Portal de Periódicos da CAPES

Pulmonary Function After Cardiac and Thoracic Surgery

1999; Lippincott Williams & Wilkins; Volume: 88; Issue: 6 Linguagem: Inglês

10.1213/00000539-199906000-00014

1526-7598

Charles Weissman,

Anesthesia and Pain Management

Perioperative care is undergoing significant change due to clinical advances and economic pressures. As more emphasis is placed on cost-containment and improved efficiency, current practices should be reviewed and critiqued to determine whether they are actually effective. Common problems, such as postoperative pulmonary complications, should be a focus of these efforts because the "bottom line" is to devise methods to reduce these complications and so decrease the costs of medical care. This requires a multifaceted approach that entails a better understanding of the causes of the complications, devising effective treatments and preventive strategies, and performing outcome studies to determine the success of currently used and new procedures. In this review, I examine the pulmonary changes after surgery involving the thoracic cavity and the consequences of modern pain management techniques. Perioperative Respiratory Function: A Brief Overview Surgery and anesthesia alter ventilatory function beginning with the induction of anesthesia and often lasting well into the convalescent period. The most frequent problem after upper abdominal and thoracic surgery is atelectasis, which reduces lung compliance and functional residual capacity. The natural history of postoperative atelectasis is usually one of spontaneous reinflation. However, failure of atelectatic regions to reinflate may lead to pneumonia. Therefore, much effort is directed at preventing and treating postoperative atelectasis, including incentive spirometry, deep breathing exercises, chest physical therapy, and, most importantly, early mobilization. Preoperative Pulmonary Function Testing Identifying patients at increased risk of developing postoperative pulmonary complications should theoretically aid in overall risk assessment and provide an opportunity to optimize pulmonary function before surgery. A critical review of 22 studies found no evidence that routine preoperative spirometry was useful or had any predictive value [1]. This led to recommendations that routine preoperative pulmonary function testing is no more useful in identifying patients at increased risk than a thorough history and physical examination [2]. It is recommended that spirometry be reserved for characterizing the type and severity of pulmonary dysfunction among patients with a history of pulmonary symptoms or tobacco use [2]. Such tests may help to determine whether any observed pulmonary dysfunction is reversed by bronchodilators. One study found that preoperative pulmonary function tests fail to completely predict which patients require extended postoperative mechanical ventilation after cardiac surgery [3]. This is not unexpected, because a major determinant of poor pulmonary outcome after cardiac surgery is poor cardiac function. The exception to these recommendations are patients undergoing lung resection, in which spirometry and arterial blood gas tension determinations are essential for predicting postresection lung function [4]. When there is doubt about the predicted amount of residual postresection function, patients may require further assessment with diffusing capacity, right heart catheterization with temporary unilateral pulmonary artery occlusion, cardiopulmonary exercise testing, quantitative computed tomographic (CT) scanning, or radionuclide quantitative ventilation-perfusion scanning [5,6]. Cardiac Surgery Cardiac surgery by its very nature-median sternotomy, cardiopulmonary bypass (CPB), depressed cardiac function, and manipulation of the thoracic contents-alters pulmonary and cardiac mechanics. Therefore, the pulmonary problems observed after such surgery include those secondary to cardiac dysfunction, e.g., pulmonary edema and congestive heart failure; those due to intrinsic pulmonary problems, such as atelectasis and pneumonia; and those resulting from CPB, specifically the "postpump" lung syndrome. Clinical manifestations span the spectrum from fever and productive cough to respiratory failure requiring prolonged mechanical ventilation. The major determinant of poor pulmonary outcome after cardiac surgery is poor cardiac function [3]. This is not unanticipated, because a low cardiac output state directly and indirectly contributes to varying pulmonary problems. A low cardiac output with increased pulmonary capillary wedge pressure results in increased lung water. Depending on its severity, a spectrum of problems from mild congestive heart failure to overt cardiogenic pulmonary edema may occur (Figure 1). Moreover, the low cardiac output state results in muscle fatigue, leading to weak coughing, reduced mobility, and lack of deep breathing. These may contribute to and worsen atelectasis and increase the possibility of pneumonia.Figure 1: Pulmonary edema after cardiac surgery may be cardiogenic-caused by decreased cardiac output, leading to increased intrapulmonary hydrostatic pressure-or