Background Healthcare governance (HG) is a quality assurance processes that aims to maintain and improve clinical practice. Clinical decisions are routinely reviewed after the outcome is known to learn lessons for the future. When the outcome is positive, then practice is praised, but when practice is suboptimal, the area for improvement is highlighted. This process requires counterfactual reasoning, where we predict what would have happened given both what happened and the possible different decisions. Causal models that capture the mechanisms that generate events can support counterfactual reasoning. Objective This study is an initial attempt to show how counterfactual reasoning with causal Bayesian networks (CBNs) can be used as a HG tool to assess what would have happened if treatments other than those occurred had been selected. Methods Motivated by the Defence Medical Services (DMS) mortality and morbidity (M&M) review meeting, in this paper we (1) extended the use of counterfactual reasoning in CBNs to review decisions, where the alternative treatment strategies and its effect belong to different stages of care, (2) placed counterfactual reasoning in a specific clinical context to examine how it can be used as a HG tool. Results Using three realistic examples, we demonstrated how the proposed counterfactual reasoning can be used to assist the DMS M&M review meetings. Conclusions Useful lessons can be learned by assessing decisions after they are made. M&M review meetings are fruitful ground for counterfactual reasoning. The use of a clinical decision support tool that can assist clinicians in assessing counterfactual probabilities will be beneficial.

Developing clinical decision support systems (CDSS) that accurately capture the way clinicians gather information and make decisions during patient’s care is challenging. The modelling problem is especially difficult for acute conditions that evolve rapidly over a short timescale. Previous researchers have proposed various time-based Bayesian networks (BNs) applicable to a stationary process rather than the evolving situation encountered in acute medical conditions. In this paper, we propose a method for developing CDSS using progressive BNs that capture the evolving nature of acute conditions, as well as the dynamics of clinical decision making. To describe and validate our methodology we developed a BN for predicting the likelihood of survival of combat casualties in the prehospital setting: field and en route care. To overcome the typical lack of data for this type of problem, we use a combination of expert knowledge and available data to produce a causally coherent BN structure. The accuracy of the model is demonstrated using a 10-fold cross validation. The model had an excellent discrimination ability with an area under the ROC curve of 0.84 and 0.82 for the BN in field and en route care, respectively. The proposed model can be used to potentially support clinical decision making in several ways, such as: patient triage, or enabling precision medicine with treatments targeted to risk.

BACKGROUND The United Kingdom government introduced lockdown restrictions for the first time on 23 March 2020 due to coronavirus disease 2019 (COVID-19) pandemic. These were partially lifted on 15 June and further eased on 4 July. Changes in social behaviour, including increased alcohol consumption were described at the time. However, there were no data available to consider the impact of these changes on the number of alcohol-related disease admissions, specifically alcohol-related acute pancreatitis (AP). This study evaluated the trend of alcohol-related AP admissions at a single centre during the initial COVID-19 lockdown. AIM To evaluate the trend in alcohol-related AP admissions at a single centre during the initial COVID-19 lockdown in the United Kingdom. METHODS All patients admitted with alcohol-related AP from March to September 2016 to 2020 were considered in this study. Patient demographics, their initial presentation with AP, any recurrent admissions, disease severity and length of stay, were evaluated using ANOVA and χ2 and Kruskal–Wallis tests. RESULTS One hundred and thirty-six patients were included in the study. The highest total number of AP admissions was seen in March–September 2019 and the highest single-month period was in March–May 2020. Admissions for first-time presentations of AP were highest in 2020 compared to other year groups and were significantly higher compared to previous years, for example, 2016 (P < 0.05). Furthermore, the rate of admissions decreased by 38.89% between March–May 2020 and June–September 2020 (P < 0.05), coinciding with the easing of lockdown restrictions. This significant decrease was not observed in the previous year groups during those same time periods. Admissions for recurrent AP were highest in 2019. The median length of hospital stay did not differ between patients from each of the year groups. CONCLUSION An increased number of admissions for alcohol-related AP were observed during months when lockdown restrictions were enforced; a fall in figures was noted when restrictions were eased.

Abstract Background Shifts in social behaviours as a consequence of the COVID-19 pandemic may make people more prone to alcohol use as a coping mechanism. This study aims to investigate whether there was an impact on acute alcohol-related pancreatitis cases given potential changes in alcohol consumption during the first COVID-19 lockdown period in the United Kingdom. It considers the time period since social establishments re-opened in July 2020, to evaluate whether increase alcoholic pancreatitis cases, as a proxy for altered drinking habits, persisted. Methods All codes relating to pancreatitis were used to identify cases between March to September (2016 to 2020). Inclusion criteria: acute alcohol-related pancreatitis only. These cases were screened by a group of independent medical juniors and analysed using Excel. Results A total of 1905 patients with pancreatitis were identified in the initial search, of which 136 were admitted with acute alcoholic pancreatitis. There was an increase in absolute number of new cases in 2020 compared to previous years (27 compared to a mean of 16 in the previous 4 years). When shops and pubs reopened, cases reduced by 37.5%. Conclusion This study highlights the issue of increased admission with acute pancreatitis relating to alcohol in the COVID-19 lock down period. During which, this group of patients are likely socially isolated with a lack of support. This calls for an increase in community support for such patients with emphasis on admission avoidance. This is especially important in an era when there is increased hospital admission due to COVID-19.

