Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Enhancing Early Childhood Mental Health Primary Care Services: Evaluation of MA Project LAUNCH

Enhancing Early Childhood Mental Health Primary Care Services: Evaluation of MA Project LAUNCH Objectives The purpose of this study was to evaluate the efficacy of an innovative early childhood mental health intervention, Massachusetts Project LAUNCH. Early childhood mental health clinicians and family partners (paraprofessionals with lived experience) were embedded within community pediatric medical homes. Methods A longitudinal study design was used to test the hypotheses that (1) children who received services would experience decreased social, emotional and behavio- ral problems over time and (2) caregivers’ stress and depressive symptoms would decrease over time. Families who were enrolled in services and who consented to participate in the evaluation study were included in analyses (N = 225). Individual growth models were used to test longitudinal effects among MA LAUNCH participants (children and caregivers) over three time points using screening tools. Results Analyses showed that LAUNCH children who scored in age-specific clinically significant ranges of social, emotional and behavioral problems at Time 1 scored in the normal range on average by Time 3. Caregivers’ stress and depressive symptoms also declined across the three time points. Results support hypotheses that the LAUNCH intervention improved social and emotional health for children and caregivers. Conclusions for Practice This study led to sustainability efforts, an expansion of the model to three additional communities across the state and develop- ment of an online toolkit for other communities interested in implementation. Keywords Early childhood · Mental health · LAUNCH · Primary care integration Significance evidence-informed approach to addressing disparities in mental health care but efficacious models of reaching Early childhood social-emotional difficulties are preva- young children are needed. This study provides evidence lent and can have negative impacts into adulthood. Inte- that enhancing mental health services in primary care grating mental health services within primary care is an through the integration of trained early childhood mental * Beth E. Molnar Deborah Allen b.molnar@northeastern.edu dallen@ph.lacounty.gov Kristin E. Lees Institute on Urban Health Research, Northeastern University, Klees@edc.org 360 Huntington Ave., Boston, MA 02115, USA Kate Roper Boston Public Health Commission, 1010 Massachusetts Ave, Kate.Roper@massmail.state.ma.us Boston, MA 02118, USA Natasha Byars Massachusetts Department of Public Health, 250 Washington nbyars@bphc.org St., Boston, MA 02118, USA Larisa Méndez-Peñate Los Angeles County Department of Public Health, 313 Larisa.Mendez-Penate@massmail.state.ma.us N. Figueroa St., Los Angeles, CA 90012, USA William McMullen Bouvé College of Health Sciences, Institute on Urban Health wmmcmulln@aol.com Research, Northeastern University, 360 Huntington Ave, M/S Jessica Wolfe 314 INV, Boston, MA 02115, USA je.wolfe@northeastern.edu Vol:.(1234567890) 1 3 Maternal and Child Health Journal (2018) 22:1502–1510 1503 health clinicians and family partners with lived experience Pediatrics n.d.). Inclusion of experience-based experts has can improve social-emotional health among young children been associated with improved vaccination rates (Justvig and their caregivers. et al. 2017), asthma control (Breysse et al. 2014), increased referrals to care coordination and improvements in patient support and education access (Matiz et al. 2014) among Introduction others. Family partners notably have a prominent role in evidence-based practices such as wrap-around care for chil- An estimated 9–14% of children aged 0–5 are affected by dren with serious emotional disorders (Bruns and Walker social and emotional difficulties (Brauner and Bowers 2006) 2008). They typically have lived experiences with relevant that limit ability to effectively engage in activities, fully ben- systems of care; they can draw on shared experiences plus efit from educational opportunities (Zbar et al. 2016), avoid shared cultural and linguistic backgrounds to engage and risky behaviors in adolescence (Thompson et al. 2011), and educate families, and have been shown to improve family prevent emotional instability in adulthood (Goodman et al. engagement, empowerment and trust in clinicians (Cournos 2011). Studies have identified multiple risk factors that affect and Goldfinger 2014). early childhood mental health (ECMH) including neighbor- This study presents results from the evaluation of Mas- hood disadvantage, witnessing violence, parental emotional sachusetts’ Project Linking Actions with Unmet Needs in distress, mental illness or substance problems, incarcerated Children’s Health (MA_LAUNCH), a preventive interven- relatives or other caregiver separation, harsh discipline, and tion housed in community pediatric medical home settings. homelessness (Bayer et al. 2011; Garner et al. 2012). Timely MA_LAUNCH addresses barriers to reducing ECMH risk interventions addressing barriers to social and emotional through integration of an ECMH clinician and family partner well-being are key to healthy development of children. team within pediatric primary care settings. This study pre- Literature on ECMH reflected in policy and practice sents evidence suggesting the efficacy of MA_LAUNCH in recommendations (Foy 2010; American Academy of Child improving social, emotional and behavioral developmental and Adolescent Psychiatry 2010), has identified burdens that progress of children and caregivers. insufficient mental health services place on medical provid- ers, especially in practices serving low income children (Foy 2010). Primary care providers report low confidence Methods in effectively screening and managing the mental health of young children, short visit time concerns, and hesitancy to Study Design refer to services with long wait times (Horwitz et al. 2007). One suggested solution is improving the integration of A longitudinal, repeated measures study design was used to mental health services into primary care (Ader et al. 2015; evaluate improvements in social, emotional, and behavio- Foy 2010; AACAP 2010). Integration can take a variety of ral development of children ages 0–8 years, and parenting- forms, including care coordination linked to mental health related stress and depressive symptoms of their caregivers, care, co-location of mental health and pediatric providers in for participants in MA_LAUNCH. The evaluation ran from same or nearby locations, or full integration of behavioral 2011 to 2015 at three treatment sites (N = 225 children). health specialists hired by primary care sites working itera- Children and caregivers were assessed at three time points: tively with pediatric care providers about cases (Tyler et al. baseline (Time 1), 6  months (Time 2), and 12  months 2017). Integration has been more thoroughly explored in (Time 3) after enrollment. All families who received MA_ adult mental health care but is gaining traction in pediatrics LAUNCH services were eligible to participate in the study; (Godoy et al. 2014; Spijkers et al. 2013; Briggs et al. 2012). records for those families who consented were transferred Early results are promising, suggesting improved identifi- to the evaluation team for analyses. Two hypotheses were cation of problems and improved screening results (Godoy tested: (1) children would demonstrate improvements in et al. 2014; Spijkers et al. 2013; Briggs et al. 2012). social, emotional and behavioral development as measured Needs of low income families often extend beyond chil- by screening tools administered three times; and (2) parent- dren’s mental health services. A new approach is to provide ing-related stress and depressive symptoms among caregiv- more family-centered case coordination with help from ers would decrease over time. “experience-based experts” (Gilkey et  al. 2011). Family partners, one type of experience-based experts, have gained Intervention Design popularity in the movement toward patient and family- centered care for underserved, culturally, and linguistically MA_LAUNCH was conducted in Boston, Massachusetts and diverse populations by community health care workers was designed to address needs of children at risk for social, (Volkmann and Castanares 2011; American Academy of emotional and behavioral problems by providing behavioral 1 3 1504 Maternal and Child Health Journal (2018) 22:1502–1510 health services and case coordination within pediatric medi- For example, a concern regarding a child’s behavior at child- cal homes. The model (Boston Public Health Commission care or preschool could stem from a variety of factors; after 2014) included the role of the “family partner,” as described careful clinical assessment and formulation, such a concern above and below, to work collaboratively with a clinician might be addressed through a combination of developmental who had masters-level training in mental health care for very psychoeducation and guidance for the caregiver, observa- young children. While the family partner role resembles that tions at the school and/or consultation with the teacher, and of other community health workers, it is distinguished in this follow-up as needed to provide the caregiver and/or teacher model by the requirement that family partners have lived with recommendations. In multiple cases, support included experience raising a child with a history of social, emotional helping the caregiver to better define and voice concerns or behavioral difficulties. The family partners were able to with the school or daycare while simultaneously supporting engage with families differently than the clinicians by draw - the relationship between caregiver and school/teacher. This ing on shared experiences, modeling effective strategies for support often, in turn, strengthened the relationship with parenting, and advocating on behalf of their children. Both and availability of the school/teacher toward the child and team members were employed by the health care site using became important to ameliorating the issue. funds from