You are on page 1of 3

Nagele

lidocaine 2% for peribulbar anaesthesia in terms of speed of onset of anaesthesia. This surprising nding was in contrast to other studies, which found no difference between the two compounds in terms of time to satisfactory extradural block5 or peribulbar block.6 In the latter study, however, large volumes of L-bupivacaine 0.75% were used as a single agent and patients received up to three injections of local anaesthetic. In this study, we used a single injection technique and equal volumes of lidocaine 2% and either bupivacaine or L-bupivacaine, thereby reducing its concentration to 0.375%. We were unable to assess the duration of motor block as patients' eyes were bandaged and covered postoperatively and they were discharged home 12 h after surgery. If a shorter duration of motor block occurs with L-bupivacaine, this could be advantageous as prolonged paralysis from local anaesthesia leaves the eye vulnerable to drying and trauma. Peribulbar anaesthesia requires relatively large volumes of local anaesthetic and concerns have been expressed about the potential for systemic toxicity. The incidence of peribulbar blocks requiring supplementary anaesthesia has been reported to be as high as 54%.8 However, Lbupivacaine is less toxic to the myocardium and central nervous system.9 10 While L-bupivacaine may have theoretical advantages in elderly patients with coexisting cardiac disease, the present study did not show any untoward effects with either drug. L-Bupivacaine did not demonstrate any advantages over racemic bupivacaine when used for peribulbar anaesthesia.

Acknowledgement
We gratefully acknowledge the support of Abbott Pharmaceuticals, who provided supplies of L-bupivacaine and nancial assistance.

References
1 Wong DHW. Regional anaesthesia for intraocular surgery. Can J Anaesth 1993; 40: 63557 2 Albright GA. Cardiac arrest following regional anaesthesia with etidocaine or bupivacaine. Anesthesiology 1979; 52: 2857 3 Heath M. Deaths after intravenous regional anaesthesia. Br Med J 1983; 285: 91314 4 McLeod GA, Burke D. Levobupivacaine. Anaesthesia 2001; 56: 33141 5 Cox CR, Faccenda KA, Gilhooly C, Bannister J, Scott NB. Extradural S()-bupivacaine: comparison with racemic RSbupivacaine. Br J Anaesth 1998; 80: 28993 6 McLure HA, Rubin AP. Comparison of 0.75% levobupivacaine with 0.75% racemic bupivacaine for peribulbar anaesthesia. Anaesthesia 1998; 53: 11604 7 Brahma AK, Pemberton CJ, Ayeko M, Morgan LH. Single medial injection peribulbar anaesthesia using prilocaine. Anaesthesia 1994; 49: 10035 8 Loots JH, Koorts AS, Venter JA. Peribulbar anesthesia. A prospective statistical analysis of the efcacy and predictability of bupivacaine and a lidocaine/bupivacaine mixture. J Cataract Refract Surg 1993; 19: 726 9 Kopacz DJ, Allen HW. Accidental intravenous levobupivacaine. Anesth Analg 1999; 89: 10279 10 Bardsley H, Gristwood R, Baker H, Watson N, Nimmo W. Comparison of the cardiovascular effects of levobupivacaine and rac-bupivacaine following intravenous administration to healthy volunteers. Br J Clin Pharmacol 1998; 46: 2459

Downloaded from http://bja.oxfordjournals.org/ by guest on March 28, 2012

British Journal of Anaesthesia 90 (4): 51416 (2001)

DOI: 10.1093/bja/aeg087

Misuse of standard error of the mean (SEM) when reporting variability of a sample. A critical evaluation of four anaesthesia journals
P. Nagele*
Department of Anesthesiology and General Intensive Care, University of Vienna, Austria and Department of Anesthesiology, Washington University School of Medicine, St Louis, MO, USA
*Address for correspondence: Department of Anesthesiology and General Intensive Care, University of Vienna, Wahringer Gurtel 1820, A-1090 Vienna, Austria. E-mail: peter.nagele@univie.ac.at
Background. In biomedical research papers, authors often use descriptive statistics to describe the study sample. The standard deviation (SD) describes the variability between individuals in a sample; the standard error of the mean (SEM) describes the uncertainty of how the sample mean represents the population mean. Authors often, inappropriately, report the SEM when describing the sample. As the SEM is always less than the SD, it misleads the reader into underestimating the variability between individuals within the study sample.

The Board of Management and Trustees of the British Journal of Anaesthesia 2003

Misuse of

SEM

Methods. The aim of this study was to evaluate the frequency of inappropriate use of the SEM in four leading anaesthesia journals in 2001. The journals were searched manually for descriptive statistics reporting either the mean (SD) or the mean (SEM), and inappropriate use of the SEM was noted. Results. In 2001, all four anaesthesia journals published articles that used the SEM incorrectly: Anesthesia & Analgesia 27.7%, British Journal of Anaesthesia 22.6%, Anesthesiology 18.7% and European Journal of Anaesthesiology 11.5%. Laboratory reports and clinical studies were equally affected, except for Anesthesiology where 90% were basic science reports. Conclusions. One in four articles (n=198/860, 23%) published in four anaesthesia journals in 2001 inappropriately used the SEM in descriptive statistics to describe the variability of the study sample. Anaesthesia journals are encouraged to provide clearer statistical guidelines on how to report data variability in descriptive statistics. Br J Anaesth 2003; 90: 51416 Keywords: statistics Accepted for publication: December 3, 2002
Downloaded from http://bja.oxfordjournals.org/ by guest on March 28, 2012

