Screening for Adverse Childhood Experiences in Children: A Systematic Review

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.³⁵  The study protocol was registered with the International Prospective Register of Systematic Reviews (registration number CRD42020196442) on August 2, 2020.

A systematic literature search was performed with the input of a professional research librarian by using the Ovid Medline, PsycINFO, CINAHL, and Cochrane Controlled Register of Trials electronic databases. The search included relevant peer-reviewed articles published between January 1, 2009, and January 31, 2021. Articles were restricted to those published in English. Key words included “adverse childhood experiences screening” OR “child adversity screening” OR “benefits child adversity screening” OR “harm child adversity screening” OR “benefits child ACEs screening” OR “harm child ACEs screening.” The full list of search terms is found in the Supplemental Information.

Studies were included if they met the following criteria: (1) children aged 0 to 12 years who are screened for current adversity; (2) randomized clinical trials or quasi-experimental designs with a control group; (3) screening for ACEs and/or screening for broader social determinants of health (SDOH), such as poverty, food insecurity, housing instability or community violence; and (4) screening was conducted by a practitioner using either a questionnaire or interview with a standard set of questions. Exclusion criteria included studies that (1) focused on screening of adolescents (age >12 years) or adults, (2) screened for only 1 ACE, (3) were set in a low- or middle-income country, or (4) screened for adversity as part of inclusion criteria for an intervention and provided no outcome measures of screening.

Studies were independently double screened by 3 authors (S.L., K.P., L.C.) for inclusion by using Covidence software (v2431).³⁶  Any conflicts about inclusion were resolved through discussion with the primary author. Screening was followed by full-text review by 2 authors (S.L. and T.H.) to assess eligibility against the inclusion and/or exclusion criteria. All articles in full-text review were double screened. Disagreement was resolved through discussion between the 2 authors (S.L. and T.H.) or by referral to a third author (H.H.). The interrater agreement was >95%.

Two authors (S.L. and T.H.) independently extracted data from all included studies using a standard data extraction template in Covidence software. Data extracted included the following information: (1) study characteristics, including country location, research design, year of publication, number of participants; (2) participant characteristics, including age, sex, education level, race, age of child, and financial support and/or Medicaid (public health insurance); (3) screening tool and measurement of ACE or adversity; and (4) outcomes, including referral rates, uptake of referrals, community service engagement, mental health service involvement, and child and parent mental health. The extraction forms were compared, and consensus was reached with discussion between 2 authors (S.L. and T.H.).

Two authors (S.L. and T.H.) independently assessed the quality of the included studies using the ROBINS-I (risk of bias in nonrandomized studies of interventions)³⁷  tool for assessing the risk of bias in nonrandomized studies and the Cochrane risk-of-bias tool for randomized studies³⁸  in Covidence software. Discrepancies were resolved with discussion between 2 authors (S.L. and T.H.).

The PRISMA flow diagram (Fig 1) outlines the search strategy and results. The initial search identified 5816 articles. After removal of duplicates, 5039 articles were screened by title and abstract, and 45 articles were identified for full-text review. Three articles met the inclusion criteria after full-text review and were included in the analysis. By hand-searching references, we identified one additional article that met inclusion criteria.

A summary of the studies included in this systematic review is presented in Table 1. All 4 studies took place in a health setting in the United States, with 3 studies based in a primary care pediatric clinic and the remaining study based in a nurse home-visiting program. In all studies, screening for ACEs occurred as part of routine well-child care. Screening for ACEs was not undertaken opportunistically when families presented for other health-related care and only occurred at specific well-child appointments.

All 4 studies enrolled families with children aged <5 years, with mean ages between 2.5 and 25 months.^39–41  There were 3178 children enrolled across the 4 studies (range 336–1149). The study populations differed in ethnicity and enrollment in Medicaid (public health insurance program). Three studies had either mixed or majority Black populations and high rates (>80%) of Medicaid enrollment^39,41,42  compared with 1 study that had a predominately White population with only 12% of families enrolled in Medicaid.⁴⁰ 

There were a range of ACEs screened across the studies. Two studies^40,41  used the parent screening questionnaire (PSQ), which screens for a range of childhood adversities, including parental depression, parental substance abuse, harsh punishment, major parental stress, IPV, and food insecurity. One study³⁹  used 12 questions from the Children’s HealthWatch survey to screen for child care needs, food security, household heat, housing, parent education, and employment and used the Personal Health Questionnaire Depression Scale to screen for maternal depression. One study⁴²  used specific tools to screen for maternal depression (Edinburgh Postnatal Depression Scale), substance use (UNCOPE [Used, Neglected, Cut Down, Objected Preoccupied, Emotional Discomfort]), and IPV (Relationship Assessment Tool). Screening for ACEs was followed-up with motivational interviewing techniques by the practitioner who undertook the screening and planned referral to community-based services in all 4 studies.

