Outlier identification and monitoring of institutional or clinician performance: an overview of statistical methods and application to national audit data.

BACKGROUND: Institutions or clinicians (units) are often compared according to a performance indicator such as in-hospital mortality. Several approaches have been proposed for the detection of outlying units, whose performance deviates from the overall performance. METHODS: We provide an overview of three approaches commonly used to monitor institutional performances for outlier detection. These are the common-mean model, the 'Normal-Poisson' random effects model and the 'Logistic' random effects model. For the latter we also propose a visualisation technique. The common-mean model assumes that the underlying true performance of all units is equal and that any observed variation between units is due to chance. Even after applying case-mix adjustment, this assumption is often violated due to overdispersion and a post-hoc correction may need to be applied. The random effects models relax this assumption and explicitly allow the true performance to differ between units, thus offering a more flexible approach. We discuss the strengths and weaknesses of each approach and illustrate their application using audit data from England and Wales on Adult Cardiac Surgery (ACS) and Percutaneous Coronary Intervention (PCI). RESULTS: In general, the overdispersion-corrected common-mean model and the random effects approaches produced similar p-values for the detection of outliers. For the ACS dataset (41 hospitals) three outliers were identified in total but only one was identified by all methods above. For the PCI dataset (88 hospitals), seven outliers were identified in total but only two were identified by all methods. The common-mean model uncorrected for overdispersion produced several more outliers. The reason for observing similar p-values for all three approaches could be attributed to the fact that the between-hospital variance was relatively small in both datasets, resulting only in a mild violation of the common-mean assumption; in this situation, the overdispersion correction worked well. CONCLUSION: If the common-mean assumption is likely to hold, all three methods are appropriate to use for outlier detection and their results should be similar. Random effect methods may be the preferred approach when the common-mean assumption is likely to be violated.

Coronary artery bypass surgery in the UK, trends in activity and outcomes from a 15-year complete national series.

OBJECTIVES: The aim of this study was to review the UK national trends in activity and outcome in coronary artery bypass graft (CABG) over a 15-year period (2002-2016). METHODS: Validated data collected (2002-2016) and uploaded to National Institute for Cardiovascular Outcomes Research were used to generate summary data from the National Adult Cardiac Surgery Audit Database for the analysis. Logistic European System of Cardiac Operative Risk Evaluation was used for risk stratification with recalibration applied for governance. Data were analysed by financial year and presented as numerical, categorical, %, mean and standard deviation where appropriate. Mortality was recorded as death in hospital at any time after index CABG operation. RESULTS: A total of 347 626 CABG procedures (282 883 isolated CABG, 61 109 CABG and valve and 4132 redo CABG) were recorded. Over this period annual activity reduced from 66.6% of workload to 41.7%. The mean age for isolated CABG was 65.7 years. The mean log European System of Cardiac Operative Risk Evaluation was 3.1, 5.9 and 23.2 for elective, urgent and emergency isolated CABG, respectively. There was a decline in the observed mortality for all procedures. Overall mortality for isolated CABG surgery is now 1.0% and only 0.6% for elective operations. CONCLUSIONS: Quality of care and risk-adjusted mortality rates have consistently improved over the last 15 years despite the increasing risk profile of patients. There have been a consistent decline in overall case volumes and a three-fold increase in elderly cases.

Cardiac surgery in older patients: hospital outcomes during a 15-year period from a complete national series.

