Risk Factor	Level of Prediction
Systemic Disease
Asthma (well controlled)	Poor
Congestive heart failure	Good
Chronic obstructive pulmonary disease	Good
Diabetes mellitus	Fair
Human immunodeficiency virus infection	I
Obstructive sleep apnea	Good
Pulmonary hypertension	Fair
Signs and Symptoms
Abnormal chest auscultation	Good
Arrhythmia	I
Functionally dependent	Good
Impaired mental status	Good
Poor exercise capacity	Fair
Weight loss	Good
Patient Features
Alcohol user	Good
Cigarette smoker	Good
Corticosteroid use	I
Objective Testing
Albumin <35 g/L	Good
Blood urea nitrogen >25 mg/dL	Good
Chest radiography, abnormal finding	Good
Spirometry	I
Surgical Site
Aortic aneurysm	Good
Abdominal	Good
Neurosurgery	Good
Vascular	Good
Hip	Poor
Gynecologic or urologic	Poor
Esophageal	I
Surgical Factors
Prolonged surgical duration	Good
Emergent surgery	Good
General anesthesia	Fair
Perioperative transfusion	Good

Good, ACP “A” or “B” rating or other data to support; Fair, ACP “C” rating or other data to support; Poor, ACP “D” rating, or is not a risk factor; I, indeterminate (insufficient data to support a rating.

Data from current American Academy of Chest Physicians (AACP) guidelines; Smetana GW et al: Cleve Clin J Med 73(Suppl 1):36–41, 2006; and other sources (see Suggested Readings).

Patient-Related Factors

In several studies, chronic obstructive pulmonary disease (COPD) has been found to double the risk of PPCs. The degree of risk is directly related to the severity of obstruction. One study of patients after upper abdominal surgery found that those with decreased breath sounds or other adventitious sounds (e.g., wheezing, rales) had a sixfold increased rate of PPCs.

In contrast, well-controlled asthma has not been found to be a risk factor for PPCs. A large retrospective analysis found that rates of bronchospasm, laryngospasm, and respiratory failure in patients with well-controlled asthma were comparable to healthy individuals. Those with poorly controlled asthma (e.g., more frequent albuterol use, recent emergency department visit for asthma) do have an increased risk for PPCs.

Pulmonary hypertension (PH), primary and secondary, may be associated with increased rates of PPCs and in-hospital mortality. In one study, 21% of patients with PH developed respiratory failure versus 3% in matched controls after noncardiac surgery.

A patient’s general health may be assessed by using the American Society of Anesthesiologists (ASA) classification system (Table 73-3). Studies have shown a direct correlation between a higher ASA class and increased pulmonary complications. Patients with an ASA Class III or higher have a twofold to threefold increase in PPCs compared with those with ASA Class II or lower. Difficulty or inability to perform activities of daily living has also been linked to increased complications.

Table 73-3 ASA Classification for Surgical Candidates

Class	Patient Status
I	Normal healthy patient
II	Patient with mild systemic disease (e.g., mild asthma)
III	Patient with severe systemic disease, but not incapacitated
IV	Patient with life-threatening illness who is incapacitated
V	Moribund patient
VI	“Brain-dead” patient eligible for organ donation

ASA, American Society of Anesthesiologists.

Obesity alone does not seem to increase the risk of PPCs. A wide review of gastric and general surgeries shows that obese and morbidly obese patients have the same risk for PPCs as nonobese patients. However, the presence of a common comorbidity in obese patients, obstructive sleep apnea (OSA), does confer increased risk. In a study of patients after hip or knee replacement, those with OSA had longer hospital stays or more frequent transfers to the intensive care unit (ICU). A more recent study found that those with five or more desaturation episodes during the night had greatly increased rates of PPCs. Therefore, patients should be asked about symptoms of OSA preoperatively.

Age is the most significant patient-related predictor for PPCs. Even after adjustment for comorbid conditions, increasing age is an independent risk factor for increased PPCs. Congestive heart failure (CHF) is also an independent risk factor and should be of particular concern in the elderly population.

Cigarette smokers have a moderately increased risk for PPCs, independent of comorbid lung conditions. Smokers have higher rates of wound infections, oxygen desaturation, laryngospasm, and increased coughing with anesthesia, all of which can lead to longer hospital stays and contribute to pulmonary complications. Quitting smoking before surgery lowers the rate of PPCs. Although early data suggested that quitting smoking close to surgery may increase the rate of PPCs, a recent meta-analysis shows this is not true. In general, the sooner a patient can quit smoking, the better it is for the patient, but smoking cessation should be encouraged at any time before surgery.

Surgery-Related Factors

The planned surgery and surgical site have an impact on risk for developing PPCs that is often greater than patient-related factors. Surgical site was the most important predictor for developing postoperative pneumonia and respiratory failure in a cohort of 160,000 U.S. military veterans undergoing major noncardiac surgery. Those undergoing abdominal aortic aneurysm (AAA) repair or thoracic surgery had the highest risk, followed by upper abdominal, neurovascular, and neck surgeries.

