Perioperative Medical Complications

I. Problem/Condition

Surgical co-management is an important component of hospitalist practice. Part of the internal medicine physician’s responsibility under this model is to manage the patien’s chronic medical problems. Much of their effort and skill, however, is directed at anticipating and treating medical complications of surgery. Many of these complications are directly related to the patients’ underlying comorbidities.

II. Diagnostic Approach

A. What is the differential diagnosis for this problem?

Below is a systematic approach to the most common conditions seen in co-management.

  • Cardiac:

    Continue Reading

    Myocardial infarction/ischemia


    Sinus tachycardia

    Atrial fibrillation

    Hypertension, uncontrolled

    Congestive heart failure exacerbation

  • Endocrine:

    Diabetes mellitus, uncontrolled

    Adrenal insufficiency, acute

    Thyroid storm



Stress ulcer



Venous thromboembolism




Urinary tract infection


Clostridium difficile infection

Wound infection



Cerebrovascular accident

Alcohol withdrawal


Obstructive sleep apnea exacerbation


Chronic obstructive pulmonary disease exacerbation

Pneumonia, nosocomial

  • Renal:

    Acute renal failure

    Electrolyte disorders



  • Rheumatologic:

    Crystalline arthropathy

    Gout, acute flare

    Pseudogout, acute flare


Urinary retention, acute

B. Describe a diagnostic approach/method to the patient with this problem

Rounds on surgical co-management patients are best approached with a particular mindset. Prevention of medical complications and early diagnosis of complications as they occur should be the primary focus. This mindset is facilitated by a knowledge of the common complications and their risk factors.

Postoperative myocardial infarction (MI) is the most hazardous postoperative medical complication and the one that receives the most preoperative attention, in the form of well-accepted screening algorithms. It is associated with approximately 25% mortality. MI occurs in 1.5-3% of typical elective noncardiac surgery patients and in 4% of postoperative patients with known coronary artery disease. These events occur most often in the first 48 hours after surgery. The risk is higher after operations, such as lower extremity arthroplasty, for conditions that limit exercise tolerance. In these types of surgeries, the inability to perform physical exertion preoperatively may mask the symptoms of coronary artery disease. Undiagnosed coronary artery disease is especially common in patients undergoing vascular surgery due to exercise intolerance and the risk factors that vascular disease shares with coronary artery disease. Recent MI is another major risk factor for postoperative MI. Approximately 30% of patients who have surgery 30 days after MI will suffer a second MI, with a mortality risk of 14%. The risk of a second MI with surgery does not return to baseline until 3 months after the first MI. The risk of postoperative cerebrovascular accident (CVA) is also highly correlated with the time since previous CVA. Within 3 months after a first CVA, the risk of suffering a postoperative CVA is 12% with mortality 7.5%. The risk of postoperative CVA does not reach a new baseline until 9 months after an initial CVA.

More than half of people over 65 years of age have hypertension. Consequently, it is quite common in the population undergoing elective surgery. Acute postoperative hypertension is most likely to occur in the immediate postoperative setting but can occur at any time perioperatively. It often develops in response to uncontrolled pain, fluid overload, hypoxia, or hypercapnea. This phenomenon is more common after cardiac, vascular, head/neck, and neurologic surgeries.

Arrhythmias are one of the most common postoperative medical complications. Arrhythmias of one type or another occur in over half of patients recovering from noncardiac surgery. Atrial fibrillation occurs most commonly on postoperative day number one and is associated with a 33% increase in length of stay. It has known association with intrathoracic surgeries and abdominal aortic aneurysm repairs. It is also more common in patients with a history of supraventricular tachycardia, asthma, or significant valvular heart disease. Postoperative ventricular arrhythmias are much less common and occur most often in patients with depressed left ventricular function.

Postoperative glycemic control is a very common surgical co-management issue, but true endocrine complications of surgery are unusual. It is not unusual, however, for patients receiving chronic glucocorticoid therapy to require elective surgery of various types. Precipitation of hypotensive collapse in these patients is a common concern, but published evidence suggests that hypotension rarely if ever occurs in patients receiving less than 16 mg of prednisone per day (or equivalent) assuming the usual daily glucocorticoid dose is continued perioperatively. Chronic glucocorticoids are one of many reasons to ensure that accurate medication reconciliation is performed postoperatively.

