Lack of association between duration of untreated psychosis and outcome in an Indian cohort

The duration of untreated psychosis (DUP), defined as the time from manifestation of the first psychotic symptom to initiation of adequate antipsychotic drug treatment, has been shown to have a robust association with poor outcome of first-episode psychosis 1. However, several methodological problems afflict research in this area. Moreover, although some Asian countries have embarked on early intervention programmes for psychosis, data on influence of DUP on outcome of psychotic disorders (particularly schizophrenia) from these countries is extremely limited. This prompted us to examine the influence of DUP on outcome among patients of first-episode schizophrenia seeking treatment at the psychiatric outpatient clinic of a multi-speciality hospital in north-India.

Patients aged 18-60 years, with a DSM-IV diagnosis of schizophrenia, in their first episode, seeking treatment for the first time, were recruited after obtaining written informed consent. Patients with comorbid psychiatric disorder, substance abuse/dependence (except nicotine), major physical illness, organic brain syndrome or mental retardation were excluded. Diagnoses were established using the Structured Clinical Interview for DSM-IV Axis I Disorders - Clinician Version. DUP was defined as the interval between the onset of psychotic symptoms and initiation of “adequate treatment”, defined as treatment with antipsychotics at adequate doses (minimum of 300 mg/day chlorpromazine equivalents), for 6 weeks or more. Onset of psychotic symptoms was determined using the Instrument for the Retrospective Assessment of Onset of Schizophrenia 2. Initiation of treatment was ascertained from information obtained from patients/relatives and scrutiny of medical records. Baseline assessments also included the Positive and Negative Syndrome Scale (PANSS); the Schedule for Assessment of Psychiatric Disability (SAPD), an Indian modification of the WHO-Disability Assessment Schedule; the Global Assessment of Functioning scale (GAF) for past-month functioning, the Lehman’s Quality of Life Interview - Brief Version (QOLI). Subsequently, patients continued their treatment and were contacted 6 months after the initial assessment. A modified version of the WHO-Life Chart Schedule (LCS; 3) was used as the primary measure of the interim course and 6-month outcome. The PANSS, GAF, SAPD and QOLI were all reapplied as secondary measures of outcome.

Consecutive sampling over 8 months yielded 38 patients with first-episode schizophrenia fulfilling selection criteria; 8 of these could not be included in the study, but their clinical/demographic characteristics were comparable to the patients recruited. Consequently, the initial sample had 30 patients; 8 more patients dropped out in the intervening 6 months; thus the follow-up sample consisted of 22 patients.

Patients were older than expected (mean 29-32 years); there was also a slight excess of females. Paranoid schizophrenia being the commonest subtype could have contributed to the later age of onset. Patients were mostly educated and from urban backgrounds. There were no differences between baseline and follow-up samples, indicating that the 8 drop-outs at this stage did not affect the results concerning outcome.

The mean DUP was 47.30 (SD 40.44) weeks for the baseline sample and 49.32 (SD 42.95) weeks for the follow-up sample. The DUP varied from 6 to 180 weeks for both samples. Predominance of urban, educated subjects could have contributed to the unusually short DUP of this sample.

The influence of DUP on outcome was examined using Spearman’s correlation coefficients. The modified LCS was the primary outcome measure; secondary measures of outcome included the PANSS, GAF, SAPD and QOLI. No significant association with DUP were evident with either primary or secondary outcome measures. Dichotomized outcome analysis was also attempted by subdividing the follow-up sample into “short DUP” (n=10) and “long DUP” (n=12) groups, using the median value of 36 weeks as the cut-off, and comparing the same outcome parameters using Mann-Whitney tests. This analysis also failed to reveal a significant association between DUP and outcome.

Although this study met most of the recommended quality control measures 1, including standardized determination of diagnosis and DUP, a reasonable follow-up rate, use of multiple measures of outcome, and of non-parametric analyses, it fell short on several aspects, including the small/restricted sample, non-blind assessments, the relatively short follow-up and the inability to control for potential confounds. However, these weaknesses could not entirely account for the lack of an association between DUP and outcome, especially since several other studies 4 with similar designs and sample sizes have yielded positive associations.

Alternative explanations for the lack of an association could be proposed. Variability in ascertaining DUP, possible influence of confounders, the likelihood of DUP being a marker, not a determinant of outcome, have all been cited as reasons for the failure to replicate a positive association between long DUP and poor outcome. Even in studies reporting a positive association, the effect is modest and correlational, not causal 5.

However, the current study highlights another related debate, that concerning a “threshold” of DUP, exceeding which it inevitably predicts poor outcome. A non-linear relationship and lack of association between effect sizes and the cut-off points of long/short DUP has led to the proposal that deleterious effects of untreated psychosis occur very early, possibly in the late prodromal phase. However, application of very short cut-offs leads to confounding between DUP, outcome and diagnosis. Samples with very short DUP are likely to be contaminated by good-prognosis cases (e.g., acute psychosis), leading to better outcome. Employing DSM-IV criteria (as in this study) to diagnose patients with schizophrenia obviates this problem to some extent, as the inclusion of a 6-month duration of illness criterion for DSM-III-R /DSM-IV diagnoses of schizophrenia eliminates much of the predictive effect of DUP. Moreover, several studies have indicated that if there is a threshold period of DUP, then this is greater than 1 year and may be much longer than that. Among all patient groups, those with schizophrenia with DUP exceeding 1 year have the poorest outcome 1,5. Finally, studies from developing countries have usually reported DUPs much longer than the average of 1-2 years found in studies from developed countries.

The results of this study do not necessarily argue against the use of measures to reduce DUP, which would be justified simply by humane reasons for ameliorating unnecessary suffering caused by untreated psychosis 6. However, they do highlight the prevailing uncertainties in this area. They also suggest the need for further research to enable (reduction of) DUP to be a reasonable goal of early prevention programmes and the need to maintain a balance between enthusiasm and proper research evidence, particularly in developing countries 7. The already limited mental health resources in these countries makes it imperative that such prevention programmes be based on sound, locally based research data linking DUP to outcome.