The present study replicates that of De Freitas and Schwartz (1979), using more typical chronic patients (on open wards rather than locked wards), and monitoring coffee intake with serum caffeine levels. The serum caffeine levels observed indicate that caffeine can be effectively manipulated on an open ward by switching the type of coffee served. Contrary to our predictions, no significant improvements in patients' behavior occurred when decaffeinated coffee was first introduced, nor was there any deterioration when regular coffee was reinstated. Only after decaffeinated coffee was introduced for the second time did any of the predicted changes occur; however, the improvements were few in number and may be accounted for by the considerable effect of time per se across all time periods. Although the findings cannot be generalized to all psychiatric patients, the results do not support recent calls for a switch to decaffeinated coffee for this population of inpatients (i.e., chronic schizophrenic patients on high doses of neuroleptics who drink large amounts of coffee).

This study was designed to identify the variables that influence a review panel's decision to discharge or detain an involuntary patient. A group of fifty patients consecutively discharged by the review panel of a provincial mental hospital were compared according to thirty-five variables, with a group of forty-five patients consecutively detained by the panel. The variable set included information on the patient's psychiatric history, current hospitalization and treatment as well as ratings of dangerousness, insight and psychopathology, as reflected in the attending physician's case summary prepared for the review panel. The released and detained groups were found to be remarkably similar. They differed on ten of the thirty-five variables measured, but they did not differ on some variables that one would expect to form the basis of the panel's decision, including diagnosis and a history of suicide attempts. On the other hand, when the predictive value of the variable set as a whole was examined using discriminant analysis, the results indicated that there was a substantial amount of predictability to the review panel process. The group membership of 77.5% of the patients can be predicted from only nine variables that contribute to the discriminant function. The results will be of interest to clinicians who deal with review panels on a regular basis and the findings have implications for other practical issues including discharge planning and readiness for community living.

The authors describe intoxicationrelated behavior patterns observed among 89 chronic schizophrenic inpatients over a 5-year period. These include caffeine intoxication, water intoxication, antihistamine intoxication, nicotine withdrawal, voluntary hypei'venUlation, and ingestion of deodorants and aerosols. Affected patients tended to abuse multiple substances in the hospital, to have generalized polydipsia, and to have histories of drug or alcohol abuse before hospitalization. Periodic intoxication in this population may be an important contributor to the refractoriness of their psychotic symptoms.

In a series of three studies, the present authors and others in this research group have examined the Review Panel process: (a) before the hearing (which patients apply for a hearing?); (b) at the hearing (how do the patients released by the Panel differ from those retained by the Panel and; (c) after the hearing, how are the Panel-released patients faring one and two years after being released by the Panel, compared to patients released by the attending psychiatrist?) In the present paper, the findings of these three studies are summarized and an attempt is made to explain the findings, some of which are surprising (for example, the 35% disagreement between Panel and attending physician with respect to suitability for discharge and the finding that Panel-released patients survive as long in the community as do physician-discharged patients). Based on the findings and the interpretation placed on them, the authors make a series of recommendations with respect to the operation of Panels, the management of defiant involuntary patients and future research in this area.

Fifty patients released by the Review Panel are compared with a matched group of 50 patients discharged by the attending physician at one and at two years after separation from hospital. The two groups did not differ with respect to readmission rate or time spent in the community. At two years the physician-discharged patients were functioning better than the Panel-discharged patients in two of the seven areas of functioning; in the other five areas of functioning the adjustment of the two groups did not differ. The implications of these findings for the operation of the Review Panel and for the timely discharge of involuntary patients by attending physicians is discussed.