a behavioural intervention with families to reduce relapse The community management of schizophrenia. A controlled trial of

NICHOLASTARRIER,CHRISTINEBARROWCLOUGH,CHRISTINEVAUGHN,J. S. BAMRAH,KATHLEENPORCEDDU,SUSANWATTS andHUGHFREEMANSchizophrenic patients were recruited into atrial of aprophylactic behavioural interventionwith families. Familieswith at least one high ExpressedEmotion (EE) relative wererandomly allocated to one of four intervention groups: Behavioural Intervention Enactive;BehaviouralIntervention Symbolic; Education Only; RoutineTreatment. Patientsfrom lowEEfamilies were randomly allocated to two groups: Education Only or Routine Treatment.Relapse rates over nine months after discharge were significantly lower for patients inthe two Behavioural Intervention, compared with Education Only and Routine Treatmentgroups. There was little difference between the two low-EE groups. Patients returningto high-EErelatives showed significantly higher relapseratesthan those returning to Iow-EErelatives, in groups not receiving active intervention. Changesfrom high to low EEoccurredin the BehaviouralIntervention groups, and similar although less extensive changesoccurredinthe EducationOnlyandRoutineTreatment groups.Changesincriticismandmarkedemotionalover-involvement(EOI)occurredgenerallyinhigh-EEgroupsbutwerelarger in magnitude in the Enactive and Symbolic groups. Reduction of hostility onlyoccurred in the Behavioural Intervention groups. These results give partial support forthe causal role of EEin relapse. There were no significant differences between the groupswith respectto contact with the psychiatricservicesor medication.The development of vulnerability—stress modelsof schizophrenia (e.g. Zubin & Spring, 1977;Nuechterlein & Dawson, 1984) has had importantimplications for the management of the disorder,especially within the broad policy of care ofthe mentally ill in the community. The modelincorporates an interaction between an enduring andinherent vulnerability on the one hand, and a set ofstressful environmental stimuli on the other, whichcan result in increasing episodes of the illness. Interms of management, the important factor is thatit may be possible to identify and modify such setsof stimuli, and hence reduce or at least delayrelapse.Following the advent of neuroleptic medicationand the move away from institutional care, a seriesof studies that investigated the relationship betweensocial factors and the recurrence of positiveschizophrenic symptoms identified a measure of therelative's level of Expressed Emotion (EE) as animportant predictive variable (Brown et al, 1962;1972; Vaughn & Leff, 1976a; Vaughn et a!, 1984;seeLeff & Vaughn, 1985;Koenigsberg& Handley,1986, for reviews); they were particularly importantbecause they identified a quantifiable measure ofenvironmental stress that was shown empirically tobe associated with relapse. Patients returning to livewith a relative who had been rated as high-EE hadmuch higher relapse rates than those who returnedto livewith a low-EE relative (Leff& Vaughn, 1985).However, patients who lived with a high-EE relativewould receive some degree of protection boththrough maintenance neuroleptic medication, andhaving low face-to-face contact (less than 35 ha week) with their relative (Vaughn & Leff,l976a).The identification of this familial factor —?highEE —?as being related to and possibly causative ofrelapse, formed the basis of a number of interventionstudies. These utilised psychosocial intervention toalleviate stress within the home environment, thusreducing the EE level of the relative, and hence therate of relapse. Intervention studies provide anopportunity both to assess the efficacy of the clinicaltechniques in altering the family environment (i.e.modifying the stimulus sets) and to determinewhether such modifications influence the course ofthe illness (Koenigsberg & Handley, 1986). Theseintervention studies have been reviewed at lengthelsewhere (Barrowclough & Tarrier, 1984; Leff,1985;Koenigsberg & Handley, 1986;Strachan, 1986;Tarrier, 1988).532