Factor analysis ofclinical datafromasbestos workers: implications fordiagnosis andscreening

Clinical datafrom624asbestos exposed workers were analysed by factor analysis. Fifteen clinical variables were foundtoreflect several underlying factors: "obstruction," "interstitial disorder," "x raychange," "air trapping," and"age/exposure." Scoring individual subjects along these axescanfacilitate identifying early casesandavoiding false diagnosis. Theanalysis suggests thatscreening shouldbemultimodal andthatradiographic abnormality neednot imply physiological impairment. Accurate diagnosis of asbestos related disease depends ontheintegration ofmuchclinical data. Whereas thelegal system mayoccasionally recognise asingle test-radiography forexample-as specific andsensitive, knowledgeable clinicians recognise the importance ofobtaining several different types of dataineachsubject. Thepulmonary system iscom- plex, having multiple components (large andsmall airways, interstitial tissue, vasculature, pleural tissue, andrespiratory muscle), andclinical information and laboratory tests relate tospecific pathological and underlying physiological abnormalities onlyindi- rectly. Thusclinical data(individual spirometry test results, symptoms, orradiographs) maybeseenas indirect reflections oftheunderlying process (or factors). Factor analysis isatechnique that maybeusedto explore theinterrelations amongobserved datato assess underlying factors thatmaynotbedirectly observable.' 2Thistechnique wasusedtoanalyse datafromagroupofheavily asbestos exposed sub- jects toassess thenature ofunderlying factors. The results suggest thepresence ofseveral different pro- cesses thatrequire different clinical andsurveillance approaches. Methods Results ofclinical evaluations of624asbestos exposed workers wereanalysed. Allhadbeenexa- minedattheoccupational medicine branchand pulmonary division oftheUCLA Medical Center