John A. Loadsman

The role of transversus abdominis plane blocks in women undergoing total laparoscopic hysterectomy: A retrospective review

Introduction:  The transversus abdominis plane (TAP) local anaesthetic block is beneficial in patients undergoing open pelvic surgery; however, there are no data on its use in women undergoing laparoscopic gynaecologic surgery.

Evidence for non‐random sampling in randomised, controlled trials by Yuhji Saitoh

A large number of randomised trials authored by Yoshitaka Fujii have been retracted, in part as a consequence of a previous analysis finding a very low probability of random sampling. Dr Yuhji Saitoh co‐authored 34 of those trials and he was corresponding author for eight of them. We found a number of additional randomised, controlled trials that included baseline data, with Saitoh as corresponding author, that Fujii did not co‐author. We used Monte Carlo simulations to analyse the baseline data from 32 relevant trials in total as well as an outcome (muscle twitch recovery ratios) reported in several. We also compared a series of muscle twitch recovery graphs appearing in a number of Saitohs publications. The baseline data in 14/32 randomised, controlled trials had p < 0.01, of which seven p values were < 0.001. Eight trials reported four ratios of the time for the return of muscle activity after neuromuscular blockade, the distributions of which were homogeneous: the p values for the observed Q statistics were 0.0055, 0.031, 0.016 and 0.0071. Comparison of graphs revealed multiple coincident or near‐coincident curves across a large number of publications, a finding also inconsistent with random sampling. Combining the continuous and categorical probabilities of the 32 included trials, we found a very low likelihood of random sampling: p = 1.27 × 10−8 (1 in 100,000,000). The high probability of non‐random sampling and the repetition of lines in multiple graphs suggest that further scrutiny of Saitohs work is warranted.

Widening the search for suspect data – is the flood of retractions about to become a tsunami?

This pessimistic assertion is from a 1988 textbook by T. W. K€orner in which he discussed the statistical methods suggested by J. B. S. Haldane [2] for recognising possibly fraudulent data. He continued his remarks thus: “The kind of tests proposed by Haldane depended on the fact that “higher order faking” required a great deal of computational work. The invention and accessibility of the computer means that the computational work involved has ceased to be a problem for the dishonest scientist” [1]. However, three decades on, we are seeing a number of high-profile cases in which dishonest scientists, apparently unaware of K€orner’s ‘advice’, have been caught out faking results, exposed by the aberrant statistical distributions of their fraudulent data. As previously noted in an editorial [3] accompanying the exposure of one author’s prolific body of fraudulent papers [4], methods being used now to detect fraud are similar to Philip and Haldane’s 1939 analysis of the subsequently discredited genetic experiments of Franz Moewus [5]. As a result of the recent cases, editors and other interested parties are now becoming far more aware of the potential for dishonest authors to submit fraudulent data. This follows a similar increase in awareness of the problem of plagiarism, and many editors are now taking a closer look at aspects like data distributions as well as textual similarity. It could be argued that journals, editors and other bodies charged with the oversight of research have been slow to learn the lessons of history, and to apply newer statistical methods to detect and analyse spurious or suspicious data, but this deficiency is now being addressed. Anaesthesia as a specialty, and particularly the journal Anaesthesia, can rightly claim with vicarious pride that one of its own, John Carlisle, is at the forefront of this effort. Carlisle’s first statistical expos e, involving data from the randomised, controlled trials (RCTs) of Yoshitaka Fujii, made the research world stand up and take notice [4]. After further refinement of the method [6], it was similarly applied to the RCTs of one of Fujii’s regular collaborators, Yuhji Saitoh [7]. Carlisle has now completed a further project of remarkable scale with arguably even more important implications – an analysis of 5087 RCTs, spanning eight journals and 16 years – published in this issue of Anaesthesia [8]. The method of Carlisle’s analysis has been published [6] and explained in detail elsewhere [3]. Briefly, in a properly conducted and accurately reported RCT, differences in baseline characteristics between groups are, by definition, due to chance. For this reason, reporting p values for demographic and other baseline data is usually discouraged. The p value is the probability of random sampling resulting in a difference as large or larger than the observed difference so, because we already know that differences in baseline characteristics occurred by chance, it is uninformative to calculate a p value. Carlisle, however, has developed and refined a novel use for the statistical analysis of baseline data to identify instances where sampling in clinical trials may not have been random, suggesting the trial was either not properly conducted or was inaccurately reported. Essentially, Carlisle’s method identifies papers in which the baseline characteristics (e.g. age, weight) exhibit either too narrow or too wide a distribution than expected by chance, resulting in an excess of p values close to either one or zero. This editorial accompanies an article by Carlisle, Anaesthesia 2017; 72: 944–52.

