The Relationship Between Social Risks and the Mental Health of School-Age Children in Primary Care

Abstract

Background and Objectives:

The study’s goal was to measure the association between social risks and the mental health of school-age children in primary care.

Methods:

We conducted a cross-sectional study in an urban safety-net hospital-based pediatric clinic using data collected from two standard screening tools administered at well child care visits for children age 6–11. Psychosocial dysfunction was measured with the Pediatric Symptom Checklist-17 (PSC-17) and six social risks (caregiver education, employment, child care, housing, food security, and household heat) were measured with the WE CARE screener. Multivariable linear and logistic regression analyses were conducted to measure the association between scores while controlling for sociodemographic characteristics.

Results:

Among N=943 patients, cumulative social risks were significantly associated with a positive PSC-17 total score (Adjusted Odds Ratio [aOR] 1.2; 95% CI [1.1–1.5]; p=0.02), indicating psychosocial dysfunction. Children with ≥3 social risks were 2.4 times more likely to have a positive PSC-17 total score compared to children with <3 social risks (95% CI [1.5–3.9]; p<0.001). Of the individual social risks measured, only food insecurity significantly predicted a positive PSC-17 total score (aOR 1.9; 95% CI [1.1–3.2]; p=0.02) and attention score (aOR 1.9; 95% CI [1.1–3.4]; p=0.03).

Conclusion:

Number of risks on a social risk screener was associated with psychosocial dysfunction in school-age children. Food insecurity was the only individual risk associated with psychosocial dysfunction, in particular attention problems. Screening tools for social risks could be used to identify at-risk children whose mental health may be adversely impacted by their social conditions.

Introduction

Psychiatric disorders are among the most prevalent conditions in childhood, affecting more than one in five children in the United States (U.S.),^1,2 and causing significant lifelong morbidity and increased mortality.^2–5 Socioeconomically disadvantaged children are at even higher risk of psychiatric conditions than their more advantaged peers.^6–8 Children raised in poverty, currently over 20% of children in the U.S.,⁹ are also less likely to access treatment¹⁰ and are more vulnerable to the detrimental consequences of mental health problems, such as academic failure and involvement with the justice system.^11–15 The increased risk for mental health problems in socioeconomically disadvantaged children is in part related to adverse living and social conditions - collectively referred to as social determinants of health (SDoH) - including food insecurity, housing instability, and lack of quality child care.⁸ There is a growing body of literature supporting the contribution of social risk in childhood to physical, developmental, and mental health outcomes. Most prior studies describe the impact of cumulative risk, demonstrating that health outcomes worsen as risks accumulate.^8,16,17 Other studies have examined individual risks. For example, homelessness and food insecurity have both been linked to higher rates of psychiatric symptoms.^18–20

Given the evidence that social risks have on lifelong health, there is increasing interest in addressing these as part of routine preventive pediatric health care. The American Academy of Pediatrics (AAP) 2016 policy statement on Poverty and Child Health recommended screening for social determinants that can be addressed at pediatric visits.¹¹ Previous work by our study team demonstrated that it is feasible to screen for certain social risks and needs in the context of well child care and increase receipt of community resources for low income families.²¹ However, while many pediatric primary care practices have begun to adopt such screening,²² it is not known whether these tools could help to identify children at risk of mental health problems due to their social adversity. Screening for social risks could also be a key component of early intervention strategies for children at risk for mental health problems, or a component of multimodal treatments for children with psychiatric symptoms.

Thus, we sought to examine the relationship between six social risks (caregiver education, employment, child care, housing, food security, and household heat) that can be identified and addressed in primary care using a common social risks screener (WE CARE), and mental health problems in school-age children, as reported by caregivers on standardized screening tools administered at well child visits in a large, urban, safety-net hospital-based pediatric practice. We hypothesized that a higher number of social risks on the WE CARE screener would be associated with higher odds of mental health problems. We were also interested in whether: 1) children with social risks also had more mental health problems, and whether this relationship was specific to certain symptom clusters (e.g. anxiety/depression vs. behavioral vs. attention issues); 2) a certain threshold of social risks could distinguish children whose mental health problems might be exacerbated by these risks; and 3) children with certain individual social risks were more likely to suffer from mental health problems. Our goal was to inform how social risks screening tools could be used in clinical practice for children with social risks, mental health needs, or both.