Duration of surgery, extent of anesthesia, and invasiveness of the surgery all influence the risk of developing PPCs. Patients undergoing emergency surgeries or surgeries lasting longer than 3 hours are at greater risk for PPCs. Surgeries requiring general anesthesia (GA) carry a greater risk for PPCs than with peripheral nerve blocks. There is questionable evidence about the relative risk of PPCs from surgeries requiring GA versus neuraxial blockade (spinal or epidural anesthesia). A previous large review of various surgeries reported increased rates of pneumonia, respiratory failure, and mortality in patients who received GA compared to those receiving neuraxial blockade. A more recent small review of orthopedic surgeries found equal rates of pneumonias in these two groups.

A study of patients undergoing AAA repair found increased rates of PPCs (16%) in open repair compared to endovascular repair (3%). Video-assisted surgeries have reduced PPC rates compared with open approaches. Limited data suggest that patients undergoing cardiothoracic robotic surgery have lower rates of PPCs than those not receiving robotic surgery.

Surgery in General

The most important part of a preoperative pulmonary evaluation is taking a complete history and performing a complete physical examination. For those without pulmonary disease or symptoms, or those with well-controlled chronic lung conditions, routine pulmonary function tests (PFTs) and chest imaging are not indicated.

During the office visit, symptoms or signs of an underlying or undiagnosed pulmonary condition may be discovered. If unexplained dyspnea, coughing, wheezing, other abnormal lung sounds, or chest pain is revealed, testing before an elective surgery may be indicated to identify their cause. Historical clues should aid in distinguishing cardiovascular dyspnea from pulmonary dyspnea and guide which tests should be performed. Pulmonary function testing will help to identify and quantitate physiologic impairment of lung function. If a cardiac etiology is suspected, an electrocardiogram, echocardiogram, brain natriuretic peptide level, and stress test may be performed. Other cardiac risk assessment is beyond the scope of this chapter.

Further testing, such as a chest radiograph, complete blood count, serum chemistries, and thyroid-stimulating hormone (TSH) level can help to determine the cause of cardiopulmonary symptoms and the need for additional testing. If the cause of dyspnea remains unclear, a cardiopulmonary exercise test may help determine its nature and quantify its severity. Additional chest imaging may also be required if a pulmonary cause of a concerning symptom is likely. Screening for PH is not recommended in the preoperative setting, particularly if respiratory symptoms are not present or are otherwise well explained. Since most, if not all, of the conditions causing dyspnea result in increased rates of PPCs, proper diagnosis and treatment optimization of that condition will improve patient outcomes.

During the history and physical examination, patients should be screened for OSA by asking about snoring, witnessed apneas, and daytime somnolence. Neck circumference and obesity should be noted. Validated questionnaires are available for identifying those at risk for OSA (Box 73-1). Evidence is limited on the usefulness of diagnosing OSA preoperatively with polysomnography, with decisions based on degree of clinical concern and urgency of the surgery.

Validated indices are available for predicting postoperative pneumonia and respiratory failure after major noncardiac surgery. The investigators from three select reviews assigned point values to corresponding risk factors, as summarized in Tables 73-4 and 73-5. Clinicians may find these tools helpful in identifying those at risk for postoperative pneumonia and respiratory failure and for selecting perioperative testing and respiratory care for those at high risk.

Table 73-4 Risk for Postoperative Respiratory Failure and Pneumonia

Table 73-5 Level of Risk Based on Point Value/Score

Lung Resection Surgery

For patients with non–small cell lung cancer (NSCLC), surgical resection with a lobectomy or pneumonectomy offers the best chance at cure. Approximately 20% of lung cancer patients are deemed to be surgical candidates. However, these resections are not without serious risk. The mortality for lobectomy is 2% to 3%, and for pneumonectomy, 4% to 6%. Overall, PPCs range from 20% to 30%. Meticulous patient selection by a multidisciplinary team for these procedures is imperative. The clinician must identify patients who have a poor prognosis without surgery, a low risk of operative mortality, and a high likelihood of tolerating well the reduced lung function after resection. Lung resection surgery is unique in that the reduction in lung function related to surgery does not occur only in the immediate postoperative period but rather persists. Thus, preoperative evaluation must consider not only PPCs but also the long-term effect of the loss of lung function on the patient’s quality of life. The following general overview recognizes that decisions about lung resection need to be individualized, considering other treatment options, patient preferences, and comorbidities.

After a complete history and physical examination, the lung resection patient should undergo a thorough pulmonary assessment focusing on pulmonary reserve. All potential lung cancer resection patients should have spirometry performed. Patients with forced expiratory volume in 1 second (FEV₁) measurements above 2 L and 80% of predicted normal values, who are free of cardiopulmonary symptoms, should tolerate resection without the need for additional testing. For all others, a diffusion capacity (DLCO) measurement should be obtained, and predicted postoperative (ppo) lung function should be calculated.

