Abstract
Schizophrenia is a severe psychiatric disorder that often imposes enormous burden on the sufferer, carers, health care system and society at large. Proper treatment leads to substantial reduction in the burden and suffering due to the illness [1]. It was established that the longer the psychotic symptoms proceed unchecked by medication the greater was the likelihood of profound clinical deterioration [2], [3]. This led to a surge of research activity on early intervention in the illness.
The rapidly growing interest in early intervention has three preventive foci. The pre-psychotic phase is when most psychosocial impairment develops but the specifics of treatment remain difficult to research and apply. The other two foci are to minimize the total duration of illness and duration of untreated psychosis. The duration of untreated psychosis (DUP) is a more realistic immediate target for strategies of early detection and intervention, although the total duration of untreated illness (that includes the overt psychotic phase as well as the nonpsychotic prodromal stages) may be more critical [4].
The concept of DUP attracted much interest because of its possible relationship to treatment outcome and implications for preventive efforts in schizophrenia [5]. Many studies demonstrated a link between DUP and both short and long-term outcome in schizophrenia [6–8], but some contested this claim [9], [10]. The relationship of DUP to outcome was strongest in the initial months of psychosis [11]. The concept of a ‘limited window of opportunity’ in the early course of illness, a critical period when putative deficit factors are at their peak and the intervention has maximum benefit, was put forth. Studies showed that outcome was significantly enhanced by more intensive treatment only if the DUP was less than 6 months [12], [13]. It was not known if this relation between DUP and treatment response would hold if the illness was untreated for many years.
It was unclear what determined DUP in schizophrenia and why long DUP predicted poor outcome [11]. The relationship between DUP and outcome is a much debated issue. Some studies state that DUP remained a significant predictor of outcome after adjusting for the effects of social and clinical variables [13], [14]. Alternatively, others argued that the association between DUP and outcome may not be a direct one, as DUP itself may be determined by demographic, clinical and premorbid factors that had a direct effect on outcome [10], [15]. The degree of direct association between DUP and outcome in patients with prolonged untreated illness is still not clear.
In this paper we present an analysis of data aiming to study the relation of DUP to treatment outcome in a cohort of never-treated schizophrenia patients ill for many years and the association of outcome with social and clinical variables.
Method
The detailed method of the data collection was presented earlier [16]. The cohort of untreated schizophrenia was derived from a twostage door-to-door survey of a population of 101 229 conducted in 1985–1986 in the city of Chennai in southern India. Schizophrenia was diagnosed using criteria defined by ICD-9 [17]. The Present State Examination [18] was used to assess symptoms. The clinical history and sociodemographic details were evaluated using the Psychiatric and Personal History Schedule (PPHS) used in an earlier study in India [19]. The Psychiatric and Personal History Schedule was developed from the Psychiatric History Schedule and Social Description Schedule (PHSD) [20]. The same instruments were used for evaluation at the end of one year of treatment. The information was gathered through interviewing the patient and the primary caregiver in the family who was living with the patient from the time preceding onset of psychosis. The survey detected 265 patients with schizophrenia, including 75 patients who never received any medical treatment during their illness. Complete data was unavailable to facilitate follow-up for three of the untreated cohort. Among the 72 subjects, treatment could be initiated in 49 and followed up for one year. They were treated as outpatients with typical antipsychotic medication (chlorpromazine/haloperidol) routinely dispensed at the public general hospital where the study was conducted. All patients lived with their families who supervised medication intake. The onset of psychosis was measured as the time the psychotic symptoms (delusions, hallucinations, disorganized behaviour, thought disorder) were first noted by the caregiver. In some cases, time anchors (time of significant events like birth of a child, marriage or death in the family, a public event) were used to time the onset of illness. The duration of untreated psychosis measured was the continuous period between onset of psychosis and the time of initiation of treatment after case identification. The duration was measured in years (rounded off to the nearest year).
Outcome was measured in three domains, clinical, social and occupational functioning, using the scores on specified items from PSE and PPHS [21]. The method of measurement of these outcomes was similar to two earlier studies on schizophrenia in India, namely, the International Pilot Study of Schizophrenia [22] and the study of factors associated with course and outcome of schizophrenia [19]. Clinical outcome was defined as a composite measure of the pattern of course and time spent in psychotic state during the follow-up period. The presence of interaction with the family, friends and extended social network as well as frequency of contact determined the social outcome. The occupational outcome took into account the duration of employment and the quality of work performance during the one year of follow-up. The home-making activity of female patients and paid employment of males were assessed similarly. The outcome in each domain was classified as best, intermediate and worst. The best and intermediate outcomes were combined to form the good outcome group [22]. For the purpose of this study a global outcome was computed to represent a composite of the outcome in the three domains. Global outcome was considered ‘good’ if there was good outcome in at least two out of the three domains.
The data was analysed using SPSS (version 5.0) [23] and Diagnostic and Agreement Statistics [24]. All the interval data were on a nominal scale and analysed using χ2 tests of significance, with Fisher's exact tests performed where appropriate. Odds ratio and their 95% confidence intervals were also reported.
