Quality Assurance in Health Core, Vol. 4. No. 3, pp. 217-224, 1992 Printed in Great Britain
1040-6166/92 S5.00 + 0.00 © 1992 Pergamon Prejj Ltd
PREOPERATIVE EVALUATION IN NONCARDIAC SURGERY: CARDIAC RISK ASSESSMENT Cristina Maggio, Alessandro Bonzano, Enrica Conte, Daniela Libertucci, Maurizio Panarelli, Marco Bobbio and Plinio Pinna Pintor Arturo Pinna Pintor Foundation Torino Italy (First submitted 14 June 1991; accepted after revision 18 December 1991)
Nine hundred and ninety patients, ages 20 years or older, undergoing non-cardiac elective surgery were prospectively studied to identify high cardiac risk preoperative factors in a case-mix population and to assess cardiological risk. The prevalence of major cardiac complications was2.3%, includingO.8% mortality. Untvariate analysis showed that: age; history of chest pain; dyspnea; hypertension; presence of systolic murmur and third sound; diastolic pressure >95 mmHg; electrocardiogram left ventricular hypertrophy; cardiothoracic ratio >0.5 and valvular calcifications are associated with cardiac complications (p = 0.001-0.02), with low sensitivity (range: 14-38%) and high specificity (range: 85-98%). Cardiological referral was required for 169 patients (17%) that showed a higher prevalence of cardiovascular diseases (85%) and of cardiac complications (5.3%). Cardiologists required further tests for 13 patients (7.7%) and modified therapy for 93 (55%). High cardiac risk patients are identified preoperatively in current practice and cardiological referral is frequent; further studies are mandatory to evaluate the most effective and efficacious procedures. Key words: Cardiac risk, preoperative evaluation.
INTRODUCTION Preoperative evaluation of general physical status in patients scheduled for surgery is current practice and focuses mainly on cardiovascular function, since this is one of the most critical factors affecting perioperative complications [1-10]. The best approach for identifying high cardiac risk in a non-selected surgical population has been discussed previously [1,6,11-17]. New insights into the value of tests in different populations and budgetary constraints have compelled health care providers to critically review preoperative assessment practice [17-23]. Moreover, Requests for reprints should be sent to: Dr Cristina Maggio, Fondazione Arturo Pinna Pintor, Via A. Vespucci, 61, 10129 Torino, Italy. 217
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
218
C. Maggio et al.
information is lacking about the practice of and the real motivations behind common preoperative assessment in different geographical areas and institutions. The aims of this prospective observational study were: to identify high cardiac risk factors in non-selected, non-cardiovascular elective surgery patients; to gain insights into the behavior of surgical staff in our common practice; to evaluate the referral rate to cardiologist consultants and to compare cardiac risk of referred and nonreferred patients in our experience. MATERIALS AND METHODS
Between May 1987 and August 1989 patients aged 20 years or older scheduled for non-cardiovascular elective surgery were recruited in three institutions (two public hospitals and one private hospital). Patients undergoing emergency or vascular surgery, or local anesthesia were excluded. A preoperative evaluation, based on history and physical examination, was carried out on each patient by five trained investigators, while laboratory, electrocardiogram (ECG) and chest X-ray information was recorded when available on surgical staff request, according to their commonly used protocols. Sixty variables were recorded for each patient. ECG was requested for 982 patients (99%) and chest X-ray for 787 (79%). Data regarding patients referred for cardiological consultation were also recorded, i.e. diagnosis and effects of consultation (request for new test and/or modification of therapy). The investigators followed each patient on a daily base until discharge in order to collect information about the end points of the study: (1) death; (2) cardiac death (sudden death, death due to refractory low output cardiac failure not associated with sepsis or acute respiratory insufficiency); (3) major cardiac complications: pulmonary cardiac edema (dyspnea, bilateral pulmonary rales—Killip m/IV—with compatible chest X-ray), angina, non-fatal myocardial infarction (new pathological Q wave on ECG and/or angina associated with ST-T depression >1 mm and/or T inversion lasting at least 72 hr not due to drugs or electrolyte imbalance and/or increase in aspartate amino transferase and CK/CK-MB in serum, and congestive heart failure (dyspnea and/or edema associated with gallop rhythm, pulmonary rales, hepatomegaly, jugular distension and clinical benefit from diuretic therapy). Statistical Analysis
Chi-squared analysis with Yates correction when requested was used to compare the univariate association between each variable and the incidence of cardiac end points. The odd ratios and their confidence intervals were calculated in order to evaluate the relative importance of each variable. Sensitivity and specificity were calculated according to standard formulas (sensitivity: true positive/true positive + false negative; specificity: true negative/true negative + false positive) [24,25]. RESULTS Nine hundred and ninety patients aged from 20 to 93 years (mean age: 54 ± 15 years), 489 (49.3%) of which were males, underwent elective surgical procedures. The characteristics and type of the elective surgery are reported in Table 1.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
Cardiac Risk Assessment
219 TABLE 1.
Population characteristics 990
Total patients Mean age (years)
54± N
%
188 391 375 36 528/539
19.0 39.0 38.0 4.0
Cardiovascular diseases Hypertension Ischemic heart disease Valvular heart disease Cerebrovascular disease Peripheral vascular disease Others
169 39 27 14 24 6
16.2 4.0 3.0 1.4 2.4 0.6
Others Chronic obstructive lung disease Diabetes meQitus Hepatic insufficiency Chronic renal insufficiency
183 43 15 8
18.5 4.3 1.5 0.8
Types of surgery Abdominal Intra thoracic Superficial (including breast) Orthopedic Neck (including thyroid)
446 245 181 66 62
45.0 25.0 18.0
Age (years) 20-39 40-59 60-79 80 or more Sex(M/F)
TABLE 2.
