| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
© 2003 European Association of Cardio-Thoracic Surgery
Outcomes in emergency redo cardiac surgery: cost, benefit and risk assessmentPapworth Hospital, Papworth Everard, Cambridge, CB3 8RE, UK
* Corresponding author. Tel.: +44-1480-364299; fax: +44-1480-364744 Received September 30, 2002; received in revised form February 7, 2003; accepted February 11, 2003
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 euro18,299 (or Euro32,147 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.
Key Words: Coronary artery surgery; Outcome; EuroSCORE; Resternotomy; Emergency
Cardiac surgery requiring resternotomy (so-called ‘redo’ surgery) is technically difficult and carries a higher operative risk than a first-time operation. The particular problems are well recognised and include difficulty with access to the heart (due to adhesions, scarring, fibrosis or calcification around the operative site) making dissection and suture placement difficult, prolonged operation times and increased postoperative mortality and morbidity. In the UK, redo cardiac surgery accounts for approximately 5% of the surgical workload [1]. The crude mortality rate for redo coronary artery surgery between 1996 and 1999 was 7.4% compared to 2.5% for first time operations [1]. In addition, mean post-operative stay was significantly longer for redo surgery (10.1 vs. 8.5 days). Likewise, emergency cardiac surgery is known to carry a much higher risk than elective procedures, mainly associated with the underlying condition that has precipitated the operation such as an acute coronary syndrome. The crude mortality rates in the UK between 1996 and 1999 for emergency coronary surgery were 11.8% compared to 2.0% for elective cases [1]. Fortunately, the combination of both emergency and redo surgery is rare, but may be required in patients with conditions such as endocarditis, unstable angina and acute aortic dissection. Although outcome might be expected to be poor in emergency redo surgery, there are no published data on this subject. We sought to explore this small but important subgroup of our practice, to assess the cost and benefit of surgical intervention and to look into the difficulties of risk stratification in such very high risk patients.
Consecutive data were collected on 65 patients from a single institution between April 1996 and October 2001. Data were collected from the hospital's computerised surgical database (entered prospectively at the time of operation) and included full risk stratification variables for both EuroSCORE and Parsonnet scoring systems. Details were checked and further information, such as the indications for surgery, was obtained from the patients’ notes. An emergency was strictly defined as an operation carried out on referral before the beginning of the following working day. Operations within 24 h of previous cardiac surgery and transplantation procedures were excluded. Predicted mortality rates were calculated using the EuroSCORE and Parsonnet systems [2,3]. In addition, the same variables were used to calculate a EuroSCORE logistic predicted mortality rate. The EuroSCORE logistic uses the full ß-coefficients of the risk factors to calculate a precise predicted mortality risk according to the equation in Appendix A. Predicted mortality rates for each of the three scoring systems were then compared with the observed rates. An estimate of the hospital costs for each patient was based on the addition of variable costs [euro per theatre minute, ITU hour, hospital day and disposables (e.g. valve prostheses, intra-aortic balloon pumps)] to fixed costs (medical staff, administration) as previously described [4]. Results are expressed as mean values (±S.D.) where appropriate.
During the study period 65 patients [51 male (78%)] underwent emergency resternotomy. Mean age at resternotomy was 64.2 years (range 28–82). Mean time from the previous operation was 6.3 years (range 1 day–24.7 years) (Table 1). In 15 patients the date of the previous operation was not precisely known, however, none of these patients had undergone surgery at our institution in the preceding month. The mean number of previous operations was 1.2 per patient (one previous operation, 58 patients; two operations, six patients; three operations, one patient).
The indications for resternotomy are summarised in Table 2. The operations performed are summarised in Table 3. Mean bypass and cross-clamp times were 107 (±50) min and 64 (±36) min, respectively. Mean theatre time was 306 min (range 140–905), 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%). The mean time to death was 10.1 days (range 0–68). Predicted mortality rates were 26% (Parsonnet), 11% (EuroSCORE) and 31% (EuroSCORE Logistic). EuroSCORE logistic predicted mortality rates ranged from 5% to 96%. The highest predicted mortality rate in a patient surviving operation was 90% The major postoperative complications are summarised in Table 4. The mean hospital cost per patient was  18,299 (or 32,147 per hospital survivor).
