Coronary artery bypass grafting (CABG) was performed primarily with the use of cardiopulmonary bypass with arrest of the heart by using cardioplegia. In patients with extensive coronary disease CABG reduces mortality. With the on pump approach perioperative mortality was 2% with an additional 5 to 7% patients having complications like stroke, myocardial infarction, renal failure requiring dialysis etc. to reduce the complications related to on pump surgeries because of cardioplegia and cross clamping of aorta the technique of performing surgery on beating heart was developed (off-pump). There were trials comparing on pump and off pump CABG . The largest of these studies was the Randomized On/Off Bypass (ROOBY) trial, where 2203 patients were enrolled from the Veterans Affairs medical system. However none of the previous trials could assess clinically important differences. Also, the skills of a surgeon could influence the outcome of a surgical procedure.[1,2]

 As a result of improving surgical strategy and the technology continually, cardiac surgery is now possible in an increasingly high-risk population. Crude mortality rates which have often been used as an indicator of quality of care have limited value without knowledge of the risk profile of the patients. It is noted by some that operative mortality is not the only important outcome measure, that patients who survive cardiac surgeries sometimes may still have high morbidity and a poor long-term outcome.

Risk scoring provides a useful tool for surgeons to make correct decisions as to whether or not coronary artery bypass graft is a suitable intervention, and which are the patients who should be carefully managed and monitored due to adverse outcomes of the operation. Furthermore, risk scoring models are also useful in surgical decision making, pre-operative patient education and consent as well as quality assurance measures. Risk prediction also allows more objective balancing of potential risks and benefits for individual patients. The most used scores initially were the additive (aES)  which was derived from an international European database of patients who had undergone cardiac surgery by the end of 1995 and logistic (lES). European System for Cardiac Operative Risk (Euroscore, ES). However, since 2012 these scores have been replaced by Euroscore 2.

Based on logistic regression analyses of 23 000 patients from 150 hospitals, the ES2 had a better predictive value in unselected cardiac surgical population using 18 different widely available preoperative clinical- and operation-related variables.[3]

 The Society of Thoracic Surgeons (STS) database was established in 1989 to assess operative mortality after cardiac surgery in adults.[4,5] This risk-stratification model has been used widely in North America and comprises over 40 clinical parameters. The original Euroscore and the STS risk algorithms have been compared for coronary bypass grafting. However, comparisons between the Euroscore II and STS in the patient populations for coronary bypass grafting are limited. The aim of the study is to compare the risk models on a prospectively collected data in a cardiac surgical cohort.

Prospective cohort study in Department of Anesthesia in Narayana Hrudayalaya, Bommasandra, Bangalore November 2019 to June 2020. The institutional ethics committee approval for the study and written informed consent from patients will be obtained.. A sample of 350 patients of either sex undergoing off pump CABG under general anesthesia

Sample Size- The sample size is calculated by taking an approximate difference in sensitivity between two studies as 10% after reviewing multiple studies.
Sample size was calculated using statistic and sample size pro software version 1.0. Based on difference in sensitivity between STS score and Euroscore as 10% with an event rate of combined mortality and morbidity as 10% , 10% of precision the minimum required sample size for OPCABG is 350 with a minimum sample size for combined mortality and morbidity is 35

Formula for sample size calculation

n= z21-alpha/2 sen(1-sen)/ d2

* Prev

n = sample size

Sen = expected sensitivity

Prev= prevalence of disease/event rate

d= precision

alpha= alpha error

Inclusion Criteria: All patients undergoing OPCABG surgeries , Elective or Emergency cases

Exclusion Criteria: Patients undergoing CABG combined with valve surgeries, Elective ON PUMP CABG cases.