6.0 5.0
Perioptrative complications N
%
Mortality Total Sudden death Cardiac death
16 4 8
1.6 0.4 0.8
Cardiac complications Non-fatal myocardial infarction Cardiac failure Cardiac pulmonary edema
1 11 4
0.1 1.0 0.4
One patient may have had more than one complication.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
220
C. Maggio et al. TABLE 3.
Risk factors associated with cardiac major events (unirariate analysis) Odd ratio (95% CI)
Prevalence (%)
P
Sensitivity Specificity (%) (%)
Age (years) >50 >60 >70
61.4 55.6 13.8
7.3 3.0 3.1
(1.7-31.3) (1.2-7 3) (1.2-7.9)
0.002 0.009 0.02
91 64 32
42 63 87
History Chest pain Dyspnea Hypertension
7.6 12.1 15.1
5.0 5.1 3.5
(1.9-13.5) (1.9-11.5) (1.4-8.7)
0.002 0.001 0.001
28 38 38
92 88 85
Physical examination Systolic murmur Third sound DiastoBc BP > 95 mmHg
15.4 1.5 7.2
3.5 10.5 5.4
(1.4-8.5) (2.7-39.5) (2.0-14.4)
0.01 0.001 0.001
38 14 28
85 98 93
4.0
6.0
(1.8-18.8)
0.004
19
96
11.0 5.0
4.3 6.6
(1.6-11.5) (2.3-19.0)
0.007 0.001
28 24
91 95
ECG (N = 982) Left ventricular hypertrophy Chest X-ray (N - 787) Cardiothoraric ratio >0.5 Valvular calcifications CI, confidence interval.
Sixteen patients died before discharge (total hospital mortality 1.6%). Cardiac mortality (8 patients) was 0.8% (of which half was due to sudden death). Major cardiac complications (cardiac death, cardiac failure, myocardial infarction and cardiac pulmonary edema) occurred in 2.3% (Table 2). In Table 3 are the reported prevalence, sensitivity and specificity of the variables significantly associated, by univariate analysis, to in-hospital major cardiac events, along with the odd ratio and confidence intervals for patients with such preoperative TABLE 4.
Preoperative cardiac referral
N
%
Diagnoses Non-cardiac disease Hypertension Ischemic heart disease Valvular disease Arrhythmias
26 66 35 23 19
15.4 39.0 20.7 13.6 11.2
Effects Further tests Therapy modifications
13 93
55.0
Total, 189 patients.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
7.7
Cardiac Risk Assessment
221
risk factors. None of the blood tests were associated with major cardiac events considered as end points of the study. Cardiological Referral Cardiological referral was required for 169 patients (17%). This population had a higher prevalence of cardiovascular diseases (85%) (Table 4). The cardiologist required further tests for 13 patients (7.7%) and modified therapy for 93 (55%). The referred population had a higher prevalence of major cardiac complications (5.3%; p = 0.017) with a relative risk of 3 (95% confidence interval 1.2-7.2) when compared with patients not referred preoperatively to the cardiologist.
DISCUSSION In our study, age is a relevant risk factor for cardiac complications, with the highest sensitivity for patients more than 50 years old. Other authors [11-15] have stressed the importance of age above 70 years as a risk factor for cardiac complications. In these studies, inclusion criteria limited the age population to subjects older than 40 years, while in our study patients aged from 20 to 39 years were included, and this may explain the different risk cut-off point that we observed. With the exception of age above 50 years, all otherriskfactors identified by current practice showed low sensitivity (range: 14-38%) and high specificity (range: 8598%). Since sensitivity represents the capacity of the test to minimize the number of patients considered healthy by the test but in fact at high cardiac risk, while specificity represents the capacity of the test to minimize the number of patients considered at high risk by the test, but in fact not at high cardiac risk, our results suggest that in a non-selected surgical elective population, at low absolute cardiac risk (2.3%), the procedure is effective in the identification of high cardiac risk, but is likely to be a failure for screening purposes, i.e. when dealing with an apparently healthy population. According to previous experience and the consensus in the literature [1-4,8,1117], in similar unselected surgical populations, preoperative symptoms and signs of heart failure or ischemic heart disease have been confirmed as being significant risk factors. In disagreement with other reports [11,14], a history of hypertension and the finding of preoperative diastolic hypertension are shown here to be significantly associated with high cardiac risk. If we consider that in our study electrocardiographic evidence of left ventricular hypertrophy was recognized as a risk factor as well, it can be suggested that the importance of hypertensive heart disease may have been underestimated and may deserve further studies. In our study, we could not confirm an association between the presence of Q-waves at baseline ECG and cardiac complications. Lack of a relationship may be explained by the impossibility of determining the age of the infarction from ECG; it has been demonstrated [7] that risk is exponentially reduced with time, with a critical reduction after 6 months. In fact, with the exception of one "silent" infarction, impossible to date, no patient in our study had suffered acute infarction previous to 6