There was moderate risk of bias in all studies on quality assessment (Tables 2 and 3), with significant missing data and low fidelity to study intervention,⁴²  lack of blinding by outcome assessors and baseline differences,^40,41  and unclear random assignment and allocation bias.³⁹ 

Meta-analysis was not possible because of the small number of studies and the significant heterogeneity in terms of interventions and outcomes reported. Therefore, a narrative description of outcomes is presented.

Screening for ACEs improved detection of some adversities in children in all studies compared with controls. Only 2 studies reported an improvement in screening rates. Dubowitz at al⁴¹  reported that screening for adversity improved by between 20% and 25%, whereas Dauber et al⁴²  reported an improvement between 4% and 12% overall. Although home visitors reported that they were more likely to discuss substance use, IPV, and maternal depression with families, there was no statistically significant improvement in risk identification for substance use or IPV.⁴² 

Two studies reported the effect of screening for ACEs on referral rates. Garg et al³⁹  reported that screening resulted in improved referrals to community-based health and social services, with mothers in the intervention group receiving significantly more referrals than those in the control group (70% vs 8%; adjusted odds ratio [OR] = 29.6; 95% confidence interval [CI] 14.7–59.6). At follow-up (at 12 months), more mothers in the intervention group were enrolled in community services than mothers in the control group (39% vs 24%; adjusted OR = 2.1; 95% CI 1.2–3.7).³⁹  This contrasted with a study by Dauber et al⁴²  in which there was minimal impact on referrals to community services (maternal depression referral OR 1.11, 95% CI 0.72–1.72; substance use referral OR 16.79, 95% CI 2.04–138.21; and IPV referral OR 2.49, 95% CI 1.27–4.91). However, there was low fidelity to the study protocol, so even when risks were identified through screening, only 32% of families received any case management including referrals.⁴² 

Two studies by Dubowitz et al^40,41  reported on rates of referral to child protection services. Screening for ACEs in a low-socioeconomic population resulted in a decrease in referral to child protection services for the intervention group (OR 1.5; 95% CI not reported).⁴¹  However, use of the same screening tool in a middle and/or high socioeconomic population had no effect on referrals to child protection services.⁴⁰ 

None of the studies reported mental health outcomes for children or parents. Most of the studies had limited follow-up between 6⁴²  and 12 months,^39,40  with one study reporting no defined length of follow-up.⁴¹  None of the studies measured harms of screening for ACEs.

This systematic review sought to identify whether screening for ACEs improves identification of adversity, referral to services, uptake of services, and better mental health outcomes for children and their parents. There is some evidence that screening for ACEs improves identification of adversity and limited evidence that screening may improve connection to community-based services and uptake of these services. There were no reported measures of mental health outcomes after screening for children or parents. This systematic review did not establish clear evidence that screening for ACEs improves any of our outcomes of interest. The evidence is limited because there were only 4 studies identified in our time period of interest, all of which were conducted in the United States and were of moderate quality.

From the studies in this systematic review, there was limited evidence that screening for ACEs improves identification of adversity, aligning with evidence from previous literature examining screening for ACEs or SDOH in children and parents.^43–47  However, it is difficult to compare screening studies because all the studies in our review screened for different ACEs, using different tools, and one study screened for SDOH in addition to screening for one ACE.³⁹  This is in part due to the lack of a universal definition of ACEs and a lack of agreement about which ACEs or other adversities for which to screen.⁷  In addition, increased rates of identification do not necessarily equate to better detection of adversity, which can be conceptualized as accurate identification of families living with adversity who would benefit from interventions. Moreover, although ACEs have been clearly demonstrated to be linked to poor health outcomes at a population level, ACEs cannot be used to predict an individual’s risk for chronic illness or mental health difficulties. Screening for ACEs in adults and children has limited ability to inform individual risk of poor health outcomes.⁴⁸ 

This systematic review revealed mixed findings for the effect of screening for ACEs on connection to community-based services. Although screening for ACEs or SDOH can lead to increased referral rates to community-based services,^43,44,49,50  referral and uptake of community services after a positive screen is dependent on health professional, parent, and systems factors. There are few current practice data on rates of referral to community services. Garg et al⁵¹  conducted a survey of pediatricians in the United States and found that 85% reported referring at least 1 family to community services in the past year; however, referrals for housing services and food support were low. From the 2 studies in this systematic review that reported on referral rates, the baseline referral rates to community services were low, between 2% and 8%.^39,42  There are a number of barriers for providers in undertaking and responding to screening. Primary care providers, in the United States, report that they have incomplete knowledge about services and, thus, are less likely to make referrals.⁵²  Primary care providers may also be ambivalent about screening for ACEs in part because of the lack of clarity about what referrals to make and concerns about lack of community resources.^19,53,54  Pediatricians are less likely to refer to services with long waiting times.⁵⁵  Uncontrolled studies have demonstrated low referral rates after positive ACEs or SDOH screening, with <30% of families having external referrals made in clinic-based or home visiting–based screening.^45,52,56,57  Other studies have shown a variable acceptance of referrals by parents. Of parents who screened positive for ACEs in uncontrolled studies, between 53% and 77% consent for community services,^52,53,58  with one study reporting rates as low as 14% of parents taking up referrals to community services.⁵²  In qualitative studies of parental engagement, researchers have identified a number of barriers, including a lack of time, stigma, and a lack of perceived benefit from the referral.⁵⁹ 

