Tools and scores for general and cardiovascular perioperative risk assessment: a narrative review

ABSTRACT The number of surgical procedures in the world is large and in Brazil it has been expressing a growth trend higher than the population growth. In this context, perioperative risk assessment safeguards the optimization of the outcomes sought by the procedures. For this evaluation, anamnesis and physical examination constitute an irreplaceable initial stage which may or may not be followed by complementary exams, interventions for clinical stabilization and application of risk estimation tools. The use of these tools can be very useful in order to obtain objective data for decision making by weighing surgical risk and benefit. Global and cardiovascular risk assessments are of greatest interest in the preoperative period, however information about their methods is scattered in the literature. Some tools such as the American Society of Anesthesiologists Physical Status (ASA PS) and the Revised Cardiac Risk Index (RCRI) are more widely known, while others are less known but can provide valuable information. Here, the main indices, scores and calculators that address general and cardiovascular perioperative risk were detailed.


INTRODUCTION
T he volume of operations in the world is vast1 and Brazil has shown a growing trend in the number of surgical procedures, which is proportionally higher than the population growth 2 .Despite this, the estimate of the need for operations 3 considerably exceeds the numbers contained in the public records, showing room for expansion.
In this context, perioperative risk assessment is necessary to mitigate the potential impacts of morbidity and health expenses arising from the growing number of surgical procedures and their complications.The risk assessment performed in the preoperative period aims to optimize outcomes from the perioperative period to the patient's full recovery in the late postoperative period.
The bases are the anamnesis and the physical examination, which are essential and irreplaceable steps to identify comorbidities, indicate additional tests, recommend clinical stabilization and possible contraindications to the operation.After this step, calculators and scores provide physicians and patients with increased objectivity of the risk-benefit assessment before joint decision-making, especially in elective procedures.
Despite the usefulness calculators and scores, they appear dispersed in the literature so that the gathering and detailing of their functioning add didactic and informative value to professionals who will use them, in addition to enabling an analytical view that allows the choice of the best tool for the patient's preoperative health status.
The systems approach facilitates the organization of the preoperative risk assessment.This literature review lists and discusses indices, scores, and calculators related to general perioperative and cardiovascular risk that receive greater focus in medical practice.We searched the electronic databases Pubmed/MEDLINE and EMBASE for manuscripts in English and Portuguese.The scope of this review does not include cardiac operations, which have specific risk assessment scores.

General risk assessment in non-cardiac surgeries
The incidence of complications resulting from non-cardiac procedures is on average between 7% and 11%, reaching 21.4% depending on the location and on the safety measures adopted 4 .The average 30-day mortality rate is between 0.8% and 1.8% 4,5 .The ASA PS score and the ACS NSQIP calculator described below are tools capable of predicting the risk of complications and mortality in general, without guidance by organ system.

Status (ASA PS)
The American Society of Anesthesiologists (ASA) classification was created in 1941 with the aim of simply determining the clinical status of surgical patients6.
The tool was revised in 1963 and became widely used in the preoperative period, due to its simplicity and reproducibility.The patient's clinical status is assigned a scale between I and VI (Table 1).

E Addition of the letter "E" denotes surgical emergency
There are criticisms of the use of the ASA PS as a surgical risk assessment, since it was not created with the aim of assigning risk and there may be interprofessional variation in the patients' classification 7 .However, the tool is simple, fast, easy to use, independent of complementary tests, can be a good predictor of risk of death in conditions of low mortality 8 , and is an independent predictor of postoperative complications and mortality 9 .poor performance for other outcomes [13][14][15] .

Cardiovascular risk for non-cardiac operation
Myocardial lesions occur in 13%5 of noncardiac surgeries and increases the risk of complications such as heart failure, stroke, and cardiac arrest, accounting for 34% of perioperative deaths 16 .Furthermore, cardiac complications determine a prolonged length of stay after the surgical procedure 17 .For these reasons, cardiovascular assessment has the largest number of validated algorithms and scores to date.

Cardiac Risk Index -Goldman index
The Cardiac Risk Index (CRI) was described in 1977 as the first multifactorial model specific for perioperative cardiac complications in non-cardiac procedures.This model categorizes the patient into four classes (I to IV) based on predefined scores for clinical, electrocardiographic, and laboratory factors, as well as type of operation (Tables 3 and 4).Outcomes considered are myocardial infarction, pulmonary edema, ventricular tachycardia within six days after surgery, and death from cardiac causes 18 .
Criteria Points