Various AI models are increasingly being considered as part of clinical decision-support tools. However, the trustworthiness of such models is rarely considered. Clinicians are more likely to use a model if they can understand and trust its predictions. Key to this is if its underlying reasoning can be explained. A Bayesian network (BN) model has the advantage that it is not a black-box and its reasoning can be explained. In this paper, we propose an incremental explanation of inference that can be applied to 'hybrid' BNs, i.e. those that contain both discrete and continuous nodes. The key questions that we answer are: (1) which important evidence supports or contradicts the prediction, and (2) through which intermediate variables does the information flow. The explanation is illustrated using a real clinical case study. A small evaluation study is also conducted.

We read with interest Vassallo et al ’s article[1][1] in which the authors describe the development of the MPTT-24, a major incident triage tool. We fully support the authors’ aims in developing a triage tool with such potential for improving the care of injured service personnel, but suggest

BACKGROUNDVarious injury severity scores exist for trauma; it is known that they do not correlate accurately to military injuries. A promising anatomical scoring system for blast pelvic and perineal injury led to the development of an improved scoring system using machine-learning techniques.METHODSAn unbiased genetic algorithm selected optimal anatomical and physiological parameters from 118 military cases. A Naïve Bayesian model was built using the proposed parameters to predict the probability of survival. Ten-fold cross validation was employed to evaluate its performance.RESULTSOur model significantly out-performed Injury Severity Score (ISS), Trauma ISS, New ISS, and the Revised Trauma Score in virtually all areas; positive predictive value 0.8941, specificity 0.9027, accuracy 0.9056, and area under curve 0.9059. A two-sample t test showed that the predictive performance of the proposed scoring system was significantly better than the other systems (p < 0.001).CONCLUSIONWith limited resources and the simplest of Bayesian methodologies, we have demonstrated that the Naïve Bayesian model performed significantly better in virtually all areas assessed by current scoring systems used for trauma. This is encouraging and highlights that more can be done to improve trauma systems not only for our military injured, but also for civilian trauma victims.

Significant lessons to inform best practice in trauma care should be learned from the last decade of conflict in Afghanistan and Iraq. This study used radiological data collated in the UK Military Hospital in Camp Bastion, Afghanistan, to investigate the most appropriate device length for needle chest decompression of tension pneumothorax (TP). We reviewed the optimal length of device and site needed for needle decompression of a tension pneumothorax in a UK military population and found no significant difference between sites for needle chest decompression (NCD). As a result, we do not recommend use of devices longer than 60mm for UK service personnel.

Aims: A Trauma damage control laparotomy (DCL) entails immediate control of haemorrhage and contamination, laparostomy and physiological stabilisation, then completion of surgery and early primary closure (EPC).Failing early primary closure, temporary abdominal closure (TAC) techniques maintain abdominal integrity until early definitive closure (EDC).The aims were to identify and compare outcomes of early definitive closure methods in these patients.Methods: NICE, Cochrane, OVID (Medline, AMED, Embase, HMIC) and PubMed databases were accessed using (traum*, damage control, abbreviated laparotomy, component separation, fascial traction, mesh closure, planned ventral hernia (PVH), and topical negative pressure (TNP)).Randomised Controlled Trials, Case Series and Cohort Studies reporting TAC and early definitive closure methods in DCL trauma patients were included.Outcomes were mortality, days to fascial closure, hospital length of stay, abdominal complications and delayed ventral herniation.Results: 26 studies identified early primary closure (DPC, acute component separation (ACS) and acute mesh repair (AMR)) and TAC methods (Whitman patch (WP), topical negative pressure (TNP), temporary mesh (TM), fascial tension, Bogota bag and skin tension).Estimates for mortality and abdominal complications in AMR and DPC groups were 0.45% and 40.85%, and 6.07%, and 16.74% respectively; AMR ventral hernia / laxity was 51.1% at one year.Days to closure were 6.30, 21.10 and 15.90 in DPC, ACS and AMR groups, whilst hospital LOS in ACS and DPC groups was 17.5 versus 23.3 days.Conclusions: Acute component separation or mesh repair are alternative early definitive closure techniques to DPC following trauma DCL.Comparing outcomes is hampered by poverty of uniform reporting and bias.Recommendations for standardised reporting nomenclature and methodology are made.

