Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Surgical site, advanced age, and comorbid conditions increase risk for postoperative pulmonary complications


ACP J Club. 2006 Sep-Oct;145:37. doi:10.7326/ACPJC-2006-145-2-037

Related Content in this Issue
• Companion Abstract and Commentary: Review: Lung expansion modalities and selective nasogastric decompression prevent postoperative pulmonary complications

Clinical Impact Ratings

Hospitalists: 6 stars

Pulmonology: 5 stars

Source Citation

Smetana GW, Lawrence VA, Cornell JE. Preoperative pulmonary risk stratification for noncardiothoracic surgery: systematic review for the American College of Physicians. Ann Intern Med. 2006;144:581-95. [PubMed ID: 16618956]



Which patient- and procedure-related factors and test results predict postoperative pulmonary complications (PPCs) after noncardiothoracic surgery?


Data sources: MEDLINE (1980 to June 2005) and bibliographies of relevant studies.

Study selection and assessment: English-language studies of any design that provided primary data on well-defined PPCs after noncardiothoracic surgery in adults. Excluded were studies with < 25 patients/group; that studied only nonclinical outcomes, identified PPCs by administrative data, or had outcomes unique to certain types of surgery; from developing countries; involving gastric pH manipulation; or involving ambulatory, cardiopulmonary, or organ transplant surgery. Studies that met the selection criteria included 27 studies (26 prospective cohort studies and 1 case–control study) with multivariate analysis (324 648 patients with 10 960 PPCs) and 83 studies of various designs with univariate analysis (173 500 patients with 11 851 PPCs). Quality assessment was done using the U.S. Preventive Services Task Force criteria.

Outcomes: PPCs including atelectasis, pneumonia, and respiratory failure.

Main results

Pooled odds ratios (random-effects model) were derived from the studies with multivariate analysis. Factors with good evidence of an association with PPCs included the patient-related factors advanced age, comorbid conditions, heart failure, chronic obstructive pulmonary disease (COPD), and functional dependence; the procedure-related factors surgical site, prolonged surgery, emergency surgery, and use of general anesthesia; and the laboratory test serum albumin level (Table). Good evidence exists that obesity; well-controlled asthma; and hip, gynecologic, and urologic surgery are not risk factors for PPCs.


Certain surgical sites (especially abdominal aortic aneurysm repair, thoracic surgery, and abdominal surgery) and patient characteristics (such as advanced age and presence of comorbid conditions) are associated with increased risk for postoperative pulmonary complications after noncardiothoracic surgery.

Source of funding: Veterans Evidence-based Research, Dissemination, and Implementation Center (VERDICT).

For correspondence: Dr. G.W. Smetana, Beth Israel Deaconess Medical Center, Boston, MA, USA. E-mail

Table. Factors associated with postoperative pulmonary complications after noncardiothoracic surgery*

Factors Number of studies Odds ratio (95% CI)†
Age 60 to 69 y 7 2.1 (1.7 to 2.6)
Age 70 to 79 y 4 3.0 (2.1 to 4.4)
ASA class ≥ II 6 4.9 (3.3 to 7.1)
ASA class ≥ III 11 2.6 (1.7 to 3.8)
Heart failure 3 2.9 (1.02 to 8.0)
Chronic obstructive pulmonary disease 8 1.8 (1.4 to 2.2)
Functional dependence (partial) 2 1.7 (1.4 to 2.0)
Functional dependence (total) 2 2.5 (2.0 to 3.2)
Abdominal aortic aneurysm repair 2 6.9 (2.7 to 17)
Thoracic surgery 3 4.2 (2.9 to 6.2)
Abdominal surgery 6 3.0 (2.4 to 3.7)
Prolonged surgery 5 2.3 (1.5 to 3.5)
Emergency surgery 6 2.2 (1.6 to 3.1)
General anesthesia 6 1.8 (1.4 to 2.5)
Serum albumin < 30 g/L 1 2.5 (2.3 to 2.8)

*ASA = American Society of Anesthesiologists; CI defined in Glossary.
†Trim-and-fill method used to adjust for publication bias if ≥ 3 studies pooled.


