Figure 15-1. Variation in surgical mortality.
These circumstances influence both institutional and overall health care delivery. The overwhelming health care spending burden adds pressure to provide efficient, high-quality care. Hospital performances on quality measures are widely available on the Internet and mandated public reporting can easily affect market share and policy making. These system-level considerations only heighten the push toward population-based management of surgical care.
Nevertheless, the direct impact of postoperative complications on an individual patient’s quality of life is critical. The consequences are often unrecognized by standard hospital-based review. In 2004, Clavien and Dindo proposed a complication grading scale based on the severity of intervention required to address a given complication, including the presence of associated disability (Table 15-3).18 This classification has been validated in a cohort of more than 6,000 patients, with subsequent evaluation showing 89% interrater agreement and correlation between perceptions by patients, nurses, and physicians.19 It is hoped that encouraging standardized complication grading will increase meaningful, patient-centered outcomes reporting.
Table 15-1 Surgical Complications: Comparison of the 10 Most Frequent Perioperative Complications Between Private Sector and VA Hospitals
Table 15-3 Grading of Surgical Complications: Clavien–Dindo Classification
Table 15-4A Logistic Regression Models for Prediction of 30-Day Operative Mortality Using Preoperative Variables
Continued development has been sponsored by the American College of Surgeons (ACS) under the ACS-NSQIP® program, combining data collection, centralized benchmarking, and regular reporting. Reports have been utilized as catalysts for institutional-level change, with substantial improvements seen following targeted quality improvement efforts.21 Following implementation, 66% of ACS-NSQIP hospitals saw reduced mortality rate and 82% saw reduction in complications. The annual benefit for a hospital enrolling in ACS-NSQIP has been estimated at 250 to 500 fewer complications and 12 to 36 fewer deaths.22 Observed improvements in outcomes appear to be sustained, with the magnitude of quality improvement increasing over time.23 Of note, improvements in risk-adjusted outcomes at ACS-NSQIP hospitals were not found to differ significantly from those concurrently seen in matched non-NSQIP hospitals, suggesting that while feedback and reporting may be critical to implementing process change, they are not by themselves required to drive improvement.24,25
Algorithm-based risk stratification based on large data sets is now available throuh the Internet or on mobile devices. In 2011, a cardiac risk calculator based on NSQIP variables compared favorably to performance of the Revised Cardiac Risk Index.26 In 2013, the ACS released an online decision support tool including a perioperative risk calculator built on the NSQIP database of 1.3 million operations (www.riskcalculator.facs.org, last accessed March 30, 2016). The calculator incorporates 21 variables from the original VA NSQIP data collection and has excellent predictive performance for mortality, morbidity, and six specific complications.27 Potential criticisms of the model include dependence on American Society of Anesthesiology (ASA) Physical Status grading, imprecise definition of functional outcomes, and lack of validation outside of a NSQIP cohort28; however, the ability to rapidly provide individualized patient risk estimation is an important advance. Figure 15-2A,B demonstrate an example of risk determination for a patient undergoing elective colon surgery, entering specific clinical variables and reporting specific complications in an individualized fashion. The use of mobile device decision support tools will also likely continue to increase; as an example, the Michigan Surgical Optimization Program’s mobile app provides risk stratification for procedures derived from the Michigan Surgical Quality Collaborative data.
GENERAL CONCEPTS OF PREOPERATIVE EVALUATION
The setting for preoperative evaluation varies with the surgical procedure and with patient circumstances. Coordination of preoperative care at a dedicated facility has led to improved efficiency for outpatient elective surgery29; however, this advantage may lessen as intensity of routine testing decreases.
Patient education remains a key aspect of the initial preoperative visit. Preoperative teaching has been shown to improve patient satisfaction and increase cooperation with postoperative recommendations.30 The wealth of information available on the Internet has enabled patients to be better educated, stronger consumers and has empowered many to participate more actively in their care. Prior to the first office visit, many will have researched their disease process, surgical procedure, as well as the scorecards of the hospital and surgeon. Forward-thinking surgeons can leverage this engagement and increase the value of the face-to-face appointment by providing online teaching materials ahead of the actual office visit.