METHODOLOGY:

 All OPCABG procedures undertaken in the study site which satisfy the inclusion criteria were included in the study. Patients were explained regarding the procedure during pre-anaesthetic visit (PAC). Patient demographic data such as Name, age, sex, weight, EuroSCORE (European System for Cardiac Risk Evaluation) were duly collected and appropriately calculated. A detailed history of ischemic heart disease (IHD), comorbid illnesses like Diabetes Mellitus (DM), Hypertension, prior surgical and anaesthetic experience, if any were carefully elicited. The drug therapy for IHD and other co-morbid illness if any, were ascertained and documented. During general examination, patient‘s general condition was assessed and other factors like pulse rate and blood pressure were measured and documented. A detailed assessment of Airway, Cardiovascular system (CVS), Respiratory System (RS), Gastrointestinal System (GIT) and Central Nervous System (CNS) were carried out and documented. Basic laboratory data including coagulation profile, ECG, Chest-X ray and 2D-ECHO, angiography was appropriately reviewed and documented. As per the institutional protocol, Angiotensin converting enzyme inhibitors and Angiotensin receptor blockers were withheld during the day of surgery. All other cardiac medications which the patient may have been receiving were continued. All patients were kept NPO for at least 6 hours prior to surgery except for emergency cases. All patients were premedicated with Diazepam 0.1mg/kg orally. As per the Institutional protocol, oral Metoprolol was administered for heart rate control as per the guidelines:

12.5mg if heart rate >50 and ejection fraction 40-50%.

25mg if heart rate >50 and ejection fraction >50%.

No beta blocking medications are initiated or stopped on the day of surgery.

After the patient was shifted to the operation theatre, monitors such as electrocardiogram (ECG), pulse oximetry (SpO2), non-invasive blood pressure (NIBP) and temperature probe were connected. Radial artery was subsequently cannulated for invasive blood pressure (IBP) monitoring. An 18G peripheral intravenous access was established and subclavian 7F Triple lumen central line was secured. Baseline vitals which include Heart rate, Blood pressure (both invasive and non-invasive), blood oxygen saturation (SpO2) and arterial blood gas (ABG) were duly documented.

After preoxygenation with 100 % oxygen for 3 minutes, GA induction was done with I/V Fentanyl 10mcg/kg and I/V Propofol 0.1mg/kg. Endotracheal intubation was done following the administration of muscle relaxant – I/V Pancuronium 0.1mg/kg. Anaesthesia was maintained with Isoflurane at minimum alveolar concentration 1.2%, Oxygen and air supplemented to maintain an FiO2 of 0.5.

Intra-operatively the complications such as conversion to CPB (cardio pulmonary bypass), presence of myocardial ischemia or infarction, any arrythmias or IABP (Intra-Aortic Balloon Pump) use were noted and documented according to the preformulated proforma. Post sternotomy closure, patients were shifted to Intensive Treatment Unit (ITU) for post- operative monitoring.

Post-operatively, the complications occurring till discharge were duly noted and documented as per the preformulated proforma such as cardiac arrhythmias, re-explorations, need for mechanical ventilation, wound infections or postoperative IABP. In hospital mortality is recorded.

Statistical Analysis

The statistical analysis was done with the statistical package for windows (SPSS Version 23). Continuous variables were described as mean and variation of each observation from the mean value (Standard deviation) represented as mean ± SD or median and interquartile range if they failed to follow a normal distribution. Categorical data was represented as frequencies and percentages the calibration of both scoring systems was assessed using the Hosmer-Lemeshow (HL) test, which compares the observed versus expected mortality by risk decile. Calibration is poor if the test is significant. The discrimination measures the capacity of a model, in this case Euroscore II and STS, to differentiate between the individuals of a sample that suffer an event, death or morbidity and those who do not. The discriminative capacity of the analyzed events of the two scales was estimated by means of receiver operating characteristic (ROC) curves. Their areas under the curve (AUC) were calculated. P value <0.05 was considered statistically significant.

A total of 367 patients were included in this study in accordance with the afore mentioned inclusion criteria. The results obtained in keeping with the previously set objectives may be presented as follows. Out of the total of 367 patients included in this study, 355 belong to the males

Table-1: Distribution of socio demographic variables

Majority of the patients belonged to the age group of 51-70 years. In our study out of 367 total sample size 28 patients were females and 339 patients were

Figure-1: Frequency distribution of in hospital mortality

In our study, the overall in hospital mortality observed was 3% and 97% were alive.

