Use of statistical process control for surveillance of pulmonary dysfunction in groups in the workplace

Workplace surveillance is an essential feature of an effective occupational health program. Unfortunately, many health care organizations are beginning to collect large quantities of clinical information without much thought to subsequent application. In this paper, we propose a screening technique to help manage this situation. Specifically, we advocate "control charts for fraction non-conforming" for use in the medical surveillance of work areas. A military installation in the mid-western United States with 63 work areas is analyzed using up to seven years worth of spirometry data. Based on the results of 6 separate tests, a classification of normal or abnormal was made for each individual and the percent of abnormalities by area and year was calculated. The results were analyzed via control charts and contrasted to a preset percentage method. Cigarette smoking was then controlled for in the analysis to account for abnormalities that may be occupationally related versus personal habits. We demonstrate the utility of control charts for the compact display of surveillance data and show how it can aid in the analysis of an extremely complex health care concern. A recommendation is made for managing a surveillance program using spirometry and control charts and supported by follow-on investigation of two areas flagged as abnormal by the procedure.