A Two-Generation, Early Childhood Advanced Primary Care Model

NYC Health + Hospitals is the largest public hospital system in the country, providing PPC to over 160 000 children, including 40 000 children ages 0 to 3 and approximately ∽14 000 births annually. Patients are racially and culturally diverse and speak over 40 languages. Children and pregnant people are primarily publicly insured through Medicaid Managed Care (Table 1).

In 2018, the Robin Hood foundation proposed a partnership with NYC H+H to “build the brains of its youngest patients.” Drawing on recommendations from the First 1000 Days Pediatric Clinical Advisory Group,¹³  an NYC H+H group of pediatric, administrative, behavioral health, and women's health leaders drafted the IMPACT model and issued an internal request for pilot proposals to all PPC sites within NYC H+H. Three facilities, Queens Hospital, Bellevue Hospital, and Gouverneur, (an NYC H+H Federally Qualified Health Center) were selected as pilot sites based on the strengths of their existing early childhood programs and site leadership’s demonstrated interest in two-generation care. Robin Hood, NYC H+H, and the NYC Mayor’s Office of Economic Opportunity (NYC Opportunity) provided funding for the proposed enhanced clinical and evaluation staff and infrastructure, and IMPACT launched on October 1, 2020.

IMPACT involves three disciplines (women’s health, behavioral health, and pediatrics); spans two generations (caregivers and their children); and has one goal: to support the long-term health and well-being of young children and their families (Fig 1). This population health PPC model serves all children aged 0 to 5 in the practice (with priority given to age 0–3), and deploys an expanded care team to meet mental health, developmental, and socioemotional needs of children and families across PPC and women’s health. Specialized staff were integrated into the traditional PPC and women’s health models, including HealthySteps specialists, Video Interaction Project (VIP) coaches, early childhood community health workers (CHWs), licensed social workers, and psychiatrists. In addition, all PPC sites had Reach Out and Read (ROR) programs. Dyadic support is designed to start prenatally and continue until the child’s fifth birthday. The program targets practice transformation at all NYC H+H PPC practices, with many components now implemented systemwide.

Building the clinical model required advocacy at every level. At the New York State level, we worked with advocacy organizations and other pediatric leaders to advocate for billing reform and alternative payment models to sustain this model. Locally, we worked with New York City agencies to align IMPACT with new initiatives under development. At the health system level, we worked with NYC H+H leadership in population health, finance, behavioral health, women’s health, and facility leadership.

IMPACT partnered with the NYC H+H Maternal Home, a maternal morbidity and mortality prevention initiative, to support care during pregnancy and though the “fourth trimester,” a critical period in the first months postpartum for both parent and child. Screening tools for depression, substance use, social determinants of health (SDOH), and adverse childhood experiences were embedded into the women's health electronic medical record (EMR), and IMPACT provided additional psychiatry and social work staff to augment Maternal Home support in pregnancy. These staff conduct assessments, refer to community partners for non-medical social needs, and manage short-term mental health needs. Based on the types and degree of need, families are assigned a “risk tier” level in the EMR. Risk tier documentation in the EMR generates a care report that is visible in both the mother and the child’s EMR records. These EMR tools, accessible systemwide, enhance visibility and facilitate longitudinal follow up. For mothers who continue PPC in IMPACT pediatric practices, prenatal period supports continue in PPC into the “fourth trimester.” When new maternal, mental, or physical health needs are identified in the postpartum period, mothers are referred back to women’s health. Results from prenatal assessments, screenings, and risk tier documentation are communicated by obstetric and pediatric teams via warm handoffs, case conferencing, and the EMR.

Key components of IMPACT’s enhanced PPC model include (1) a strong link to women’s health, with service coordination for families with identified needs; (2) universal screening across multiple domains; (3) timely referral to and improved engagement with early intervention; (4) a robust, practice-based, early childhood community health worker (CHW) program; and (5) integration of three standardized, evidence-based programs to improve child developmental and socioemotional outcomes.

The program is a risk tiered model, founded on the HealthySteps risk model.¹⁴  Tier assignment is a clinical decision by the HealthySteps specialist, factoring in prenatal risk tiering, screening results, and clinical assessment (Fig 2). Following the HealthySteps model, all children engaged in primary care are in tier 1, and have access to HealthySteps consultation, learning resources from the program, Reach Out and Read, and VIP. Those who receive short-term interventions or limited support from HealthySteps, CHWs, or from mental health are designated “elevated risk” (tier 2) and receive follow up every four months, or as “high risk” (tier 3) and receive follow up more closely (every two months) if more intensive or longer-term support is deemed necessary.

Universal screening of all patients in the practice is performed using an EMR-linked, tablet-based platform, using validated screening tools available in multiple languages. Screening includes SDOH, postpartum depression, developmental delay, and socioemotional development, performed on a set schedule agreed upon by NYC H+H ambulatory pediatric leaders (Fig 3) with the goal of minimizing screening burden at any one visit. Results are reviewed and referrals are made by the primary care provider if apprproaite and if desired by the family.

