Healthcare resource utilization of people with depressive disorder and co-occurring incident diabetes mellitus: A 20-year population-based propensity score-matched cohort study.

Background: Depression with co-occurring diabetes is associated with elevated risk of diabetes complications, driving substantial healthcare expenditure. However, previous research investigating healthcare utilization associated with depression in patients with diabetes is hampered by important limitations. We examined whether pre-existing depression increased healthcare utilization and costs in patients with incident diabetes.

Method: This population-based propensity score (PS)-matched (1:10) cohort study identified 12,165 patients with pre-existing depression and incident-diabetes (depression-diabetes group) and 116,853 patients with incident-diabetes only (diabetes-only group) between 2002 and 2021 in Hong Kong, using territory-wide medical-record database of public healthcare services. A comprehensive array of covariates was included in PS-matching, comprising age, sex, calendar-year period, catchment-area, physical and psychiatric comorbidities, and medication prescriptions. Service-specific utilization per patient-year was compared by negative binomial regression between groups, with stratified analyses by service types, sex, and age groups.

Results: Depression-diabetes group was significantly associated with 38 % elevated healthcare utilization relative to diabetes-only group across emergency (Odds ratio 1.16, 95 %CI[1.12-1.20]; HK$1906[US$243] versus HK$1648[US$210]), inpatient (1.47[1.27-1.72]; HK$17,333[US$2208] versus HK$12,356[US$1574]), and outpatient services (1.29[1.27-1.31]; HK$11,974[US$1526] versus HK$8670[US$1105]), leading to additional HK$8540 annual healthcare costs per-patient. Additional analyses demonstrated that increased healthcare utilization was observed in both men and women with depression and co-occurring diabetes relative to their diabetes-only counterparts, across three age-groups, and in both psychiatric and non-psychiatric resources.

Conclusion: Patients with pre-existing depression and co-occurring incident-diabetes are at heightened risk for increased healthcare resource utilization and cost compared to diabetes-only patients. Further research on effective interventions is warranted to alleviate the burden associated with both conditions.