Asian Cardiovasc Thorac Ann 2005;13:345-350
© 2005 Asia Publishing EXchange Ltd
Effect of Prolonged Intensive Care Stay on Survival Following Coronary Surgery
Sanjay V Ghotkar, FRCS,
Antony D Grayson, BSc1,
Walid C Dihmis, FRCS
Department of Cardiothoracic Surgery
1 Department of Research and Development, The Cardiothoracic Centre, Liverpool NHS Trust, Liverpool, United Kingdom
For reprint information contact: Walid C Dihmis, FRCS Tel: 44 151 293 2309 Fax: 44 151 220 8573 Email: Walid.Dihmis{at}ctc.nhs.uk, The Cardiothoracic Center, Liverpool NHS Trust, Thomas Drive, Liverpool L14 3PE, United Kingdom.
 |
ABSTRACT
|
|---|
The aim of the study was to examine midterm survival in patients who required prolonged recovery in the intensive care unit. The 5,186 consecutive patients who underwent isolated coronary surgery between April 1997 and March 2002 were retrospectively analyzed. Patients were classified as having prolonged ( > 3 days) or normal ( 
3 days) stay in the intensive care unit. Patient records were matched to the National Health Service Strategic Tracing Service which records all-cause mortality in the UK. Case-mix was controlled for by constructing a propensity score from core patient characteristics, which was included along with the comparison variable in a multivariable analysis of outcome. Prolonged intensive care unit stay was recorded in 475 (9.16%) patients. Mortality was 9.14% during the study period with a total follow-up of 19,618 patient-years (mean, 3.8 years). Adjusted 5-year survival was 78.0% for prolonged intensive care unit stay vs. 90.7% for normal stay, with an adjusted hazard ratio for midterm mortality of 2.6 ( p < 0.001). Midterm mortality was significantly higher in patients with a prolonged intensive care unit stay following coronary bypass.
|
INTRODUCTION
|
|---|
Coronary artery bypass graft (CABG) surgery is one of the most commonly performed operations in the developed world. The majority of patients undergoing this operation are low risk, with a high probability of successful outcomes including significant improvement in symptomatic status and life expectancy.1 With better results being achieved, the surgical indications for this operation have become increasingly broader. At the same time, overall improvements in general healthcare and recent progress in interventional cardiology have altered the profile of patients referred for CABG. Thus, the proportion of high-risk patients referred for CABG has increased, with a greater number of the very elderly, more with severe comorbid conditions, and an increase in those requiring re-operation.2 In most instances, these high-risk patients have a rapid postoperative recovery and are discharged from hospital, although some require extended postoperative treatment in the intensive care unit (ICU). However, their survival in the community has not been adequately evaluated. We aimed to evaluate the midterm survival of patients who required prolonged ICU stay compared to those with a normal length of stay, after adjusting for differences in patient and disease characteristics.
|
PATIENTS AND METHODS
|
|---|
The 5,186 consecutive patients who underwent isolated CABG between 1st April 1997 and 31st March 2002 at The Cardiothoracic Center, Liverpool, UK, were retrospectively analyzed. Those undergoing CABG incidental to heart valve repair or replacement, resection of a ventricular aneurysm, or other similar surgical procedures were not included, to exclude any bias that may arise due to combined surgical procedures. Data on the following variables were routinely collected prospectively: age, sex, body mass index (BMI), urgency of operation, prior cardiac surgery, New York Heart Association functional class, Canadian Cardiovascular Society angina class, the extent of coronary disease, and left ventricular ejection fraction. History of myocardial infarction, smoking, diabetes, hypercholesterolemia, hypertension, peripheral vascular disease, cerebrovascular disease, respiratory disease, and renal dysfunction were also noted. Procedural data collected included use of cardiopulmonary bypass (CPB), duration of CPB and aortic crossclamp time, as well as type and number of grafts. Definitions and data collection methods have been previously published.3
Criteria for discharge from the ICU included cardiovascular stability, minimal or no respiratory assistance, evidence of adequate renal function with normal serum electrolyte levels, and evidence of adequate neuropsychological function. Days spent in the ICU were counted by patient census at midnight each day. Patients who stayed in the ICU for more than 3 consecutive days on the initial admission were classified as having a prolonged ICU stay, while patients staying 3 days or less were classified as having a normal ICU stay. Hospital morbidity was also collected during the patient’s admission and included myocardial infarction, stroke, acute renal failure, deep sternal wound infection, re-exploration for bleeding, and duration of ventilation. Postoperative myocardial infarction was defined as new Q-waves postoperatively in 2 or more contiguous leads on an electrocardiogram, or a significant rise in postoperative creatine kinase-MB level with hemodynamic and echocardiographic signs of myocardial infarction. Postoperative stroke was defined as a new focal neurological deficit and/or comatose state occurring postoperatively that persisted for more than 24 hours after its onset. We excluded confused states, transient cerebral events, and intellectual impairment from our study to avoid any subjective bias. Acute renal failure was defined as the need for postoperative dialysis. Criteria for diagnosing deep sternal wound infections were in accordance with the guidelines published by the Centers for Disease Control and Prevention.4 Postoperative bleeding was defined as bleeding that required surgical re-exploration after initial departure from the operating theatre. Patient records were linked to the National Health Service Strategic Tracing Service (NSTS) which records all deaths in the UK. To establish current vital status for the patient population through 30th September 2003, patients were matched to the NSTS database by name, National Health Service number, date of birth, sex, and postcode.
