Postoperative complications and mortality: Are they unavoidable?

As surgery inherently involves injury to the human body as a result of its invasive nature, it may result in postoperative complications.1 Further, in patients who do not recover from the morbidity, worsening of their condition could ultimately result in mortality. Although numerous efforts have aimed at reduction in surgical mortality, zero postoperative mortality has been achieved in only a very limited number of hospitals. Most of the reports of zero mortality involved retrospective studies in single centers. However, as the results of small studies are believed to be less reliable, it is difficult to explore universal best practices. Furthermore, even a hospital that has achieved zero postoperative mortality at some point, might encounter operative death at a later date owing to expansion of operative indications. Therefore, the results of risk analysis carried out at a particular hospital often may not be valid in other hospitals. Recently, several multicenter nationwide cohort studies on risk factors for operative morbidity and mortality, consisting of over thousands of patients, have been published. In Japan, an extensive nationwide database of clinical factors from surgical patients (National Clinical Database [NCD]) was launched in 2011. This database covers about 1.2 million cases a year, accounting for approximately 95% of general surgical cases.2 The total number of cases accumulated exceeds 4 million cases from 3745 hospitals.3 A total of 1 494 934 cases underwent 115 types of gastroenterological surgeries from 2011 to 2013. The 30-day mortality was 1.5% (n = 22 334), and 90-day mortality was 2.8% (n = 42 048). Of them, 14 000 patients died of postoperative complications per year. In addition, several nationwide database studies have been reported, such as that of the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP).4 Such studies have provided much reliable information from all over the world on risk analysis for postoperative complications and deaths, which is difficult to obtain from single-center retrospective studies. Surgical morbidity rates vary according to the type of procedure (Table 1).5-23 The lowest morbidity rate of 5.6%-7.2% was observed with laparoscopic cholecystectomy for acute cholecystitis, whereas the highest rate of 54.4% was observed with pancreaticoduodenectomy (PD). Operative mortality showed the same tendencies. The lowest mortality of 0.4%-0.5% was seen with laparoscopic cholecystectomy, and the highest with esophagectomy, hepatectomy and PD.6, 17, 21, 22 Interestingly, mortality rates were approximately one-tenth the morbidity rates for most operative procedures. The fact that one-tenth of patients with morbidities show a decline in their condition, resulting in death, might imply hidden factors related to vulnerability to surgical stress, such as single nucleotide polymorphisms (SNP) of inflammatory cytokines, neutrophil activation, and immunoglobulins. Elderly patients have double the rate of postoperative complication/mortality ratio than non-elderly patients (20.1% vs 10.8%) 14(Table 1). This disparity in mortality may indicate that elderly patients are at greater risk for difficulty of recovery from severe complications. However, low anterior resection for rectal cancer represents a deviation from the above-mentioned "10% rule". Its morbidity and mortality rates in the NCD study were 26.3% and 0.9%, respectively. The ACS NSQIP also indicated a similar tendency, with a morbidity of 30.1% and a mortality of 1.1%.6 This may be explained by the use of diverting ileo- or colostomy, a potentially effective evacuation of severe abdominal abscess as a result of anastomotic leakage. 4.0 (local) 3.5 (systemic) Numerous risk factors for postoperative morbidity and mortality have been reported, which can be broadly classified as patient factors (age, Charlson comorbidity score, hypoalbuminemia etc.), therapeutic factors (preoperative chemotherapy, blood transfusion etc.), and hospital factors (number of operative cases, participation of board-certified surgeons etc.). With regard to hospital factors, quality of perioperative care is associated not only with the skill of the surgeon, but also with that of other medical staff. Recovery from postoperative complications, in particular, involves a comprehensive surgical team including non-surgical staff, such as interventional radiologists, intensive care staff, and the infection control team. High-risk operations have a decreased rate of postoperative complications when carried out at high-volume centers (HVC). The rates of most postoperative complications varied nearly 2-fold across hospital volume groups.24 For example, a Danish nationwide database study reported on hospitals that were categorized by