Methods

Research Design and Setting

We conducted a cross-sectional analysis of de-identified electronic medical record (EMR) data for children age 6–11 years old seen for well child visits at a large, urban, safety-net hospital-based pediatric clinic between September 1, 2016 and August 31, 2017. We extracted scores on two brief standardized questionnaires that caregivers complete as a part of well child visits for school-age children: 1) the Pediatric Symptom Checklist-17 (PSC-17),^23,24 which assesses psychosocial functioning; and 2) the Well Child Care, Evaluation, Community Resources, Advocacy, Referral, Education (WE CARE) screener,^21,25 which measures six social risks. Both screening tools were offered in multiple languages (English, Spanish, Portuguese, Haitian Creole) and given to caregivers at check-in. During the study period, results on both screeners were collected and entered electronically into the EMR by practice medical assistants. We also extracted sociodemographic data from the EMR including patient’s age, gender, race, ethnicity, language, and primary health insurance.

The study was approved by the Boston University Medical Campus Institutional Review Board.

Measures and Variables

Psychiatric Symptoms.

Psychiatric symptoms were measured with the PSC-17, a caregiver-report measure of children’s psychosocial functioning. The PSC-17 is widely used as a psychosocial screening instrument in pediatrics, has been validated in many populations and languages,^24,26–28 and has both high internal consistency (Cronbach’s α = 0.87) and test-retest reliability (ICC = 0.85).^23,24 It was derived via exploratory factor analysis from the original 35-item psychiatric symptoms questionnaire.²² Caregivers rate symptoms on a three-point Likert scale, with 0 = never, 1 = sometimes, and 2 = often. Scores for each item are summed and reported as a total score (all 17 items) and three subscales: internalizing (5 items), externalizing (7 items), and attention (5 items). Higher scores indicate higher levels of impairment, and a positive total score (≥15) indicates psychosocial dysfunction and need for further evaluation.^23,26 Children with an elevated externalizing subscale score (≥7), attention subscale score (≥7), or internalizing subscale score (≥5) usually have clinically significant problems with behavior, attention, or anxiety and depression, respectively.^23,24

Social Risks.

Social risks were measured by the WE CARE screener, a caregiver-report questionnaire written at a third grade reading level which assesses six basic risks in the following order: caregiver education, employment, child care, housing, food security, and household heat.^21,25 The original questionnaire was developed using an interdisciplinary collaborative approach to include only items for which community resources were available. The questionnaire has established face and content validity as well as high 2-week test-retest reliability (r=0.92).²⁵ Previous research has demonstrated that WE CARE is feasible at well child visits with low-income children, and that it can lead to the receipt of more community resources and reduced social risks for families.^21,25 Table 1 lists the WE CARE questions and answers, including risk represented by each question and scoring information.

Table 1.

WE CARE Survey Questions and Coding.

Risk	Survey question (yes/no answers)*	Answer indicating risk present**
Education	Do you have a high school degree?	No
Employment	Do you have a job?	No
Child care	Do you risk daycare for your child?	Yes
Housing	Do you think you are at risk of becoming homeless?	Yes
Food	Do you always have enough food for your family?	No
Household heat	Do you have trouble paying your heating bill for the winter?	Yes

^*Questions are listed in same order as the survey administered to families.

^**A score of 1 was assigned to each answer indicating a risk in order to calculate a cumulative risks score.

Data Analysis

We used data only from children who had complete answers to both questionnaires (WE CARE and PSC-17). Cumulative social risks were calculated by assigning a value of “1” to a risk, and “0” to no risk; therefore, items for education, employment, and food were reverse-coded (Table 1). The resulting score (ranging from 0 to 6) indicates the number of current, caregiver-reported social risks. In order to examine the association between PSC-17 scores and cumulative and individual social risks, we performed chi-squared tests and calculated odds ratios. We plotted a Receiver Operating Characteristics (ROC) curve for WE CARE score to determine a cut-off that best distinguished children with a positive PSC-17 total score from those with a negative score. We performed multiple linear regression analyses to examine the association between cumulative social risks and PSC-17 scores after adjusting for the following potential confounding variables – age (in years), gender (male vs. female), race/ethnicity (4 categories as listed in Table 2, with Non-Hispanic White as reference), language (English vs. other), and insurance (public vs. other).⁵ We performed multivariable logistic regression analyses to assess whether one or more social risks predicted positive PSC-17 scores¹ after adjusting for the same potential confounders. We performed additional multivariable logistic regression analyses with all six individual social risks as predictors to examine the association between individual risks and PSC-17 scores, adjusting for potential sociodemographic confounders and checking for interactions among the six individual risks. For all statistical analyses, significance was set to α = 0.05. We used Statistical Analytical Systems (SAS) version 9.4 for data management and analysis.²⁹

Table 2.

