A Bootstrap Approach for Predicting Methoxyflurane Occupational Exposure in Paramedicine

Background - Methoxyflurane was used as an anaesthetic agent from 1958 until it was withdrawn around 1974 after dose-related patient nephrotoxicity was identified. It is now available only for administration via the Medical Developments International Penthrox Inhaler device for analgesia. When administering methoxyflurane, ambulance officers will be exposed to some methoxyflurane vapour. A previous study has extrapolated data from anaesthetized patients to suggest that such occupational exposure is within safe limits. There is a need to explore the robustness of these thresholds. Approach - A model is created to investigate the range of possible regression lines from repeated bootstrap samples of the same patient data, and to describe the probability distribution of those regression lines. Findings - The model shows a wide range of possible extrapolations due to the limited patient data set and the extrapolation being conducted over 2 - 5 orders of magnitude of exposure. With the range of ambulance officer exposures reported elsewhere, 95.7% of these regressions fall within an identified safe limit, suggesting that one-off exposures of this nature are safe. This model does not account for repeated exposures over days or weeks as would be seen in occupational settings. Originality - Bootstrapping methods are applied to test the statistical robustness of extrapolation. Results indicate that ambulance crews could be safe if exposed between the limits currently in place.