OBJECTIVES: The objective was to review national trends in activity and hospital outcomes in older patients having cardiac surgery over a 15-year time period. METHODS: Data were collected prospectively and uploaded to the National Institute for Cardiovascular Outcomes Research electronically. Data were validated, cleaned and processed using reproducible algorithms. Mortality was death in hospital after index operation. RESULTS: A total of 227 442 cardiac procedures were recorded in patients aged ≥70 years of which 46 354 were in those aged ≥80 years. Overall patients aged ≥70 years represented 43% of all adult cardiac surgery in the most recent study year. The annual proportion of surgery in patients ≥80 years increased from 4.1% to 10.8% between the first and last study years. There has been a significant linear increase in octogenarian valve [ß 67.44, 95% confidence interval (CI) 55.04 to 79.83, P < 0.001] and coronary artery bypass graft surgery (ß 32.53, 95% CI 6.16 to 58.90, P = 0.020) patients. In-hospital mortality reduced significantly for patients aged 70-79 years (ß -0.17, 95% CI -0.20 to -0.13, P < 0.001) and all patients aged ≥80 (ß -0.37, 95% CI -0.45 to -0.30, P < 0.001). The median length of hospital stay was 7 days for 70-79 and 9 days for ≥80 group, compared with 7 days for the whole cohort <70 years. CONCLUSIONS: This study represents the largest complete validated national dataset of cardiac surgery in the entire population of older patients. Octogenarians represent 11% of adult patients having cardiac surgery by the end of the study period, a three-fold increase from the start. In-hospital mortality in patients aged ≥80 years halved during study period to only 4% despite high logistic EuroSCORE of 15%. Cardiac surgery in octogenarians places a higher demand on resources, however, with an increased postoperative length of stay.

Aortic valve surgery in the UK, trends in activity and outcomes from a 15-year complete national series.

OBJECTIVES: Since the turn of the century, cardiac surgery has evolved quite notably. This study sought to investigate the trends in aortic valve surgery activity and subsequent outcomes in the UK by using a mandatory national cardiac surgical clinical database within the context of a comprehensive public healthcare system (National Health Service). METHODS: The UK National Adult Cardiac Surgery Audit database provided data for aortic valve surgery procedures performed between 2002 and 2016, and the data were validated and cleaned using reproducible algorithms. The findings and trends in in activity and outcomes were then analysed by financial year. RESULTS: During the study period, a total of 148 862 procedures were performed, with the number of cases per year peaking in 2014/2015 at 12 483. The mean in-hospital mortality rate for all aortic valve surgery has fallen from 5.6% to 3.4%, despite an increase in patient age and mean logistic EuroSCORE. While the number of isolated aortic valve replacements has remained stable at around 5000 per year, aortic valve replacement and coronary artery bypass graft have increased to over 3200 with transcatheter aortic valve implantation activity continuing to increase. CONCLUSIONS: This study demonstrates that despite an increase in patient risk profile, there has been a consistent reduction in in-hospital mortality within all aortic valve surgery procedures performed in the UK over a 15-year period. Increasing catheter-based interventions have not yet resulted in a significant decrease in surgical aortic valve replacements in the UK.

Indirect Impact of the COVID-19 Pandemic on Activity and Outcomes of Transcatheter and Surgical Treatment of Aortic Stenosis in England.

[Figure: see text].

Revascularisation strategies in patients with significant left main coronary disease during the COVID-19 pandemic.

BACKGROUND: There are limited data on the impact of the COVID-19 pandemic on left main (LM) coronary revascularisation activity, choice of revascularisation strategy, and post-procedural outcomes. METHODS: All patients with LM disease (≥50% stenosis) undergoing coronary revascularisation in England between January 1, 2017 and August 19, 2020 were included (n = 22,235), stratified by time-period (pre-COVID: 01/01/2017-29/2/2020; COVID: 1/3/2020-19/8/2020) and revascularisation strategy (percutaneous coronary intervention (PCI) vs. coronary artery bypass grafting (CABG). Logistic regression models were performed to examine odds ratio (OR) of 1) receipt of CABG (vs. PCI) and 2) in-hospital and 30-day postprocedural mortality, in the COVID-19 period (vs. pre-COVID). RESULTS: There was a decline of 1,354 LM revascularisation procedures between March 1, 2020 and July 31, 2020 compared with previous years' (2017-2019) averages (-48.8%). An increased utilization of PCI over CABG was observed in the COVID period (receipt of CABG vs. PCI: OR 0.46 [0.39, 0.53] compared with 2017), consistent across all age groups. No difference in adjusted in-hospital or 30-day mortality was observed between pre-COVID and COVID periods for both PCI (odds ratio (OR): 0.72 [0.51. 1.02] and 0.83 [0.62, 1.11], respectively) and CABG (OR 0.98 [0.45, 2.14] and 1.51 [0.77, 2.98], respectively) groups. CONCLUSION: LM revascularisation activity has significantly declined during the COVID period, with a shift towards PCI as the preferred strategy. Postprocedural mortality within each revascularisation group was similar in the pre-COVID and COVID periods, reflecting maintenance in quality of outcomes during the pandemic. Future measures are required to safely restore LM revascularisation activity to pre-COVID levels.

