Abstract
There is a paucity of outcome data for critically ill patients with combined acute liver and kidney injury secondary to paracetamol overdose (POD) requiring renal replacement therapy (RRT). We retrospectively reviewed all admissions over a 6-year period to the intensive care unit (ICU) at a university teaching hospital which supports an active liver transplant program. Of the 5582 admissions over this period, 73 patients were admitted with combined liver and kidney injury requiring RRT, and of these 10 patients went on to receive a liver transplant. Overall mortality was 58%, being lower at 20% for transplant recipients. Transplant recipients were younger than non-transplanted patients with similar global disease severity scores [Model for End-Stage Liver Disease (MELD) and Acute Physiology and Chronic Health Evaluation II (APACHE II)]. Patients with a higher MELD or APACHE II score fared worse and patients fulfilling the King’s College Hospital transplant criteria on admission had an odds ratio (OR) for death of 3.8 (1.3–10.6). Logistic regression modeling found that only a higher admission bilirubin OR 1.6 (1.1–2.3) mg/dL and a lower creatinine OR 0.52 (0.3–0.9) mg/dL were predictive of mortality. Of the ICU survivors, 41% remained RRT dependant at the time of ICU discharge; all regained independent renal function by 1 month. Combined severe acute liver and kidney injury secondary to POD requiring RRT is associated with a high mortality. The majority of survivors recover independent kidney function by 1 month. Standard disease severity scores appear to reflect prognosis in these patients.
INTRODUCTION
Paracetamol overdose (POD) is one of the commonest causes of non-accidental poisoning seen in secondary careCitation1 and accounts for a significant number of deaths in the United Kingdom every year.Citation2 The associated liver injury is well recognized,Citation3 but the potential for nephrotoxicity perhaps less so. Concomitant acute kidney injury (AKI) can occur in as many as 50% of severe PODs,Citation4 even in the absence of fulminant liver failure.Citation5
There is a paucity of published data on the outcomes of patients with combined severe acute liver and kidney injury secondary to paracetamol toxicity. We report on the outcomes of these patients admitted to our intensive care unit (ICU) over a 6-year period and examine which of the parameters measured at the time of admission to ICU were predictive of mortality. Additionally, we were interested in comparing those patients who went on to receive a liver transplant with those who did not. More recently, there has been increasing awareness of the longer term sequelae following an episode of AKI in terms of residual kidney injury and the risk of progressive chronic kidney diseaseCitation6–8 and we report here on the renal outcomes for this group of patients.
Table 1. ICU admission characteristics and principal outcomes for critically ill patients with combined liver and renal toxicity secondary to paracetamol toxicity RRT.
MATERIALS AND METHODS
A single-center retrospective study of all adult ICU admissions over a 6-year period was undertaken. Patients were initially identified from the Intensive Care National Audit and Research (ICNARC) database. Those patients admitted due to paracetamol toxicity and who received renal replacement therapy (RRT) during their ICU admission were selected for further analysis. In addition to basic demographic data, information was obtained on length of hospital and ICU stay, number of days of RRT, discharge destination from ICU, admission Acute Physiology and Chronic Health Evaluation II (APACHE II) score, Model for End-Stage Liver Disease (MELD) score, and ICU and hospital mortality.
The ICU serves a tertiary regional liver unit and supports both liver and kidney transplantation. All RRT on the ICU for this period was performed using predilation continuous venovenous hemofiltration with a substitution rate of 35 mL/kg/h of bicarbonate-buffered replacement fluid. Dialysis access was by means of a dual lumen catheter inserted in either the femoral or internal jugular vein. Patients were anticoagulated with systemic heparin aiming for an activated partial thromboplastin time between 1 and 2 times the control value unless they had an elevated international normalized ratio (INR) > 2.0, had active bleeding, were at high risk of hemorrhage, or had thrombocytopenia in which case either no anticoagulation or prostacyclin (5 ng/kg/min) was used. The initiation of RRT was commenced at the discretion of the attending nephrologists and intensivists. The most frequent indication for initiation was worsening electrolyte or acid–base abnormalities in the setting of progressive oliguria. RRT was prescribed and reviewed by the nephrology team on consultation with the intensivists on a daily basis and delivered by ICU nursing staff.
Statistical analysis was performed using Students t-test, chi-square test, or Mann–Whitney U test as appropriate. Logistic regression modeling was used to look for significant predictors of hospital mortality based on the variables recorded at the time of ICU admission. A p-value of less than 0.05 was considered significant.