Detsky Index
Developed in 1986 as an adaptation for the Goldman's risk (CRI), it included variables considered clinically important by the authors, in addition to simplifying the scoring scheme, as shown in Table 5.
The type of operation was also removed from the index as it was not a patient's characteristic, and validation included minor procedures, such as cataract extraction or prostate resection 18 .Expected outcomes are myocardial infarction, acute pulmonary edema, tachycardia or ventricular fibrillation requiring electrical cardioversion, death from cardiac causes, and worsening or onset of heart or coronary failure 22 .
The assessment by this method requires knowledge of the pre-test risk of complication of the operation to be performed, which, combined with the Detsky score, determines the posttest risk.The authors propose the use of a nomogram to detail the posttest risk according to the score.In summary, scores below 10 mean that the patient's risk is less than the pre-test probability of complications from that operation.A score equal to 10 means equal pre-and posttest risk, and greater than 10 expresses that the estimated risk is above the mean 22,23 .The Detsky index has already been shown to be equivalent to other perioperative cardiac risk assessment scores but may be inferior to the Revised Cardiac Risk Index (RCRI), described below 24 , in predicting death or stroke, wound complications, and minor neurological complications 25 . .

Revised Cardiac Risk Index (RCRI)
The index proposed in 1999 24 was based on the Cardiac Risk Index 18 and aims at carrying out a simple assessment of the perioperative risk of major cardiac complications in patients aged 50 years and over undergoing major non-cardiac surgeries.Major cardiac complications were defined as acute myocardial infarction, acute pulmonary edema, ventricular fibrillation or primary cardiac arrest, and complete atrioventricular block.The variables independently associated with the increased risk of major cardiac complications were six: high-risk operation, ischemic heart disease, heart failure, history of cerebrovascular disease, insulin-dependent diabetes mellitus, and creatinine >2mg/dL, with an odds ratio between 1.9 and 3.0.High-risk operations were defined as intraperitoneal, intrathoracic, or suprainguinal vascular procedures 24 .For each of the variables, 1 point is attributed, and the classification is made as shown in Table 6.The predictive capacity of this scheme was confirmed in further studies 26,27 .It has been one of the most widely used risk assessment scores.
RCRI is well suited for stable patients who will undergo major, non-urgent, noncardiac surgeries, but limited for vascular procedures such as abdominal aortic aneurysm repair 24,26 , small surgeries, and very high-risk populations -as in emergency situations 24 .It should be noted that this score predicts cardiac complications and mortality, but it is not a good predictor of overall mortality 26 .One of its limitations is the exclusion of some factors considered clinically important, such as age, functional tolerance, and aortic stenosis 26 ..

Multicenter Perioperative Evaluation Study (EMAPO)
EMAPO is a Brazilian classification published in 2007 that assesses 27 variables to estimate perioperative risk.Each of these variables is assigned a specific score and the result of the sum of the points of the present variables classifies the patient into one of five risk levels (Tables 7 and   8).On the positive side, the study included validation for the Brazilian population, the inclusion of diseases not addressed by previous risk assessment guidelines, and modern treatment options in its objectives, to determine new variables associated with cardiovascular complications 34 .The index requires a large amount of information for the application, which can be a limitation 34 .On the other hand, it remains among the indices highlighted by the perioperative cardiovascular assessment guideline of the Brazilian Society of Cardiology, since it was developed and validated for the Brazilian population.The guideline recommends its use in patients without previous severe cardiovascular disease -which must be treated before the operation -and in non-urgent procedures 30.

MICA)
NSQIP MICA is a calculator created in 2011 from an extensive database (more than 400,000 patients), multicentric (more than 250 hospitals) and prospective, which aimed to assess risk factors associated with myocardial infarction or cardiac arrest in the peri and postoperative period (up to 30 days after the operation), as this would be a weak point of the risk scores developed so far.These outcomes are considered relevant because, despite being rare (less than 1% in the peri or postoperative period), when they occur, they result in death in 61% of cases within 30 days after the procedure 35 .
The variables associated with an increased risk of myocardial infarction or cardiac arrest were ASA class, dependent functional status (partially or totally), elevated creatinine (>1.5mg/dL), age, and type of operation.The consideration of dependent functional status in the assessment is a differential of this tool, as it did not appear in other previously published systematized assessments.
As this is a more complex calculation, it is used on a website, available at: http://www.surgicalriskcalculator.
com/miorcardiacarrest, which can be downloaded or used on the online platform 35 .
Compared to the RCRI, the MICA risk assessment benefits from greater specificity in relation to the procedure performed, but there is no significant association of heart failure with the primary outcomes not covered by the high ASA class and functional dependence 35 , and it remains limited for vascular operations 36 .
A retrospective observational study found a disagreement between MICA and RCRI assessments in classifying patients at low risk for adverse cardiac events in 30% of cases; the two tools look for different primary outcomes, but disagreement can be problematic, as low-risk patients often undergo surgery without further evaluation 37 .Even so, MICA is among the risk indices recommended by the American (ACC/AHA) 31 and European (ESC/ESA) 32 guidelines for perioperative risk assessment.

Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (POSSUM)
Published in 1991, POSSUM was developed with 1,372 patients undergoing elective or emergency operations in Liverpool in the years 1988-1989.It is a dual scoring system that combines a 14-item physiological score and a six-item operative severity score, which allows for a more accurate differentiation of risk by type of procedure.The study that originated it showed a good relationship between the predicted risk and the mortality and morbidity outcomes found.However, it was limited to a small population and developed with the aim of assisting in surgical auditing and evaluating quality of care, not validated for the process of decision making 38 .The POSSUM assessment has already undergone some adaptations 39 , among which the Portsmouth-POSSUM (P POSSUM) 40 stands out.It has been observed that the original POSSUM overestimates the prediction of mortality, especially in low-risk patients [41][42][43][44] , while the P POSSUM is more accurate in predicting postoperative mortality in various surgical scenarios 39,41,42,44 .

Index (VSG-CRI)
The VSG-CRI was proposed in 2010 with the objective of predicting cardiac events specifically for nonemergency vascular operations, seeking efficacy superior to the RCRI in this group, as the latter underestimates the risk of cardiac events in vascular procedures.The proposal is similar in logic to the RCRI, assigning points to a simple score, though the risk factors used are partially different (Tables 9 and 10).The outcomes considered in this evaluation are myocardial infarction, clinically significant arrhythmia, and in-hospital congestive heart failure45.
Currently, the VSG-CRI calculator is also available online at http://www.qxmd.com/calculate-online/vascular-surgery,where one can select the specific assessment for each type of vascular procedure.
The original work proposing the VSG-CRI found greater accuracy than the RCRI in the assessment of risk for vascular operations 45 .Subsequent studies in substantially smaller groups evaluated the VSG-CRI compared to the RCRI in arterial vascular operations and found low accuracy for the RCRI, as expected, but disparity in the results of the VSG-CRI.On the other hand, these studies agree that the VSG-CRI was not adequately accurate in the risk assessment for endovascular repair of abdominal aortic aneurysms (EVAR) [46][47][48] .All tools detailed here were developed and should be used for general non-cardiac operations.Some consider the type of operation within the evaluation, which may be of interest to the evaluator, namely ACS Calculator NSQIP, Goldman (CRI), EMAPO, MICA, VSG-CRI (this one specific for vascular operations), and the Model for Stroke and Cardiac Risk After Surgery.It should also be noted that the Goldman, RCRI, and MICA models have limited accuracy for vascular procedures, which is why the VSG-CRI is preferred in the risk assessment of this type of operation.Finally, the combined use of more than one tool can be a strategy adopted by the physician to compose the assessment.

Study Limitations
It is noteworthy that the narrative review, the format chosen for aggregating and discussing the information contained herein, is subject to some degree of subjectivity.However, physicians who perform the preoperative assessment will be able to take advantage of this information to adapt the decision-making process about performing a procedure, use calculators and risk scores to complement their assessment, and guide preoperative clinical interventions and the joint decision with the patient.

National Surgical Quality Improvement Program Risk Calculator (ACS NSQIP
11his calculator was initially developed between 2009 and 2012 in the United States based on data from 393 hospitals and about 1.4 million patients, with the objective of becoming a universal tool for estimating surgical risk11.It uses 21 patient variables, including the type of operation intended, and delivers the risk of nine main outcomes within 30 days of the procedure, which are summarized in Table2.It is currently available online, free of charge, and in English (https://riskcalculator.facs.

Table 2 .
ACS NSQIP calculator variables and outcomes.The type of operation is added to these variables to calculate the risk.

Table 4 .
18I classes and respective risks of complications and cardiac death18.

Table 6 .
24riables, classes, and risk of cardiac complications according to RCRI24.Risk factors considered in the preoperative evaluation are known ischemic heart disease, heart failure, high-risk operation (as in RCRI), diabetes mellitus, renal failure, and inadequate functional status.If all these factors are absent, the authors do not recommend 32rioperative cardiovascular risk assessment indices included in the guidelines of the Brazilian Society of Cardiology (SBC)30, American College of Cardiology, American Heart Association (ACC/AHA)31, European Society of Cardiology, and European Society of Anesthesiology (ESC/ESA)32.Fleisher-EaglePublished in 2001, this assessment resembles the RCRI in the evaluated parameters.However, it does not assign a score, but proposes a flowchart that indicates measures according to the number of risk factors found, to avoid cardiac complications (myocardial infarction, death from cardiac causes).

Table 8 .
37assification of cardiovascular risk according to the EMAPO assessment score37.