David Prieto-Merino

Publications (27) View all

• Article: The CRASH-2 trial: a randomised controlled trial and economic evaluation of the effects of tranexamic acid on death, vascular occlusive events and transfusion requirement in bleeding trauma patients.

  I Roberts, H Shakur, T Coats, B Hunt, E Balogun, L Barnetson, L Cook, T Kawahara, P Perel, D Prieto-Merino, M Ramos, J Cairns, C Guerriero
  [show abstract] [hide abstract]
  ABSTRACT: Among trauma patients who survive to reach hospital, exsanguination is a common cause of death. A widely practicable treatment that reduces blood loss after trauma could prevent thousands of premature deaths each year. The CRASH-2 trial aimed to determine the effect of the early administration of tranexamic acid on death and transfusion requirement in bleeding trauma patients. In addition, the effort of tranexamic acid on the risk of vascular occlusive events was assessed. Tranexamic acid (TXA) reduces bleeding in patients undergoing elective surgery. We assessed the effects and cost-effectiveness of the early administration of a short course of TXA on death, vascular occlusive events and the receipt of blood transfusion in trauma patients. Randomised placebo-controlled trial and economic evaluation. Randomisation was balanced by centre, with an allocation sequence based on a block size of eight, generated with a computer random number generator. Both participants and study staff (site investigators and trial co-ordinating centre staff) were masked to treatment allocation. All analyses were by intention to treat. A Markov model was used to assess cost-effectiveness. The health outcome was the number of life-years (LYs) gained. Cost data were obtained from hospitals, the World Health Organization database and UK reference costs. Cost-effectiveness was measured in international dollars ($) per LY. Deterministic and probabilistic sensitivity analyses were performed to test the robustness of the results to model assumptions. Two hundred and seventy-four hospitals in 40 countries. Adult trauma patients (n = 20,211) with, or at risk of, significant bleeding who were within 8 hours of injury. Tranexamic acid (loading dose 1 g over 10 minutes then infusion of 1 g over 8 hours) or matching placebo. The primary outcome was death in hospital within 4 weeks of injury, and was described with the following categories: bleeding, vascular occlusion (myocardial infarction, stroke and pulmonary embolism), multiorgan failure, head injury and other. Patients were allocated to TXA (n = 10,096) and to placebo (n = 10,115), of whom 10,060 and 10,067 patients, respectively, were analysed. All-cause mortality at 28 days was significantly reduced by TXA [1463 patients (14.5%) in the TXA group vs 1613 patients (16.0%) in the placebo group; relative risk (RR) 0.91; 95% confidence interval (CI) 0.85 to 0.97; p = 0.0035]. The risk of death due to bleeding was significantly reduced [489 patients (4.9%) died in the TXA group vs 574 patients (5.7%) in the placebo group; RR 0.85; 95% CI 0.76 to 0.96; p = 0.0077]. We recorded strong evidence that the effect of TXA on death due to bleeding varied according to the time from injury to treatment (test for interaction p < 0.0001). Early treatment (≤ 1 hour from injury) significantly reduced the risk of death due to bleeding [198 out of 3747 patients (5.3%) died in the TXA group vs 286 out of 3704 patients (7.7%) in the placebo group; RR 0.68; 95% CI 0.57 to 0.82; p < 0.0001]. Treatment given between 1 and 3 hours also reduced the risk of death due to bleeding [147 out of 3037 patients (4.8%) died in the TXA group vs 184 out of 2996 patients (6.1%) in the placebo group; RR 0.79; 95% CI 0.64 to 0.97; p = 0.03]. Treatment given after 3 hours seemed to increase the risk of death due to bleeding [144 out of 3272 patients (4.4%) died in the TXA group vs 103 out of 3362 patients (3.1%) in the placebo group; RR 1.44; 95% CI1.12 to 1.84; p = 0.004]. We recorded no evidence that the effect of TXA on death due to bleeding varied by systolic blood pressure, Glasgow Coma Scale score or type of injury. Administering TXA to bleeding trauma patients within 3 hours of injury saved an estimated 755 LYs per 1000 trauma patients in the UK. The cost of giving TXA to 1000 patients was estimated at $30,830. The incremental cost of giving TXA compared with not giving TXA was $48,002. The incremental cost per LY gained of administering TXA was $64. Early administration of TXA safely reduced the risk of death in bleeding trauma patients and is highly cost-effective. Treatment beyond 3 hours of injury is unlikely to be effective. Future work [the Clinical Randomisation of an Antifibrinolytic in Significant Head injury-3 (CRASH-3) trial] will evaluate the effectiveness and safety of TXA in the treatments of isolated traumatic brain injury (http://crash3.lshtm.ac.uk/). Current Controlled Trials ISRCTN86750102, ClinicalTrials.gov NCT00375258 and South African Clinical Trial Register DOH-27-0607-1919. The project was funded by the Bupa Foundation, the J P Moulton Charitable Foundation and the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 17, No. 10. See HTA programme website for further project information.
  Health technology assessment (Winchester, England). 03/2013; 17(10):1-79.
• Article: ASCORE: an up-to-date cardiovascular risk score for hypertensive patients reflecting contemporary clinical practice developed using the (ASCOT-BPLA) trial data.

