If unrecognised, unintended pelvic roll may lead to clinically-meaningful changes in final acetabular component version through perceptual distortion during terminal insertion. Such deviation from target position may compromise the mechanical characteristics of the final construct and has previously been associated with instability/dislocation, edge loading/bearing wear, squeaking, range-of-movement limitations and increased revision rate [7, 9, 15]. Especially while under sterile exclusion draping, the ability of surgeons or surgical teams to recognise such movement can be understandably challenging [7, 8].

For this hypothesis-generating investigation, we made a prospective decision to capture ‘final’ pelvic position data at the end of the surgical procedure—immediately prior to the commencement of soft tissue closure—such that the overall magnitude of patient roll could be determined. It was our two-fold intent to be able to demonstrate that clinically-meaningful unintended pelvic movement was indeed occurring during routine decubitus surgery and also to then quantify the overall magnitude of such movement in a uniaxial plane. We believe that we have been able to achieve both of these considerations in the current work with respect to gender differences (or rather lack thereof). We accept that pelvic roll throughout the case is unlikely to follow a linear progression, and that such movement is likely to be influenced substantially by force-imparting procedural steps such as primary retractor placement/leverage and acetabular reaming/cup impaction. We acknowledge that the amount of positional change likely has the greatest clinical relevance immediately prior to definitive acetabular component insertion—whereby such understanding might permit the surgeon to make corrective decisions regarding cup placement such that the intended final orientation appropriately considers/factors in pelvic roll from the starting position. Having demonstrated in our previous published work the common occurrence of anteriorly-biased pelvic roll during primary THA surgery in the lateral position [8], and having shown herein that patient gender does not appear to be a statistically-relevant confounder to the magnitude of such roll, we believe our presented evidence provides a scientifically-robust impetus to support future investigations aimed at more precisely describing the roll increments at sequential steps in the performance of a routine lateral THA. We hope to explore this element with targeted future studies.

Despite recognised differences in pelvic shape and proportions between genders [1], surprising little published work relates to such considerations being factored into pelvic position support design and/or application. There are four recognised pelvic ‘morpho-types’: gynecoid, android, anthropoid and platypelloid (Fig. 2) [4, 16]. The changing bony structure associated with these variants suggests that—when it comes to pelvic positioning clamps—that ‘one size may not fit all’. Despite this, none of the mainstream pelvic positioning supports marketed for THA use are offered as ‘gender specific’ nor with clear application differences for male and female patients. Despite the underpinning logic, in the absence of evidence of previous published work on the topic, we aimed to provide some initial data to explore whether or not there was a performance difference using standard positioning clamps during THA, as determined by the loss of angular position across the course of the procedure. We have previously shown—as have others—that unintended pelvic movement around a central sagittal axis (i.e. pelvic roll) commonly occurs during THAs performed in the lateral position [8, 12] and that such movement is almost universally anterior in vector [8]. We have also previously published work suggesting that BMI alone was not an independent correlate for the anticipated magnitude of pelvic roll [8]. While the ad hoc analysis of the relationship between the magnitude of intra-operative pelvic roll and patient BMI in the current study did not achieve statistical significance, we acknowledge that this consideration was not a primary evaluation goal of this work and the data pool itself was substantially underpowered to confidently demonstrate a true difference, if present. Future targeted studies of appropriate a priori size are indicated to further elucidate if a true difference between genders exists with regard to patient BMI at the time of surgery and the subsequent vector and magnitude of intra-operative pelvic movement in the lateral decubitus position.

Recognised pelvic morpho-types, as described by Caldwell and Moloy [4]

This study has several potential limitations that require consideration. Firstly, despite the prospective collection of all of the data utilised in this study we cannot exclude a potentially biasing effect of the retrospective nature of the analyses we have performed. Secondly, while recognising that this investigation likely represents the first such reported initiative in the field (and hence cohort data do not exist to define ‘clinically meaningful difference’ ranges), it may well prove in time that the endeavour was under-powdered to unmask a true gender difference in pelvic roll that may otherwise reasonably exist—future work to validate our findings will add merit to this in due time. Thirdly, we have not considered potential ethnic nor racial variations in pelvic morphology and any possible influence this may have had on our results. While our recruitment was from a consecutive (i.e. non-selected) cohort of patients presenting to a high-volume arthroplasty centre, the potential for under-representation of particular racially-diverse groups cannot be excluded. Fourthly, we have deliberately limited our study to primary THAs performed by an experienced senior arthroplasty surgeon within a high-volume arthroplasty health network. The results we report here may therefore not be representative of those seen with more complex (and/or longer) procedures such as revision operations, when dealing with more challenging anatomy such the setting of dysplasia, or when performed by lower volume surgeons. Finally, while we elected to employ the standardised use of a commercially-available, off-the-shelf, set of paired anterior and posterior pelvic positional clamps for our study to maintain high internal validity we cannot claim generalisability necessarily to all other available brands/types used for the same purpose. A follow up investigation, using different supports may therefore yield different final results. We are happy with this reality for the current study as we are confident that any differences between genders (or lack thereof) represent patient variation, not confounded by support device differences. Based on the above, an opportunity does exist to externally validate our work in diverse patient populations and using different support devices.