Variations in Mental Illness treatment in US Adolescents

 The sources indicate that there is a great deal of small-area variation in the treatment American adolescent patients receive following an initial mental illness claim, both in terms of receiving follow-up care and the type of care provided.

Key areas of variation documented across zip codes (small areas) include:

1. Follow-up Treatment Rate: The percentage of children who receive any follow-up treatment within three months varies widely. While only 70.8% receive follow-up treatment overall, this rate ranges from 50% to over 90% depending on the zip code.
2. Treatment Modality: There is substantial variation in the specific type of care received.
   • The fraction of children receiving therapy alone varies from 17% to 62%.
   • The fraction receiving both therapy and drugs varies from 0% to 17%.
   • The fraction receiving drugs only treatment—which runs counter to guidelines that often recommend combination treatment—varies from 0% to 45% across areas.
3. Use of Questionable Drugs ("Red Flags"): The variation in the prescription of drugs that raise a "red flag" is particularly wide. Overall, 45.15% of children who receive drug treatment receive benzodiazepines, tricyclic antidepressants (TCAs), or a drug not Food and Drug Administration (FDA)-approved for their age.
   • The rate of children receiving a red-flag drug treatment ranges from 0% to 100% across zip codes.
   • Specifically, rates of receiving benzodiazepines or non-FDA-approved drugs vary from 0% to 50%, while the rate of receiving TCAs varies from 0% to 33%.

Drivers of Variation:

While shortages in the supply of mental health professionals may impact treatment choices, the sources conclude that very little of the overall variation in these treatment outcomes can be explained by supply-side variables. Furthermore, much of the variation in treatment is occurring within small areas rather than across them. These findings suggest that other factors, such as physician practice style and the "idiosyncratic beliefs of physicians," may be important determinants of the types of treatment children receive.

The sources confirm that the availability of mental health professionals significantly impacts the type of treatment children receive, even though these supply factors account for little of the overall variation in treatment outcomes across zip codes.

Specific findings from the regression models on supply-side variables include:

1. Psychiatrist Supply: An increase in the supply of psychiatrists per 1,000 BCBS children is associated with undesirable treatment choices. As the psychiatrist supply rises, the probability of drug-only treatment and the use of red-flag drugs rise. Conversely, the probability of receiving therapy (alone or in conjunction with drugs) falls sharply.
2. Therapist Supply: A greater number of therapists per 1,000 BCBS children is strongly associated with guideline-concordant care. Increased therapist availability increases the probability of therapy (alone or with drugs) and decreases the probability of drugs-only treatment and red-flag treatments. Moving from the lowest to the highest quartile of therapist supply implies a nearly 25% decline in the use of drugs alone.

Despite these statistically and economically significant effects, comparing regression models shows that adding supply-side variables (psychiatrists and therapists) adds little explanatory power (low R²) once child-level characteristics are accounted for. This suggests that much of the variation observed occurs within zip codes rather than across them, pointing to factors other than shortages.

The sources provide strong evidence suggesting that physician practice style and idiosyncratic beliefs are crucial drivers of the wide variation observed in adolescent mental illness treatment, particularly because supply-side shortages explain so little of the overall variation.

Evidence for Practice Style as the Main Driver:

1. Within-Area Variation: The analysis shows that much of the variation in treatment occurs within zip codes, rather than across them. Adding fixed effects for zip codes (which control for fixed area characteristics like average income or urban status) explains at most half of the observed variation in treatment outcomes. This means characteristics specific to the local geographic area do not fully account for the observed differences, suggesting provider-level choices are paramount.
2. Provider Beliefs and Knowledge: The existence of these vast, unexplained variations aligns with observations cited in the literature regarding the influence of the "idiosyncratic beliefs of physicians [and] the parochial character of much clinical practice". Provider knowledge and preferences for different treatments are considered important determinants. Previous research suggests psychiatrists may have "favorite drugs" for most conditions, and their practice styles exhibit cohort-effects.
3. Implausibility of Patient Demand: While differences in patient demand (such as parental preferences or attitudes) might explain some variation in getting follow-up care, the sources find it less plausible that parental demand could be the main driver for variation in the prescription of specific, problematic treatments. It is unlikely that large numbers of parents are demanding that their children be initially treated with non-FDA approved drugs, benzodiazepines, or tricyclic antidepressants (TCAs) (the "red-flag" drugs).

The sources conclude that further exploration of individual clinician treatment patterns could illuminate the variations observed for children with emerging mental illness.

The study, upon documenting vast variations and poor adherence to evidence-based practices, concludes that its findings leave open the question of whether existing guidelines from professional associations are adequate to protect the interests of these vulnerable children.

While the study does not definitively answer this question, the data collected raise significant concerns about the practical implementation of these guidelines in the U.S..

Specific Guideline Concerns Highlighted by the Data:

• Follow-up Care: Guidelines emphasize prompt follow-up treatment, yet 29.4% of children received no treatment within the three months following an initial claim, and this rate varied dramatically by zip code.
• Combination Treatment: Current guidelines generally suggest that children should receive therapy either with or without drug treatment. However, only 5.9% of children received both therapy and drugs, and the fraction receiving only drug treatment varied widely, from 0% to 45% across areas.
• "Red Flag" Drugs: Guidelines strongly caution against certain drugs for initial treatment due to safety and efficacy concerns in adolescents. Yet, 45.15% of children receiving drug treatment received benzodiazepines, tricyclic antidepressants (TCAs), or drugs not FDA-approved for their age. The rate of receiving such treatments varied from 0% to 100% across zip codes.

The study implies that if existing guidelines were adequate and uniformly applied, such massive and medically questionable variations driven by physician practice style would likely not occur.