Estimating postoperative lung function can help determine the risks related to lung resection surgery and identify patients likely to tolerate resection who may not qualify by other measures. An anatomic segment method or radionuclide quantitative perfusion imaging may be used for patients undergoing surgery up to a lobectomy; the latter is preferred for those likely to require a pneumonectomy.

To calculate ppo lung function with the anatomic segment method, the fraction of total lung segments that will remain after the resection is multiplied by the preoperative lung function value. The two equations typically used are based on pulmonary segment total (19) or subsegment total (42). Using the former, the predictions correlate well with a lobectomy but less so with a pneumonectomy. Overall, these calculations tend to underestimate actual postoperative lung function, particularly in those with baseline impairment of lung function.

Radionuclide quantitative perfusion studies estimate the relative function of the lung to be resected by quantifying perfusion to the portion of lung to be resected as a fraction of total lung perfusion. Postoperative lung function can be calculated similar to the anatomic segment methods. Radionuclide studies also consistently underestimate FEV₁ values in patients with lower baseline values. Although more accurate than segment methods for postpneumonectomy lung function prediction, the overall predictive value of quantitative perfusion studies is less than ideal.

Pulmonary Function Measures

Although clinicians cannot rely solely on pulmonary function measures to predict the risk of PPCs, perioperative mortality, or unacceptable levels of long-term dyspnea, the FEV₁ value does correlate with these outcomes. As a general rule, a preoperative FEV₁ value of 2 L or greater indicates that a patient without cardiopulmonary symptoms will tolerate a pneumonectomy, and with a value of 1.5 L or greater, a lobectomy. Because absolute values do not consider age, gender, or size, percent of predicted normal values are more useful. Values greater than 80% of predicted normal in those without symptoms support proceeding directly to resection without additional testing. As stated earlier, ppo values as a percentage of normal are also helpful in identifying patients capable of undergoing resection who might have otherwise been rejected based on absolute FEV₁ values.

Guidelines recommend that patients with FEV₁ values below the general parameters just listed and any symptomatic patient should have ppoFEV₁ values calculated. It is difficult and probably inappropriate to apply cutoff values for ppoFEV₁ to an individual patient. As expected, the lower the ppoFEV₁ value, the higher is the rate of complications or unacceptable outcomes. One study noted that no postoperative deaths occurred in patients with ppoFEV₁ greater than 40% after resection, whereas those with ppoFEV₁ less than 40% had a 50% chance of postoperative mortality. Another study found that all patients with a ppoFEV₁ of 30% or lower developed respiratory failure or died after resection.

The DLCO and ppoDLCO are also independent predictors for PPCs and mortality. The majority of patients undergoing lung resection should have DLCO measured. DLCO values may be low despite normal spirometry. As with FEV₁, one cutoff value for resection should not be applied equally to all patients. Lower values correlate with worse outcomes. Those with an absolute DLCO of 60% predicted normal or lower had greater rates of postoperative respiratory failure and death in one study; ppoDLCO values of 40% predicted normal or greater have been used as cutoffs in published guidelines. One study found that patients with ppoFEV₁ × ppoDLCO (predicted postoperative product) of less than 1650 have a high mortality rate.

Exercise Testing

When questions remain after standard measures of lung function have been performed and ppo values have been calculated, measures of exercise capacity can be helpful. Cardiopulmonary exercise testing (CPET) with cycle ergometry has been used to assess a patient’s fitness for resection. An individual’s exercise capacity can be measured as the peak oxygen consumption (VO₂ peak). Values greater than 20 ml/kg/min have been used to suggest the ability to tolerate pneumonectomy, 15 ml/kg/min for lobectomy, and less than 10 ml/kg/min for a high-risk group. Many studies have found this value to be an independent predictor for PPCs and mortality. Formal oxygen desaturation studies may be performed as well. A 4% oxygen desaturation is associated with increased PPCs, including respiratory failure, increased ICU stays, and home oxygen requirement. Stair climbing, a 6-minute walk test, and shuttle walking also provide measures of cardiopulmonary capacity that can assist with decisions about resectability.

Box 73-2 summarizes the American College of Chest Physicians (ACCP) recommendations for the preoperative pulmonary evaluation of the lung resection candidate. Figure 73-1 outlines the ACCP and European Respiratory Society and European Society of Thoracic Surgeons (ERS/ESTS) recommended steps to follow when assessing a patient’s pulmonary fitness for lung resection. The differences in these guidelines highlight the lack of consensus in this area and the need to individualize patient decisions.

Box 73-2

Summary of ACCP Guidelines for Lung Cancer Resection

Patients should not be denied lung resection surgery based on age alone.