Article Text

Download PDFPDF

Personal therapy reduced adverse outcomes in patients with schizophrenia living with family but increased psychotic relapse rates in those living independently; personal therapy improved social adjustment but increased anxiety
  1. Gary Marlowe, MBBS, MRCGP
  1. St Leonards London, UK

Statistics from

OpenUrlCrossRefPubMedWeb of Science

OpenUrlCrossRefPubMedWeb of Science

Question In patients with schizophrenia living with or independent of family, can personal therapy in addition to antipsychotic medication forestall relapse and improve personal and social adjustment?


2 randomised controlled trials with 3 years of follow up.


Western Psychiatric Institute and Clinic, Pittsburgh, Pennsylvania, USA.


151 patients between 16 and 55 years of age admitted for an index episode of a Research Diagnostic Criteria of schizophrenia or schizoaffective disorder who were subsequently discharged. Exclusion criteria were IQ ≤75, organic brain syndrome, serious alcohol or drug abuse or dependence in the previous 6 months that impaired adjustment, or contraindications to antipsychotic medication. 97 patients (mean age 29 y, 58% men) living with family were included in trial 1 and 54 (mean age 33 y, 56% women) living independent of family were included in trial 2.


In trial 1, 97 patients who resided with family were allocated to personal therapy (n=23), family psychoeducation/management (family therapy, n=24), a combination of personal and family therapy (n=26), or supportive therapy (n=24). In trial 2, 54 patients who either lived alone or shared quarters with non-relatives were allocated to personal therapy (n=25) or supportive therapy (n=29). All patients in both trials were prescribed antipsychotic medication. Personal therapy sought to forestall the relapse and to enhance personal and social adjustment through the identification and effective management of affect dysregulation. Personal therapy encouraged patients to identify the affective, cognitive, and physiological experience of stress. Family therapy followed the 3 broad phases of joining, survival skills training and reintegration within the home, and reintegration into the community.

Main outcome measures

Psychotic or affective relapses, treatment related termination, and personal and social adjustment.

Main results

The mean number of treatment sessions/month/patient over 3 years in the personal, family, and supportive therapy groups were 2.4, 1.4, and 1.6, respectively. Over 3 years, 44 of 151 patients (29%) had a combined total of 66 recurrent episodes of schizophrenia, and 24 patients (16%) had 28 non-psychotic affective relapses. 27 patients (18%) prematurely terminated treatment, most from the no personal therapy treatment conditions. All but 2 of the 66 recurrent psychotic episodes required readmission to hospital. Among patients living with family and based on life table analyses that accounted for the number and speed of relapse, the personal therapy alone condition had a lower psychotic relapse rate than the family therapy condition (p=0.02), and although not statistically significant, than the supportive therapy condition (p=0.10). There were no statistically significant differences regarding affective episodes. When all adverse outcomes were tested by survivorship analyses (time to first psychotic or affective episode or treatment related early termination), there was a borderline overall effect of personal therapy on delaying adverse outcomes (p=0.06). The greatest protection offered by personal therapy against poor outcome occurred in the first year after discharge. Among the 4 treatment conditions, the largest difference was between personal therapy alone and supportive therapy (p=0.01). Among patients living independently of family, those in personal therapy had more psychotic relapses than those in supportive therapy (p=0.02). The difference between the groups was not statistically significant for affective relapses (p=0.09). When all adverse outcomes were analysed (psychotic and affective episodes and treatment related terminations), there were no statistically significant differences between the personal and supportive therapy groups.

In terms of personal and social adjustment, the most important main effects occurred in the second and third year of treatment. Personal therapy had positive effects on broad components of social adjustment (role performance, p=0.02) which included normalisation of functioning in expressive (relationship) and instrumental (activity) roles, decreasing levels of manifest illness, and the resolution of negative symptoms. Personal therapy had few differential effects on symptoms, and patients receiving personal therapy remained more anxious than patients who received family or supportive therapy. For patients living with family, personal therapy led to better outcomes in overall performance than did family or supportive therapy. Family therapy had only 1 positive effect on social adjustment but personal adjustment improved more in these patients than in those receiving family or supportive therapy (p=0.03). For patients not living with family, personal therapy compared with family or supportive therapy improved work performance (p<0.001) and relationships out of the home (p=0.02).


