Peer reviewed Studies: Delaying Treatment leads to poorer recovery and prognosis

Edited by Dr. Fuller Torrey.

Prolonged recovery in first-episode psychosis.

Edwards J, Maude D, McGorry PD, Harrigan SM, and Cocks JT (1998). British Journal of Psychiatry (Supplement) 172 (Suppl 33): 107-116.


This study found that 6.6% of a sample of individuals experiencing their first episode of psychosis were treatment resistant, and that between 9 and 17% of individuals with schizophrenia, schizophreniform, and schizoaffective disorder were treatment resistant. These rates are much lower than the 30% reported in more chronic samples. In assessing differences between treatment resistant individuals and those who responded to treatment, it was found that the duration of psychotic symptoms before treatment approached significance, which suggests that untreated psychosis may lead to treatment resistance, and supports the argument for early intervention as soon as possible following the onset of positive symptoms. The lack of a difference between the groups on a scale assessing premorbid functioning further suggests that premorbid factors are less important than duration of untreated psychosis in preventing treatment refractoriness.


The aim of this study was to identify individuals experiencing treatment resistance early in the course of their psychotic disorder. It has been suggested that approximately 30% of individuals with schizophrenia have a less than adequate response to antipsychotic medications; however, it is not clear how many of these individuals are treatment refractory at the beginning of their illness and how many become so subsequently. Studies have suggested that the critical period for the development of chronic impairment is approximately one year.

The sample in this study consisted of 227 individuals experiencing their first psychotic episode, assessed at three time points (at admission, at stabilization of symptoms, at three or six months after stabilization, and at 12 months after stabilization) over a 12 month period. Thirty-six percent had schizophrenia, 22.5% had schizophreniform disorder, 11% had schizoaffective disorder, 2.2% had delusional disorder, 12.8% had bipolar disorder, 8.4% had depression with psychotic features, 0.4% had brief reactive psychosis, and 6.6% had psychotic disorder NOS. Individuals were identified as treatment resistant when they attained threshold on at least one positive symptom item of the Brief Psychiatric Rating Scale (BPRS) at stabilization, at the second time point (three or six months after stabilization), and at 12 months after stabilization.

The authors found that 6.6% of all the individuals with first-episode psychosis experienced psychotic symptoms at these three time points. When the analysis was restricted to people initially diagnosed with schizophrenia, schizophreniform, or schizoaffective disorder, the percentage of treatment-refractory individuals was 8.9%. When only schizophrenia and schizophreniform disorder were considered, the total rose to 11.4%. These numbers are much lower than the estimated 30% of treatment resistant schizophrenics in the larger schizophrenic population.

Treatment refractory individuals were found to have a significantly longer duration of psychotic symptoms during their first hospitalization, a greater severity of depressive symptoms 12 months after stabilization, and poor psychosocial functioning 12 months after stabilization. Interestingly, the treatment-refractory patients did not differ from the treatment responders on measures of premorbid functioning. However, the duration of psychotic symptoms before treatment also approached statistical significance, which suggests that untreated psychosis may lead to treatment resistance. The authors conclude that, in order to shorten the duration of active psychosis, there is a need to shorten not only the duration of untreated psychosis prior to entering treatment, but also to reduce the duration of psychosis after treatment has begun; the global aim can thus be seen as shortening the duration of active psychosis.

Course and outcome for schizophrenia versus other psychotic patients: a longitudinal study.

Harrow M, Sands JR, Silverstein ML, and Goldberg JF (1997). Schizophrenia Bulletin 23:287-303.


When looking at three groups of patients—those with schizophrenia, patients with other psychotic disorders, and patients with nonpsychotic disorders—the authors found that patients with schizophrenia show worse outcomes than other psychotic patients. In recent years, there has been speculation over whether patients with schizophrenia who respond to antipsychotic medications show relatively favorable or unfavorable outcomes over time. This study showed that, for individual patients with schizophrenia, the pattern of psychopathology and the level of functioning over time are related to previous psychopathology. It therefore suggests that preventing poor functioning at any given time for an individual with schizophrenia could be integral to preventing poor functioning in the future.


