Variation in Trauma Admission During the COVID-19 Pandemic in a Major Trauma Unit - An Observational Study

Purpose: Following the World Health Organization declaration of a COVID-19 pandemic, there was a severe impact on UK trauma provision. Our aim in this paper was to examine the effect of alteration in working patterns and ‘lockdown’ on trauma and orthopaedic referral and management in a major trauma centre. Methods: All trauma referrals to the Liverpool Orthopaedic and Trauma Service were analysed over a 33-week period, divided into ‘Pre-lockdown’, ‘Lockdown’, and ‘Post-lockdown’. Results: 1661 orthopaedic trauma operations were undertaken - 967 in ‘pre-lockdown’, 193 in ‘lockdown’ and 501 in ‘post-lockdown’. An average weekly reduction in numbers of referrals to virtual fracture clinic and trauma cases undergoing surgery was observed, with numbers in virtual fracture clinic and in trauma cases undergoing surgery falling by similar amounts from prelockdown to lockdown (51.71% drop in virtual fracture clinic referrals and 54.38% reduction in trauma cases undergoing surgery). Hip fractures occupied significantly more theatre time compared to other trauma in the lockdown period. Post-lockdown, there was a rebound effect in the trauma numbers of lower limb surgery (excluding hip) and central surgery. Conclusion: The COVID-19 pandemic lockdown resulted in a significant change in practice and a reduction in both the presenting trauma numbers and those undergoing surgery in our trauma and orthopaedic department. The effect of the lockdown on trauma numbers has had a major mitigating effect on the consequences of rationing of resources. Variation in Admission During the COVID-19 Pandemic in a Major Trauma Unit - An Observational Study. Biomed J Sci & Tech Res 34(5)-2021.

patients and medical personnel being largely unknown, nonemergency services were reduced [4]. Guidelines were issued by groups including The British Orthopaedic Association (BOA) and the Federation of Surgical Specialty Associations advising how to apportion services with priorities set to maintain emergency surgery provision; protecting the surgical workforce; fulfilling alternative surgical and non-surgical roles; and how to rank cases according to importance [5][6][7]. This created ethical dilemmas, that reasonable decisions were to be made based on best evidence available at the time and in a collaborative way as much possible [8]. The Liverpool University Hospitals NHS Foundation Trust contains the region's major trauma centre, serving a population of 2.5 million, where the Liverpool Orthopaedic and Trauma Service (LOTS) resides. Due to the COVID-19 pandemic, LOTS produced local policies that supported triaging of urgent procedures and the implementing of alternative management based on the BOAST guidelines [5,7]. This included the non-surgical treatment of cases where possible, and preservation of surgical resources where significant morbidity or mortality was predicted if surgery was not performed. Our aim in this paper was to examine the effect of alteration in working patterns in 'lockdown' on orthopaedic trauma referral and management in a major trauma centre.

Methods
This study was a retrospective, observational study of prospectively collected data by LOTS. The study was conducted under the auspices of a service evaluation, therefore no ethical approval was required. All trauma patients referred to LOTS have their data prospectively collected and added to our database (Bluespier, Droitwich, UK). This includes inpatient and outpatient trauma referrals, with all accident and emergency attendances either referred directly to the on-call team or via the Virtual Fracture Clinic (VFC), where all acute ambulatory trauma is triaged by a consultant trauma surgeon and specialist physiotherapist.
The data collected includes standard demographic data, trauma categorisation, surgical intervention, post-operative outcomes and complications. A 33-week period was retrospectively analysed with inclusion criteria being all trauma referred to LOTS between the 2 nd December 2019 and 16 th July 2020. Trauma cases were characterised by AO/OTA classification and anatomical subspecialties of upper limb, central surgery (pelvis and rib), hip fractures (including all proximal femur fractures), and lower extremity surgery excluding hip fractures [9]. Our unit does not perform spinal surgery with all relevant cases transferred to the regional spinal unit, and were therefore not included. Where more than one anatomical site required surgical intervention, this was included as more than one procedure. Major trauma related information was identified via the prospectively collected database submitted to the Trauma Audit and Research Network and the National Hip Fracture database was also analysed to ensure all data was complete [10,11]. Initially, SARS-CoV-2 testing was not undertaken unless patients were symptomatic with routine testing introduced on 1 st June. Three time intervals were defined, totalling 33-weeks; 'Pre-lockdown' covering 16 weeks before 23 rd March; 'Lockdown' covering nine weeks from 23 rd March to 24 th May; and 'Post-lockdown' covering eight weeks from 24 th May to 16 th July as in previous studies [12]. Time periods were dictated by the UK Government announcements of national lockdown and easing of lockdown measurements [13]. The 16 week 'pre-lockdown' period was defined from when the merger of the hospitals occurred; prior to then, the hospitals acted separately but not equally and hence comparison could not be made. This was expressed as a percentage of the total number of cases performed from the date.