Trends and outcomes for cardiac surgery in the United Kingdom from 2002 to 2016.

Objectives: Cardiac surgery has evolved significantly since the turn of the century. The objective of this study was to investigate trends in cardiac surgery activity and outcomes in the United Kingdom utilizing a mandatory national cardiac surgical clinical database in the context of a comprehensive public health care system (ie, the UK National Health Service). Methods: Data for all cardiac surgery procedures performed between 2002 and 2016 were extracted from the UK National Adult Cardiac Surgery Audit database. Data are validated and cleaned using reproducible algorithms. Trends in activity and outcomes were analyzed by fiscal year using linear regression. Results: A total of 534,067 procedures were performed during the study period with the number of cases per year peaking in 2008/2009 at 41,426. Despite an increase in patient age and mean logistic European System for Cardiac Operative Risk Evaluation score, the in-hospital mortality rate for all cardiac surgery has fallen from 4.0% to 2.8% (P < .001). The number of isolated coronary artery bypass graft procedures has steadily declined but the total number of valve procedures has steadily increased (both P values < .001). The number of thoracic aortic procedures performed each year has doubled (P < .001), but the incidence of redo procedures has steadily declined. The proportion of emergency and salvage procedures has remained stable. Conclusions: This study, which covers all cardiac surgery procedures performed in the United Kingdom for fiscal years between 2002 and 2016, demonstrates that despite an increase in patient risk profile, there has been a consistent reduction in in-hospital mortality. A number of other markers associated with quality have also improved.

Impact of COVID-19 on cardiac procedure activity in England and associated 30-day mortality.

AIMS: Limited data exist on the impact of COVID-19 on national changes in cardiac procedure activity, including patient characteristics and clinical outcomes before and during the COVID-19 pandemic. METHODS AND RESULTS: All major cardiac procedures (n = 374 899) performed between 1 January and 31 May for the years 2018, 2019, and 2020 were analysed, stratified by procedure type and time-period (pre-COVID: January-May 2018 and 2019 and January-February 2020 and COVID: March-May 2020). Multivariable logistic regression was performed to examine the odds ratio (OR) of 30-day mortality for procedures performed in the COVID period. Overall, there was a deficit of 45 501 procedures during the COVID period compared to the monthly averages (March-May) in 2018-2019. Cardiac catheterization and device implantations were the most affected in terms of numbers (n = 19 637 and n = 10 453), whereas surgical procedures such as mitral valve replacement, other valve replacement/repair, atrioseptal defect/ventriculoseptal defect repair, and coronary artery bypass grafting were the most affected as a relative percentage difference (Δ) to previous years' averages. Transcatheter aortic valve replacement was the least affected (Δ -10.6%). No difference in 30-day mortality was observed between pre-COVID and COVID time-periods for all cardiac procedures except cardiac catheterization [OR 1.25 95% confidence interval (CI) 1.07-1.47, P = 0.006] and cardiac device implantation (OR 1.35 95% CI 1.15-1.58, P < 0.001). CONCLUSION: Cardiac procedural activity has significantly declined across England during the COVID-19 pandemic, with a deficit in excess of 45 000 procedures, without an increase in risk of mortality for most cardiac procedures performed during the pandemic. Major restructuring of cardiac services is necessary to deal with this deficit, which would inevitably impact long-term morbidity and mortality.

