Modeling Phase I Postanesthesia Care Unit Patient Transport Times Among Multiple Destinations

Paul Cover; Franklin Dexter; Yasser Ms El-Hattab

doi:10.1213/ANE.0000000000007455

Although the importance of transporter availability to the operating room and phase I postanesthesia care unit workflow has been known since Marcon et al Anesth & Analg 2003, no detailed data have been published about patient transport times. Nearly all facilities know the times of postanesthesia care unit (PACU) discharges, but transport time through when the transporter returns consumes porter resources.BACKGROUNDAlthough the importance of transporter availability to the operating room and phase I postanesthesia care unit workflow has been known since Marcon et al Anesth & Analg 2003, no detailed data have been published about patient transport times. Nearly all facilities know the times of postanesthesia care unit (PACU) discharges, but transport time through when the transporter returns consumes porter resources.Epic's Rover was implemented to track all PACU transports, 22,846 from July 2022 through April 2024. Destinations were grouped based on distance traveled. Transport counts and total times were calculated by 4-hour period and day of the week. χ2 and Kruskal-Wallis tests were used for inference. Number of staff starting shifts, transports per transporter per hour, and transport time per transporter per 4-hour period were determined, with χ2 and Kruskal-Wallis tests for inference. Bland-Altman plots compared (i) total transport time for each day and 4-hour period with (ii) estimates from the product of PACU discharges and the overall mean time per transport. Delayed transports were identified and evaluated for successive delays. Standardized normal probability plots and Shapiro-Wilk tests of normality were used to examine the probability distribution of total transport time among workday for 4-hour periods.METHODSEpic's Rover was implemented to track all PACU transports, 22,846 from July 2022 through April 2024. Destinations were grouped based on distance traveled. Transport counts and total times were calculated by 4-hour period and day of the week. χ2 and Kruskal-Wallis tests were used for inference. Number of staff starting shifts, transports per transporter per hour, and transport time per transporter per 4-hour period were determined, with χ2 and Kruskal-Wallis tests for inference. Bland-Altman plots compared (i) total transport time for each day and 4-hour period with (ii) estimates from the product of PACU discharges and the overall mean time per transport. Delayed transports were identified and evaluated for successive delays. Standardized normal probability plots and Shapiro-Wilk tests of normality were used to examine the probability distribution of total transport time among workday for 4-hour periods.Transports to the phase II recovery were faster than to other common transport destinations (all adjusted P ≤ .0001). Number of transports and total hours of transport time differed among 4-hour periods and destination categories (all P < .0001). Weekday was inconsequential. Approximately half of all 4-hour periods had ≥1 delay, and runs of delays were significant (P < .0001). Prediction of transport workload using the count of transports for each combination of day and 4-hour period multiplied by the overall mean time per transport proved insufficient, differing from actual time by -1.04 to +1.05 hours per 4-hour period. Total hours of transport time per 4-hour period were normally distributed for the busiest 2 periods (Shapiro-Wilk W > 0.99), allowing the mean plus the standard deviation to be used to choose the number of transporters needed for each period.RESULTSTransports to the phase II recovery were faster than to other common transport destinations (all adjusted P ≤ .0001). Number of transports and total hours of transport time differed among 4-hour periods and destination categories (all P < .0001). Weekday was inconsequential. Approximately half of all 4-hour periods had ≥1 delay, and runs of delays were significant (P < .0001). Prediction of transport workload using the count of transports for each combination of day and 4-hour period multiplied by the overall mean time per transport proved insufficient, differing from actual time by -1.04 to +1.05 hours per 4-hour period. Total hours of transport time per 4-hour period were normally distributed for the busiest 2 periods (Shapiro-Wilk W > 0.99), allowing the mean plus the standard deviation to be used to choose the number of transporters needed for each period.PACU transport times differed significantly among destinations. Therefore, when >1 patient is waiting for transport, target the patient who will have the briefest transport, especially if to the phase II PACU. Approximation of transport workload based on the distribution of PACU discharges by time of day is insufficient for accurate staffing. Instead, rely on the mean and standard deviation of the workday's total transport times during the period.CONCLUSIONSPACU transport times differed significantly among destinations. Therefore, when >1 patient is waiting for transport, target the patient who will have the briefest transport, especially if to the phase II PACU. Approximation of transport workload based on the distribution of PACU discharges by time of day is insufficient for accurate staffing. Instead, rely on the mean and standard deviation of the workday's total transport times during the period.

Modeling Phase I Postanesthesia Care Unit Patient Transport Times Among Multiple Destinations

View Online

Abstract

Details

Metrics