OBJECTIVES: Presently, schizophrenia guidelines recommend waiting for 8 weeks before considering a patient as non-responder. This study aims to assess whether the early response (week 2,3 and 4) can predict the outcome at 8^th week for the treatment of first episode schizophrenia (FES) using olanzapine.

METHODS: A prospective observational study was conducted in 36 adult FES patients with four points follow up (week 2,3,4 and 8). Patients with any other psychiatric co-morbidities and/or taking any antipsychotics were excluded from the study. A reduction of 25% in PANNS score from the base line at week 3 and 30% at week 4 were considered as early response. A reduction of 50% at week 8 was considered as responders. Receiver Operating Characteristics (ROC) curves were implemented to detect the optimal threshold of early response (AUC>80%) at 2^nd, 3^rd and 4^th weeks. To identify the variables that are predictive of treatment responses at Week 8, binary logistic regression was performed.

RESULTS: The mean total PANSS baseline score of the study population was 106.66 (95% C.I, 100.4-112.9). Week 2 (AUC=50.5, P=0.964) and week 3 (AUC=64.9, P=0.136) responses failed to predict the 8^th week response. Week 4 response (AUC = 92%, P<0.001) can be taken for the prediction of 8th week response (specificity=72%, sensitivity=100%, Positive Predictive Value=61.1%, Negative Predictive Value=100% and OER was <29.4% reduction). 25 patients (69%) achieved more than 50% reduction (responders) in PANSS score after 8 weeks of treatment. Independent predictors for good treatment responses at Week 8 were female gender (P=0.032), a higher educational level(P=0.021) and non-smoking(P=0.014) and non-alcoholics (P=0.048).

CONCLUSIONS: Our study suggests that patients with early response at week 4 are likely to achieve positive response after 8 weeks. This may help in appropriate clinical decision making for early non-responders.