When reporting data in biomedical research papers, authors often use descriptive statistical methods to describe their study sample. Descriptive statistics aim to describe a given study sample without regard to the entire population; inferential statistics generalize about a population on the basis of data from a sample of this population. If normally distributed, the study sample can be described entirely by two parameters: the mean and the standard deviation (SD). The SD represents the variability within the sample; the larger the SD, the higher the variability within the sample.1 Although it is clear that samples should always be summarized by the mean and SD,25 authors often use the standard error of the mean (SEM) to describe the variability of their sample. The SEM is used in inferential statistics to give an estimate of how the mean of the sample is related to the mean of the underlying population. As the SEM is always smaller than the SD, the unsuspecting reader may think that the variability within the sample is much smaller than it really is. Although the SD and the SEM are related (SEM=SD/ n), they give two very different types of information.6 Whereas the SD estimates the variability in the study sample, the SEM estimates the precision and uncertainty of how the study sample represents the underlying population.1 7 In other words, the SD tells us the distribution of individual data points around the mean, and the SEM informs us how

precise our estimate of the mean is.3 It is therefore inappropriate and incorrect to present data only as the mean (SEM). This evaluation was designed to identify the frequency of this statistical error in articles published in 2001 in four leading anaesthesia journals: two from the USA (Anesthesiology and Anesthesia & Analgesia), and two from Europe (British Journal of Anaesthesia and European Journal of Anaesthesiology).

Methods and results


All articles published in Anesthesiology, Anesthesia & Analgesia, British Journal of Anaesthesia or European Journal of Anaesthesiology in 2001 were searched manually for descriptive statistics reporting either mean (SD) or mean (SEM). Inappropriate use of the SEM in text, gures and tables was noted when the SEM was used to describe the variability of the study sample (instead of SD). Excluded from this analysis were articles using median and range, and articles that solely used inferential statistics such as condence intervals (CI). Case reports and review articles were not considered, except for case series comprising several cases and using descriptive statistics to describe the study sample.

Table 1 Frequency of use of standard error of the mean (SEM) and standard deviation (SD) in four anaesthesia journals, listed in order of decreasing percentage misuse. Data are numbers of articles (%). *Some of these articles used both the SD and the SEM to describe the study sample Incorrect use of SEM;total* Anesthesia & Analgesia British Journal of Anaesthesia Anesthesiology European Journal of Anaesthesiology 112 31 48 7 (27.7) (22.6) (18.7) (11.5) Laboratory studies using SEM incorrectly 66 15 43 3 (59) (47) (90) (43) Correct use of
SD

Total

293 106 209 54

(72.3) (77.4) (81.3) (88.5)

405 137 257 61

515

Nagele

A total of 257 articles fullled the search criteria in Anesthesiology, 405 articles in Anesthesia & Analgesia, 137 in the British Journal of Anaesthesia, and 61 in the European Journal of Anaesthesiology. Detailed results are given in Table 1, where the four journals are listed in order of decreasing percentage misuse of SEM. Eight articles each in Anesthesiology and Anesthesia & Analgesia even failed to state which parameter was used. It must be noted that in some of the articles that incorrectly used the SEM, both parameters, SEM and SD, were used. In these articles, the SD was mostly found in the text and the SEM in the gures.

statistical error by requiring authors to adhere to statistical recommendations, for instance through a more stringent statistical review process. The goal should be to have one standard method to describe the distribution of a study sample, thereby reducing confusion among the readers of biomedical research papers.

Acknowledgements
The author wishes to thank C. Michael Crowder MD PhD for his helpful comments on the manuscript, and Dennis M. Fisher MD and Doug Altman DSc for valuable comments on biomedical statistics. The author was supported by the Fonds zur Forderung der Wissenschaftlichen Forschung (FWF)Austrian Science Fundas the recipient of an Erwin-Schrodinger research fellowship.

Discussion
This evaluation of four leading anaesthesia journals shows clearly that a signicant number of published articles (mis-)use the SEM in descriptive statistics, which may be misinterpreted as showing the variability within the study sample. This use is not only statistically inappropriate, it also makes the reader assume a much smaller variability of the sample. In general, the use of the SEM should be limited to inferential statistics where the author explicitly wants to inform the reader about the precision of the study, and how well the sample truly represents the entire population. Thus, in inferential statistics, the use of SEM is valid but the CI is more valuable. In graphs and gures, use of SD is preferable to the SEM but the SEM can be used to improve the interpretation of the gure if the number of individuals/ experiments and the CI are clearly stated. In conclusion, in spite of clear recommendations, the SEM is still widely and inappropriately used in the anaesthesia literature. Anaesthesia journals could easily avoid this

References
1 Glantz S. Primer of Biostatistics, 4th Edn. New York: McGraw-Hill, 1997 2 Fisher DM. Research design and statistics in anesthesia. In: Miller RD, ed. Anesthesia, 5th Edn, Vol. 1. Philadelphia: Churchill Livingston, 2000; 75392 3 Streiner DL. Maintaining standards: differences between the standard deviation and standard error, and when to use each. Can J Psychiatry 1996; 41: 498502 4 Altman DG, Gore SM, Gardner MJ, et al. Statistical guidelines for contributors to medical journals. Br Med J 1983; 286: 148993 5 Lang TASM. How to report statistics in medicine: annotated guidelines for authors, editors, and reviewers. Philadelphia: American College of Physicians, 1997 6 Carlin JB, Doyle LW. Basic concepts of statistical reasoning: standard errors and condence intervals. J Paediatr Child Health 2000; 36: 5025 7 Webster CS, Merry AF. The standard deviation and the standard error of the mean. Anaesthesia 1997; 52: 183

Downloaded from http://bja.oxfordjournals.org/ by guest on March 28, 2012

516