the grant and participated in ongoing trainings When adjustment was a concern, whether due to a new run jointly by the local and State public health depart- sibling, new schools, immigration, a separation or loss, ments on evidence-based early childhood development, or other cause, the team worked with the family, pairing mental health, and parenting interventions. At least one play-based, dyadic intervention with caregiver guidance. MA_LAUNCH team member at each site was multi-lingual Additionally, interventions often incorporated supporting in languages relevant to populations served. In addition to the caregiver’s reflective functioning related to the child’s trainings, MA_LAUNCH teams benefitted from biannual subjective experience, emotions, and responses, also known cross-site/cross-project learning collaboratives and monthly as parental mentalization. Parental mentalization is thought meetings with medical and behavioral health staff from each to not only support the child’s experience, regulation, and site and the MA_LAUNCH team (Oppenheim et al. 2016). social emotional development, but to also support the car- Clinical consultation, technical assistance and administrative egiver’s own regulation and subjective experience as a supervision was provided by the local public health team parent in the face of challenges (Sharp and Fonagy 2008). throughout to assist in integration of MA_LAUNCH ser- Another significant opportunity presenting in primary care vices into each center and in keeping fidelity to the model. is seeing caregiver(s) during the perinatal period. Perinatal Massachusetts pediatric practices receiving MassHealth depression and emotional dysregulation represent the most (Medicaid) reimbursement, including the three MA_ common complications in the perinatal period and have LAUNCH sites, are required to implement behavioral impacts on the caregiver, baby, and family system (Meltzer- screenings at each well-child visit. Based on screening Brody 2011). Often complicated by additional stressors such results, clinician judgement or family concerns, warm as income loss and housing, perinatal mental health con- handoffs were made to the MA_LAUNCH teams during the cerns were commonly referred to MA_LAUNCH teams by intervention period, who introduced families to the program primary care providers and supporting families during this and enrolled them if appropriate. Subsequent steps in the critical time became an area of focus. When child or fam- service delivery process were (1) completion of intake and ily challenges necessitated interventions beyond the scope informed consent processes, (2) administration of social and of MA_LAUNCH, the teams made referrals to in-house or mental health needs assessments; (3) collaborative develop- external referral sources, but typically continued to be part ment of a care plan based on child needs and family priori- of the family’s longer-term supportive pediatric care. ties; (4) initiation of case management and related referrals; Trainings built into MA_LAUNCH benefited whole and, as needed, (5) child mental health and/or parenting centers. For example, in one health center, 25 health interventions. center staff from diverse disciplines and roles—includ- Multiple child and family factors were explored in clini- ing primary care physicians, nurses, medical assistants, cal assessments—socioeconomic, relational, immigration, social workers, and interpreters—engaged in a two-day perinatal, traumatic, developmental, cultural, and more— MA_LAUNCH-funded training on supporting budding and became important aspects of the team’s formulation relationships between caregivers and their newborns, and guiding framework for care plans. The in-depth clinical and the caregiver’s own subjective sense of competency assessment included consultation with the primary care pro- in their new role. In addition to these patient- and pro- vider, objective screening tools, caregiver interviews gath- vider-specific activities, MA_LAUNCH teams developed ering family history, observations of the child and family activities within the medical home setting to engage and when possible, and play-based interaction when indicated; educate MA_LAUNCH families and overall pediatric cli- interventions were then tailored to the family’s unique needs. entele. Activities led by the family partners and clinicians 1 3 Maternal and Child Health Journal (2018) 22:1502–1510 1505 from MA_LAUNCH were family-centered and encom- Table 1 Demographic characteristics of MA_LAUNCH evaluation study participants passed both health promotion and prevention activities to engage whole families, including health center-wide Children N = 225 events such as family game nights, kindergarten registra- Age Mean (SD) tion workshops, caregiver support groups, playgroups, Age at intake—months 39.09 (24.07) and field trips. Age intake—years 3.26 (2.01) For the evaluation study, assessment, screening and service data were entered by site teams into databases cre- N (%) ated by the evaluators. After families gave consent, these Gender project records were transferred to the evaluators for anal-  Male 140 (62) yses. Four Institutional Review Boards (IRB) reviewed  Female 85 (38) and approved all procedures and protocols of the evalua- Race tion study: Northeastern University, MA Department of  Biracial 10 (4) Public Health, Boston University and one participating  Black 77 (34) site’s IRB.  Hispanic 119 (53)  Multiracial 4 (2) Participants  Other/unknown 5 (2)  White 9 (4) Three pediatric practices implemented MA_LAUNCH from  Asian 1 (0.4) 2010 to 2015; two were community health centers and one Primary caregiver N = 186 was a hospital-based pediatric clinic. All three served pri- Age Mean (SD) marily low-income residents of surrounding Boston neigh- borhoods, where the majority of patients are families of Age at intake- years 30.19 (7.84) color. All sites offered both pediatric and behavioral health N (%) care. Demographics of families are displayed in Table 1. Gender (2 missing)  Male 4 (2) Measures  Female 180 (97) Race (6 missing) Two parent-report screening tools were used to assess social  Biracial 1 (0.55) emotional and behavioral concerns of children, with improve-  Black/African American 70 (38) ment defined as participant movement from clinically con-  Multiracial 3 (2) cerning ranges into healthy ranges. The Ages & Stages Social  White 11 (6) and Emotional questionnaire (ASQ-SE) was used with chil-  Asian 2 (1) dren age 5 and younger and the Child Behavior Checklist  Hispanic 84 (46) (CBCL) with children 6–8 years. For children who began  Unknown 9 (5) the study at age 5 and ended at age 6, they were screened with the ASQ-SE first and the CBCL later. The ASQ-SE is a screening and monitoring tool to identify social emotional subscale of the full PHQ based on DSM-IV diagnostic cri- problems in children 0–6-years-old, with 19–30 items and teria, with demonstrated validity as a screening tool with good internal consistency (α = 0.82) and test–retest reliability diverse primary care patients, strong internal (α = 0.86–0.92) (0.94) (Squires et al. 2001). Clinical cutoff scores for each age and test–retest (0.83–0.84) reliability (Huang et al. 2006), group indicate follow-up and monitoring needs (Squires et al. and 88% sensitivity/specificity for depression diagnoses 2002). The CBCL contains 120 items assessing emotional or (Kroenke et al. 2010). Each item is scored from 0 (not at all) behavioral problems in the past 6 months, with scores above to 3 (nearly every day) and summed to obtain scores from 0 63 considered clinically concerning; strong reliability and to 27, with ≥10 representing clinically significant depressive validity has been demonstrated in many populations (Achen- symptoms. The PSI-SF is a 36-item abbreviated version of bach and Rescorla 2008; Nakamura et al. 2008). a 120-item scale used widely to measure overall parental Two tools were used to measure caregiver functioning: stress (Abidin 1995). Scores from 16 to 84th percentiles are the Patient Health Questionnaire-9 (PHQ-9) and the Parent- considered within the normal range while 85th percentile ing Stress Index-Short Form (3rd Edition) (PSI-SF). The indicates clinically significant stress (Reitman et al. 2002). PHQ-9 is a self-administered, nine-item depressive symptom 1 3 1506 Maternal and Child Health Journal (2018) 22:1502–1510 Statistical Analyses Results Screening was done at the MA_LAUNCH sites across three Social, Emotional and Behavioral Improvements time points. Individual growth modeling was used to analyze for Children ASQ-SE and CBCL scores for MA_LAUNCH children, and the PSI and PHQ-9 scores for their caregivers. Multilevel Results from individual growth models (Table 2) assessing models with restricted maximum likelihood estimations change in MA_LAUNCH children’s ASQ-SE scores over for mixed models were performed in SAS version 9.3. This time (Fig. 1) show significant declines in social, emotional procedure allows for the inclusion of all available data even and behavioral problems for children under age 5 (N = 188 at for those growth records that were incomplete (Wolfinger baseline); children who started above the clinical cutoff score and Chang 1995). Not all children and caregivers completed on average scored below (in the healthy range) by Time 3 the screening tools at all time points (see figure captions (Χ = 74.73, p < 0.0001). Children ages 1–5 experienced a for sample sizes); sensitivity analyses were conducted with reduction in ASQ-SE scores putting them on average below subjects who completed two and three time points for each the cutoff score by Time 3 (12-months post-enrollment), with measure to confirm that growth model results were not the greatest change in children ages 4–5 years at baseline as biased by missing data. Table 2 Individual growth model results Estimate (SE) Covariance Parameter ASQ-SE CBCL PSI PHQ-9 Individual 1032.62 (154.28)** 414.29(85.53)** 291.57 (44.76)** 12.97 (2.18)** Residual 798.31(78.33)** 200.66(32.23)** 225.98 (21.11)** 14.98(1.37)** Final models 2 2 2 Parameter ASQ-SE (Model X = 74.73 CBCL (Model X = 46.22 PSI (Model X = 46.92 PHQ-9 (Model p < 0.0001) p < 0.0001) p < 0.0001) X = 15.41 p < 0.0001) Intercept 