The authors studied 276 young, early-phase patients longitudinally, beginning at the acute phase and continuing through followups at roughly two years, four and a half years, and seven and a half years. Seventy-four of the patients were schizophrenic, 74 had other psychotic disorders, and 128 suffered from nonpsychotic mental illnesses. Patients were evaluated initially and at each successive followup for overall functioning; psychotic, anxiety, and affective symptoms; work and social functioning; potential rehospitalization; and medications.

The authors found that patients with schizophrenia functioned significantly more poorly than patients with other psychotic disorders at each of the three followups. The combined group of nonpsychotic and “other psychotic” patients tended to show improvement in overall outcome following the index hospitalization; in contrast, patients with schizophrenia showed more consistent psychopathology and evidence of psychotic symptoms, and had been rehospitalized significantly more frequently at each of the followups than the other patient groups. Although 32% of the schizophrenic sample showed a complete remission at one of the followups, less than 5% had a complete remission at all three of the followups. In contrast, at each followup after the first, more of the other psychotic patients had evidence of a complete remission than those who had uniformly poor outcomes.

Furthermore, schizophrenia patients deemed to be doing poorly at the initial evaluation showed significantly slower recovery at each followup than the other psychotic patients who had been doing poorly early on. There was a similar tendency for both schizophrenic and other psychotic patients with good functioning at one followup to show good functioning at the next followup. Schizophrenics also had the poorest outcome in terms of work functioning, although their social functioning was comparable to both the other patient groups. These results suggest that, during the early course of the illness, patients with schizophrenia still show relatively poor outcomes when compared with other mentally ill patients. Although a small number of them enter into complete remission, they tend to recover more slowly than other psychotic patients.

Research and treatment strategies in first-episode psychoses: the Pittsburgh experience.

Keshavan MS, Schooler NR, Sweeney JA, Haas GL, Pettegrew JW (1998). British Journal of Psychiatry 172 (Suppl 33): 60-65.


Studies of individuals with first-episode schizophrenia help avoid the confounding influences of chronicity, medication effects, and institutionalization on neurobiological alterations, as well as on the course of illness. In this study, baseline neurobiological evaluations of first-episode patients revealed structural and functional brain abnormalities consistent with a neurodevelopmental model of schizophrenia. The authors conclude that such changes probably take place prior to the onset of illness; however, because such changes occur progressively over time, early intervention may halt some of the expected changes.

Preliminary data also supported the value of risperidone as an antipsychotic medication for individuals suffering from their first episode of schizophrenia. Until now, atypical antipsychotic medications have been used for predominantly chronic patients.


The authors begin by reviewing the often confusing findings regarding whether there are changes in the brains of individuals with schizophrenia and, if such changes exist, they occur before or after the onset of the illness. The authors’ review of the literature concludes that such changes do exist, and that they probably occur prior to illness onset.

This article then summarizes preliminary findings from ongoing magnetic resonance imaging (MRI) and spectroscopy studies of first-episode schizophrenia patients.Patients were studied for a period of two years. In this study, the authors focused on longitudinal changes in the structure of the basal ganglia. In a series of 16 patients with first-episode schizophrenia, baseline neurobiological evaluations revealed structural and functional brain abnormalities consistent with a neurodevelopmental model of schizophrenia. Significant reductions in caudate volume were seen prior to initiation of treatment, compared with people who had a non-schizophrenic psychosis and healthy controls. In a subset of 11 people with first-episode, treatment-naive schizophrenia treated with typical antipsychotic medications, follow-up MRI scans revealed a significant increase in caudate volume. Past studies have suggested that structural changes in basal ganglia related to antipsychotic medications may be confined to typical medications.

Because some studies have suggested that delay from onset of illness to first treatment is associated with more severe psychotic symptoms, poorer functioning, and less adequate response to antipsychotic medications, 19 patients who had previously untreated first-episode schizophrenia were also studied. An inverse relationship was found between illness duration and the volume of the left superior temporal gyrus. Other studies looking at the correlation between various brain abnormalities and duration of psychosis have produced mixed results.