Sample characteristics.

	N=943
Age in years: Mean ± SD	8.5 ± 1.7
Sex: n (%)
Male	472 (50.1%)
Female	471 (49.9%)
Race/Ethnicity: n (%)
Hispanic/Latino/Spanish	154 (16.3%)
Non-Hispanic Black	486 (51.5%)
Non-Hispanic White	47 (4.9%)
Non-Hispanic Other/Declined	256 (27.2%)
Language: n (%)
English	610 (64.7%)
Non-Englishǂ	333 (35.3%)
Haitian Creole	193 (20.5%)
Port Creole/Cape Verdean	42 (4.5%)
Spanish	37(3.9%)
Other	61 (6.5%)
Insurance: n (%)
Public	700 (74.2%)
Commercial	236 (25.0%)
Self-pay	6 (0.6%)
Other	1 (0.1%)
Positive PSC-17ǂ ǂ scores: n (%)
Internalizing (≥ 5)	67 (7.1%)
Externalizing (≥7)	86 (9.1 %)
Attention (≥7)	99 (10.5 %)
Total (≥15)	98 (10.4%)
WE CAREǂ ǂ ǂ social risks: n (%)
Parent Education	217 (23.0%)
Employment	289 (30.7%)
Child care	106 (11.2%)
Housing	50 (5.3%)
Food security	135 (14.3%)
Household heat	228 (24.2%)
Number of social risks: n (%)
0	402 (42.6%)
1	236 (25.0%)
2	175 (18.6%)
3	85 (9.0%)
4	41 (4.4%)
5	4 (0.4%)
6	0 (0%)

^ǂNon-English language includes Haitian Creole, Port Creole/Cape Verdean, Spanish, and others.

^{ǂ ǂ}PSC-17 = Pediatric Symptom Checklist, 17 item version.

^{ǂ ǂ ǂ}WE CARE = Well Child Care, Evaluation, Community Resources, Advocacy, Referral, Education.

Results

Sample characteristics

There were 3345 patients age 6–11 years old seen for well child visits during the study time period, and 943 (28.1 %) of these patients had complete results on both questionnaires entered into the EMR. The sample of patients with both completed questionnaires did not differ from the remainder of the sample based on sociodemographic characteristics (see Table 2).

Among the 943 children with complete WE CARE and PSC-17 screeners, 50.1% were male with a mean age of 8.5 years (SD = 1.7) (Table 2). The majority identified as non-Hispanic Black (51.5 %); 16.3% as Hispanic/Latino/Spanish, and 4.9% as non-Hispanic White. English was listed as the family’s primary language in the EMR for most patients (64.7%) followed by Haitian (20.5%), Port Creole/Cape Verdean (4.5%), Spanish (3.9%), and other (6.5%). Most children (74.2%) were publicly insured.

A quarter of the sample had at least one social risk. Table 2 shows the percentage of children with individual and cumulative risks (e.g. one, two, three risks, etc). The most common social risks were employment (30.7%), household heat (24.2%), and caregiver education (i.e., lack of high school diploma or GED) (23.0%).

Approximately 10% of children had a positive PSC-17 total score, indicating psychosocial dysfunction (Table 2). The externalizing and attention subscales identified about 10% of the sample with clinically significant symptoms in that realm, consistent with previously published risk rates.²⁴ The internalizing score identified about 7% of children with clinically significant symptoms, which was less than rates reported in large epidemiologic samples.²⁴

Association between cumulative social risks and psychosocial dysfunction

In the multivariable logistic regression model, cumulative social risks predicted significantly increased odds of a positive PSC-17 total score (Wald X² = 5.16; adjusted Odds Ratio [aOR] 1.2; 95% CI [1.1 – 1.5]; p-value = 0.02) (Table 3), indicating psychosocial dysfunction. Plotting a ROC curve with WE CARE against the PSC-17 total score demonstrated that we could most reliably distinguish children with a positive PSC-17 total score at a cut-off of 3 social risks on WE CARE (Figure 1A). Children with 3 or more social risks had 2.4 times the odds of having a positive PSC-17 total score compared to children with less than 3 risks (X² = 13.90; 95% CI [1.5–3.9]; p-value <0.001) (Figure 1B). Likewise, in the multivariable logistic regression model, children with 3 or more social risks had 2.4 times the odds of having a positive PSC-17 total score compared to children with less than 3 social risks (Wald X² = 9.34; aOR 2.3; 95% CI [1.4 – 3.9]; p-value = 0.002), after adjusting for sociodemographic characteristics.