Thyroid storm is a truly rare condition, but the precipitating event may be a surgery. The mortality associated with thyroid storm approaches 30%, even in patients who receive appropriate treatment.

Stomach complaints, including constipation, are frequently encountered in surgical co-management. These complaints don’t always get the attention they deserve. Gastrointestinal bleeding that develops in hospitalized patients is associated with mortality in excess of 30%. This complication occurs most commonly in patients who need intensive care, especially if they are mechanically ventilated. Having said that, over half of nosocomial gastrointestinal bleeding occurs outside of the ICU. Bleeding is more likely in patients with prolonged hospital stay, multiple comorbidities, and those undergoing organ transplantation.

Postoperative ileus complicates approximately 4% of all surgeries and prolongs length of stay by about 4 days in affected patients. It occurs most commonly after gastrointestinal operations but can follow any type of surgery. For example, acute gastroparesis occurs after more than half of Whipple procedures while colonic pseudo-obstruction develops following 1-2% of orthopedic and spine surgeries.

Venous thromboembolism (deep venous thrombosis and pulmonary embolus), is an important cause of postoperative morbidity and mortality. In the absence of prophylaxis, patients develop asymptomatic deep venous thrombosis (DVT) following 15-30% of general surgeries. Fatal pulmonary embolus (PE) occurs in 0.2-0.9%. The risk of asymptomatic DVT following major orthopedic, major trauma, and thoracic surgery approaches or exceeds 50% without prophylaxis. Other risk factors for postoperative venous thromboembolism include malignancy, previous DVT or PE, obesity, impaired mobility, age, duration of surgery, and postoperative infection. Postoperative anemia is almost ubiquitous, occurring after 85-90% of surgeries that require hospitalization. Anemia tends to be more common and more severe in patients undergoing surgery for cancer, especially gastrointestinal malignancies.

Fever is a common occurrence after surgery, with an incidence of 15-90% depending on the definition of fever and the type of surgery. Infection is present in less than 10% of patients with postoperative fever.

Delirium is a disorder of inpatients, including those who are recovering from surgery. It complicates 11% of general surgeries, 36% of vascular surgeries, and 42% of surgeries that last longer than 10 hours. Delirium is present in 80% of ICU patients and often extends beyond their stay in the ICU. Delirium risk factors include age, dementia, and personal history of delirium. Delirium is associated with increased mortality and longer lengths of stay.

One in ten general surgery patients has alcohol dependence. The prevalence rises to over 25% in those admitted due to an injury. Alcohol dependency is associated with several postoperative complications other than the withdrawal syndrome. It increases the risk of infection, acute respiratory distress syndrome, arrhythmias, congestive heart failure, and postoperative bleeding. Untreated alcohol withdrawal syndrome carries a mortality rate of 15%, but treatment reduces the mortality to below 5%.

The burden of perioperative complications associated with obstructive sleep apnea (OSA) may be larger than for any other chronic medical condition. OSA is highly prevalent in patients undergoing surgery and is associated with frequent and serious postoperative complications. OSA is present in over 20% of general surgery patients. It is probably even more common in patients planning major orthopedic surgery and was found in over 75% of patients presenting for obesity surgery. Patients with OSA, particularly if not receiving preoperative treatment with continuous positive airway pressure (CPAP), have more than three times as many serious complications as matched controls. They are more likely to require transfer to the intensive care unit and have longer lengths of stay. Despite the growing recognition of OSA as a significant public health problem, studies show that well over half of patients with the disorder remain undiagnosed.

1. Historical information important in the diagnosis of this problem

For the diagnosis of postoperative MI, vigilant history-taking is often required. Daily cardiac systems review for angina or dyspnea or even gastrointestinal upset is appropriate, especially in patients at higher risk. MI in the postoperative setting usually presents with subtle or atypical symptoms.

Patients with elevated postoperative blood pressures may or may not have previously-diagnosed hypertension. A review of preoperative blood pressures and antihypertensive medications is often fruitful. Patients with uncontrolled preoperative hypertension are, of course, at risk for postoperative elevations in blood pressure. Failure to resume preoperative antihypertensive agents, especially clonidine, can result in rebound hypertension 24-48 hours postoperatively.

Acute adrenal insufficiency can be precipitated by the stress of surgery in patients with Addison’s disease or, more commonly, in patients with iatrogenic adrenal insufficiency due to chronic glucocorticoid administration. In patients with hypotension that does not respond adequately to fluids, any history of preoperative glucocorticoid use should be carefully re-explored.

