OUCRU has a research programme dedicated to developing quantitative methods in biology and medicine. The research programme is headed by Dr Maciej Boni and Dr Marcel Wolbers and includes projects and foci in mathematical modelling, biostatistics, and bioinformatics. The team includes one post-doc, seven PhD students, and several research assistants and technicians, and the group’s activities include support for OUCRU’s clinical, laboratory, and epidemiological studies as well as development of novel methods that can be applied more generally.
Quantitative methods are crucial at each stage of a scientific investigation from study conception and design to data collection, modelling and data analysis and, finally, communication and dissemination of results. At OUCRU, we strive to achieve a synthesized understanding of our study outcomes by having a single team of investigators with complementary expertise in basic and clinical science, quantitative methods, and project management involved in all stages of the scientific studies we run.
Biostatistical analysis permeates all OUCRU studies as this fundamental field is critical for the foundations of all scientific studies. Mathematical modelling techniques have been used at OUCRU for ‘predictive’ modelling of the effectiveness of future malaria policy, and retrospective modelling to help understand past influenza epidemics in Viet Nam. Bioinformatic analysis has been used to help design sampling studies and perform phylogenetic and recombination analyses in norovirus, influenza, dengue virus, and other pathogens. Several of these projects are described below.
Understanding Influenza Dynamics in the Tropics
Influenza epidemics are seasonal wintertime phenomena in the world’s temperate zones. In tropical countries, however, influenza epidemics are irregular and do not usually follow a consistent season pattern. In addition, tropical regions may support persistent circulation of human influenza viruses, but the extent and duration of these persistence patterns is not yet known. The mathematical modelling group at OUCRU has been focused on designing epidemiological studies that will yield the most information on past influenza dynamics in Viet Nam. These studies have foundations in seroepidemiology and community surveillance, and are designed so that mathematical modelling analyses can be maximally informative in uncovering something about the ‘influenza season’ in Viet Nam or typical attack rates for waves of different influenza strains or subtypes. The modelling analyses use both stochastic and deterministic dynamical models with likelihood methods to evaluate how close of a fit the models are to the data. The first results of these studies will be available in 2014.
Prognostic Models for Infectious Diseases
Prognostic models combine patient data from clinical and laboratory examinations to predict clinical outcomes. For example, in children diagnosed with dengue fever, it is important to identify children with a high risk of progressing to severe diseases at an early disease stage and to adapt patient triage and clinical management accordingly. OUCRU is currently developing and validating such a risk prediction model based on a cohort of more than 2500 hospitalized children with dengue.
The development of reliable prognostic models requires both biostatistical and clinical expertise and the biostatistics group at OUCRU is involved in multiple diagnostic and prognostic studies across several disease areas including dengue fever and central nervous system infections. In addition, we are also developing novel statistical methodology for prognostic models with competing risks and for dynamic prediction models.
Modeling Malaria Treatment Strategies
Malaria exerts a tremendous health burden on billions of people living the world’s malaria-endemic regions. Artemisinin-combination therapies (ACTs) are currently the only highly effective antimalarial drugs, and effective stewardship of these drugs is necessary to prevent the rapid emergence and spread of artemisinin resistance. One approach to extending the useful therapeutic life of these valuable drugs involves changing the distribution pattern of the drugs, on a large-scale population level, so that no single therapy is overused thereby minimizing the ‘pressure’ to evolve drug resistance. In order to explore these options, the OUCRU modelling team has been developing individual-based modelling tools to explore the future of drug resistance in malaria, and to compare different national level policies of recommending which first-line antimalarials should be used and in what manner.
Design of Randomized Controlled Trials
Randomized controlled trials (RCT) are the gold standard for evaluating the effectiveness of clinical interventions and a fundamental component of OUCRU’s research programme with major ongoing trials in central nervous system infections, HIV opportunistic infections, typhoid fever, dengue, and hand, foot, and mouth disease. Biostatistics plays a key role in all stages of an RCT from design to analysis and dissemination of results. The biostatistics group at OUCRU is particularly interested in methods for adaptive clinical trials which offer an opportunity for a more rational and efficient evaluation of interventions but have not yet been widely established in infectious or tropical diseases.
National Institutes for Hygiene and Epidemiology, Hanoi, Vietnam.
MRC Centre for Outbreak Analysis and Modelling, Imperial College, London.
Pritzker School of Medicine, University of Chicago.
Mahidol-Oxford Research Unit (MORU), Bangkok, Thailand.
School of Biological Sciences, University of Sydney, Sydney, Australia.
Le QM, Lam HM, Cuong VD, Lam TT-Y, Haplin RA, Wentworth DE, Hien NT, Thanh LT, Phuong HVM, Horby P, Boni MF. Migration and persistence of human influenza viruses in Vietnam. Emerg Infect Dis. 2013 Nov;19(11):1756-65. doi: 10.3201/eid1911.130349. PMID: 24188643
Lam PK, Tam DHT, Dung NM, Hanh Tien NT, Kieu NTT, Simmons CP, Farrar J, Wills B, Wolbers M. A prognostic model for development of profound shock among children with dengue shock syndrome. Clin Infect Dis. 2013 Dec;57(11):1577-86. doi: 10.1093/cid/cit594. Epub 2013 Sep 17. PMID: 24046311
Wolbers M, Kleinschmidt I, Simmons CP, Donnelly CA. Considerations in the design of clinical trials to test novel entomological approaches to dengue control. PLoS Negl Trop Dis. 2012; 6: e1937. PMID: 23209869
Boni MF, Nguyen TD, de Jong MD, van Doorn HR. Virulence attenuation during an influenza A/H5N1 pandemic. Philos Trans R Soc Lond B Biol Sci. 2013 Feb 4;368(1614):20120207. doi: 10.1098/rstb.2012.0207. Print 2013 Mar 19. PMID: 23382429