More information about text formats
Riveros Perez et al. analysed 636 turnover times from a surgical suite with 16 theatres before and after intervention, a dedicated nurse anaesthetist for each of four theatres . Their “overall” statistical analysis, reported in their abstract, seems to treat all turnovers as statistically independent events (i.e., treated the sample size as 16 x # analysed days). That probably was incorrect because the 636 turnover times likely were correlated among theatres on the same day . The authors’ Wilcoxon-Mann-Whitney overall P-value (0.0121) likely is an underestimate of the correct result .
To understand, consider that the authors’ intervention was one extra nurse anaesthetist for four theatres. If there were more than one turnover among the four theatres simultaneously, the nurse anaesthetist’s efforts would be diluted. That is precisely what happens routinely (e.g., for housekeeping staff [3,4]). Earlier, we showed validity and reliability of choosing the optimal number of shared personnel (e.g., nurse anaesthetist) by analysing those simultaneous turnovers .
The authors state in their paper that the turnover times were skewed. Analysis methods tested by Monte-Carlo simulation for accurate P-values and confidence intervals are to take the mean of the turnovers among the 16 theatres, for each day, although the median could be used . There then is one number per day as a summary measure. By central limit theorem (and in practice [2,5]), those means g...
The authors state in their paper that the turnover times were skewed. Analysis methods tested by Monte-Carlo simulation for accurate P-values and confidence intervals are to take the mean of the turnovers among the 16 theatres, for each day, although the median could be used . There then is one number per day as a summary measure. By central limit theorem (and in practice [2,5]), those means generally are normally distributed within groups. Compare among days, before and after the intervention (i.e., between the two groups), either using Wilcoxon-Mann-Whitney or Student’s two-group t-test.
What are the authors’ mean (SD) or median (IQR) among days, before/after intervention, when analysed using the mean or median among theatres on each day? What is the P-value testing the authors’ primary result? If the authors use Student’s t-test, they can report the corresponding confidence interval for the mean difference.
1. Riveros Perez E, Kerko R, Lever N, White A, Kahf S, Avella-Molano B. Operating room relay strategy for turnover time improvement: a quality improvement project. BMJ Open Qual 2022; 11:e001957.
2. Dexter F, Epstein RH, Marcon E, Ledolter J. Estimating the incidence of prolonged turnover times and delays by time of day. Anesthesiology 2005; 102:1242-1248.
3. Dexter F, Marcon E, Aker J, Epstein RH. Numbers of simultaneous turnovers calculated from anesthesia or operating room information management system data. Anesth Analg 2009; 109:900-905.
4. Wang J, Dexter F, Yang K. A behavioral study of daily mean turnover times and first case of the day start tardiness. Anesth Analg 2013; 116:1333-1141.
5. Austin TM, Lam HV, Shin NS, Daily BJ, Dunn PF, Sandberg WS. Elective change of surgeon during the OR day has an operationally negligible impact on turnover time. J Clin Anesth 2014; 26:343-349.