Results
Clinical outcome at the end of one year was good in 14 patients (29%) with 13 recovering with or without residual personality change. A good social outcome was measured in 17 patients (35%). The occupational outcome was good in 25 (51%) with seven of the patients unemployed at intake gaining employment. The global outcome was good in 15 patients (31%). Table 1 shows comparison of patients of good with poor global outcome on sociodemographic and clinical factors measured at intake. There was no significant difference between the two groups on the socio-demographic variables. Patients with poor global outcome were significantly more likely to have PSE syndromes of formal thought disorder and delusion.
Comparison of patients with good and poor global outcome
The patients were divided into four groups based on DUP: less than 2 years (n = 6); 2–5 years (n = 13); 6–15 years (n = 12) and more than 15 years (n = 18). The percentage of patients with good outcome in each domain for the four DUP groups, respectively, were: clinical outcome, 50, 46, 25 and 11 (Fisher's exact, p = 0.093); social outcome, 17, 23, 50 and 39 (Fisher's exact, p = 0.443); occupational outcome, 83, 62, 50 and 33 (Fisher's exact, p = 0.155); and global outcome, 50, 46, 33 and 6 (Fisher's exact, p = 0.026). When the patients were divided into two groups, one with DUP of 5 years or less (n = 19) and the other with longer duration (n = 30) and compared, the group with shorter duration were more likely to have good clinical outcome than the longer duration group (47%
Discussion
This is one of the few studies that examined subjects with DUP of so long a duration. The presence of a significant number of never-treated schizophrenia patients living in the community in the study region facilitated this study.
Measurement of DUP
We measured ‘untreated psychosis’ rather than ‘untreated illness’ as the informants were surer of the beginnings of obvious psychotic manifestations than more subtle non-psychotic ones. The ability of the informant to estimate accurately the time of onset after many years of illness could still be doubted. We adopted some measures described earlier to reduce this error. We were of the opinion that the timing of the onset of illness by the caregiver living with and involved in the care of the patient from the outset would be reasonably accurate.
Prolonged DUP and response to treatment
In the developing world the issue of treatment delay is often one of a permanent nature, that is patients never accessing any treatment [16], [25]. These populations provide grounds to explore the relevance of prolonged DUP to treatment response, as done here. An encouraging observation in the study was the notable degree of success with treatment despite many years of untreated illness. The steep fall in the proportion of patients with good outcome with increasing DUP, significantly so after 5 years of illness, suggested that though the illness had been present for many years there was a distinct benefit of treatment in the ‘early chronic’ stages of the illness.
Relation between DUP and outcome
In discussing the association between DUP and outcome and the role of other factors that influence this association, it would be relevant to present here our earlier observations made on this never-treated cohort [16], [21]. We presented the extended family structure as the important factor associated with the never-treated status more than other social or clinical characteristics [16]. In the subsequent follow-up report [21] the factor of extended/joint nature of the family (along with the family's awareness of nature of illness) again featured as the prominent reason associated with failure to take treatment. Factors like premorbid adjustment were not studied but the level of education was related to nontreatment [16]. This may be taken as an indicator of premorbid functioning but in the community studied, factors like social and economic limitations, played a significant role in determining opportunities for education. Thus it seemed that familial factors were crucial in determining whether a patient accessed treatment or not.
The association between prolonged DUP and outcome evidenced in this study concurred with studies on shorter DUP. The relation between DUP and outcome held good even after many years of illness. The lack of relation between DUP and social and occupational outcome could be due to the prevailing social milieu. Many of the subjects in the cohort lived with large families that were joint, extended or ‘functionally extended’ (family members in separate households but functioned as one family) in structure. Such a scenario could have influenced social outcome when measured by the two indicators – presence of interaction with family and its frequency. The limited social demands from the large family also could have minimized social dysfunction [26]. The relatively high rates of employment could be related to the necessity to engage in remunerative activity because of economic pressures and the absence of social security or disability benefits, as noted in a study on schizophrenia patients from India [26].
The independence of relationship between prolonged DUP and outcome could not be explored in this study. The size of the sample available restricted application of statistical methods that would help find an answer to this issue. No doubt, it would be an enormous task to recruit patients untreated for many years in sufficient number to cater to such a need. However, some direction could be obtained from this study. But for presence of delusions and thought disorder at intake, patients with poor outcome did not differ significantly from the good outcome group. This suggested the possibility that prolonged DUP was related, at least partially, to outcome independent of other confounders, but no firm conclusions could be drawn.
Conclusions
We conclude that in patients with schizophrenia untreated for many years the treatment response was related to duration of untreated psychosis and presence of active psychotic symptoms at intake but not to any sociodemographic variable. In community-based health services every effort should be taken to initiate treatment to untreated patients as early as possible, as such treatment makes a difference even though they were ill for many years.
Footnotes
Acknowledgements
This study was supported by the Indian Council of Medical Research (India) and Schizophrenia Research Foundation (India).