Quality control aspects of expired air carbon monoxide measurement in the Multiple Risk Factor Intervention Trial.

The determination of carbon monoxide levels in the expired air of MRFIT participants was used to monitor the level of cigarette smoking in all participants and to assess the adherence of the SI participants to the MRFIT protocol, which required that they quit or reduce their cigarette smoking. Two problems affected the quality of the data. First, as with the measurements of pulmonary function (Chapter 12), the initial lack of training in adequate maintenance was responsible for inoperative instruments. Centralized training, both in the performance of the test and in equipment maintenance, helped to solve this problem. For future clinical trials, the importance of centralized training and certification of all technicians at the start of the study cannot be overemphasized. The second problem, the instability of the CO calibration gas in steel tanks over long periods, was not anticipated. The solution was to purchase small amounts of freshly prepared calibration gas, or to have such gas available to check the major gas supply. Many of the observed maintenance problems might have been avoided by the use of a checklist to ensure that preventive measures were taken before the equipment became inoperative.