Sir, we would like to draw attention to an observed differential rate of eye injury between International Security Assistance Force (ISAF) and Afghan National Security Force (ANSF) troops in Afghanistan. Until recently, most conflicts had seen an increased rate of fragment-related penetrating eye

Recent military operations have resulted in a small but significant number of military personnel suffering severe perineal injuries. In association with lower limb amputation and pelvic fracture, this complex is described as the 'signature injury' of the current conflict in Afghanistan. There are significant consequences of surviving severe perineal injury but the experience of managing these casualties is limited. This article gives an overview of the processes developed to meet these challenges and introduces a series of articles which examine the subject in finer detail.

OBJECTIVES:Management of blunt splenic injury (BSI) in battlefield casualties is controversial. Splenectomy is the traditional treatment, as setting the conditions for selective non-operative management (SNOM) is difficult in the operational environment. On mature operations, it may be feasible to adopt a more conservative approach and manage the patient according to civilian protocols. The aim of this study was to document the contemporary practice of deployed military surgeons when dealing with BSI and to compare this against a matched cohort of civilian BSI patients.METHOD:The Joint Theatre Trauma Registry held at the Royal Centre for Defence Medicine, Birmingham, was thoroughly examined to yield patients with BSI. The study encompassed a 55-month period ending September 2009. Data abstracted included patient demographics, injury epidemiology, grade of splenic injury, treatment and outcome. These data were compared with a registry database from a UK civilian major trauma centre.RESULT:Of 1516 military trauma patients, 16 (1%) had a splenic injury, of which five were excluded either because of fatalities due to overwhelming injury or penetrating trauma. The remaining 11 had a blunt component. Median (IQR) injury severity score (ISS) was 17 (15-21). Nine underwent a splenectomy with median (IQR) ISS of 17 (12-18). Of this group, organ injury grades were documented in 10 patients (four Grade V injuries, three Grade IV and three Grade II). All patients survived surgery. There were no complications in survivors as a result of splenic conservation in the military group. Data from the civilian major trauma centre database showed 160 (2%) patients sustained a splenic injury, of which 131 (82%) had a blunt mechanism, 43/160 (27%) and 9/160 (6%) patients underwent splenectomy and angio-embolisation, respectively.CONCLUSIONS:Patients with BSI, an uncommon finding in combat casualties, are occasionally selected for conservative management, contrary to previous military surgical paradigms but in keeping with the civilian shift to SNOM. Guidelines to clarify the place of SNOM are required to assist surgical decision making on deployed operations.

Injury patterns in Afghanistan have altered from ballistic trauma in 2006 to blast trauma in 2010–2011. Surgeons have had to alter their surgical resuscitation strategies. Improvised explosive device (IED) yields have increased, typically causing bilateral high transfemoral amputations and increasing the likelihood of pelvic and perineal injury.1 Forty per cent of bilateral transfemoral amputations in 2009 had an associated pelvic fracture. This led to a UK military policy of applying a pelvic binder to all IED victims in the pre-hospital environment.

Background Improvised explosive device (IED) yields in Afghanistan have increased resulting in more proximal injuries. The injury severity score (ISS) is an anatomic aggregate score of the three most severely injured anatomical areas but does not accurately predict severity in IED related pelvi-perineal trauma patients. A scoring system based on abbreviated injury score (AIS) was developed to reflect the severity of these injuries in order to better understand risk factors, develop a tool for future audit and improve performance.Method Using standard AIS descriptors, injury scales were constructed for the pelvis (1, minor to 6, maximal). The perineum was divided into anterior and posterior zones as relevant to injury patterns and blast direction with each soft tissue structure being allocated a score from its own severity scale. A cumulative score, from 1 to 36 for soft tissue, or a maximum of 42 if a pelvic fracture was involved, was created for all structures injured in the anterior and posterior zones.Results Using this new scoring system, 77% of patients survived with a pelvi-perineal trauma score (PPTS) below 5. There was a significant increase in mortality, number of pelvic fractures and amputations with increase in score when comparing the first group (score 1-5) to the second group (score 6-10). For scores between 6 and 16 survival was 42% and 22% for scores between 17 and 21. In our cohort of 62 survivors, 1 patient with an IED related pelvi-perineal injury had a 'theoretically un-survivable' maximal ISS of 75 and survived, whereas there were no survivors with a PPTS greater than 22 but this group had no-one with an ISS of 75 suggesting ISS is not an accurate reflection of the true severity of pelvi-perineal blast injury.Conclusions This scoring system is the initial part of a more complex logistic regression model that will contribute towards a unique trauma scoring system to aid surgical teams in predicting fluid requirements and operative timelines. In austere environments, it may also help to prevent futile resuscitations. Better correlation between measurement of severity and outcome would aid performance improvement monitoring. In the longer term it will also allow benchmarking of current survival rates and comparisons in the future.