42.90 (5.13)** 32.70 (2.57)** 65.71 (1.76)** 4.04 (0.33)** Time − 0.37 (4.51) − 2.40 (1.72) 0.17 (1.19) 0.022 (0.27) Age 10.80 (1.76)** Baseline risk 39.98 (4.86)** 34.64 (2.61)** 9.58 (0.61)** Time*age − 3.13 (1.47)* Time*baseline risk − 10.86 (3.12)** − 8.87 (1.84)** − 2.94 (0.50)** **p < 0.001; *p < 0.05 Fig. 1 Changes in ASQ-SE scores over time (Time 1 N = 188; Time 2 N = 127; Time 3 N = 73) by age. ASQ-SE clinical cutoff scores vary by age and are indicated by the red line. Children aged 1–5 scored on average, above the cutoff scores at Time 1 and below the cutoff score at Time 3 40 <1 1234 5 Baseline Age in Years Time 1 Time 2 Time 3 Age-specific Clinical Cutoff 1 3 Average ASQ-SE Score Maternal and Child Health Journal (2018) 22:1502–1510 1507 indicated by a statistically significant (p = 0.03) age-time inter- Time 1 to Time 3, bringing average scores for this group action estimate (− 3.13). within the healthy range. Caregivers whose scores were For children ages 6–8 years (N = 75 at baseline), individual within the healthy range did not change significantly Time growth models were used to compare changes in CBCL total 1–3. The difference between the groups was statistically sig- problems scores over the 1 year study period for children who nificant as indicated by the time-baseline risk interaction scored above the clinical cutoff at Time 1 compared to chil- estimate of − 8.87 (p < 0.001). dren who scored in the healthy range at Time 1. While total For depressive symptoms as measured by the PHQ-9 problem scores of children who scored below the cutoff at (N = 181 caregivers at baseline), individual growth mod- Time 1 remained on average below the cuto, ff the scores of els compared MA_LAUNCH caregivers with clinical-level children who scored above it at Time 1 dropped by over 25 depressive symptoms at Time 1 (scores > 10) with those points, a 37% decline (Fig. 2). The difference between the caregivers with scores in the non-clinical range. The fitted two groups was statistically significant as indicated by the growth model (Table 2; Fig. 4) estimating change in depres- time-baseline risk interaction estimate of − 10.86 (p < 0. 001). sive symptoms over time indicates that caregivers with high depressive symptoms at Time 1 had a decline of almost 6 Stress and Depressive Symptoms Among Caregivers points with scores on average falling within the non-clinical of MA_LAUNCH Children range by Time 3. Differences between the two groups were statistically significant as indicated by the time-baseline risk Results from individual growth models of PSI-SF scores interaction estimate of − 2.94 (p < 0.001). of MA_LAUNCH caregivers (N = 167 at baseline) com- pared caregivers with high parental stress levels at Time 1 (scores > 85) with caregivers within the healthy range (scores = 16–85) (Table 2; Fig. 3). Caregivers with scores above the cutoff showed an average decline of 17.38 points Fig. 2 Comparison of children 80 who scored above versus below the clinical cutoff score at Time 1 (Time 1 N = 75; Time 2 N = 45; Time 3 N = 28). On average, children who scored below the cutoff at baseline retained their healthy range status whereas children who scored above the cutoff at base- line, dropped below the cutoff by Time 3 Time 1Time 2Time 3 Measurement Occasion Clinical Cutoff Healthy range at baseline Clinical range at baseline Fig. 3 Comparison of caregiv- 110 ers who scored above versus below the clinical cutoff score at Time 1 (Time 1 N = 167; Time 2 N = 131; Time 3 N = 97). On average, caregivers who scored below the cutoff at baseline retained their healthy-range status whereas caregivers who 75 scored above the cutoff at base- line, dropped below the cutoff by Time 3 Time 1Time 2Time 3 Measurement Occasion Clinical Cutoff Healthy range at baseline Clinical range at baseline 1 3 PSI Score CBCL Score 1508 Maternal and Child Health Journal (2018) 22:1502–1510 Fig. 4 Comparison of caregiv- 16 ers who scored above versus below the clinical cutoff score at Time 1 (Time 1 N = 181; Time 2 N = 142; Time 3 N = 90). On average, caregivers who scored 10 below the cutoff at baseline retained their healthy-range status whereas caregivers who scored above the cutoff at base- line, dropped below the cutoff 4 by Time 3 Time 1Time 2Time 3 Measurement Occasion Clinical Cutoff Healthy range at Baseline Clinical range at Baseline sharing of resources and experiences. Situating the interven- Discussion tion within the medical home allowed families and health care professionals to maximize time together and make use Children and caregivers served by MA_LAUNCH expe- of family partners’ shared experiences and cultural back- rienced significant improvements on common screening grounds. MA_LAUNCH teams promoted the medical home tools for social, emotional and behavioral health, especially model’s ideal of family-centered approaches and confronted for those participants with scores in clinically concerning barriers to providers’ ability to address ECMH by providing ranges initially. Children had on average clinically signifi- immediate availability, support and center-wide activities for cant screening results for all but the youngest age group providers, families and to their clinics overall. upon study entry, and 1 year later they were (on average) in As with all studies, findings should be interpreted consid- healthy ranges. Similarly, caregivers with clinically signifi- ering limitations. The first is that the ASQ-SE is designed to cant parenting-related stress and depressive symptom scores be used as a screening and monitoring tool rather than for were on average below clinical cutoffs after 1 year. Children diagnostic outcome measurement; this tool was used in this and caregivers whose screening results did not indicate clini- study to monitor children’s changes in problems over time. cal risk at the beginning of the study retained healthy status. Results should not be interpreted as diagnostic. Secondly, Evaluations of models that promote universal behavioral for families in crisis, screenings were sometimes delayed health screening and embed mental health clinicians within until after the crisis subsided and thus after the family began primary care sites reveal mixed results (Godoy et al. 2014; receiving services. This delay may have resulted in underes- Spijkers et al. 2013; Briggs et al. 2012). For example, one timations of problems. Some subsequent assessments were model embedding an infant/toddler specialist in primary also completed later than ideal due to scheduling dic ffi ulties. care settings reported improved ASQ-SE scores (Briggs Some families were lost to follow up or did not complete et al. 2012) whereas one offering brief counseling found no assessments at all three time points. Reasons for loss to fol- improvements over usual care (Spijkers et al. 2013). Results low up included that the child aged out of MA_LAUNCH from Project LAUNCH in Rhode Island, an initiative with- services, the family moved outside of the service area, or out the family partner role, resulted in improved identifica- staff were unable to maintain contact with the family. How - tion of young children at risk but found that subsequently ever, sensitivity analyses showed the same results whether younger children were less likely than older children to families completed two or three assessments. Sample size utilize subsequent mental health consultation (Godoy et al. limited the researchers’ ability to include large numbers of 2014). These results highlight needs for innovative family covariates. This limitation was not expected to influence engagement approaches, particularly families with very results due to the homogeneity of the sample; sensitivity young children in stressful circumstances. analyses revealed no effect of excluded demographic vari- The MA_LAUNCH intervention addresses gaps in ables. While a comparison site was recruited, budgetary ECMH service delivery for families by integrating trained and time constraints allowed for data collection only among family partners and mental health clinicians working as children, and for only one follow-up screening assessment. teams in primary care settings. MA_LAUNCH teams Although the same criteria of exposure to a list of risk worked across sites in learning collaboratives, attended factors was utilized, very few children from the compari- monthly meetings together, and received supervision from son group had ASQ-SE scores above the clinical cutoff at the state and local public health departments, facilitating 1 3 PHQ-9 Score Maternal and Child Health Journal (2018) 22:1502–1510 1509 (IUHR), at Northeastern University, Boston, MA. The authors of this baseline. The lack of differences and limited data precluded paper would like to acknowledge and thank the participating com- meaningful comparisons of change between groups. Lack of munity health center pediatric practices, the MA_LAUNCH provider meaningful comparison site data is a limitation of this study. teams, the Massachusetts early childhood mental health network and Given these evaluation results, future research examin- parent council, and all the families who participated in this study. ing the effectiveness of integrating behavioral health ser - Open Access This article is distributed under the terms of the Crea- vices into primary care is warranted, with attention on the tive Commons Attribution 4.0 International License (http://creat iveco impact of specific aspects of the MA_LAUNCH model. mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- Although not included in the data presented here, process tion, and reproduction in any medium, provided you give appropriate evaluation results showed the intervention to be widely credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. accepted by caregivers who reported nearly unanimous satisfaction when interviewed by the evaluation team as to whether the MA_LAUNCH team helped them under- stand their children’s emotions and behaviors and whether References they helped them help their children express feelings in more positive ways. Data collected from non-LAUNCH Abidin, R. R. (1995). Parenting stress index, third edition: professional providers within the clinical sites and the MA_LAUNCH