During the two-year period, patients received treatment with either typical antipsychotic medications or the atypical antipsychotic medication risperidone. All 16 patients had not previously been treated with antipsychotic medications. Preliminary data support the value of risperidone as an antipsychotic medication for individuals suffering from their first episode of schizophrenia. Both groups showed significant improvement in negative and positive symptoms, as well as in global functioning. Furthermore, significantly fewer people taking risperidone needed concomitant administration of anticholinergic agents.

First-episode schizophrenia with long duration of untreated psychosis: pathways to care.

Larsen TK, Johannessen JO, Opjordsmoen S (1998). British Journal of Psychiatry 172 (Suppl 33): 45-52.


Long duration of untreated psychosis has been associated with worse outcome in schizophrenia. This study divided a group of 34 first-episode patients with schizophrenia into two groups: those with a long and short duration of untreated psychosis. The main obstacles to receiving treatment were found to be withdrawal and having a poor social network, suggesting that, in order to identify people earlier, a system of detection must be attentive to the early symptoms of psychosis and include provisions for follow-up care. The social network related to the individual should also be educated as to the importance of early treatment.


The goal of this study was to assess the impact of duration of untreated psychosis in individuals suffering from first-episode schizophrenia. Most studies of first-episode schizophrenia have suggested that treatment delay is a serious problem because the beneficial effects of treatment may be strongest in the earliest phases of the illness. The authors assumed that people with schizophrenia who had a long duration of untreated psychosis would have poorer premorbid functioning and more negative symptoms at onset, and that the main obstacles to receiving treatment would be social withdrawal, a poorly functioning social network, and the misinterpretation of early symptoms of psychosis by primary health services.

Individuals studied were 34 people with schizophrenia admitted to Rogaland Psychiatric Hospital in Norway. All were recruited as part of a first-episode study of non-affective psychosis. Functioning was assessed using the Positive and Negative Syndrome Scale (PANSS), the Structured Clinical Interview for DSM-III-R (SCID), the Premorbid Adjustment Scale (PAS), and the Global Assessment of Functioning Scale (GAF). Duration of untreated psychosis was defined as the time interval between the onset of psychotic symptoms and hospitalization for psychosis or initiation of adequate treatment (defined as sufficient quantities of antipsychotic medications). The mean value of duration of untreated psychosis was 130 weeks; the median value of duration of untreated psychosis was 54 weeks. Participants were thus split into two groups: short duration of untreated psychosis was considered less than 54 weeks (mean: 15 weeks), and long duration of untreated psychosis 54 weeks or longer (mean: 244 weeks).

No statistically significant differences between the two groups were found for age at onset of psychosis, age at hospitalization, premorbid functioning (as measured by the PAS), or GAF scores in the last week before hospitalization. There was, however, a trend for the long duration of untreated psychosis group to have poorer premorbid functioning in late adolescence and adulthood than the group with short duration of untreated psychosis. Not working and having few friends were found to significantly increase the risk of having a long duration of untreated psychosis. Furthermore, there was a trend for the long duration of untreated psychosis group to have more negative symptoms at hospitalization (as measured by the PANSS), and that group had the worst scores on the three variables of the PANSS that describe withdrawal (emotional withdrawal, passive/apathetic social withdrawal, and active social avoidance).

When reviewing individual cases, the authors found that for most cases, treatment was offered long before adequate treatment was initiated; in only one case was the individual not identified because of poor assessment. For most of the 34 cases, withdrawal was rated as the most important reason for late detection. Poor social network was the second strongest factor. Differences in how schizophrenia presented itself did emerge between the two groups, however. Most patients in the short duration of untreated psychosis group had a more obvious change in functioning from the prodromal stage to psychosis, although these individuals also seemed to have a better social network and more insight regarding their symptoms.

