Screening for psychiatric morbidity. The pattern of psychological illness and premorbid characteristics in four chronic pain populations.

Three hundred and seventy-eight patients from 4 chronic pain populations have been examined by self-assessment questionnaire methods to estimate the amount of psychiatric morbidity present. Using the General Health Questionnaire-28 scaled version (GHQ-28) the findings for probable psychiatric illness by clinic were: for anaesthetists' pain clinics serving a mixed urban and rural population--37%; for an oral medicine facial pain clinic--30%; for a rural hospital pain clinic--37%; and for a psychiatrists' pain assessment and treatment service--51%. These findings demonstrate the effects of selection upon the psychiatric characteristics of different pain populations. On the subscales of the General Health Questionnaire the psychiatric clinic patients were significantly more depressed than those in the other 3 groups (P less than 0.001) and also showed more social dysfunction (P less than 0.001). On the Irritability/Depression and Anxiety Questionnaire (IDA), depression and inward irritability were higher in the psychiatric clinic patients (P less than 0.001) but the amount of anxiety did not differ by clinic or by diagnosis (P greater than 0.05). These findings are taken to indicate that the extent of somatic complaints and anxiety does not differentiate the majority of pain patients in pain services with psychological illness from others attending for treatment. However, in the patients who have definite psychological symptoms, depression, social dysfunction and irritability provide a characteristic pattern. The psychiatric clinic patients were demonstrably more introverted or obessional on the Hysteroid/Obsessoid Questionnaire (HOQ) than those in other clinics. Childhood experience as seen by the patients did not differ by clinic or diagnosis and did not correlate significantly with personality as measured by the HOQ. It did correlate very significantly with measures of the current mood represented by the IDA. This effect was relatively weak, permitting the inference that the major portion of those psychological abnormalities which were found to be present was related to other factors such as the occurrence of painful lesions.