Manubrium-limited ministernotomy versus conventional sternotomy for aortic valve replacement (MAVRIC): study protocol for a randomised controlled trial.

BACKGROUND: Aortic valve replacement is one of the most common cardiac surgical procedures performed worldwide. Conventional aortic valve replacement surgery is performed via a median sternotomy; the sternum is divided completely from the sternal notch to the xiphisternum. Minimally invasive aortic valve replacement, using a new technique called manubrium-limited ministernotomy, divides only the manubrium from the sternal notch to 1 cm below the manubrio-sternal junction. More than one third of patients undergoing conventional sternotomy develop clinically significant bleeding requiring post-operative red blood cell transfusion. Case series data suggest a potentially clinically significant difference in red blood cell transfusion requirements between the two techniques. Given the implications for National Health Service resources and patient outcomes, a definitive trial is needed. METHODS/DESIGN: This is a single-centre, single-blind, randomised controlled trial comparing aortic valve replacement surgery using manubrium-limited ministernotomy (intervention) and conventional median sternotomy (usual care). Two hundred and seventy patients will be randomised in a 1:1 ratio between the intervention and control arms, stratified by baseline logistic EuroSCORE and haemoglobin value. Patients will be followed for 12 weeks from discharge following their index operation. The primary outcome is the proportion of patients who receive a red blood cell transfusion post-operatively within 7 days of surgery. Secondary outcomes include red blood cell and blood product transfusions, blood loss, re-operation rates, sternal wound pain, quality of life, markers of inflammatory response, hospital discharge, health care utilisation, cost and cost effectiveness and adverse events. DISCUSSION: This is the first trial to examine aortic valve replacement via manubrium-limited ministernotomy versus conventional sternotomy when comparing red blood cell transfusion rates following surgery. Surgical trials present significant challenges; strengths of this trial include a rigorous research design, standardised surgery performed by experienced consultant cardiothoracic surgeons, an agreed anaesthetic regimen, patient blinding and consultant-led patient recruitment. The MAVRIC trial will demonstrate that complex surgical trials can be delivered to exemplary standards and provide the community with the knowledge required to inform future care for patients requiring aortic valve replacement surgery. TRIAL REGISTRATION: International Standard Randomised Controlled Trial Number (ISRCTN) ISRCTN29567910 . Registered on 3 February 2014.

Manubrium-limited sternotomy decreases blood loss after aortic valve replacement surgery.

OBJECTIVES: Minimally invasive surgical approaches for aortic valve replacement (AVR) are growing in popularity in an attempt to decrease morbidity from conventional surgery. We have adopted a technique that divides only the manubrium and spares the body of the sternum. We sought to determine whether patients benefit from this less-invasive approach. METHODS: We retrospectively analysed our prospectively maintained database to review all isolated aortic valve replacements performed in an 18-month period from November 2011 to April 2013. RESULTS: One hundred and ninety-one patients were identified, 98 underwent manubrium-limited sternotomy (Mini-AVR) and 93 had a conventional median sternotomy (AVR). The two groups were well matched for preoperative variables and risk (mean logistic EuroSCORE mini-AVR 7.15 vs AVR 6.55, P = 0.47). Mean cardiopulmonary bypass and aortic cross-clamp times were 10 and 6 min longer, respectively, in the mini-AVR group (mean values 88 vs 78 min, P = 0.00040, and 66 vs 60 min, P = 0.0078, respectively). Mini-AVR patients had significantly less postoperative blood loss, 332 vs 513 ml, P = 0.00021, and were less likely to require blood products (fresh-frozen plasma and platelets), 24 vs 36%, P = 0.042. Postoperative complications and length of stay were similar (discharge on or before Day 4; mini-AVR 15 vs AVR 8%, P = 0.17). Valve outcome (paravalvular leak mini-AVR 2 vs AVR 1%, P = 1.00) and survival (mini-AVR 99 vs AVR 97%, P = 0.36) were equal. CONCLUSIONS: A manubrium-limited approach maintains outcomes achieved for aortic valve replacement by conventional sternotomy while significantly reducing postoperative blood loss and transfusion of blood products.