Factors influencing treatment response and outcome of first episode schizophrenia: implications for understanding the pathophysiology of schizophrenia.

Lieberman JA, Koreen AR, Chakos M, Sheitman B, Woerner M, Alvir JMJ, and Bilder R (1996). Journal of Clinical Psychiatry (Suppl) 57 (Suppl 9): 5-9.


This study indicates that, although treatment resistance may be present in a small proportion of patients suffering from their first episode of schizophrenia, it develops in other patients over the course of their illness. This suggests that, although certain aspects of the illness in terms of its severity and course are predetermined, a number of other risk and treatment factors can exert favorable and unfavorable effects on the course of illness. There may be an active pathologic process that occurs during periods of acute psychosis which, if not counteracted, can produce persisting morbidity and impair patients’ ability to respond to treatment. If so, then the early use of antipsychotic medications could be integral to preventing relapses and thus subsequent deterioration in patients with schizophrenia.


The authors’ premise is that, in schizophrenia, specific disease-related factors are either associated with or influence the course of the illness, as well as the response to antipsychotic medications. Most studies have found that first episode patients with schizophrenia differ from more chronic patients. The authors cite their recent study, conducted with Nina Schooler, which found that, with treatment, acute psychotic symptoms in first-episode patients were more likely to subside than those of more chronic patients, and that first-episode patients required lower doses of antipsychotic medications for both acute and maintenance treatment. However, it is not clear whether these differences are the result of disease progression or attrition patterns in the patient population, with the first-episode patients who have the least severe course recovering and thus not being present in analyses of more chronic populations.

In order to address the issue of what causes the difference between first-episode and chronic patients with schizophrenia, the authors analyzed the relapse pattern and recovery time for a group of 70 schizophrenic patients followed prospectively over 10 years. All patients were initially treated with fluphenazine, and most (70%) had had no prior exposure to antipsychotic medications. The authors found that 86% (n=57) of this sample recovered (either fully or partially) from their first episode of illness; however, 22 of these patients had a subsequent relapse, and another six had a second relapse (or third episode). For the group of 22 patients, the median time to remission was 8.4 weeks for the first, and 11.9 weeks for the second episode. Whereas all 22 patients remitted from the first episode, only 20 remitted from their first relapse. The median time to remission for the six patients who had three episodes were four, seven, and 24.1 weeks, respectively. Although the authors caution that their results are preliminary, the findings show a clear pattern of decreasing responsiveness to treatment over subsequent episodes of illness; however, they cannot determine to what degree this diminished response is due to the disease or because there is something about the treatment that produced it.

The authors also discuss the risk factors associated with outcome in schizophrenia. Factors thought to negatively influence the course of the illness include: longer duration of illness; more numerous episodes; duration of psychosis prior to the first episode of illness; the presence of extrapyramidal side effects (EPS) or tardive dyskinesia (either idiopathic or drug-induced); and enlargement of the lateral and third ventricles in the brain. Taken together, these suggest that there is an active pathologic process that occurs during periods of acute psychosis which, if not counteracted, can produce persisting morbidity and impair patients’ ability to respond to treatment.

Factors thought to positively influence the course of illness include: early treatment with antipsychotic medications, and higher pretreatment levels of plasma homovanillic acid, one of the major metabolites of dopamine.

Early intervention, untreated psychosis and the course of early schizophrenia.

Linszen D, Lenoir M, de Haan L, Dingemans P, Gersons B (1998). British Journal of Psychiatry 172 (Suppl 33): 84-89.


Results from a previous study by these authors had shown that an intensive 15-month early intervention program improved the course of illness for individuals with recent-onset schizophrenia. In this follow-up study, however, the authors found that the benefits of this intervention did not last once patients had been referred to other mental health agencies. The favorable effects of the early and intensive treatment disappeared rapidly, suggesting that the only way to prevent poor outcome in schizophrenia seems to be the continuation of medication compliance, stress management, and case management for an as yet unknown critical period.