The three evidence-based programs that form the foundation for IMPACT are: HealthySteps, the Video Interaction Project (VIP), and ROR. HealthySteps colocates a licensed child development expert (psychologist or social worker) in the PPC practice to foster healthy cognitive, perceptual, motor, physical, and socioemotional development, and language and literacy skills.^15 –20  VIP is delivered at routine PPC visits by a trained coach who videotapes parent-child dyad interactions and provides developmentally-focused positive feedback and coaching to the parent. The coach also provides developmentally-appropriate toys, books, and resources to enhance daily activities that promote early child development and literacy (eg, pretend-play, dyadic reading, and daily routines).^21 –23  ROR is a nationally-disseminated program in which providers counsel parents on the importance of reading aloud to their infants and young children. PPC providers model dyadic reading and provide an age-appropriate book to children under age five at each well-child visit.^24,25  In addition to these three programs, PPC providers distribute age-specific positive parenting handouts based on evidence-informed VROOM© materials.²⁶ 

Expanding on the model developed for IMPACT, NYC H+H implemented an early childhood CHW program at all PPC practices across the health system. CHWs embedded in PPC are trained in a two-generation, early childhood model, providing community resource navigation for young children and families. These include connections to early intervention (EI), preschool special education, subsidized day care, benefits, food access, legal services, postpartum care, and evidence-based parenting support. Supported by these CHWs, the EI referral process was developed in collaboration with the NYC Department of Health and Mental Hygiene, and is now bidirectional, closed loop, and integrated into the EMR. Finally, CHWs support care transitions from women’s health to pediatrics and ongoing family connection to services throughout age five.

The IMPACT care model requires complex and collaborative implementation and clinical workflows and an EMR build that supports this complexity. During its initial phase, leaders at the three IMPACT pilot facilities met weekly to discuss workflows and program implementation. This collaboration led to workflow and staffing model modifications that were essential to successful implementation. The team built simple, flexible, and user-friendly EMR tools that supported agreed-upon workflows, producing real-time data for program evaluation and quality improvement. A limited number of key structured data elements were integrated into various provider documentation templates and pull into EMR summary tools, to enable enhanced chart review by monitoring family engagement with tier 2 and 3 services (eg, consults with HealthySteps and initial visits with CHWs). Specific structured data elements were also shared across provider documentation templates to enable more collaborative documentation and clinical care (eg, risk tier and referrals). These tools are designed to be simple (ie, have only a few structured elements allowing better enforcement of universal use): for example, they allow free text (supporting more nuanced individualized assessments), and both structured elements and free text components are viewable in summary tools that allow efficient review across multiple disciplines. For infants born in our health system, mother and baby charts are linked through five years postpartum. EMR reports built to support clinical care and quality improvement also provided required reporting to funders, evaluators, and National Offices for HealthySteps, VIP, and ROR.

PPC models like IMPACT are predominantly funded through private and public grant funding, with fewer than 50% of its services reimbursable through current Medicaid payment structures. Vitally important yet often unbillable model components include some interdisciplinary team members (eg, CHWs, VIP coaches), preventive mental health services, resource navigation, many screenings, and other parental support and education activities. A key focus for IMPACT is advocating for improvements to current billing revenue capture, and implementing an advanced payment model to cover nonbillable services.

There have been several recent opportunities to improve revenue capture for preventive behavioral health billing and licensed clinical social work services. Provider training and improved EMR automation and documentation did double our billing revenue for IBH services, but from a very low baseline. Restrictions on diagnosis codes made only a fraction of care provided by IMPACT social workers reimbursable. A statewide advocacy collaboration focused on primary care transformation led to the clarification of New York State Medicaid billing guidance for behavioral health services, using Z codes to document family risk. This allows billing for preventive interventions in the absence of a Diagnostic and Statistical Manual of Mental Disorders, fifth edition diagnosis, as well as for dyadic services on the child’s Medicaid. Through a Special Plan Amendment (SPA) and budget authority approvals, billing was authorized for these mental health services in New York. The Special Plan Amendment also authorized billing by CHWs working under a licensed care provider in women’s health, with expansion to pediatrics expected in 2024. We await rule and rate code definitions that will allow us to apply these new billing opportunities.