Should you stand on the left or the right of a patient with dextrocardia who needs coronary surgery?

A best evidence topic in cardiac surgery was written according to a structured protocol. The question addressed was which side of the operating table you should stand on when carrying out surgical revascularization on a patient with dextrocardia. Altogether 40 papers were found using the reported search, of which 19 represented the best evidence to answer the clinical question. The authors, journal, date and country of publication, patient group studied, study type, relevant outcomes, and results of these papers are tabulated. The side on which the operating surgeon stood was mentioned in 20 out of the 24 cases. Surgery was carried out from the conventional right side of the patient in 5 cases, while in 10 cases, it was carried out from the left side. The surgeon needed to switch sides to facilitate surgery in three cases. In addition, the right internal mammary artery (RIMA) was anastomosed to the left anterior descending artery (LAD) in 16 cases. Of these, surgery was carried out from the left side in 11 cases. The left internal mammary artery (LIMA) to LAD anastomosis was carried out in two cases, one of which was a free LIMA graft. In six cases, only vein grafts were used. Fourteen cases were carried out using cardiopulmonary bypass while 10 cases were carried out as off-pump cases with one conversion. The majority of patients were operated on from the left of the table. More cases were performed with the RIMA as the conduit of choice to the LAD.

Right pulmonary artery occlusion by an acute dissecting aneurysm of the ascending aorta.

We describe the case of a 76-year old female who presented with a Type A aortic dissection requiring repair with an interposition graft and aortic valve replacement. Post-operatively she had clinical features and computerised tomographic images suggestive of a pulmonary embolus and died 24 hours later. The extremely rare finding of intramural thrombus occluding the right pulmonary artery was seen at post mortem.

Clinical outcome versus post-mortem finding in thoracic surgery: a 10-year experience.

OBJECTIVES: The role of post-mortem following thoracic surgery has not been previously studied. Most importantly, the clinical diagnosis of thoracic surgical mortality cannot be certain unless post-mortem analysis has been performed. METHODS: Consecutive post-mortem data were collected on 110 patients between 1992 and 2001 (66.3% of all in-hospital thoracic surgical deaths). Clinically attributed causes of death were compared with post-mortem findings. RESULTS: A total of 4302 thoracic operations were performed during the 10-year period, with overall mortality 3.8%. The mean age was 63.6 years (range 21-87) with 73.6% male. In the 110 patients undergoing post-mortem examination, the operations performed were pneumonectomy 24.5%, lobectomy 14.5%, oesophagectomy 12.7%, lung biopsy 8.2%, pleurectomy/bullectomy 6.4%, decortication 4.5%, lung volume reduction 1.8%, other thoracic 13.6%, other oesophageal 9.1%, and other procedures 4.5%. The mean time to death was 12.5 days (range 0-85). The causes of death were respiratory 47.3%, cardiac 16.4%, multiple organ failure 8.2%, sepsis 6.4%, gastrointestinal 4.5%, haemorrhage/technical failure 10%, and others 7.3%. Post-mortem revealed an unsuspected cause of death in 34 (31%) patients, comprising pulmonary 17, cardiac 5, gastrointestinal 3, haemorrhage/technical failure 2, multiple organ failure 2 and other 5. CONCLUSION: Post-mortem determined unsuspected diagnoses in a high proportion of patients undergoing thoracic surgery. Post-mortem continues to be the 'gold standard' method for attributing the cause of death. Accurate outcome data following thoracic surgery are essential for proper audit, and hence for improvements in clinical practice to occur.

Outcomes in emergency redo cardiac surgery: cost, benefit and risk assessment.

