ABSTRACT
A cornerstone argument supporting early detection is drawn from studies that have investigated the links between duration of untreated psychosis (DUP) and prognosis and reported that subjects with a long DUP have a poorer clinical outcome than those with a shorter delay between onset of psychosis and first treatment.1,2 The rationale for early intervention is that reducing DUP may have a favourable impact on the subsequent course of the illness. However, the existence of a causal link between DUP and outcome is not perfectly established, as these findings are from observational studies where the delay between onset of psychosis and first treatment was not randomly distributed with regard to the other characteristics of the disease. Thus, factors delaying treatment seeking, such as poor premorbid adjustment or severe negative symptoms, may also independently predict poor outcome. Hence, the association between DUP and poor outcome may be a spurious one, confounded by these factors.2-5
Characteristics associated with duration of untreated psychosis
The first question is to assess what characteristics are associated with DUP in naturalistic settings. This issue was explored by us in the Bordeaux First Episode study.6 This population-based study was carried out on first-admitted subjects presenting with psychotic symptoms consecutively hospitalized in the acute wards of Bordeaux’s psychiatric hospital serving an urban geographical area with 250 000 inhabitants. DUP was categorized according to the median into ‘short’ DUP (< 3 months) versus ‘long’ (≥ 3 months) DUP. Univariate analyses showed that patients with a long DUP were more likely to have a low educational level, to have had poorer global functioning in the previous year, and to present with a more severe global clinical state and more prominent negative symptoms at admission. A long delay was also associated with a diagnosis of schizophrenia broadly defined, and with a family history of psychiatric hospitalization in first and/or second-degree relatives. Multivariate analyses were then used to explore which characteristics independently predicted a long delay between onset of first psychotic symptoms and first treatment. Four such characteristics were found
independently: family history of psychiatric hospitalization (adjusted odds ratio (OR) = 12.1, 95% confidence interval (CI) 1.15 to 9 7.0, P = 0.02), a low educational level (adjusted OR = 7.7, 95% CI 1.0 to 50.0, P = 0.05), a low Global Assessment of Functioning Score during the previous year (adjusted OR = 0.93, 95% CI 0.86 to 0.99, P = 0.04) and, at trend level, a high Clinical Global Impression score (measuring severity of illness) at admission (adjusted OR = 4.0, 95% CI 0.87 to 18.3, P = 0.07). No interaction was found between these variables. Thus subjects with a long DUP are already different from the other subjects when presenting for treatment for the first time, and importantly, the characteristics differentiating those subjects have also been found to be associated with poor outcome in previous studies.