Continuous variables are shown as median with 25th and 75th percentiles, and categorical variables are shown as a percentage. Comparisons were made with the Wilcoxon rank sum test or chi-squared test, as appropriate. Deaths occurring as a function of time were described actuarially using the product-limit methodology of Kaplan and Meier.5 To control for differences in patient characteristics, we developed a propensity score that was entered into a Cox proportional hazards analysis to calculate the adjusted hazard ratio (HR) and to risk-adjust the Kaplan-Meier survival curves.6–7 The propensity score is the probability that a patient will have a prolonged ICU stay, and it is constructed from offering all the preoperative variables listed in Table 1
(C statistic = 0.71). Once the propensity score is constructed for each patient, there are 3 ways of using the score for comparisons: matching, stratification, and multivariable adjustment. We used multivariable adjustment because matching would have reduced the study size, and stratification can be difficult to interpret. The propensity score is then included along with the comparison variable (prolonged vs. normal ICU stay) in the Cox proportional hazards analysis. The propensity score adjusts for the treatment selection bias, which is evident in Table 1, between the two groups.6 In all cases, a p value < 0.05 was considered significant. All statistical analyses were performed retrospectively with SAS for Windows version 8.2 (SAS Institute, Cary, NC, USA).
|
RESULTS
|
|---|
Overall, 475 (9.2%) patients were classified as having a prolonged ICU stay, while 4,711 (90.8%) were classified as having a normal ICU stay. Table 1
lists preoperative characteristics based on length of ICU stay. Operative characteristics are shown in Table 2. In the patients who underwent CABG on cardiopulmonary bypass, the mean cardiopulmonary bypass time was 118.5 min (25th and 75th percentiles: 97 and 145 min) for those with a prolonged ICU stay, compared to 106 min (25th and 75th percentiles: 89 to 128 min) for those with a normal ICU stay ( p < 0.001). The mean aortic crossclamp times were 66 min (25th and 75th percentiles: 51 to 85 min) and 60 min (25th and 75th percentiles: 46 to 76 min), respectively, ( p < 0.001). The mean length of stay in the ICU for patients with a prolonged stay was 6 days (25th and 75th percentiles: 4 to 8 days), with the longest stay being 117 days. Hospital mortality and morbidity outcomes depending on the length of ICU stay are shown in Table 3.
There were 474 (9.1%) deaths among the 5,186 patients during the study period, with a mean (± standard deviation) follow-up of 3.8 ± 1.6 years. Total patient follow-up was 19,618 years. The crude HR of midterm mortality for prolonged ICU stay was 3.6 ( p < 0.001). Freedom from death in patients with prolonged ICU stay at 30 days, 1, 2, 3, 4, and 5 years was 91.8%, 84.6%, 81.2%, 78.1%, 75.6%, and 72.5%, respectively, compared with 98.4%, 96.7%, 95.2%, 93.7%, 92.2%. and 90.6% for those with normal ICU stay (Figure 1).
After adjustment for baseline differences in patient and disease characteristics with the propensity score (shown in Table 4), the adjusted HR of midterm mortality for prolonged ICU stay was still significantly different at 2.6 ( p < 0.001). The adjusted Kaplan-Meier survival curves are shown in Figure 2. The adjusted freedom from death in patients with a prolonged ICU stay at 30 days, 1, 2, 3, 4, and 5 years was 95.7%, 91.2%, 88.0%, 84.7%, 81.5% and 78.0%, respectively, compared with 98.3%, 96.4%, 95.1%, 93.7%, 92.3% and 90.7% for those with normal ICU stay. Due to concerns that the significant survival disadvantage was due solely to the increased hospital mortality for patients with a prolonged ICU stay, we repeated our risk-adjusted analyses on all patients who were discharged alive from their initial hospitalization. The adjusted HR of midterm mortality for prolonged ICU stay was still significantly different at 1.9 ( p < 0.001). The adjusted freedom from death in patients with a prolonged ICU stay at 5 years was 84.9% compared with 91.9% for those with a normal ICU stay.
