Physiogenic Bias as Invalidity in Psychiatric Symptom Scales

Although it has been suggested that physical illness may compromise the results of epidemiological and social surveys that are based on psychiatric symptom checklists, the problem has rarely been investigated. This study presents a methodfor identifying physiogenic bias in psychiatric symptomatology scales and analyzes the consequences of such bias. Using an adaptation of recently developed clinical psychiatric screening procedures, occurrences of symptoms common to Langnerand Macmillan-type scales of psychological distress are classified as either physiogenic or psychogenic. Removing reported symptoms due to organic or physiological causes is shown to halve serious implications for both (I) estimates of prevalence of psychological disorder in community populations and (2) estimates of relations between psychiatric symptomatology and traditional sociocultural correlates of symptom level.