noncardiogenic-caused by increased capillary permeability.The incidence of pneumonia after coronary artery bypass surgery ranges from 3% to 16% and from 5% to 7% after valvular surgery [7,8]. However, lesser complications, such as atelectasis and pleural effusions, occur more frequently. In a large series, 63% of patients had atelectasis and/or pleural effusion detected on postoperative chest radiographs [9]. Left lower lobe atelectasis, the most frequently observed radiological abnormality after cardiac surgery, occurs in 73% of patients after internal mammary artery grafting and in 54% when only vein grafts were used [10]. CT scans of patients with normal chest radiography but reduced oxygenation reveal crescent-shaped bilateral densities in the dependent portions of the lungs consistent with atelectasis. The causes of atelectasis are many (Table 1). Discriminant analysis shows an increase in the severity of atelectasis with a larger number of grafts, longer operative and bypass times, violation of the pleural space, lack of a phrenic nerve insulating pad during surface myocardial cooling, and a lower body temperature during bypass [11].Table 1: Causes of Postcardiac Surgery AtelectasisAfter cardiac surgery, there are decreases in forced vital capacity (FVC), expiratory volume in the first second of forced expiration (FEV1), peak expiratory flow rate (PEFR), and maximum voluntary ventilation (MVV). These changes may persist for >3 mo after surgery [17]. When saphenous vein grafts are used exclusively, Functional Residual Capacity (FRC) and FEV1 decrease less than when an internal mammary artery graft is also used [13]. Pleural changes on chest radiographs after internal mammary artery grafting are associated with larger decreases in pulmonary function than when no pleural changes are noted and when only saphenous veins are used. This has been ascribed to the extensive dissection of the interior portion of the anterior chest wall and violation of the pleural cavity. As expected, there is a greater restrictive defect after bilateral than after unilateral mammary artery harvesting [13]. Median sternotomy alters the spinocostal angles, thus reducing the mobility of the ribs [14]. These structural changes, along with incisional pain, may contribute to a breathing pattern characterized by small tidal volumes and increased respiratory rates. The degree of postoperative diaphragmatic dysfunction is unclear, although elevated diaphragms are seen on chest radiographs in >or=to13% of patients after surgery. Clergue et al. [15] speculated that these diaphragmatic elevations might be due not only to atelectasis, but also to diaphragmatic dysfunction caused by either phrenic nerve damage or reflex-mediated decreases in diaphragmatic function. Phrenic nerve injury most often occurs on the left side, and it is almost always caused by irrigating the pericardial space with cold solution for myocardial preservation during CPB. Electrophysiological evidence of nerve injury is seen in 32% of patients when ice slush is used and in 2%-6% when cold saline is used [16]. The prevalence of clinically significant diaphragmatic dysfunction is 2.1% when the heart is cooled without an insulation pad protecting the phrenic nerve and approximately 0.5% without any topical cooling [17]. It is rarely seen when only intracoronary cardioplegia is used [18]. Although it is often short-lived, diaphragmatic dysfunction may interfere with discontinuation of mechanical ventilation. Unfortunately, in some patients, the dysfunction may persist for >or=to6 mo and interfere with activities of daily living. Rarely, phrenic injury may occur during internal mammary artery harvesting because the nerve crosses the path of the artery on the left side and the two run in parallel on the right side [19]. Despite the high rate of radiographic atelectasis, the incidence of clinically significant pulmonary complications after cardiac surgery is relatively low [7]. This is likely due to early mobilization and pain control. This low incidence and low morbidity of pulmonary complications are reflected by the inability of a randomized study to identify any advantage of adding single-handed percussions to early mobilization and deep breathing exercises after valvular surgery [7]. Similarly, routine physical therapy does not benefit patients after elective coronary artery surgery [20]. There are less common causes of postcardiac surgery respiratory failure. Noncardiogenic pulmonary edema has long been of concern because early CPB techniques were associated with postoperative hypoxemia and noncardiogenic pulmonary edema-post-pump lung. Yet, the incidence of this complication has decreased likely because of changes in techniques, such as the introduction of membrane oxygenators. This was demonstrated by a recent study, which revealed that adult respiratory distress syndrome occurred in only 1% (38 of 3848) of patients after coronary artery bypass and valve replacement surgery [21]. Changes in clinical management have altered the pattern of postoperative complications so that supraventricular arrhythmias (incidence 17%-20%), not serious pulmonary problems, are