This Article Full Text (PDF) Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Goodman, N. W Articles by Mills, J. Search for Related Content PubMed PubMed Citation Articles by Goodman, N. W Articles by Mills, J.

Good advice on airways, but measurements are not parameters

Bristol, Uk

To the Editor:

Hung and Mills' editorial¹ about predicting difficult intubation was useful, but it is a shame they misused their terms. Karkouti et al.² correctly referred throughout to measures and variables; Hung and Mills wrote "no single airway parameter". Hung and Mills then wrote of the analysis identifying "three simple bedside tests... as [the] most useful airway parameters to predict ...". They should have written "three simple bedside tests... as the most useful to predict ...", and they confused meanings more by writing in the very next sentence that the investigators compared their predictions "in terms of three parameters". These were not, as one might suppose at first reading, the three tests, but instead the positive predictive value, sensitivity, and specificity, which are technically proportions, but more loosely statistics.

Hung and Mills are not alone in using parameter incorrectly.³ They will not be the last, but that does not stop the usage being unwise. According to dictionaries, variable is one of the meanings of parameter, but scientists need to be more precise. A word such as parameter, which can used for almost anything that can be measured or calculated even if only vaguely - and risks being a scientific thingummy (or, in French, le machin) - is best avoided except when it is unambiguously correct. Otherwise we will need to find a new word when we speak of the parameters of the normal distribution (the mean and standard deviation) and of clearance, half-life and volume of distribution as pharmacokinetic parameters.

References

1 Hung O, Mills J. Predictions and clinical decisions: a fine balance. Can J Anesth 2000; 47: 721–4.[Free Full Text]

2 Karkouti K, Rose K, Wigglesworth D, Cohen MM. Predicting difficult intubation: a multivariable analysis. Can J Anesth 2000; 47: 730–9.[Abstract/Free Full Text]

3 Goodman NW. Paradigm, parameter, paralysis of mind. Br Med J 1993; 307: 1627–9.

Halifax, Nova Scotia

We would like to thank Dr. Goodman for his comments. Generally, in the field of statistics, one usually thinks of a parameter as being linked to a particular model. In the third paragraph of our editorial which Dr. Goodman has critiqued, we are not referring to a particular statistical model and consequently (but perhaps ambiguously) using "parameter" synonymously with "measurement". For the sake of clarity we would therefore be content to replace "parameter" with "measurement" in the locations Dr. Goodman has highlighted.

This Article Full Text (PDF) Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Goodman, N. W Articles by Mills, J. Search for Related Content PubMed PubMed Citation Articles by Goodman, N. W Articles by Mills, J.