A previous study by these authors had shown that, in a group of patients with recent-onset schizophrenia, an intensive early intervention program had a beneficial effect on relapse and the course of psychosis, which lasted until the end of the 15-month period. Treatment consisted of an initial inpatient hospitalization which included psychoeducation for patients’ relatives, followed by outpatient treatment with a day hospital and community care. Seventy-six patients completed both the inpatient and outpatient phases of the clinical trial (55 percent had schizophrenia, 21 percent schizoaffective disorder, 13 percent schizophreniform disorder, and 11 percent other psychotic disorders such as delusional disorder and atypical psychosis). Half the group randomly received a year-long behavioral family intervention program combined with the patient-oriented psychosocial intervention (including antipsychotic medications) that all patients received. All patients were between the ages of 15 and 26 and were in close contact with their parents or other relatives. During the 15-month period, relapse rate was equally low (15 percent) for both groups. No differences were found in relapse rate or clinical course for those patients who received the additional family intervention.

This study was a follow-up of the initial one, and lasted 17 to 55 months. Seventy of the initial 76 patients completed the follow-up. Relapse, clinical symptomatology, and medication compliance were assessed for all patients. The authors found that the benefits conferred by the early intervention in the first study did not last. Fifteen percent of the group had relapsed during the 15 month intervention, whereas 64 percent had relapsed during follow-up. Duration of untreated psychosis did not predict the course of the illness, although only 12 individuals had a duration of untreated psychosis lasting more than one year.

The main conclusion drawn from the initial study was that an early, intensive intervention that included both treatment with antipsychotic medications and a psychosocial treatment program had a favorable effect on the course of schizophrenia. The results from this follow-up study, however, suggest that referral to other mental health agencies after early intervention is not sufficient, and that early benefits may disappear rapidly if not continued. Thus, more support is required to continue disease management, medication compliance, and stress management for an as yet unknown critical period.

Analysis of the initial treatment phase in first-episode psychosis.

Power P, Elkins K, Adlard S, Curry C, McGorry P, and Harrigan S (1998). British Journal of Psychiatry 172 (Suppl 33): 71-76.


The results of the EPPIC are one of the first “real world” or naturalistic studies of early intervention, in which principles thought to apply to the detection and prevention of the earliest episodes of psychosis are actually used as part of a treatment plan. The sample size in this study was reasonably large, and patients were followed for three months. Results from the EPPIC study seem to indicate that early detection and intervention may reduce both the number and duration of hospitalizations, that a high number of patients (80 percent) responded to treatment, and that low-dose medication management appeared to be an effective way to treat these illnesses.


Most studies of first-episode psychosis have involved hospitalized patients. Studies suggest, however, that treatment should begin as early as possible for first-episode patients because, with optimal treatment, recovery from the first episode of psychosis can be remarkably good. In this article, the authors describe the Early Psychosis Prevention and Intervention Center (EPPIC), in Melbourne, Australia. The goal of the EPPIC is to manage first-episode psychosis, with an emphasis on early detection, the use of low-dose antipsychotic medications, a less restrictive environment, and psychological intervention, including a strong social support system. The aim of this study was to present findings from a naturalistic or ‘real world’ assessment of the initial treatment phase in early psychosis rather than construe inferences from review studies or trials in hospitalized patients.

Data were examined from the first three months of treatment for 231 consecutive admissions with first-episode psychosis. All patients were between the ages of 16 and 30 and presented with symptoms of first-episode psychosis (eg, hallucinations, delusions, disorganized behavior and/or formal thought disorder). Diagnoses included both affective and non-affective psychosis. Clinical ratings were assessed through the Brief Psychiatric Rating Scale (BPRS), the Health of the Nation Outcome Scale (HoNOS), and the Scale for Assessment of Negative Symptoms (SANS). Ratings were obtained immediately following entry to treatment, as well as at the time of remission of active positive symptoms or at three months, whichever occurred first.

