The Ten Country Study - Schizophrenia

The IPSS was followed by another study coordinated by the WHO, the Ten Country Study described above (Sartorius et al., 1986; Jablensky et al., 1992). Data on outcome after 2 years were obtained for 78% (n=1078) of the original sample. The outcome for patients diagnosed with broad schizophrenia was more favourable in developing countries than in developed countries, for example, on five of the six measures of 'best outcomes' examined.

Though limited, these reports support the view that patients fare better, on average, in the developing countries. Additional evidence from follow-up at the Chandigarh sites raises the possibility that there may be greater variability in outcomes in developing countries (Mojtabai et al., 2001).

In Chandigarh, there was a group of subjects that did poorly in the first 2 years (n=15); nearly half had died by the time of the 15-year follow-up. In two patients, malnutrition was a contributing cause of death. These findings are generally consistent with another study conducted in Brazil. Menezes and Mann (1996) carried out a 2-year follow-up of a prevalent sample of 120 patients with schizophrenia who had been consecutively admitted to psychiatric hospitals of a catchment area of Sao Paulo, Brazil and found an eightfold increased risk of dying compared with the general population (standardized mortality ratio 8.4), and a dramatically increased risk for suicide.

Several explanations of the fundamental differences in outcome between developed and developing countries in the Ten Country Study beg consideration.

{embed="ref/55"}

Artifact is one potential explanation. It is possible that the disease 'mix' in developing countries (higher incidence of non-S+ schizophrenia, higher proportion NARP) is at the foundation of the improved prognosis. However, improved or better course and outcomes apply to both acute and nonacute onset disease, and that broadly and narrowly defined (Jablensky et al., 1992).

The 'mix' may be different, but it does not explain all of the difference in prognosis. Other explanations relating to background characteristics of the underlying samples and ascertainment bias have been considered (Jablensky et al., 1992; Hopper and Wanderling, 2000); however, none has been shown to account for the difference.

{embed="ref/54"}