Prevalence and predictors of patient-directed discharge in patients with opioid use disorder.

BACKGROUND

Patients with opioid use disorder (pOUD) are at risk of patient-directed discharge (PDD). We hypothesize that multiple risk factors (RF) contribute to this risk.

METHODS

PDD prevalence at an urban community hospital was assessed using aggregate data between January 2022-December 2023. Retrospective chart review was conducted of pOUD with PDD (n = 362) over those two years and pOUD discharged conventionally (n = 108) between January 2023-March 2023. Nested multivariable logistic regression models were fit for potential predictors of PDD.

RESULTS

In patients without OUD, PDD prevalence was 6.8% compared with 23.8% in pOUD (OR 4.28, [95% CI 3.75, 4.87], p < 0.0001). RFs for PDD were cocaine use disorder (aOR 1.62 [95% CI 1.00, 2.65], p = 0.05) and non-opioid-based withdrawal treatment (aOR 2.44 [95% CI 1.30, 4.57], p = 0.005). Age (aOR 0.98 [95% CI 0.96, 0.99], p = 0.01), being domiciled (aOR 0.41 [95% CI 0.22, 0.74], p = 0.007), and methadone maintenance program enrollment (MMT) (aOR 0.54 [95% CI 0.31, 0.93], p = 0.03) protected against PDD. Non-confirmation of MMT enrollment was a RF for PDD (p < 0.0001).

CONCLUSIONS

This study corroborates elevated risk of PDD in pOUD and provides novel insight into RFs for PDD, reinforcing the importance of identifying OUD and comorbid substance use, confirming MMT enrollment, and managing withdrawal. Future directions include education on withdrawal screening and provision of opioid-based treatment.