The authors found that about one third of those treated did not require hospitalization. Hospitalizations for the remaining patients were brief, and most (68 percent) occurred within the first day of contact with the EPPIC. Only 16 percent required a second hospitalization, and only four percent required a third hospitalization within the three-month evaluation period; second and third admissions were shorter than the first. Individuals with bipolar disorder were more likely to be hospitalized. Individuals who were hospitalized received higher doses of antipsychotic medications, but they also had a greater rate of reduction on the BPRS.

Hospitalized patients had higher mean HoNOS aggression scores, as well as higher BPRS excitement scores than non-hospitalized patients, although these higher ratings may reflect the fact that more individuals with bipolar disorder were hospitalized. By the end of the three-month period, 80 percent of the sample had responded to treatment, and 63 percent were in remission. Little change was observed in negative symptom scores. The results of this study suggest that both hospitalized and non-hospitalized people with first-episode psychosis can be managed with low-dose antipsychotic medications and experience significant reductions in their symptom ratings within the first three months of treatment.

Lithium maintenance treatment of depression and mania in bipolar I and bipolar II disorders.

Tondo L, Baldessarini RJ, Hennen J, and Floris G (1998). American Journal of Psychiatry 155:638-645.


Most of the literature concerning lithium in bipolar illness has focused on the immediate benefits of treatment, as well as on lithium’s ability to reduce the risk of suicide. Here, the authors found strong evidence that longer delays in initiating sustained treatment with lithium were associated with fewer reductions in morbidity during subsequent treatment. This suggests that patients with bipolar disorder who receive early treatment with lithium may have less long-term morbidity. The article also suggests that, although both subtypes showed marked improvement from lithium, patients with the bipolar II subtype might have the most lasting benefits.


This paper adds to a number of studies that have demonstrated the benefits of lithium maintenance treatment for bipolar patients. The authors wanted to clarify the benefits of lithium in patients with bipolar disorder in general, but more specifically compare its benefits in bipolar type I and bipolar type II disorders.

To do so, the authors studied the clinical research records of patients with DSM-IV diagnosed bipolar disorder (188 with type I and 129 with type II) to determine the frequency and duration of affective episodes and hospitalizations before (mean=8.38 years) and during (mean=6.35 years) maintenance treatment with lithium. The main difference between the subtypes of bipolar disorder, as defined by the DSM-IV, is the presence of mania in type I, and hypomania in type II. The authors point out that bipolar II syndrome is not merely a milder form of bipolar disorder, but is highly associated with morbidity, including suicide. In this study, bipolar II patients were more likely to be women, have a later age of onset (26.8 and 33.3 years for types I and II, respectively), be married and employed, and have a longer time before the onset of lithium treatment.

The authors found that lithium had a protective effect for both forms of bipolar disorder. The time from the end of an episode to the start of the next episode was longer when the patients were receiving lithium. For both subtypes, lithium was also associated with reductions in various measures of morbidity, such as having fewer episodes per year and spending less time ill. Lithium, however, conferred greater benefit to bipolar II patients, reducing both mania and depressive morbidity. Bipolar II patients also had a significantly greater reduction of episodes per year and spent less time ill. During treatment, bipolar II patients had almost 6-fold longer intervals between episodes and were twice as likely as type I patients to have no new episodes.

The authors also studied factors influencing successful lithium treatment and found that gender, family history, education, marital and employment status, age of onset, number of episodes or percentage of time ill, manic, or depressed before lithium treatment, and the presence of a rapid-cycling course before lithium treatment did not affect patients’ improvement. Duration of treatment and serum lithium concentrations were weakly related to improvement. However, the duration of illness before starting lithium maintenance was strongly negatively associated with clinical improvement—starting lithium maintenance earlier predicted greater improvement.

Natural course of schizophrenic disorders: a 15-year follow-up of a Dutch incidence cohort.

Wiersma D, Nienhuis FJ, Slooff CJ, and Giel R (1998). Schizophrenia Bulletin 24: 75-85.


The study’s finding of a relationship between delays in mental health treatment and a longer duration of psychosis supports existing evidence that early intervention with antipsychotic medications improves the long-term course of schizophrenia. Because patients with more relapses tended to do worse, the study further suggests that preventing relapses may be important because it could have the power to change a patient’s mental health outcome by preventing further damage and deterioration.