Ileus can be distinguished from small bowel obstruction on the basis of history and examination in most circumstances. The main differentiating historical element is pain. Patients with ileus typically have little or no abdominal pain while those with small bowel obstruction usually suffer from severe colicky abdominal pain.

Differentiating benign fever from true infection is an essential skill of the hospitalist performing surgical co-management. The timing of fever is a useful indicator of the presence or absence of infection. Only 20% of patients with fever on the first postoperative day have infection as opposed to 90% of those who develop fever on the fifth postoperative day or later. In general, fever that develops after the second postoperative day should be considered to represent infection until proven otherwise. Elements of the patients’ histories are the most important clues to the source of infection. Surgical site pain that fails to improve as expected can suggest wound infection which accounts for over 40% of postoperative infections. Fever due to surgical site infections usually develops after the fourth postoperative day. Current or recent urinary catheterization increases the risk of urinary tract infection, which accounts for almost a third of postoperative infections. Early removal (or complete avoidance) of urinary catheters is the most important method for preventing catheter-associated urinary tract infections. Nosocomial pneumonia is the cause of approximately 10% of postoperative infections and occurs more commonly in patients with chronic obstructive pulmonary disease. Any indwelling central venous catheter should be investigated as the cause of a fever developing after postoperative day number two.

Noninfectious causes must be considered in the evaluation of postoperative fever. Malignant hyperthermia usually becomes evident intraoperatively due to high fever and intense muscle contraction, but can present within a few hours postoperatively. Thyroid storm presents with fever, vomiting, diarrhea, and neurologic changes including seizure or coma. These are two causes of fever within the first 48 hours after surgery that do require immediate action. Acute gout attacks complicate up to 15% of surgeries in patients with a history of gout; in the postoperative setting, fever is a usual feature of gout attacks and polyarticular involvement is common. Lastly, drug fever should be considered in patients receiving antibiotics, especially beta-lactams or sulfonamides.

Delirium can be differentiated from depression and dementia based on its natural history. It develops over hours to days as opposed to months for depression and years for the development of dementia. Delirium also tends to have a fluctuating course, often worsening at night and improving or resolving during the day. The history is also key to uncovering the etiology of delirium. Medication adverse effect is the most common underlying etiology of delirium, so a thorough medication history is vital. Opioids, sedatives, glucocorticoids, and anticholinergics are well-known causes of delirium, but any new medication in a delirious patient may be partly responsible. Alcohol withdrawal syndrome is a diagnosis of exclusion, but must be considered in patients who develop hyperactive delirium postoperatively.

2. Physical Examination maneuvers that are likely to be useful in diagnosing the cause of this problem

Elevations in systolic blood pressure above 180 mmHg or diastolic blood pressure above 110 mmHg require physician confirmation of blood pressure and evaluation for signs of end organ damage. This evaluation should include mental status examination, palpation of pulses in all four extremities, ophthalmoscopic examination, and auscultation for pulmonary crackles. Regular, narrow-complex tachycardia can sometimes be further evaluated or even terminated with carotid sinus massage. Care should be taken to avoid massage of a carotid artery containing unstable plaque.

Standard physical examination techniques can be very helpful in the evaluation of postoperative gastrointestinal complications. Despite the predominance of stress ulcers among the causes of inpatient gastrointestinal bleeding, 13% of inpatient bleeding is due to a lower gastrointestinal source. For this reason, rectal examination is an important early step toward identifying the bleeding source. Ileus can usually be distinguished from small bowel obstruction based on history and examination, as mentioned above. The main differentiating exam element is bowel sound auscultation. Patients with ileus typically have absent bowel sounds while those with small bowel obstruction have hyperactive bowel sounds often accompanied by rushing or tinkling noises.

Since its symptoms may be outshined by postoperative pain, routine lower extremity examination of surgical co-management patients is vital to avoid a delay in the diagnosis of postoperative DVT. Although Homans’ sign has notoriously low sensitivity, physical exam findings, when combined and used in the Wells score, are predictive of DVT.

Delirium is characterized by a change in level of consciousness, disturbance in cognition, and fluctuating course. Use of a bedside instrument, such as the confusion assessment method (CAM), facilitates the timely diagnosis of delirium by non-psychiatric health care providers and may be used as part of a screening protocol for high risk patients.

