Researchers at University Hospital Bonn have carried out the first clinical use of an expanded provocation protocol designed to determine, in a single patient, whether reported drug side effects stem from the drug's pharmacology or from the patient's own negative expectations. The proof of concept n-of-1 study, published in Frontiers in Pharmacology on September 18, 2026, describes a "mixed open blinded crossover balanced placebo provocation protocol": five structured sessions in which the patient received an active substance or a placebo, paired with true or false information about which one it was, under direct clinician supervision throughout.
The design extends a decades-old experimental method, the balanced placebo design, that researchers have long used in laboratories to separate a drug's pharmacological effect from the effect of simply believing one has taken it. The Bonn team, led by Martin Coenen, Gunther Hartmann and Thomas Buttner, says this is the first time that method has been adapted into a repeatable clinical procedure and applied to help a specific patient understand their own symptoms, rather than to study nocebo responses in groups of research volunteers.
The same group had previously described a simpler version of this idea: a 2025 case report in Frontiers in Cardiovascular Medicine in which a woman with reported statin intolerance underwent a six week, single-blind crossover test with atorvastatin and a matched placebo in unmarked containers. Her pain ratings did not differ meaningfully between the placebo weeks and the drug weeks, and she was able to resume statin therapy. The new five-session protocol was built for situations where that simpler test would not be considered a firm enough answer, adding repeated exposures and deliberate mismatches between what patients were told and what they actually received.
The authors are explicit about the limits of what one patient can show. They write that a single case cannot establish the protocol's diagnostic utility or how well it would generalize to others, and they note the design's own demands, multiple visits, clinician supervision at each session, and the use of deception, may make it impractical for routine use. For many patients, they suggest, the simpler single-blind version already reported may be diagnostically sufficient; the more elaborate protocol is offered as a proof of concept that a more rigorous version exists, not as a tool ready for widespread clinical rollout.