Because most studies of schizophrenia do not control for patients’ stage of illness, patients with a heterogeneous course, or those with a chronic form of the illness, tend to be overrepresented in these studies. In contrast, this study looked at the course of schizophrenia over 15 years in a group of 82 first-contact cases in the Netherlands. Thus, the authors had a chance to determine whether the illness gradually worsens or improves over time, and they were also able to determine an overall pattern for the course of illness, and whether any factors predicted outcome.

They used a broad definition of schizophrenia (ICD-9 295, 297, 298.3-9), and found a pattern of chronicity and relapses with a high risk of suicide. In this study, 12% of the patients experienced a complete remission after one episode; another 14.6% had two or more episodes followed by a complete remission; 17% had one or more episodes followed by partial remission, usually characterized by anxiety and depression; 11% had only one episode but remained chronically psychotic because they never fully recovered from that episode over the 15-year evaluation period; 33% developed a negative syndrome after one or more episodes; and 12% were lost to follow-up.

The study also found that two-thirds of the subjects had at least one relapse, and after each relapse one out of every six subjects did not recover; one out of every 10 committed suicide; and one out of every seven had at least one episode with affective psychotic symptoms that started on average six years after the onset of schizophrenia. One third of the patients had three or more psychotic episodes, occurring six years after their first episode; with each episode there was an increased probability of chronic psychosis or persistent negative symptoms. With the exception of the first episode, there was also a gradual increase in mean duration of each subsequent psychotic episode (second episode=9 months, third episode=15 months, and fourth episode=27 months).

The authors also studied risk and predictive factors, and found that demographic, illness, and treatment variables at onset of the illness had very little predictive ability, although being married, not dependent on social assistance, having good premorbid functioning, an acute onset and immediate treatment predicted shorter first and subsequent psychotic episodes. In contrast, insidious onset of illness and delays in mental health treatment were risk factors that predicted the longer duration of first or subsequent episodes.

Long-term morbidity associated with delayed treatment of first admission schizophrenic patients: a re-analysis of the Camarillo State Hospital data.

Wyatt RJ, Green MF, and Tuma AH (1997). Psychological Medicine 27: 261-268.


This study found that patients with schizophrenia who were not treated with antipsychotic medications during their first admission were doing worse at followup than patients who had initially been treated with antipsychotic medications; both groups had been released from the hospital within six months. The study thus suggests that early treatment with antipsychotic medications both decreases the immediate morbidity associated with schizophrenia, and prevents detrimental changes possibly related to prolonged untreated psychosis.


The authors conducted a retrospective analysis of records from the Camarillo State Hospital study, which was conducted by Philip May and colleagues in the late 1950s and early 1960s, a time when the issue of whether there were immediate benefits from antipsychotic medications were still being debated. That study randomly assigned first-episode patients with schizophrenia to one of five treatment groups: milieu therapy alone, psychotherapy alone, ECT, antipsychotic medications alone, and antipsychotic medications plus psychotherapy.

This study re-examined some of these data, reasoning that those first-episode patients initially treated with milieu- or psychotherapy alone who were stable enough to be discharged from the hospital within six months might be considered to have a spontaneously remitting form of schizophrenia, and thus be expected to do well on followup. Patients were thus divided into two groups: those who had initially been treated with antipsychotic medications (the antipsychotic medications as well as the antipsychotic medications plus psychotherapy groups) and those who had initially been treated without antipsychotic medications (milieu- and psychotherapy alone groups); data from patients initially treated with ECT were not analyzed. Because patients in the study by May and colleagues who had not been discharged within six months were often reassigned to different treatment groups, the authors chose six months as the cutoff date for defining spontaneously remitting patients. Out of 89 patients given either milieu- or psychotherapy alone, 25 were discharged within six months. In contrast, 71 of 91 patients given antipsychotic medications were discharged in the same time period. Today, this improved immediate outcome from antipsychotic medication is, of course, well documented.

