A National Examination of Child Psychiatric Telephone Consultation Programs’ Impact on Children’s Mental Health Care Utilization

To the Editor:

Estimates are that half of children with mental health problems do not receive needed treatment.¹ One of the barriers they face is that there are not enough child mental health specialty providers to meet their needs.² Pediatric providers, who might partially fill this gap, often feel they lack the training to treat children’s mental health disorders without consulting a specialist.³

In response, the majority of states have established child psychiatric telephone consultation programs. Modeled after the Massachusetts Child Psychiatry Access Project implemented in 2004, these programs typically provide a dedicated hotline for primary care clinicians to consult immediately with a child psychiatrist, to arrange expedited face-to-face psychiatric consultation if required, and to get help with referrals to community mental health services.^4,5 Evaluations of such programs have found that they are widely used, have increased pediatric providers’ reported ability to meet their patients’ mental health needs, and have enhanced clinicians’ perceived ability to deliver mental health care consistent with families’ preferences.^4–6 A Washington State study found an increase in treatment among a subset of Medicaid-enrolled children.⁶ However, no multistate studies have examined the effectiveness of such programs.

To address this gap, we used multiple waves of the National Survey of Children’s Health (NSCH), supplemented with information about the establishment of telephone consultation programs to examine how such programs affect children’s mental health service utilization. We used weighted information on children ages 5 to 17 years old from the 2003, 2007, 2011, and 2016 NSCH to identify receipt of mental health services, and merged the NSCH data by year and state with information about state child psychiatric telephone access programs from the National Network of Child Psychiatry Access Programs. We adopted a difference-in-differences approach to take advantage of the natural experiment created by the child psychiatric telephone access programs implemented by different states in different years, estimating a multivariate logistic model to examine the impact of such programs on the likelihood of a child receiving any mental health care, controlling for child sociodemographic characteristics, insurance status, parental education, and NSCH year.

We observed a total of 245,512 children 5 to 17 years old. The percentage receiving mental health treatment was 8.4% in 2003, 9.5% in 2007, 11.1% in 2011, and 11.4% in 2016. Because the first child psychiatric telephone access program was implemented in Massachusetts in 2004, in 2003 all children lived in states without such programs. There was a growth in programs over time (Figure 1), so that by 2016, 28.1% of children resided in states with a statewide child psychiatric telephone access program, 48.6% of children resided in states with a partial child psychiatric telephone access program (defined as a state in which some but not all of the counties in a state could use the program), and 23.3% of children resided in a state with no such programs. The percentage of children residing in states with statewide child psychiatric telephone access programs who received mental health services (12.3%) was significantly higher than the percentage of children residing in states with partial child psychiatric telephone access program receiving mental health services (10.9%), or in states with no such programs (9.5%) (χ² test, p < .0001).

FIGURE 1.

Adoption of Child Psychiatric Telephone Consultation Programs Over Time

Note: Please note color figures are available online.

In the multivariate regression (Table 1), we found that, compared to 2003, children were more likely to receive mental health services: children in 2007 (adjusted odds ratio [aOR] = 1.16, 95% confidence interval [CI] = 1.08 – 1.26), 2011 (aOR = 1.34, 95% CI = 1.24 – 1.46), and 2016 (aOR = 1.37, 95% CI = 1.25 – 1.49). Male patients and children 12 to 17 years of age were less likely to receive mental health treatment than female patients (aOR = 0.79, 95% confidence interval = 0.74 – 0.84) and children 5 to 11 years (aOR = 0.66, 95% CI = 0.62 – 0.70), respectively. Compared to non-Hispanic white children, non-Hispanic black children (aOR = 0.64, 95% CI = 0.58 – 0.71), Hispanic children (aOR = 0.61, 95% CI = 0.55 – 0.68), and children of other racial/ethnic groups (aOR = 0.73, 95% CI = 0.66 – 0.80) were all less likely to receive mental health services. Uninsured children were less likely to receive mental health services than privately insured children (aOR = 0.73, 95% CI = 0.62 – 0.86). Medicaid-enrolled children were more likely to receive mental health services than privately insured children (aOR = 2.12, 95% CI = 1.98 – 2.28). Children in households who had a parent with more than a high school education were significantly more likely to receive mental health services than children whose parents were high school graduates only (aOR = 1.17, 95% CI = 1.08 – 1.26). Children’s receipt of mental health services did not differ between households in which parents had high school diplomas and households in which parents did not have diplomas.

TABLE 1.