now the leading cause of postcardiac surgery morbidity [22]. There are a number of causes of hypoxemia after cardiac surgery. The incidence of severe hypoxemia (PaO2 75 yr, body mass index >30 kg/m2, mean pulmonary artery pressure >20 mm Hg, reduced stroke volume, decreased serum albumin, a history of cerebrovascular disease, emergency surgery, and extended bypass time [23]. The amount of bilateral dependent atelectasis seen on CT scans correlates with the degree of venous admixture [24]. The hypoxemia is mainly due to increased shunt fraction, although low mixed venous PO2 secondary to reduced cardiac output also is a contributing factor [24]. As expected, patients with left lower lobe atelectasis have lower arterial oxygen tension when placed in the left lateral decubitus position. In a subsequent study, no differences were found between intra- and postoperative intrapulmonary shunt and ventilation-perfusion relationships when patients undergoing coronary bypass and mitral valve replacement were compared. This does not support the hypothesis that patients with mitral valve disease have more intrapulmonary shunting after surgery due to residual lung water from chronically increased preoperative lung water [25]. Zin et al. [26] observed that, before surgery, patients with valvular heart disease had increased lung and respiratory elastances and lung resistance than patients with ischemic disease, probably because of the higher incidence of left ventricular failure in the former group. These differences decreased postoperatively. Long-term respiratory outcome after valve surgery is favorable, as pulmonary function actually improves after surgery. However, after mitral valve surgery, FVC, FEV1, and MVV are all reduced below preoperative levels on discharge from the hospital, and after 3 mo, all increased above preoperative levels, but remained below predicted values [27]. Fast-Tracking The increasing popularity of early postoperative extubation has confirmed the advantages of allowing patients to cough and ambulate soon after surgery. A retrospective matched cohort study demonstrated that patients tracheally extubated early after surgery have significantly less atelectasis than those extubated later. Additionally, on Postoperative Day 5, vital capacity (VC) and FEV1/FVC were increased after early extubation [28]. Fast-tracking is also associated with significantly lower rates of nosocomial pneumonia. In one study, 3.4% of patients 70 yr developed pneumonia [29]. Minimally Invasive Cardiac Surgery Minimally invasive cardiac surgery performed through a limited thoracotomy or thorascopically, at times without CPB, has recently been introduced. In a preliminary report, Chitwood et al. [30] reported no cases of pneumonia after video-assisted mitral valve surgery and three cases after sternotomy. After port-access coronary artery bypass with a left mammary artery graft, 3 of 42 patients had pleural effusions and 2 had left lower lobe atelectasis. No prospective randomized study has compared the incidence of pulmonary complications after minimally invasive cardiac surgery with that after sternotomy. Pain Management After cardiac surgery, patients have significant thoracic pain, and patients who underwent coronary artery bypass also have pain in the area of the venous graft harvest. However, it is not surprising that the pain is less than that after upper abdominal surgery and lateral thoracotomy, because median sternotomy does not entail dividing muscles (Table 2). Once the effects of anesthesia dissipate, IV opioids are used most commonly. Until extubation, this tends to be nurse-administered IV morphine, and after extubation, nurse-administered or patient-controlled analgesia are used. Within a day or two, many patients require only oral analgesics, such as oxycodone plus acetaminophen. Yet, several studies have shown that pain relief is often inadequate, a situation especially problematic when early extubation is planned [31].Table 2: Pain Scores After Thoracic SurgeryThe intrathecal administration of morphine before the induction of anesthesia has been used with some success. However, the dose must be small enough not to interfere with early extubation, while providing adequate pain relief. In a prospective, randomized, double-blinded, placebo-controlled trial, the time from intensive care unit (ICU) arrival to extubation was prolonged among patients who received 10 [micro sign]g/kg subarachnoid morphine [32]. Respiratory depression was observed in 1.9% of those receiving 30 [micro sign]g/kg intrathecal morphine [33]. Swenson et al. [34] used small-dose intrathecal morphine (0.5 mg) combined with 50 [micro sign]g of sufentanil, and 8 of their 10 patients were extubated within 8 h of ICU admission. Some investigators have reported the use of thoracic epidural anesthesia and analgesia despite the issue of intraoperative anticoagulation and postoperative coagulopathy. No neurological complications have been reported, and compared with IV analgesia, extubation was earlier and there were lesser reductions in FEV1 and PEFR on the second and third days after surgery [35]. Despite these reports, spinal and epidural