manual. Odessa, FL: Psychological Assessment Resources. teams themselves also reported many benefits includ- Achenbach, T., & Rescorla, L. (2008). Manual for the ASEBA School- ing care coordination successes, integration of services, Age Forms & Profiles: an integrated system of multi-informant assessment. Burlington, VT: University of Vermont. and increased access to school registration and school Ader, J., Stille, C. J., Keller, D., Miller, B. F., Barr, M. S., & Perrin, J. and child care services. Components to explore further M. (2015). The medical home and integrated behavioral health: include whether dose–response effects exist for families Advancing the policy agenda. Pediatrics, 135(5), 909–917. who receive services more frequently or for longer periods American Academy of Child and Adolescent Psychiatry. (2010). A guide to building collaborative mental health care partnerships of time; impacts of the family partner role versus the clini- in pediatric primary care. Retrieved March 15, 2017, from https cian role; the influence of full integration of the model into ://www .aacap .or g/App_Theme s/AA C AP /docs/clini cal_pr act primary care on service acceptance and adherence; and ice_center/guide _t o_building_colla bor ativ e_mental_healt h_car e_ physicians’ perceived abilities to incorporate behavioral partn ershi ps.pdf. American Academy of Pediatrics. (n.d.). Salud para todos improving health treatment into their practice. health through medical homes. Retrieved from https ://medic alhom Project LAUNCH in Massachusetts was delivered in einfo .aap.org/tools -resou rces/Docum ents/RI%20Fac tShee t.pdf. community-based pediatric primary care practices in low Bayer, J. K., Ukoumunne, O. C., Lucas, N., Wake, M., Scalzo, K., & income urban settings. Children served at these centers are Nicholson, J. M. (2011). Risk factors for childhood mental health symptoms: National longitudinal study of Australian children. primarily recipients of MassHealth, the state Medicaid pro- Pediatrics, 128(4), e865–e879. gram. Families in this study experienced community- and Boston Public Health Commission. (2014). Early childhood mental family-level risk factors, and health and/or education dis- health toolkit: Integrating mental health services into the pediat- parities. The results of the evaluation of the MA_LAUNCH ric medical home. Retrieved March 15, 2017, from http://www. ecmhm atter s.org/ForPr ofess ional s/Pages /Medic alHom e.aspx. project show that both children and caregivers experienced Brauner, B. C., & Bowers, S. C. (2006). Estimating the prevalence improvements in social, emotional and behavioral difficul- of early childhood serious emotional/behavioral disorders: ties. Given these results, it is likely MA_LAUNCH helped Challenges and recommendations. Public Health Reports, 121, to overcome barriers to service delivery and attainment of 303–310. Breysse, J., Dixon, S., Gregory, J., Philby, M., Jacobs, D. E., & Krieger, social and emotional well-being. Interventions are needed J. (2014). Effect of weatherization combined with community that put children on a healthy developmental pathway. The health worker in-home education on asthma control. American results of this study can be used to inform the development Journal of Public Health, 104, e57–e64. of similar programs and policies to address disparities in Briggs, R. D., Stettler, E., Silver, E., Schrag, R., Nayak, M., Chinitz, S., & Racine, A. D. (2012). Social-emotional screening for infants access to behavioral health care and well-being for very and toddlers in primary care. Pediatrics, 129(2), E377–E384. young children. Bruns, E. J., & Walker, J. S. (Eds.) (2008). The resource guide to wrap- around. Portland, OR: National Wraparound Initiative, Research Acknowledgements Massachusetts Linking Actions for Unmet Needs and Training Center for Family Support and Children’s Mental in Children’s Health (MA_LAUNCH) is an initiative of the Massa- Health. Retrieved March 15, 2017, from https ://n wi.pdx.edu/NWI- chusetts Department of Public Health, funded by the U.S., Substance book/Chapt ers/COMPL ETE-RG-BOOK.pdf. Abuse and Mental Health Services Administration (SAMHSA), (Grant Cournos, F., & Goldfinger, S. M. (2014). Family partners improve early No. SM 059334). The original intervention was carried out in col- childhood mental health services. Psychiatric Services, 65(11), laboration with the Boston Public Health Commission. The evalua- tion study was carried out by the Institute on Urban Health Research 1 3 1510 Maternal and Child Health Journal (2018) 22:1502–1510 Foy, J. M. (2010). Enhancing pediatric mental health care: Report Oppenheim, J., Stewart, W., Zoubak, E., Donato, I., Huang, L., & from the American Academy of Pediatrics Task Force on Mental Hudock, W. (2016). Launching forward: The integration of Health. Introduction. Pediatrics, 3, S69–S74. behavioral health in primary care as a key strategy for promot- Garner, A. S., Shonkoff, J. P., Siegel, B. S., Dobbins, M. I., Earls, M. F., ing young child wellness. American Journal of Orthopsychiatry, Garner, A. S., … Wood, D. L. (2012). Early childhood adversity, 86(2), 124–131. toxic stress, and the role of the pediatrician: Translating develop- Reitman, D., Currier, R. O., & Stickle, T. (2002). A critical evaluation mental science into lifelong health. Pediatrics, 129(1), e224–e231. of the parenting stress index-short form (PSI-SF) in a Head Start https ://doi.org/10.1542/peds.2011-2662. population. Journal of Clinical Child and Adolescent Psychology, Gilkey, M., Garcia, C. C., & Rush, C. (2011). Professionalization and 31(3), 384–392. the experience-based expert: Strengthening partnerships between Sharp, C., & Fonagy, P. (2008). The parent’s capacity to treat the child health educators and community health workers. Health Promo- as a psychological agent: Constructs, measures and implications tion Practice, 12(2), 178–182. for developmental psychopathology. Social Development, 17(3), Godoy, L., Carter, A. S., Silver, R. B., Dickstein, S., & Seifer, R. 737–754. (2014). Mental health screening and consultation in primary care: Spijkers, W., Jansen, D. E., & Reijneveld, S. A. (2013). Effectiveness of The role of child age and parental concerns. Journal of Devevel- primary care Triple P on child psychosocial problems in preven- opmental and Behavior Pediatrics, 35(5), 334–343.9. tive child healthcare: A randomized controlled trial. BMC Medi- Goodman, A., Joyce, R., & Smith, J. P. (2011). The long shadow cast cine. https ://doi.org/10.1186/1741-7015-11-240. by childhood physical and mental problems on adult life. Proceed- Squires, J., Bricker, D., Heo, K., & Twombly, E. (2001). Identification ings of the National Academy Sciences of the United States of of social-emotional problems in young children using a parent- America, 108(15), 6032–6037. completed screening measure. Early Childhood Research Quar- Horwitz, S. M., Kelleher, K. J., Stein, R. E., Storfer-Isser, A., Young- terly, 16, 405–419. strom, E. A., Park, E. R.,… Hoagwood, K. E. (2007). Barriers Squires, J., Bricker, D., & Twombly, E. (2002). The ASQ:SE user’s to the identification and management of psychosocial issues in guide: For the Ages & Stages Questionnaires: Social-emotional. children and maternal depression. Pediatrics, 119(1), e208–e218. Baltimore: Paul H Brookes. Huang, F., Chung, H., Kroenke, K., Delucchi, K., & Spitzer, R. (2006). Thompson, R., Tabone, J. K., Litrownik, A. J., Briggs, E. C., Hussey, Using the Patient Health Questionnaire-9 to measure depression J. M., English, D. J., & Howard, D. (2011). Early adolescent risk among racially and ethnically diverse primary care patients. Jour- behavior outcomes of childhood externalizing behavioral trajecto- nal of General Internal Medicine, 21(6), 547–552. ries. The Journal of Early Adolescence, 31(2), 234–257. Justvig, S. P., Li, J., Caravella, G., Chen, M., Wang, H., Benz Scott, L. Tyler, E. T., Hulkower, R. L., & Kaminski, J. W. (2017). Milbank A., & Pati, S. (2017). Improving adherence to care recommenda- memorial fund. Behavioral health integration in pediatric pri- tions using a community health worker (CHW) intervention with mary care: Considerations and opportunities for policymakers, the pediatric medical home. Journal of Community Health, 42, planners, and providers. Retrieved March 15, 2017, from https :// 444–452.www.milba nk.org/publi catio ns/behav ioral -healt h-integ ratio n-in- Kroenke, K., Spitzer, R. L., Williams, J. B. W., & Löwe, B. (2010). pediatr ic-primar y-care-consider ations-and-oppor tunit ies-f or-polic The patient health questionnaire somatic, anxiety, and depressive ymake rs-plann ers-and-provi ders/. symptom scales: A systematic review. General Hospital Psychia- Volkmann, K., & Castanares, T. (2011). Clinical community health try, 32(4), 345–359. workers: linchpin of the medical home. The Journal of Ambula- Matiz, L. A., Peretz, P. J., Jacotin, P. G., Cruz, C., Ramirez-Diaz, E., & tory Care Management, 34(3), 221–233. Nieto, A., R (2014). The impact of integrating community health Wolfinger, R., & Chang, M. (1995) Comparing the SAS® GLM workers into the patient-centered medical home. Journal of Pri- and MIXED procedures for repeated measurements analysis. mary Care and Community Health, 5(4), 271–274. Retrieved March 15, 2017, from https://suppo r t.sas.com/rnd/app/ Meltzer-Brody, S. (2011). New insights into perinatal depression: stat/paper s/abstr acts/mixed glm.html. Pathogenesis and treatment during pregnancy and postpartum. Zbar, A., Surkan, P. J., Fombonne, E., & Melchior, M. E. (2016). Emo- Dialogues in Clinical Neuroscience, 13(1), 89–100. tional and behavioral difficulties and adult educational attainment: Nakamura, B. J., Ebesutani, C., Bernstein, A., & Chorpita, B. F. (2008). An 18-year follow-up of the TEMPO study. European Child and A psychometric analysis of the child behavior checklist DSM- Adolescent Psychiatry, 25(10), 1141–1143. oriented scales. Journal of Psychopathology and Behavioral Assessment, 31(3), 178–189. 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Maternal and Child Health Journal Springer Journals

Enhancing Early Childhood Mental Health Primary Care Services: Evaluation of MA Project LAUNCH

Loading next page...