Table 3:

Association between cumulative social risks measured by WE CARE^ǂ and domains of psychiatric risk measured by positive PSC-17^{ǂ ǂ} total and subscale scores.

Outcome: Positive PSC-17 Score	Wald X2ǂ ǂ ǂ	p-value	aOR (95% CI)ǂ ǂ ǂ ǂ
Positive Total Score	5.16	0.02*	1.2 (1.1–1.5)
Positive Internalizing Score	0.28	0.59	1.1 (0.9–1.3)
Positive Attention Score	3.65	0.06	1.2 (1.0–1.4)
Positive Externalizing Score	1.46	0.23	1.1 (0.9–1.4)

Note: Multivariable logistic regression model with cumulative social risks as the main predictor and positive PSC-17^{ǂ ǂ} score (total or one of 3 subscales) as the outcome. Models were adjusted for the following confounders: age, sex, race/ethnicity, language, and insurance.

^*Denotes significant difference at p-value = 0.05.

^ǂWE CARE = Well Child Care, Evaluation, Community Resources, Advocacy, Referral, Education.

^{ǂ ǂ}PSC-17 = Pediatric Symptom Checklist, 17 item version.

^{ǂ ǂ ǂ}Wald X² indicates Chi-squared test statistics.

^{ǂ ǂ ǂ ǂ}aOR = adjusted Odds Ratio and CI = Confidence Interval.

Figure 1a.

Receiver Operating Characteristic curve of positive PSC-17 total score at possible WE CARE cut-offs

Figure 1b.

Odds of positive PSC-17 total score at possible WE CARE cut-offs

Multivariable logistic regression models with the three PSC-17 subscales as outcomes revealed that higher numbers of social risks did not increase the odds of having a positive attention subscale score (Wald X² = 2.79; aOR 1.2; 95% CI [0.9–1.4]; p-value = 0.09), a positive externalizing subscale score (Wald X²=2.43; aOR 1.2; 95% CI [0.9–1.4]; p-value = 0.12) or a positive internalization score (Wald X² = 0.28; aOR 1.1; 95% CI [0.9–1.3]; p-value = 0.59).

Association between cumulative social risks and psychiatric symptoms

Due to previous literature demonstrating that PSC-17 scale cut-offs may be specific but not sensitive for psychiatric disorders³⁰ - and thus concern that a multivariable logistic regression model using these cut-offs would miss clinically relevant associations - we also performed a multivariable linear regression model to assess the relationship between cumulative risks and continuous PSC-17 total and subscale scores. In this model, cumulative social risks were significantly associated with the PSC-17 total score (t-statistics = 2.51; β_estimate = 0.39; p-value = 0.01) as well as the PSC-17 attention subscale score (t-statistics=2.57; β_estimate = 0.18; p-value = 0.01) and the internalizing subscale score (t-statistics= 2.05; β_estimate = 0.10; p-value = 0.04), but not with the externalizing subscale score (t-statistics = 1.56; β_estimate = 0.11; p-value = 0.12).

Association between individual social risks and psychosocial dysfunction

Of the six individual social risks, food security and employment were significantly associated with a positive PSC-17 total score (Table 4 and Figure 2). Children with food insecurity were 2.1 times as likely to have a positive PSC-17 total score compared to children with food security (X² = 9.22; 95% CI [1.3–3.5]; p-value = 0.002). Children with an unemployed caregiver were 1.6 times as likely to have a positive PSC-17 total score (X² = 4.31; 95% CI [1.1–2.4]; p-value = 0.03).

Table 4.

Odds of positive total PSC-17^ǂ score in children with vs. without individual social risks on WE CARE^{ǂ ǂ}.

	Percent (%) of children with a positive PSC-17 total score		X2ǂ ǂ ǂ	p-value	OR (95% CI)ǂ ǂ ǂ ǂ
	No (−) Risk	With (+) Risk
Individual risks:
Parent Education	10.3%	10.6%	0.01	0.91	1.0 (0.6–1.7)
Employment	9.0%	13.5%	4.31	0.03*	1.6 (1.1–2.4)
Child care	10.0%	13.2%	1.02	0.31	1.3 (0.7–2.5)
Housing	10.1%	16.0%	1.78	0.18	1.7 (0.7–3.7)
Food security	9.2%	17.8%	9.22	0.002*	2.1 (1.3–3.5)
Household heat	9.7%	12.7%	1.75	0.19	1.4 (0.9–2.2)

^*Denotes significant difference at p-value = 0.05.

