Objectives: Acute kidney injury is a significant cause of morbidity after orthotopic liver transplant. Early extubation after liver transplant may have a beneficial effect on postoperative renal function. This may be
the result of reduction in vasopressor-mediated vasoconstriction used to counteract the hypotension associated with sedative use and the effects of positive-pressure ventilation. Previous studies explored advantages of early extubation after liver transplant but focused on resource usage rather than clinical benefit. This study was designed to determine the association between fast-track extubation and reduction in postoperative vasopressor requirement and whether this had any association with acute kidney injury incidence or renal replacement therapy requirement.
Materials and Methods: Data were collected from 144 orthotopic liver transplants. A propensity-matched case-control analysis was conducted on a subgroup of 33 patients who were fast-track extubated and with 33 propensity score-matched control patients who were not. The primary outcome was median days of postoperative vasopressor use, and secondary outcomes included incidence of acute kidney injury, renal replacement therapy requirement, and critical care admission duration.
Results: The fast-track extubation group had a shorter postoperative vasopressor requirement (0 vs 2 days; P < .01) and a reduced need for renal replacement therapy (3% vs 21.2%; P = .05). Median critical care admission duration (3 vs 4 days; P = .03) and hospital admission duration (14 vs 19 days; P = .04) were shorter in the fast-track extubation group.
Conclusions: This is the first study to reveal a significant association between fast-track extubation and reduced postoperative vasopressor requirement. Additionally, this was associated with a trend toward reduced renal replacement requirement after liver transplant. It suggests that early extubation may not just be a resource benefit to an institution but may convey a clinical benefit to patients through a reduction in organ failure and requirement for organ support.
Key words : Early extubation, Noradrenaline, Renal failure
Acute kidney injury (AKI) is a serious yet common complication that affects up to 70% of patients after orthotopic liver transplant (OLT), and as many as 25% of patients may require renal replacement therapy (RRT).1,2 Rapid initiation of tracheal extubation after OLT has a number of physiological benefits that theoretically support renal perfusion, glomerular filtration, and kidney function.3-5 One important mechanism postulated for this includes the reduction in vasopressor-mediated renal vasoconstriction that is required to treat the hypotension associated with sedative use and the effects of positive-pressure ventilation on normal physiology. Early extubation also reduces the risk of sepsis-driven hypotension associated with prolonged intubation in patients who receive high-dose immunosuppression.6,7
Over the past decade, several centers have demonstrated that, in selected subgroups,8-10 tracheal extubation immediately after OLT is safe and can be associated with significant cost and resource benefits,4,11-20 especially when combined with enhanced recovery protocols.21-23 However, the emphasis on early extubation has always been demonstration of safety compared with prolonged ventilation8 or focus on critical care and hospital length of stay or cost, rather than determination of physiological or clinical benefits. It has been demonstrated in patients who have undergone cardiac surgery that postoperative duration of intubation may be an independent risk factor for AKI.24,25 However, no study has investigated a possible independent association between early extubation, postoperative vasopressor requirement, the development of AKI, or the need for postoperative RRT in the liver transplant population.
This study was designed to assess whether fast-track extubation (FTE, within 4 hours of the end of operation) was associated with reduction in the duration of vasopressor use for critical care. Secondary outcomes included the association of FTE with a reduction in (1) the incidence of stage ≥1 Acute Kidney Injury Network (AKIN) AKI criteria after OLT, (2) the requirement for RRT in the first postoperative week after OLT, and (3) the duration of critical care and hospital admission.
Materials and Methods
Previous data from our institution had shown an increase in the number of patients with extubation in the immediate postoperative period after OLT. Data were collected from all adult patients who underwent OLT from May 2016 to January 2018 at the Royal Free NHS Foundation Trust. Recipients who received status 1, domino, or dual transplants, as well as those who were never extubated (died before extubation or received a tracheostomy before extubation) or required ventilatory or renal replacement support pretransplant, were excluded from the study. For propensity score matching of patients for likelihood of FTE, binomial logistic regression analysis was performed with backward conditional elimination to generate a model to predict the probability of FTE. From these data, a propensity-matched case-control study was then conducted to compare patients who underwent FTE after OLT versus those who did not. As all data were collected retrospectively and anonymized, the need for ethical approval or individualized consent was not required by the Royal Free NHS Foundation Trust Research and Development Office, in accordance with the Helsinki Declaration of 1975.
Fast-track extubation was defined as extubation immediately after OLT in operating theater or within 4 hours of the end of surgery. All patients in our institution, regardless of whether or not they were extubated in the operating room, were admitted directly to critical care postoperatively. The attending anesthetist was charged with the decision to extubate the patient at the end of the operation (in theater) or to extubate within 4 hours after the end of the operation, based on a FTE protocol (Figure 1). At our institution, standard first-line sedation for those arriving intubated to critical care consisted of propofol (0.25-1 mg/kg/h) and fentanyl (1-3 μg/kg/h) infusions. First-line vasopressor regimen both in theater and on critical care admission was standardized to a norepinephrine infusion (0.01-1 μg/kg/min). Extubation failure was defined as reintubation for a physiological deterioration (respiratory, cardiovascular, metabolic, neurological) as opposed to a return to theater for a purely surgical complication.