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
222
C. Maggio et al.
months before the elective surgical procedure. On the other hand, Q-wave and nonQ-wave infarction carry the same risk [10]. In our study, we were also not able to confirm that arrhythmias (mainly atrial fibrillation) at preoperative ECG are associated with cardiac complications; the role of arrhythmias in surgical cardiac risk has been discussed previously [3,10-16]. Both the radiological signs we considered (cardiothoracic ratio >0.50 and valvular calcifications) were significant risk factors, according to the literature [11-16]: it is possible that both signs reflect advanced cardiopathy. In our series, preoperative ECG was required for 99% of patients, but the prevalence of abnormalities was low (9.4%) when compared with previous reports [26,27]. This may be explained by differences in the populations considered and by the fact that our study took into account only the few abnormalities that were expected to be relevant for cardiac risk. There is agreement in the literature [17,21,28] that preoperative ECG should be limited to older patients (aged more than 45, 50 and 60 years in different reports), both for risk evaluation purposes and for baseline reference. Nevertheless, in Italy, ECG is recommended in preoperative evaluation by the law regulating induced abortion, and this may account for its overuse in general surgery as well. The role of routine preoperative ECG, especially for baseline reference and complication monitoring and follow up, requires further study. In the literature [29,30], there is agreement that there should not be a generalized use of chest X-ray preoperatively, since the diagnosis of unexpected abnormalities is rare (1%). In our study, where more than half of the patients were older than 50 years, i.e. with a recommended preoperative indication [21], such a test was required for two-thirds of the patients. We found a high proportion of abnormalities (16%), considering that only two abnormalities, cardiomegaly and valvular calcifications, were registered. In a previous report that studied 3698 chest X-rays performed routinely in hospital on patients older than 40 years, a prevalence of 14% for cardiomegaly only is reported [31]. Although we cannot specify in our population the prevalence of unexpected abnormalities, we want to stress that both abnormalities we considered, cardiomegaly and the presence of valvular calcifications, were shown to be risk factors for cardiac complications. Cardiological Referral
To our knowledge, no data regarding the prevalence of cardiological referral in the preoperative evaluation of surgical patients is available, although a growing demand for such data from surgical staff is perceptible. Information about both the motivations (clinical and/or "defensive" medicine) and the purposes (cardiac risk stratification, diagnosis, possibility of improving treatment and/or monitoring in the perioperative period) are lacking as well. In this study, the group of patients referred to the cardiologist (17%) had a prevalence of cardiovascular diseases of 85% and a higher risk of cardiac complications (odd ratio = 3). In our opinion, this may mean that in common practice preoperative cardiological consultation is relatively frequent and, indirectly, that the surgical staff involves the cardiologist in the evaluation of high cardiac risk patients. Unfortunately, it has not been possible in this study to explain the motivations and the purposes of referral to the cardiologist, and which among the possible factors affecting decision-making throughout the preoperative
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
Cardiac Risk Assessment
223
evaluation was the most relevant for the surgical staff: this is the aim of an ongoing study. Ultimate exploitation of cardiological expertise in surgical patient care depends on a better understanding of the above two issues. When preoperative diagnostic and prognostic evaluation is the main purpose, it should be kept in mind that, according to Bayes' theorem, tests are more effective when applied to intermediate risk [32]. Because major cardiac events are so infrequent (2.3% in our sample) most of the prognostic information is contained in the a priori probability alone. Our observational results suggest that clinical examination is indeed useful in high cardiac risk identification in a non-selected surgical population, since it resulted in high cardiac risk patients being submitted to the cardiologist. The cardiologist, in most instances, did not require further tests for diagnostic purposes and was involved mainly in therapy modification. This observation is consistent with the hypothesis that the prevalent demand may for an interaction between surgical staff and cardiologist in order to optimize high risk patient management, as considered previously [9,33]. Unfortunately, there have been no studies, to our knowledge, considering the efficacy of intervention for or the effect of its omission on the prognosis of high risk cardiac patients. In conclusion, this study shows that, in a case-mix elective surgical population at a low absolute risk of mortality and cardiac complication, clinical preoperative evaluation, including ECG and chest X-ray for most patients, is unable to identify high cardiac risk patients. Further studies are mandatory to evaluate the procedures that are most efficacious and effective in high cardiac risk identification, in order mainly to improve sensitivity in the apparently healthy subjects. Cardiological intervention seems aimed mainly to optimize patient management. Specific studies are needed to evaluate whether such intervention modifies the prognosis of high cardiac risk surgical patients. Acknowledgement: We thank Paolo Maggio for his assistance in the elaboration of the results of this study.