Emergency redo surgery is rare, but may be required in patients with conditions such as endocarditis, unstable angina and acute aortic dissection. To date there are no published data on the outcome of these difficult patients. Prospective consecutive data were collected from a single institution on 65 patients (51 male) undergoing coronary artery bypass grafts (27), mitral valve replacement or repair (13), aortic valve replacement (7), aortic surgery (13), and other (combined procedures) (5). Indications for surgery were unstable angina (12), endocarditis (11), resuscitation/catheter lab complications (11), torn prosthetic leaflet (6), aortic dissection (4), paraprosthetic leak (2), other (19). Mean ITU stay was 45 h (0-284) and hospital stay was 13.3 days (0-68). There were 14 intra-operative deaths and 14 further in-hospital deaths (overall mortality 43%). Predicted mortality rates were 26% (Parsonnet), 11% (EuroSCORE) and 31% (EuroSCORE logistic). Mean hospital cost per patient was 18,299 euros (or 32,147 euros per hospital survivor). In conclusion, the mortality in these difficult patients is very high, however, often no other treatment option is available. More sophisticated models, such as EuroSCORE logistic, may allow better prediction of risk in very high risk cases.

Differential effects of endothelin-1 on isolated working rat hearts before and after ischaemia and reperfusion.

Endothelin (ET) may have both detrimental (reduced coronary flow) and beneficial effects (positive inotrope, reduced arrhythmogenesis) following ischaemia. We examined the effects of ET on cardiac function during reperfusion following prolonged hypothermic cardioplegic arrest in a protocol mimicking cardiac transplantation. Isolated working rat hearts were perfused with Krebs buffer to which increasing concentrations of ET-1 or sarafotoxin S6c had been added. Identical experiments were performed after 4 h of cardioplegic arrest at 4 degrees C. Under pre-ischaemic conditions ET-1 caused a dose-dependent decrease in cardiac function compared with controls. In contrast, following ischaemia low doses of ET-1 (10(-10) M) caused a significant and beneficial increase in cardiac output (109.1% versus 81.3%), dP/dt i.e. the rate of change of pressure with time (94.7% versus 75.6%) and stroke volume (100.3% versus 77.5%) compared with controls (P<0.05). At higher doses of ET-1 there was a detrimental effect on cardiac output, dP/dt and stroke volume similar to that seen prior to ischaemia. Sarafotoxin S6c had no significant effect pre or post ischaemia on any of the parameters measured compared with controls (P=not significant). ET-1 at low concentrations during reperfusion can improve the recovery of cardiac function mediated via ET(A) receptors. ET may play an important physiological role in the recovery of cardiac function following prolonged ischaemia.

Up-regulation of endothelin-converting-enzyme mRNA expression following cardioplegic arrest.

Endothelin-1 (ET-1) is believed to play an important role in cardiac ischaemia/reperfusion injury. ET-1 is synthesized from preproET-1 by the action of ET-converting enzyme (ECE). It is unclear to what extent the ET system is activated following prolonged ischaemia. In this study we used a model mimicking the conditions of the donor heart during transplantation. Isolated rat hearts perfused with Krebs-Henseleit buffer were subjected to 30 min of normothermic perfusion, then 4 h of cardioplegic arrest at 4 degrees C with St Thomas' Hospital solution, followed by reperfusion for 2 h. Hearts were freeze-clamped at different time points during the protocol. Using quantitative reverse transcription-PCR, relative levels of ET-1 and ECE mRNA expression were measured and compared with a housekeeping gene (ribosomal protein L32). During reperfusion there was a consistent decrease in coronary flow to approx. 85-90% of pre-ischaemic flow. There was no significant alteration in preproET-1 mRNA expression during 2 h of reperfusion. However, ECE mRNA expression was increased by 77.5% at 1 h and by 74.6% at 2 h following ischaemia compared with pre-ischaemic values (P<0.05). Thus we conclude that ECE mRNA expression is increased following prolonged hypothermic cardioplegic arrest. Elevations in the expression of this enzyme may help to explain the role of the ET system in the pathogenesis of ischaemia/reperfusion injury following cardiac surgery and transplantation.