 
/lp/springer-journals/enhancing-early-childhood-mental-health-primary-care-services-w0cwWk049w

References (44)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Public Health; Sociology, general; Population Economics; Pediatrics; Gynecology; Maternal and Child Health
ISSN
1092-7875
eISSN
1573-6628
DOI
10.1007/s10995-018-2548-4
Publisher site
See Article on Publisher Site

Abstract

Objectives The purpose of this study was to evaluate the efficacy of an innovative early childhood mental health intervention, Massachusetts Project LAUNCH. Early childhood mental health clinicians and family partners (paraprofessionals with lived experience) were embedded within community pediatric medical homes. Methods A longitudinal study design was used to test the hypotheses that (1) children who received services would experience decreased social, emotional and behavio- ral problems over time and (2) caregivers’ stress and depressive symptoms would decrease over time. Families who were enrolled in services and who consented to participate in the evaluation study were included in analyses (N = 225). Individual growth models were used to test longitudinal effects among MA LAUNCH participants (children and caregivers) over three time points using screening tools. Results Analyses showed that LAUNCH children who scored in age-specific clinically significant ranges of social, emotional and behavioral problems at Time 1 scored in the normal range on average by Time 3. Caregivers’ stress and depressive symptoms also declined across the three time points. Results support hypotheses that the LAUNCH intervention improved social and emotional health for children and caregivers. Conclusions for Practice This study led to sustainability efforts, an expansion of the model to three additional communities across the state and develop- ment of an online toolkit for other communities interested in implementation. Keywords Early childhood · Mental health · LAUNCH · Primary care integration Significance evidence-informed approach to addressing disparities in mental health care but efficacious models of reaching Early childhood social-emotional difficulties are preva- young children are needed. This study provides evidence lent and can have negative impacts into adulthood. Inte- that enhancing mental health services in primary care grating mental health services within primary care is an through the integration of trained early childhood mental * Beth E. Molnar Deborah Allen b.molnar@northeastern.edu dallen@ph.lacounty.gov Kristin E. Lees Institute on Urban Health Research, Northeastern University, Klees@edc.org 360 Huntington Ave., Boston, MA 02115, USA Kate Roper Boston Public Health Commission, 1010 Massachusetts Ave, Kate.Roper@massmail.state.ma.us Boston, MA 02118, USA Natasha Byars Massachusetts Department of Public Health, 250 Washington nbyars@bphc.org St., Boston, MA 02118, USA Larisa Méndez-Peñate Los Angeles County Department of Public Health, 313 Larisa.Mendez-Penate@massmail.state.ma.us N. Figueroa St., Los Angeles, CA 90012, USA William McMullen Bouvé College of Health Sciences, Institute on Urban Health wmmcmulln@aol.com Research, Northeastern University, 360 Huntington Ave, M/S Jessica Wolfe 314 INV, Boston, MA 02115, USA je.wolfe@northeastern.edu Vol:.(1234567890) 1 3 Maternal and Child Health Journal (2018) 22:1502–1510 1503 health clinicians and family partners with lived experience Pediatrics n.d.). Inclusion of experience-based experts has can improve social-emotional health among young children been associated with improved vaccination rates (Justvig and their caregivers. et al. 2017), asthma control (Breysse et al. 2014), increased referrals to care coordination and improvements in patient support and education access (Matiz et al. 2014) among Introduction others. Family partners notably have a prominent role in evidence-based practices such as wrap-around care for chil- An estimated 9–14% of children aged 0–5 are affected by dren with serious emotional disorders (Bruns and Walker social and emotional difficulties (Brauner and Bowers 2006) 2008). They typically have lived experiences with relevant that limit ability to effectively engage in activities, fully ben- systems of care; they can draw on shared experiences plus efit from educational opportunities (Zbar et al. 2016), avoid shared cultural and linguistic backgrounds to engage and risky behaviors in adolescence (Thompson et al. 2011), and educate families, and have been shown to improve family prevent emotional instability in adulthood (Goodman et al. engagement, empowerment and trust in clinicians (Cournos 2011). Studies have identified multiple risk factors that affect and Goldfinger 2014). early childhood mental health (ECMH) including neighbor- This study presents results from the evaluation of Mas- hood disadvantage, witnessing violence, parental emotional sachusetts’ Project Linking Actions with Unmet Needs in distress, mental illness or substance problems, incarcerated Children’s Health (MA_LAUNCH), a preventive interven- relatives or other caregiver separation, harsh discipline, and tion housed in community pediatric medical home settings. homelessness (Bayer et al. 2011; Garner et al. 2012). Timely MA_LAUNCH addresses barriers to reducing ECMH risk interventions addressing barriers to social and emotional through integration of an ECMH clinician and family partner well-being are key to healthy development of children. team within pediatric primary care settings. This study pre- Literature on ECMH reflected in policy and practice sents evidence suggesting the efficacy of MA_LAUNCH in recommendations (Foy 2010; American Academy of Child improving social, emotional and behavioral developmental and Adolescent Psychiatry 2010), has identified burdens that progress of children and caregivers. insufficient mental health services place on medical provid- ers, especially in practices serving low income children (Foy 2010). Primary care providers report low confidence Methods in effectively screening and managing the mental health of young children, short visit time concerns, and hesitancy to Study Design refer to services with long wait times (Horwitz et al. 2007). One suggested solution is improving the integration of A longitudinal, repeated measures study design was used to mental health services into primary care (Ader et al. 2015; evaluate improvements in social, emotional, and behavio- Foy 2010; AACAP 2010). Integration can take a variety of ral development of children ages 0–8 years, and parenting- forms, including care coordination linked to mental health related stress and depressive symptoms of their caregivers, care, co-location of mental health and pediatric providers in for participants in MA_LAUNCH. The evaluation ran from same or nearby locations, or full integration of behavioral 2011 to 2015 at three treatment sites (N = 225 children). health specialists hired by primary care sites working itera- Children and caregivers were assessed at three time points: tively with pediatric care providers about cases (Tyler et al. baseline (Time 1), 6  months (Time 2), and 12  months 2017). Integration has been more thoroughly explored in (Time 3) after enrollment. All families who received MA_ adult mental health care but is gaining traction in pediatrics LAUNCH services were eligible to participate in the study; (Godoy et al. 2014; Spijkers et al. 2013; Briggs et al. 2012). records for those families who consented were transferred Early results are promising, suggesting improved identifi- to the evaluation team for analyses. Two hypotheses were cation of problems and improved screening results (Godoy tested: (1) children would demonstrate improvements in et al. 2014; Spijkers et al. 2013; Briggs et al. 2012). social, emotional and behavioral development as measured Needs of low income families often extend beyond chil- by screening tools administered three times; and (2) parent- dren’s mental health services. A new approach is to provide ing-related stress and depressive symptoms among caregiv- more family-centered case coordination with help from ers would decrease over time. “experience-based experts” (Gilkey et  al. 2011). Family partners, one type of experience-based experts, have gained Intervention Design popularity in the movement toward patient and family- centered care for underserved, culturally, and linguistically MA_LAUNCH was conducted in Boston, Massachusetts and diverse populations by community health care workers was designed to address needs of children at risk for social, (Volkmann and Castanares 2011; American Academy of emotional and behavioral problems by providing behavioral 1 3 1504 Maternal and Child Health Journal (2018) 22:1502–1510 health services and case coordination within pediatric medi- For example, a concern regarding a child’s behavior at child- cal homes. The model (Boston Public Health Commission care or preschool could stem from a variety of factors; after 2014) included the role of the “family partner,” as described careful clinical assessment and formulation, such a concern above and below, to work collaboratively with a clinician might be addressed through a combination of developmental who had masters-level training in mental health care for very psychoeducation and guidance for the caregiver, observa- young children. While the family partner role resembles that tions at the school and/or consultation with the teacher, and of other community health workers, it is distinguished in this follow-up as needed to provide the caregiver and/or teacher model by the requirement that family partners have lived with recommendations. In multiple cases, support included experience raising a child with a history of social, emotional helping the caregiver to better define and voice concerns or behavioral difficulties. The family partners were able to with the school or daycare while simultaneously supporting engage with families differently than the clinicians by draw - the relationship between caregiver and school/teacher. This ing on shared experiences, modeling effective strategies for support often, in turn, strengthened the relationship with parenting, and advocating on behalf of their children. Both and availability of the school/teacher