Up-down study--is the analysis adequate?
Although a formal dose range study was not available, a pilot study could have been conducted or a standard deviation and likely range of doses estimated from the data of their previous study (3).
The value of up-down study designs was challenged at the 2005 Obstetric Anesthesiology and Perinatology Conference4. As with any issue, there are differing opinions about the clinical value of up-down studies but we consider the authors should have subjected their data to probit and logit regression as a back up or sensitivity test. Given the limitations of the study, we wonder whether the results contained are robust enough to withstand review.
(1.) Columb MO, Lyons G. Determination of the minimum local analgesic concentrations of epidural bupivacaine and lidocaine in labor. Anesth Analg 1995; 81:833-837.
(2.) Mohta M, Aggarwal D, Gupta LK, Sethi AK, Tyagi A. Potency of mephentermine for prevention of post-spinal hypotension. Anaesth Intensive Care 2009; 37:568-570.
(3.0 Mohta M, Aggarwal D, Gupta LK, Tyagi A, Gupta A, Sethi AK. Comparison of potency of ephedrine and mephentermine for prevention of post spinal hypotension in caesearean section. Anaesth Intensive Care 2008; 36:360-364.
(4.) D'Angelo R, Shafer S, Clyne B, Eisenach JC. Efficiency and accuracy of up-down vs. random allocation to determine analgesic potency and dose response. Anesthesiology SOAP Abstract Supplement 2005; 102:A-24.
Up-down study--is the analysis adequate?--Reply
I thank Makkar and Wig for showing their interest in our study. However, I suggest that they review our first study1 which was an initial attempt to compare the [ED.sub.50] of ephedrine and mephentermine for prevention of hypotension in women undergoing caesarean section under spinal anaesthesia. The study to which they refer (2) was the follow-up study, using exactly the same methodology and therefore the detailed methodology was referenced but not repeated.
It was clearly mentioned in the initial study that the sample size was based on the up-and-down method of Dixon, according to which at least six independent pairs of patients with no hypotension/hypotension (response/no response pairs) should provide reliable estimates of the minimum effective dose of vasopressor [([ED.sub.50]).sup.3]. Makkar and Wig suggested using the likely range of doses from the data of that initial study--indeed, the initial dose and the dose interval of mephentermine used in the most recent study were based on those results.
Makkar and Wig also raised an interesting point regarding use of probit and logit regression. These regression analyses have been used as a back-up test by many workers but their value is controversial. In 1999, Vagero and Sundberg (4) warned against conventional logit or probit regression used in combination with up-and-down designs. This issue was again raised by Pace and Stylianou (5), who consider that standard probit or logit regression is likely to produce biased estimators. Considering the lack of familiarity of many anaesthetists with probit or logit regression, in their opinion up-and-down studies should no longer include logit or probit regression analysis. Thus, we believe that the results of both the studies we conducted are robust enough to withstand review.
(1.) Mohta M, Agarwal D, Gupta LK, Tyagi A, Gupta A, Sethi AK. Comparison of potency of ephedrine and mephentermine for prevention of post-spinal hypotension in caesarean section. Anaesth Intensive Care 2008; 36:360-364.
(2.) Mohta M, Agarwal D, Gupta LK, Sethi AK, Tyagi A. Potency of mephentermine for prevention of post-spinal hypotension. Anaesth Intensive Care 2009; 37:568-570.
(3.) Dixon WJ. The up-and-down method for small samples. Journal of American Statistical Association 1965; 60:967-978.
(4.) Vagero M, Sundberg R: The distribution of the maximum likelihood estimator in up-and-down experiments for quantal dose-response data. J Biopharm Stat 1999; 9: 499-519.
(5.) Pace NL, Stylianou MP. Advances in and limitations of up-and-down methodology: a precis of clinical use, study design, and dose estimation in anesthesia research. Anesthesiology 2007; 107:144-152.
|Printer friendly Cite/link Email Feedback|
|Author:||Makkar, J.K.; Wig, J.|
|Publication:||Anaesthesia and Intensive Care|
|Article Type:||Letter to the editor|
|Date:||Jan 1, 2010|
|Previous Article:||The ethics of anaesthesia learning curves (comment).|
|Next Article:||The Blease Datum anaesthesia vaporiser and the backbar.|