Statistical Analysis
The study was completed according to STROBE guidelines for observational studies [14]. Continuous variables were tested for normality distribution and presented as means with 95% confidence intervals. Categorical and qualitative variables are expressed as numbers and percentages. The Student t-test was used for continuous variables if the criteria for normality and equality of variances were fulfilled. Alternatively, the Mann-Whitney U test was performed. Categorical variables were analysed using the Chisquare test for sample sets greater than 5, otherwise the Fisher's exact test was used. The one-way analysis of variance (ANOVA) was used where comparison between the means of three or more independent groups was required. A p value less than 0.05 was considered significant. All data were assessed using SPSS Version 26.0 (SPSS Inc., IBM, Chicago, IL).

Overall Orthopaedic Trauma Surgical Procedures
In the 33-week analysis period, 1661 orthopaedic trauma operations were undertaken with 967 procedures in pre-lockdown, 193 procedures in lockdown and 501 cases in the post-lockdown periods. The number of cases performed per week showed a significant difference across the time periods (p<.001) with much higher volumes being performed pre-lockdown as compared to lockdown and post-lockdown periods (Table 1).

Surgical Procedures by Anatomical Location
In the entire cohort there were 337 upper extremity procedures, 54 central surgeries, 552 hip fracture procedures, 532 lower extremity procedures and 186 'otherwise' classified.
There were significant decreases in the average cases per week for all anatomical locations between pre-lockdown and lockdown (Table 1 & Figure 1). Comparing average percentage of procedures undertaken per week, hip fractures occupied significantly more theatre space during lockdown due to priority given to them with non-operative treatment being potentially more appropriate for other anatomical locations (Table 1

Virtual Fracture Clinic
The mean total number of VFC referrals per week reduced significantly once lockdown started (221.13 vs 109.00) and remained lower in the post-lockdown period (203.13 p/w) ( Table   2). Regardless of the overall reduction in numbers, the percentage VFC outcomes remained relatively unchanged with the percentage referred to fracture clinic, discharged, unable to contact and referred to physiotherapy being the same throughout the periods analysed (Table 2). However, there was a significant reductions in patients sent to elective or extended scope practitioner clinics, which is most likely due to the reduction in resources that were available due to re-allocation or shielding of staff (Table 3).   [5,15]. Studies in the UK and in Germany also found similar reduction in both case load and resources to our findings [16,17]. In our unit, the emphasis was on a more consultant-led, non-surgical treatment of cases where possible. We have not analysed as part of this study the effect on the change of management on the clinical outcome of the affected patients. However, in a recent paper from our unit on foot and ankle surgery during the COVID-19 pandemic, it was shown that in 12 cases management was changed due to the COVID-19 pandemic resulting in three undergoing surgery at a later date [12]. There has been concern regarding the long term effect such rationing of care will have on patients who have had trauma, however it is clear in our study that the numbers of trauma referrals reduced proportionally to the number of undertaken surgical cases.
Comparing the average weekly reduction in case numbers in VFC showed that the referral of cases in VFC fell by 51.71% from pre-lockdown to lockdown, compared to the 54.38% reduction in surgical cases from pre-lockdown to lockdown. This illustrates that the main factor for our reduction in surgical cases was not pandemic may illustrate these changes in practice further.
Our study has limitations. This was a retrospective study of observational data. This study included all patients undergoing orthopaedic trauma surgery in an operating theatre, however during the lockdown period a number of patients may have had interventions outside of an operating theatre. As we did not assess the management decisions in this paper, we may not have included patients who had sedation in the emergency department or fracture clinic. Similarly, we have not analysed the complexity of cases seen in our department, and all inferences have been made on total numbers of cases which can create bias. As a retrospective series, some datasets were incomplete and there is a higher chance of errors in dates recorded.

Conclusion
The COVID-19 pandemic lockdown resulted in a significant change in practice and a reduction in both the presenting trauma numbers and those undergoing surgery in our orthopaedic department. The effect of the lockdown on reducing trauma numbers has mitigated the consequences of rationing of resources.