ARTICLE IN PRESS doi:10.1510/icvts.2009.208371 Editorial
www.icvts.org
Institutional report - Cardiac general
Cardiothoracic Surgical Unit, Papworth Hospital, Cambridge CB3 8RE, UK b Wythenshawe Hospital Manchester, Manchester M23 9LT, UK
a
Abstract
2009 Published by European Association for Cardio-Thoracic Surgery
• • •
not preventable preventable (technical error) preventable (system error)
Brief Case Report Communication
*Corresponding author. Tel.: q44 1480 364 299; fax: q44 1480 364 744. E-mail address:
[email protected] (S.A.M. Nashef).
Historical Pages
We know very little about the mechanism or the sequence of events leading to low-risk death and we have no idea whether any such deaths are preventable, and if so, what
Nomenclature
•
Best Evidence Topic
•
an unavoidable and unpredictable event (pulmonary embolus, stroke etc.) a patient who truly is high risk but whose risk factors are not adequately represented by the risk model used failure in achieving a satisfactory cardiac outcome (FIASCO)
All cardiac surgical patients at Papworth Hospital are prospectively risk-stratified using Parsonnet, the additive EuroSCORE and the logistic EuroSCORE by the anaesthetist in charge of the case. Data on risk, operation details and outcomes are recorded contemporaneously in a computer database. Patients with a logistic EuroSCORE 0–2 who died in hospital after cardiac surgery were identified from the database and their case notes were obtained for review. The cases were assessed internally by two surgeons and by an independent external reviewer with considerable experience in reviewing cases for the National Confidential Enquiry into Postoperative Deaths (NCEPOD), a central organisation which has studied postoperative deaths in the UK since 1988. Deaths were categorized as cardiac or noncardiac and further subclassified into three categories:
State-of-the-art
•
2. Patients and methods
Follow-up Paper
Assessment of perioperative risk is possible using a number of risk models. Stratification of patients into risk groups generally correlates very well with immediate survival. High-risk subgroups have often been studied and institutions including ours w1x have reported on achievable outcomes in these patients. Low-risk patients have not been studied. We know that low-risk patients have a low mortality, but no operation is free of risk and there are occasional deaths in the lowest risk subgroup. In the absence of any identifiable risk factors of any kind, the mortality of coronary artery bypass grafting (CABG) is 0.4% and that of single valve surgery around 1% w2x. When death does happen in low-risk patients, we can postulate that the cause is likely to be one of the following three possibilities:
proportion can be prevented. This study aims to determine the mortality rate in our low-risk patients and to examine the cause of death. We sought to determine if death was an unforeseen catastrophe, or occurred in a high-risk patient who was not recognized as such by the risk model, or resulted from technical or system errors and thus was potentially preventable.
Negative Results
1. Introduction
Proposal for Bailout Procedure
Keywords: Statistics; Risk analysisymodeling; Perioperative care; Surgery complications
ESCVS Article
Death in low-risk patients is not studied as frequently as it is in other cardiac patients. We, therefore, sought to determine why some low-risk patients die after cardiac surgery. All low-risk patients (EuroSCOREF2) who died after cardiac surgery in one institution between 1996 and 2005 were included and meticulously studied by internal and independent external review of preoperative, operative and postoperative information from the case-notes and post-mortem findings. Deaths were classified into non-cardiac and cardiac and further subclassified into unavoidable deaths or due to failure in achieving a satisfactory cardiac outcome (FIASCO). Between 1996 and 2005, there were 16 deaths in 4294 low-risk patients (mortality 0.37%). Internal and external review agreed that nine deaths were non-preventable (CVA, bronchopneumonia, etc.) and that avoidable FIASCO accounted for seven deaths. Of the deaths considered to be preventable, all had probable errors of technique and three also had additional system errors. No cardiac operation is without risk. Mortality, though fortunately rare, can still occur, even in low-risk patients. Despite an extremely low mortality in the low-risk group FIASCO still accounts for nearly one-half of deaths. This suggests that mortality may be reduced even further as part of a quality improvement programme. 2009 Published by European Association for Cardio-Thoracic Surgery. All rights reserved.