Initial evaluation includes a standard medical history and complete physical examination. The presence of allergies, current medications, chronic illnesses related and unrelated to the surgical condition, prior procedures, as well as staging and acuity of disease all contribute to the risk and benefit determinations.
Routine preoperative laboratory testing is no longer recommended. The majority of preoperative laboratory tests are rarely indicated and are of limited value.31 It has been demonstrated that, even when abnormal, preoperative testing often does not lead to change in perioperative management or adverse postoperative outcomes.32 Despite this, preoperative testing continues to be overused in both low risk33 and hospital settings.34
Selective laboratory testing has potential value in certain elective cases, as well as in remote telemedicine evaluations35 and as part of risk assessment in elderly patients.36 In all cases, testing should be based on clinical likelihood of finding and acting on abnormal results. Baseline hemoglobin should be assessed in elderly patients, patients undergoing procedures with expected major blood loss, and patients with underlying hematologic disorders. Renal function can be assessed in older patients, in younger patients with suspicion of renal disease or with associated diabetes, cardiac, vascular, or pulmonary disease, and in patients undergoing major surgery. Electrolytes, glucose, coagulation studies, and liver function tests should be performed selectively. Adjunctive cardiopulmonary evaluation should also be performed selectively.
Routine electrocardiogram (ECG) is not recommended before low-risk surgery. Patients with known coronary disease or elevated risk should be evaluated, unless the procedure itself is low risk. More invasive testing, including exercise testing, echocardiography, and cardiac catheterization should be based on the presence of symptoms, overall functional level, procedural factors, and calculated coronary risk (see Cardiac section). Routine chest x-ray, pulmonary function tests, and arterial blood gases are not indicated unless specific pulmonary symptoms are identified (see Pulmonary section).
For otherwise healthy patients, selective, rather than routine, evaluation by an anesthesiologist is recommended. Use of a screening questionnaire may facilitate appropriate referral for a formal anesthesia consultation.37 A nursing instrument identifying high-risk conditions (history of prior or familial anesthetic difficulty, difficult airway/limited neck motion, decreased exercise tolerance, stroke, thyroid, cardiac, asthma, and liver and renal disease) has been validated as a trigger for formal anesthesia consult.38
Figure 15-2. A,B: Risk stratification: ACS-NSQIP risk calculator. (Source: www.riskcalculator.facs.org, American College of Surgeons.)
The ASA Physical Status classification (Table 15-15) addresses the overall physical status of the patient, exclusive of specific surgical considerations. Despite its relative imprecision and inconsistent application, ASA class remains strongly associated with other predictors of surgical outcomes39 and continues as a standard adjunct during preoperative evaluation. The choice of anesthetic approach (local, monitored anesthesia care/sedation, neuraxial, and general) varies on the basis of the indicated procedure as well as patient considerations, with general anesthesia carrying a higher risk for pulmonary complications.40 Systematic review suggests that the use of neuraxial blockade in surgery is associated with overall mortality reduction and decreased complications; however, it is not possible to ascertain whether these findings were due to salutary effects of regional blockade or reduction in the use of general anesthesia.41 The use of epidural anesthesia may decrease perioperative cardiac events in patients with hip fracture and abdominal aortic aneurysm surgery,28 but it is not clear that intraoperative neuraxial anesthesia universally reduces perioperative cardiac events.
Each surgical procedure carries its own set of specific risks. The overall physiologic impact varies between procedures. Procedures involving the thoracic or abdominal cavity, neurosurgery, head and neck, or vascular system carry higher risk42 than those limited to the extremities or subcutaneous tissue. Procedural duration, often associated with complexity and extent of resection, is also associated with increased risk.40 The extent of surgical exposure has also been correlated with physiologic impact. Minimally invasive approaches have, in general, been associated with lower hospital length of stay, earlier return to function, reduced postoperative pain, and patient satisfaction, but in many cases encumber the risk of increased procedural duration and costs without clear mortality benefit. The urgency of the procedure affects risk assessment (Table 15-16).28 Emergency surgery carries the additional challenges of life- or limb-threatening condition without the opportunity for medical optimization. The underlying processes are often catastrophic, with increased risk of infection, hemorrhage, and uncorrected physiologic derangement adding to the risk of postoperative complications43 and mortality.42
Disparities in surgical outcomes based on racial and socioeconomic considerations still persist. The relative rates of postoperative complications, including respiratory failure, physiologic and metabolic derangements, are generally increased for minority patients over Caucasian patients.44 The causes are multifactorial and, likely, patient-related factors, as well as chronic and acute health care resource availability, play a role. In a survey of the US Nationwide Inpatient Sample involving more than 3 million oncologic procedures performed between 1999 and 2009, African American, but not Hispanic patients had higher rates of vascular, wound, gastrointestinal (GI), and infectious complications when compared to white patients.45 These variations in patterns of use suggest that unexploited opportunities may exist to reduce disparities. Cultural and economic barriers to surgical care may not be readily visible to the surgeon and should be actively considered.