The CAM is made up of one historical feature, acute onset and fluctuating course, and three exam features, inattention, disorganized thinking, and altered level of consciousness. Delirium is present if the patient displays an acute onset and a fluctuating course, is inattentive, and exhibits either disorganized thinking or an altered level of consciousness. The physical exam can also help guide the search for underlying etiologies of delirium. Fever suggests an infectious cause. Asterixis suggests hepatic or renal failure. Nystagmus or opthalmoplegia suggest Wernicke’s encephalopathy.

3. Laboratory, radiographic and other tests that are likely to be useful in diagnosing the cause of this problem

The diagnosis of postoperative acute coronary events is made using the same testing strategy used in other settings. An initial electrocardiogram and an initial set of cardiac enzymes, if normal, are followed serially. One unique aspect of postoperative cardiac ischemia is that it is rarely accompanied by chest pain. For this reason, inpatient monitoring for ischemia or infarction has been investigated. Both continuous ST-segment evaluation and daily troponin measurement have shown promise in detecting otherwise silent ischemia/infarction events. Consensus guidelines don’t go as far as recommending either one, but they might be appropriate for higher risk patients.

The new onset of atrial fibrillation/flutter or its occurrence in a patient who was felt to be in good rhythm control preoperatively should result in a search for precipitating causes. Hypo/hyperkalemia, hypomagnesemia, hyperthyroidism, anemia, acidosis, hypoxia, and hypercapnea should be considered. Atrial fibrillation may be a sign of pulmonary embolism or pneumonia. When sustained or symptomatic ventricular tachycardia or any ventricular fibrillation occurs in the postoperative setting, patients must be evaluated for cardiac ischemia/infarction as well as electrolyte derangements.

The goals of postoperative glycemic control are under some debate, but there is no debate that glucose control can’t be achieved if diabetes remains undiagnosed. Over 25% of patients without a previous diagnosis of diabetes who present for elective noncardiac surgery can be expected to have elevated blood glucose levels. For this reason, routine perioperative screening with a fasting blood glucose level is recommended for all patients without a previous diabetes diagnosis.

Ileus can usually be diagnosed without radiographic testing, but when it is persistent or when the diagnosis is in doubt, ready confirmation is available in the form of a flat plate abdominal x-ray. Computed tomography (CT) scanning is useful if small bowel obstruction is suspected over ileus.

When combined with a low-risk Wells score (see Table I), a normal d-dimer level excludes the diagnosis of DVT. Patients with a high-risk Wells score should undergo compression ultrasonography to confirm the diagnosis of DVT.

Clinical Characteristic Points
Active cancer 1
Paralysis, paresis, or lower extremity casting 1
Bedridden >3 days or major surgery in the previous 12 weeks 1
Tenderness along the deep venous system 1
Swelling of the entire leg 1
Asymmetric calf diameter >3 cm 1
Pitting edema confined to one leg 1
Visible venous collaterals 1
History of DVT 1
Alternative diagnosis is at least as likely as DVT -2

0 or 1 = low risk for DVT

2 or greater = high risk for DVT

Concurrent with discontinuation of suspect medications, laboratory evaluation of delirium should be initiated. Basic chemistries and a complete blood count are reasonable for all patients with delirium to exclude hyponatremia, hypercalcemia, anemia, and disorders associated with thrombocytopenia as well as to inform the suspicion of infection. Urine drug screening can be informative in the appropriate situation. Elevated mean corpuscular volume, aspartate aminotransferase, or gamma-glutamyl transpeptidase lend weight to the suspicion of alcohol withdrawal.

Infectious causes of delirium can be further investigated with blood cultures, urine culture, and chest x-ray. Arterial blood gas evaluation is appropriate in patients with pulmonary disease who develop delirium to rule out hypoxia and hypercapnea. In patients who are suspected of having advanced liver disease, an elevated arterial ammonia level can confirm hepatic encephalopathy.

Because fever and unusual patterns of joint involvement are the rule in postoperative crystalline arthropathy attacks, differentiation from septic arthritis is often difficult based on history and exam alone. Joint aspiration with examination for crystals under a polarizing microscope is necessary unless the attack is typical for the patient.

