Preoperative Testing

Medical consultants are generally asked to assess preoperative risk in most patients who are to undergo surgery. The focus of discussion in this article applies to healthy people who are to undergo an elective surgical procedure.

Despite a low risk of perioperative complications, the use of laboratory tests before surgery became ingrained in clinical practice not only across the United States but also across the world in the latter half of the 20th century. At that time, clinicians thought it logical to order tests to detect abnormalities that might lead to increased morbidity or mortality in the perioperative period. Despite its widespread use, however, systematic evaluations of the clinical effectiveness and cost-effectiveness of routine laboratory testing were often lacking.

In the early and mid 1980s, several investigators published a number of papers demonstrating that routine preoperative testing was not cost-effective and did not benefit the patient. For example, in the mid 1980s, Kaplan and colleagues, in a retrospective review of the charts of 2000 patients who underwent elective surgery, demonstrated that 60% of these patients had laboratory tests ordered for no apparent reason, and that only 0.22% of the abnormal results influenced preoperative management. In another study, Turnbull and colleagues reviewed the charts of 2570 patients undergoing elective surgery, finding that only 104 of 5003 laboratory test results were abnormal and significant, and that only 4 patients would have benefited from "routine" laboratory testing.

To compound the problem, it appears that physicians are poor at evaluating the tests ordered. For example, in a study in which the records of 3782 elective surgery patients were reviewed, only 10 of 160 patients with abnormal test results were treated for such abnormalities. The lack of treatment of identified abnormalities therefore raises the issue of increased legal liability.

In the last 20 years, a progressive challenge to the use of gratuitous routine laboratory testing has developed, especially within the environment of cost-containment and managed care. What, then, should physicians do? A good history and physical examination followed by a review of a patient's chart are undoubtedly the most important routine tests needed.

For example, Narr and colleagues reviewed the charts of 1044 healthy patients who did not undergo any preoperative laboratory testing before their elective surgeries. These patients did not experience any significant perioperative morbidity or mortality. The use of previous laboratory results, performed within 4 months before elective surgery, was supported by a study in which 7549 laboratory results of 1109 patients were reviewed. This study showed that 47% of the laboratory test results duplicated those obtained within 1 year. Of the 3096 normal laboratory test results, only 13 (0.4%) repeated values were abnormal, most of which could have been predicted on the basis of patient history and physical findings.

Furthermore, 5% of healthy people have abnormal test results. This is due to arbitrary cut points that define the range of normal laboratory values to include 2 standard deviations with a 95% confidence interval. For example, the chance that the results of 1 of the 6 tests included in a basic metabolic profile will be abnormal is 26%; hence, the predictive value of the test will be low, especially if the prevalence of the disease is low. For example, based on the Bayes theorem, the positive predictive value of an abnormal hemoglobin test finding is 16.1%, since the prevalence of anemia in healthy individuals is approximately 1%. Accordingly, such abnormal laboratory values, with very low predictive values, may result in further unnecessary workup and delays in surgery.

In a review of studies of routine preoperative testing by Smetana et al (2003), the positive likelihood ratio was modest (>3) for hemoglobin, electrolytes, and renal dysfunction but had a low impact for change on preoperative management. Normal test results did not reduce the likelihood of postoperative complications. In a recent single center study, the incidence of unindicated preoperative screening tests was found to be more than 50%, but it did not add to any benefit to support this persistence of unwarranted testing.

Advancing age, especially older than 70 years, is associated with increased hospital stay and perioperative morbidity and mortality. However, most people in this age group have comorbid conditions, and it remains unclear if complications are secondary to comorbid conditions or age itself.

Contrary to the common belief, obesity does not increase postoperative complications. In a prospective cohort of 6336 patients undergoing general elective surgery, Dindo et al (2003) did not find obesity to be a risk factor for the development of postoperative complications.

For excellent patient education resources, visit eMedicine's Procedures Center. Also, see eMedicine's patient education article Understanding X-rays.

