Clinic-Based Financial Coaching and Missed Pediatric Preventive Care: A Randomized Trial

We conducted this randomized controlled trial of clinic-based financial coaching among low-income parents at the pediatric primary care clinic at Harbor-UCLA, an academically affiliated safety net medical center and second largest in Los Angeles County’s Department of Health Services. Data for this study of early findings were collected from November 2018 to January 2020, before the COVID-19 pandemic, and participants accrued complete outcome data for this study’s endpoints by January 2020.

English- and Spanish-speaking adult parents and their infants younger than 4 months of age presenting to care for scheduled, nonacute primary care visits at the study clinic were eligible to participate. These parent-infant dyads were approached, recruited, and enrolled from clinic waiting areas or examination rooms by research staff between fall 2018 and summer 2019. Informed consent was obtained in English or Spanish by study staff who randomized participants to either intervention (clinic-based financial coaching plus usual care) or control (usual care) arm using computer-generated simple randomization. Participants received $20 incentive cards at enrollment and the 6-month follow-up. More than one-half of families were enrolled 1 to 4 weeks after birth, three-quarters enrolled before infants’ 2-month visits, and the remainder enrolled at the 2-month visit. Usual care included standardized social needs screening forms (Supplemental Fig 4) and clinician referral to community- based resources for identified needs, as well as routine clinic visit reminder calls, text messages, and mailings. Participant well-child visit adherence and vaccinations were measured during visits after enrollment up to the 6-month preventive visit or child age 9 months, whichever came first.

This study presents early findings of a secondary outcome (preventive pediatric care visit adherence rate) from the first year of a 5-year randomized controlled trial of clinic-based financial coaching. For the broader trial, parent-reported survey outcomes are collected every six months, meaning follow-up results accumulate more slowly than the preventive pediatric visit attendance outcome, which has at least four data points per participant by the time the first parent-reported measures are collected. The visit adherence outcome is, therefore, expected to be better-powered than the primary parent-reported health measures, which have not been analyzed at this early stage of the broader study at the time of this publication.

Intervention arm parents received financial coaching services as part of their infants’ well-child visits, with coaching usually taking place in examination rooms while parents waited for clinicians. Financial coaches delivering the intervention were in social work training programs and received at least 16 hours of additional financial coaching training before working with participants under direct supervision thereafter by a licensed clinical social worker and pediatrician. The coaches received an ongoing curriculum on motivational interviewing, social determinants of health, community partnerships, integrated health care, and more. Coaches were paired with intervention participants at enrollment and followed up with them in-person during periodicity schedule preventive care clinic visits and remotely throughout the study period at least monthly (often more depending on parent need/interest).

The intervention was developed based on the well-established financial coaching model that has grown widely in the financial services field but has not been applied in health care.^32,33  The approach was specifically modeled after that implemented by LIFT, a national financial coaching organization and the study’s community partner agency. The intervention’s clinical integration feasibility, model fidelity, and acceptability were confirmed through workflow time motion analyses, participant focus groups, and clinic-wide surveys indicating high demand for clinic-based financial services (Bell O, Wu O, Perales L, et al., unpublished data). Consistent with prior literature, we confirmed underuse of public benefits, such as supplemental nutrition assistance and utilities discounts, among clinic families, despite universal eligibility by virtue of their Medicaid enrollment (A.S., L.P., and M.H., unpublished data).

The financial coaching model prioritizes longitudinal client-coach relationships as the foundation for trust and motivation toward client financial behavior change in alignment with client-directed financial goals, strengths, and needs. The study financial coaching intervention consisted of interrelated components furthering parent financial goal plans to address areas of greatest financial potential jointly determined by participant and coach. Goals and action plans were developed through motivational interviewing and shared decision-making with coaches at each visit. Coaching also included assessment of income (including public benefits eligibility/receipt), savings, credit, debt, and taxes, as well as expense reduction opportunities through household budgets. Coaches identified parent strengths and navigated participants to and through cost-saving services and public benefits, such as affordable child care, transportation and utilities discounts, nutrition assistance, and free tax preparation to access child tax credits. Financial coaching encounters in clinic lasted approximately 20 to 40 minutes (often before/after primary pediatricians met with the families), and parents were provided written referral materials and diapers when requested. Parents communicated with their coaches between visits via phone and text-based messaging at least monthly to advance financial goals.

The control group received usual clinical care. Usual care included telephonic clinic visit reminder calls from clinic staff 1 to 2 days in advance (with voicemails left and 3 call attempts if calls were unanswered), a text message visit reminder from the health system roughly 3 days before the appointment, a mailed postcard visit reminder, and paper-based social needs screenings at each preventive visit (the same reminders and social needs screenings received by the intervention group) as part of a universal clinic protocol. Screens were reviewed by pediatricians and nurse practitioners providing external resource referrals (most commonly for acute material hardships); those with positive screens were instructed to take resource sheets available in every clinic examination room. These referrals did not include financial coaching, however.