C. Criteria for Diagnosing Each Diagnosis in the Method Above


D. Over-utilized or “wasted” diagnostic tests associated with the evaluation of this problem

With few exceptions, potential perioperative medical complications deserve a thorough diagnostic evaluation.

In the case of uncontrolled postoperative hypertension, a complete evaluation for causes of secondary hypertension is usually unnecessary. Neither a 24-hour urine collection for catecholamines and metanephrines nor serum metanephrine measurement should be undertaken in patients without symptoms of pheochromocytoma. Screening for renal artery stenosis with a retroperitoneal ultrasound is unnecessary unless there is evidence of resistant hypertension preceding surgery.

In postoperative patients with regular tachycardia, an EKG tracing is not always helpful. An EKG is indicated if the tachycardia persists, especially without an identifiable cause such as pain, nausea, or anemia. EKG evaluation is also warranted if the pulse exceeds 120 bpm, to differentiate sinus tachycardia from atrial flutter with 2:1 block.

Evaluation of postoperative fever probably leads to more fruitless diagnostic effort than all other postoperative complications combined. Fever in the postoperative setting is not necessarily indicative of a pathologic process. The elevated temperature is caused by unknown mechanisms, but is probably not related to atelectasis. Chest x-rays, urine studies, and blood cultures need not be performed unless the fever occurs more than 48 hours after surgery or is accompanied by symptoms of infection (cough/dyspnea, dysuria, rigors, etc.).

Finally, brain imaging with CT or MRI is not useful in the evaluation of postoperative delirium, unless focal neurologic symptoms are present or can’t be ruled out by physical exam.

III. Management while the Diagnostic Process is Proceeding

A. Management of Clinical Problem Perioperative Medical Complications

Great attention is paid to the prevention of postoperative myocardial infarction in the preoperative arena, but postoperative prevention should also be considered. Because of the high associated mortality and the limited therapeutic options, postoperative myocardial infarction should be managed in consultation with a cardiologist.

In high-risk patients, tachycardia appears to increase the risk of myocardial ischemia and thus, evaluation and management of underlying causes of tachycardia is essential. Sinus tachycardia is a common occurrence after virtually all types of surgery. It can usually be addressed through attention to pain, nausea, anemia, and hypovolemia. When sinus tachycardia persists, is severe (pulse >130), or is associated with fever, consideration should be given to serious underlying causes such as sepsis, alcohol withdrawal, pulmonary edema, or pulmonary embolus.

Severe perioperative elevations in blood pressure require prompt evaluation, as discussed above. Any preoperative antihypertensives that were discontinued, especially beta-blockers or clonidine, should be reinstituted. Volume overload, pain, nausea, hypothermia, hypoxia, and hypercapnea can acutely increase the blood pressure perioperatively. Resumption of preoperative antihypertensives and attention to these common exacerbating factors is usually adequate to slowly normalize the blood pressure. In the interim, intravenous antihypertensives may be necessary. Caution should be employed to avoid rapid blood pressure normalization as it can precipitate or worsen end organ damage.

The management of postoperative atrial fibrillation should focus initially on rate control. As mentioned above in relation to sinus tachycardia, pain, nausea, anemia and hypovolemia can also contribute to rapid ventricular response in patients with atrial fibrillation. While those issues are being addressed, patients with heart rates over 100 should be treated with a beta-blocker or a non-dihydropyridine calcium channel blocker, which may be administered either orally or intravenously. Prophylaxis against systemic embolization must be considered within 48 hours of the onset of atrial fibrillation. Patients who are newly-diagnosed in the perioperative setting likely have paroxysmal atrial fibrillation that was unrecognized preoperatively. Anticoagulation for these patients may be immediately indicated, based on the CHADS-VASc score, but the risk of embolization must be balanced against the risk of surgical site bleeding.

Acute adrenal crisis can be prevented or treated with the intravenous administration of an agent with glucocorticoid and mineralocorticoid activity. Hydrocortisone is often employed at doses between 75 and 300 mg divided several times per day. The management of thyroid storm involves a combination of antithyroid medication, iodine, beta-blockade and sometimes glucocorticoids. The sequence in which these agents are administered is critical. Endocrinology consultation should always be sought.