analgesia are not commonly used after cardiac surgery because patients can usually be adequately treated with an IV/oral analgesic regimen. Thoracic Surgery Lobectomy and Pneumonectomy Surgery for the removal of diseased lung parenchyma is usually performed in patients with varying degrees of chronic lung disease due to smoking. Surgery is usually performed via a posterolateral thoracotomy, which is a painful incision associated with marked changes in respiratory function [36]. These changes include significant reductions in maximal expiratory force, VC, FEV1, and PEFR. The decrease in VC exceeds the volume of lung removed during surgery [36,37]. Among patients receiving postoperative epidural opioid analgesia, VC is reduced by 43%-58% [37]. Persistent intrapleural air spaces, pneumothorax, and mechanical compression of lung parenchyma during surgery contribute to atelectasis formation. Additionally, using double-lumen endotracheal tubes with the cessation of ventilation to the operative lung may accentuate atelectasis formation. Breathing patterns characterized by shallow breathing, increased respiratory rate, and a lack of sighs further accentuate atelectasis formation. Karlson et al. [38] observed decreased lung compliance immediately after thoracotomy (while the patient was still anesthetized) that resolved in the weeks after surgery. Lung compliance was also noted to decrease in proportion to the amount of tissue removed, the amount of airway secretions, and the degree of atelectasis. There was an association between postthoracotomy complications and increased work of breathing, the latter likely due to increased elastic work secondary to reduced lung compliance. Respiratory muscle strength (maximal inspiratory and expiratory pressures) is also reduced after thoracotomy and, along with pain, reduces the effectiveness of coughing, which is vital in preventing and reexpanding atelectasis. The reductions in muscle strength are greater in patients older than 70 yr. Atelectasis, direct parenchymal injury, and redistribution of pulmonary blood flow increase shunt fraction (14% to 21%) and ventilation-perfusion mismatch, leading to hypoxemia [39]. The effect of thoracotomy on respiratory patterns and muscle function remains unclear. Reduced motion of the chest wall is predictable because of muscle injury and spasm, as well as pain. However, posterolateral thoracotomy does not change the relative contributions of the chest wall and abdomen to tidal volume [40]. The lack of an effect on chest wall motion is not immediately apparent but could be due to pain relief restoring a normal respiratory pattern, the fact that most thoracotomy incisions are posterior and splinted by the scapula, or diaphragmatic dysfunction caused by supradiaphragmatic stimulation necessitating compensatory chest wall motion. Maeda et al. [41] observed an increase in respiratory rate, but not tidal volume, after thoracotomy. During quiet breathing, there were suggestions of preserved diaphragmatic function but evidence of increased intercostal/accessory muscle function likely due to expiratory activity of the abdominal muscles or resistance to lung inflation. At maximal inspiration, there were suggestions of reduced maximal diaphragmatic strength. Similarly, after pneumonectomy, transdiaphragmatic pressure seems to be preserved [42]. Using sonomicrometry crystals in sheep, Torres et al. [43] found that diaphragmatic shortening was depressed for at least 4 wk after thoracotomy. Human studies showed decreased shortening of the costal portion of the diaphragm, which is unaffected by lidocaine epidural anesthesia. However, the epidural medication is associated with greater tidal volumes, VCs, and transdiaphragmatic pressures [44]. It seems that surgery above the diaphragm reduces diaphragmatic function, although not to the same degree as upper abdominal surgery. Pulmonary complications are still a major cause of morbidity after thoracotomy. In a prospective observational study, Amar et al. [45] reported that respiratory failure developed in 3 of 47 postlobectomy and 3 of 39 postpneumonectomy patients. Pneumonia was observed in three lobectomies and no pneumonectomies. In a report on 7099 thoracic procedures performed in Japan, there was a 1.3% 30-day mortality rate (3.2% after pneumonectomies, 1.2% after lobectomies). Forty-eight deaths were due to pneumonia and respiratory failure, and five were due to bronchopleural fistulae and empyema [46]. Alternative Incisions The traditional posterolateral thoracotomy involves a long muscle-splitting incision that results in pain and muscle stiffness after surgery. Surgeons have developed approaches, such as limited thoracotomies, axillary thoracotomies, and muscle-sparing incisions to decrease this discomfort. Limited thoracotomies cause less reduction in respiratory muscle strength and FVC than standard thoracotomies. Using a non-seratus-sparing anterioaxillary thoracotomy with disconnection of the anterior rib cartilage causes less of a reduction in FEV1 and VC 1 wk after surgery than posterolateral thoracotomy [47]. Despite the differences in pulmonary function, there are no differences