Among patients with schizophrenia who were living with family, personal therapy reduced adverse outcomes. Among patients with schizophrenia who were living independently, personal therapy increased psychotic relapse rates compared with supportive therapy but this difference was no longer statistically significant when all adverse outcomes were considered. Personal therapy had positive effects on broad components of social adjustment but had few differential effects on symptoms, and these patients remained more anxious than those receiving family or supportive therapy. For patients living with family, personal therapy led to better outcomes in overall performance. For patients living independently, personal therapy compared with supportive therapy led to improvement in work performance and relationships outside of the home.


It is often taught that personal psychotherapy for schizophrenia has no beneficial effect on outcome and may have deleterious effects, even increasing the risk of suicide. However, “the evidence that psychodynamic treatment worsens the outcome of schizophrenia is indirect and debatable.”1 These 2 studies by Hogarty et al are therefore welcome.

I am impressed by the care that was taken with the design of the trials. The inclusion criteria were clear and adhered to research design centre standards. Where possible, analyses were done on an intention to treat basis. The important confounding factor of medication variation was addressed both by attention to consistency of prescription and by after trial analysis searching for systematic differences. Consistency of the intervention was attempted by having a small and stable number of therapists. Other confounding factors reflecting patients' needs were addressed by the primary physician. The problem of rapid decay of efficacy after an initially intensive but brief intervention is addressed by the long (3 years) trial period. Most studies in this area extend over 6–12 months. The statistical analyses were rigorous and conventionally displayed.

The important points are that, firstly, for patients living with their families personal therapy reduced relapse rates. Secondly, the effects of personal therapy, when considering all adverse outcomes, would seem to continue to increase at the end of the study unlike the effects of family and supportive therapies which appear to have reached a plateau. In fact the improvements in social adjustment, especially important in the management of negative symptoms, begin to appear in the second and third year of treatment. Thirdly, personal therapy for patients living independently results in significantly more relapses than supportive therapy. However, this is against a background of no statistically significant difference when all adverse outcomes are considered.

Any quibbles I have are relatively minor. The mean number of treatment sessions varied across the groups so that those receiving personal therapy saw their therapist more often (approximately 40% more often). Of the patients with schizophrenia admitted to the inpatient unit, 75% were not eligible for the study. Although the criteria for exclusion were not archived it is suggested that many of these would be because of drug and alcohol problems. This substantial and difficult subgroup continues to defy our therapeutic attempts. There was an absence of independent and blinded clinical assessors. To their credit, these points are clearly acknowledged by the authors.

Not all that can be counted counts. It is easier to measure changes in areas of practical adaptation but more difficult to assess areas of intrapscyhic function. The second paper attempts this by looking at the rather more nebulous components of social adjustment. The improvements that were measured were statistically and almost certainly personally meaningful. Patients receiving personal therapy remained more anxious than those receiving family or supportive therapy. This would perhaps be an obvious outcome to any psychoanalyst; after all, one would expect any personal therapy worth its salt to lead to stirring of unconscious, repressed forces in the self. The worry may be that while pursuing these changes there might be a potential for precipitating deterioration in other dimensions.

Interestingly, although the trials were primarily aimed at assessing personal therapy, the often advocated family therapy2 does not seem to be any better than supportive therapy (this is an inference rather than analysis). Another useful comparison would be with cognitive behavioural therapy (CBT). Recently, there have been reports of various studies suggesting that CBT is an effective measure in the treatment of schizophrenia. Some authors have suggested that positive outcome in CBT is due in part to specific effects upon delusional thinking.3 These studies of CBT are, however, of limited duration, with small samples, and have tended to have relatively circumscribed outcomes, mainly positive symptoms. Personal therapy encourages patients to identify affective, cognitive, and physiological experiences and may affect a broader range of schizophrenic symptoms, including negative symptoms.

No economic analysis was mentioned, but I would imagine that such an intensive and specialised therapy would be costly. Bearing in mind, however, the enormous burden that schizophrenia causes both in patients and their families, any therapy that reduced such an important outcome measure as relapse should be considered. For patients living with families who have achieved a reasonable degree of symptom stability it would have my wholehearted recommendation. However, I would suggest much more cautious consideration for patients living independently. Sadly, this represents most of the patients with schizophrenia that I see.


View Abstract


  • Source of funding: National Institute of Mental Health.

  • For correspondence: Professor G E Hogarty, Western Psychiatric Institute and Clinic, 3811 O'Hara Street, Pittsburgh, PA 15213, USA. Fax +1 412 624 3429.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.