topeeng.gif (8416 bytes)

[ Scientific Activity - Actividad Científica ] [ Brief Communications - Temas Libres ]

Quantifying Risk in Cardiovascular Surgery

Bonaccorsi Héctor; Sosa Pamela; Sgrosso José; Ameriso José; Dogliotti Ariel.

Instituto Cardiovascular de Rosario
Rosario. Argentina

Introduction
Objectives
Patients and Methods
Results
Discussion
Conclusion
References

Introduction

In present days world demand in quality and efficiency, both in the manufacture of the products and services, grows day by day. Medicine is not an exception. Although this is stronger in developed countries, it seems to be widening all over the world. The forementioned it is specially true with cardiovascular surgery since this is an expensive and very frequent procedure.

However, since long time ago, there is agreement in thinking that mortality because of this kind of procedures depends on not only the quality of care, but also the stage of the patient´s previous illness and hazard [1,2]. The subsequent inference that worse population means worse results, led concerned people to design systems with which taking into account the patient´s preoperative characteristics one could anticipate the surgical outcome and in that way make the quality control more accurate.

Therefore cardiovascular surgery results are required by the patients themselves, by the doctors and by the institutions that render the service and by the ones which pay for it.

To summarize, the risk stratification in cardiovascular surgery is useful to make clinical decisions, assessing quality of care and the administrative management related to the procedure.

Objectives

To quantify the surgical risk in patients who were sent to cardiovascular surgery using a mathematical model developed in our country based on a multicentric study.

Patients and Methods

49 consecutive patients who were sent to our institution for coronary artery bypass graft surgery, were taken into consideration in a prospective manner.

Each patient´s preoperative variables were entered through a personal computer in a software named "Score". This program allows the calculation of the hospital mortality risk using the risk prediction formula that came out the mathematical model developed in our country through the multicentric study CONAREC III. This study was done in 1992 and 1993, over 1293 patients from 41 medical institutions from all over the country[3]. Therefore its expected results can be considered a national average.

After obtain the death probability according to its preoperative features, the outcome of each patient was evaluated. Postoperative major complications taking into account were myocardial infarction, severe ventricular arrhythmia, bradyarrhythmia, mediastinal bleeding, reoperation for bleeding, reoperation for other causes, acute renal failure, dialysis need, cardiogenic pulmonary edema, non-cardiogenic pulmonary edema, pneumothorax, prolonged mechanical ventilation, low cardiac output syndrome, intra-aortic balloon pump need, encephalopathy, stroke, coma, sepsis, mediastinitis, atrial fibrillation and in-hospital death. It was also considered the long of stay in Cardiac Surgical Intensive Care Unit and the total long of stay in hospital.

According to the original report of CONAREC III the expected mortality was stratified in five categories: low risk < 4%, moderate risk 4-8%, high risk 8-12%, very high risk 12-30%, too high risk > 30%.

Results

Preoperative risk stratification:

The following risk stratification was found, according to expected mortality: patients with low risk were 14%(7/49) of the studied sample, those with moderate risk were 22%(11/49) and those with high, very high and too high risk were 63%(31/49).

Death risk:

The global hospital mortality shown in the sample was 8,2%(4/49) (confidence interval of 95%: 0,5%-15,8%). 75%(3/4) of the dead patients were in too high risk category. The average expected mortality for the same group of patients was 18,4%.

Risk of complications:

57%(28/49) of the patients had complications and/or died. The average rate of complications per patient who survived was of 1,5(69/45).

The average rate per patient who survived according to the risk category was: low risk 0,7(5/7), moderate risk 1(11/11), high risk 1,1 (11/10), very high risk 1,6(16/10), too high risk 3,6(25/7).

Risk of prolonged long of stay:

53% of the patients were discharged from Cardiac Surgical Intensive Care Unit within 48 hours since their admission. 86% of the low risk patients and 91% of the moderate risk ones were discharged from this unit within 48 hours, while 40% of the very high risk patients and 50% of the too high risk ones remained in that unit more than 72 hs.