View larger version (10K):
[in this window]
[in a new window]
|
Figure 2. Adjusted midterm survival following coronary artery bypass surgery, adjusted for the propensity score.
|
|
|
DISCUSSION
|
|---|
Despite continuous improvements in surgical techniques for CABG and better outcomes for patients, some complications are inevitable.1–2 Several patient characteristics have been associated with prolonged lengths of ICU stay, due to the increased risk of postoperative morbidity.8 With an increasing number of high-risk cases undergoing CABG, it is not surprising that the proportion of patients having a prolonged ICU stay has also increased.2 Patients who require prolonged ICU stay are known to have higher hospital mortality and increased hospital costs. In this study, we observed that patients with a prolonged stay have a hospital mortality of 10.7% compared to 1.4% for other patients ( p < 0.001). An examination of costs at our institution imply that, on average, patients with a prolonged ICU stay cost 7,428 GBP more than other CABG patients.
Prolonged stays in the ICU not only carry a higher hospital mortality for the patient and increased costs for the hospital, but we have demonstrated in this study that survivors of prolonged ICU stay also carry a higher mortality in the subsequent years after CABG. The risk-adjusted mortality rate at 5 years was 15.1% for patients with a prolonged ICU stay compared to 8.1% for other patients ( p < 0.001). This highlights the importance of identifying those at greater risk of requiring a prolonged stay in ICU, prior to undergoing CABG surgery, with the potential for risk-factor modification when possible. To risk-adjust our survival curves, we developed a propensity score which was the probability of requiring a prolonged ICU stay. The variables identified as predictors of a prolonged ICU stay (Table 4) give us some ideas on how we can attempt to reduce the likelihood of a poor outcome for these patients.
Although most of the identified risk factors remain unamenable to change, the first obvious solution would be to tackle the patient’s obesity and smoking habits prior to CABG. However, experience has shown that despite advice on the need to lose weight or quit smoking, it comes down to patient choice, and most are unable or reluctant to change their habits. The treatment of diabetes could potentially have an impact on reducing the chances of requiring a prolonged ICU stay. Furnary and colleagues9 demonstrated that tight control of blood glucose levels with intravenous insulin infusion throughout the perioperative period can reduce the incidence of wound complications. One of the reasons for patients staying in the ICU for more than 3 days is the development of sternal wound infections (Table 3), and this complication has been linked to a poor midterm prognosis.10 Therefore, tighter control of diabetes might reduce the chances of such patients requiring a prolonged ICU stay and improve their midterm survival. McAlister and colleagues11 argued that tighter control of blood glucose levels would help improve outcomes following CABG, because of the adverse outcomes associated with hyperglycemia on the first postoperative day.
At the base of Table 4 is the prediction equation used to calculate the propensity score which was used to risk-adjust the survival curves. This works in exactly the same way as the logistic EuroSCORE, except that it predicts the chances of a prolonged ICU stay and not the probability of hospital mortality. With modern technology, this prediction tool can be easily programmed into appropriate software resident on desktops and hand-held computers. Such a tool may prove useful in assessing the risks to the patient prior to CABG, and may even help plan surgical activity, with patients who have a high probability of prolonged ICU stay being evenly spread out on the operating list during a given week’s surgery. The tool also provides a useful benchmark to aid in assessing improvements made in reducing the need for prolonged ICU stay, although we recommend modifying the tool according to an individual institution’s performance, as discussed by Wynne-Jones and colleagues3 (visit: www.nwheartaudit.nhs.uk for examples of this when monitoring hospital mortality).
This study has highlighted the potential role of off-pump CABG surgery and total arterial revascularization in reducing length of ICU stay. It is important to note that the off-pump patients may be influenced by an element of selection bias; however, previous work from our institution has shown in a propensity-matched cohort that off-pump total arterial revascularization can significantly lower ICU and hospital lengths of stay compared to conventional on-pump surgery.12 Bashour and colleagues13 found that in patients who stayed for a long time in the ICU and survived to discharge, the mortality during a mean follow-up of 30.6 months was almost 50%. They also described how nearly 50% of all costs for the 2,618 patients in their study were accounted for by the patients who stayed longer in the ICU. It is important to note that Bashour classified anyone with a stay in ICU of 10 or more days as prolonged. Definitions of prolonged ICU stay are not clearly defined in the literature. The definition has varied from > 48 hours to > 14 days.14–16 The cut-off of 3 days was chosen for our study because of a recognized national measure used for quality of ICU care, and also because of contractual arrangements (for every CABG performed, our institution receives payment for 3 days’ ICU care).
There are some limitations that need to be considered for this study. Firstly, patients who have a prolonged ICU stay are naturally sicker patients, and thus are more likely to have a higher mortality rate. However, by excluding hospital mortality and risk-adjusting the survival curves through propensity score methods, we have attempted to reduce such bias. A further limitation is that we only had all-cause mortality available and so we are unclear about the mode of death in both groups during the follow-up period. Although it must be said that there is no reason to believe that noncardiac-related deaths should occur more frequently in one group compared to another. Finally, mortality is not always the best method for judging follow-up prognosis of patients after CABG. Unfortunately, we do not have information regarding such data as quality of life or functional capacity.