REFERENCES 1. Mangano D T, Perioperative cardiac morbidity. Anesthesiology 72: 152,1991. 2. Cohen M M and Duncan P G, Physical status score and trends in anesthetic complications. J Clin Epidemiol 41: 83, 1988. 3. Salem D N, Chuttani K and Isner J M, Valutazione e trattamento di malattia cardiaca nel paziente chirurgico. In: Current problems in cardiology—Raccolta 1989, p. 133, Centro Scientifico Editore, Torino, 1990. 4. Rose S D, Corman L C and Mason D T, Cardiac risk factors in patients undergoing non cardiac surgery. Med Clin North Am 63: 1271, 1979. 5. Weitz H H and Goldman L, Non cardiac surgery in the patient with heart disease. Med Clin North Am 71: 413, 1987. 6. Editorial, Accounting for perioperative death. Lancet i: 1369, 1987. 7. Rao T L, Jakobs K H, El Etz et al., Reinfarction following anesthesia in patients with myocardial infarction. Anesthesiology 59: 499,1983. 8. Eagl K A and Boucher C A, Cardiac risk of non cardiac surgery. N EnglJ Med 321: 1330, 1989. 9. Wells P H and Kaplan J A, Optimal management of patients with ischemic heart disease for non cardiac surgery by complementary anesthesiologist and cardiologist interaction. Am Heart J 102: 1029, 1981.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
224
C. Maggio et al.
10. Shub C, Management of cardiac patient undergoing non cardiac surgery. Learning Center Highlights 6: 1,1991. 11. Goldman L, Caldera D, Nussbaum S R et al., Multifactorial index of cardiac risk in non cardiac surgical procedures. N EnglJMed 197: 845,1977. 12. Waters J, Wilkinson C, Golman M et al., Evaluation of cardiac risk in non cardiac surgical patients. Anesthesiology 55: A34,1981. 13. Detsky A S, Abrams H B, McLaughJin J R et al., Predicting cardiac complications in patients undergoing non cardiac surgery. J Gen Int Med 1:211,1986. 14. Zeldttn R A, Assessing cardiac risk in patients who undergo non cardiac surgical procedures. Can J Surg 27: 402, 1984. 15. Larsen S F, Olesen K H, Jakobsen Eetal., Prediction of cardiac risk in non cardiac surgery. Ew Heart / 8 : 179,1987. 16. Detsky A S, Abrahams H B, Forbath Setal., Cardiac assessment of patients undergoing non cardiac surgery. Multifactorial clinical risk index. Arch Intern Med 146: 2131, 1986. 17. Robbins J A and Mushlin A I, Preoperative evaluation of the healthy patient. Med Oin North Am 63: 1145,1979. 18. Editorial, Routine pre-operative investigations are expensive and unnecessary. Lancet H: 1466,1983. 19. Blery C, Chastang C and Gandy J H, Critical assessment of routine preoperative investigations. Effective Health Care 1: 111, 1983. 20. Turnbull J M and Buch C, The value of preoperative screening investigations in otherwise healthy individuals. Arch Intern Med 147: 1101,1987. 21. Haberer J P, Guelon D and Bichet G, Examen preoperatoir et Evaluation du risk op6ratoir, Encyd Mid Our (Paris, France) Anesthesie-Rianimation 36377: A05,1989. 22. Kaplan E B, Boeckmann A S, Roizen M F et al., Elimination of unnecessary pre-operative laboratory tests. Anesthesiology 57: A445,1982. 23. Blery C, Charpak Y, Szatan Metal, Evaluation of a protocol for selective ordering of pre-operative tests. Lancet I: 139,1986. 24. Bobbio M, Corso introduttivo ai metodi statistid in epidemiologia. G Ital Cordial 14: 206,1984. 25. Bobbio M, Corso introduttivo ai metodi statistic! in epidemiologia. G Ital Cordial 14: 941,1984. 26. Elston R A and Taylor D J E, The preoperative electrocardiogram. Lancet i: 349,1984. 27. Jakobson A and White T, Routine preoperative electrocardiograms. Lancet I: 972,1984. 28. Goldberger A L and O'Konski M, Utility of the routine electrocardiogram before surgery and on general hospital admission—critical review and new guidelines. Ann Int Med 105: 552, 1986. 29. National Study by the Royal College of Radiologists, Pre-operative chest radiology. Lancet ii: 83, 1979. 30. U.S. Department of Health and Human Services, The selection of patients for X-ray examinations. In: Presurgical chest X-ray screening examinations (April 1986). HHS Publications FDA 86-8261. 31. Sagel S S, Evans R G, Forrest JVetal., Efficacy of routine screening and lateral chest radiographs in a hospital based population. N EnglJMed 291: 1001,1074. 32. Bobbio M, Pollock B H, Cohen I and Diamond G A, Comparative accuracy of clinical tests for diagnosis and prognosis of coronary artery disease. Am J Cordial 62: 862,1988. 33. Downing J W and Tinker J H, Fitness for anesthesia: who decides? Lancet I: 387,1987.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
1040-6166/92 S5.00 + 0.00 © 1992 Pergamon Prejj Ltd
PREOPERATIVE EVALUATION IN NONCARDIAC SURGERY: CARDIAC RISK ASSESSMENT Cristina Maggio, Alessandro Bonzano, Enrica Conte, Daniela Libertucci, Maurizio Panarelli, Marco Bobbio and Plinio Pinna Pintor Arturo Pinna Pintor Foundation Torino Italy (First submitted 14 June 1991; accepted after revision 18 December 1991)