Institutional Report
Received 6 April 2009; received in revised form 28 June 2009; accepted 29 June 2009
Protocol
Darren H. Freeda, Andrew J. Draina, Jago Kitcata, Mark T. Jonesb, Samer A.M. Nashefa,*
Work in Progress Report
Death in low-risk cardiac surgery: the failure to achieve a satisfactory cardiac outcome (FIASCO) study
New Ideas
Interactive CardioVascular and Thoracic Surgery 9 (2009) 623–625
ARTICLE IN PRESS D.H. Freed et al. / Interactive CardioVascular and Thoracic Surgery 9 (2009) 623–625
624
3. Results Between 1996 and 2005, 4294 patients with a logistic EuroSCOREF2 were operated on at Papworth Hospital. There were 16 deaths in this cohort giving an actual mortality rate of 0.37%. All were true low-risk deaths and no patients in this group had preoperatively documented rare or obscure risk factors not recognized by EuroSCORE. Of the 16 deaths, 13 (81%) were considered to be cardiac. Nine of 16 deaths were considered unavoidable. There was complete agreement between internal and external reviewers on the cause and preventability of death. The results are summarized in Table 1. Of the unavoidable deaths, there were two strokes, two sudden deaths on the ward where no technical problem was identified at post-mortem, one aortic dissection (starting from the cross-clamp site) that presented on postoperative day 3, one instance of preoperative malignant arrhythmias that persisted postoperatively despite revascularization, anti-arrhythmic drugs and ICD placement, one instance of bronchopneumonia, one instance of preoperative right ventricular infarction and one late haemorrhage in a patient who had cystic medial necrosis of the aorta. Of the seven deaths considered preventable, all had technical errors and three also had system errors. There were five perioperative myocardial infarcts (MI) identified at post-mortem. All grafts were patent and there was no evidence of preoperative MI, haemodynamic instability, ECG changes or chest pain preoperatively. We concluded that these were preventable deaths due to suboptimal myocardial protection (four patients had a single dose of 500 ml cold crystalloid cardioplegia with relatively long cross-clamp times (up to 55 min) and the 5th patient had intermittent cross-clamping with fibrillatory arrest). All
five were classified as technical errors as myocardial preservation strategy forms part of the surgical technique. In addition, two of these also had system errors, one because the level of input by the consultant surgeon was unclear and the other because the patient was returned to theatre by another consultant and the degree of communication between the consultants was unclear. Another patient sustained an epicardial tear on cardiac retraction; this was repaired but the patient bled profusely after extubation and further repair was unsuccessful. This was classified as a technical error for obvious reasons, but we also classified it as a system error because the level of consultant input was unclear. The last patient had a VF arrest on arrival in the ITU and was returned to theatre, placed on CPB and could not be successfully separated from CPB. The postmortem examination revealed significant left ventricular hypertrophy. We concluded this death was preventable because myocardial protection consisted of a single dose of crystalloid cardioplegia with a 58-min cross-clamp time. Again, this was classified as a technical error because the myocardial preservation strategy was probably insufficient for the length of ischaemic time. 4. Comment Quality control is an essential element in assuring optimal results in any process, whether in industry or in health care delivery. Medicine, particularly surgery, is no different. Mortality, while a crude outcome measure, is easy to track and quantify. Commonly, individual surgeon or institutional results are compared against actual rates that are derived from large retrospective analyses (registry data) or results are compared to predicted results that use any one of several risk scoring systems w3–6x. This practice is helpful
Table 1 Causes of low-risk cardiac surgical deaths Patient
Operation
Cause of death
Cardiac death
Preventable
Problem identified
1 2
CABG CABG
YES YES
YES – technical YES – technical
5 6
MV repair CABG
Perioperative MI; MOF Perioperative MI; pneumonia, ARDS VF arrest, LV hypertrophy Perioperative MI
YES YES
YES – technical YES – technical, system
7
CABG
Perioperative MI
YES
8
CABG
LV rupture
YES
9 3 4
CABG CABG CABG
YES YES YES
10 11 12 13 14
CABG CABG CABG CABG CABG
Perioperative MI Arrest on the ward – no PM Late aortic bleed (cystic medial necrosis) Arrest on the ward Aortic dissection Stroke Stroke Ischaemic heart disease
YES – technical, system YES – technical, system YES – technical NO NO
Myocardial preservation Myocardial preservation, obtuse marginal graft Myocardial preservation Myocardial preservation; problem with insertion of IABP, resulting in ruptured iliac artery Myocardial preservation; apparently minimal consultant input Epicardial tear led to rupture; apparently minimal consultant input Myocardial preservation
YES YES NO NO YES
NO NO NO NO NO
15 16
CABG CABG
Pneumonia, ARDS Persistent VFyVT
NO YES
NO NO
Bled from SV side branch; preoperative RV infarct
CABG, coronary artery bypass grafting; LV, left ventricle; RV, right ventricle; MV, mitral valve; MI, myocardial infarct; PM, post-mortem; MOF, multi-organ failure; VF, ventricular fibrillation; IHD, ischaemic heart disease; IABP, intra-aortic balloon pump.
ARTICLE IN PRESS D.H. Freed et al. / Interactive CardioVascular and Thoracic Surgery 9 (2009) 623–625
provide valuable information from which all patients may stand to benefit.
Work in Progress Report Protocol Institutional Report ESCVS Article
This is a retrospective, observational study. We have implicated cardioplegia strategy as a causative factor in perioperative MI. We have come to this conclusion by a process of elimination and consensus. There may be other factors that could account for perioperative MI that we have not taken into account. We have also identified potential human factors leading to unsatisfactory outcomes, however, analysis of specific factors would have to be undertaken in tens of thousands of low-risk patients before firm conclusions could be drawn. The study period is relatively long (nine years), and cardiac surgical techniques have changed over this period, particularly with respect to myocardial protection strategies, intraoperative assessment of aortic pathology, etc. Further studies are ongoing to determine the impact of these techniques on outcomes.