^ǂPSC-17 = Pediatric Symptom Checklist, 17 item version.

^{ǂ ǂ}WE CARE = Well Child Care, Evaluation, Community Resources, Advocacy, Referral, Education.

^{ǂ ǂ ǂ}X² = Chi-squared test statistic.

^{ǂ ǂ ǂ ǂ}OR = Odds Ratio and CI = Confidence Interval.

Figure 2.

Bivariate analysis comparing percent of children with positive PSC-17 total scores with and without individual social risks.

In the multivariable logistic regression model including the six individual social risks (Table 5), only food security was significantly associated with positive PSC-17 total score after adjusting for sociodemographic characteristics and other social risks (Wald X² = 5.48; p-value = 0.02). Children with food insecurity had increased odds of a positive PSC-17 total score compared to food secure children (aOR 1.9; 95% CI [1.1–3.4]). Children with food insecurity also had increased odds of a positive PSC-17 attention subscale score (Wald X² = 4.71; aOR 1.9; 95% CI [1.1 – 3.4]; p = 0.03) compared to food secure children.

Table 5.

Association between individual social risks on WE CARE^ǂ and positive total PSC-17^{ǂ ǂ} scores.

Predictor	Wald X2ǂ ǂ ǂ	p-value	aOR (95% CI)ǂ ǂ ǂ ǂ
Parent Education	0.06	0.80	0.9 (0.5–1.6)
Employment	0.35	0.56	1.2 (0.7–1.9)
Child care	0.38	0.54	1.2 (0.7–2.3)
Housing	0.35	0.55	1.3 (0.5–2.9)
Food security	5.48	0.02*	1.9 (1.1–3.4)
Household heat	0.05	0.82	1.1 (0.6–1.7)

Note: Multivariable logistic regression outcome is a Positive Total Score on the 17-item Pediatric Symptom Checklist (PSC-17). There are six main predictors. For each of the main predictors, the model was adjusted for the other social risks and potential confounders – age, sex, race/ethnicity, language, and insurance.

^*Denotes significant difference at p-value = 0.05.

^ǂWE CARE = Well Child Care, Evaluation, Community Resources, Advocacy, Referral, Education.

^{ǂ ǂ}PSC-17 = Pediatric Symptom Checklist, 17 item version.

^{ǂ ǂ ǂ}Wald X² indicates Chi-squared test statistics.

^{ǂ ǂ ǂ ǂ}aOR = adjusted Odds Ratio and CI = Confidence Interval.

Discussion

Over half of the diverse children at this urban, hospital-based, primary care practice had at least one social risk, and the number of social risks were significantly associated with psychosocial dysfunction. School-age children with food insecurity and those with 3 or more social risks - identified using a simple screen of six social risks in primary care - had over twice the rate of psychosocial dysfunction as their peers. These findings are novel in that they suggest that urban children with 3 or more risks on the WE CARE screener are at particularly high risk for psychosocial dysfunction, and distinguish food insecurity from other risks as most problematic for child mental health, particularly with regard to attention problems. These results extend the current understanding of the relationship between social risks and mental health of school-age children, and how social risk screening can be used in primary care to identify children at high risk for mental health problems related to these risks.

We found that a higher number of social risks were associated with more psychiatric symptoms and higher rates of psychosocial dysfunction, consistent with past literature supporting a cumulative risk model of the deleterious effects of psychosocial stressors on psychiatric outcomes.^{16,17,31–33} However, while existing literature largely supports a linear cumulative risk model of the association between childhood psychosocial stressors and psychiatric problems in both childhood and adulthood, our study findings offer several important new insights. First, we measured only specific social risks that could be addressed in pediatric primary care using the validated screening tool, WE CARE.²¹ This is the first study to show that children with higher WE CARE scores are at increased odds of mental health problems, and that a score of 3 on WE CARE could be used to identify children most vulnerable to psychosocial dysfunction. Children with 3 or more risks reported by caregivers (13.8% of our sample) had more than twice the odds of having a positive PSC-17 total score compared to peers with lower scores. Thus, children with a score of 3 or higher on WE CARE may need more intensive monitoring for mental health problems, and may be important candidates for preventive mental health interventions. Furthermore, our findings suggest that children with positive PSC-17 scores should be screened for social risks that can be addressed as part of comprehensive treatment planning.