Nine hundred and ninety patients, ages 20 years or older, undergoing non-cardiac elective surgery were prospectively studied to identify high cardiac risk preoperative factors in a case-mix population and to assess cardiological risk. The prevalence of major cardiac complications was2.3%, includingO.8% mortality. Untvariate analysis showed that: age; history of chest pain; dyspnea; hypertension; presence of systolic murmur and third sound; diastolic pressure >95 mmHg; electrocardiogram left ventricular hypertrophy; cardiothoracic ratio >0.5 and valvular calcifications are associated with cardiac complications (p = 0.001-0.02), with low sensitivity (range: 14-38%) and high specificity (range: 85-98%). Cardiological referral was required for 169 patients (17%) that showed a higher prevalence of cardiovascular diseases (85%) and of cardiac complications (5.3%). Cardiologists required further tests for 13 patients (7.7%) and modified therapy for 93 (55%). High cardiac risk patients are identified preoperatively in current practice and cardiological referral is frequent; further studies are mandatory to evaluate the most effective and efficacious procedures. Key words: Cardiac risk, preoperative evaluation.
INTRODUCTION Preoperative evaluation of general physical status in patients scheduled for surgery is current practice and focuses mainly on cardiovascular function, since this is one of the most critical factors affecting perioperative complications [1-10]. The best approach for identifying high cardiac risk in a non-selected surgical population has been discussed previously [1,6,11-17]. New insights into the value of tests in different populations and budgetary constraints have compelled health care providers to critically review preoperative assessment practice [17-23]. Moreover, Requests for reprints should be sent to: Dr Cristina Maggio, Fondazione Arturo Pinna Pintor, Via A. Vespucci, 61, 10129 Torino, Italy. 217
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
218
C. Maggio et al.
information is lacking about the practice of and the real motivations behind common preoperative assessment in different geographical areas and institutions. The aims of this prospective observational study were: to identify high cardiac risk factors in non-selected, non-cardiovascular elective surgery patients; to gain insights into the behavior of surgical staff in our common practice; to evaluate the referral rate to cardiologist consultants and to compare cardiac risk of referred and nonreferred patients in our experience. MATERIALS AND METHODS
Between May 1987 and August 1989 patients aged 20 years or older scheduled for non-cardiovascular elective surgery were recruited in three institutions (two public hospitals and one private hospital). Patients undergoing emergency or vascular surgery, or local anesthesia were excluded. A preoperative evaluation, based on history and physical examination, was carried out on each patient by five trained investigators, while laboratory, electrocardiogram (ECG) and chest X-ray information was recorded when available on surgical staff request, according to their commonly used protocols. Sixty variables were recorded for each patient. ECG was requested for 982 patients (99%) and chest X-ray for 787 (79%). Data regarding patients referred for cardiological consultation were also recorded, i.e. diagnosis and effects of consultation (request for new test and/or modification of therapy). The investigators followed each patient on a daily base until discharge in order to collect information about the end points of the study: (1) death; (2) cardiac death (sudden death, death due to refractory low output cardiac failure not associated with sepsis or acute respiratory insufficiency); (3) major cardiac complications: pulmonary cardiac edema (dyspnea, bilateral pulmonary rales—Killip m/IV—with compatible chest X-ray), angina, non-fatal myocardial infarction (new pathological Q wave on ECG and/or angina associated with ST-T depression >1 mm and/or T inversion lasting at least 72 hr not due to drugs or electrolyte imbalance and/or increase in aspartate amino transferase and CK/CK-MB in serum, and congestive heart failure (dyspnea and/or edema associated with gallop rhythm, pulmonary rales, hepatomegaly, jugular distension and clinical benefit from diuretic therapy). Statistical Analysis
Chi-squared analysis with Yates correction when requested was used to compare the univariate association between each variable and the incidence of cardiac end points. The odd ratios and their confidence intervals were calculated in order to evaluate the relative importance of each variable. Sensitivity and specificity were calculated according to standard formulas (sensitivity: true positive/true positive + false negative; specificity: true negative/true negative + false positive) [24,25]. RESULTS Nine hundred and ninety patients aged from 20 to 93 years (mean age: 54 ± 15 years), 489 (49.3%) of which were males, underwent elective surgical procedures. The characteristics and type of the elective surgery are reported in Table 1.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
Cardiac Risk Assessment
219 TABLE 1.
Population characteristics 990
Total patients Mean age (years)
54± N
%
188 391 375 36 528/539
19.0 39.0 38.0 4.0
Cardiovascular diseases Hypertension Ischemic heart disease Valvular heart disease Cerebrovascular disease Peripheral vascular disease Others
169 39 27 14 24 6
16.2 4.0 3.0 1.4 2.4 0.6
Others Chronic obstructive lung disease Diabetes meQitus Hepatic insufficiency Chronic renal insufficiency
183 43 15 8
18.5 4.3 1.5 0.8
Types of surgery Abdominal Intra thoracic Superficial (including breast) Orthopedic Neck (including thyroid)
446 245 181 66 62
45.0 25.0 18.0
Age (years) 20-39 40-59 60-79 80 or more Sex(M/F)
TABLE 2.
