Objectives: To compare alternative methods of interpreting the response of finger skin temperature (FST) to cold provocation for the detection of the abnormal cold response observed in vibration-induced white finger (VWF).
Method: The FST response to cold provocation was measured in 36 male subjects: 12 office workers, 12 manual workers and 12 manual workers with symptoms of VWF. The FSTs were monitored continuously on the distal phalanges of all five fingers of a test hand for 2 min before, for 5 min during, and for 10 min following, immersion of the test hand in water at 15 degrees C. Of the fingers investigated, 147 were reported not to exhibit blanching and 33 were reported to exhibit blanching. Twenty-one alternative methods of interpreting the response of FSTs to cold provocation were assessed. These were grouped as: (1) areas above the response profile (i.e. the area above the curve showing the FSTs as a function of time during cooling and recovery), (2) areas below the response profile, (3) absolute temperatures during and following cold provocation, (4) percentage differences in FSTs, (5) the times taken for FSTs to rise by specified amounts and (6) rates of change of FSTs. Differences in the response to cooling between those fingers reported to blanch and the fingers not reported to blanch were tested, and receiver operating characteristics (ROCs) were used to compare the sensitivity and specificity of the various measures to symptoms of VWF.
Results: The areas above the response profile, areas below the response profile, percentage FSTs, absolute FSTs and rates of change of FSTs tended to discriminate between healthy and unhealthy subjects on a group basis. However, some of these methods of interpreting the FST response to cold provocation did not show a high sensitivity or specificity to vascular dysfunction on individual fingers. The area above the response profile, the percentage of initial temperature at the fifth minute of recovery and the maximum temperature during the 10-min recovery period, were found to show the highest sensitivity and specificity to symptoms of vascular dysfunction.
Conclusions: The method chosen to interpret the FST response to cold provocation affects the ability of the test to detect an abnormal cold response. The area above the response profile, the percentage of initial temperature at the fifth minute of recovery and the maximum temperature achieved during a 10-min recovery period appear to be the most suitable measures for monitoring vascular function in workers exposed to hand-transmitted vibration. It is suggested that the FST response to cold provocation should be interpreted with respect to the state of initial blood flow.