Tools for rapid preoperative cardiovascular risk assessment

,

Commentary

Tools for rapid preoperative cardiovascular risk assessment

J Fam Pract. 2003 April;52(4):313-320

By

Gary N. Fox, MD

Prediction is very difficult, especially about the future.
—Niels Bohr

PDF Download

References

1. Eagle KA, Berger PB, Calkins H, et al. ACC/AHA guideline update for perioperative cardiovascular evaluation for noncardiac surgery—executive summary: a report of the American College of Cardiology/American Heart Association Task Force onPractice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). J Am Coll Cardiol 2002;39:542-53.

2. Guidelines for assessing and managing the perioperative risk from coronary artery disease associated with major noncardiac surgery-American College of Physicians. Ann Intern Med 1997;127:309-12.Also available at: http://www.acponline.org/journals/annals/15aug97/ppcad1.htm.

3. Detsky AS, Abrams HB, Forbath N, Scott JG, Hilliard JR. Cardiac assessment for patients undergoing noncardiac surgery. A multifactorial clinical risk index. Arch Intern Med 1986;146:2131-4.

4. Mangano DT, Goldman L. Preoperative assessment of patients with known or suspected coronary disease. N Engl J Med 1995;333:1750-6.

5. Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 1999;100:1043-9.

6. Detsky AS, Abrams HB, McLaughlin JR, et al. Predicting cardiac complications in patients undergoing non-cardiac surgery. J Gen Intern Med 1986;1:211-9.

7. Steyerberg EW, Kievit J, de Mol Van Otterloo JC, van Bockel JH, Eijkemans MJ, Habbema JD. Perioperative mortality of elective abdominal aortic aneurysm surgery. A clinical prediction rule based on literature and individual patient data. Arch Intern Med 1995;155:1998-2004.

8. L’Italien GJ, Paul SD, Hendel RC, et al. Development and validation of a Bayesian model for perioperative cardiac risk assessment in a cohort of 1,081 vascular surgical candidates. J Am Coll Cardiol 1996;27:779-86.

9. Goldman L, Caldera DL, Nussbaum SR, et al. Multifactorial index of cardiac risk in noncardiac surgical procedures. N Engl J Med 1977;297:845-50.

10. Wasson JH, Sox HC, Neff RK, Goldman L. Clinical prediction rules. Applications and methodological standards. N Engl J Med 1985;313:793-9.

11. Romero L, de Virgilio C. Preoperative cardiac risk assessment: an updated approach. Arch Surg 2001;136:1370-6.

12. Hollenberg SM. Preoperative cardiac risk assessment. Chest 1999;115(suppl):51S-57S.

13. Paul SD, Eagle KA. A stepwise strategy for coronary risk assessment for noncardiac surgery. Med Clin North Am 1995;79:1241-62.

Stating that a patient is “cleared for surgery” is no longer considered adequate; rather, the patient should be assessed for surgical risk, a major component of which is cardiovascular risk. The risk level helps the surgical team determine which preoperative, intraoperative, and postoperative assessments and therapies are indicated.¹

Clinical decision tools to assist physicians perform preoperative cardiovascular assessment have been developed and validated, and some have been adapted to handheld computers (Table 1).^1-8 Yet few prospective or randomized studies have been conducted to establish the value of various preoperative assessment strategies on outcome.¹ In this report, we review the studies underlying these rules and the corresponding software programs for handheld computers (also known as personal digital assistants, or PDAs). Our goal is to find readily applicable software that assists physicians in performing rapid preoperative cardiovascular assessment of patients in the office or hospital.

Cardiovascular Decision Tools

Some cardiovascular decision tools, such as those developed by the American College of Physicians (ACP) and American College of Cardiology/American Heart Association (ACC/AHA), are algorithmic approaches that make direct recommendations about whether to pursue cardiac testing. These tools are designed to be widely applicable to potential candidates for noncardiac surgery.

Other decision tools provide a risk score or index, which the user must interpret and translate into perioperative recommendations. The physician must also assure that the index is appropriate to the patient being evaluated by considering the original study’s patient selection criteria, the setting in which the rule was validated (eg, referral center), and which outcomes the rule predicts. Table 2provides more detailed information about each rule, as well as an assessment of each rule’s methodologic rigor (adapted from methodologic standards for clinical prediction rules developed by Wasson et al¹⁰).