Table 2: Category wise distribution of EuroSCORE II

Majority (77.3%) of the patients had a EuroSCORE of 1-2, 24.5% of the patients had a score between 3-5 and 1% had score >5.

Table-3: Euro Score II in relation to various variables

Table-4: Category –wise distribution of STS SCORE

Majority of the patients were categorized into STS score 0-2.

Table-5: STS scores in relation to various variable.

Table-6 : Univariable analysis for Factors affecting mortality

The results of the study indicated that the age of the patient played a significant role in determining the mortality outcome with the mean age among the surviving and expired population standing at 58.74 and 64.08 respectively (Figure 2). Comparison of the AGE (years) between the two groups shows that AGE (years) is higher in the expired group with a t-value of -2.191 which is statistically significant with a p value of 0.029. Assessment of the relationship between Gender and mortality yielded a P-value of 0.92 indicating a non-significant correlation. In our study for prolonged ventilation observed morbidity was 3% and predicted morbidity with STS score was 2.23%. AUC was 0.626 with p value 0.154 which is non-significant.

Figure-2: ROC for STS score Prolonged Ventilation

In our study for DSW infections, observed morbidity was 0.3% and predicted morbidity with STS score was 0.12%. AUC was 0.247 with P value 0.384 which is not significant.

Figure-3: ROC for STS and DSWI

In our study observed morbidity REOPERATION was 1.40 and predicted morbidity with STS score was 0.71%. AUC was 0.612 with a P value of 0.389 which is not significant.

Figure-4: ROC for STS Score for Reoperation.

In our study observed morbidity Renal failure was 1.40 and predicted morbidity with STS score was 0.71%. AUC was 0.893 with a P value of 0. 003 which was found to be significant.

Figure-5: ROC for STS score for Renal Failure

Table-7: A table depicting comparison of observed and predicted mortality based on Euro Score:

In our study we divided the population based on EUROSCORE II into low risk (<2%), moderate risk (2-5%) and high risk (>5%) respectively. A one look superficial side by side comparison of 3 groups shows that the mean values of observed mortalities are 1.07%, 10.97% and 0 respectively while the predicted values are 1.31%, 2.71% and 5.47% respectively. The area under curve is 0.89 with p value <0.001 which is statistically significant.

Table-8: A table depicting comparison of observed and predicted mortality based on STS Score:

In our study patients were divided based on STS scores into three groups namely low risk (<2), intermediate risk (2-5%) and high risk respectively (>5%). A side-by-side comparison of three groups showed observed mortality as 2.2%, 50% and 0 respectively while the predicted mortalities are 0.53%, 2.65% and 0. AUC was 0.93 and p value <0.001 which is statistically significant.

Figure-6: A demonstration of ROC comparision between EuroSCORE and STS predicted risk.

The present study being prospective, demonstrates that using Hosmer-Lemeshow test, EuroSCORE II and STS risk-score have good calibration power (P = 0.417 and P = 0.858, respectively) indicating satisfactory model fit. The area under the ROC curve 0.869 and 0.973 for EuroSCORE II and STS risk-score, respectively, could achieve statistical significance (P = <0.001 and P = <0.001 for EuroSCORE II and STS risk-score respectively) indicating good discriminatory power in the present cohort.

Our study was done among off pump CABG patients to compare Euroscore II and STS score. Both the scores were comparable and had good sensitivity and specificity in our sample size. We have divided the patients into three risk groups namely low risk <2% moderate risk (2-5%) and high risk (>5%). In our study Euroscore II has significantly overestimated mortality in high-risk group. Euro Score II and STS risk-score have good calibration power (P = 0.417 and P = 0.858, respectively) indicating satisfactory model fit. The area under the ROC curve 0.869 and 0.973 for Euroscore II and STS risk-score respectively could achieve statistical significance (P = <0.001 and P = <0.001 for Euroscore II and STS risk-score respectively) indicating good discriminatory power in the present cohort.