toward the child and team members were employed by the health care site using became important to ameliorating the issue. funds from the grant and participated in ongoing trainings When adjustment was a concern, whether due to a new run jointly by the local and State public health depart- sibling, new schools, immigration, a separation or loss, ments on evidence-based early childhood development, or other cause, the team worked with the family, pairing mental health, and parenting interventions. At least one play-based, dyadic intervention with caregiver guidance. MA_LAUNCH team member at each site was multi-lingual Additionally, interventions often incorporated supporting in languages relevant to populations served. In addition to the caregiver’s reflective functioning related to the child’s trainings, MA_LAUNCH teams benefitted from biannual subjective experience, emotions, and responses, also known cross-site/cross-project learning collaboratives and monthly as parental mentalization. Parental mentalization is thought meetings with medical and behavioral health staff from each to not only support the child’s experience, regulation, and site and the MA_LAUNCH team (Oppenheim et al. 2016). social emotional development, but to also support the car- Clinical consultation, technical assistance and administrative egiver’s own regulation and subjective experience as a supervision was provided by the local public health team parent in the face of challenges (Sharp and Fonagy 2008). throughout to assist in integration of MA_LAUNCH ser- Another significant opportunity presenting in primary care vices into each center and in keeping fidelity to the model. is seeing caregiver(s) during the perinatal period. Perinatal Massachusetts pediatric practices receiving MassHealth depression and emotional dysregulation represent the most (Medicaid) reimbursement, including the three MA_ common complications in the perinatal period and have LAUNCH sites, are required to implement behavioral impacts on the caregiver, baby, and family system (Meltzer- screenings at each well-child visit. Based on screening Brody 2011). Often complicated by additional stressors such results, clinician judgement or family concerns, warm as income loss and housing, perinatal mental health con- handoffs were made to the MA_LAUNCH teams during the cerns were commonly referred to MA_LAUNCH teams by intervention period, who introduced families to the program primary care providers and supporting families during this and enrolled them if appropriate. Subsequent steps in the critical time became an area of focus. When child or fam- service delivery process were (1) completion of intake and ily challenges necessitated interventions beyond the scope informed consent processes, (2) administration of social and of MA_LAUNCH, the teams made referrals to in-house or mental health needs assessments; (3) collaborative develop- external referral sources, but typically continued to be part ment of a care plan based on child needs and family priori- of the family’s longer-term supportive pediatric care. ties; (4) initiation of case management and related referrals; Trainings built into MA_LAUNCH benefited whole and, as needed, (5) child mental health and/or parenting centers. For example, in one health center, 25 health interventions. center staff from diverse disciplines and roles—includ- Multiple child and family factors were explored in clini- ing primary care physicians, nurses, medical assistants, cal assessments—socioeconomic, relational, immigration, social workers, and interpreters—engaged in a two-day perinatal, traumatic, developmental, cultural, and more— MA_LAUNCH-funded training on supporting budding and became important aspects of the team’s formulation relationships between caregivers and their newborns, and guiding framework for care plans. The in-depth clinical and the caregiver’s own subjective sense of competency assessment included consultation with the primary care pro- in their new role. In addition to these patient- and pro- vider, objective screening tools, caregiver interviews gath- vider-specific activities, MA_LAUNCH teams developed ering family history, observations of the child and family activities within the medical home setting to engage and when possible, and play-based interaction when indicated; educate MA_LAUNCH families and overall pediatric cli- interventions were then tailored to the family’s unique needs. entele. Activities led by the family partners and clinicians 1 3 Maternal and Child Health Journal (2018) 22:1502–1510 1505 from MA_LAUNCH were family-centered and encom- Table 1 Demographic characteristics of MA_LAUNCH evaluation study participants passed both health promotion and prevention activities to engage whole families, including health center-wide Children N = 225 events such as family game nights, kindergarten registra- Age Mean (SD) tion workshops, caregiver support groups, playgroups, Age at intake—months 39.09 (24.07) and field trips. Age intake—years 3.26 (2.01) For the evaluation study, assessment, screening and service data were entered by site teams into databases cre- N (%) ated by the evaluators. After families gave consent, these Gender project records were transferred to the evaluators for anal-  Male 140 (62) yses. Four Institutional Review Boards (IRB) reviewed  Female 85 (38) and approved all procedures and protocols of the evalua- Race tion study: Northeastern University, MA Department of  Biracial 10 (4) Public Health, Boston University and one participating  Black 77 (34) site’s IRB.  Hispanic 119 (53)  Multiracial 4 (2) Participants  Other/unknown 5 (2)  White 9 (4) Three pediatric practices implemented MA_LAUNCH from  Asian 1 (0.4) 2010 to 2015; two were community health centers and one Primary caregiver N = 186 was a hospital-based pediatric clinic. All three served pri- Age Mean (SD) marily low-income residents of surrounding Boston neigh- borhoods, where the majority of patients are families of Age at intake- years 30.19 (7.84) color. All sites offered both pediatric and behavioral health N (%) care. Demographics of families are displayed in Table 1. Gender (2 missing)  Male 4 (2) Measures  Female 180 (97) Race (6 missing) Two parent-report screening tools were used to assess social  Biracial 1 (0.55) emotional and behavioral concerns of children, with improve-  Black/African American 70 (38) ment defined as participant movement from clinically con-  Multiracial 3 (2) cerning ranges into healthy ranges. The Ages & Stages Social  White 11 (6) and Emotional questionnaire (ASQ-SE) was used with chil-  Asian 2 (1) dren age 5 and younger and the Child Behavior Checklist  Hispanic 84 (46) (CBCL) with children 6–8 years. For children who began  Unknown 9 (5) the study at age 5 and ended at age 6, they were screened with the ASQ-SE first and the CBCL later. The ASQ-SE is a screening and monitoring tool to identify social emotional subscale of the full PHQ based on DSM-IV diagnostic cri- problems in children 0–6-years-old, with 19–30 items and teria, with demonstrated validity as a screening tool with good internal consistency (α = 0.82) and test–retest reliability diverse primary care patients, strong internal (α = 0.86–0.92) (0.94) (Squires et al. 2001). Clinical cutoff scores for each age and test–retest (0.83–0.84) reliability (Huang et al. 2006), group indicate follow-up and monitoring needs (Squires et al. and 88% sensitivity/specificity for depression diagnoses 2002). The CBCL contains 120 items assessing emotional or (Kroenke et al. 2010). Each item is scored from 0 (not at all) behavioral problems in the past 6 months, with scores above to 3 (nearly every day) and summed to obtain scores from 0 63 considered clinically concerning; strong reliability and to 27, with ≥10 representing clinically significant depressive validity has been demonstrated in many populations (Achen- symptoms. The PSI-SF is a 36-item abbreviated version of bach and Rescorla 2008; Nakamura et al. 2008). a 120-item scale used widely to measure overall parental Two tools were used to measure caregiver functioning: stress (Abidin 1995). Scores from 16 to 84th percentiles are the Patient Health Questionnaire-9 (PHQ-9) and the Parent- considered within the normal range while 85th percentile ing Stress Index-Short Form (3rd Edition) (PSI-SF). The indicates clinically significant stress (Reitman et al. 2002). PHQ-9 is a self-administered, nine-item depressive symptom 1 3 1506 Maternal and Child Health Journal (2018) 22:1502–1510 Statistical Analyses Results Screening was done at the MA_LAUNCH sites across three Social, Emotional and Behavioral Improvements time points. Individual growth modeling was used to analyze for Children ASQ-SE and CBCL scores for MA_LAUNCH children, and the PSI and PHQ-9 scores for their caregivers. Multilevel Results from individual growth models (Table 2) assessing models with restricted maximum likelihood estimations change in MA_LAUNCH children’s ASQ-SE scores over for mixed models were performed in SAS version 9.3. This time (Fig. 1) show significant declines in social, emotional procedure allows for the inclusion of all available data even and behavioral problems for children under age 5 (N = 188 at for those growth records that were incomplete (Wolfinger baseline); children who started above the clinical cutoff score and Chang 1995). Not all children and caregivers completed on average scored below (in the healthy range) by Time 3 the screening tools at all time points (see figure captions (Χ = 74.73, p < 0.0001). Children ages 1–5 experienced a for sample sizes); sensitivity analyses were conducted with reduction in ASQ-SE scores putting them on average below subjects who completed two and three time points for each the cutoff score by Time 3 (12-months post-enrollment), with measure to confirm that growth model results were not the greatest change in children ages 4–5 years at baseline as biased by missing data. Table 2 Individual growth model results Estimate (SE) Covariance Parameter ASQ-SE CBCL PSI PHQ-9 Individual 1032.62 (154.28)** 414.29(85.53)** 291.57 (44.76)** 12.97 (2.18)** Residual 798.31(78.33)** 200.66(32.23)** 225.98 (21.11)** 14.98(1.37)** Final models 2 2 2 Parameter ASQ-SE (Model X = 74.73 