the number of PD carried out annually as fewer than 5, between 5 and 10, 11-19, and 20 or more, and were accordingly referred to as very low-, low-, medium-, and high-volume centers, respectively. During the 6-year study period, mortality rates for very low-, low-, medium-, and high-volume centers were 14.7%, 9.8%, 6.3%, and 3.3%, respectively (P < .001).17 Even in HVC, operative morbidity and mortality rates are significantly different between hospitals. The Japanese Society of Hepato-Biliary Pancreatic (HBP) Surgery has launched a system for recognition of institutions where more than 50 high-level HBP surgeries are carried out annually as board-certified A training institutions (A training institutions), and for recognition of institutions where between 30-50 high-level HBP surgeries are carried out annually as board-certified B training institutions (B training institutions). There is a significant difference between A and B training institutions in 90-day mortality rates for right hepatectomy or extended right hepatectomy (2.2% vs 3.6%) and PD (1.1% vs 1.5%), and in overall mortality rate for all HBP surgeries.25 Furthermore, operative mortality varied from 0% to over 5%. Although this study provides valuable feedback, it provides little insight into why mortality rates vary to such a great extent. Elucidating the mechanisms underlying these variations is important for the development of more effective preventive strategies. In addition, board-certified training institutions are obliged to provide annual reports on their surgical cases and surgical mortality cases. If the surgical mortality rates exceed 5%, board-certified training institutions must submit case reports and are required to undergo an audit of medical and surgical records of deceased patients. Interestingly, the number of hospitals with a 90-day mortality of over 5% has decreased from 30 hospitals in 2012 to 8 hospitals in 2015 after the introduction of this system. This might mean that this system serves as a deterrent against operative mortality by promoting appropriate patient selection. There are three possible countermeasures for reduction of operative mortality. The first is appropriate patient selection using risk calculators constructed by including many possible factors that affect the incidence of postoperative morbidities, such as patient age, nutritional status, and comorbidities, as with the P-POSSUM score.26 However, as there is still a tendency to overestimate the risk despite these objective scores, strict application of the risk calculator may deter performance of major surgeries even in patients who can be expected to be cured by the surgery. The NCD also provides a risk-calculation service. However, operative indications and surgical procedures have been changing over the years. Therefore, it is mandatory to revise risk calculators based on further information from nationwide databases. Second, several preventive interventions to reduce operative morbidity and mortality have been reported, such as immunonutrition, synbiotic administration, early removal of surgical drains, rehabilitation, and oral care before major surgery. Nutritional interventions during neoadjuvant chemotherapy might also improve postoperative course. Third, the method of recovery from postoperative morbidities once it occurs is important. Some investigators reported that hospitals with high rates of complications do not necessarily also have high mortality rates.27 Recovery rates from postoperative complications vary between hospitals. Actually, Ghaferi reported similar complication rates at the worst and best hospitals (36.4% vs 32.7%), although "failure to rescue" rates were much higher at the worst compared with the best hospitals (16.7% vs 6.8%).27 This indicates that high-mortality hospitals might not be as proficient at recognizing and managing serious complications. Adequate training material on early detection of postoperative complications and recovery strategies in patients with deteriorating status, such as guidelines for perioperative care, edited by the best hospitals, may contribute to reduction of mortality. As approximately one-tenth of patients with morbidity ultimately succumb to their illness, surgical teams should devote every effort to reduction of operative morbidity and mortality. Patient selection based on appropriately updated risk calculators is mandatory. Furthermore, availability of guidelines for prevention and rescue edited by the best hospitals on nationwide clinical databases might be useful. Moreover, establishment of reliable strategies that are validated by the nationwide clinical database may encourage surgeons to proceed with more aggressive procedures when appropriate. Conflict of Interest: Authors declare no conflicts of interest in relation to this article.