Logo image
Pollyanna Phenomenon
Letter/Communication   Peer reviewed

Pollyanna Phenomenon

Richard J. Hamilton and Lewis R. Goldfrank
Clinical toxicology (Philadelphia, Pa.), v 36(1-2), pp 141-141
1998
pdf
Hamilton-PollyannaPhenomenon-ClinTox-199855.70 kB
Published, Version of Record (VoR) Access upon request

Abstract

While it certainly is onerous to have to play “professional worriers” to Dr. Robertson’s . Pollyanna, his comments warrant clarification of our point. The rule of three (an established statistical tool) states that 3/n is an upper 95% confidence bound for binomial probability p when in n independent trials no events occur! Indeed, for the set of data in question, the rule of three would set the upper limit of life-threatening complications at 1.5% (if indeed everyone were exposed to the loperamide). However, clinician/scientists use this data to report with 95% confidence that the life-threatening complications could be as little as zero or as much as 1.5%. Cautious clinicians use the upper 95% confidence probability to define the worst possible outcome for their patient. A report of this data with the upper 99% confidence probability would drive that number even higher. As toxicologists, we must not report data as “soft” or “semisoft,” we cannot measure “worry titers” and state that they “plummet to infinitesimal,” and we should not have “increasing confidence that the sky will not fall” based on the statistically biased, retrospective data collected from poison centers. We should strive toward precision. The 19th decimal place is unnecessary, but certainly one or two would be helpful. In conclusion, it was a pleasure to receive Dr. Robertson’s “couldn’t agree more with the authors” support as well as his reflections on Sir Bradford Hill’s concerns about our 1997 data. [1st paragraph]

Metrics

Details

UN Sustainable Development Goals (SDGs)

This publication has contributed to the advancement of the following goals:

#3 Good Health and Well-Being

InCites Highlights

Data related to this publication, from InCites Benchmarking & Analytics tool:

Web of Science research areas
Toxicology
Logo image