Impact of automated dispensing cabinets on medication administration efficiency in the emergency department

Timely medication administration in emergency departments (EDs) is essential for optimal patient outcomes. Automated Dispensing Cabinets (ADCs) may improve point-of-care medication access, yet their impact on administration timeliness is uncertain in ED settings.

To evaluate the association between ADC implementation and the timeliness of stat medication administration in a high-volume ED in Taiwan.

We conducted a retrospective observational study analyzing 16,450 adult ED visits from July-August 2019 (pre-ADC) and July-August 2020 (post-ADC). Order-to-administration (OTA) times for stat medications were compared. Multivariable logistic and linear regression analyses evaluated the association between ADC implementation and timely medication delivery. Sensitivity analyses, including a difference-in-differences model and survival analysis using Kaplan-Meier curves and multivariable Cox regression, were performed to evaluate the robustness of findings.

After adjusting for potential confounders, ADC implementation was independently associated with higher odds of achieving OTA times ≤ 30 min (adjusted odds ratio [aOR], 1.59; 95% confidence interval [CI], 1.42-1.78). Linear regression showed a mean reduction in OTA time of 3.09 min per prescription (β = - 3.09; 95% CI, - 3.58 to - 2.60). Subgroup analyses showed that ADC implementation was associated with higher odds of achieving OTA ≤ 30 min and shorter OTA times for antibiotics (aOR, 2.19; 95% CI, 1.54-3.12; β = - 4.99; 95% CI, - 7.03 to - 2.94), as well as during daytime (aOR, 1.87; 95% CI, 1.58-2.21; β = - 4.34; 95% CI, - 5.26 to - 3.41) and weekend shifts (aOR, 1.66; 95% CI, 1.34-2.05; β = - 3.48; 95% CI, - 4.30 to - 2.65). Findings from sensitivity analyses were consistent with the primary results.

ADC implementation In the ED was associated with modest but operationally meaningful improvements in medication administration timeliness. These findings may be context-dependent and warrant confirmation across diverse workflows and with patient-centered and safety outcomes.

Not applicable.