These patients were followed for two years after their initial discharge. The study found that, by the second year of followup, those patients initially treated with antipsychotic medications required significantly fewer rehospitalization days than the group of initially non-medicated patients. And, although in the first year of followup the initially medicated patients required significantly more medication in chlorpromazine equivalents, this was not true by the second year of followup.

In order to assess patients’ continued ability to function, 11 patients from each group were matched for age, educational status at first admission, race, and gender, and their Global Assessment of Functioning (GAF) scores were estimated over a period of six to seven years following discharge. Again, those patients initially treated with antipsychotic medications were found to be functioning at a significantly higher level, as measured by GAF scores, than patients initially not treated with antipsychotic medications.

Prediction of psychosis: A step towards indicated prevention of schizophrenia.

Yung AR, Phillips LJ, McGorry PD, McFarlane CA, Francey S, Harrigan S, Patton GC, and Jackson HJ (1998). British Journal of Psychiatry (Supplement) 172 (Suppl 33): 14-20.


Although there were only a small number of at-risk individuals in this sample, 40% of them made the transition from being at-risk for psychosis to being psychotic within a six month followup period. This high rate raises the question of whether specific early interventions—including antipsychotic medications and psychosocial intervention—should be targeted at individuals showing these at-risk symptoms, especially if their baseline testing scores indicate them to be particularly vulnerable. The fact that five of the eight individuals who became psychotic in this sample did so within the first month of followup further suggests that this period is one of particularly high risk. At the least, close followup of identified vulnerable individuals could minimize the duration of untreated psychosis, hence improving outcome for these individuals.


It has long been acknowledged that people who become psychotic often experience some changes in behavior in the period preceding the onset of psychosis; this period has been known as the prodrome. The goal of this study was to identify individuals experiencing prodromal symptoms and thus likely to become psychotic within a followup period, with the eventual goal of either preventing full-blown psychosis or minimizing their symptoms and disability.

The study recruited individuals with psychological features which seemed to indicate the presence of a prepsychotic symptoms, e.g., individuals at risk for psychosis. Individuals were recruited if they showed either “attenuated” or “transient” psychotic symptoms. The former was defined as the presence of ideas of reference, odd beliefs or magical thinking, perceptual disturbance, odd thinking and speech, paranoid ideation, or odd behavior or appearance, apparent for at least a week. “Transient” psychotic symptoms were identified in individuals who had already experienced psychotic symptoms—hallucinations, delusions, unusual thought content, suspiciousness or formal thought disorder—but whose symptoms had lasted less than a week and resolved spontaneously. In addition, the authors considered individuals who had a family history of a first-degree relative with a psychotic disorder, as well as individuals with schizotypal personality disorder as being at-risk for developing psychosis.

The authors hypothesized that the rate of transition to psychosis would be 20-30% of the cases over a 12-month period, and that people who became psychotic would display more aberrant results on testing measures than people who did not; furthermore, they did not expect the two groups to differ on baseline measures. In order to assess psychopathology, the authors administered the Brief Psychiatric Rating Scale (BPRS), the Scale for the Assessment of Negative Symptoms (SANS), the Hamilton Rating Scale for Depression (HRSD), and the mania rating scale. When admitted to the program, individuals also received the Quality of Life (QLS) scale and the Global Assessment of Functioning (GAF), both of which measure functioning and disability. All of these tests were administered before the development of frank psychosis.

The sample consisted of the first 20 people recruited into the study, who were followed for six months. At the end of the six months, eight individuals (40%) had become psychotic—five within the first month of followup, one between months three and four, and two between months five and six. This suggests that the first month after seeking help is a period of particularly high risk for at-risk individuals. Significantly, differences were found on baseline scores for the BPRS, HRSD, and GAF between the eight individuals who became psychotic and the 12 who did not. This suggests that testing measures that assess psychopathology and functioning can actually distinguish between individuals likely to become psychotic within a six-month period and individuals who are simply at-risk for psychosis.