6.0 5.0
Perioptrative complications N
%
Mortality Total Sudden death Cardiac death
16 4 8
1.6 0.4 0.8
Cardiac complications Non-fatal myocardial infarction Cardiac failure Cardiac pulmonary edema
1 11 4
0.1 1.0 0.4
One patient may have had more than one complication.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
220
C. Maggio et al. TABLE 3.
Risk factors associated with cardiac major events (unirariate analysis) Odd ratio (95% CI)
Prevalence (%)
P
Sensitivity Specificity (%) (%)
Age (years) >50 >60 >70
61.4 55.6 13.8
7.3 3.0 3.1
(1.7-31.3) (1.2-7 3) (1.2-7.9)
0.002 0.009 0.02
91 64 32
42 63 87
History Chest pain Dyspnea Hypertension
7.6 12.1 15.1
5.0 5.1 3.5
(1.9-13.5) (1.9-11.5) (1.4-8.7)
0.002 0.001 0.001
28 38 38
92 88 85
Physical examination Systolic murmur Third sound DiastoBc BP > 95 mmHg
15.4 1.5 7.2
3.5 10.5 5.4
(1.4-8.5) (2.7-39.5) (2.0-14.4)
0.01 0.001 0.001
38 14 28
85 98 93
4.0
6.0
(1.8-18.8)
0.004
19
96
11.0 5.0
4.3 6.6
(1.6-11.5) (2.3-19.0)
0.007 0.001
28 24
91 95
ECG (N = 982) Left ventricular hypertrophy Chest X-ray (N - 787) Cardiothoraric ratio >0.5 Valvular calcifications CI, confidence interval.
Sixteen patients died before discharge (total hospital mortality 1.6%). Cardiac mortality (8 patients) was 0.8% (of which half was due to sudden death). Major cardiac complications (cardiac death, cardiac failure, myocardial infarction and cardiac pulmonary edema) occurred in 2.3% (Table 2). In Table 3 are the reported prevalence, sensitivity and specificity of the variables significantly associated, by univariate analysis, to in-hospital major cardiac events, along with the odd ratio and confidence intervals for patients with such preoperative TABLE 4.
Preoperative cardiac referral
N
%
Diagnoses Non-cardiac disease Hypertension Ischemic heart disease Valvular disease Arrhythmias
26 66 35 23 19
15.4 39.0 20.7 13.6 11.2
Effects Further tests Therapy modifications
13 93
55.0
Total, 189 patients.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
7.7
Cardiac Risk Assessment
221
risk factors. None of the blood tests were associated with major cardiac events considered as end points of the study. Cardiological Referral Cardiological referral was required for 169 patients (17%). This population had a higher prevalence of cardiovascular diseases (85%) (Table 4). The cardiologist required further tests for 13 patients (7.7%) and modified therapy for 93 (55%). The referred population had a higher prevalence of major cardiac complications (5.3%; p = 0.017) with a relative risk of 3 (95% confidence interval 1.2-7.2) when compared with patients not referred preoperatively to the cardiologist.
DISCUSSION In our study, age is a relevant risk factor for cardiac complications, with the highest sensitivity for patients more than 50 years old. Other authors [11-15] have stressed the importance of age above 70 years as a risk factor for cardiac complications. In these studies, inclusion criteria limited the age population to subjects older than 40 years, while in our study patients aged from 20 to 39 years were included, and this may explain the different risk cut-off point that we observed. With the exception of age above 50 years, all otherriskfactors identified by current practice showed low sensitivity (range: 14-38%) and high specificity (range: 8598%). Since sensitivity represents the capacity of the test to minimize the number of patients considered healthy by the test but in fact at high cardiac risk, while specificity represents the capacity of the test to minimize the number of patients considered at high risk by the test, but in fact not at high cardiac risk, our results suggest that in a non-selected surgical elective population, at low absolute cardiac risk (2.3%), the procedure is effective in the identification of high cardiac risk, but is likely to be a failure for screening purposes, i.e. when dealing with an apparently healthy population. According to previous experience and the consensus in the literature [1-4,8,1117], in similar unselected surgical populations, preoperative symptoms and signs of heart failure or ischemic heart disease have been confirmed as being significant risk factors. In disagreement with other reports [11,14], a history of hypertension and the finding of preoperative diastolic hypertension are shown here to be significantly associated with high cardiac risk. If we consider that in our study electrocardiographic evidence of left ventricular hypertrophy was recognized as a risk factor as well, it can be suggested that the importance of hypertensive heart disease may have been underestimated and may deserve further studies. In our study, we could not confirm an association between the presence of Q-waves at baseline ECG and cardiac complications. Lack of a relationship may be explained by the impossibility of determining the age of the infarction from ECG; it has been demonstrated [7] that risk is exponentially reduced with time, with a critical reduction after 6 months. In fact, with the exception of one "silent" infarction, impossible to date, no patient in our study had suffered acute infarction previous to 6