Recovery from postoperative ileus can be facilitated by correcting electrolyte disturbances, addressing infection, and avoiding medications that decrease gastrointestinal motility, such as opioids, anticholinergics, and calcium channel blockers. Early enteral feeding may decrease the duration of ileus. Pro-motility agents such as erythromycin and metoclopramide may provide short-term benefit. The chewing of gum is a nonpharmacologic modality that increases gastrointestinal motility and has shown some promise in treating ileus.

Pharmacologic prophylaxis for venous thromboembolism is indicated after most surgeries that require hospitalization. The decision to forgo pharmacologic prophylaxis should be made with the assistance of a validated risk score (Caprini score >2). In patients with malignancy or a history of previous venous thromboembolism who undergo general surgery, pharmacologic thromboprophylaxis should be continued after discharge for a total of 4 weeks. The mean time to development of DVT is 7 days following total knee arthroplasty and 17 days following total hip arthroplasty. Because of the extended risk, pharmacologic thromboprophylaxis must be continued after discharge in these patients as well. Following bariatric surgery, the usual prophylaxis doses of heparin, low-molecular-weight heparins, and fondaparinux may not be fully effective. If low-molecular-weight heparins or subcutaneous unfractionated heparin are employed, anti-Xa levels should be followed. For surgeries that are associated with higher postoperative bleeding risk such as trauma, neurosurgery, and prostate surgery, decisions about thromboprophylaxis must be individualized. In general, even after these surgeries, the benefits of pharmacologic thromboprophylaxis are felt to outweigh the risks.

Due to the risks of exposure to allogenic blood and the failure of early transfusion to improve meaningful outcomes, postoperative transfusion thresholds should remain strict. Current evidence supports a postoperative transfusion threshold of 7-8 g/dl. Erythropoietin and oral iron have shown benefit when administered preoperatively in selected patients, but neither appears to have a role in the postoperative setting. In contrast, postoperative administration of intravenous iron has been shown to reduce the need for transfusion in certain populations, such as those recovering from orthopedic surgery. Postoperative intravenous iron should be considered in patients with a known or suspected element of iron deficiency.

The mainstay of delirium treatment is correction of underlying medical causes. There are usually multiple causes of delirium in individual patients, all of which should be addressed. General measures should also be undertaken in delirious patients to optimize environmental stimulation. Use of patients’ glasses and hearing aids should be facilitated. Hospital rooms should have a clearly visible wall clock and an area for displaying the day of the week and the date. Keeping window shades open during the day helps delirious patients reorient. Noise, including alarms, should be kept to a minimum, especially at night. Restraints should be avoided. Finally, family and friends should be encouraged to visit frequently. While underlying causes are being addressed, antipsychotics may be useful to treat the symptoms of delirium. Haloperidol may be employed on a prn basis for agitation. Some evidence suggests that a scheduled atypical antipsychotic should be used in place of or in addition to haloperidol.

Continuous positive airway pressure is the cornerstone of OSA management, but its utility in preventing postoperative complications is based on limited evidence. Standard practice is to initiate CPAP promptly after surgery using patients’ own equipment. In contrast to how it is used as an outpatient, current guidelines recommend that CPAP be worn continuously with minimal interruption for the first 24-48 hours after surgery. In addition to CPAP, several less-invasive measures should also be taken. Patients with OSA recovering from surgery should be maintained in a semi-upright position in bed and should be monitored with continuous oximetry. Avoidance of opioid analgesia is ideal for patients with OSA. Where possible, opioids should be replaced with NSAIDs or acetaminophen. When opioids must be used in patients with OSA, efforts to find the minimum effective dose are worthwhile. The management of patients with suspected OSA who have not yet been evaluated by a polysomnogram is more challenging. Institution of the less-invasive measures noted above is appropriate for patients suspected of having OSA. Postoperative initiation of CPAP might be considered for those patients considered to be at the highest risk for OSA or in those with witnessed postoperative apneas.

Embolization Score

See Table II.

Table II.
Clinical Characterstic Points
Congestive heart failure 1
Hypertension 1
Age >75 1
Diabetes mellitus 1
History of stroke or transient ischemic attack 2

2 or greater = anticoagulation recommended, annual stroke risk exceeds 4%

B. Common Pitfalls and Side-Effects of Management of this Clinical Problem

The mortality associated with nosocomial gastrointestinal bleeding and the availability of powerful acid-suppressing medications has resulted in the over-utilization of stress ulcer prophylaxis outside of the ICU. Proton pump inhibitors are commonly employed for this purpose despite limited evidence of their benefit. Side-effects are rare but proton pump inhibitors do increase the risk of nosocomial infections and they are associated with significant cost.

