Schizophrenia Bulletin, Vol. 22, No. 2, 1996
National Institute of Mental Health


Secondary prevention in schizophrenia is still not good enough, and the authors of the articles in this issue of Schizophrenia Bulletin present important new knowledge that may increase earlier detection and treatment. This final article discusses some of their main findings in relation to clinical consequences and new research. We need more knowledge about the clinical characteristics of the prodromal phase and about the factors that influence the help-seeking behavior of patients and families in the early phases of schizophrenia. The new models presented for early detection and treatment are promising and should be replicated in other countries and on other continents.

Schizophrenia Bulletin, 22(2): 347-351, 1996.

Because we still lack an effective primary prevention of schizophrenia, we need to make great efforts to improve our secondary prevention. As long as this disorder persists as one of our most costly medical disorders — economically as well as in terms of human suffering – any clinically significant reduction in the severity and/or duration of schizophrenia is surely worth any investment.

Theoretically, secondary prevention is improved if we can apply a more effective treatment at an earlier point in the course of most disorders. For schizophrenia, there is some empirical evidence, although inconclusive, that this is the case. The work to improve the secondary prevention of schizophrenia will therefore have two aims: to find strategies and methods for earlier detection and to develop treatment methods and programs that are tailor-made for patients in the early phases of schizophrenia.

What kind of knowledge do we skill need to reach these aims? Concerning earlier detection, we obviously need to know more about the very early signs and symptoms indicating the development of a schizophrenic disorder with a reasonably high degree of probability. We also need to know the biological, psychological, and social factors that influence the treatment-seeking behavior of subjects in the earlier phases of schizophrenia. Although the earliest signs and symptoms of schizophrenia may be universal and thereby easily generalizable, treatment-seeking behavior may be differently influenced by individual, family, social, and health service factors in different cultures and countries. Therefore, general knowledge about the most influential factors on treatment seeking must be based on studies from different countries and continents.

Concerning early treatment intervention, we need to know whether the treatment modalities that have a significant impact on schizophrenia patients under the usual treatment-seeking situation, such as psychopharmacological, psychoeducational, and assertive outreach programs (Lehman et al. 1995), can be applied with the same or better results and without modifications to an earlier detected group. We should also explore whether patients detected earlier may benefit even more from other interventions. It is also important to prove empirically that an early intervention by itself is more beneficial for the patients and their families than a later one. These questions should be explored by comparing earlier and later detected patients who receive the same treatment program, at least concerning the core elements. We are in fact initiating such a prospective study in Norway in collaboration with Danish and American groups.

Several of the research questions mentioned here are addressed by the contributors to this issue of the Schizophrenia Bulletin. In this final article, I will summarize and discuss some of the answers that these authors have given to these questions. I will also indicate some clinical consequences of the findings, and some research questions that should be studied further.

What Are the Earliest Signs and Symptoms of Schizophrenia?

The time interval from the onset of a schizophrenic disorder to the establishment of an adequate therapy is often several years (Larsen et al. 1996b, this issue), even if we estimate this time from the point when the patient fulfills the DSM-III-R (American Psychiatric Association 1987) criteria for schizophrenia and not from the (retrospectively estimated) beginning of the prodromal phase. Because most of our efforts to identify and treat patients in the prodromal phase will also influence the identification and treatment of patients in the early manifest psychotic phase, I will focus mostly on early detection and treatment in the prodromal phase.

Schizophrenia usually starts with a prodromal phase in which the patient does not fulfill all the criteria of the disorder, but in which observable and experiential changes in behavior, thinking, and feelings occur. An early treatment should be established to identify (and treat) these patients in the prodromal phase. To accomplish this we need to know the “true” symptoms and signs of the prodromal phase. Most of our knowledge, as reflected in the DSM-III-R criteria, has been collected retrospectively by interviewing patients and families about the history of the psychotic development. These investigations, like the study reported here by Yung et al. (1996, this issue), have found that the length of the prodromal phase may vary from days to many years, and that most of the reported symptoms and signs seem to be rather nonspecific, even for psychosis. These findings also indicate that not all schizophrenia patients will have a long enough prodromal phase to be identified as such within this phase, and that cross-sectionally it is very difficult to pick out those who will eventually develop schizophrenia among persons having more moderate problems like anxiety, depression, or sleeping disorders. Even among persons with a substantial vulnerability for schizophrenia, some will fail to proceed from the prodromal phase to a full-blown syndrome even without treatment. Furthermore, the prodromal phase may be more or less the same for many types of psychoses (Yung et al. 1996, this issue). We need to know which features in the prodromal phase can predict the type and course of the subsequent psychosis. Ideally this should be studied prospectively, but it can also be explored retrospectively by comparing patient groups that each have a different first-onset psychosis.

