LETTER

Misleading Decimal Places To the Editor: I suspect there may be errors in table 4 in the article by Pandya and Edmonds1 in December’s issue. These affect the figures and probabilities in the conclusions and the abstract. I prefer percentages of small numbers of patients to be expressed as integers, as decimal points give a spurious sense of precision when judging whether the results of a small sample from a retrospective case series can be generalized to a wider population.

TO THE

EDITOR

However, if one does use decimal places, they should be accurate. For one case of 14 open fractures, 7% or 7.1%, but not 7.0% and one of 12 closed fractures 8% or 8.3%. Assuming the bone-healing complications to represent the sum of the patients in the first 4 lines of table 4, I would expect between 5 patients (36%) and 5+3+3 = 11 patients (79%) with open fractures, not 21.0%. For bone-healing complications in closed fractures, I calculate either 1 or 2 patients (8% or 17%), not 4.0%, which represents just under half a patient. Infectious/wound-healing problems should be 7% or 14% of 14 open

fractures, with an overall complication rate of at least 36%. Infectious/ wound-healing problems should be 8% of 12 closed fractures, with an overall complication rate of at least 16.7 or 17%. Robert S. Jeffery, MB, BChir, FRCS(orth) Honorary University Fellow, Peninsula School of Medicine and Dentistry Plymouth University, Plymouth, UK

REFERENCE 1. Pandya NK, Edmonds EW. Immediate intramedullary flexible nailing of open pediatric tibial shaft fractures. J Pediatr Orthop. 2012;32:770–776.

The author declares no conflict of interest.

e18 | www.pedorthopaedics.com

J Pediatr Orthop



Volume 35, Number 2, March 2015

Misleading decimal places.

Misleading decimal places. - PDF Download Free
57KB Sizes 0 Downloads 12 Views