• Hallucination change score [ Time Frame: Measured at every week ] [ Designated as safety issue: No ]
  
• Clinical Global Improvement Scale [ Time Frame: Measured at every week ] [ Designated as safety issue: No ]
  
• Frequency subscale of Auditory Hallucinations Rating Scale [ Time Frame: Measured at baseline and every week ] [ Designated as safety issue: No ]
  

• Summed scores of Auditory Hallucination Rating Scale [ Time Frame: Measured at baseline and every week ] [ Designated as safety issue: No ]
  
• PANSS composite positive symptoms scale [ Time Frame: Measured at baseline and every week ] [ Designated as safety issue: No ]
  
• PANSS composite negative symptom scale [ Time Frame: Measured at baseline and every week ] [ Designated as safety issue: No ]
  
• PANSS total score [ Time Frame: Measured at baseline and every week ] [ Designated as safety issue: No ]
  
• California Verbal Learning Test (CVLT) [ Time Frame: Measure at baseline and after Week 4 ] [ Designated as safety issue: Yes ]
  

This study is an extension of an ongoing clinical trial (ClinicalTrials.gov identifier NCT 00308997) that was initiated in 2006. The primary objective of the ongoing clinical trial (hereafter called the "parent trial") is to determine efficacy of rTMS in curbing auditory hallucinations when delivered to a part of the left temporal lobe called Wernicke's area and a corresponding region in the right temporal lobe. The parent trial appears to show robust effects for active rTMS compared to effects of sham stimulation. However, observed responses following active rTMS have often been incomplete. Moreover, in some cases there has been a subsequent return of symptoms 1 to 6 months after the trial ended.

We consequently have initiated a re-enrollment trial where patients who have participated in the parent trial and demonstrated an incomplete response or a subsequent return of symptoms may return to receive additional active rTMS. We hypothesize that efficacy of suppressive rTMS will be enhanced if directed simultaneously to right/left Wernicke's area (the site used in the parent trial) as well as to a second site located in the opposite middle temporal cortex. Roughly half of subjects in the re-enrollment will be randomized to receive active rTMS to right/left Wernicke's area plus active rTMS to opposite hemisphere middle temporal region, while half of subjects will be randomized to receive active rTMS to right/left Wernicke's area plus sham rTMS to opposite hemisphere middle temporal region. The position of the middle temporal regions will be determined by two recently completed brain imaging studies of auditory hallucinations suggesting that activation in these sites triggers auditory hallucinations. The two-position design will allow us to determine if active rTMS delivered to the middle temporal cortex is superior in amplifying efficacy of active rTMS targeting Wernicke's area and in reducing auditory hallucinations to sham stimulation to the same site. The re-enrollment protocol will utilize two rTMS devices simultaneously where one directly triggers the other.

Inclusion Criteria:

• Previously enrolled in our "parent" rTMS trial with passage of at least six months since last received active rTMS

Exclusion Criteria:

• Active substance abuse or alcohol abuse
• Pregnancy
• Dose or type of psychiatric medication changed within the 4 weeks prior to study entry
• Recent head trauma, seizures, or significant unstable medical condition