Discussion

Since long time ago it´s clear for the medical community that unwanted outcome in cardiovascular surgery depends on, not only, the quality of care, but also the stage of the cardiac illness, the associated illness and hazard [1,2]. Therefore it is very important to estimate the surgical risk of the patients who were sent to cardiovascular surgery, according to their preoperative characteristics. This estimation can be done in different ways. Without doubt the oldest one consists in just using the doctor´s intuition or reasoning. Unfortunately as there are too many variables to take into account, this method is less efficient than the mathematical models [1].

The mathematical models or clinical predictions rules are, in general, equations found through statistical techniques after having analyzed a sample of people. Each variable with predictable capacity found by the model is given a weight if it´s found in patient. As a result the probability, that the unfavourable event occurs, is obtained.

There has been made and published a great number of surgical risk models [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15].

Risk of death

The use of mortality in the prediction models has dome advantages lide that is is a concrete fact easy to identify that requires no further definition and ist is in mostly all medical records or even administrative ones. The cons ar that is is not directly related to the costs and in many cases the facts which predict it are very different from the ones that predict costs, morbility or long of stay [1].

The first relevant data given by this study was the high percentage of patients from the sample who were in the high risk group (high risk, very high risk and too high risk), that is 63% of the patient population. This could be because our institution as a reference center, receives very ill patients from smaller institutions from an important area.

The global hospital mortality of the sample was 8,2% while the average expected mortality obtained through the mathematical model was 18,4%. The last one is out of the confidence interval of 95% of the first, so it can be considered statistically different. There are two reasons that can explain this. First, the features of the patient population of our institution and those taken into account by the CONAREC III trial were very different. Unfortunately, the comparative analysis was extremely difficult because lack of data in CONAREC III original report. However, since the observed mortality in our study is less than the expected according to the model, if the populations were different, the patients from our institution shuld be less ill than the ones included in the mentioned trial.

Unfortunately, in the original report of the CONAREC III study doesn't appear the percentage of risk stratification so as to make comparisons. However, in absolute values the amount of 63% of patients in the high risk group in our population is hihg enough to be surpassed by a study that can be considered as a national average. There is also the epidemiologic data that a mixture of patients (some came from centers like ours that receive very ill patients and others came from smaller centers wich have only less ill patients) was taken in consideration for the mulicentric study. That is probably why bias direction is opposite and the amount of more seriously ill patients is he one evaluated in the present study.

The second reason why the observed mortality in the present sample was minor than the one expected by the mathematical model, could be the better quality of care given.

Risk of complications

The factors which estimate complications are frequently related to the predictors of long of stay, quality of life and costs. The disadvantages of using morbidity in the prediction models are that the the definitions of complications change from one center to another and that they are more difficult to obtain [1,2,3,4].

With reference to complications in post surgical patients we can say that although the mathematical model used was designed to predict only mortality risk, it appeared a trend for patients to present a higher average of more serious complications as the mortality risk goes up.

Risk of prolonged long of stay:

The estimation or prediction of long of stay is a way to verify the use or resources. Its relation to the costs is obvious and it was used to infer the last ones[1].

With reference to the risk of prolonged long of stay in Cardiac Surgical Intensive Care Unit something similar happened to what it was observed with complications. A trend appeared in which patients classified in mortality risk categories had a longer long of stay in that unit as greater were the death risk calculated according to the preoperative variables.

Conclusion

To use a mathematical model to quantify the surgical risk was useful to detect a high percentage of patients in the high risk group in the sample studied. It allowed comparison between mortality in our institution and the expected results of a model that can be considered a "national average". Some relation was found between the quantification of mortality risk and the probability of complications and long of stay in Cardiac Surgical Intensive Care Unit.