Acute kidney injury was diagnosed according to the AKIN26 criteria and was defined as AKIN stage 1 or higher. Chronic kidney disease was judged according to estimated creatinine clearance (CrCl), as calculated with the Cockcroft-Gault equation26: CrCl = ([140 - age] × IBW)/(serum creatinine × 72) × (0.85 if female), where creatinine clearance is in milliliters per minute, age is in years, ideal body weight (IBW) is in kilograms, and serum creatinine is in milligrams per deciliter.
Data were assessed for normality with the Kolmogorov-Smirnov test with the Lilliefors significance correction. All metrics other than the score for Acute Physiology, Age, Chronic Health Evaluation II (APACHE II) were nonparametric. The t test, the Mann-Whitney U test, and the chi-square test were applied, as appropriate. An a priori sample size calculation was performed with a 2-tailed Wilcoxon signed-rank test for matched pairs. The test was performed with a reduction in the primary outcome of duration of vasopressor requirement by 1 day, from 2.34 days in the non-FTE patients (SD 2.06 days). With an alpha of .05, a power of 0.95, and a presumed correlation of 0.5, this generated a sample size calculation of 60 patients, with 30 in each group.
To minimize the effects of variables, binary logistic regression analysis was performed with backward conditional elimination to generate a model to predict the probability of a patient undergoing FTE. Only patient, intraoperative, and postoperative risk factors that were independently associated with FTE were retained in this model. This model probability was then used to propensity-match FTE patients and non-FTE patients based on their likelihood of undergoing FTE. This model probability was used to match patients who underwent FTE to those with a comparable probability of FTE, and hence suitability for FTE, but who were not extubated within 4 hours. The patient matches were performed manually, and all patients were matched to within 10% FTE probability.
Preoperative factors for the recipient analyses were as follows: patient sex, body mass index (calculated as weight in kilograms divided by height in meters squared), Model for End-Stage Liver Disease (MELD) score, United Kingdom End-Stage Liver Disease (UKELD) score, liver disease etiology, international normalized ratio, serum bilirubin, serum creatinine, and derived creatinine clearance. The analyzed donor information included the determinants of the donor risk index score,27 which are donor age, height, ethnicity, and cause of death; donation type (after brainstem or cardiac death); graft type (partial or full), location, and cold ischemic time; and graft steatosis (none, mild, moderate, severe).
The following intraoperative information was analyzed: surgical technique (piggyback vs caval replacement) (note that venovenous bypass was not in use at our institution), total operative time, and units of transfused red blood cells. The postoperative variables included the following: APACHE II score on intensive care unit (ICU) arrival; time to extubation from the end of the surgery; creatinine level 24 hours after ICU arrival; vasopressor requirement beyond postoperative day 1; RRT requirement in the first postoperative week; and creatinine values at days 3, 5, and 7 after ICU arrival.
The final model only incorporated variables that were independently associated with each outcome.
There were 187 OLTs conducted from May 1, 2016, to January 1, 2018. There were 16 sets of incomplete notes, and these were excluded from our analyses. Of the 171 remaining patients, 22 were excluded because they received status 1 transplants (14 after primary status 1 transplants and 8 after failed primary transplants), 2 were never extubated, 2 received domino transplants, and 1 underwent a dual-organ transplant.
This left 144 patients from which to generate the propensity-matched case-control groups; 43 of these 144 patients (30%) were extubated within 4 hours. Of these FTE patients, 33 were matched to 33 control patients with a comparable probability (within 10%) of FTE but who were not extubated within 4 hours. All experimental cases and control cases were deceased donations rather than living donations. All 144 patients survived to the time of critical care discharge; however, 4 patients died before hospital discharge. None of these patients was extubated within 4 hours, and none was included in the propensity-matched control group. Baseline characteristics of the entire group of 144 patients on whom data were collected are shown in Table 1.
The baseline characteristics of the 66 patients included in the propensity-matched case-control study are shown in Table 2. The predicted probability of FTE was not statistically different between the 2 groups (40% vs 43%; P = .878). The only significant baseline difference between the case and control groups was the median extubation time (0 vs 19 hours; P < .01). No patient in either case or control group failed the initial trial of extubation.
Outcome comparisons of the 2 groups are shown in Table 3. The median value for days of vasopressor use was significantly lower in the FTE group compared with the control group (0 vs 2 days; P < .01). Although the incidence of AKIN stage ≥1 at days 1, 3, 5, and 7 was statistically similar between groups, the need for RRT in the first postoperative week showed a trend toward a lower requirement in the FTE group (21% vs 3%, absolute risk reduction 18%; P = .05). In addition, median length of ICU stay was reduced by 1 day (3 vs 4 days; P = .03) and median hospital length of stay was reduced by 5 days (14 vs 19 days; P = .04) in the FTE group.
This case-control analysis demonstrates that FTE is associated with a significant reduction in vasopressor use within the first week after OLT with a reduction in median ICU and hospital admission duration. Requirement for RRT was less in the FTE group, although this did not reach statistical significance. A significant difference in the incidence of AKIN stage 1 or stage 2 AKI between the 2 groups was not demonstrated.
