There are four phases.

Phase I: In the initial treatment phase (Phase 1), patients will be randomized to one of the three atypical antipsychotics or placebo in the ratio 100:100:100:150 respectively. After two weeks, the investigator can move the patient to the next phase because of lack of efficacy or tolerability. At week 12, the investigator can decide whether the current medication is sufficiently optimal or it would be more beneficial to try another randomized medication.

Phase 2: Phase 2 starts when the patient is randomized to a second medication, i.e., olanzapine, quetiapine, risperidone, or citalopram. Patients will be randomized from an antipsychotic treatment to another antipsychotic treatment or citalopram in the ratio 3:3:2, or from placebo to an antipsychotic treatment or citalopram in the ratio 1:1:1:3 respectively. Therefore, 50% of patients who took placebo in Phase 1 will be randomized to an antipsychotic in Phase 2, and 50% will be randomized to citalopram in Phase 2. After the initial two weeks in Phase 2, the investigator can move the patient to the next phase, due to lack of efficacy or tolerability. After the patient has been on the Phase 2 study drug for approximately 12 weeks, the investigator can decide whether the current medication is sufficiently optimal or whether it would be more beneficial to try another randomized medication.

Phase 3: Phase 3 is randomized open-label treatment of one of the medications not previously received, i.e., olanzapine, quetiapine, risperidone, or citalopram. Treatment failures to the second treatment can be switched to a third open-label treatment. During Phase 3 patients will be maintained on their treatments openly and managed clinically until week 36.

If the investigator determines that the patient's response is not sufficiently optimal to the randomized open-label medication, then after the first two weeks of Phase 3, the investigator can prescribe another medication (of the investigator's choice) to the patient. If this occurs then patients are classed as being in the Open-Choice Phase.

Open-Choice Phase: The Open-Choice Phase can be entered at anytime during the 36-week study and directly from any of the three phases. There are four reasons a patient can enter the open choice phase:

• Withdrawal from Phase 1 or Phase 2 with the patient or surrogate decision-maker refusing to proceed to the next randomized phase;
• Withdrawal from Phase 3;
• Withdrawal from current study drug from any of the three previous phases due to antipsychotic medication no longer being required in the opinion of the investigator; or
• Withdrawal due to concomitant treatment with an exclusionary medication.

The Open-Choice Phase is designed to keep patients monitored in the trial for the 36-week duration.