Pulmonary Artery Catheterization: Indications

The practice guidelines and consensus statements that have been generated over the past several years best summarize the current indications for PAC monitoring.^{[318] [559] [584] [585] [586]} The most recent of these practice guidelines is the American Society of Anesthesiologists practice guideline published in 2003.^[318] The guideline task force considered PAC monitoring to be appropriate or necessary, or both, in selected surgical patients undergoing procedures associated with a high risk of complications from hemodynamic changes (e.g. cardiac surgery) or in those with advanced cardiopulmonary diseases who would be at increased risk for adverse perioperative events because of their preoperative medical condition. Furthermore, the characteristics of the practice setting should be considered, including the proficiency and experience of clinicians who would be using perioperative PAC monitoring. In other words, three interdependent variables should help a physician determine when PAC monitoring is indicated: the patient, the surgical procedure, and the practice setting.

Many high-risk patients undergo low-risk operations that have little physiologic impact, and these individuals should not receive PAC monitoring. One example would be a patient with advanced ischemic cardiomyopathy who needs lower extremity amputation that could be performed with a regional anesthesia technique. In contrast, a relatively low-risk patient undergoing a very high-risk procedure may benefit from perioperative PAC monitoring. An example might be a patient with stable ischemic heart disease scheduled to undergo extensive abdominal cancer surgery. Finally, the third factor that determines whether PAC monitoring is appropriate relates to the individual practice setting.^[318] Keats has termed this feature "the role of environment in the outcome of operation," by which he meant all the important unmeasurable aspects of each clinical setting, including among other things, surgical skills and experience.^[587] Clearly, all physicians and nurses using the PAC must have the requisite knowledge and skills to use it safely and effectively. However, there is something that at the present time still defies objective assessment, and a short postgraduate educational program alone seems to be ineffective in improving PAC interpretation skills.^[418]

To provide better scientific evidence to define appropriate indications for PAC monitoring, well-designed randomized clinical trials are required.^{[318] [559] [588] [589] [590] [591] [592] [593]} Such trials remain difficult to perform because of problems in gaining informed consent from critically ill patients, difficulty ensuring that physicians remain committed to trial guidelines, and obstacles encountered in patient selection bias and crossover from non-PAC monitoring to PAC monitoring. ^{[549] [588] [589] [594]} The National Institutes of Health has recently undertaken large multicenter trials focused on PAC use in patients with acute respiratory distress syndrome and severe congestive heart failure.^{[321] [595]} Although these trials will help answer clinically important questions in these patient cohorts, they will have limited applicability for perioperative patients.

In 2003, Sandham and coauthors published the largest prospective randomized clinical trial to date focusing on the use of PACs in high-risk elderly surgical patients.^{[324] [596]} Nearly 2000 patients were randomized to receive goal-directed therapy guided by a PAC or standard care without a PAC. PAC monitoring did not lead to any benefit in terms of morbidity or mortality, and the PAC-monitored group had a higher rate of pulmonary embolism. The authors found that it was difficult to achieve the "supra-normal" physiologic goals desired in the PAC-treated group, a finding that was consistent with other trials. Perhaps of greatest importance, the authors demonstrated the ability to perform a well-controlled randomized clinical trial, considered to be a significant milestone in our search to determine the appropriate role of PAC monitoring in clinical practice.^[596]