Patients are living longer and with more substantial comorbidities. The world’s population of people aged 60 years and older has doubled since 1980 and is forecast to reach 2 billion by 2050.46 Demand for surgical services continues to increase faster than the rate of population growth; in 2006, surgery on elderly patients accounted for 35% of inpatient and 32% of outpatient surgical procedures.47
Table 15-6 Procedural Urgency
Normal aging reduces the ability to maintain homeostatic responses in the face of acute physiologic stress48 such as surgery. In a series of over a million Medicare patients, operative mortality for patients aged 80 years and older was observed to be more than twice that for patients aged 65 to 69 years.3 Compared with patients younger than 60 years, the risk of postoperative pulmonary complication was doubled for patients between the ages of 60 and 69 years and tripled for patients aged 70 to 79 years.49 Optimization of surgery in the elderly requires recognition of their decreased tolerance for complications. An initial infectious or pulmonary complication is associated with significantly higher risk-adjusted failure-to-rescue rates in elderly patients when compared with younger patients.50 Attending to prevention and early recognition of complications potentially will have great impact on these outcomes in the elderly population.
While surgical risk increases with advancing age, chronologic age alone should not be considered a contraindication to surgery in otherwise healthy patients. The association of age with other factors such as decreased functional status, an increased number of comorbidities, and more complex pharmacologic history51 renders simple conclusions unfeasible. For example, in the Revised Cardiac Risk Index model for cardiac mortality, other risk factors become more significant than age alone,52 while in cases of advanced age, adding age to the revised cardiac risk index (RCRI) model improves its utility for prediction.53
The ACS and American Geriatric Society have published best practice Guidelines for the Optimal Preoperative Assessment of the Geriatric Patient.47 The guidelines highlight 13 geriatric-specific domains to identify and address during the preoperative period (Table 15-7). Areas of focus include traditional preoperative evaluation (cardiopulmonary risk assessment, directed preoperative testing) and added evaluations of cognition (baseline cognition, decision-making ability, depression, risk for postoperative delirium), functional status (mobility, fall risk, and frailty), nutrition, medication (polypharmacy, alcohol/substance abuse screening), support systems (family and social), as well as discussion of goals and expectations of care.
Table 15-7 Perioperative Evaluation of the Geriatric Patient
Few of these domains are covered in detail during traditional preoperative surgical evaluation and further integration into routine practice is needed. Additional recommendations for geriatric patients include screening for cognitive deficits, with referral for support if screening is positive; recognition of risks for postoperative delirium (inadequate pain control, sleep deprivation), avoidance of polypharmacy (specifically avoiding drugs that contribute to delirium); encouraging early return to function with mobilization and physical therapy; and dedicating adequate time for counseling with patient, family, and surrogate decision makers.
Obesity/Obstructive Sleep Apnea
The prevalence of adult obesity is increasing in the United States, with one-third of adults in 2009–2010 considered to be obese (body mass index [BMI] >30 kg/m2).54 Paradoxically, in nonbariatric general surgery, a J-shaped relationship of lower mortality has been reported in overweight and moderately obese patients.55 Nevertheless, an increased risk of postoperative complications and the need for additional operative planning is present for the bariatric patient. For this chapter, discussion centers on risk of airway difficulty and ventilation in the high BMI patient with sleep apnea.