Perioperative outcomes after total laparoscopic hysterectomy compared with fast‐track open hysterectomy – A retrospective case–control study

Aims:  To examine perioperative outcomes after total laparoscopic hysterectomy compared with fast‐track open hysterectomy using a retrospective case–control study.

Direct hospital costs of total laparoscopic hysterectomy compared with fast‐track open hysterectomy at a tertiary hospital: a retrospective case‐controlled study

To assess the direct intraoperative and postoperative costs in women undergoing total laparoscopic hysterectomy and fast‐track open hysterectomy.

Seeking and reporting apparent research misconduct: errors and integrity - a reply

References 1. Carlisle JB. Data fabrication and other reasons for non-random sampling in 5087 randomised, controlled trials in anaesthetic and general medical journals. Anaesthesia 2017; 72: 944–52. 2. Carlisle JB, Dexter F, Pandit JJ, Shafer SL, Yentis SM. Calculating the probability of random sampling for continuous variables in submitted or published randomised controlled trials. Anaesthesia 2015; 70: 848–58. 3. Devlin H. Statistical vigilantes: the war on scientific fraud – Science Weekly podcast. https://www.theguardian.com/ science/audio/2017/sep/14/statisticalvigilantes-the-war-on-scientific-scienceweekly-podcast (accessed 04/10/2017). 4. Senn SJ. Covariate imbalance and random allocation in clinical trials. Statistics in Medicine 1989; 8: 467–75.

Peri-operative lidocaine infusion for open radical prostatectomy.

We would like to invite Weinberg et al. to answer some questions that we have about their paper [1], which reports an improvement in average length of stay, as well as a number of secondary outcomes, when lidocaine was infused for 24 h peri-operatively for open radical prostatectomy. We are concerned that the difference in primary outcome, 1.3 days, appears to be almost entirely due to three outliers in the control group (lengths of stay: 9 days, 14 days and 19 days), and no details are provided to explain why these patients had a much longer hospital stay. Figure 2 in the paper shows a minimal difference in median length of stay and the rates of discharge were similar out to 5 days. In small studies such as this, it cannot be assumed the two groups were matched for confounders. The paper does not report important potential confounders such as duration of surgery and transfusion rates. We are left wondering if the prolonged stay in three of the control group patients may have been due to significant intra-operative or postoperative complications unrelated to the lack of lidocaine. Similarly, the study protocol allowed inclusion of patients taking up to 72 mg of oral morphine per day for up to a month pre-operatively. That amount of opioid could have a very important influence on a study such as this, but no information is provided regarding pre-operative use of opioids or other analgesics. It appears to us that the mean length of stay was compared with a t-test. If so, is that appropriate given the highly unequal variance between groups and the clearly non-normal distribution in the control group? Also, the standard deviation in the control group is several times larger than in the 2009 sample upon which the power calculation was based, suggesting the control group was, for some reason, not representative of the institutions’ usual patients. Furthermore, regarding the pain score comparison, was this analysis performed with a parametric test? The Author Guidelines for Anaesthesia indicate parametric testing of visual analog scale (VAS) data is likely to be inappropriate for samples less than 50. The pain assessment tool is described as a ten millimetre VAS. We suggest it would not be possible for a patient to mark such a tiny scale, and therefore we have concerns about the accuracy of these data. It is stated in the Methods that pain was assessed hourly for the first 4 h then 4-hourly for the next 20 h. It is therefore perplexing that Fig. 3 includes pain scores every hour for the first 24 h. The paper does not indicate who administered a VAS hourly throughout the night. Were those additional data actually numerical pain scores, rather than VAS, collected by ward nurses as part of routine PCA observations and subsequently recovered for this study from the nursing observation charts? If so, that could seriously compromise the reliability of the pain score data. The authors’ statement that lidocaine reduced pain at rest over the first 24 h does not appear to be supported by their data. They claim a difference of 1.8 mm.h . We do not understand how pain can be quantified in units of velocity. If the statistic was actually 1.8 units on the vertical scale of Figure 3 (whatever those units

Dilemmas in biomedical research publication: are we losing the plot?

Purpose of review Some recent and very controversial developments in the sphere of scientific publication, with significant implications for biomedical research, are posing a significant threat to traditional publication models. Many of these developments appear to be the result of a vicious circle that has developed from publication pressure on researchers, corporate financial exploitation of those pressures, and an apparent perception that individual and institutional reputations are to be promoted at all cost. Recent findings The detrimental effects of these developments have led to a groundswell of academic discontent and a ‘researcher uprising’ that may result in precipitous changes. Summary In many different respects, biomedical research publication is now in an unprecedented state of flux. Traditional models are being strongly challenged, probably with good reason, and alternative models of both funding and delivery need to be settled upon with some urgency. While individuals, institutions and corporate bodies who resist the current evolution may find themselves in line for extinction, at the same time it is important for the entire ‘industry’ to revert to some more traditional values and not allow self-interest to prevail.