Selective Workup

Laboratory studies:

• CBC count
  • Several studies reported a wide range of hemoglobin abnormality among elective surgery patients, based on different study populations. In healthy individuals undergoing elective surgery, the variation is estimated to be less than 1%. A mild hemoglobin abnormality was not associated with an increase of perioperative morbidity or mortality. Recent guidelines recommend preoperative hemoglobin testing if the history is suggestive of underlying anemia or if a significant blood loss is anticipated during the operation.
  • The prevalence of severe leukopenia or leukocytosis is extremely low and rarely leads to a change in patient management. Similarly, thrombocytopenia is found in fewer than 1% of healthy elective surgery patients; thus, routine preoperative WBC or platelet count is not recommended unless the cost of a CBC count is not substantially higher than that of hemoglobin. The cost of falsely pursuing an abnormal WBC or platelet count may not be substantial, although no studies in that regard are available except for platelets.
• Electrolytes
  • Unanticipated electrolyte abnormality (sodium, potassium, bicarbonate, chloride) ranges from 0.2-8.0% among surgery patients. A recent systemic literature review reported that unsuspected electrolyte abnormality is 1.4% among healthy elective surgery patients.
  • Although hypokalemia is considered a minor risk factor for perioperative cardiac complications based on the Goldman risk index, no study showed a relation between hypokalemia and perioperative morbidity and mortality.
  • Postoperative hyponatremia is common in certain types of surgeries, such as transurethral resection of prostate and neurosurgical procedures; however, it is still unclear how baseline electrolyte abnormality may affect physicians' decisions in postoperative management. Accordingly, electrolyte determination is not routinely recommended for elective surgery in healthy individuals.
• Creatinine
  • The prevalence of elevated creatinine levels in asymptomatic patients ranges from 0.2-2.4% and increases with age. Approximately 9.8% of patients aged 46-60 years have elevated creatinine levels.
  • Patients with mild-to-moderate renal insufficiency are usually asymptomatic but have an increased risk of perioperative morbidity and mortality. Accordingly, testing renal function with serum creatinine level is recommended for all patients older than 40 years, especially if hypotension or use of nephrotoxic medications is anticipated.
• Blood sugar
  • The frequency of abnormal glucose laboratory results in asymptomatic patients ranges from 1.8-5.5%. The frequency increases with age, so that nearly 25% of patients older than 60 years have a fasting blood sugar level above 120 mg/dL.
  • Only in certain operations, such as vascular surgery and coronary artery bypass grafting, was diabetes associated with higher perioperative risks; hence, routine blood sugar determination is not recommended unless the patient has high risk for diabetes (eg, obesity, steroids, strong family history) or will be undergoing vascular or bypass surgery.
• Liver enzymes
  • The frequency of a hepatic aminotransferase enzyme (aspartate aminotransferase [AST], alanine aminotransferase [ALT]) abnormality is estimated to be approximately 0.3%. Although Powell-Jackson and colleagues showed that severe liver test abnormalities may lead to an increase in surgical morbidity and mortality risk, no evidence confirms that mild elevation in liver enzymes is associated with such an increased risk.
  • Because most patients with severe aminotransferase enzyme elevation are likely to be symptomatic, and jaundice may be detected by physical examination, routine preoperative screening is not recommended for healthy individuals.
• Hemostasis
  • In the absence of a history of bleeding diathesis in elective surgery patients, abnormal bleeding time, prothrombin time (PT), and activated partial thromboplastin time (aPTT) results are estimated to be less than 1%.
  • Suchman and colleagues showed that in low-risk patients, per history and physical examination, aPTT does not predict the risk of perioperative bleeding. Similarly, the bleeding time has no predictive value on the incidence of perioperative bleeding in healthy elective surgery patients. Accordingly, PT, aPTT, and bleeding time are not recommended for routine screening.
• Urinalysis
  • The primary rationale for ordering urinalysis (UA) preoperatively is to detect either asymptomatic renal disease or underlying urinary tract infection (UTI). To detect unsuspected renal insufficiency, serum creatinine measurement is recommended for any elective surgery patient older than 40 years, although it is unclear if any correlation exists between asymptomatic UTI and surgical wound infection.
  • One study that included 200 patients undergoing orthopedic procedures showed that physicians addressed only 5 of 27 abnormal urine test results. A further economic analysis showed that in order to prevent a single wound infection, approximately $1.5 million must be spent on UA; therefore, UA is not recommended routinely for asymptomatic patients.
• Fecal occult blood
  • The prevalence of positive fecal occult blood findings among healthy individuals undergoing elective surgery is unknown. In addition, the benefits of routine screening are unclear.
  • A decision-analysis study showed no benefit of routine screening; therefore, insufficient evidence exists to support routine screening for fecal occult blood.

