We have tested deep rTMS treatment as an add-on to medication for refractory auditory hallucinations. The main finding of this study is that low-frequency deep transcranial magnetic stimulation to the left temporoparietal cortex given in the protocol mentioned above had no statistically significant effect on auditory hallucinations.
These findings are in agreement with Loo et al.'s findings , which suggested based on examinations of individually-controlled trials that a substantial proportion of rTMS studies of the treatment of auditory hallucinations did not find rTMS to be superior to sham stimulations. Our findings also correlate with a double-blind, randomized, sham-controlled study of 62 patients completed in March 2011 , which concluded that low-frequency rTMS administered to the left temporoparietal cortex (shown to be the site of maximal hallucinatory activation) is not more effective for medication-resistant auditory hallucinations than sham treatment. Tranulis et al.  conducted a meta-analytic study of 10 sham-controlled studies of low-frequency rTMS over the left temporoparietal cortex and found a medium statistical effect size. They reached the conclusion that rTMS is an effective tool as a complementary treatment for auditory hallucinations. On account of this contradictory data, the efficacy of low frequency rTMS over the left temporoparietal region is a matter that deserves further investigation.
It is important to note that the distribution of the H1 coil's magnetic field is both deeper and wider than the traditional figure-eight coil used in previous studies . The H coils induce an effective field at a depth of approximately 3 cm below the skull, compared to less than 1.5 cm for the standard TMS figure-eight coil . We would expect the H coil used in this study to be more effective than a figure-eight coil in treating auditory hallucinations due to its deeper penetration into the brain and because subcortical structures (the thalamus, for example) may be involved in the generation of auditory hallucinations . However, we failed to achieve statistically or clinically meaningful improvements. One possible theoretical explanation may be that low frequency stimulation of subcortical structures counteracts effective stimulation of cortical zones (for example, the temporoparietal region).
The results of this study contradict our original open study  using the same deep TMS coil applied over the same location and using similar treatment parameters. This contradiction may stem from the fact that the placebo effect probably caused the positive results in the open study . The placebo effect was weakened in this study because patients were oblivious to the kind of treatment (real/sham) they were given.
There are a few possible explanations for the inefficacy of the TMS presented in this study. The left temporoparietal cortex may not be the origin of auditory hallucinations in patients with schizophrenia, a possibility that may be supported by the fact that functional imaging studies showed increased activation of the right brain area in some patients suffering from auditory hallucinations. Other reasons may be inadequate/insufficient treatment parameters (including frequency, session duration, number of sessions, total pulses, or stimulation pattern).
The major limitation of our study is the small sample size. The negative result of this study may be false on account of the small sample size.