It was concluded that patients who require a prolonged ICU stay have increased midterm mortality. Methods of reducing the incidence of prolonged ICU stay should be strongly encouraged. Furthermore, we have provided within this paper a prediction equation that might prove useful in assessing the probability of a patient requiring prolonged ICU stay.
|
ACKNOWLEDGMENTS
|
|---|
We would like to acknowledge the co-operation given to us by all the Consultant Cardiac Surgeons at the Cardiothoracic Centre-Liverpool: Mr JAC Chalmers, Mr BM Fabri, Miss EM Griffiths, Mr N Mediratta, Mr RD Page, Mr DM Pullan, Mr A Rashid, and Mr WI Weir. We would like to thank Janet Deane, who maintains the quality and ensures the completeness of data collected in our Cardiac Surgery Registry.
Presented at the 12th Annual Meeting of the Asian Society for Cardiovascular Surgery, Istanbul, Turkey, April 19 22, 2004.
|
REFERENCES
|
|---|
- Sundt TM III, Gersh BJ, Smith HC. Indications for Coronary Revascularization. In: Cohn LH, Edmunds LH Jr, editors. Cardiac Surgery in the Adult. New York: McGraw-Hill, 2003:541–59.
- Ferguson TB Jr, Hammill BG, Peterson ED, DeLong ER, Grover FL; STS National Database Committee. A decade of change – risk profiles and outcomes for isolated coronary artery bypass grafting procedures, 1990–1999: a report from the STS National Database Committee and the Duke Clinical Research Institute. Society of Thoracic Surgeons. Ann Thorac Surg 2002;73:480–90.[Abstract/Free Full Text]
- Wynne-Jones K, Jackson M, Grotte G, Bridgewater B. Limitations of the Parsonnet score for measuring risk stratified mortality in the north west of England. The North West Regional Cardiac Surgery Audit Steering Group. Heart 2000;84:71–8.[Abstract/Free Full Text]
- Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 1999;20:250–78.[Medline]
- Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:547–81.
- Blackstone EH. Comparing apples and oranges. J Thorac Cardiovasc Surg 2002;123:8–15.[Free Full Text]
- Cox DR. Regression models and life-tables. J R Stat Soc 1972;34:187–220.
- Bucerius J, Gummert JF, Walther T, Doll N, Falk V, Schmitt DV, Mohr FW. Predictors of prolonged ICU stay after on-pump versus off-pump coronary artery bypass grafting. Intensive Care Med 2004;30:88–95.[Medline]
- Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 1999;67:352–62.[Abstract/Free Full Text]
- Lu JC, Grayson AD, Jha P, Srinivasan AK, Fabri BM. Risk factors for sternal wound infection and mid-term survival following coronary artery bypass surgery. Eur J Cardiothorac Surg 2003;23:943–9.[Abstract/Free Full Text]
- McAlister FA, Man J, Bistritz L, Amad H, Tandon P. Diabetes and coronary artery bypass surgery: an examination of perioperative glycemic control and outcomes. Diabetes Care 2003;26:1518–24.[Abstract/Free Full Text]
- Pandey R, Grayson AD, Pullan DM, Fabri BM, Dihmis WC. Total arterial revascularisation: effect of avoiding cardiopulmonary bypass on in-hospital mortality and morbidity in a propensity-matched cohort. Eur J Cardiothorac Surg 2005;27:94–8.[Abstract/Free Full Text]
- Bashour CA, Yared JP, Ryan TA, Rady MY, Mascha E, Leventhal MJ, et al. Long-term survival and functional capacity in cardiac surgery patients after prolonged intensive care. Crit Care Med 2000;28:3847–53.[Medline]
- Treasure T, Holmes L, Loughead K, Gallivan S. Survival and quality of life in patients with protracted recovery from cardiac surgery. Can we predict poor outcome? Eur J Cardiothorac Surg 1995;9:426–32.[Abstract]
- Wahl GW, Swinburne AJ, Fedullo AJ, Lee DK, Bixby K. Long-term outcome when major complications follow coronary artery bypass graft surgery. Recovery after complicated coronary artery bypass graft surgery. Chest 1996;110:1394–8.[Abstract/Free Full Text]
- Williams MR, Wellner RB, Hartnett EA, Thornton B, Kavarana MN, Mahapatra R, et al. Long-term survival and quality of life in cardiac surgical patients with prolonged intensive care unit length of stay. Ann Thorac Surg 2002;73:1472–8.[Abstract/Free Full Text]
TOP
ABSTRACT
INTRODUCTION