The systematic reviews by Smetana and Lawrence and their colleagues are important contributions to the field of perioperative care and serve as the basis for a new practice guideline by the American College of Physicians (ACP) (1). Both reviews used well-designed search strategies and inclusion and exclusion criteria, and standardized level-of-evidence ratings for their recommendations. The authors are to be commended for excluding studies that reported only physiologic endpoints, as these are often of unclear clinical importance.

The review by Smetana and colleagues serves as a wake-up call, reminding clinicians that PPCs are an important cause of morbidity and mortality and are at least as common as postoperative cardiac complications. Finding new ways of assessing and reducing perioperative risk is especially important now: It was recently found that, for all but the highest-risk patients, administration of perioperative β-blockers is less beneficial, and more clearly harmful, than previously thought (2).

The authors found a crude PPC rate of 3.4% across 27 cohort studies. Use of the Bayes' nomogram allows clinicians to estimate an individual patient's risk for PPCs using the odds ratios (ORs) provided (3). For example, assuming the risks are multiplicative, a 65-year-old man (OR 2.1) with a serum albumin < 30 g/L (OR 2.5) having abdominal surgery (OR 3.0) has a combined OR for PPCs of 15.75 (2.1 × 2.5 × 3.0). Using 3.4% as an estimate of baseline risk, a straight-edge nomogram calculation reveals that this patient has > 25% risk for PPCs.

Patient-specific estimates of risk, and use of the new ACP guideline, will help identify those who will benefit most from techniques shown to prevent PPCs, such as lung expansion modalities and selective nasogastric decompression. Although no clear difference among lung expansion modalities is evident, CPAP should be reserved for patients who cannot participate in less-intensive and cheaper methods, such as incentive spirometry, deep breathing exercises, and chest physiotherapy.

Routine preoperative chest radiography and spirometry are not recommended, adding little value in patients with an unremarkable history and physical examination. However, because interval radiographic changes can be important in the early diagnosis of pneumonia, I would consider obtaining a preoperative chest film in patients at high risk for PPCs. Smetana and colleagues point out that, in patients with dyspnea of unclear cause or in select patients with COPD or asthma, preoperative spirometry may be reasonable to determine if airflow obstruction has been maximally treated.

Insufficient evidence exists to answer many important questions. As these debates continue, it is important to remember that lack of evidence is not evidence of lack of effect. Further studies are needed to define the role of spirometry, obstructive sleep apnea, and poor exercise capacity in preoperative pulmonary risk stratification. Also still unsettled are whether preoperative smoking cessation reduces (or increases) risk for PPCs, whether laparoscopic surgery causes fewer PPCs than open surgery, and whether epidural anesthesia and analgesia reduce PPCs. RCTs of adequate power to detect important differences in clinical outcomes are needed.

Daniel I. Steinberg, MD
Hospital of the University of Pennsylvania
Philadelphia, Pennsylvania, USA


1. Qaseem A, Snow V, Fitterman N, et al. Risk assessment for and strategies to reduce perioperative pulmonary complications for patients undergoing noncardiothoracic surgery: a guideline from the American College of Physicians. Ann Intern Med. 2006;144:575-80. [PubMed ID: 16618955]

2. Devereaux PJ, Beattie WS, Choi PT, et al. How strong is the evidence for the use of perioperative beta blockers in non-cardiac surgery? Systematic review and meta-analysis of randomised controlled trials. BMJ. 2005;331:313-21. [PubMed ID: 15996966]

3. Page J, Attia J.Using Bayes' nomogram to help interpret odds ratios [Editorial]. ACP J Club. 2003;139(2): A11-2. [PubMed ID: 12954046]