A retrospective cohort study examining health care utilization patterns in individuals diagnosed with an eating disorder in childhood and/or adolescence

Abstract

Objective: This study examined a 2-year period following an eating disorder (ED) diagnosis in order to determine patterns of health care utilization. Method:We conducted a retrospective cohort study of children (n = 1560) diagnosed with an ED between 2000 and 2017. The ED diagnosis was made at a tertiary level hospital for children and adolescents presenting for outpatient assessment by specialist adolescent medicine physicians and recorded in a program database over this period of time. We then created three sex-and age-matched comparison cohorts using provincial health administrative databases including: a general population cohort, a diabetes cohort (to compare nonmental health care utilization) and a mood disorder cohort (to compare mental health care utilization). Outcomes included hospitalizations, emergency department visits, as well as general practitioner, psychiatrist, and pediatrician visits. Odds ratios (dichotomous outcomes) and rate ratios (continuous outcomes) were calculated.Results: Compared to the general population cohort, the ED cohort had higher odds and rates of all types of health care utilization. Compared to the diabetes cohort, the ED cohort had higher odds of nonmental health-related admissions (OR 1.45, 95% CI 1.09-1.95) and higher rates of nonmental health-related emergency department visits (RR 1.59, 95% CI 1.18-2.13). Compared to the mood disorder cohort, the ED cohort had higher rates of pediatrician visits, which were mental health-related (RR 14.88, 95% CI 10.64-20.82), however most other types of mental health service utilization were lower.Discussion: These patterns indicate that the service needs of young people diagnosed with EDs are higher than those with diabetes with respect to nonmental health admissions and emergency department visits, while in terms of mental health service utilization, there appears to be a lack of use of mental health services compared to youth with mood disorders with the exception of pediatrician mental health visits. These findings must be interpreted in the context of under-detection and undertreatment of EDs.

A diagnostic subgroup comparison of health care utilization patterns in individuals with eating disorders diagnosed in childhood and/or adolescence

“…These subgroups were matched to the general population cohort in a 1:5 match ratio in order to compare health care utilization. This is a reanalysis of our data reported in a recent paper examining health care utilization in our eating disorder cohort using similar methodology (Couturier et al, 2022) however, the current paper examines diagnostic subgroups. After comparing to the general population, a direct comparison of the ARFID subgroup to the other subgroups was made.…”

“…Healthcare utilization studies in the field of eating disorders are uncommon and fraught with challenges of under‐detection and under‐treatment (Simon et al, 2005). Our recently published retrospective cohort study examined a group of 1560 children and adolescents diagnosed with an eating disorder and found that in the 2 years following diagnosis the eating disorder cohort had higher odds and rates of all types of health care utilization compared to the general population (Couturier et al, 2022). We also compared our cohort to a diabetes cohort for non‐mental health‐related utilization and a mood disorder cohort for mental health‐related utilization.…”

“…We also compared our cohort to a diabetes cohort for non‐mental health‐related utilization and a mood disorder cohort for mental health‐related utilization. Compared to the cohort with diabetes, the eating disorder cohort had higher odds of non‐mental health‐related admissions (Couturier et al, 2022). Compared to the mood disorder cohort, the eating disorder cohort had higher rates of pediatrician visits which were mental health‐related, however most other types of mental health service utilization were lower (Couturier et al, 2022).…”