The American Society of Anesthesiologists updated guidelines for perioperative management of patients with obstructive sleep apnea in 2014.56 In this report, the prevalence of sleep-disordered breathing was 9% in women and 24% in men. Patients often present without formal confirmation of the diagnosis by polysomnography. Increases in perioperative risk are noted to occur in proportion to the severity of sleep apnea.57 Sleep apnea should be suspected in patients with history of apparent airway obstruction during sleep, BMI >35 kg/m2, large neck circumference, craniofacial abnormalities, snoring, frequent arousals from sleep, and unexplained daytime somnolence or fatigue. A trial of preoperative continuous positive airway pressure may be attempted to improve nighttime airway obstruction, particularly if obstructive sleep apnea is severe. Other potential interventions include the use of mandibular advancement devices, oral appliances, and preoperative weight loss; however, evidence was judged to be insufficient to make formal recommendations.
Intraoperative considerations include choice of anesthetic technique, airway management, and patient monitoring. For appropriate procedures, regional anesthesia may be preferred over general anesthesia and continuous monitoring should be employed because of the potential for airway obstruction. Patients at risk for obstructive sleep apnea may have associated difficult airway and require special attention during induction and emergence from anesthesia. Sedation and neuromuscular blockade should be carefully monitored and fully reversed prior to extubation.
Table 15-8 METs Equivalents
Anemia is independently associated with mortality in noncardiac surgery,58 yet it is not clear that correction of anemia will alter mortality. Several randomized trials examining transfusion strategies in relatively stable patients failed to demonstrate mortality benefit with higher transfusion triggers in the ICU,59 cardiac60 and orthopedic surgery61 settings. At present, recommendations for perioperative transfusion are general, with consideration for higher thresholds given to patients who are symptomatic with chest pain, hypotension, or fatigue, those with known cardiac ischemia or those receiving massive transfusion for hemorrhagic shock. The use of erythropoietin stimulating agents and intravenous iron supplementation to facilitate perioperative transfusion management remains controversial,62 secondary to increased risk in oncologic and nonanemic patients. Current evidence does not support aggressive preoperative transfusion in patients with sickle cell disease.63 Conflicting evidence has been published regarding the relationship between age of transfused cells and mortality.64,65
Preoperative Functional Status
The impact of preoperative functional status has been recognized for years. Self-reported exercise intolerance (to walking four blocks or two flights of steps) is independently associated with higher risk of perioperative cardiovascular, myocardial ischemia, and neurologic events after elective major noncardiac surgery.66 The relationship is inversely related to the total number of blocks that could be walked or flights of stairs that could be climbed. Functional ability can be quantified in terms of metabolic equivalents (METs), with 1 MET defined as the amount of oxygen consumed while sitting at rest and equal to 3.5 mL O2 uptake/kg/min.67 Functional status can be categorized as excellent, >10 METs; good, 7 to 10 METs; moderate, 4 to 6 METs; and poor, <4 METs. Risk is increased if unable to perform at a level of 4 METs (Table 15-8).28 The ACS recommends a four-question survey of activities of daily living (ADLs) as a screen to identify unsuspected functional compromise in geriatric patients (Table 15-9A).68
Separate from comorbidity, elderly patients may suffer from frailty and sarcopenia.69 In the Medicare population, additional mortality has been related to frailty70 and the relationship has been documented in a broad population, including patients undergoing gastric surgery,71 gynecologic oncology,72 and cardiac surgery.73 CT-based morphomic quantitation of trunk muscle volume is associated with decreases in functional performance on ADLs and independent activities as measured by the Vulnerable Elder Survey74 as well as poorer outcomes in liver transplantation75 and aortic surgery.76 An operational frailty score (Table 15-9B) has been validated in elderly surgical patients and is associated with higher incidence of postoperative adverse events, increased length of hospital stay, and higher likelihood of discharge to a skilled or assisted-living facility.77
Table 15-9A Short Simple Screen for Functional Assessment
Direction of Care
One of the most important discussions between a surgeon and the patient is the understanding of the patient’s desires regarding direction of care. There is little question of aggressive interventions for a healthy patient expecting rapid return to their preoperative level of function after elective surgery. Nevertheless, patients with severe underlying comorbidity or limited life expectancy should have the opportunity to direct overall goals of their care. One approach that incorporates principles of geriatric and palliative care is to frame surgical care for these patients in the context of time-limited trials.78,79 Overall prognosis, patient priorities, and specific milestones defining improvement or deterioration are discussed with the patient and family. A mutual decision for surgical care is crafted, outlining a specific time-limited trial of care and potential actions that could be undertaken at the end of the trial or in the event that complications arise. In this fashion, treatment decisions can be matched to the individual patient’s overall priorities and goals.