The mortality and morbidity in patients undergoing emergency cardiac surgery requiring resternotomy is high. The mortality in our series was 43%, with half of these occurring intra-operatively. The most important morbidity in survivors was stroke in 9%. The stroke rate is significantly higher than that expected following routine cardiac surgery, where rates of 1–3% are more usual. However, the stroke rate is perhaps not surprising given the high proportion of patients requiring resuscitation pre-operatively (17%), or undergoing surgery on the thoracic aorta (20%). Although the mortality in these patients was high, it probably compares very favourably with the outcome without surgery. Patients with conditions such as acute aortic dissection and prosthetic valve endocarditis and patients undergoing salvage procedures following cardiopulmonary resuscitation face a mortality rate that may approach 100% without surgical intervention, although this cannot be proven without a control group. Despite the high mortality rate, the hospital costs per life saved (euro 32,147) are acceptable and roughly equivalent to three first-time CABG or two valve replacement procedures in the UK. In the current era of media and public access to surgeon-specific outcome data in many countries, there is increasing concern that high-risk patients such as these may be denied life-saving surgery because of their potential impact on the published results. The ‘cardiac surgical paradox’, however, remains true: the higher the risk of surgery, the more an individual patient stands to gain from the procedure. It is vital, therefore, that surgeons are not penalised for being prepared to take on these high risk patients. Additive risk scoring models tend to underestimate the risk in very high risk cases. This has been confirmed in this study where both Parsonnet and EuroSCORE underestimated the risk of death (25% and 11%, respectively, vs. an actual mortality of 43%). The logistic regression score based on the same variables as EuroSCORE had a predicted mortality of 31% and may therefore allow better prediction of risk in these very high risk cases, although further work is required to confirm this. In conclusion, the mortality in these difficult patients undergoing emergency redo cardiac surgery is very high. However, there is often no other treatment option available and these operations can save lives at a cost that is not prohibitive. The two most widely used risk scoring systems do not adequately predict outcome in these patients. More sophisticated models, such as EuroSCORE logistic, may allow better prediction of risk in very high risk cases.
EuroSCORE logistic regression equation
Beta coefficients for EuroSCORE logistic equation
Conference discussion Dr T. Aberg (Umea, Sweden): I have some problems with this paper. You have collected two things that are risk factors, emergency and reoperation, however, I feel that also other risk factors perhaps should be addressed. And also I would like to ask you why doesn't it fit, and what is your definition for emergency? Mr Ooi: At Papworth Hospital emergency is defined and set out before the study: if there is any surgical referral, we carry out the operation before the next working day. Dr Aberg: So it may entail at least a 24-hour waiting period before? Mr Ooi: Yes, at most. The next working day is 8 o'clock. If they are referred at 3 o'clock or midnight, we have to operate on them before the next working day, yes. Dr M. Antunes (Coimbra, Portugal): Any procedure that carries a mortality of 43% has to be scrutinized very carefully. And I am not really concerned about the impact it has on surgeons but the impact it has on patients. That is 43% of patients who died. I am not entirely sure that in 2002 some of these cases that are considered emergencies, whether redos or first operations, are not adequately treatable by conservative (medical) treatment temporarily, in order to improve the condition of the patient. And I wasn't entirely elucidated by the author's presentation to the exact condition of patients when they were taken to surgery as emergency cases, apart from the cases who came as life-rescue procedures from the cath lab. For example, I am not sure that a patient with a periprosthetic leak, or even a torn leaflet, does require emergency procedures, but I am not sure what was the percentage of emergency procedures with regards to your overall experience. Can you give us some more information about that? Mr. Ooi: We say endocarditis and valve leaflet leakage is because the patient went into acute heart failure and pulmonary edema and medical treatment has failed, and so they have to refer to a surgeon urgently, and because this is a mechanical problem, eventually you need a mechanical solution such as a surgeon to repair the valve and to operate on them. As far as the figures are concerned, we showed here that an acute mechanical problem was presented in about 30 to 50% of them, including infective endocarditis. Dr V. DiSesa (West Chester, Pennsylvania): You listed intra-aortic balloon pump placement as a complication in about nine patients, and the largest single percentage of your patients are those with coronary artery disease. I would rather look at the balloon pump, and this was related to the previous question, as a way to convert a patient who may be an emergency to a patient who could be done under less urgent circumstances. Clearly the condition of the patients when they arrive in the operating room has a lot to do with the outcome. I was wondering why you didn't use more balloon pumps or other ways of stabilizing the patients before taking them to the operating room. Mr Ooi: This study was done before the year 2001. At our institute there is an increasing usage of intra-aortic balloon pumps nowadays before we operate. Yes, we do that. I mean, this study started in 1996, before that we tended to use the balloon pump mainly for patients with ischaemic left ventricular. In this study, the patients also suffered from heart failure, which they needed intra-aortic balloon pumps. So, we had to use them. Mr S. Nashef (Cambridge, United Kingdom): I would just like to add another response to the two questioners who asked questions just now. Of course, many patients are stabilized with an intra-aortic balloon and operated semi-electively, and many patients are treated conservatively and stabilized. These patients that we presented today are those in whom all such efforts failed and there was no option but to take them to the operating room. So the decision was not taken lightly. Dr T. Aberg (Umea, Sweden): And that refers to my first question, that the definition for including the patients here are more than those given in the paper, because otherwise, I mean, with the definition given here, I don't think we would accept that high a mortality. So I think that there is some matter of the definition of what you call emergency. Mr S. Nashef: Yes. As Dr. Ooi alluded to, these are patients who either have a major structural defect that is causing them to go into severe hemodynamic dysfunction which cannot be treated by conservative means or, in the case of coronary surgery, they are patients who are clearly beginning to evolve infarcts and in whom the balloon has not successfully stabilized the situation. They are sick patients.
Presented at the 16th Annual Meeting of the European Association for Cardio-thoracic Surgery, Monte Carlo, Monaco, September 22–25, 2002. doi:10.1016/S1569-9293(03)00041-0
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
Top
Abstract
1. Introduction