Tope

References

[1] Jones RH. In search of the optimal surgical mortality. Circulation 1989;79(suppl I):I-132-I-136.
[2] Jencks SF, Daley J, Draper D y col. Interpreting hospital mortality data. JAMA 1988;260:3611-16.
[3] Ciruzzi M, Aranda G, Bozovich G y col. CONAREC III. Predicción del riesgo de mortalidad intrahospitalaria en pacientes sometidos a cirugía coronaria. Rev Argent Cardiol 1996;64:79-90.
[4] Lee KL, Pryor DB, Harrell FE y col. Predicting outcome in coronary disease. Statistical models versus expert clinicians. Am J Med 1986;80:553-60.
[5] O’Connor G T, Plume S K, Olmstead E M y col. Multivariate prediction of in-hospital mortality associated with coronary artery bypass graft surgery. Circulation 1992;85:2110-18.
[6] Tu J V, Jaglal S B, Naylor C D y col. Multicenter validation of a risk index for mortalty, intensive care unit stay, and overall hospital lengyh of stay after cardiac surgery. Circulation 1995;91:677-84.
[7] Geraci J M, Rosen A K, Ash A S y col. Predicting the occurrence of adverse events after coronary artery bypass surgery. Ann Intern Med 1993;118:18-24.
[8] Lahey S J, Borlase B C, Lavin P T y col. Preoperative risk factors that predict hospital length of stay in coronary artery bypass patients >60 years old. Circulation 1992;86(S II):II-181-II-185.
[9] Tuman K J, McCarthy R J, March R J y col. Morbidity and duration of ICU stay after cardiac surgery. Chest 1992;102:36-44.
[10] Ferraris V A, Ferraris S P, Singh A. operative outcome and hospital cost. J Thorac Cardiovasc Surg 1998;593-603.
[11] Smith P K, Smith L R, Muhlbaier L H. Risk stratification for adverse economic outcomes in cardiac surgery. Ann Thorac Surg 1997;64:S 61-S 63.
[12] Smith L R, Milano C A, Molter B S y col. Properative determinants of postoperative costs associated with coronary artery bypass graft surgery. Circulation 1994;90:II-124-II-128.
[13] Kennedy JW, Kaiser GC, Fisher LD y col. Multivariate discriminat analysis of the clinical and angiographic predictors of operative mortality from the Collaborative Study in Coronary Artery Surgery (CASS). J Thorac Cardiovasc Surg 1980;80:876-87.
[14] Hannan EL, Kilburn H, O'Donnell JF y col. Adult open heart surgery in New York State. JAMA 1990;264:2768-74.
[15] Parsonnet V, Dean D, Bernstein A. A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 1989;79(S I):I-3-I-12.
[16] Shroyer AL, Plomondon ME, Grover FL y col. The 1996 Coronary Artery Bypass Risk Model: The Society of Thoracic Surgeons Adult Cardiac National Database. Ann Thorac Surg 1999;67:1205-8.
[17] Grover FL, Hammermeister KE, Burchfiel Cy col. Initial report of the Veterans Administration Preoperative Risk ssessment Study for Cardiac Surgery. Ann Thorac Surg 1990;50:12-28.
[18] Higgins TL, Estafanous FG, Loop FD y col. Stratification of morbidity and mortality ourcome by preoperative risk factors in coronary artery bypass patients. JAMA 1992;267:2344-48.
[19] Higgins TL, Estafanous FG, Loop FD y col. ICU admission score for predicting morbidity and mortality risk after coronary artery bypass grafting. Ann Thorac Surg 1997;64:1050-8.
[20] Higgins TL. Quantifying risk and assessing outcome in cardiac surgery. J Cardiothor Vasc Anesth 1998;12:330-40.
[21] Geraci J M, Rosen A K, Ash A S y col. Predicting the occurrence of adverse events after coronary artery bypass surgery. Ann Intern Med 1993;118:18-24.
[22] Hammermeister K E, Burchfiel C, Johnson R y col. Identification of patients at greatest risk for developing major complications at cardiac surgery. Circulation 1990;82(S IV):IV-380-IV-389.
[23] Higgins TL, Estafanous FG, Loop FD y col. Stratification of morbidity and mortality ourcome by preoperative risk factors in coronary artery bypass patients. JAMA 1992;267:2344-48.
[24] Tuman K J, McCarthy R J, March R J y col. Morbidity and duration of ICU stay after cardiac surgery. Chest 1992;102:36-44.
[25] Tu J V, Jaglal S B, Naylor C D y col. Multicenter validation of a risk index for mortalty, intensive care unit stay, and overall hospital lengyh of stay after cardiac surgery. Circulation 1995;91:677-84.

 

Questions, contributions and commentaries to the Authors: send an e-mail message (up to 15 lines, without attachments) to surgery-pcvc@pcvc.sminter.com.ar , written either in English, Spanish, or Portuguese.

Top


© CETIFAC
Bioengineering
UNER

Update
Oct/30/1999