CBCL (Model X = 46.22 PSI (Model X = 46.92 PHQ-9 (Model p < 0.0001) p < 0.0001) p < 0.0001) X = 15.41 p < 0.0001) Intercept 42.90 (5.13)** 32.70 (2.57)** 65.71 (1.76)** 4.04 (0.33)** Time − 0.37 (4.51) − 2.40 (1.72) 0.17 (1.19) 0.022 (0.27) Age 10.80 (1.76)** Baseline risk 39.98 (4.86)** 34.64 (2.61)** 9.58 (0.61)** Time*age − 3.13 (1.47)* Time*baseline risk − 10.86 (3.12)** − 8.87 (1.84)** − 2.94 (0.50)** **p < 0.001; *p < 0.05 Fig. 1 Changes in ASQ-SE scores over time (Time 1 N = 188; Time 2 N = 127; Time 3 N = 73) by age. ASQ-SE clinical cutoff scores vary by age and are indicated by the red line. Children aged 1–5 scored on average, above the cutoff scores at Time 1 and below the cutoff score at Time 3 40 <1 1234 5 Baseline Age in Years Time 1 Time 2 Time 3 Age-specific Clinical Cutoff 1 3 Average ASQ-SE Score Maternal and Child Health Journal (2018) 22:1502–1510 1507 indicated by a statistically significant (p = 0.03) age-time inter- Time 1 to Time 3, bringing average scores for this group action estimate (− 3.13). within the healthy range. Caregivers whose scores were For children ages 6–8 years (N = 75 at baseline), individual within the healthy range did not change significantly Time growth models were used to compare changes in CBCL total 1–3. The difference between the groups was statistically sig- problems scores over the 1 year study period for children who nificant as indicated by the time-baseline risk interaction scored above the clinical cutoff at Time 1 compared to chil- estimate of − 8.87 (p < 0.001). dren who scored in the healthy range at Time 1. While total For depressive symptoms as measured by the PHQ-9 problem scores of children who scored below the cutoff at (N = 181 caregivers at baseline), individual growth mod- Time 1 remained on average below the cuto, ff the scores of els compared MA_LAUNCH caregivers with clinical-level children who scored above it at Time 1 dropped by over 25 depressive symptoms at Time 1 (scores > 10) with those points, a 37% decline (Fig. 2). The difference between the caregivers with scores in the non-clinical range. The fitted two groups was statistically significant as indicated by the growth model (Table 2; Fig. 4) estimating change in depres- time-baseline risk interaction estimate of − 10.86 (p < 0. 001). sive symptoms over time indicates that caregivers with high depressive symptoms at Time 1 had a decline of almost 6 Stress and Depressive Symptoms Among Caregivers points with scores on average falling within the non-clinical of MA_LAUNCH Children range by Time 3. Differences between the two groups were statistically significant as indicated by the time-baseline risk Results from individual growth models of PSI-SF scores interaction estimate of − 2.94 (p < 0.001). of MA_LAUNCH caregivers (N = 167 at baseline) com- pared caregivers with high parental stress levels at Time 1 (scores > 85) with caregivers within the healthy range (scores = 16–85) (Table 2; Fig. 3). Caregivers with scores above the cutoff showed an average decline of 17.38 points Fig. 2 Comparison of children 80 who scored above versus below the clinical cutoff score at Time 1 (Time 1 N = 75; Time 2 N = 45; Time 3 N = 28). On average, children who scored below the cutoff at baseline retained their healthy range status whereas children who scored above the cutoff at base- line, dropped below the cutoff by Time 3 Time 1Time 2Time 3 Measurement Occasion Clinical Cutoff Healthy range at baseline Clinical range at baseline Fig. 3 Comparison of caregiv- 110 ers who scored above versus below the clinical cutoff score at Time 1 (Time 1 N = 167; Time 2 N = 131; Time 3 N = 97). On average, caregivers who scored below the cutoff at baseline retained their healthy-range status whereas caregivers who 75 scored above the cutoff at base- line, dropped below the cutoff by Time 3 Time 1Time 2Time 3 Measurement Occasion Clinical Cutoff Healthy range at baseline Clinical range at baseline 1 3 PSI Score CBCL Score 1508 Maternal and Child Health Journal (2018) 22:1502–1510 Fig. 4 Comparison of caregiv- 16 ers who scored above versus below the clinical cutoff score at Time 1 (Time 1 N = 181; Time 2 N = 142; Time 3 N = 90). On average, caregivers who scored 10 below the cutoff at baseline retained their healthy-range status whereas caregivers who scored above the cutoff at base- line, dropped below the cutoff 4 by Time 3 Time 1Time 2Time 3 Measurement Occasion Clinical Cutoff Healthy range at Baseline Clinical range at Baseline sharing of resources and experiences. Situating the interven- Discussion tion within the medical home allowed families and health care professionals to maximize time together and make use Children and caregivers served by MA_LAUNCH expe- of family partners’ shared experiences and cultural back- rienced significant improvements on common screening grounds. MA_LAUNCH teams promoted the medical home tools for social, emotional and behavioral health, especially model’s ideal of family-centered approaches and confronted for those participants with scores in clinically concerning barriers to providers’ ability to address ECMH by providing ranges initially. Children had on average clinically signifi- immediate availability, support and center-wide activities for cant screening results for all but the youngest age group providers, families and to their clinics overall. upon study entry, and 1 year later they were (on average) in As with all studies, findings should be interpreted consid- healthy ranges. Similarly, caregivers with clinically signifi- ering limitations. The first is that the ASQ-SE is designed to cant parenting-related stress and depressive symptom scores be used as a screening and monitoring tool rather than for were on average below clinical cutoffs after 1 year. Children diagnostic outcome measurement; this tool was used in this and caregivers whose screening results did not indicate clini- study to monitor children’s changes in problems over time. cal risk at the beginning of the study retained healthy status. Results should not be interpreted as diagnostic. Secondly, Evaluations of models that promote universal behavioral for families in crisis, screenings were sometimes delayed health screening and embed mental health clinicians within until after the crisis subsided and thus after the family began primary care sites reveal mixed results (Godoy et al. 2014; receiving services. This delay may have resulted in underes- Spijkers et al. 2013; Briggs et al. 2012). For example, one timations of problems. Some subsequent assessments were model embedding an infant/toddler specialist in primary also completed later than ideal due to scheduling dic ffi ulties. care settings reported improved ASQ-SE scores (Briggs Some families were lost to follow up or did not complete et al. 2012) whereas one offering brief counseling found no assessments at all three time points. Reasons for loss to fol- improvements over usual care (Spijkers et al. 2013). Results low up included that the child aged out of MA_LAUNCH from Project LAUNCH in Rhode Island, an initiative with- services, the family moved outside of the service area, or out the family partner role, resulted in improved identifica- staff were unable to maintain contact with the family. How - tion of young children at risk but found that subsequently ever, sensitivity analyses showed the same results whether younger children were less likely than older children to families completed two or three assessments. Sample size utilize subsequent mental health consultation (Godoy et al. limited the researchers’ ability to include large numbers of 2014). These results highlight needs for innovative family covariates. This limitation was not expected to influence engagement approaches, particularly families with very results due to the homogeneity of the sample; sensitivity young children in stressful circumstances. analyses revealed no effect of excluded demographic vari- The MA_LAUNCH intervention addresses gaps in ables. While a comparison site was recruited, budgetary ECMH service delivery for families by integrating trained and time constraints allowed for data collection only among family partners and mental health clinicians working as children, and for only one follow-up screening assessment. teams in primary care settings. MA_LAUNCH teams Although the same criteria of exposure to a list of risk worked across sites in learning collaboratives, attended factors was utilized, very few children from the compari- monthly meetings together, and received supervision from son group had ASQ-SE scores above the clinical cutoff at the state and local public health departments, facilitating 1 3 PHQ-9 Score Maternal and Child Health Journal (2018) 22:1502–1510 1509 (IUHR), at Northeastern University, Boston, MA. The authors of this baseline. The lack of differences and limited data precluded paper would like to acknowledge and thank the participating com- meaningful comparisons of change between groups. Lack of munity health center pediatric practices, the MA_LAUNCH provider meaningful comparison site data is a limitation of this study. teams, the Massachusetts early childhood mental health network and Given these evaluation results, future research examin- parent council, and all the families who participated in this study. ing the effectiveness of integrating behavioral health ser - Open Access This article is distributed under the terms of the Crea- vices into primary care is warranted, with attention on the tive Commons Attribution 4.0 International License (http://creat iveco impact of specific aspects of the MA_LAUNCH model. mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- Although not included in the data presented here, process tion, and reproduction in any medium, provided you give appropriate evaluation results showed the intervention to be widely credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. accepted by caregivers who reported nearly unanimous satisfaction when interviewed by the evaluation team as to whether the MA_LAUNCH team helped them under- stand their children’s emotions and behaviors and whether References they helped them help their children express feelings in more positive ways. Data collected from non-LAUNCH Abidin, R. R. (1995). Parenting