The utilization of preoperative mortality risk algorithms to determine the best treatment modality for a given patient with cardiovascular disease is a common practice among cardiac surgeons. In practice at the bedside, the use of risk scoring algorithms serves two purposes beyond simply giving both the surgeon and patient a predicted numerical outcome profile of the upcoming surgery: 1) the patient is more fully informed of the risks inherent in their upcoming risky procedure and 2) provide the surgeon with an objective, measurable risk profile for comparison purposes should the surgery be delayed, thereby allowing time for a diminished risk profile.

In our study we are comparing two models that are most frequently used in cardiac surgical procedure. There are important differences between the STS and Euroscore II. The Euroscore models are simple, derived from a limited set of variables and applicable to the entire spectrum of cardiac surgical procedures. They are based on a fixed and relatively small number of patients operated years ago. Therefore, they may fail to capture the on-going changes in patient profiles and practice patterns. In contrast, the STS ACSD risk-prediction models are available for only five surgical procedures (although accounting for 70–80% of the average practice), are based on large amount of data, and are updated periodically [6]. The most recent version of the STS ACSD risk algorithms were developed based on 775,000 operative records. In addition to mortality, the STS ACSD provides important additional risk- adjusted predictive outcomes scores, including overall morbidity and specific complications affording the opportunity to calculate comprehensive procedural composite scores. In contrast, the two Euroscores provide prediction of procedural mortality only.

 We observed a good correlation and model discrimination among the two scores. These findings suggest that the models performed equally well in the cohort tested. This study served its purpose in comparing Euroscore II and STS score in terms of sensitivity and specificity. In our study we observed that for a Euroscore II cut-off for 2.7 the sensitivity is 100% and specificity is 92.8% and for STS score cut-off 0.93% the specificity is 91.7%.

  In a prospective comparative study among 1769 patients over a period of 4 years conducted at National University Heart Singapore, it was found that the Society of Thoracic Surgeons (STS) (C-statistic 0.774, 95% CI 0.697-0.851) and Euroscore (C-statistic 0.766, 95% CI 0.690-0.843) have the best predictive value for patients in South East Asian origin and overall mortality was 2.48%. This study mainly evaluated the efficacy of risk scoring systems in South East Asian population as most of the cardiac surgery risk models are based on predominantly Caucasian population. Our study yielded similar results in Indian population. [7]

In a comparison of the STS score and the Euroscore II for specific cardiac surgeries, a Pakistani study found that Euroscore II was superior in isolated valve surgeries, whereas the STS score was superior in combined CABG and valve procedures . However in our study we only compared both the scores in off pump CABG surgeries. In relation to studies that concentrated on the Indian population, a study conducted by Deepak Borde et al[8]  focused on the relation between Euroscore II and complications of OPCABG namely mortality rate and conversion to on pump. This was conducted at two separate institutions at Aurangabad and Bangalore between 2009 to 2014. The study concluded that Euroscore II risk scoring has satisfactory calibration and discrimination power to predict the outcomes that they studied (p value of 0.472 and area under curve 0.706). This finding concurs with our study result ascertaining the discriminatory power of Euroscore II with a ROC area under the curve of 0.869.

Another study was conducted by Kar et al[9]  at Nizams Institute of Medical Sciences, Hyderabad in 2011-12. The study was conducted in a rural population. The study revealed agood discriminatory power of Euroscore II in all cardiac surgeries in general for predicting the mortality rate. But specifically, the study concluded that the predictive power of Euroscore II was very good in low-risk groups but however under-predicted mortality in all the other risk groups. Our study also revealed similar results with Euroscore II bring a good predictor of mortality in low-risk groups but over estimating mortality in high-risk group.

