LAUSR

A recent high-profile paper in JAMA reported on a caseseries with neurological and cognitive symptoms due to an undetermined cause [1]. Under study were US government personnel who had been working in Havana, and had reported auditory and sensory phenomena of unknown origin. Six of the 21 people in the study were assessed with a neuropsychological battery of 37 tests divided into 10 cognitive domains. In each domain, the person was considered impaired if their score fell below the 40th percentile in at least one of the tests. The authors reported that all six people showed impairment in at least one domain, supporting the idea that they had sustained neurological injury in the absence of any head trauma. Concerns over possible ‘health attacks’ have led to the ongoing withdrawal of staff from the US embassy in Havana. By definition, 40% of people score at or below the 40th percentile on any task. Using this threshold for impairment, and without any correction for the number of test scores examined (37 tests*6 people = 222), we should expect around 89 false positives. We ran a simulation to explore the range of outcomes, using the authors’ system for counting impairments. Our method was to substitute their test scores with random numbers from a uniform distribution between 1 and 100, and to count how many people showed at least one impairment per domain (< 40th percentile). We did this 1000 times, and compared the distributions of our outcomes with their findings (see Fig. 1). The only deviations were where the number of impairments reported by Swanson and colleagues was fewer than expected by chance (possibly because the people tested were smarter than average).