New Ideas
5. Limitations
Editorial
References
Negative Results Follow-up Paper State-of-the-art Best Evidence Topic Nomenclature Historical Pages
w1x Stoica S, Balaji H, Helmy A, Kitcat J, Freeman C, Sharples L, Nashef SA. Against the odds: long-term outcome of drastic-risk cardiac surgery. J Thorac Cardiovasc Surg 2006;132:1226–1228. w2x Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E, Cortina J, David M, Faichney A, Gabrielle F, Gams E, Harjula A, Jones MT, Pintor PP, Salamon R, Thulin L. Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 1999;15:816–822. w3x Michel P, Roques F, Nashef SA. Logistic or additive EuroSCORE for highrisk patients? Eur J Cardiothorac Surg 2003;23:684–687. w4x Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9–13. w5x Asimakopoulos G, Al-Ruzzeh S, Ambler G, Omar RZ, Punjabi P, Amrani M, Taylor KM. An evaluation of existing risk stratification models as a tool for comparison of surgical performances for coronary artery bypass grafting between institutions. Eur J Cardiothorac Surg 2003;23:935– 941. w6x Parsonnet V, Dean D, Bernstein AD. A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 1989;79(6 Pt 2):I3–I12. w7x Cohen G, Borger MA, Weisel RD, Rao V. Intraoperative myocardial protection: current trends and future perspectives. Ann Thorac Surg 1999;68:1995–2001. w8x Guru V, Omura J, Alghamdi AA, Weisel R, Fremes SE. Is blood superior to crystalloid cardioplegia? A meta-analysis of randomized clinical trials. Circulation 2006;114(1 Suppl):I331–I338. w9x Nicolini F, Beghi C, Muscari C, Agostinelli A, Maria BA, Spaggiari I, Gherli T. Myocardial protection in adult cardiac surgery: current options and future challenges. Eur J Cardiothorac Surg 2003;24:986–993. w10x Goodwin AT, Birdi I, Ramesh TP, Taylor GJ, Nashef SA, Dunning JJ, Large SR. Effect of surgical training on outcome and hospital costs in coronary surgery. Heart 2001;85:454–457. w11x Baskett RJ, Kalavrouziotis D, Buth KJ, Hirsch GM, Sullivan JA. Training residents in mitral valve surgery. Ann Thorac Surg 2004;78:1236–1240. w12x Baskett RJ, Buth KJ, Legare JF, Hassan A, Hancock FC, Hirsch GM, Ross DB, Sullivan JA. Is it safe to train residents to perform cardiac surgery? Ann Thorac Surg 2002;74:1043–1048.
Proposal for Bailout Procedure
in identifying outliers although the cause of excess risk may not be readily apparent. Examination of results of surgery on very high-risk patients is a popular practice, and many authors enjoy reporting excellent results (much better than predicted) with difficult and complex procedures. However, the results of low-risk patients undergoing straightforward operations have not been examined. For that reason, we chose to examine our results in this latter patient group to determine what the exact risk is, and try to identify areas that could be improved so that the risk could be even further minimized. We examined the actual mortality rate of patients with logistic EuroSCOREF2 receiving cardiac surgery at Papworth hospital and conclude that death is infrequent in this group (0.37%), however, over a third of deaths are potentially preventable. This suggests that further improvement in outcomes is possible through modification of individual technique or a change in the systematic delivery of cardiac surgical care or training. Defining a death as preventable is challenging, and for the purposes of this study, we relied on consensus opinion from individual experienced cardiac surgeons who were not directly involved in the cases in question. Identifying myocardial preservation as a problem is an important finding because a change in practice would benefit all patients undergoing cardiac surgery in our hospital. It is interesting to note that no patient who received multi-dose cold blood cardioplegia appeared in this patient group. Myocardial preservation techniques have evolved substantially since the beginning of open heart surgery, with a range of techniques from intermittent cross-clamping to single dose or multi-dose cold crystalloid and blood cardioplegia w7x. Although there is no clearly superior strategy, the length of warm ischaemia directly correlates with clinical outcomes. We, and others w7–9x suggest that multi-dose cold blood cardioplegia is one of the safer strategies. This study also highlights the potential problems associated with surgical training. Training of junior surgeons has previously been shown not to be associated with poor outcomes w10–12x. Independent operating is an important component of the training process; however, the degree of supervision varies between institutions and between consultant surgeons in the same institution. Where we found system errors related to the degree of supervision, they reflected a lack of clarity in the requirement for senior surgical input and in the identity of the surgeon who is ultimately responsible and must be kept informed. With hospital mortality at 0.37%, these are excellent results in low-risk patients, but we have found that even here, there is some potential room for improvement. Our study highlights the importance of regular institutional review processes in identifying problems and correcting them as part of a continual quality improvement programme for patient outcomes. Perhaps, other institutions will also find that review of their own low-risk deaths will
625
Brief Case Report Communication
