For personal use. Only reproduce with permission from The Lancet Publishing Group. ARTICLES 558 THE LANCET • Vol 359 • February 16, 2002 • www.thelancet.com Summary Background Although the King’s College Hospital (KCH) selection criteria for emergency liver transplantation in paracetamol-induced acute liver failure are widely used, strategies to improve sensitivity and facilitate earlier transplantation are required. We investigated the use of arterial blood lactate measurement for the identification of transplantation candidates. Methods In a single-centre study, we measured arterial blood lactate early (median 4 h) and after fluid resuscitation (median 12 h) in patients admitted to a tertiary-referral intensive-care unit. Threshold values that best identified individuals likely to die without transplantation were derived in a retrospective initial sample of 103 patients with paracetamol-induced acute liver failure and applied to a prospective validation sample of 107 patients. Predictive value and speed of identification were compared with those of KCH criteria. Findings In the initial sample, median lactate was significantly higher in non-surviving patients than in survivors both in the early samples (8·5 [range 1·7–21·0] vs 1·4 [0·53–7·9] mmol/L, p<0·0001) and after fluid resuscitation (5·5 [1·3–18·6] vs 1·3 [0·26–3·2], p<0·0001). Applied to the validation sample, a threshold value of 3·5 mmol/L early after admission had sensitivity 67%, specificity 95%, positive likelihood ratio 13, and negative likelihood ratio 0·35; the corresponding values for a threshold of 3·0 mmol/L after fluid resuscitation were 76%, 97%, 30, and 0·24. Combined early and postresuscitation lactate concentrations had similar predictive ability to KCH criteria but identified non-surviving patients earlier (4 [3–13] vs 10 [3·5–19·5] h, p=0·01). Addition of postresuscitation lactate concentration to KCH criteria increased sensitivity from 76% to 91% and lowered negative likelihood ratio from 0·25 to 0·10. Interpretation Arterial blood lactate measurement rapidly and accurately identifies patients who will die from paracetamol- induced acute liver failure. Its use could improve the speed and accuracy of selection of appropriate candidates for transplantation. Lancet 2002; 359: 558–63 Introduction Acute liver failure resulting from severe hepatotoxic effects of paracetamol is a short intense illness, with rapid progression from initial hepatic failure to involvement of multiple organ systems, with acute renal failure, haemodynamic instability, and encephalopathy. 1,2 In patients with advanced acute liver failure, the only therapy of proven benefit is emergency transplantation. 2,3 Current clinical management of paracetamol-induced acute liver failure therefore relies on early identification and transplantation for those patients who will die without it, and the provision of supportive medical care to those in whom a spontaneous recovery of liver function is possible. 2,4 Although there are no standard clinical selection criteria for transplantation in use worldwide, the King’s College Hospital (KCH) criteria 5 are the most widely applied. 3,6 Validation studies have consistently shown that without transplantation, survival in patients who meet the KCH criteria is less than 15%. 6–9 However, most of the patients who meet these criteria do not undergo transplantation. 9 At the time of fulfilling the criteria, many patients are already too unwell for transplantation to be contemplated, and a substantial proportion of those listed for transplantation deteriorate before grafts become available. 9,10 Furthermore, although the KCH criteria seem to have a clinically acceptable specificity in identifying patients with a poor prognosis, their sensitivity may be limited, in that they fail to identify a significant proportion of those who will die. 6,9,11 Earlier, more accurate identification of patients who require transplantation is therefore important if survival is to be improved. In critical illness other than acute liver failure, prolonged high blood lactate concentrations are associated with a poor prognosis. 12–15 Hyperlactataemia in this setting has been accepted as a marker of systemic tissue dysoxia or microcirculatory perfusion abnormalities and increased anaerobic metabolism. 15,16 Since circulating lactate is metabolised mainly by the liver, high blood lactate concentrations may also reflect decreases in clearance secondary to impaired hepatic function. 16–18 The injured liver may itself also act as a source of lactate. 19 In acute liver failure, hyperlactataemia may therefore indicate the severity of both the hepatic injury sustained and the accompanying multiple organ failure. Blood lactate can be rapidly and accurately measured by point- of-care testing and thus could provide a practical early indicator of outcome. We assessed the clinical use of blood lactate measurements to identify patients likely to die without transplantation. We examined the clinical and bio- chemical correlates of blood lactate in a large cohort of patients with paracetamol-induced acute liver failure, and investigated threshold values that best identified non- survivors. These threshold values were applied to a second prospective sample of patients and compared with the KCH criteria for prognostic accuracy and speed of identification of patients who will not survive without transplantation. Blood lactate as an early predictor of outcome in paracetamol- induced acute liver failure: a cohort study William Bernal, Nora Donaldson, Duncan Wyncoll, Julia Wendon Institute of Liver Studies, King’s College Hospital (W Bernal MRCP, J Wendon FRCP); Clinical Research Statistics, Department of Research and Development, King’s College Hospital (N Donaldson PhD); and Department of Intensive Care, St Thomas’ Hospital (D Wyncoll FRCA); Guy’s, King’s, and St Thomas’s School of Medicine, London, UK Correspondence to: Dr Julia Wendon, Institute of Liver Studies, King’s College Hospital, Denmark Hill, London SE5 9RS, UK (e-mail: julia.wendon@kcl.ac.uk)