A prospective study conducted at Max Heart and Vascular Institute by Baloria et al[10] for predictive accuracy of Euroscore II in an urban Indian population, suggests that the Euroscore-II model in its present form may not be entirely validated for use in the Indian population with high P values stating good predictive mortality, except in high-risk group (p = 0.175). Our study also enabled the determination of extent of influence of various factors on the mortality outcome of the study subjects as well as determine the extent of occurrence of various related complications.

A study conducted by Poullis et al.[11] stated that it does not appropriately account for age and hence may not be useful for patients more than 70 years of age . In this study, we have taken into account the age of all 367 patients and attempted to determine its impact on the mortality outcome. Our study has shown a significant relationship between advancing age and mortality indicating that the mortality rate increases with advancing age. This may be attributed to the fact that with increasing age, the occurrence of associated comorbidities as well as related complications increase proportionally. Age has been included as a contributive factor towards mortality.

Patient Gender is another factor included in 2 scores with a slightly higher risk of mortality attributed to the Female Gender. However, in our study which comprised of a total of 27 females and 338 males, we were able to establish a non-significant correlation between mortality and gender. This brings into question the inclusion of gender into the scoring system as a reliable/significant contributory factor. However, the true nature of this relationship may be only determined by an extensive study considering the strength and nature of all other confounding factors as well as using a study population comparable in all respects.[12,13]

The next criteria used was the presence or absence of Chronic Obstructive Pulmonary Disease. Our study population consisted of 14 patients with chronic lung disease and established a significant relationship with mortality. A significant relationship has been established with patients who are alcoholics and mortality. This can be attributed to the fact that chronic alcohol use accelerates the disease process and may cause significant morbidity.  Patients with sleep apnea have a significant relationship with mortality according to our study. This variable is in the STS score and not Euroscore II.

The respective NYHA Class of each individual patient was included as a relevant criterion in the latest Euroscore II model and STS model. In order to study the significance of this criterion, we classified patients into 4 separate groups based on their respective NYHA Class (I, II, III, IV). Based on the application of statistical analysis to the aforementioned data and its evaluation, we were able to determine a significant relationship. Percentage stenosis of left main artery and proximal LAD also has a significant relationship with mortality according to our data analysis as the mortality was higher in patients with higher percentage of stenosis of the artery.  Patients with third degree heart block which was included in the STS score have a significant relationship with mortality. In our study we recorded the type of arrhythmias present in our population. This is not included in Euroscore II but included in STS  score. Recent MI (Myocardial Infarction within 90 days) is one of the criteria which forms the Cardiac Related factors in STS  score and Euro SCORE II. Our study comprised of 73 patients with a positive history of recent MI. Evaluation of the obtained results, however failed to recognize a statistically significant relationship between Recent MI and Mortality. Our study therefore rules out Recent MI as a significant contributor towards mortality outcome.

This study also grouped the study population into 2 based on the need for intra- operative Intra-aortic Balloon Pump (IABP) requirement. 11 out of the total 367 patients in our study population required the usage of intra-operative IABP insertion. Statistical analysis of this data revealed a significant relationship between intra- operative IABP insertion and mortality outcome. Our study therefore recognizes Intra- op IABP insertion is a significant indicator of mortality. Our study also recognizes a significant relationship between patients on extra corporeal membrane oxygenation and mortality.

The relationship between prolonged mechanical ventilation and mortality has been recognized. Statistical analysis determined a significant relationship between post- operative ventilation and mortality outcome with mortality significantly increasing with the duration of ventilation.

LIMITATIONS OF STUDY

The following may however be considered the limitations that may have curtailed the development of a more comprehensive and conclusive study result. Certain limitations of this study included those inherent to all prospective cohort studies. Lack of requisite amount of study patients for in-depth analysis of various subcategories. Inability of standardization of study subjects with respect to individual contributory factors owing to the multitude of cross-confounding factors. Suboptimal number of study subjects due to various aforementioned factors thereby confining the study to the minimum required sample size..

