Parent Preferences on Screening for Social Determinants of Health in Hospitalized Children

We performed a single-center, descriptive, survey-based study using a convenience sample of participants to assess caregiver preferences on screening for social needs in hospitalized children. The survey instrument was distributed to families of children admitted to the medical unit of a university-affiliated, freestanding children’s hospital in the mid-Atlantic region. Our inpatient setting sees a wide variety of diagnoses and pathology with >14 448 admissions per year to the hospital. The medical unit contains patients mostly from the hospital medicine service with few subspecialty patients (Adolescent, Pulmonary, and Gastroenterology). The overall hospital average length of stay (LOS) is 7.74 days. The insurance mix is 20% private and 73% public. The hospital medicine racial/ethnic breakdown is 44% Black, 23% Hispanic, and 8% White, with 30% other. The most common cause of hospitalization is respiratory failure. The hospital medicine service had 4852 discharges during the study period and a 4-day median LOS (6.63-day average LOS) for the study period in 2021.

Recruitment efforts captured all days of the week, at times amenable to the recruiters’ schedules. The recruitment team consisted of 3 trained medical students, as well as a pediatric resident (AL) and nurse (WN) from the study team. They sequentially approached each patient room on the unit, inviting each caregiver who identified themself as a parent or guardian of the patient to participate. An information sheet was reviewed with the caregiver and once consented they were instructed to access the survey, which was in English and Spanish, via a QR code. Patient diagnoses, LOS, and other identifiers were not known to the study team. We did not track the total number of families approached and did not collect sociodemographic or clinical information on those who declined to participate. Therefore, the results should be interpreted with those limitations in mind.

Surveys were distributed between March 2021 and January 2022 and administered by using a QR code, which opened the web-based survey on the caregiver’s smartphone (due to coronavirus disease 2019 infection precautions). An iPad was available for use if the caregiver did not have access to a smartphone. The participant answered the survey without the assistance or presence of the study team unless specifically requested by the caregiver for technology assistance. The survey was piloted in English and Spanish versions. Spanish translation of the survey was provided by Cyracom International and when available, a bilingual nurse. The Spanish language survey was available 3 months after the English language version, delayed by funding issues. Caregivers received a $25 gift card as compensation for their participation. Study data were collected and managed by using Research Electronic Data Capture (REDCap, Inc.).

The study, including a written consent waiver, was approved by the hospital’s institutional review board.

The survey assessed caregiver preferences on screening of social needs (See Survey Instrument in Supplemental Information). The specific social needs domains included food insecurity, housing and housing conditions, safety, utilities, access to health insurance, transportation, health literacy, neighborhood violence, and stress. We first asked caregivers if they felt it was important for the hospital health care team to know about adverse circumstances for each of the listed domains and, separately, if they would feel comfortable sharing these circumstances with their inpatient health care team. We assessed caregiver perspectives on importance and comfort in answering questions regarding social needs through a 3-point Likert-type scale (Important, Somewhat Important, Not Important and Comfortable, Somewhat Comfortable, Not Comfortable). The survey asked caregivers how they would prefer to answer these questions (in private versus with a specific health care team member [definitions of specific team members were provided]), their preference of modality (electronically versus on paper versus verbally), and when during the hospitalization, if at all, they would want to be screened for social needs (in the emergency department versus arrival to the inpatient unit). The survey also asked about the acceptability of screening for social needs if the hospital was unable to provide resources that met the needs of the families.

We collected caregiver demographic information, including age, sex, race/ethnicity, education level, and whether the child had been to their pediatrician within the last year. All survey questions were optional. At the end of the survey, all caregivers were provided a Web site (https://www.findhelp.org) that outlines a multitude of domain-specific resources by zip code; in addition, participants were also offered additional support from the health care team to address any unmet needs.

Survey development was based on existing social needs screening tools and used the survey design framework described by Gehlbach, Artino, and Durning.^9,25,26  The survey was then reviewed by experts, including the Medical Director for our Child Health Advocacy Institute, the Medical Director for Community Affairs and Population Health, the Associate Director for the Center for Translational Research, the Research Division Chief of Biostatistics and Study Methodology, the Social Work Team Lead at the study site, and the Director of the Center for Child and Health Policy and Advocacy at an outside institution. The survey instrument was sent to each expert, and pertinent feedback was incorporated into the final draft of the survey. Next, cognitive interviews were conducted with 3 members of our institution’s parent navigator program (a program that provides guidance and support to families to access services), to obtain response process validity.²⁷  The cognitive interviews prompted the individual to think aloud while reading the questions on the survey to share how they interpreted the survey questions. This resulted in several questions being rephrased or answer choices reworded to improve readability. Finally, 12 members of the hospital’s Patient/Parent and Family Advisory Council (a council of patients and parents who advocate for quality care and system changes to enhance or improve the family experience) then piloted the electronic survey instrument. The pilot involved these members responding to the survey using the QR code. The Patient/Parent and Family Advisory Council also evaluated the survey for understandability, readability, and ease of use.

