BACKGROUND

Sleep Disturbances (SDs) are a symptom common to mental health disorders (MHD) and substance use disorders (SUD). We aimed to identify the value of SD as a predictor for subsequent treatment of illicit drug and alcohol use disorders (SUDs) in primary care and relative to the predictive value of mental health disorders (MHDs).

METHODS

We used electronic health records data from ambulatory primary care in a safety net Boston area healthcare system from 2013 to 2015 (n = 83,920). SUD (separated into illicit drug use disorder and alcohol use disorder) and MHD were identified through ICD-9 codes and medical record documentation. We estimated Cox proportional hazard models to examine the risk of SUD across four comparison groups (SD only, SD and MHD, MHD only, and neither SD nor MHD).

RESULTS

Compared to patients with no sleep or MHD, patients with SD had a greater risk for subsequent SUD treatment. Approximately one-fifth of patients with SD were treated for an illicit drug use disorder and approximately 12% were treated for alcohol use disorder. Risk for SUD treatment, estimated at over 30% by the end of the study, was greatest for patients with a MHD, either alone or comorbid with SD. Risk was greater for older patients and men, and lower for minority patients.

CONCLUSIONS

SD and MHD, individually and comorbid, significantly predict subsequent treatment of illicit drug and alcohol use disorder in primary care. Screening and evaluation for SD should be a routine practice in primary care to help with identifying potential SUD risk.