Jenkins PL, Knight M, Hirabayashi L, Weil R, Echard J, Krupa N, Sorensen JA. Am. J. Ind. Med. 2019; ePub(ePub): ePub.
Affiliation
The Northeast Center for Occupational Health and Safety: Agriculture, Forestry and Fishing, Bassett Healthcare Network Research Institute, Cooperstown, New York.
BACKGROUND: Falls overboard are the most common cause of fatalities in commercial fishing. As a result, interventions aimed at increasing and measuring the use of personal flotation devices (PFDs) are a high priority. The focus of this study was to explore the use of accelerometers as a means for objectively measuring PFD use on lobster fishing vessels.
METHODS: For participating vessels, researchers attached an accelerometer to a PFD worn by a crewmember and another to the vessel's wheelhouse. GoPro videos were also employed to record crewmember activities so these could be synchronized with accelerometer outputs. Accelerometer outputs included two distinct measures, the proportional integration mode (PIM) and zero crossing mode (ZCM). Data were fitted to various equations to identify the best method for predicting PFD use.
RESULTS: Seven lobster fishing vessels participated in the trial. Data indicated that accelerometers could predict PFD use with a fairly high degree of accuracy. In particular, a logistic equation incorporating PIM values from the PFD and the absolute value of the difference between the PFD PIM readings and the PIM readings from the stationary accelerometer demonstrated the highest degree of accuracy, with correct classifications for 73.3% to 77.6% of the 10-second data intervals. Accuracy was highest when crew members were moving versus stationary. The predictive value of ZCM was comparatively limited.
CONCLUSIONS: PIM accelerometer readings can be used to measure PFD use with a considerably high degree of accuracy, especially for sternmen who are moving regularly and have the highest risk for falling overboard.