The study revealed that both the scores are comparable in predicting mortality with good calibration power while STS score additionally could predict morbidity outcomes. However due to the ease of use , our study recommends Euro SCORE II in assisting the determination of mortality outcome in patients undergoing OPCABG. By achieving the goals set out by the current study, it should provide a stronger platform to base and evolve current practices that would be more appropriate for the Indian population.

1. Novitzky D, Shroyer AL, Collins JF, McDonald GO, Lucke J, Hattler B, Kozora E, Bradham DD, Baltz J, Grover FL, VA# 517 Randomized On/Off Bypass (ROOBY) Study Group. A study design to assess the safety and efficacy of on-pump versus off- pump coronary bypass grafting: the ROOBY trial. Clinical Trials. 2007 Feb;4(1):81- 91.
2. Shroyer AL, Grover FL, Hattler B, Collins JF, McDonald GO, Kozora E, Lucke JC, Baltz JH, Novitzky D. On-pump versus off-pump coronary-artery bypass surgery. New England Journal of Medicine. 2009 Nov 5;361(19):1827-37.

1. Nashef SA, Roques F, Sharples LD, Nilsson J, Smith C, Goldstone AR, Lockowandt U. Euroscore ii. European journal of cardio-thoracic surgery. 2012 Apr 1;41(4):734- 45.
2. Shahian DM, O'Brien SM, Filardo G, Ferraris VA, Haan CK, Rich JB, Normand SL, DeLong ER, Shewan CM, Dokholyan RS, Peterson ED. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: part 1—coronary artery bypass grafting surgery. The Annals of thoracic surgery. 2009 Jul 1;88(1):S2-2.
3. O'Brien SM, Shahian DM, Filardo G, Ferraris VA, Haan CK, Rich JB, Normand SL, DeLong ER, Shewan CM, Dokholyan RS, Peterson ED. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: part 2—isolated valve surgery. The Annals of thoracic surgery. 2009 Jul 1;88(1):S23-42.
4. Jin Huang and C. X. Ling, "Using AUC and accuracy in evaluating learning algorithms," in IEEE Transactions on Knowledge and Data Engineering, 2005:vol. 17, no. 3, pp. 299-310.
5. Liang S, Zheng Daniel Lim Y, Ma R, Shing Yow K. Quality of care and outcomes session title: Acute and chronic coronary artery disease: quality care and outcomes III. 2015;132(A):9972
6. Borde D, Asegaonkar B, Apsingekar P, Khade S, Futane S, Khodve B et al. Risk Stratification in Off-Pump Coronary Artery Bypass (OPCAB) Surgery—Role of EuroSCORE II. Journal of Cardiothoracic and Vascular Anesthesia. 2015;29(5):1167-1171.
7. Kar P, Geeta K, Gopinath R, Durga P. Mortality prediction in Indian cardiac surgery patients: Validation of European System for Cardiac Operative Risk Evaluation II. Indian Journal of Anaesthesia. 2017;61(2):157.
8. Baloria K, Selot N, Pillai B. Validation of the European System for Cardiac Operative Risk Evaluation-II model in an urban Indian population and comparison with three other risk scoring systems. Annals of Cardiac Anaesthesia. 2015;18(3):335
9. Qadir I, Alamzaib SM, Ahmad M, Perveen S, Sharif H. EuroSCORE vs. EuroSCORE II vs. Society of Thoracic Surgeons risk algorithm. Asian Cardiovascular and Thoracic Annals. 2014 Feb;22(2):165-71.
10. Poullis M, Pullan M, Chalmers J, Mediratta N. The validity of the original EuroSCORE and EuroSCORE II in patients over the age of seventy. Interactive cardiovascular and thoracic surgery. 2015 Feb 1;20(2):172-7.
11. Puskas JD, Kilgo PD, Thourani VH, Lattouf OM, Chen E, Vega JD, Cooper W, Guyton RA, Halkos M. The society of thoracic surgeons 30-day predicted risk of mortality score also predicts long-term survival. The Annals of thoracic surgery. 2012 Jan 1;93(1):26-35.