stress index, third edition: professional providers within the clinical sites and the MA_LAUNCH manual. Odessa, FL: Psychological Assessment Resources. teams themselves also reported many benefits includ- Achenbach, T., & Rescorla, L. (2008). Manual for the ASEBA School- ing care coordination successes, integration of services, Age Forms & Profiles: an integrated system of multi-informant assessment. Burlington, VT: University of Vermont. and increased access to school registration and school Ader, J., Stille, C. J., Keller, D., Miller, B. F., Barr, M. S., & Perrin, J. and child care services. Components to explore further M. (2015). The medical home and integrated behavioral health: include whether dose–response effects exist for families Advancing the policy agenda. Pediatrics, 135(5), 909–917. who receive services more frequently or for longer periods American Academy of Child and Adolescent Psychiatry. (2010). A guide to building collaborative mental health care partnerships of time; impacts of the family partner role versus the clini- in pediatric primary care. Retrieved March 15, 2017, from https cian role; the influence of full integration of the model into ://www .aacap .or g/App_Theme s/AA C AP /docs/clini cal_pr act primary care on service acceptance and adherence; and ice_center/guide _t o_building_colla bor ativ e_mental_healt h_car e_ physicians’ perceived abilities to incorporate behavioral partn ershi ps.pdf. American Academy of Pediatrics. (n.d.). Salud para todos improving health treatment into their practice. health through medical homes. Retrieved from https ://medic alhom Project LAUNCH in Massachusetts was delivered in einfo .aap.org/tools -resou rces/Docum ents/RI%20Fac tShee t.pdf. community-based pediatric primary care practices in low Bayer, J. K., Ukoumunne, O. C., Lucas, N., Wake, M., Scalzo, K., & income urban settings. Children served at these centers are Nicholson, J. M. (2011). Risk factors for childhood mental health symptoms: National longitudinal study of Australian children. primarily recipients of MassHealth, the state Medicaid pro- Pediatrics, 128(4), e865–e879. gram. Families in this study experienced community- and Boston Public Health Commission. (2014). Early childhood mental family-level risk factors, and health and/or education dis- health toolkit: Integrating mental health services into the pediat- parities. The results of the evaluation of the MA_LAUNCH ric medical home. Retrieved March 15, 2017, from http://www. ecmhm atter s.org/ForPr ofess ional s/Pages /Medic alHom e.aspx. project show that both children and caregivers experienced Brauner, B. C., & Bowers, S. C. (2006). Estimating the prevalence improvements in social, emotional and behavioral difficul- of early childhood serious emotional/behavioral disorders: ties. Given these results, it is likely MA_LAUNCH helped Challenges and recommendations. Public Health Reports, 121, to overcome barriers to service delivery and attainment of 303–310. Breysse, J., Dixon, S., Gregory, J., Philby, M., Jacobs, D. E., & Krieger, social and emotional well-being. Interventions are needed J. (2014). Effect of weatherization combined with community that put children on a healthy developmental pathway. The health worker in-home education on asthma control. American results of this study can be used to inform the development Journal of Public Health, 104, e57–e64. of similar programs and policies to address disparities in Briggs, R. D., Stettler, E., Silver, E., Schrag, R., Nayak, M., Chinitz, S., & Racine, A. D. (2012). Social-emotional screening for infants access to behavioral health care and well-being for very and toddlers in primary care. Pediatrics, 129(2), E377–E384. young children. Bruns, E. J., & Walker, J. S. (Eds.) (2008). The resource guide to wrap- around. Portland, OR: National Wraparound Initiative, Research Acknowledgements Massachusetts Linking Actions for Unmet Needs and Training Center for Family Support and Children’s Mental in Children’s Health (MA_LAUNCH) is an initiative of the Massa- Health. Retrieved March 15, 2017, from https ://n wi.pdx.edu/NWI- chusetts Department of Public Health, funded by the U.S., Substance book/Chapt ers/COMPL ETE-RG-BOOK.pdf. Abuse and Mental Health Services Administration (SAMHSA), (Grant Cournos, F., & Goldfinger, S. M. (2014). Family partners improve early No. SM 059334). The original intervention was carried out in col- childhood mental health services. Psychiatric Services, 65(11), laboration with the Boston Public Health Commission. The evalua- tion study was carried out by the Institute on Urban Health Research 1 3 1510 Maternal and Child Health Journal (2018) 22:1502–1510 Foy, J. M. (2010). Enhancing pediatric mental health care: Report Oppenheim, J., Stewart, W., Zoubak, E., Donato, I., Huang, L., & from the American Academy of Pediatrics Task Force on Mental Hudock, W. (2016). Launching forward: The integration of Health. Introduction. Pediatrics, 3, S69–S74. behavioral health in primary care as a key strategy for promot- Garner, A. S., Shonkoff, J. P., Siegel, B. S., Dobbins, M. I., Earls, M. F., ing young child wellness. American Journal of Orthopsychiatry, Garner, A. S., … Wood, D. L. (2012). Early childhood adversity, 86(2), 124–131. toxic stress, and the role of the pediatrician: Translating develop- Reitman, D., Currier, R. O., & Stickle, T. (2002). A critical evaluation mental science into lifelong health. Pediatrics, 129(1), e224–e231. of the parenting stress index-short form (PSI-SF) in a Head Start https ://doi.org/10.1542/peds.2011-2662. population. Journal of Clinical Child and Adolescent Psychology, Gilkey, M., Garcia, C. C., & Rush, C. (2011). Professionalization and 31(3), 384–392. the experience-based expert: Strengthening partnerships between Sharp, C., & Fonagy, P. (2008). The parent’s capacity to treat the child health educators and community health workers. Health Promo- as a psychological agent: Constructs, measures and implications tion Practice, 12(2), 178–182. for developmental psychopathology. Social Development, 17(3), Godoy, L., Carter, A. S., Silver, R. B., Dickstein, S., & Seifer, R. 737–754. (2014). Mental health screening and consultation in primary care: Spijkers, W., Jansen, D. E., & Reijneveld, S. A. (2013). Effectiveness of The role of child age and parental concerns. Journal of Devevel- primary care Triple P on child psychosocial problems in preven- opmental and Behavior Pediatrics, 35(5), 334–343.9. tive child healthcare: A randomized controlled trial. BMC Medi- Goodman, A., Joyce, R., & Smith, J. P. (2011). The long shadow cast cine. https ://doi.org/10.1186/1741-7015-11-240. by childhood physical and mental problems on adult life. Proceed- Squires, J., Bricker, D., Heo, K., & Twombly, E. (2001). Identification ings of the National Academy Sciences of the United States of of social-emotional problems in young children using a parent- America, 108(15), 6032–6037. completed screening measure. Early Childhood Research Quar- Horwitz, S. M., Kelleher, K. J., Stein, R. E., Storfer-Isser, A., Young- terly, 16, 405–419. strom, E. A., Park, E. R.,… Hoagwood, K. E. (2007). Barriers Squires, J., Bricker, D., & Twombly, E. (2002). The ASQ:SE user’s to the identification and management of psychosocial issues in guide: For the Ages & Stages Questionnaires: Social-emotional. children and maternal depression. Pediatrics, 119(1), e208–e218. Baltimore: Paul H Brookes. Huang, F., Chung, H., Kroenke, K., Delucchi, K., & Spitzer, R. (2006). Thompson, R., Tabone, J. K., Litrownik, A. J., Briggs, E. C., Hussey, Using the Patient Health Questionnaire-9 to measure depression J. M., English, D. J., & Howard, D. (2011). Early adolescent risk among racially and ethnically diverse primary care patients. Jour- behavior outcomes of childhood externalizing behavioral trajecto- nal of General Internal Medicine, 21(6), 547–552. ries. The Journal of Early Adolescence, 31(2), 234–257. Justvig, S. P., Li, J., Caravella, G., Chen, M., Wang, H., Benz Scott, L. Tyler, E. T., Hulkower, R. L., & Kaminski, J. W. (2017). Milbank A., & Pati, S. (2017). Improving adherence to care recommenda- memorial fund. Behavioral health integration in pediatric pri- tions using a community health worker (CHW) intervention with mary care: Considerations and opportunities for policymakers, the pediatric medical home. Journal of Community Health, 42, planners, and providers. Retrieved March 15, 2017, from https :// 444–452.www.milba nk.org/publi catio ns/behav ioral -healt h-integ ratio n-in- Kroenke, K., Spitzer, R. L., Williams, J. B. W., & Löwe, B. (2010). pediatr ic-primar y-care-consider ations-and-oppor tunit ies-f or-polic The patient health questionnaire somatic, anxiety, and depressive ymake rs-plann ers-and-provi ders/. symptom scales: A systematic review. General Hospital Psychia- Volkmann, K., & Castanares, T. (2011). Clinical community health try, 32(4), 345–359. workers: linchpin of the medical home. The Journal of Ambula- Matiz, L. A., Peretz, P. J., Jacotin, P. G., Cruz, C., Ramirez-Diaz, E., & tory Care Management, 34(3), 221–233. Nieto, A., R (2014). The impact of integrating community health Wolfinger, R., & Chang, M. (1995) Comparing the SAS® GLM workers into the patient-centered medical home. Journal of Pri- and MIXED procedures for repeated measurements analysis. mary Care and Community Health, 5(4), 271–274. Retrieved March 15, 2017, from https://suppo r t.sas.com/rnd/app/ Meltzer-Brody, S. (2011). New insights into perinatal depression: stat/paper s/abstr acts/mixed glm.html. Pathogenesis and treatment during pregnancy and postpartum. Zbar, A., Surkan, P. J., Fombonne, E., & Melchior, M. E. (2016). Emo- Dialogues in Clinical Neuroscience, 13(1), 89–100. tional and behavioral difficulties and adult educational attainment: Nakamura, B. J., Ebesutani, C., Bernstein, A., & Chorpita, B. F. (2008). An 18-year follow-up of the TEMPO study. European Child and A psychometric analysis of the child behavior checklist DSM- Adolescent Psychiatry, 25(10), 1141–1143. oriented scales. Journal of Psychopathology and Behavioral Assessment, 31(3), 178–189. 1 3

Journal

Maternal and Child Health JournalSpringer Journals

Published: Jun 16, 2018

There are no references for this article.