  D Prieto-Merino, J Dobson, A K Gupta, C-L Chang, P S Sever, B Dahlöf, H Wedel, S Pocock, N Poulter
  [show abstract] [hide abstract]
  ABSTRACT: A number of risk scores already exist to predict cardiovascular (CV) events. However, scores developed with data collected some time ago might not accurately predict the CV risk of contemporary hypertensive patients that benefit from more modern treatments and management. Using data from the randomised clinical trial Anglo-Scandinavian Cardiac Outcomes Trial-BPLA, with 15 955 hypertensive patients without previous CV disease receiving contemporary preventive CV management, we developed a new risk score predicting the 5-year risk of a first CV event (CV death, myocardial infarction or stroke). Cox proportional hazard models were used to develop a risk equation from baseline predictors. The final risk model (ASCORE) included age, sex, smoking, diabetes, previous blood pressure (BP) treatment, systolic BP, total cholesterol, high-density lipoprotein-cholesterol, fasting glucose and creatinine baseline variables. A simplified model (ASCORE-S) excluding laboratory variables was also derived. Both models showed very good internal validity. User-friendly integer score tables are reported for both models. Applying the latest Framingham risk score to our data significantly overpredicted the observed 5-year risk of the composite CV outcome. We conclude that risk scores derived using older databases (such as Framingham) may overestimate the CV risk of patients receiving current BP treatments; therefore, 'updated' risk scores are needed for current patients.Journal of Human Hypertension advance online publication, 14 February 2013; doi:10.1038/jhh.2013.3.
  Journal of human hypertension 02/2013; · 2.80 Impact Factor
• Dataset: Crash2 TARN Pablo

  Pablo Perel, David Prieto-Merino, Haleema Shakur, Tim Clayton, Fiona Lecky, Omar Bouamra, Rob Russell, Mark Faulkner, Ewout W Steyerberg, Ian Roberts
• Article: Using Electronic Health Care Records for Drug Safety Signal Detection: A Comparative Evaluation of Statistical Methods.

  Martijn J Schuemie, Preciosa M Coloma, Huub Straatman, Ron M C Herings, Gianluca Trifirò, Justin Neil Matthews, David Prieto-Merino, Mariam Molokhia, Lars Pedersen, Rosa Gini, Francesco Innocenti, Giampiero Mazzaglia, Gino Picelli, Lorenza Scotti, Johan van der Lei, Miriam C J M Sturkenboom
  [show abstract] [hide abstract]
  ABSTRACT: BACKGROUND:: Drug safety monitoring relies primarily on spontaneous reporting, but electronic health care record databases offer a possible alternative for the detection of adverse drug reactions (ADRs). OBJECTIVES:: To evaluate the relative performance of different statistical methods for detecting drug-adverse event associations in electronic health care record data representing potential ADRs. RESEARCH DESIGN:: Data from 7 databases across 3 countries in Europe comprising over 20 million subjects were used to compute the relative risk estimates for drug-event pairs using 10 different methods, including those developed for spontaneous reporting systems, cohort methods such as the longitudinal gamma poisson shrinker, and case-based methods such as case-control. The newly developed method "longitudinal evaluation of observational profiles of adverse events related to drugs" (LEOPARD) was used to remove associations likely caused by protopathic bias. Data from the different databases were combined by pooling of data, and by meta-analysis for random effects. A reference standard of known ADRs and negative controls was created to evaluate the performance of the method. MEASURES:: The area under the curve of the receiver operator characteristic curve was calculated for each method, both with and without LEOPARD filtering. RESULTS:: The highest area under the curve (0.83) was achieved by the combination of either longitudinal gamma poisson shrinker or case-control with LEOPARD filtering, but the performance between methods differed little. LEOPARD increased the overall performance, but flagged several known ADRs as caused by protopathic bias. CONCLUSIONS:: Combinations of methods demonstrate good performance in distinguishing known ADRs from negative controls, and we assume that these could also be used to detect new drug safety signals.
  Medical care 08/2012; 50(10):890-897. · 3.24 Impact Factor
• Article: The science of risk models.

  David Prieto-Merino, Stuart J Pocock
  [show abstract] [hide abstract]
  ABSTRACT: An individual's overall cardiovascular risk should guide appropriate therapy and patient management. Several risk assessment scores are available; however, further development of risk algorithms is necessary to account for changes in available treatments and patient lifestyles, to make use of emerging risk factors and more accurate methods for measuring outcomes, and to provide more targeted measurement of risk for different patient subpopulations. When developing a risk model it is important to clearly define the outcome that the risk will predict, the period of follow up, the patient population, and the predictors to be used and how they will be combined. An appropriate statistical model is specified with the aim of finding the weighted combination of the candidate risk factors that best predicts the disease outcome. Stepwise regression is used to systematically search through candidate risk factors to produce a final model with an acceptable number of highly relevant variables. Possible non-linear effects of continuous variables and interactions between variables must be considered. However, the selection of variables requires not just statistical criteria but also clinical, biological and epidemiological judgement. In general, relatively simple, clinically reasonable and easy-to-use models that can be generalized to other settings are preferred to complex mathematical models that fit the sample data perfectly. There is a permanent need for updating cardiovascular risk scores to reflect advances in our clinical knowledge over time and changes in population risk. Development of a risk model requires both statistical expertise and a sound knowledge of the clinical and epidemiological aspects of cardiovascular disease.
  European journal of preventive cardiology. 08/2012; 19(2 Suppl):7-13.