Some authors propose that attenuated psychotic symptoms like attentional or perceptual distortions may increase the low specificity of the prodromal signs and symptoms (see Yung et al. 1996, this issue). Based on their own clinical experiences and the findings from research on vulnerability markers, Olin and Mednick (1996, this issue) and Yung et al. (1996, this issue) propose that in the 16-30 year-old cohort population we can use a combination of different predictors of schizophrenia to identify persons with moderate psychopathological disturbances who should get early treatment for schizophrenia. McGorry et al. (1996, this issue) and Yung et al. (1996, this issue) are now exploring the value of combining risk factors such as family history of psychosis, prodromal symptoms from the DSM-III-R list, history of transient psychotic symptoms, and other neurobiological and neuropsychological predictors of schizophrenia. This is based on the notion that persons with prodromal symptoms of psychosis run an increased risk for advancing further into manifest schizophrenia if the above characteristics are present. The extensive list of predictors of schizophrenia (Olin and Mednick 1996, this issue) could be used to decide whether a person has a preschizophrenic prodromal syndrome. However, in clinical practice, subjecting patients with only vague prodromal symptoms to extensive biological and neuropsychological examinations may not be feasible. Therefore, we need to explore further which predictors best indicate a forthcoming schizophrenic syndrome among prodromal subjects. The ongoing study in Australia (Yung et al. 1996, this issue; McGorry et al. 1996, this issue) may give us new answers to this question.

Prodromal symptoms for relapses have been studied in several follow-along studies of schizophrenia cohorts. As reviewed by Falloon et al. (1996, this issue), these prodromal symptoms are often idiosyncratic for each patient. If this is also the case in the first prodromal phase, we should focus more on the specificity of individual changes (e.g., degree, type, speed, stress relation) than on the symptom state to identify the important differences between a prodromal phase of psychosis and other changes that may occur in adolescence and young adulthood.

In general, it might be best to follow the advice from McGorry et al. (1996, this issue) to study different phases of the psychotic process more separately with regard to clinical picture — influencing biological, psychological, and social factors — and therapeutic interventions. As in substance abuse disorders, different factors may be important for the prodromal, the active psychotic, the relapsing, and the chronic phases.

Which Factors Influence the Duration of Untreated Psychosis (DUP)?

In general, treatment seeking is influenced by a combination of factors relating to sociodemographics, disorder, personality, family, and health services. DUP is one way of measuring some important aspects of the treatment-seeking behavior in the early phases of schizophrenia. An effort to reduce DUP time must be based on knowledge of which factors both strongly influence DUP and are changeable. These factors may be different in the prodromal and the early manifest psychotic phases, so these phases should be explored separately. The studies in this issue do not clearly separate these phases, and the forthcoming discussion is therefore relevant to both phases. The studies in this issue (Larsen et al. 1996b; McGorry et al. 1996) strongly support previous findings showing that there is a wide variation in DUP and therefore much to gain in secondary prevention if we can reduce the duration. The Larsen et al. study (1996b, this issue) found that women had a significantly lower DUP than men (39 weeks vs. 154 weeks). If this sex difference remains when one controls for other predictors of DUP and is replicated in studies from other countries, new research should explore reasons for this sex difference. In programs for early detection, one should be especially aware of the risk for delayed treatment in men.

Age at onset and age at first hospital treatment do not seem to have any definite relationship to DUP. The importance of social class or ethnicity is still unknown. Since schizophrenia most often occurs in young age and in lower socioeconomic groups, programs for early detection should focus on adolescents and young adults and be available in low socioeconomic areas.

