff education and to foster a culture of follow-through29  in which all the members of the NICU health care team believe they share the responsibility to identify and address SDH to improve health outcomes of infants and their families. Although the majority of survey respondents believed that social workers were primarily responsible for assessing SDH among NICU families, a large proportion of them also agreed or strongly agreed that physicians and/or nurses share that responsibility. However, in practice, only 14% of physicians and 26% of nurses participated of SDH screening and referral. Notably, almost 90% of leaders of NICUs with standardized SDH screening believed that this practice built provider-family trust, and this belief was strongly associated with having standardized processes. Previous literature has shown that caregivers of hospitalized children have overall favorable opinions of inpatient SDH screening conducted by physicians.30  This literature suggests that physicians should be confident in the acceptability of SDH screening in acute settings. Quality improvement and implementation science methodologies can be applied to further assess feasibility, acceptability as well as effectiveness of SDH screening and referral interventions in the NICU context.31,32 

Strengths of this study include the sampling strategy and weighting procedures, which enabled the estimate of national prevalence of SDH screening and referral in level 2, 3, and 4 NICUs in the United States. Limitations of this study may include the type of respondent provider. Up to 29% of clinical leaders responded don’t know to the question on standardized screening, which prompted the need to reach out to social workers in those hospitals. It is possible that clinical leaders that reported not having standardized screening and referral were simply not aware of this practice in their units. However, prevalence of this practice was similar among the subsample of hospitals in which the social worker was the primary respondent (26% vs 23%). Despite our efforts, overall response rate was 34% (28%–40% across US regions), which may reflect the increased administrative and clinical workload of clinical leaders during the coronavirus disease 2019 pandemic. Nonetheless, our response rate is higher than other recent national surveys of neonatal providers.3335  An incentive may have boosted response rate. A larger sample size may have enabled the detection of observable differences in hospital characteristics. Although we weighted our results for nonresponse, there is a possibility of selection bias if nonrespondents were less interested in the subject of the survey and thus more likely to not have standardized processes for SDH screening and referral. Inclusion of such individuals would have led to an even lower than estimated national prevalence of standardized SDH screening and referral, further emphasizing the need to support NICUs in the adoption of standardized SDH screening and referral processes.

Despite recommendations for standardized SDH screening and referral in clinical settings by professional organizations such as the AAP, only 23% of US level 2 to 4 NICUs have implemented this practice. Given the extended opportunities for provider-family interaction over the prolonged neonatal hospitalization, this represents a missed opportunity to address the high burden of unmet social needs among families of high-risk infants. Further investigation of optimal implementation strategies of SDH screening in US NICUs is needed.

We acknowledge Emma Jacobs, Selena Chen, Andrea Kuriyama, Farah Delgado, and Mariana Serrano for their critical assistance in obtaining clinical leader’s contact information that enabled electronic survey administration.

Drs Cordova-Ramos and Parker contributed to the conceptualization, design, and analysis, drafted and edited the manuscript, and approved the final manuscript as submitted; Drs Heeren, Drainoni, and Garg contributed to the conceptualization, design, and edited the manuscript; Dr Kerr contributed to the conceptualization and performed the analysis; and all authors approved the final manuscript as submitted.

FUNDING: This work was supported by the W.K. Kellogg Foundation (P0131665 PI Parker) and the Evans Center for Implementation and Improvement Sciences (CIIS Department of Medicine, Boston University School of Medicine. Dr Cordova-Ramos is supported by BU-CTSI grant 1KL2TR001411. This work used Research Electronic Data Capture (REDCap) database, which was made available by the Boston University CTSI supported by 1UL1TR001430.

CONFLICT OF INTEREST DISCLOSURES: The authors have no conflicts of interest relevant to this article to disclose.

