Absence of Association between Preoperative Estimated Glomerular Filtration Rates and Postoperative Outcomes following Elective Gastrointestinal Surgeries: A Prospective Cohort Study

Background Preoperative risk stratification and optimising care of patients undergoing elective surgery are important to reduce the risk of postoperative outcomes. Renal dysfunction is becoming increasingly prevalent, but its impact on patients undergoing elective gastrointestinal surgery is unknown although much evidence is available for cardiac surgery. This study aimed to investigate the impact of preoperative estimated glomerular filtration rate (eGFR) and postoperative outcomes in patients undergoing elective gastrointestinal surgeries. Methods This prospective study included consecutive adult patients undergoing elective gastrointestinal surgeries attending preassessment screening (PAS) clinics at the Queen Elizabeth Hospital Birmingham (QEHB) between July and August 2016. Primary outcome measure was 30-day overall complication rates and secondary outcomes were grade of complications, 30-day readmission rates, and postoperative care setting. Results This study included 370 patients, of which 11% (41/370) had eGFR of <60 ml/min/1.73 m2. Patients with eGFR < 60 ml/min/1.73 m2 were more likely to have ASA grade 3/4 (p < 0.001) and >2 comorbidities (p < 0.001). Overall complication rates were 15% (54/370), with no significant difference in overall (p=0.644) and major complication rates (p=0.831) between both groups. In adjusted models, only surgery grade was predictive of overall complications. Preoperative eGFR did not impact on overall complications (HR: 0.89, 95% CI: 0.45–1.54; p=0.2). Conclusions Preoperative eGFR does not appear to impact on postoperative complications in patients undergoing elective gastrointestinal surgeries, even when stratified by surgery grade. These findings will help preassessment clinics in risk stratification and optimisation of perioperative care of patients.


Introduction
With increasing ageing population, chronic kidney disease (CKD) is becoming increasingly prevalent [1,2]. CKD is recognised as a risk factor for cardiovascular morbidity and mortality in the general population [3,4]. A recent metaanalysis estimated the global rates of CKD to be 13%, with majority of patients having stage 3 CKD (i.e., eGFR 30-60%) [5,6]. In parallel, there are rising numbers of patients undergoing elective surgery, and the impact of renal dysfunction on postoperative outcomes following surgery remains unclear. Current guidelines from the Association of Anaesthetists of Great Britain and Ireland (AAGBI) do not recommend perioperative risk stratification and management of patients with reduced estimated glomerular filtration rate (eGFR) undergoing surgery, since evidence in this area is limited.
Much data are available in the context of cardiac and vascular surgeries to demonstrate association between renal dysfunction and postoperative outcomes following cardiac and vascular surgeries. A recent systematic review and meta-analysis of 46 studies demonstrated a threefold increased risk of 30-day postoperative complications and acute kidney injury (AKI) in patients with eGFR <60 ml/min/1.73 m 2 undergoing cardiac and vascular surgeries [7]. Furthermore, this review also demonstrated that patients with eGFR < 60 ml/min/1.73 m 2 had increased risk of major cardiovascular events and allcause mortality during long-term follow-up.
With strong, convincing evidence available for cardiac and vascular surgeries, this is however limited and nondefined for patients undergoing gastrointestinal (GI) surgeries. Recently, two studies evaluated the impact of preoperative renal dysfunction on mortality and cerebrovascular events in patients undergoing noncardiac surgeries [8,9]. Although both studies included a large proportion of patients with gastrointestinal (GI) and hepatopancreatobiliary (HPB) surgery, these studies did not evaluate the impact of renal dysfunction in this subgroup of patients. To further define the role of preoperative renal dysfunction in GI and HPB surgery, this study sought to explore the relationship between preoperative eGFR levels and overall complication rates of patients undergoing elective GI and HPB surgeries.
is study will provide important information for preoperative counselling on risk of postoperative complications of patients undergoing elective GI and HPB surgeries.

