Integrated Care Cuts Pediatric Behavioral Issues by 30%

Expanding pediatric behavioral health care to include integrative services that encompass meetings with behavioral health clinicians (BHCs) and psychotropic prescriptions led to reductions in scores on the 17-item Pediatric Symptom Checklist (PSC-17), indicating improved overall behavioral health outcomes, according to new research.1

Contributions from meetings with community health workers (CHWs) were also evaluated, but were not shown to make a statistically significant difference in the children in the study cohort compared with the children in the control group. These results were published online today in JAMA Network Open.

The study authors used electronic medical records from June 2020 through April 2023 from 4 federally qualified health centers that implemented the Transforming and Expanding Access to Mental Health Universally in Pediatrics, or TEAM UP, model of integrated behavioral health care. For study inclusion, children had to be aged 4 to 18 years, have completed the PSC-17 at the time of or after their index screening for a behavioral health concern, have at least 1 follow-up PSC-17 between 6 and 18 months later, and have a note in their record of a behavioral health concern from their primary care physician.

A score of 15 or higher on the PSC-17 indicates concerns, with total scores ranging from 0 to 34 and each item being scored a 0 (never), 1 (sometimes), or 2 (often).

These investigators explain that despite calls for greater availability of integrating behavioral health care services for a pediatric population with their primary care services, data demonstrating current resources are insufficient for these patients, and considering previous research on the effectiveness of integrated behavioral health care,2-5 indirect evidence is still needed “that supports the “minimal, sequential clinical assumptions that must be verified using empirical evidence.”1

Overall, 368 children had behavioral health encounters and were matched to a control cohort of 528 children. Of this group, 58.4% were female patients, their mean (SD) age at the baseline visit was 11.7 (3.5) years, most were Hispanic (43.8%) or non-Hispanic White (27.7%), the primary language was English (51.9%), and the top 5 health concerns were depression (32.9%); anxiety (31.0%); hyperactivity, inattention, or disruptive behavior (29.6%); parent or caregiver mental health concern (17.9%); and emergency services (15.5%). Food was the most common health-related social need in 7.9%.

Just 2.4% reported a previous encounter with a CHW and 1.6%, prior use of psychotropic medication.

The children with behavioral health encounters were older than the control cohort (11.7 vs 10.9 years), more likely to speak Spanish as a primary language (29.9% vs 19.5%), and more likely to have behavioral health concerns identified by a health care practitioner (32.9% vs 14.0%, depression; 31.0% vs 12.9%, anxiety).

There were 3 regression analyses in regard to changes measured by the PSC-17, and each compared results between children who did and did not receive the treatment, respectively. Baseline PSC-17 scores were 2.06 points (95% CI, 1.03-3.09) higher for at least 1 BHC encounter, 1.46 points (95% CI, 0.29-2.63) higher for at least 1 CHW encounter, and 4.06 points (95% CI, 2.76-5.36) higher with at least 1 medication prescription. Improvements by way of a 1.51 (95% CI, −2.65 to −0.37)–point reduction from least 1 BHC encounter and a 2.21 (95% CI, −3.89 to −0.54)–point reduction with prescription receipt were seen in the treated group. The point reduction from a CHW encounter was 0.53 points (95% CI, −1.86 to 0.80).

Children in the control group did not see any significant changes in their PSC-17 scores.

The study authors also considered total BHC and CHW encounters. Having at least 2 BHC encounters was linked to a PSC-17 improvement of 2.17 points (95% CI, –4.03 to –0.31) and 3 or more encounters, 1.70 points (95% CI, –3.03 to –0.37). Again, the results for the control group were not statistically significant.

Externalizing scores also improved in the treated group of patients vs the controls: by 0.77 points (95% CI, –1.26 to –0.28) for at least 1 BHC encounter and by 0.92 points (95% CI, –1.63 to –0.03) for receipt of medication, with the authors noting, “Descriptive analyses suggest that symptom improvements generally aligned with the type of presenting concern and treatment received.”

The authors explain that their findings back efforts of integrating pediatric behavioral health care at federally qualified health centers and that there is a dose effect, too, for BHC encounters. Their study is also the first to use a nonrandomized trial to evaluate the impact of integrated pediatric behavioral health care services on behavioral health symptoms. Additional study strengths are that their findings on the TEAM UP model echo previous research on the effectiveness of psychotherapy and psychotropic medications and they focused on specific elements of behavioral health care.

However, there are also limitations. The electronic medical records did not include information on outside health care received, results could be biased from regression to the mean, clinical significance was difficult to estimate because the PSC-17 does not have a minimal clinically important difference, and results were included on all children identified as having behavioral health concerns, not just elevated screening scores.