Referrals to child protection services is also a factor when considering ACEs screening. Dubowitz et al^40,41  reported lower rates of child protection notifications in their studies included in this review, which was reported as a positive finding. However, one could also expect that ACE screening may lead to increased reporting to child protection. This could impact on the veracity of the screening tests because parents are less likely to report adversity if they feel that child protection may be notified.⁶⁰ 

None of the studies identified in this systematic review reported on child or parent mental health outcomes. Early exposure to ACEs is associated with poor emotional regulation and aggression in the first 3 years of life,⁶¹  yet none of the studies assessed socioemotional health of the children enrolled. Dubowitz et al^40,41  assessed harsh parenting in both studies included in this review and reported an improvement in the Conflict Tactics Scale, Parent-Child (CTSPC), in the intervention group compared with the control group. This is an interesting finding because there are a number of parenting factors that are associated with poor child mental health, namely, abusive parenting, aversiveness, and interparental conflict.⁶²  In the studies by Dubowitz et al,^40,41  the improvement in parent-child relationship may be a proxy for an improvement in parental mental health.

An important consideration when screening for a disease or condition is to assess the harm of the screening test. All screening tests do harm; however, it is important to balance whether the harm is offset by the effectiveness of the screening.⁶³  Other research has indicated that there may be harms associated with just asking about ACEs. Although not explicitly an aim of this review, none of the included studies reported any assessment of harm. This contrasts with the consistent trend that ∼3% to 5% of parents report that ACEs screening is uncomfortable or harmful.^9,20,23,64  This reported harm from screening is not negligible and may outweigh the potential benefits from ACEs screening. The focus on adversities may lead to stigmatization of the family and families may be less willing to engage.¹³  This is particularly relevant for families who are from an ethnic minority group who tend to have a high burden of ACEs.⁶⁵  Labeling a child as high risk for a mental health disorder may lead to negative psychological effects and confirmation bias.²⁴ 

This is the first review, to our knowledge, that has examined the impacts of ACEs screening on identification, referrals made, uptake of referrals, and child and parent mental health. We conducted a rigorous systematic search across 4 databases and hand searched the references of included articles. We followed a rigorous process to select studies (with a high interrater agreement), and 2 authors independently assessed for bias. There are some limitations of this review. As with all systematic reviews, there is a potential publication bias with only studies with positive results being published. In this review, we intentionally limited studies to randomized controlled trials (RCTs) or controlled trials in which researchers assessed the effect of screening for ACEs. This was to limit bias so that the outcome of screening for ACEs could more accurately be assessed. However, this limited the number of studies that were identified because there are few controlled trials published. All included studies were conducted in the United States, which limits the generalizability of these results to other health systems and populations. We included only English-language articles; however, given we found no ACEs screening trials implemented outside of the United States, we believe that there is unlikely to be similar trials conducted and published in languages other than English.

This systematic review has implications for health policy and health service delivery. A number of jurisdictions across the United Kingdom, United States, and Australia are implementing population-based ACEs screening, whether through universal platforms (eg, well-child care) or more targeted platforms (eg, pediatricians).^9–13  From this systematic review, there is evidence that screening can improve detection of adversity and may improve referrals to community services, but there are no data on whether ACE screening improves child or parent mental health outcomes. All of the studies in this review were conducted within the health sector and undertaken during routine well-child assessments. This has implications for other health services that do not have funded yearly well-child assessments outside of infancy. This also has implications about generalizability of the results to population screening. None of the available evidence has demonstrated the efficacy, safety, and cost effectiveness of screening for ACEs at a population level. Furthermore, there is no current evidence about how ACE screening could be integrated into the current health system. This is a case in which health policy has outpaced evidence and caution needs to be employed. Further robust research, comprising rigorous, adequately powered RCTs is needed to determine if screening for ACEs can lead to improved outcomes, or unintentional harms, before widespread screening is advisable.

There is limited evidence, all arising from the United States, that screening for ACEs improves identification of childhood adversity and referrals to community services but no evidence for improvements in mental health outcomes for children or parents. Further RCTs are needed, in settings outside the United States, that measure both harms and benefits of screening. The relative merits of screening for ACEs, with and without screening for SDOH, should also be established through robust trials. More research is needed to examine the impact of screening for ACEs on parent-child relationships and mental health. Finally, understanding the barriers to referrals and how best to support families to take up referrals to community services once ACEs are identified is paramount if we are to realize the hypothesized benefits of ACEs screening on child and parent mental health.

ACE

adverse childhood experience

CI

confidence interval

CTSPC

Conflict Tactics Scale, Parent-Child

IPV

intimate partner violence

OR

odds ratio

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analysis

PSQ

parent screening questionnaire

RCT

randomized controlled trial

SDOH

social determinants of health