Methods
is prospective study identified consecutive adult (≥18 years) patients attending preassessment screening clinic (PAS) for GI and HPB surgeries from August 2016 to September 2016 at the Queen Elizabeth Hospital Birmingham (QEHB). Eligible procedures were those involving elective surgery on any part of the gastrointestinal tract or biliary tree, involving a hospital admission with an overnight stay. Patients undergoing daycase urological, gynaecological, vascular, or transplant procedures were excluded. is study was registered and approved by the local audit department. Patients' medical records were reviewed from the online patient notes, and the data were extracted on to a uniform database (Microsoft ® Excel 2010) that was designed to include all relevant details pertinent to this study. ree coauthors were involved with data extraction (Sivesh K. Kamarajah, Behrad Barmayehvar, and Mustafa Sowida), and the data were then validated for accuracy by an independent fourth coauthor (Amirul Adlan).

Preassessment Clinics.
At QEHB, all patients undergoing surgical procedures are referred by the surgeon to dedicated preassessment clinics according to surgery grade and comorbidities. Clinics are divided into low risk and high risk; numerically, these correspond to levels 1 and 2A and levels 2B and 3. Low-risk clinics are led and delivered by trained preassessment nurses, whereas high-risk clinics are led and delivered by more experienced nursing staff and consultant anaesthetists. Currently, there is no systematic way of allocating patients into these clinics as patients are risk-assessed individually by the surgeons.

Main Explanatory
Variable. eGFR as a measure of renal function was the main explanatory variable. e eGFR was estimated using the Modification of Diet in Renal Disease-4 (MDRD-4) equation, which is one of the accepted and validated methods of evaluating kidney function used worldwide [10]. e unit used for the eGFR variable was ml/min/1.73 m 2 . Preoperative renal dysfunction was defined by using the cutoff eGFR value of ≤60 ml/min/1.73 m 2 , as it is the common cutoff that has been widely used and reported in the literature [7].

Explanatory Variables.
Explanatory variables were collected to provide a risk-adjusted estimate. Variables were predefined and selected based on clinical plausibility. To account for comorbidities, both the American Society of Anesthesiologists (ASA) fitness grade and the number of comorbidities (0 versus 1-2 versus >2) were measured. e ASA grade considers disease severity and is a reliable metric for the measurement of postoperative mortality and complications [11]. Grade of surgery is a category that indicates a combination of complexity and amount of tissue injury in the surgical procedure. Exact definitions used are similar to that used in a recent publication using definition from the European Surgical Outcomes Study, and these are provided in Supplementary Table 1 [9]. Surgical approach was defined as open, laparoscopic, or endoscopic/ultrasound (for minor surgical grade only).

Outcome Measures.
e primary outcome in this study was the 30-day overall complication rates. Secondary outcome measures were grade of complications according to the Clavien-Dindo classification system [12], 30-day readmission rates, and postoperative care setting. According to the Clavien-Dindo classification system, complications were graded from grade I (any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic, and radiological interventions) to Grade V (death of patient following surgery). is allows comparison of the grades of complications in patients with and without renal dysfunction.

Statistical Analysis.
Continuous variables were expressed as mean ± standard deviation or median (interquartile range) and analysed using t-test or Mann-Whitney test, where appropriate. Categorical variables were expressed as percentages and analysed using chi-square test or Fisher's exact test, where appropriate. Univariate logistic regression and multivariate logistic regression were used to determine the strength association between risk factors for postoperative overall complications. Clinically, important variables with p < 0.25 on univariate analysis were entered into multivariate analysis. e Hosmer-Lemeshow goodness-of-fit test and c-statistics were used to evaluate the model's performance and discriminative ability. Results are presented as hazard ratios (HRs) with 95 percent confidence interval (CI 95% ). In all analyses, a p value of <0.05 was maintained as statistically significant. Data analysis was undertaken using R Foundation Statistical Software (R 3.2.1, R Foundation for Statistical Computing, Vienna, Austria).
A flow diagram of patients included in the study is presented in Figure 1.
e baseline characteristics of included patients are presented in Table 1. Patients with eGFR ≤ 60 ml/min/1.73 m 2 were significantly older than patients with eGFR > 60 ml/min/1.73 m 2 (69 versus 54 years; p < 0.001).
ere were also no significant differences in the length of hospital stay between the two groups.
Binary multivariate logistic regression was used to identify factors predictive of overall complications in this cohort ( is nonsignificance remained even when the cohort was stratified by surgical grade. Analyses were also repeated with eGFR cutoff as ≥45 ml/min/1.73 m 2 (Supplementary Table 1). However, there were no significant differences in the rate of postoperative complications between patients with eGFR ≥45 ml/min/1.73 m 2 and eGFR <45 ml/min/1.73 m 2 (15% versus 0%; p � 0.338).
ere were also no differences in the 30-day readmission rates and length of hospital stay (Supplementary Table 2).