Factors Associated With Receiving Any Mental Health Treatment

Effect	N	OR	95% CI
State presence of a child psychiatric telephone consultation program
No child psychiatric telephone consultation program	202,122	Ref
Partial child psychiatric telephone consultation program	20,039	0.986	0.884	1.10
Statewide child psychiatric telephone consultation program	23,351	1.093	1.002	1.193
Sex
Male	126,959	0.788	0.742	0.836
Female	118,553	Ref
Age, y
12–17	126,297	Ref
5–11	119,215	0.66	0.621	0.701
Race/ethnicity
White	169,540	Ref
Black	22,268	0.64	0.579	0.707
Hispanic	28,761	0.612	0.547	0.684
Other	23,746	0.731	0.663	0.806
Unknown	1,197	0.825	0.496	1.372
Insurance
Commercial	176,777	Ref
Medicaid	50,938	2.124	1.98	2.279
Other	2,429	1.104	0.809	1.507
Uninsured	15,368	0.732	0.622	0.861
Highest Family Education Level
Less than 12 years/high school	18,354	1.002	0.874	1.147
12 Years/high school graduate	55,384	Ref
More than 12 years/high school	171,774	1.169	1.084	1.261
Survey Year
2003	73,858	Ref
2007	67,727	1.163	1.077	1.256
2011	66,806	1.344	1.236	1.46
2016	37,121	1.367	1.252	1.494

Note: OR = odds ratio; Ref = referent.

Controlling for child and family factors and secular trends, children from states with statewide child psychiatric telephone access programs were significantly more likely to receive mental health services than children residing in states without such programs (aOR = 1.09, 95% CI = 1.00 – 1.19, p < .05). The likelihood of receiving mental health services did not differ for children living in states with partial child psychiatric telephone access programs versus children living in states without such programs.

In this first national study of child psychiatric telephone consultation programs, we found that children in states with statewide programs were significantly more likely to receive mental health services than children in states without such programs. Given that most children who need mental health services do not receive them,¹ the child mental health workforce is inadequate to meet demand,² and many individuals experience a multi-year delay between symptom onset and receipt of services,⁷ it is encouraging that the effect of child psychiatric telephone consultation programs, albeit modest, may help to increase access to child mental health services.

We also found that, over time, more children have been receiving mental health services. Policies such as the Mental Health Parity and Addiction Equity Act of 2008 and the 2010 Affordable Care Act with Medicaid Expansion in 2014 appear to have lowered barriers to receipt of mental health care. Given that untreated mental health disorders in children can have a long-term impact,⁸ the greater use of child mental health services in more recent years is promising.

Many of our other findings, such as lower rates of child mental health service use among racial/ethnic minorities, uninsured individuals, and children living in households with less educated parents are consistent with prior research.^9,10 Unfortunately, differences in access and use of child mental health services among these populations have been consistent for decades. They are also consistent with patterns seen in health care more broadly, reflecting longstanding societal barriers to accessing mental health services related to stigma, discrimination, and low socioeconomic status,¹⁰ which will likely require wide-ranging persistent efforts to address.

Our study had limitations. The NSCH is a parental self-report and provides no information on the type, quality, or effectivess of mental health services being received by the children. The public-use NSCH does not have county identifiers, and we are therefore unable to examine how county characteristics may influence the impact of a child psychiatric telephone consultation program. Because the NSCH is not an annual survey, we could not assess trends in mental health service use before and after implementation of the child psychiatric telephone consultation programs, nor do our data allow us to directly connect services to the timing of the programs. We have only the 2003 data point before the child psychiatric telephone consultation programs first started in Massachussetts in 2004, making it impossible to test whether trends in child mental health services use were parallel between states with and without the child psychiatric telephone consultation programs before the programs started.

To the extent that the telephone consultation programs were implemented as part of broader state efforts to enhance access to child mental health services, our results could overstate the effects of the programs themselves. However, this concern actually speaks to one of the strengths of our difference-in-difference approach, which exploits the natural experiment created by the telephone consultation programs implemented by different states in different years. Essentially, each state becomes its own control, and the main comparison is a before versus after the implementation of a telephone consultation program in the state, which occurs in multiple states but at different times. Because the timing and scope of telephone consultation program implementation differed across states, we can differentiate program effects from larger national trends such as the Domenici–Wellstone federal parity law, which has been implemented nationwide for private insurance since January 2011.

Despite these limitations, our findings suggest that recent federal investments to substantially expand child psychiatric telephone consultation programs could significantly increase the number of children receiving mental health services. The programs likely complement other efforts such as tele-psychiatry programs for children and families with no nearby provider and loan forgiveness designed to increase the number of child mental health providers. Helping families to find mental health care for their children is likely to require the full range of such efforts.

Key Points.

Question:

Was the establishment of a child psychiatric telephone consultation program in a state associated with increased use of child mental health services?

Findings:

This difference-in-difference analysis based on multiple years of national survey data showed that children living in states with a statewide child psychiatric telephone consultation program had significantly greater parentreported child mental health service use than children in states without such programs.

Meaning:

Child psychiatric telephone consultation programs may be an important tool in efforts to increase children’s mental health service use.