RESULTS
There were a total of 5582 admissions to the ICU over the 6-year period of which 73 patients met the inclusion criteria. Of the 73 patients, 10 received a liver transplant during their admission. The admission characteristics and principal outcomes for all patients are shown in . Comparison is also shown on the basis of whether patients subsequently received a liver transplant.
Table 2. Outcomes stratified on basis of patients receiving RRT within the first 24 h in ICU.
Overall mortality was high with 58% of patients dying in hospital. The mortality was lower for patients who received a liver transplant at 20%. In terms of admission characteristics, the transplanted and non-transplanted groups were similar except those patients who went on to receive a liver transplant were significantly younger (p = 0.04) and had a higher alanine transaminase (ALT) level (p = 0.01). Patients were well matched in terms of the MELD and APACHE II global disease severity scores and for other laboratory parameters previously shown to be of prognostic significance such as bilirubin, kidney function, and INR.Citation9
At the time of admission, 29 patients fulfilled the King’s College Hospital (KCH) criteriaCitation10 for liver transplant eligibility and these patients had an odds ratio (OR) of 3.8 (1.3–10.6) for death compared with those patients who did not meet the criteria. Of the 29 patients who fulfilled the KCH criteria, 23/29 (79%) did not receive a transplant and 22/29 (76%) died before discharge. A total of six patients who fulfilled the KCH criteria at the time of ICU admission went on to receive a liver transplant; four out of these six survived to hospital discharge.
Univariate analysis revealed that both a higher APACHE II score [OR 1.1 (1.0–1.2)] and a higher MELD score [OR 1.1 (1.0–1.1)] were associated with an increased risk of death. Multivariate logistic regression analysis of the variables measured at the time of ICU admission found that only higher bilirubin [OR 1.6 (1.1–2.3) mg/dL] and lower serum creatinine [OR 0.52 (0.3–0.9) mg/dL] were statistically associated with inpatient mortality. The patient age, sex, admission INR, platelet count, ALT, sodium, hemoglobin level, and pH on admission had no statistical bearing on the outcome. In terms of pH, all patients with an admission pH < 7 (n = 8) died. However, multivariate logistic regression analysis did not find a significant association between pH and mortality when entered into the analysis as either an absolute value or a dummy variable coding for pH < 7.3 in line with the KCH criteria. The hospital and ICU mortality rates were similar suggesting that the majority of deaths occur while patients are in the ICU and if they survived to ICU discharge their subsequent risk of death appeared to be low.
Despite a higher serum creatinine on admission being associated with a lower mortality, stratifying patients on the basis of whether they received RRT within 24 h of ICU admission revealed no differences in mortality, number of days of RRT required, or length of stay ().
At the time of ICU discharge, 13/32 (41%) of ICU survivors remained RRT dependant. Follow-up data were available for 8/13 of these patients, all of whom had regained independent renal function by 1 month. One patient died after ICU discharge and before being discharged from hospital. The mean creatinine for dialysis-independent patients at the time of ICU discharge was 1.8 mg/dL (SD ± 1.0) which had fallen significantly to 1.2 mg/dL (SD ± 0.7) (p < 0.05) by 1 month for the 18 hospital survivors, 8 of whom had received a liver transplant, with data available.
DISCUSSION
Admission to the ICU with paracetamol-induced liver injury and severe AKI necessitating RRT is associated with a high mortality. Mortality is lower for patients who receive a liver transplant. This is an expected finding as the patients selected for transplant are likely to be the subset of patients most likely to do well. Interestingly in our cohort, with the exception of age, it is hard to differentiate between the two groups on the basis of standard disease severity scores (APACHE II and MELD) and biochemical characteristics at the time of ICU admission. Although originally developed for assessing chronic liver disease, there is evidence that the MELD score has utility in predicting mortality in both non-paracetamol-Citation9 and paracetamol-Citation11 induced acute liver injury. In this cohort a high MELD score was also found to be predictive of mortality. Patients who fulfilled the KCH criteria for transplant listingCitation10 at the time of ICU admission fared worse. This is an expected finding as these criteria were developed to identify patients with an adverse prognosis. Comparing patients on the basis of who went on to receive a liver transplant revealed marked similarities between the groups with the exception of ALT at the time of ICU admission which was higher in the transplant group (p = 0.01). Liver transplant recipients were younger and had lower mortality rates and longer hospital stays than patients who did not receive a transplant.