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
222
C. Maggio et al.
months before the elective surgical procedure. On the other hand, Q-wave and nonQ-wave infarction carry the same risk [10]. In our study, we were also not able to confirm that arrhythmias (mainly atrial fibrillation) at preoperative ECG are associated with cardiac complications; the role of arrhythmias in surgical cardiac risk has been discussed previously [3,10-16]. Both the radiological signs we considered (cardiothoracic ratio >0.50 and valvular calcifications) were significant risk factors, according to the literature [11-16]: it is possible that both signs reflect advanced cardiopathy. In our series, preoperative ECG was required for 99% of patients, but the prevalence of abnormalities was low (9.4%) when compared with previous reports [26,27]. This may be explained by differences in the populations considered and by the fact that our study took into account only the few abnormalities that were expected to be relevant for cardiac risk. There is agreement in the literature [17,21,28] that preoperative ECG should be limited to older patients (aged more than 45, 50 and 60 years in different reports), both for risk evaluation purposes and for baseline reference. Nevertheless, in Italy, ECG is recommended in preoperative evaluation by the law regulating induced abortion, and this may account for its overuse in general surgery as well. The role of routine preoperative ECG, especially for baseline reference and complication monitoring and follow up, requires further study. In the literature [29,30], there is agreement that there should not be a generalized use of chest X-ray preoperatively, since the diagnosis of unexpected abnormalities is rare (1%). In our study, where more than half of the patients were older than 50 years, i.e. with a recommended preoperative indication [21], such a test was required for two-thirds of the patients. We found a high proportion of abnormalities (16%), considering that only two abnormalities, cardiomegaly and valvular calcifications, were registered. In a previous report that studied 3698 chest X-rays performed routinely in hospital on patients older than 40 years, a prevalence of 14% for cardiomegaly only is reported [31]. Although we cannot specify in our population the prevalence of unexpected abnormalities, we want to stress that both abnormalities we considered, cardiomegaly and the presence of valvular calcifications, were shown to be risk factors for cardiac complications. Cardiological Referral
To our knowledge, no data regarding the prevalence of cardiological referral in the preoperative evaluation of surgical patients is available, although a growing demand for such data from surgical staff is perceptible. Information about both the motivations (clinical and/or "defensive" medicine) and the purposes (cardiac risk stratification, diagnosis, possibility of improving treatment and/or monitoring in the perioperative period) are lacking as well. In this study, the group of patients referred to the cardiologist (17%) had a prevalence of cardiovascular diseases of 85% and a higher risk of cardiac complications (odd ratio = 3). In our opinion, this may mean that in common practice preoperative cardiological consultation is relatively frequent and, indirectly, that the surgical staff involves the cardiologist in the evaluation of high cardiac risk patients. Unfortunately, it has not been possible in this study to explain the motivations and the purposes of referral to the cardiologist, and which among the possible factors affecting decision-making throughout the preoperative
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
Cardiac Risk Assessment
223
evaluation was the most relevant for the surgical staff: this is the aim of an ongoing study. Ultimate exploitation of cardiological expertise in surgical patient care depends on a better understanding of the above two issues. When preoperative diagnostic and prognostic evaluation is the main purpose, it should be kept in mind that, according to Bayes' theorem, tests are more effective when applied to intermediate risk [32]. Because major cardiac events are so infrequent (2.3% in our sample) most of the prognostic information is contained in the a priori probability alone. Our observational results suggest that clinical examination is indeed useful in high cardiac risk identification in a non-selected surgical population, since it resulted in high cardiac risk patients being submitted to the cardiologist. The cardiologist, in most instances, did not require further tests for diagnostic purposes and was involved mainly in therapy modification. This observation is consistent with the hypothesis that the prevalent demand may for an interaction between surgical staff and cardiologist in order to optimize high risk patient management, as considered previously [9,33]. Unfortunately, there have been no studies, to our knowledge, considering the efficacy of intervention for or the effect of its omission on the prognosis of high risk cardiac patients. In conclusion, this study shows that, in a case-mix elective surgical population at a low absolute risk of mortality and cardiac complication, clinical preoperative evaluation, including ECG and chest X-ray for most patients, is unable to identify high cardiac risk patients. Further studies are mandatory to evaluate the procedures that are most efficacious and effective in high cardiac risk identification, in order mainly to improve sensitivity in the apparently healthy subjects. Cardiological intervention seems aimed mainly to optimize patient management. Specific studies are needed to evaluate whether such intervention modifies the prognosis of high cardiac risk surgical patients. Acknowledgement: We thank Paolo Maggio for his assistance in the elaboration of the results of this study.