in either short- or long-term pulmonary or other morbidity between the two incisions. Thoracoscopy Video-assisted thoracoscopy with two or three stab incisions further reduces postoperative discomfort and attenuates postoperative reductions in pulmonary function. This is confirmed by the findings that thoracoscopy causes less pain, decreases opioid requirements and results in lesser reductions in respiratory muscle strength and lung volumes compared with posterolateral thoracotomy [48]. When thorascopic lobectomies are compared with those performed through muscle-sparing incisions, there are no differences in postoperative pulmonary function, although there is less pain. This is not unexpected because similar amounts of lung tissue are removed; however, these findings reduce the strength of the claim that postoperative pain plays an important role in reducing postoperative pulmonary function. Similarly, in sheep, the diaphragmatic shortening fraction recovers slightly faster after thoracoscopy, but this does not translate into a functional advantage compared with lateral thoracotomy [49]. Long-Term Consequences Long-term consequences of pulmonary resection include the effects of reduced lung tissue. FEV1 and FVC are reduced immediately after surgery due to the removal of lung tissue, pain, and atelectasis. Over the next 6 mo, there is slow improvement in pulmonary function after lobectomy, but not pneumonectomy. FEV (1), VC, and total lung capacity increase as atelectatic areas re-expand and as ventilation-perfusion relations improve. Exercise capacity 6 mo after lobectomy is unchanged from preoperative capacity in most patients, whereas it is reduced by 20% after pneumonectomy. Similar findings are observed 12 mo after pulmonary resection [50]. Pain Management Posterolateral thoracotomy is reported to be among the most painful of surgical incisions because major muscles are transected and rib(s) are removed (Table 2). Additionally, chest tube insertion sites are often very painful. Ameliorating this pain is essential for patient comfort and to facilitate coughing and maneuvers designed to prevent atelectasis, such as deep breathing exercises and incentive spirometry. The salutary effects of contemporary postoperative pain management techniques after thoracotomy remain unproven. Specifically, the ability of superior analgesia to decrease the incidence of pulmonary complications must be balanced against the respiratory depressive effects of opioids. The most frequently used analgesic modality is continuous epidural infusion of opioids and local anesthetic via a midthoracic level catheter [51]. Some [52], but not all [53], studies have reported that epidural opioids are superior to IV opioids in attenuating postoperative reductions in lung volume. However, these studies involved small numbers of patients and focused on analgesia scores and immediate side effects with respiratory measurements limited to spirometry and blood gas analysis. No large randomized study has yet examined whether epidural analgesia reduces complications and improves outcome after thoracotomy. A meta-analysis that included both postabdominal and postthoracotomy patients concluded that, compared with systemic opioids, epidural opioids reduce the incidence of atelectasis, pulmonary infections, and pulmonary complications [54]. Additionally, patients who received epidural local anesthetics had reduced incidences of pulmonary complications compared with systemic opioids. The epidural administration of opioids after thoracotomy presents a particular challenge because of the need to deliver medication to the upper thoracic region without causing respiratory compromise. Initial studies found significant PaCO2 elevations (>50 mm Hg) when 5 mg and 0.15 mg/kg morphine were injected into the lumbar epidural space. This led to the recommendation that the dose of lumbar epidural morphine be reduced to 2.5-3.0 mg. Subsequent observations with these smaller doses found that the possibility of respiratory depression with epidural opioids is always present, but the incidence is low. Only 0.6% of 504 postthoracotomy patients receiving epidural analgesia with bupivacaine plus fentanyl or morphine have respiratory rates <10 bpm [55]. Other investigators have found the incidence to be <0.1% [56]. Epidural analgesic regimens using mixtures of local anesthetics and opioids are currently popular. Whether such mixtures result in less respiratory depression is unclear, although they do decrease the incidence of hypotension. In postthoracotomy patients, adding bupivacaine to a thoracic epidural fentanyl infusion does not alter the degree of reduction in FVC, FEV1, or PEFR, but it reduces the PaCO2[57]. Similarly, adding epidural morphine to a continuous infusion of thoracic epidural bupivacaine does not influence PEFR, FEV1, or FVC, nor is there a difference in the suppression of the ventilatory response to CO2 between sufentanil alone and in combination with bupivacaine. However, recovery after stopping the infusion is faster in the combination group [58]. Nonsteroidal antiinflammatory drugs, especially ketorolac tromethamine, are used to supplement