The study by Larsen et al. (1996a, this issue) found that many, but not all, patients have had poor or declining social functioning for years before either the prodromal or the first manifest psychotic phase. Poor social functioning and/or a decline in social functioning from childhood to late adolescence can be related to long DUP, as can an insidious onset of psychosis and the presence of negative symptoms (Larsen et al. 1996a, this issue). Together, these findings indicate that a slow deterioration in function over years seems to increase the threshold that the person (or the family) has to surmount to seek help and therefore increases DUP. An early detection program could teach general practitioners (GPs), families, schools, and others to identify changes in social functioning and insidiously developing negative symptoms. The study by Olin and Mednick (1996, this issue) underlines how competent teachers already are at identifying pupils at risk for becoming psychotic. As mentioned by McGorry et al. (1996, this issue), certain positive symptoms such as persecutory ideas may also delay treatment- seeking. These positive symptoms might be easier for the patient’s network persons to identify, but they may need to know how to overcome such patient’s resistance to getting help.

In the prodromal phase, one would also expect the presence of premorbid personality features of the paranoid, schizoid, schizotypal, avoidant, narcissistic, or antisocial type to increase DUP. There is, however, no clear empirical evidence of this. A great proportion of schizophrenia patients may have premorbid personality disorders, and we should explore further how such disorders influence both DUP and early treatment compliance. Again, an early detection program could train GPs and social welfare workers on how to establish a working alliance with patients having such personality disorders.

Without doubt, the family plays an important role in the treatment-seeking and treatment-receiving process of psychotic patients, but we have no clear evidence as to how strongly and in what ways they influence DUP. McGorry et al. (1996, this issue) report that early in the help-seeking phase, subjects frequently sought help themselves (40% at first contact), but later their family and others tended to seek help on their behalf. To help families detect possible signs of psychosis early and establish adequate help early, we need to know what the problems are on a family level. An early intervention program should also focus on how families tend to cope with their prodromal or manifest psychotic members in accordance with family traditions (Vaglum 1972).

Health service-related factors also have a big influence on DUP. The studies collected in this issue point to several problems with both the lack of recognition of prodromes and early psychotic symptoms, and the lack of adequate treatment in the health services in several countries. In the Norwegian study, one third of those with DUP of more than 1 year had been identified early but did not receive adequate treatment (Larsen et al. 1996b, this issue). In the Australian study (McGorry et al. 1996, this issue), only 5 percent of the referrals to psychiatry came from GPs. These and other studies show that secondary prevention of schizophrenia could be increased with improvements in early diagnosis and management among health and social personnel. Falloon et al. (1996, this issue) and McGorry et al. (1996, this issue) have presented research from such efforts in England and Australia and indicate the possibility of reducing DUP by a tailor-made program involving a broad spectrum of people able to detect and treat subjects with a beginning psychosis. Their findings, however, need replication from different countries with different health service systems such as those in Scandinavia and the United States.

Do Early Detected Patients Need Special Treatment?

Regardless of how early they are detected, patients who meet the criteria of a schizophrenic disorder should receive an optimal treatment program as soon as possible. It should contain as core elements psychopharmacological treatment, a psychoeducational family interventional program, and a long-term outpatient program (Lehman et al. 1995). The important question here is whether this type of program may also be used for a group of patients in the prodromal phase or in the early manifest phase of psychosis, or whether one has to modify the treatment approach.

McGorry et al. (1996, this issue) and Yung et al. (1996, this issue) point to several important changes that are needed. They have found that assessment should be performed in the least threatening environment (e.g., home, school, local GP’s office), over an extended period of time, and by the same team that can build up a working alliance with the subject and the family. The assessment should be offered very soon, because it is easier to establish a working relationship before the patient becomes manifestly psychotic. The assessment procedure must be conducted knowing that some of the supposed prodromal subjects will be false positives and others will be true positives, who never will proceed to a fully developed psychosis.

The experiences of McGorry et al. (1996, this issue) and Falloon et al. (1996, this issue) suggest that providing interventions on an outpatient/daypatient level (e.g., ad modum case management) and in combination with a family-supporting intervention is possible and effective. In psychopharmacological treatment, it may be possible to use lower doses for shorter periods of time than usual (Yung et al. 1996, this issue). The research of this Australian group may give us new data and ideas about the psychopharmacological treatment of the prodromal phase. Although the Falloon et al. (1996, this issue) and McGorry et al. (1996, this issue) groups report very promising results in their exploratory studies, the substance of an optimal treatment program for patients detected in the prodromal phase skill needs research (McGlashan 1996, this issue; McGlashan and Johannessen 1996, this issue).


An optimal secondary prevention of schizophrenia demands detection as early as possible and adequate treatment opportunities for subjects who are in the early phases of the disorder.

