First episode psychosis and outcome : findings from a Swedish multi-centre study

Abstract

The Parachute Project was a Swedish multi-centre project that included 175 First Episode Psychosis patients who were followed over five years. The aim was to provide need adapted care with low medication and based on out-patient support in order to positively affect long-term outcome. One historical and one prospective treatment-as-usual group provided comparison follow-up data. The aim of this thesis is to describe outcome from various perspectives in FEP patients. Study I compared symptoms, functional outcome, care consumption and costs between the Parachute patients with a baseline schizophrenia syndrome and the comparison groups at three-year follow-up. Study II examined the relation of background variables to patients satisfaction with care (PSC) at 1-year follow up. Study III and IV explored the relation between background variables, baseline status gender and 5-year outcome. Study V examined the associations between perceived financial strain and qualitative and quantitative network to level of recovery after 5-years.

After three years the consumption of inpatient care had been significantly lower in the Parachute group than in the prospective comparison group. In spite of increased outpatient care in the Parachute patients group the total mean cost was about half of the cost in the Prospective comparison group. Symptoms and functional outcome were better compared to the Historical group but there were no differences compared to the Prospective group. At one year follow-up the Parachute Project patients showed a high satisfaction with the received care. Factors that were shown to be predictors of PSC were educational level , social network , duration of untreated psychosis (DUP) and Global Assessment of Functioning (GAF) the year prior to onset together with negative symptoms and lack of hope at admission. Twenty-nine percent of the variance in PSC was explained by these factors. At five year follow-up 85 patients (37 males, 48 females) showed a good outcome and 68 patients (44 males, 24 females) a poor outcome. The relative contribution of the baseline characteristics to the 5-year outcome was analyzed. The variables found to have unique contributions were: GAF the year before first admission ≤ 70 , highest educational level is compulsory school , GAF at first admission ≤30 , male gender and contact with friends ≤ 2-3 times/month . Furthermore, gender differences in the predictors were examined. GAF the year before first admission ≤ 70 and GAF at first admission ≤30 explained most of the variance of a poor 5-year outcome for males, whereas for females the corresponding items were highest educational level is compulsory school , living with parents and contact with friends ≤ 2-3 times/month . A predictive rating scale (PRS) was constructed from the results. After five years the Parachute Project patients answered a questionnaire from the PART study, a population-based study of mental ill health, which included questions about qualitative and quantitative network and experience of financial strain. The relation between these variables and outcome group (PART participants, recovered FEP patients and non-recovered FEP patients) was examined. The results showed that financial strain, quantitative social-network and qualitative social-network explained 16% of the variance in the outcome groups.

In summary, the results showed that it is possible to provide need-adapted care to persons suffering from psychosis, but further studies with a naturalistic design are needed to increase knowledge regarding what ingredients are effective and how various interventions interact. In addition it is important with early intervention to decrease the prodromal period and DUP. Outcome depended on various factors present before the first contact with psychiatry and those factors differ between the genders. Thus outcome measures have to be comprehensive and cover various aspect of a person s life. It is not enough to measure outcome, i.e. only with symptom levels or diagnosis. More factors have to be examined and taken into account, such as gender, age, quality of life, amount of medication, social adversity and financial strain, both in treatment choices as well as in outcome measures.