Discussion
is prospective single-centre cohort study demonstrated no association between preoperative renal dysfunction and postoperative overall complication following elective GI and HPB  surgeries. Instead, high surgical grade and the presence of T2DM were significant predictive factors of the 30-day overall complications. is demonstrates that risk stratification of patients according to eGFR may not be warranted from this study.
Although the association between chronic kidney disease and adverse postoperative outcomes has been described previously, most research have focused on high-risk patient groups where specific risk factors for renal dysfunction are Anesthesiology Research and Practice 5 more prevalent [7,13]. Recently, a number of observational studies have examined preoperative eGFR and postoperative survival in noncardiac surgery [8,9,[14][15][16]. e largest of these studies examined a retrospective cohort of over 250,000 patients undergoing noncardiac surgery from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQUIP) data sets for 2005-2007 [14]. is study demonstrated that CKD and mineral bone disorders in patients undergoing general and vascular surgeries were independent predictors of early postoperative outcomes. However, this study included a broad range of procedures which may limit its applicability to GI and HPB surgeries. Furthermore, more recent evidence suggests that proteinuria may be a more important marker than eGFR for risk stratification of CKD [17,18].
Despite these varying evidences demonstrating impact of eGFR, preassessment services do not consider eGFR as a risk factor during triage. Current literature is limited in the context of preassessment service in the UK and worldwide, advocating that a nurse-led and consultant/specialist-led PAS clinic is feasible but this remains nondefined considering the range of surgical specialities [19][20][21]. At QEHB, this two-tier clinic was recently introduced to allow assessment of patients by risk groups based on their comorbidities and ASA grade. is service allows risk assessment of patients for their comorbidities such as diabetes and cardiovascular disease prior to surgery, which may explain the nonsignificant findings in postoperative complications in patients with and without renal dysfunction. e impact of renal dysfunction in normal physiology is often complex. Clinically, studies have demonstrated that patients with CKD are associated with higher rates of comorbidities such as T2DM, hypertension, hyperlipidemia, and cardiovascular disease, as is also seen in this study [4,22]. Despite this, these diseases are not solely responsible for the adverse effects of CKD since strong associations remain after accounting for other factors on multivariable analysis from previous studies in patients undergoing cardiac surgeries [3,4]. Furthermore, patients with CKD may have other risk factors such as elevated levels of inflammatory markers, homocystinemia, albuminuria, and elevated levels of uric acid which are often difficult to account for in studies [23,24]. Alterations in these homeostatic mechanisms and others likely complicate normal recovery from abdominal surgery, specifically major surgeries.
One of the limitations of this study is that this is a singlecentre study with a relatively small sample size. It was difficult to provide a formal sample size estimate as the complication rates of patients with and without preoperative renal dysfunction undergoing GI and HPB surgery is relatively unknown. However, this study provides early baseline data on 30-day overall complication rates between these two groups. Furthermore, this study only included patients undergoing elective procedures, and hence results may not be comparable to patients undergoing emergency GI and HPB surgery.

Conclusion
Preoperative eGFR level did not impact on postoperative overall complication rates even when stratified by surgical grades. Findings from this study suggests that the inclusion of patients with eGFR of <60 ml/min/1.73 m 2 in high-risk PAS clinics for risk assessment based purely on eGFR alone for elective GI surgeries may not be warranted. Larger multicentre prospective studies should aim to better define the stage of impairment in renal function in the groups and assess the level of risk accordingly. is can potentially lead to a risk stratification scoring system to be used by clinicians in the preoperative care setting, which would help in identification, prioritisation, and optimising management of high-risk patients.

Conflicts of Interest
e authors declare that they have no conflicts of interest.