The studies presented in this issue of Schizophrenia Bulletin add important knowledge to what we already know about improving secondary prevention. They increase our understanding of the early phases of psychosis, especially the clinical signs and symptoms and the state and trait markers that may influence the change from prodromal to manifest psychotic states. They also propose clear definitions of the different phases and points of time in the course of schizophrenia, theoretical work that is important for further clinical work and research.

Of special importance are the demonstrations, mainly by Falloon et al. (1996, this issue) and McGorry et al. (1996, this issue), of how one can design programs for early detection that extend into the prodromal phase and can reduce DUP. These groups have given us new clinical knowledge of how to organize and manage very early assessments and have proposed modifications to the usual treatment approach to schizophrenia for recruiting a prodromal group.

In the future, several important clinical questions should be pursued in clinical trials and research, including the following:

  • What psychobiological factors influence the precipitation of the prodromal phase and which factors increase the risk of proceeding from this phase to manifest psychosis?
  • Are there schizophrenic specific and/or psychosis specific signs or symptoms in the prodromal phase?
  • What are the most important personality, family, and health service factors that influence the illness behavior and DUP in the first onset of schizophrenia?
  • What are the benefits to the patient, the family, and society of a treatment instituted in the prodromal phase or at the point of the first manifest psychotic episode?
  • Which changes in the usual optimal treatment program for first-episode schizophrenia can improve the outcome, to a clinically significant degree, in cases identified in the prodromal phase?

The research groups represented in this issue of Schizophrenia Bulletin are all working intensively on one or several of these questions.

This research is, however, methodologically as well as administratively very complicated, demanding close collaboration with relatively extensive treatment service organizations and qualified researchers. The work must continue over a long time, and a combination of studies on clinical population and populations drawn from the general population may be necessary. These studies will therefore need firm and long-term support from government and local health authorizes if they are to reach their goals.


  • American Psychiatric Association. DSM-III-R Diagnostic and Statistical Manual of Mental Disorders. 3rd ed., revised, Washington, DC: The Association, 1987.
  • Falloon, I.R.H.; Kydd, R.R.; Coverdale, J.H.; and Laidlaw, T.M. Early detection and intervention for initial episodes of schizophrenia. Schizophrenia Bulletin, 22(2):271-282, 1996.
  • Larsen, T.K.; McGlashan, T.H.; Johannessen, J.O.; and Vibe-Hansen, L. First-episode schizophrenia: II. Premorbid patterns by gender. Schizophrenia Bulletin, 22(2):257-269, 1996a.
  • Larsen, T.K.; McGlashan, T.H.; and Moe, L.C. First-episode schizophrenia: I. Early course parameters. Schizophrenia Bulletin, 22(2):241-256, 1996b.
  • Lehman, A.F.; Thompson, J.F.; Dixon, L.B.; and Scott, J.E. Schizophrenia: Treatment Outcomes Research — Editors’ Introduction. Schizophrenia Bulletin, 21(4):561-566, 1995.
  • McGlashan, T.H. Early detection and intervention in schizophrenia: Research. Schizophrenia Bulletin, 22(2):327-345, 1996.
  • McGlashan, T. H., and Johannessen, J.O. Early detection and intervention with schizophrenia: Rationale. Schizophrenia Bulletin, 22(2):201-222, 1996.
  • McGorry, P.D.; Edwards, J.; Mihalopoulos, C.; Harrigan, S.M.; and Jackson, H.J. EPPIC: An evolving system of early detection and optimal management. Schizophrenia Bulletin, 22(2):305-326, 1996.
  • Olin, S.S., and Mednick, S.A. Risk factors of psychosis: Identifying vulnerable populations premorbidly. Schizophrenia Bulletin, 22(2):223-240, 1996.
  • Vaglum, P. Contrasting multigenerational attitudes toward psychosis in two Norwegian families. Family Process, 11:311-320, 1972.
  • Yung, A.R.; McGorry, P.D.; McFarlane, C.A.; Jackson, H.J.; Patton, G.C.; and Rakkar, A. Monitoring and care of young people at incipient risk of psychosis. Schizophrenia Bulletin, 22(2):283-303, 1996.

The Author

Per Vaglum, M.D., Ph.D., is Professor, Department of Behavioural Sciences in Medicine, University of Oslo, Norway.

This article, originally from the Schizophrenia Bulletin, is in the public domain and may be reproduced or copied without requesting the author’s permission.