Association between fast-track extubation, vasopressor requirement, and renal function
Despite the high incidence of AKI and RRT postoperatively after OLT, no published study has yet investigated an association between early extubation after OLT and the incidence of AKI or the need for RRT. Alcaraz and colleagues (abstract)28 demonstrated in a retrospective analysis of 113 patients that immediate postoperative extubation was associated with a lower acute renal failure rate (40% vs 68.8%; P = .04) compared with extubation in critical care. However, there was no further delineation of the exact timing of critical care extubation in their control group, and so it was difficult to interpret their results. One mechanism postulated for the development of AKI in intubated patients includes an increase in intrathoracic pressure that leads to a reduction in right ventricular function and cardiac output, as well as hypotension and poor renal perfusion.24 Although our study did not examine mean arterial pressures in recipients, the need for maintenance vasopressors to mitigate hypotension may support our hypothesis that prolonged intubation and consequent sedation may have an adverse effect on renal perfusion and function. We suggest that future trials to study the association between FTE and postoperative renal function should also investigate sedative-free hours in addition to the need for vasopressor.
Although there was a trend in reduction of RRT requirement in the FTE group, this did not reach statistical significance. Interestingly, this did not correspond to any reduction in incidence of AKI between the groups. There may be a number of reasons for this. First, our study was small and not adequately powered to demonstrate these differences because there were low numbers of patients with postoperative AKIN stage 3 AKI in our data set. Second, RRT in itself may have been a confounder in the analysis. As the timing of RRT was not examined in our analysis, there may have been potential for early filtration for those developing immediate (or judged clinically to have high risk of postoperative renal failure) to lower postoperative creatinine values, precluding demonstration of an association between FTE and AKI. Third, there may have been post-operative reasons for RRT commencement other than an increase in creatinine (thereby fulfilling our AKIN definition of AKI). For example, poor liver graft function and consequent severe acidosis may have been an indication for RRT. Previous studies have suggested an association among postoperative graft blood flow, graft oxygenation, and positive-pressure ventilation.3-5 By noting a trend between FTE and reducing need for RRT but no relationship between FTE and AKI, our results suggest that FTE may convey a measurable clinical benefit to the new graft rather than simply the preservation of renal function. However, this is hypothetical, and larger trials will be needed to confirm this assertion.
Our case-control analysis revealed that FTE was associated with a reduction in median critical care length of stay, which supports the view that early extubation can lead to improvements in resource utilization. The reason for this may be twofold, that is, the reduced requirement for vasopressor as well as improved renal and hepatic graft function and thereby the reduced need for organ support. This seems to lead to overall shorter postoperative hospital stays. Our study is the first to investigate any clinical benefit of early extubation after FTE and will provide a basis for larger trials to investigate organ failure after OLT.
This was a retrospective observational study, and therefore the authors can only state association rather than causation between the variables. Although there were 43 patients extubated within 4 hours in our cohort of 144 patients, it was only possible to match 33 patients adequately to 33 control patients in order to perform the case-control analysis. Although this allowed for greater accuracy of matches and comparisons, overall numbers in the case-control study were small. In addition, despite adequate power to demonstrate our primary outcome, the study did not have adequate power to draw definite conclusions regarding the link between FTE and secondary outcomes. Binomial logistic regression was performed with a variety of perioperative data and donor information, but, as with any regression analysis, there may have been relevant factors that were not included.
The baseline characteristics of our case-control subgroup (Table 2) showed that, for the variables included, the 2 groups were matched appropriately. We argue that our model has produced 2 groups that are similar (apart from the time to extubation) and therefore comparable. We do acknowledge, however, that the exclusion criteria of the study may disallow the application of these findings to complex liver transplants.
Our study supports the view that FTE is safe, is an improvement in resource utilization, and is clinically beneficial to the liver transplant recipient in terms of the level of required postoperative organ support. This is the first study to demonstrate a significant association between FTE, postoperative vasopressor requirement, and a trend toward reduction in RRT after OLT. This study will form a basis for larger trials to investigate the clinical benefits rather than just the resource benefits of early extubation after OLT.
Volume : 19
Issue : 4
Pages : 339 - 344
DOI : 10.6002/ect.2020.0422
From the Royal Free Perioperative Research Group, Royal Free NHS Foundation Trust, Anaesthetics Department, Hampstead, London, United Kingdom
Acknowledgements: We thank P. Tsang and H. Chang for help with medical record collection. J. Fabes received an academic clinical fellowship from the National Institute for Health Research during this study. Other than described, the authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no further declarations of potential conflicts of interest.
Corresponding author: Ravi Bhatia, Royal Free Perioperative Research Group, Royal Free NHS Foundation Trust, Pond Street, Hampstead, London NW3 2QG, UK
Phone: +44 020 7794 0500 ext. 36503
Figure 1. Protocol for Fast-Track Extubation After Orthotopic Liver Transplant
Table 1. Baseline Characteristics for Whole Study Population
Table 2. Baseline Characteristics for Case-Control Subgroup
Table 3. Case-Control Comparison