Table 15-9B Frailty Scoring (Operational Definition)
Medication management in the perioperative period is subject to general guidelines of practice. The pharmacokinetics of drug administration, distribution, metabolism, and excretion may all be affected by surgery. Major GI procedures limit oral intake and may require alternate routes of administration. Patients receiving volume resuscitation may have increased drug volume of distribution. Fluctuations in hepatic and renal function and addition of multiple new medications may alter drug metabolism and excretion. New drug interactions may lead to increased toxicity; appropriate monitoring by drug levels may improve drug management.
A careful medication history includes prescription and nonprescription medications and nutritional and herbal supplements. Patients may be taking over-the-counter herbal supplements with substantial physiologic effects80 but may not perceive these as being relevant to the conduct of their surgery. In one series, more than 70% of patients failed to disclose herbal medicine use during routine preoperative assessment.81 Herbal supplements are presently not subject to FDA regulation; contents may vary substantially from manufacturer to manufacturer and frank adulteration has been documented.82 Recommendations for commonly taken herbal supplements are summarized in Table 15-10, keeping in mind that variations in supplement content may make a 7-day holding period prudent.83
In the absence of clear evidence-based guidelines for perioperative medication management, a general approach is to continue drugs that would cause withdrawal within the perioperative period, and to discontinue unnecessary drugs or those with potential for adverse events, restarting them as soon as the patient condition permits.83 Medications that are generally continued include β-blockers, clonidine, statins, and selective serotonin reuptake inhibitors (SSRIs). Medications that can be discontinued include diuretics, metformin, and, possibly, angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) (Table 15-11).
Chronic β-blocker therapy should not be held perioperatively, as abrupt discontinuation may lead to rebound tachycardia and hypertension. Patients not previously receiving β-blockers should not have them initiated in anticipation of planned surgery, as this practice has been associated with an increase in cardiovascular events and is no longer recommended. Statins should be continued in the perioperative period when feasible. Other than β-blockers, antihypertensive medications are generally held perioperatively and should be considered individually. ACE inhibitors and ARBs are associated with perioperative hypotension84 and renal risk. Diuretics are associated with chronic volume depletion and electrolyte abnormalities and should be held perioperatively.
Management of anticoagulation is of critical importance to surgeons and the indication for anticoagulation should be weighed against the risk of procedural bleeding. Agents should be held for the minimum interval felt safe after invasive procedure. Patients on chronic long-acting anticoagulation who are undergoing elective operation can use preoperative bridging with unfractionated or low-molecular-weight heparin prior to surgery.85 In the emergent setting, the American College of Chest Physicians (ACCP) has published guidelines covering reversal of vitamin K competitors based on urgency of procedure and risk of life-threatening hemorrhage.86 Reversal of newer direct thrombin inhibitors may not be possible other than waiting for drug clearance; dabigatran may be removed with dialysis; however, rivaroxaban and apixaban are highly protein bound and unlikely to be removed with this modality. In a large series, perioperative aspirin administration was associated with an increased risk of postoperative hemorrhage87; however, a recent meta-analysis concluded that aspirin should not be stopped in the perioperative period unless bleeding risk outweighed thrombotic risk of holding the drug.88 Dual antiplatelet therapy in the setting of coronary stent placement is specifically addressed in current American College of Cardiology/the American Heart Association (ACC/AHA) guidelines28 and is discussed separately in the Cardiac section.
Table 15-10 Potential Interactions for Some Common Herbal Medicines
Perioperative glucose control reduces mortality rate, hospital length of stay, and deep wound infections in cardiac surgery.89 An adjusted dose of short-acting insulin is administered perioperatively, with intravenous insulin infusion intraoperatively if needed. The oral hypoglycemic, metformin, is associated with lactic acidosis and should be held prior to surgery.