The primary outcome of missed preventive care pediatric visits was measured as the rate of attendance at recommended 1-, 2-, 4-, and 6-month visits per the AAP Periodicity Schedule (ie, number of visits attended out of number recommended, at each Periodicity Schedule interval and overall through the 6-month visit period). Data regarding preventive pediatric visits at the study clinic site were available from the electronic health record for all participants. Appointments were considered missed if they were not attended and the child aged into the window at which a subsequent visit was recommended (eg, a child would be marked as having missed a 6-month preventive visit if he or she did not attend a primary care visit after 6 months of age and the child turned 9 months old, and he or she did not have an early preventive care visit before 6 months to administer 6-month vaccines). Visits were included in outcome analyses if they occurred after the time of enrollment and before the end of the 6-month visit period in the Periodicity Schedule, regardless of follow-up time elapsed.

Because children who miss preventive pediatric visits early in life (especially at 2, 4, and 6 months, when AAP recommends primary series poliovirus, diphtheria and pertussis, rotavirus, hepatitis B, and Haemophilus influenzae vaccinations) would be expected to also miss needed vaccinations; this would have a large impact on health in its own right. Post hoc, we explored the number of missed vaccine doses to capture the impact of missed visits. The vaccination outcome was assessed (1) based on the total count of incomplete recommended vaccinations at the end of the child’s eighth month of life and (2) as a binary indicator of whether the child was up to date on vaccinations at the end of each recommended preventive visit child age window according to the harmonized Centers for Disease Control Advisory Committee on Immunization Practices and AAP schedules.³⁴  Vaccinations were verified through electronic health record data from the California Immunization Registry, allowing us to measure vaccinations given anywhere in California.

Other measures collected at baseline and follow-up included child age, parent sex, household size, employment, race/ethnicity, language, income, savings, and debt.

We used t tests and tests of proportions to compare participant baseline characteristics. To assess missed-visit and vaccine outcomes, we examined zero-inflated negative binomial regression models for these zero-inflated and overdispersed count outcomes (Stata, zinb), and incidence rate ratios were estimated using an exposure variable of the number of preventive pediatric care visits recommended in the AAP Periodicity Schedule after participant enrollment through the end of the 6-month preventive visit period. Differences in rates and counts between intervention and control groups were assessed via postestimation (Stata, margins). Longitudinal generalized estimating equation regression models interacting the study arm variable with a follow-up time period variable for each visit were used to estimate the difference in rates attending 1-, 2-, 4-, and 6-month well-child visits, as well as differences in rates of vaccination completion and counts of vaccines missed. All models used intent- to-treat analytic frames and 2-sided alphas, using Stata 14 (StataCorp, College Station, TX).

The study was approved by UCLA and Harbor-UCLA/Lundquist Research Institute institutional review boards and registered on ClinicalTrials.gov (NCT03736590; https://clinicaltrials.gov/ct2/show/NCT03736590).

Eighty-one parent participants were randomized to receive financial coaching plus social needs screening and referral (intervention, N = 35) or usual care (including social needs screening and referral but no financial coaching; control, N = 46; Fig 1). The intervention and control groups had similar characteristics (Table 1). At baseline, parent participants were mostly female, half were Latina, just over half were employed or pursuing educational programs, a majority were English-speaking, and their average income was just above the federal poverty level for a family of 4.

On average, the intervention participants identified 3.8 (SD ± 2.1) financial goals and made progress on or completed 3.2 (SD ± 2.1) during the study period. The most common goals were finding child care, full-time employment, and/or enrolling in public benefits. Intervention participants significantly increased average monthly household income 6 months after enrollment compared with their baseline income (change of $1761; 95% confidence interval [CI], $457–$3065) and their increase in monthly income was $876 (95% CI, –$717 to $2469) greater than control participants. Average savings increased by $842 (95% CI, –$640 to 2325) among intervention participants by 6 months, with a relative change of $1504 (95% CI, –$873 to $3882) compared with controls.

In the primary intent-to-treat analyses, financial coaching parents missed 0.61 (95% CI, 0.16-1.05) fewer preventive pediatric visits relative to control parents in the time period from the 1-month visit to the 6-month visit recommended by the AAP Periodicity Schedule (0.46 vs 1.07 visits missed; P = .01). A similar gap in mean number of missed visits was observed between the groups when the outcome was limited to the 2-, 4-, and 6-month visits at which primary immunization series are administered (0.48 visits; 95% CI, 0.04-0.91; P = .03). The likelihood of attending each preventive care visit between 1 and 8 months of age was more than 20% higher for financial coaching parents relative to the control group (relative rate, 1.28; 95% CI, 1.06–1.54; P = .009). The effect was strongest at the 2 and 4-month visits (Fig 2).