Second, we found that all six modifiable social risks were not equal predictors of psychosocial dysfunction in children. Food insecurity - present in 1 of every 7 families in our sample - was the only individual risk associated with increased psychosocial dysfunction once we controlled for sociodemographic and other social risks. Food insecurity was associated with twice the odds of a positive PSC-17 total score, and with increased odds of a positive PSC-17 attention score as well. This is consistent with existing literature that food insecurity negatively impacts general cognitive development and psychological functioning.^19,34–37 Others have also found a specific association between food insecurity and attention/hyperactivity problems in young children,^18,38 and our findings suggest that this association continues into the school-age, which could negatively impact academic success and peer relationships. Food insecurity may impact child mental health and increase attention problems via a variety of hypothesized mechanisms. First, inadequate nutrition may lead to low dietary intake of essential vitamins, iron and fatty acids, which may worsen attention and cognition.¹⁸ Second, food insecurity may cause high levels of stress, anxiety and depression in caregivers, which may in turn cause mental health problems and inattention/hyperactivity in children.³⁹ Second, food insecurity may be associated with other risk factors that we did not measure, such as single parenthood, severe poverty, or unwell caregivers. Although more work is needed to understand the contributions of these factors to our results, our findings nonetheless support the importance of detecting and addressing food insecurity as a critical part of preventive pediatric health care. Children with food insecurity should also be monitored for mental health problems - particularly attention and hyperactivity - and could be candidates for preventive mental health interventions initiated in primary care.

Importantly, social risk screening is not synonymous with, nor a substitute for, mental health screening. As expected, the majority of children with multiple social risks did not have positive PSC-17 scores. In addition, the majority of children with positive PSC-17 scores did not have multiple social risks. However, our findings highlight the fact that these tools could be used together to enhance treatment planning, and that the mental health of children with social risks may need to be monitored closely.

Our study has some limitations. This was a retrospective chart review of cross-sectional data, and thus we cannot establish temporal relationships or causality of the associations found. The screeners were administered as part of standard clinical practice at well child visits in a large, urban, safety net hospital-based pediatric practice, and only about one-third of children had both completed questionnaires in their medical records. This was largely due to a low completion rate of WE CARE, which had been recently implemented, and highlights the challenges of beginning social risks screening in a busy primary care practice. Despite this limitation, we were still able to retain a large sample with information on social risks and psychosocial functioning for analysis, and there were no meaningful differences between children with complete and incomplete questionnaires. We also did not have data on whether caregivers thought they needed help related to each risk, and so we do not know whether families’ social needs matched these measured risks, and whether this would more or less robustly predict psychosocial functioning. We also had very few children with four or more social risks on the WE CARE screener, and thus we were not able to reliably report on the effect of social risk accumulation past three risks. Children in our study were mostly publicly insured and non-white; thus, our findings may not generalize to other populations. We did not have information about income, because income is not included in WE CARE and the EMR does not otherwise contain this information. Thus, we used public insurance as a proxy, which may have obscured income differences that would have clarified the role of income in our findings. Finally, due to the sample size, we may have been inadequately powered to detect other clinically meaningful differences. Despite these limitations, our study extends the current literature on the association between social risks and psychosocial dysfunction in children and their detection in primary care.

Conclusions

In this study, we found a positive association between cumulative social risks and psychosocial dysfunction in school-age children. Individually, food insecurity was the only risk examined that was positively associated with psychosocial dysfunction, in particular attention problems. Our results suggest that screening for food insecurity and other social risks using a primary care based tool such as WE CARE can be combined with well-validated psychosocial screening tools such as the PS-17 to identify and appropriately treat children whose mental health may be impacted by adverse social conditions.

What’s New:

School-age children with food insecurity and those with 3 or more social risks - identified using a screen of six social risks in primary care - had over twice the rate of psychosocial dysfunction as their peers.

Abbreviations:

PSC-17

Pediatric Symptom Checklist, 17-item version

WE CARE

Well Child Care, Evaluation, Community Resources, Advocacy, Referral, Education

SDoH

Social Determinants of Health

ACEs

Adverse Childhood Experiences

ICC

Intraclass Correlation Coefficient