Although metoclopramide may have some place in the treatment of postoperative nausea and vomiting, it must be avoided in patients with Parkinson’s disease due to its anti-dopaminergic effect.

Fever that develops after the second postoperative day should not be ascribed to atelectasis; rather, it should be considered to represent a potential infection. The best evidence to date suggests that atelectasis is not a cause of postoperative fever.

Management of delirium with antipsychotics can result in prolongation of the QT interval and torsades de pointes ventricular arrhythmias. Antipsychotics should be avoided in patients taking other agents that prolong the QT interval (e.g., antiarrhythmics, fluoroquinolones, macrolides, methadone). Antipsychotics should not be used alone in the treatment of alcohol withdrawal as they are known to decrease the threshold for seizures.

VII. What's the evidence?

Fleisher, LA, Fleischmann, KE. “2014 ACC/AHA Guideline on Perioperative Cardiovascular Evaluation and Management of Patients Undergoing Noncardiac Surgery”. J Am Coll Cardiol. vol. 64. 2014. pp. e77-137.

Devereaux, PJ, Goldman, L, Cook, DJ, Gilbert, K. “Perioperative cardiac events in patients undergoing noncardiac surgery: a review of the magnitude of the problem, the pathophysiology of the events and methods to estimate and communicate risk”. CMAJ. vol. 173. 2005. pp. 627-34.

Haas, CE, LeBlanc, JM. “Acute postoperative hypertension: a review of therapeutic options”. Am J Health-Syst Pharm. vol. 61. 2004. pp. 1661-75.

Hatzakorzian, R, Bui, H, Carvalho, G, Shan, WLP. “Fasting blood glucose levels in patients presenting for elective surgery”. Nutrition. vol. 141. 2011. pp. 298-301.

Marik, PE, Varon, J. “Requirement of perioperative stress doses of corticosteroids: a systematic review of the literature”. Arch Surg. vol. 143. 2008. pp. 1222-6.

Johnson, MD, Walsh, RM. “Current therapies to shorten postoperative ileus”. Cleve Clin J Med. vol. 76. 2009. pp. 641-8.

Carson, JL. “Perioperative blood transfusion and postoperative mortality”. JAMA. vol. 279. 1998. pp. 199-205.

Carson, JL. “Liberal or restrictive transfusion in high-risk patients after hip surgery”. N Engl J Med. vol. 365. 2011. pp. 365-2453.

Gould, MK, Garcia, DA, Wren, SM, Karanicolas, PJ. “Prevention of vte in nonorthopedic surgical patients: antithrombotic therapy and prevention of thrombosis, 9th ed: american college of chest physicians evidence-based clinical practice guidelines”. Chest. vol. 141. 2012. pp. e227S-e277S.

Galck-Ytter, Y. “Prevention of vte in orthopedic surgery patients: antithrombotic therapy and prevention of thrombosis, 9th ed: american college of chest physicians evidence-based clinical practice guidelines”. Chest. vol. 141. 2012. pp. e278S-e325S.

Ward, DT, Hansen, EN, Takemoto, SK, Bozic, KJ. “Cost and effectiveness of postoperative fever diagnostic evaluation in total joint arthroplasty patients”. J Arthroplasty. vol. 25. 2010. pp. 43-8.

Engoren, M. “Lack of association between atelectasis and fever”. Chest. vol. 107. 1995. pp. 81-4.

Cavaliere, F, D’Ambrosio, F, Volpe, C, Masieri, S. “Postoperative delirium”. Curr Drug Targets. vol. 6. 2005. pp. 807-14.

Inouye, SK, van Dyck, CH, Alessi, CA, Balkin, S. “Clarifying confusion: the confusion assessment method. A new method for detection of delirium”. Ann Intern Med. vol. 113. 1990. pp. 941-8.

Gupta, RM, Parvizi, J, Hanssen, AD, Gay, PC. “Postoperative complications in patients with obstructive sleep apnea syndrome undergoing hip or knee replacement: a case-control study”. Mayo Clin Proc. vol. 76. 2001. pp. 897-905.

“Practice guidelines for the perioperative management of patients with obstructive sleep apnea”. Anesthesiology. vol. 120. 2014. pp. 268-86.