Imaging studies:

• ECG
  • The prevalence of abnormal ECG findings among healthy elective surgery patients ranges from 14-53% and increases with age in a continuous fashion.
  • The rationale for obtaining ECG preoperatively is to identify high-risk patients with prior myocardial infarction or arrhythmia. Detecting a silent myocardial infarction is of main clinical benefit because numerous investigators showed an association between preoperative myocardial infarction and surgical mortality. One study showed that 25% of 708 myocardial infarctions in the Framingham study were detected by ECG. In addition, any rhythm other than sinus, including frequent premature ventricular contraction, is associated with an increase in surgical risk. Accordingly, routine ECG is recommended for all patients older than 40 years undergoing elective surgery.
  • In a recent retrospective study of 23,036 patients who underwent 28,457 surgical procedures, multivariate logistic regression was used to evaluate the relationship between ECG abnormalities and cardiovascular death. A total of 199 in-hospital cardiovascular deaths (0.7%) occurred. A higher incidence of cardiovascular death was observed in patients with an abnormal ECG than in those with normal ECG results (1.8% vs 0.3%; adjusted odds ratio 4.5, 95% confidence interval 3.3-6.0). However, there was no significant difference (0.5%) in the incidence of cardiovascular death in patients, with or without ECG abnormality, who underwent low-risk or low- to intermediate-risk surgery.
  • Noordzij et al (2006) concluded that preoperative ECG provides prognostic information in addition to clinical characteristics and the type of surgery. However, the usefulness of routine ECG testing in lower risk surgery is questionable.
• Chest radiograph
  • The frequency of abnormal chest radiograph (CXR) findings increases with age. One study showed that 0.3% of patients younger than 60 years had unsuspected abnormal CXR results or clinical findings suggestive of underlying cardiac or pulmonary disease compared to 22% of patients older than 60 years.
  • In addition, Goldman reported that CXR in patients with congestive heart failure does not independently add to the risk of perioperative mortality and morbidity. A meta-analysis of 21 studies that included 14,390 routine CXR showed that only 140 of 1444 abnormal results were not clinically expected and that only 14 affected physicians' decisions in managing their patients. Accordingly, routine CXR is recommended only for patients older than 60 years unless underlying heart or lung disease is a possibility.

Summary

Routine preoperative screening of healthy people undergoing elective surgery is not recommended. Instead, a selective strategy, as outlined above, is safe and cost-effective as long as a complete history and physical examination are obtained. Based on the available evidence, the authors recommend the following preoperative tests:

1. Hemoglobin level for major surgery with significant expected blood loss or CBC count if the cost is not substantially increased
2. Serum creatinine level for people older than 40 years
3. ECG in patients older than 40 years
4. CXR in patients older than 60 years

No laboratory test must be repeated if results were normal within 4 months of the surgery and no change in the patient's clinical status occurred. Finally, this strategy applies only to healthy, asymptomatic patients undergoing elective surgery. Patients with suspected pulmonary or cardiac disease or those undergoing urgent operation require additional evaluation that is beyond the scope of this article.