On average, control group children had missed or delayed 1.45 (95% CI, 0.48–2.42) more recommended vaccinations at the study site clinic than those of parents in the intervention group through the 6-month well-child visit period (2.52 vaccines versus 3.98; P = .002). For each well-child visit period, parent-child dyads receiving financial coaching were more than 25% more likely to be up to date on recommended child vaccines (relative rate, 1.26; 95% CI, 1.05–1.51; P = .01).

Children in financial coaching families had higher rate of completion of all vaccines for any given visit overall and greater average number of vaccines received at each visit compared with controls (Fig 3), with an average difference of 0.8 vaccinations received per visit at 2-, 4-, and 6-month visits (95% CI, 0.4–1.21; P < .001). Intervention children remained more likely to have received recommended vaccines even when accounting for vaccinations administered California-wide according to the California Immunization Registry (average difference in vaccines missed 0.5 per visit; 95% CI, 0.26–0.73; P < .001).

In this randomized trial, we found clinically integrated financial coaching led to lower rates of missed preventive pediatric visits and fewer missed immunizations through 6 months of child age among low-income parent-child dyads. The effects were sizable at a difference of approximately half a visit of the 4 recommended between the child age of the 1- and 6-month visit window. Our findings suggest clinic-based financial services through MFPs can influence primary care use and vaccinations.

Missed appointments are a persistent, costly, and health care-disrupting problem for primary pediatrics clinics serving low-income families expected to benefit most from MFPs, with no-show rates reaching 50% in some studies.^28,35  Poverty-related social needs and other measures of financial hardship are among the most consistent predictors of no-shows,^29,36  yet interventions to increase visit attendance have tended not to focus on underlying financial needs of families.^37,38  Our study suggests that addressing financial goals and needs can improve preventive visit care adherence and vaccinations. This may partially offset resource costs to clinics of MFP programs, in addition to optimizing the host of health, developmental, and psychosocial benefits of preventive pediatric care.^39–41 

This study was not designed to determine mechanisms by which MFPs affect visit attendance, though it highlights possibilities. Although financial coaches in our study communicated with participants between clinic visits (like many health care team members), those communications focused on financial stressors and parent-centered goal progress, not clinic visit adherence or reminders. In fact, the coaches were trained specifically to take the long view of building supportive, trusting, nonjudgmental relationships with parents rather than to attempt thwarting no-shows, which were accepted as normal deprioritizations of visits by parents juggling family, work, financial, and social responsibilities. Much of the financial coaches’ work focused on supporting parent financial capabilities and strengths while addressing parent-identified economic hardships known to create barriers to visit attendance, perhaps preventing those barriers from occurring or persisting. Therefore, the MFP may have created greater trust in the health care team by virtue of the coaching approach and new motivation to attend clinic visits by offering clinic-based resources of interest to parents aligned with their self-identified financial goals.⁴²  This may have led to “spillover” effects on visit adherence. Future studies should explore these potential mechanisms.

Limitations of our study are its small size at a single site. There was no feasible way to blind participants or clinic staff, and we cannot rule out unmeasured collateral intervention effects spreading through clinic staff to controls. We did not access information on whether parents declined vaccines, which is uncommon in this clinic. There was no difference in attrition between study arms. Impacts driven by individual coaches or program practices may account for client successes that might not persist in future implementations.

Clinically integrated financial coaching through an MFP increased adherence to preventive pediatric care visits and vaccination in low-income parent-child dyads. This suggests a dual benefit of such programs to patients and clinics alike, which may incentivize health system implementation despite the new personnel and service integration required. The impact of this pilot intervention appeared strongest at the earliest preventive care visits, which might suggest an optimal intervention duration or window for implementation. Future studies should evaluate mechanisms and effects of such programs over the medium to long run, including child health impacts, which may be driven by increased visit adherence and vaccinations as well as increased family financial security.

The authors thank LIFT Communities for partnership, the model of financial coaching used in this study, and training for the coaches who delivered the intervention. We thank the 2018–2020 Harbor-UCLA Medical- Financial Partnership coaches Denise Cerda, Angel Garcia, Erika Jasper, Christina Lozano, Nicol Molina Torres, and Lorraine Perales. We thank research team members Nicole Morales, Jazelle Bautista, and Jacqueline Plaza, as well as the clinical staff at Harbor-UCLA Primary Pediatrics. We thank the parents and families who participated in this study.

AAP

American Academy of Pediatrics

CI

confidence interval

MFP

medical-financial partnership