Although these efforts will substantially increase IMPACT’s revenue, opportunities to bill for prevention-focused care in primary care and women’s health settings remain limited. Reimbursement does not cover the cost of the expanded IMPACT team model. Moreover, NYC H+H is in nearly full-risk capitated arrangements with most of its contracted managed care organizations (MCOs), which limits the impact of billing revenue for our system. Also, components of care—such as non-face-to-face care coordination by billing providers and support from supervised, but unlicensed, care providers (eg, VIP)—will likely remain nonbillable.²⁷  Given evidence to support the high value and cost effectiveness of these nonbillable services,²⁸  a national movement is growing to support these PPC models through alternative payment models (APMs), a payment structure that focuses on population level payments when a practice implements high-quality, cost-effective care for a particular condition or population.^29 –31 

For an APM to support a PPC-focused, primary prevention and two-generation care, stakeholders must look beyond traditional value-based payment models.³²  Models that effectively support a program like that recommended by the NY State’s Pediatric Advisory Group¹³  must look beyond short-term (one-year) quality and cost outcomes to realize the full value of these services and resource investments in early childhood. APMs must use data structures that measure quality and cost outcomes across the dyad or family, and premium risk adjustments must include social, emerging, and dyadic risk. Ideally, to facilitate and assess population-level savings, all of a state’s managed care plans should participate and share accountability, and data sharing agreements should be in place to identify cross-sector savings in a state’s budget. Nationally, leaders are advocating for these types of APMs leveraging Medicaid opportunities, such as 1115 Waivers, pilots similar to “in-lieu of services'' or Patient-Centered Medical Home or Center of Excellence, state plan amendments, and state budgets.^33 –35 

As the largest public hospital system in the country with nearly full-risk arrangements with MCOs, NYC H+H is well-positioned to pilot a population-focused APM model with MCO partners to support early childhood and dyadic care long-term. With two MCOs, NYC H+H is cocreating an early childhood APM pilot that we expect to implement in 2024.

From October 2020 to September 2022 at its three pilot sites, IMPACT served over 3365 expectant mothers and 3786 children aged 0 to 3. In year two (Y2), 87% of prenatal patients were screened for psychosocial risk and 57% were reached by enhanced IMPACT services (CHW, HealthySteps, and/or integrated behavioral health). In pediatrics, children were screened at all sites across five domains (see Table 2 for screening rates). 47 percent of pediatric patients were reached by at least one enhanced IMPACT service, including CHW support services (30%), integrated behavioral health (4%), or at least one evidence-based support program (HealthySteps or VIP; 29%). In accordance with a tier-based approach, patients identified as having elevated risk in pediatrics (N = 1139) and women’s health (N = 468) in Y2, received nearly five times the number of contacts and touchpoints by enhanced IMPACT services than those without any risk identified. Data are from the NYC H+H EMR.

For the patient satisfaction survey, two members of the evaluation team were positioned in the waiting room of the women’s health and pediatrics departments at the three sites from May to December 2022. To identify eligible participants while they waited for their appointments or had gaps between providers, the team obtained daily appointment lists. Eligible patients included English- or Spanish- speaking pregnant patients and caregivers of one or more children between 0 and 3 years of age who attended a pediatric well-child visit, new patient visit, or regular obstetrician and gynecologist checkup in the women’s health or pediatric setting. These patients were provided with hard copies of the survey, which were completed during appointments and returned to the research team. Of the 2813 patients eligible, a total of 331 patients completed the survey for a response rate of 11.8%. Between April and May 2022, the clinic staff and administration received a survey via a Qualtrics link that captured their perspectives about the model and its implementation. Eligible participants included leadership, administrative staff, primary care providers, and expanded care team providers (such as nurses, community health workers, and integrated behavioral health providers) within the women’s health and pediatrics departments across the three sites. A total of 492 surveys were initially distributed via e-mail, and 138 completed the survey, resulting in a 28% response rate.

Caregivers and staff surveys reflected high satisfaction with the model. Based on survey data, most caregivers (N = 331) found IMPACT services very helpful (82%), very easy to obtain (62%), and matched their needs very well (61%). A majority of staff surveyed (N = 138) also reported that services were very helpful to families (68%) and met families’ needs very well (57%). Based on analyses of open-text responses from staff, major facilitators to implementation included shared decision-making, case conferencing, and warm handoffs between the primary and expanded care teams. Areas to strengthen included enhanced communication among the care teams, and clarifying workflows and referral pathways.

Over three years, the IMPACT initiative was fully implemented in three NYC H+H practices with many core components scaled systemwide, reaching over 11 000 children aged 0 to 3. The early childhood-trained CHW program is in place in 16 NYC H+H PPC practices (65+ early childhood CHWs funded by Public Health Corps, a Mayor’s office initiative to transform the coronavirus disease 2019 outreach workforce); tablet-based, multilingual, universal screening at 16 NYC H+H PPC practices (funded by NYC H+H), with HealthySteps implemented at 12 sites by the end of 2023 (funded by NY State Office of Mental Health).