Patients with worse kidney function at the time of ICU admission, as defined by a higher serum creatinine, had a lower mortality rate. This is at odds with a previously published report of 522 patients with paracetamol poisoning referred to the Scottish Liver Transplant Unit in whom kidney dysfunction at presentation was a marker of poor prognosis.Citation12 However, in our analysis the study cohort is different in that it comprises only patients who went on to receive RRT in the ICU setting. It is not clear how to explain these findings and is unlikely to be due to early initiation of RRT as there was no mortality difference between those who received RRT within 24 h of ICU admission and those who started later. The optimum time to commence RRT in critically ill patients is yet to be clearly defined.Citation13 Our findings may simply represent a type II statistical error owing to the sample size as there was a lower mortality in those patients receiving RRT earlier although it did not reach statistical significance. An alternative explanation may be that worse kidney function prompts earlier critical care involvement with timely resuscitation and specialist input thus improving the outlook for these patients.
Despite 41% of patients being RRT dependant at the time of ICU discharge, all of them for whom data were available had regained independent kidney function by 1 month. For those who were RRT independent prior to leaving the ICU, the creatinine had fallen further by 1 month compared with ICU discharge values. These data indicate that the majority of patients with paracetamol-induced acute liver and kidney injury requiring RRT on the ICU recover independent kidney function.
There are a number of limitations to this retrospective single-center analysis which include the unavailability of data on ventilatory and inotropic requirements, serum lactate measurements, delivered RRT dose (as compared with prescribed dose), baseline kidney function, and biochemical and urine output parameters at the time of RRT commencement. However, we have shown that AKI requiring RRT in the setting of POD-induced liver failure is associated with a high mortality and that the majority of survivors recover independent kidney function. We have also demonstrated that standard disease severity scores such as APACHE II and MELD seem to have prognostic use in this cohort of critically ill patients. Interestingly, variables such as bilirubin, pH, and INR which are known to have prognostic value in all comers with paracetamol-induced liver injury seem to have less bearing on eventual patient outcome for the subset of patients who go on to receive RRT.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References
- Health Protection Agency. National Poisons Information Service Annual Report 2007/8. London, UK: Health Protection Agency; 2008.
- Office for National Statistics. Deaths Related to Drug Poisoning in England and Wales, 2009. Newport, UK: Office for National Statistics; 2010.
- Bailey B, Amre DK, Gaudreault P. Fulminant hepatic failure secondary to acetaminophen poisoning: A systematic review and meta-analysis of prognostic criteria determining the need for liver transplantation. Crit Care Med. 2003;31(1):299–305.
- Larson AM, Polson J, Fontana RJ, . Acetaminophen-induced acute liver failure: Results of a United States multicenter, prospective study. Hepatology. 2005;42(6):1364–1372.
- Eguia L, Materson BJ. Acetaminophen-related acute renal failure without fulminant liver failure. Pharmacotherapy. 1997;17(2):363–370.
- Wald R, Quinn RR, Luo J, . Chronic dialysis and death among survivors of acute kidney injury requiring dialysis. J Am Med Assoc. 2009;302(11):1179–1185.
- Lo LJ, Go AS, Chertow GM, . Dialysis-requiring acute renal failure increases the risk of progressive chronic kidney disease. Kidney Int. 2009;76(8):893–899.
- Amdur RL, Chawla LS, Amodeo S, Kimmel PL, Palant CE. Outcomes following diagnosis of acute renal failure in U.S. veterans: Focus on acute tubular necrosis. Kidney Int. 2009;76(10):1089–1097.
- Dhiman RK, Jain S, Maheshwari U, . Early indicators of prognosis in fulminant hepatic failure: An assessment of the Model for End-Stage Liver Disease (MELD) and King’s College Hospital criteria. Liver Transpl. 2007;13(6):814–821.
- O‘Grady JG, Alexander GJ, Hayllar KM, Williams R. Early indicators of prognosis in fulminant hepatic failure. Gastroenterology. 1989;97(2):439–445.
- Schmidt LE, Larsen FS. MELD score as a predictor of liver failure and death in patients with acetaminophen-induced liver injury. Hepatology. 2007;45(3):789–796.
- Pakravan N, Simpson KJ, Waring WS, Bates CM, Bateman DN. Renal injury at first presentation as a predictor for poor outcome in severe paracetamol poisoning referred to a liver transplant unit. Eur J Clin Pharmacol. 2009;65(2):163–168.
- Bagshaw SM, Cruz DN, Gibney RN, Ronco C. A proposed algorithm for initiation of renal replacement therapy in adult critically ill patients. Crit Care. 2009;13(6):317.