1.
World Health Organization
.
Social determinants of health
.
2.
Marmot
M
.
Social determinants of health inequalities
.
Lancet
.
2005
;
365
(
9464
):
1099
1104
3.
Braveman
P
,
Egerter
S
,
Williams
DR
.
The social determinants of health: coming of age
.
Annu Rev Public Health
.
2011
;
32
:
381
398
4.
Council on Community Pediatrics
,
Gitterman
BA
,
Flanagan
PJ
,
Cotton
WH
, et al
.
Poverty and child health in the United States
.
Pediatrics
.
2016
;
137
(
4
):
e20160339
5.
Sokol
R
,
Austin
A
,
Chandler
C
, et al
.
Screening children for social determinants of health: a systematic review
.
Pediatrics
.
2019
;
144
(
4
):
e20191622
6.
Wallace
AS
,
Luther
BL
,
Sisler
SM
,
Wong
B
,
Guo
JW
.
Integrating social determinants of health screening and referral during routine emergency department care: evaluation of reach and implementation challenges
.
Implement Sci Commun
.
2021
;
2
(
1
):
114
7.
Walter
LA
,
Schoenfeld
EM
,
Smith
CH
, et al
.
Emergency department-based interventions affecting social determinants of health in the United States: A scoping review
.
Acad Emerg Med
.
2021
;
28
(
6
):
666
674
8.
Parker
MG
,
de Cuba
SE
,
Rateau
LJ
, et al
.
Household unmet basic needs in the first 1000 days and preterm birth status
.
J Perinatol
.
2022
;
42
(
3
):
389
396
9.
Parker
MG
,
Garg
A
,
Brochier
A
, et al
.
Approaches to addressing social determinants of health in the NICU: a mixed methods study
.
J Perinatol
.
2021
;
41
(
8
):
1983
1991
10.
American Academy of Pediatrics Committee on Fetus And Newborn
.
Levels of neonatal care
.
Pediatrics
.
2012
;
130
(
3
):
587
597
11.
United States Census Bureau
.
Quick facts
.
Population estimates 2021. Available at: https://www.census.gov/quickfacts/fact/table/US/PST045221. Accessed May 5, 2021
12.
United States Census Bureau
.
National population by characteristics: 2010-2019
.
13.
Kenyon
C
,
Sandel
M
,
Silverstein
M
,
Shakir
A
,
Zuckerman
B
.
Revisiting the social history for child health
.
Pediatrics
.
2007
;
120
(
3
):
e734
e738
14.
Billioux
A
,
Verlander
K
,
Anthony
S
,
Alley
D
.
Standardized screening for health-related social needs in clinical settings: the accountable health communities screening tool
.
15.
Protocol for Responding and Assessing Patients’ Assets
.
Risks and experiences (PRAPARE)
.
Available at: https://prapare.org/. Accessed June 27, 2022
16.
Safe Environment for Every Kid
.
The SEEK Project, LLC
.
Available at: https://seekwellbeing.org/seek-materials/. Accessed June 27, 2022
17.
Parker
MG
,
de Cuba
SE
,
Rateau
LJ
, et al
.
Household unmet basic needs in the first 1000 days and preterm birth status
.
J Perinatol
.
2022
;
42
(
3
):
389
396
18.
Parker
MG
,
Garg
A
,
McConnell
MA
.
Addressing childhood poverty in pediatric clinical settings: the neonatal intensive care unit is a missed opportunity
.
JAMA Pediatr
.
2020
;
174
(
12
):
1135
1136
19.
Horbar
JD
,
Edwards
EM
,
Ogbolu
Y
.
Our responsibility to follow through for NICU infants and their families
.
Pediatrics
.
2020
;
146
(
6
):
e20200360
20.
Leary
JC
,
Rijhwani
L
,
Bettez
NM
, et al
.
Clinical stakeholder perspectives on pediatric inpatient screening for social needs
.
Acad Pediatr
.
2022
;
22
(
3
):
470
477
21.
Garg
A
,
Toy
S
,
Tripodis
Y
,
Silverstein
M
,
Freeman
E
.
Addressing social determinants of health at well child care visits: a cluster RCT
.
Pediatrics
.
2015
;
135
(
2
):
e296
e304
22.
Gottlieb
LM
,
Adler
NE
,
Wing
H
, et al
.
Effects of in-person assistance vs personalized written resources about social services on household social risks and child and caregiver health: a randomized clinical trial
.
JAMA Netw Open
.
2020
;
3
(
3
):
e200701
23.
Kaiser Family Foundation
.
A view from the states: key Medicaid policy changes: results from a 50-state Medicaid budget survey for state fiscal years 2019 and 2020
.
24.
Starfield
B
,
Gérvas
J
,
Mangin
D
.
Clinical care and health disparities
.
Annu Rev Public Health
.
2012
;
33
:
89
106
25.
Tanenbaum
SJ
.
Can payment reform be social reform? The lure and liabilities of the “triple aim”
.
J Health Polit Policy Law
.
2017
;
42
(
1
):
53
71
26.
Taylor
LA
,
Berchuck
CM
,
Barnett
KG
.
A strategic tension for hospitals moving upstream: Cede control but maintain accountability
.
Prev Med
.
2018
;
112
:
207
208
27.
Tobin-Tyler
E
,
Ahmad
B
.
Marrying value-based payment and the social determinants of health through Medicaid ACOs: implications for policy and practice
.
28.
Liu
PY
,
Beck
AF
,
Lindau
ST
, et al
.
A framework for cross-sector partnerships to address childhood adversity and improve life course health
.
Pediatrics
.
2022
;
149
(
Suppl 5
):
e2021053509O
29.
Beck
AF
,
Edwards
EM
,
Horbar
JD
,
Howell
EA
,
McCormick
MC
,
Pursley
DM
.
The color of health: how racism, segregation, and inequality affect the health and well-being of preterm infants and their families
.
Pediatr Res
.
2020
;
87
(
2
):
227
234
30.
Colvin
JD
,
Bettenhausen
JL
,
Anderson-Carpenter
KD
,
Collie-Akers
V
,
Chung
PJ
.
Caregiver opinion of in-hospital screening for unmet social needs by pediatric residents
.
Acad Pediatr
.
2016
;
16
(
2
):
161
167
31.
Parker
MG
,
Hwang
SS
.
Quality improvement approaches to reduce racial/ethnic disparities in the neonatal intensive care unit
.
Semin Perinatol
.
2021
;
45
(
4
):
151412
32.
Kaplan
HC
,
Walsh
KE
.
Context in Implementation Science
.
Pediatrics
.
2022
;
149
(
Suppl 3
):
e2020045948C
33.
Chiruvolu
A
,
Mallett
LH
,
Govande
VP
,
Raju
VN
,
Hammonds
K
,
Katheria
AC
.
Variations in umbilical cord clamping practices in the United States: a national survey of neonatologists
.
J Matern Fetal Neonatal Med
.
2022
;
35
(
19
):
3646
3652
34.
Hwang
JS
,
Friedlander
S
,
Rehan
VK
,
Zangwill
KM
.
Diagnosis of congenital/perinatal infections by neonatologists: a national survey
.
J Perinatol
.
2019
;
39
(
5
):
690
696
35.
Lieschke
A
,
Dahlstrom
JE
,
Kent
AL
,
Sethna
F
.
Uncertainty over implications of placental histopathological findings: A survey of Australian and New Zealand neonatologists
.
J Paediatr Child Health
.
2020
;
56
(
2
):
259
264