Psychiatric medications should be managed on the basis of concern for withdrawal. Abrupt withdrawal of SSRIs may lead to nausea, vomiting, and lethargy, with symptoms seen more quickly in drugs with shorter half-life.90 The use of SSRIs may increase the risk of bleeding, particularly when used in conjunction with nonsteroidal anti-inflammatory drugs (NSAIDs).83 Benzodiazepines may be continued perioperatively. Older agents, including tricyclic antidepressants and monoamine oxidase inhibitors, may precipitate hypertensive crisis in conjunction with indirect sympathomimetics.
Patients receiving daily maintenance buprenorphine will not have sufficient surgical analgesia from usual postoperative opioid dosing regimens. Preoperative consultation with their prescriber and with the anesthesiology pain service is recommended prior to surgery. Alternative strategies, including the use of regional anesthesia and nonsteroidal and nonnarcotic regimens, may facilitate postoperative management.
Immunosuppressant mediations may be reduced or discontinued prior to surgery, depending on the extent and the presence of associated infection, although a systematic review of immunomodulators in patients with inflammatory bowel disease has suggested no clear increased risk of total or infectious complications with azathioprine, cyclosporine, or infliximab.91 Steroid use carries a dose-dependent relationship with an increased risk of postoperative complications, both infectious and overall.92 Meta-analysis of randomized controlled trials suggests that perioperative stress dose steroids are not required unless there is dysfunction of the hypothalamic-pituitary-adrenal axis.93 Anti-VEGF inhibitors are associated with an increased risk of GI perforation and delayed wound healing and their use is not recommended for 28 days before or after surgery or until surgical wounds are healed.
SPECIFIC ORGAN SYSTEM GUIDELINES
Prediction of cardiac adverse events following noncardiac surgery was pioneered in the 1970s1 and risk stratification is most refined in this area.
The current Revised Cardiac Risk Index (Table 15-12)52 is based on risk of surgery, history of ischemic heart disease, history of congestive heart failure, history of cerebrovascular disease, preoperative insulin-requiring diabetes mellitus, and preoperative renal insufficiency and can be used to identify patients at higher risk for major cardiac complications (myocardial infarction [MI], pulmonary edema, ventricular fibrillation, primary cardiac arrest, complete heart block). The RCRI has been validated as a predictor of increased risk, although, disappointingly, targeted interventions in high-risk groups have not yet conclusively shown benefit. In 2007, evaluation of ACS NSQIP data in 211,410 patients demonstrated a 0.65% rate of perioperative myocardial infarction or cardiac arrest. Five predictors were identified: type of surgery, dependent functional status, abnormal creatinine, ASA class, and increasing age.26 Active angina and signs of left ventricular dysfunction, such as dyspnea or worsening heart failure; syncope; significant arrhythmias; severe valvular disease; and surgery should prompt further preoperative evaluation.
The ACC and the AHA issued joint recommendations for perioperative cardiovascular evaluation and management for noncardiac surgery in 2014.28 The guideline covers preoperative evaluation, risk assessment, perioperative management, and post–percutaneous coronary intervention (PCI) recommendations. Outcomes are defined in terms of a composite Major Adverse Cardiac Event (MACE), which includes death or MI. Low-risk procedures are defined as carrying a predicted MACE risk of <1% and elevated-risk procedures carrying a MACE risk of ≥1%.
Stable patients undergoing low-risk surgery and those with good exercise tolerance rarely need further cardiac evaluation. For those at higher risk, cardiology evaluation is recommended. In urgent or emergent situations where surgical intervention must proceed prior to any meaningful management changes, delay for cardiac evaluation will not be of benefit and care is focused on prompt recognition and treatment of any perioperative events. In this paradigm, decision-making for higher-risk patients can be simplified to (1) initial determination if the procedure can be delayed to permit additional evaluation and management which could significantly alter the outcome and (2) further risk stratification and intervention if the patient condition permits. The overall ACC/AHA algorithm is reproduced in Algorithm 15-1 and specific recommendations are summarized later.
Table 15-12 Revised Cardiac Risk Index