Iboga research remains controversial (Reference VastagVastag, 2005). Alper & Glick refer to clinical evidence from ‘two larger studies’. However, in these papers it is acknowledged that there is no substantial properly conducted trial available. These studies are difficult to interpret and mainly report treatment of acute opiate detoxification – not the topic of our review. It remains unclear how the patients were selected from the sampling frame. The first series includes a subset of 33 patients treated in hotel rooms or apartments in the USA and The Netherlands between 1962 and 1963 and 1989 and 1993 respectively (Reference Alper, Lotsof and FrenkenAlper et al, 1999). These are referred to as case reports in our review. The second series of 32 patients was equally difficult to rate. Thirty-two patients were treated in a private facility in ‘offshore studies’ but substantially more patients may have been treated (Reference VastagVastag 2005) and information on inclusion and exclusion criteria for the individual patients is not available. In an evidence-based review, this would be difficult to accept. For instance, in meta-analyses of randomised controlled trials, open-label trials would not be included even if they were positive. With regard to longer-term abstinence, Mash et al (Reference Mash, Kovera and Pablo2001) suggested that ‘many’ patients were successful but the supporting data were not presented. Notably, in a series of 27 patients, possibly a subset of the 32, reported 1 year earlier, longer-term outcomes were not presented either (Reference Mash, Kovera and PabloMash et al, 2000).

With regard to safety concerns, cerebellar degeneration has been reported in rat experiments (Reference Xu, Chang and SikkerXu et al, 2000). The absence of similar findings in mouse experiments does not imply that ibogaine is safe in humans. Reports of eight deaths after ibogaine use between 1990 and 2006 have been compiled (http://myeboga.com/fatalities.html). This source notes that more deaths might have occurred but might have not been reported owing to the ‘underground nature of ibogaine treatment’. One death occurred at a dose of 4.5 mg/kg orally, a much lower dose than used in the rat experiments. Health problems associated with substance misuse or potentiation of ibogaine toxicity when used with heroin have been implicated. Alper et al (Reference Alper, Lotsof and Frenken1999) when referring to one of the fatalities quoted Vocci, the Director of the Medication Development Division, National Institute on Drug Abuse (MDD–NIDA), that ‘this incident was a significant factor in the decision not to pursue a clinical trial of ibogaine following the NIDA review meeting held in March of 1995’.

The antagonism at the α₃β₄ nicotinic receptor should have been highlighted as a therapeutic target for the modulation of drug seeking but this would not have changed our conclusions since we have not doubted the potential efficacy of ibogaine. However, we maintain that ibogaine and iboga extracts may not be safe and thus should not be recommended. Ibogaine derivatives with an improved therapeutic index may prove clinically useful in the future. These are likely to be synthetic, thereby leaving the realm of complementary medicine.