Survey analysis consisted of reviewing downloaded descriptive and summary data from REDCap. For each question, total respondents, missing responses, and breakdown of question responses in table format were reviewed. All data were recorded as total number of responses and percentages.

We enrolled 160 participants. Thirty-six percent were aged 35 to 44 years old, with 25% of participants aged 25 to 34 years old and ∼26% aged 45 to 55 years old. Most participants were female (77.5%). vThirty-one percent of respondents identified as Black or African American, 28% identified as White, and 28% identified as Hispanic or Latino (see Table 1). Eleven percent had not completed high school. Eighty-nine percent had taken their child to their pediatrician in the past year. Ten percent of the caregivers used the Spanish language version of the survey.

When asked about how they would prefer to be screened for social needs, 45% preferred to be screened by themselves in private, 9% preferred to be screened with a health care team member, but 37% were comfortable being screened either in private or with a health care team member. Only 3% of participants stated they would not want to answer any questions about screening for social needs. (See Table 2).

When asked by which method they would wish to be screened, 44% preferred electronic methods, 19% preferred speaking with someone, 7% preferred paper screening, and 29% were comfortable with any method. If answering social needs screening questions with a health care member, social workers were the most preferred (50%), followed by case managers (30%), nurses (20%), doctors (15%), and registration workers (0%), although 41% responded that it would be acceptable for any member of the team to perform the screening. When asked the reverse question, “Is there a member of the team that you would not want to ask you these questions,” 29% responded that they would not want the registration worker, 26% responded that they would not want the doctor, and 12% reported they would not want the nurse to ask the social needs screening questions. However, 60% reported it would be acceptable for anybody to perform the screening questions. It should be noted that >50% of respondents did not respond to either question regarding who to screen them because of the branching logic within the survey. There was not a strong preference as to when during the hospitalization to undergo screening, although most preferred not to be screened on the first day of hospitalization or while in the emergency department.

Apart from the domains of problems for paying utility bills, finding childcare, and neighborhood safety, >60% of respondents felt that it was important to be screened for each of the listed needs (Fig 1). More than 70% of respondents felt it was important to be screened for problems getting appointments because of medical insurance, problems paying for appointments or medications because of cost, having a place to call home, and problems with safety in the home.

In addition to perceived importance, caregivers also reported their comfort with being screened for social needs (Fig 2). More than 60% of respondents were comfortable responding to questions for all domains. In all but the question asking about problems with mold or pests in the home, participants’ level of comfort with social needs screening questions exceeded their reported perceptions of the importance of screening for the same domains. Between 40% and 50% of participants responded it would be acceptable to be screened for each of the social needs domains even if the hospital team could not offer immediate resources to meet their needs (Fig 3).

In our pilot study, we found that caregivers were overall comfortable being screened for social needs in the inpatient setting, and most felt it was important for their health care team to do this screening. Participants had mixed preferences about whether it was acceptable to be screened for social needs domains if there were no resources immediately available. Although almost one-half of our respondents preferred to be screened in private, 37% were open to being screened either in private or with a health care team member. Similarly, although many preferred the option of electronic screening, nearly one-third were accepting of either electronic or face-to-face screening with a health care team member. Social workers and case managers were the preferred health care providers to perform face-to-face screening in person, and only 3% of participants stated they would not want to answer any screening questions about social needs.

These data add to qualitative work by Leary et al that highlighted parental thoughts on the importance of screening for social needs and preferences toward electronic screening or social worker and case management engagement.²¹  Social needs reflect sensitive and private issues for families that they might not want to disclose to their children or others directly on their inpatient health care team. Electronic or paper screening allows caregivers to have privacy around these topics. Should in-person questions need to be asked, parental preference was for social work and case management team members to complete the screening. This is in contrast to previous research by Colvin et al, which revealed that 71% of caregivers on an inpatient pediatric unit wanted to be screened by a physician for social needs.¹⁸  Our findings suggest the importance of additional resource allocation to support social workers and other trained staff to aid care teams in SDOH screening to meet the preferences of all caregivers. Approximately one-quarter of caregivers in our study who responded to this question did not want to be screened for social needs by a physician. Potential reasons might include the lack of a preexisting relationship with the inpatient physician team or the expectation that the focus of the physician should be on the child’s acute illness. Other studies have revealed increased parental disclosure of social needs when using electronic formats, potentially due to concerns for social desirability or fear of judgment by their provider.^28,29  Despite the sensitive nature of these questions, our participants were largely in favor of being screened for social needs. Our results lend support for inpatient social needs screening for hospitalized children at our institution.