Limitations

Even when applied correctly, these indices have substantial limits. The positive predictive values (percentage of patients predicted to have complications who actually suffer a complication) are generally low and, especially in the case of higher-risk surgery, a low risk score does not eliminate the risk of complications.²

It is therefore important to consider the overall risk of adverse outcomes for patients undergoing a particular kind of surgery (the “pretest probability”): the same patient planning cataract surgery and vascular surgery has 2 distinctly different risks, despite having the same score on the biomedical components of the index. Also, these decision support tools were developed using information from patients enrolled during past decades, and their results may not be directly applicable to current surgical risks.⁴ Finally, these indices cannot be used on all patients. For example, a patient with a rare or unusual problem such as left atrial myxoma may be at a higher risk than would otherwise be indicated by one of the decision support tools.³

TABLE 1
Program information

Algorithm	Program	Version	Size	Cost	Source
ACC/AHA¹ 2002	STAT Cardiac Clearance	1.1	97 KB	Free	http://www.statcoder.com
Detsky³	MedRules	2.5	324 KB	Free	http://pbrain.hypermart.net/
Detsky⁶	InfoRetriever	4.2 (beta)	2.1 MB	Beta is free;cost of final version unknown	http://www.infopoems.com

TABLE 2
Selected characteristics of studies of perioperative risk assessment

Decision rule	Study population characteristics	Derivation and validation set sample size; years of enrollment	Prospective data collection and reviewer blinding	Major outcomes measured and number of outcomes	Pretest probabilities†
ACC/AHA¹	All noncardiac surgery, major & minor	N/A / N/A Evidencebased guideline	N/A	N/A	N/A
ACP²	All noncardiac surgery, major & minor	N/A / N/A Evidence based consensus guideline	N/A	N/A	N/A
Detsky^3,6	All noncardiac surgery, major & minor, at a large teaching hospital. Patients were 40 years and older, referred by the surgical services for consultation because of a question of cardiac risk or chronic disease.^a	Not specified/ 455; Enrollment years not specified	Blinded,* prospective collection of validation data	Cardiac death, nonfatal myocardial infarction, ventricular tachycardia or fibrillation requiring counter shock, and nonfatal alveolar pulmonary edema;30 outcomes occurred³	Major surgery^‡ Vascular 13.2%/21% Orthopedic 13.6%/18.2% Intrathoracic/intraperitoneal 8.0%/12.6% Head and neck 2.6%/7.8%
Detsky^3,6		Not specified/ 455; Enrollment years not specified	Blinded,* prospective collection of validation data		Minor surgery 1.6%/2.1% (eg, TURP, cataracts)
Mangano/Goldman^4,9	Consecutive unselected patients at a large teaching hospital. Patients were 40 years and older; patients with angina or those undergoing minor surgery were excluded^b	1001/None; 1975–1976	Derivation set data not uniformly prospective; blinding not mentioned	Myocardial infarction,pulmonary edema, ventricular tachycardia; 58 outcomes occurred	5.8%
Lee5	All patients at a large teaching hospital, 50 years and older, with an anticipated surgical length of stay ≥ 2 days^c	2893/1422; 1989–1994	Blinded, prospective collection of derivation and validation set data	Myocardial infarction,pulmonary edema, ventricular fibrillation or cardiac arrest, complete heart block; 36 outcomes occurred	2.5%
Steyerberg⁷	Consecutive patients for primary elective abdominal aortic aneurysm surgery at a university hospital in the Netherlands^d	238/None; 1977–1988	Unclear whether data collection was prospective and whether reviewer was blinded	Surgical mortality; 18 deaths	7.6%
L’Italien⁸	Consecutive vascular surgery patients at 5 teaching hospitals referred to their respective institutions’ nuclear cardiology laboratory for preoperative ipyridamole-thallium testing.^e	567/514; 1988–1991	Derivation data collection was retrospective; validation data collection was prospective. Unclear whether reviewer was blinded	Cardiac death, fatal/nonfatal myocardial infarction, not pulmonary edema or congestive heart failure; 39 outcomes occurred	Surgical type	Training/Validation sets
					Aortic	6% / 6%
					Infrainguinal	13% / 10%
					Carotid	6% / 6%
					Total	8% / 8%
All studies described the mathematical model used in constructing the decision rule. None of the studies reported measuring the clinical effect of applying the decision rule. It is evident most rules are based on studies with methodologic concerns and relatively small numbers of target outcomes. The small numbers of outcome events in any 1 subgroup, group, or study means that a difference of 1 or 2 more or fewer outcomes could substantially change the reported results, especially the pretest probability.
*Blinding of postoperative evaluators to preoperative information and classification.
^†Overall rate in study of target outcomes when more specific data are not available.
^‡First percentage: Rate of cardiac death, myocardial infarction, pulmonary edema; second percentage additionally includes worsened coronary insufficiency and congestive heart failure without pulmonary edema.
a. "Not a sample of consecutive patients undergoing surgery, and, therefore, our pretest probabilities will be higher than those that would be found in [a consecutive series]."
b. Because 1977 data excluded minor surgery, unclear how 1995 revision was constructed to include patients undergoing minor procedures. Subject to ascertainment bias because all patients were not studied postoperatively for silent myocardial infarction and other complications.
c. Validation set performed poorly for abdominal aortic aneurysm surgery; authors speculated low number of patients (~100) in derivation set responsible.
d. Patient sample from Netherlands. 238 patients represents "core" because authors used "unique" methods ("We used a new statistical method to quantify the combined effect...") that may be subject to question to "expand" their sample size beyond the actual number of patients studied.
e. Applicable to vascular surgery candidates who require preoperative nuclear stress tests.
N/A, not applicable, none or not specified, not available; TURP, transurethral prostatectomy.