REFERENCES 1. Mangano D T, Perioperative cardiac morbidity. Anesthesiology 72: 152,1991. 2. Cohen M M and Duncan P G, Physical status score and trends in anesthetic complications. J Clin Epidemiol 41: 83, 1988. 3. Salem D N, Chuttani K and Isner J M, Valutazione e trattamento di malattia cardiaca nel paziente chirurgico. In: Current problems in cardiology—Raccolta 1989, p. 133, Centro Scientifico Editore, Torino, 1990. 4. Rose S D, Corman L C and Mason D T, Cardiac risk factors in patients undergoing non cardiac surgery. Med Clin North Am 63: 1271, 1979. 5. Weitz H H and Goldman L, Non cardiac surgery in the patient with heart disease. Med Clin North Am 71: 413, 1987. 6. Editorial, Accounting for perioperative death. Lancet i: 1369, 1987. 7. Rao T L, Jakobs K H, El Etz et al., Reinfarction following anesthesia in patients with myocardial infarction. Anesthesiology 59: 499,1983. 8. Eagl K A and Boucher C A, Cardiac risk of non cardiac surgery. N EnglJ Med 321: 1330, 1989. 9. Wells P H and Kaplan J A, Optimal management of patients with ischemic heart disease for non cardiac surgery by complementary anesthesiologist and cardiologist interaction. Am Heart J 102: 1029, 1981.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
224
C. Maggio et al.
10. Shub C, Management of cardiac patient undergoing non cardiac surgery. Learning Center Highlights 6: 1,1991. 11. Goldman L, Caldera D, Nussbaum S R et al., Multifactorial index of cardiac risk in non cardiac surgical procedures. N EnglJMed 197: 845,1977. 12. Waters J, Wilkinson C, Golman M et al., Evaluation of cardiac risk in non cardiac surgical patients. Anesthesiology 55: A34,1981. 13. Detsky A S, Abrams H B, McLaughJin J R et al., Predicting cardiac complications in patients undergoing non cardiac surgery. J Gen Int Med 1:211,1986. 14. Zeldttn R A, Assessing cardiac risk in patients who undergo non cardiac surgical procedures. Can J Surg 27: 402, 1984. 15. Larsen S F, Olesen K H, Jakobsen Eetal., Prediction of cardiac risk in non cardiac surgery. Ew Heart / 8 : 179,1987. 16. Detsky A S, Abrahams H B, Forbath Setal., Cardiac assessment of patients undergoing non cardiac surgery. Multifactorial clinical risk index. Arch Intern Med 146: 2131, 1986. 17. Robbins J A and Mushlin A I, Preoperative evaluation of the healthy patient. Med Oin North Am 63: 1145,1979. 18. Editorial, Routine pre-operative investigations are expensive and unnecessary. Lancet H: 1466,1983. 19. Blery C, Chastang C and Gandy J H, Critical assessment of routine preoperative investigations. Effective Health Care 1: 111, 1983. 20. Turnbull J M and Buch C, The value of preoperative screening investigations in otherwise healthy individuals. Arch Intern Med 147: 1101,1987. 21. Haberer J P, Guelon D and Bichet G, Examen preoperatoir et Evaluation du risk op6ratoir, Encyd Mid Our (Paris, France) Anesthesie-Rianimation 36377: A05,1989. 22. Kaplan E B, Boeckmann A S, Roizen M F et al., Elimination of unnecessary pre-operative laboratory tests. Anesthesiology 57: A445,1982. 23. Blery C, Charpak Y, Szatan Metal, Evaluation of a protocol for selective ordering of pre-operative tests. Lancet I: 139,1986. 24. Bobbio M, Corso introduttivo ai metodi statistid in epidemiologia. G Ital Cordial 14: 206,1984. 25. Bobbio M, Corso introduttivo ai metodi statistic! in epidemiologia. G Ital Cordial 14: 941,1984. 26. Elston R A and Taylor D J E, The preoperative electrocardiogram. Lancet i: 349,1984. 27. Jakobson A and White T, Routine preoperative electrocardiograms. Lancet I: 972,1984. 28. Goldberger A L and O'Konski M, Utility of the routine electrocardiogram before surgery and on general hospital admission—critical review and new guidelines. Ann Int Med 105: 552, 1986. 29. National Study by the Royal College of Radiologists, Pre-operative chest radiology. Lancet ii: 83, 1979. 30. U.S. Department of Health and Human Services, The selection of patients for X-ray examinations. In: Presurgical chest X-ray screening examinations (April 1986). HHS Publications FDA 86-8261. 31. Sagel S S, Evans R G, Forrest JVetal., Efficacy of routine screening and lateral chest radiographs in a hospital based population. N EnglJMed 291: 1001,1074. 32. Bobbio M, Pollock B H, Cohen I and Diamond G A, Comparative accuracy of clinical tests for diagnosis and prognosis of coronary artery disease. Am J Cordial 62: 862,1988. 33. Downing J W and Tinker J H, Fitness for anesthesia: who decides? Lancet I: 387,1987.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/217/1797159 by Insead user on 11 July 2018