Screening for social needs in the pediatric outpatient setting has been demonstrated to improve the referral to and utilization of resources.^10,11,13  The potential for effective screening in other clinical sites has been demonstrated. At the Boston Children’s Hospital Hematology Clinic, 66% of patients with sickle cell disease reported at least 1 social need, and at a 2-week follow-up call, almost one-half of the families had contacted referrals for resources.³⁰  The inpatient setting may provide a unique screening opportunity in which different health care professionals have multiple opportunities to interact with, solicit information from, and provide resources addressing social needs to patients and their families. These additional screening opportunities over the days of hospitalization may impact postdischarge care.¹⁵  The inpatient setting also captures patients who may not have been seen in the outpatient setting because of a host of barriers, such as lapses in insurance coverage or transportation difficulties. Hospitalization itself may give rise to unanticipated social needs, such as caregivers missing work or lack of other supports.¹⁴ 

Growing evidence indicates that the families of hospitalized children have unmet social needs. A study by Banach published in 2016 revealed that ∼25% of hospitalized children suffered from food insecurity and 31% were candidates for supplemental nutrition assistance but were not connected to services.³¹  More recent studies reveal that 38% of pediatric inpatient families were food insecure or had unmet social needs.^14,32 

Our study adds to the growing body of literature that inpatient SDOH screening is valued by the families of hospitalized children.²¹  Our findings with respect to parental preferences, along with the prevalence of unmet social needs, reveal the imperative to screen hospitalized children for social needs.¹⁷  Our data suggest that caregivers are generally comfortable with screening and find screening for social needs important. Social needs, such as the ability to obtain housing and food, intersect with the mental and physical health of caregivers and their children. Some studies, such as by Vaz et al, have revealed that social needs among hospitalized children are uniformly common.¹⁵ 

The concept of screening for social needs without available resources for families has been the subject of concern and described as potentially “unethical”.³³  Meanwhile, providers may be hesitant to screen for social conditions when they are not able to offer help.^16,20,34  Others recommend as a core principle that social needs screening be initiated after available resources are identified.^35,36  Some experts counter that screening even without available resources improves public health surveillance, can identify patients who may need additional help in managing their condition, or help inform resource allocation.^34,37  Our findings indicate that only 40% of participants find SDOH screening to be acceptable even if resources are not available. As such, although some caregivers in our study may be comfortable with SDOH screening, it is important for the health care team to be equipped to provide support when a social need is identified.

Perhaps the most important limitation of our study was the small, single-center sample of caregivers surveyed. Our patient demographics may not be representative of all hospital admissions or other hospital systems. The demographic information we collected focused on parent/caregiver report rather than the demographics of the child, which could account for some of this discrepancy. Of note, ∼30% of our sample reported having a graduate degree, which could limit the generalizability to populations with different educational backgrounds. However, there is data that 30% of parents who have college or more education are considered low-income.³⁸  In addition, only a few participants completed the survey in Spanish. This may be due in part to a delay in the translation of our survey instrument. Likewise, the survey instrument was limited to caregivers with English or Spanish proficiency, so there may have been caregivers/patients who preferred other languages that were unable to participate. Our survey methods limited the extent to which we were able to ask follow-up probing questions about participant responses or nonresponses. Moreover, the reliance on QR code technology may have discouraged participants who had technical challenges. We attempted to address this by ensuring that study recruiters were available in person to help caregivers with any technical issues. It is possible that caregivers completed the survey >1 time, during the same or subsequent admissions. Finally, we did not collect any information about those who declined to participate in the study; therefore, we are unable to calculate an overall response rate to our survey. It is possible that those who chose not to respond may be different from the study group with respect to their demographics, comfort with screening, and perceived importance of screening. This may bias our results by either overestimating or underestimating the comfort/importance of screening and limits generalizability to the hospital-wide population.

Despite these limitations, our results support the notion that caregivers find the inpatient setting to be an acceptable time to be screened for SDOH and provide insights into the preferences of caregivers that may inform future widespread hospital-based social needs screening efforts. Once social needs screening is implemented, other areas for investigation include how to best share this information with the primary care provider, how to integrate this information into the electronic medical record based on caregiver preference, and how to optimize the impact of inpatient screening for social needs on referral and use of resources by caregivers.

The inpatient setting represents another opportunity to screen for social needs. Caregivers in this study report both being comfortable and finding it important for their health care team to perform this screening. Health care teams should consider screening in private using electronic means. Future studies should evaluate the impact of inpatient screening on referral and resource utilization, which may ultimately improve long-term patient outcomes.

We wish to thank the Patient Family Advisory Council at Children’s National, and the future physicians from George Washington University that aided in patient recruitment.