opioid analgesia. These drugs work synergistically with opioids and have no respiratory depressive effects. Disadvantages include platelet dysfunction and renal dysfunction. IV ketorolac is a better supplement for patient-controlled epidural hydromorphone analgesia than adding bupivacaine [59]. Ketorolac is also more effective (requires less on-demand opioid administration) than extrapleural intercostal nerve blocks with intermittent bupivacaine [60]. Current analgesic practice includes using epidural opioids with or without local anesthetics, supplemented with nonsteroidal antiinflammatory drugs. Although these regimens provide excellent pain relief, it is still unclear whether they reduce pulmonary morbidity. Esophagectomy Resection of the esophagus is usually performed to either cure or palliate an esophageal malignancy. Esophagectomies are performed via an abdominal incision, with or without an additional cervical incision or a lateral thoracotomy. Regardless of whether the chest is opened, there is trauma to the intrathoracic structures, especially when the stomach or jejunum is used to replace the esophagus. Therefore, the degree of respiratory compromise may be greater than that assumed from the location and extent of the skin incision. Postoperative respiratory complications in these patients may be due to a number of causes. There may be direct trauma to lung tissue during surgery, and afterward, the interposed stomach may compress adjacent lung tissue, especially if it becomes distended. The combination of upper abdominal surgery with its resultant diaphragmatic dysfunction and the pain and muscle spasms secondary to a lateral thoracotomy increases the likelihood that atelectasis will develop. Surgical complications, such as anastomotic leak and pneumothorax, may also cause respiratory compromise. In one series, pneumonia occurred in 12% (9 of 75) of patients after esophagogastrectomy. In another study, 50% of deaths were due to pneumonia [61]. Patients are often mechanically ventilated after esophagogastrectomy. The introduction of epidural analgesia has facilitated early extubation either immediately or soon after the completion of surgery. In patients who receive 2 mg of epidural morphine both at T6-7 and at T3-4 at the time of incision, the addition of a continuous epidural infusion of 0.25% bupivacaine (at 3 mL/h) results in earlier extubation (4.4 vs 13.7 h) than when isotonic sodium chloride solution is infused [62]. Early extubation after esophagectomy is associated with reduced morbidity and decreased length of ICU stay [63]. Some investigators have credited superior analgesia with improving outcome after esophagogastrectomy. Patients given either epidural, patient-controlled analgesic, or continuous IV morphine have fewer (13% vs 25%) respiratory complications, shorter hospital stays, and fewer (21% vs 43%) cardiovascular complications, as well as lower mortality (8% vs 14%), than those given intramuscular meperidine [64]. Similar reductions in hospital stay have been reported among patients who receive epidural rather than IV morphine [65]. Conclusions The perioperative period is marked by successive changes in the structure and function of the respiratory system. Atelectasis forms on the induction of general anesthesia, and it may be weeks until function has been restored to preoperative levels. This overview of postcardiac and thoracic surgery and the consequences of modern pain management demonstrates that much is unexplained and many questions remain unanswered. Pain management has long been considered part of the defense against postoperative atelectasis and pneumonia. Yet the evidence is sparse and disappointing because most studies involve small numbers of patients and focus on analgesia scores and immediate side effects. Only a few investigations have examined outcome. Interestingly, there is more evidence that epidural and other advanced pain modalities may reduce morbidity after esophagectomy than after lobectomy or cardiac surgery. The impression based on available evidence is that opioid analgesia, even when administered in novel ways, has only a limited ability to reduce or attenuate the postoperative alterations in pulmonary function and may even contribute to pulmonary complications by causing respiratory depression. If this is true, then more analgesics that do not depress respiration, e.g. nonsteroidal antiinflammatory drugs, should be used in this setting. There is limited evidence regarding whether the newer approaches to pain management are truly cost-effective. The cost-benefit (reduced length of hospital stay, earlier recovery of gastrointestinal function) of superior pain relief has been demonstrated in colonic surgery [66], and similar studies are needed in cardiac and thoracotomy patients. It is also currently unknown whether analgesia improves coughing effectiveness and facilitates the treatment and prophylaxis of atelectasis and pneumonia with incentive spirometry and deep breathing exercises. The future challenge will be to evaluate the efficacy of current practices while simultaneously devising new methods that reduce complications and improve outcome.