A closed loop referral system with the NYC EI program was implemented successfully at IMPACT pilot sites and then spread to all NYC H+H PPC practices; it is now a model for expansion to three additional healthcare systems in NYC. Preliminary data for IMPACT sites demonstrates earlier ages at completed referral over the first year (decreased from 24–29 months to 18–23 months), and mitigation of poverty and race and ethnicity-related disparities in the referral process, with increased referrals over the first year for the lowest income patient cohort (50% vs 40%) and for Black and Latino children (11% increase vs 3% citywide) (NYC Department of Health and Mental Hygiene, Personal communication, 2022).

IMPACT’s complex service model, population-health focus, and sustainability challenges require a multi-pronged approach to evaluation and ongoing quality improvement. Currently underway, this includes: (1) a mixed-methods implementation evaluation focused on access, fidelity, and capacity; (2) a three-year outcomes evaluation that uses New York State Medicaid Claims data to track maternal and child core Medicaid outcomes from claims-based utilization, quality, expenditures, and enrollment data; (3) national program fidelity reporting to ensure IMPACT sites reach fidelity to established HealthySteps, ROR, and VIP models; and (4) internally-derived key performance indicators to inform quality improvement work.

With support from both NYC H+H and philanthropic funders, the IMPACT population health model has been fully implemented at three pilot sites, sparking substantial practice transformation across all NYC H+H PPC practices. A key innovation of the IMPACT initiative and related advocacy was naming social-emotional well-being as the outcome of interest for pediatric population health during early childhood—and underscoring that PPC practice must support mothers and families to achieve this goal. This dyadic focus helped us to effectively advocate internally and externally for system-level resources to address population health in primary care—resources that were previously focused on the care of adults with chronic illness. The population health focus also drove the implementation of multiple evidence-based interventions to scale across our large public hospital system.

Our successful efforts depended on our collaborative approach to program design. Team members worked with the New York State First 1000 Days group to define the model of Advanced Primary Care for Early Childhood. IMPACT leads worked closely with health system leadership who supported advocacy with MCOs and New York State and provided critical systemwide resources, specifically scaling the Early Childhood CHW model and implementing tablet screening in all PPC practices. NYC H+H also invested substantial funds to support new clinical lines, matching the large philanthropic contribution from Robin Hood and from the mayoral office, NYC Opportunity. Philanthropic funds supported start-up costs and nonclinical care team members in our pilot sites.

Attempting substantial change at both the practice and system levels—and spanning the period of the coronavirus disease 2019 pandemic—was difficult: more work remains to make the changes last. We built expanded care teams but sustained substantial challenges related to space constraints and difficulty establishing clear workflows across overlapping roles (eg, social work and CHW). We also faced workforce challenges recruiting mental health roles. For the scaled systemwide components, uptake varies. This year, we are focused on quality improvement to ensure quality implementation systemwide.

As mental health and SDOH-related needs in women’s health were more systematically identified, we realized that significantly more resources are needed to effectively support maternal health. Dyadic needs can be addressed in PPC, but mental health needs specific to the caregiver cannot. Referral of pregnant and new mothers to mental health services is difficult and successful engagement is uncommon. More robust services must be embedded in women’s health departments to address a wide range of mental health needs and offer facilitated access from pediatrics. A truly two-generational model of care will require practice transformation in women’s health and focused efforts to break down silos across pediatrics and women’s health clinical practice lines.³⁶ 

Sustaining this model of care at NYC H+H will require an alternative payment model with our largest contracted MCOs. We anticipate implementing such a model in 2024. National expansion will require further advocacy and work to define payment models that truly center the mother-child dyad for the medium- and long-term. Further evaluation is needed to assess child, family, and health system-level outcomes. Our belief is that this approach can acheive synergy between multiple evidence-based and evidence-informed interventions that individually have relatively small effect sizes, but combined could substantially improve child and maternal health outcomes.

The IMPACT care model ensures all families have access to preventive, dyadic, and integrated services; preliminary results indicate that successful delivery of this care model is feasible in a large public healthcare system. Long-term sustainability of this or similar APC models will depend on the implementation of early childhood-focused advanced payment models. Sustainability will require the pediatric community to agree on an advanced pediatric primary care model that may vary for different age groups. Advocacy should focus on these areas.

We wish to gratefully acknowledge the vision of Robin Hood, The NYC Mayor's Office for Economic Opportunity, and NYC Health + Hospitals leadership in supporting such an ambitious project, the specific contributions of Shauntée Henry, Nura Anwar and Shivali Choxi who each led the implementation of important program components and the facility-based staff of IMPACT whose dedication and thoughtful participation nurtured the vision to become reality.

APM

advanced payment model

CHW

community health worker

EMR

electronic medical record

ERH

early relational health

IMPACT

Integrated Model for Parents and Children Together

MCO

managed care organization

NYC H+H

NYC health and hospitals

PPC

pediatric primary care

ROR

Reach Out and Read

SDOH

social determinants of health

VIP

Video Interaction Project