Prof Ben Cooper
|Research Area:||Bioinformatics & Stats (inc. Modelling and Computational Biology)|
|Technology Exchange:||Bioinformatics, Computational biology and Medical statistics|
|Scientific Themes:||Tropical Medicine & Global Health and Immunology & Infectious Disease|
Ben Cooper holds an MRC senior non-clinical research fellowship and is based at the Mahidol-Oxford Tropical Medicine Research Unit in Bangkok.
His work uses mathematical modelling and statistical techniques to help understand infectious disease dynamics and evaluate potential control measures. This involves developing mathematical models to help evaluate the likely impact and cost-effectiveness of control measures, developing and applying new statistical approaches based on mechanistic models for the analysis of longitudinal infectious disease data (increasingly making use of whole genome sequence data), and designing and analysing epidemiological studies. The major focus of this work is on antibiotic-resistant bacteria in resource-limited hospital settings.
Additional projects include within-host dynamics of Plasmodium vivax, ecological interactions of the nasopharyngeal flora, cost-effectiveness of seasonal influenza vaccination in Thailand, and dynamics and control of Hepatitis E infections in refugee camps.
|Dr Paul Turner||Tropical Medicine||University of Oxford||United Kingdom|
|Prof Lisa J White||Tropical Medicine||University of Oxford||United Kingdom|
|Dr Direk Limmathurotsakul||Tropical Medicine||University of Oxford||United Kingdom|
|Dr Sharon J Peacock||Tropical Medicine||University of Oxford||United Kingdom|
|Dr Jonathan D Edgeworth||Guy’s and St Thomas’ NHS Foundation Trust||United Kingdom|
|Prof Marc Bonten||University Medical Center Utrecht||Netherlands|
|Prof Angela McLean||Department of Zoology||University of Oxford||United Kingdom|
|Dr Craig MacLean||Department of Zoology||University of Oxford||United Kingdom|
|Prof Nicholoas Graves||QUT||Australia|
|Dr Aronrag C Meeyai||Mahidol University||Thailand|
Influenza epidemiology differs substantially in tropical and temperate zones, but estimates of seasonal influenza mortality in developing countries in the tropics are lacking. We aimed to quantify mortality due to seasonal influenza in Thailand, a tropical middle-income country. Time series of polymerase chain reaction-confirmed influenza infections between 2005 and 2009 were constructed from a sentinel surveillance network. These were combined with influenza-like illness data to derive measures of influenza activity and relationships to mortality by using a Bayesian regression framework. We estimated 6.1 (95% credible interval: 0.5, 12.4) annual deaths per 100,000 population attributable to influenza A and B, predominantly in those aged ≥60 years, with the largest contribution from influenza A(H1N1) in 3 out of 4 years. For A(H3N2), the relationship between influenza activity and mortality varied over time. Influenza was associated with increases in deaths classified as resulting from respiratory disease (posterior probability of positive association, 99.8%), cancer (98.6%), renal disease (98.0%), and liver disease (99.2%). No association with circulatory disease mortality was found. Seasonal influenza infections are associated with substantial mortality in Thailand, but evidence for the strong relationship between influenza activity and circulatory disease mortality reported in temperate countries is lacking. Hide abstract
BACKGROUND: Experimental treatments for Ebola virus disease (EVD) might reduce EVD mortality. There is uncertainty about the ability of different clinical trial designs to identify effective treatments, and about the feasibility of implementing individually randomised controlled trials during an Ebola epidemic. METHODS AND FINDINGS: A treatment evaluation programme for use in EVD was devised using a multi-stage approach (MSA) with two or three stages, including both non-randomised and randomised elements. The probabilities of rightly or wrongly recommending the experimental treatment, the required sample size, and the consequences for epidemic outcomes over 100 d under two epidemic scenarios were compared for the MSA, a sequential randomised controlled trial (SRCT) with up to 20 interim analyses, and, as a reference case, a conventional randomised controlled trial (RCT) without interim analyses. Assuming 50% 14-d survival in the population treated with the current standard of supportive care, all designs had similar probabilities of identifying effective treatments correctly, while the MSA was less likely to recommend treatments that were ineffective. The MSA led to a smaller number of cases receiving ineffective treatments and faster roll-out of highly effective treatments. For less effective treatments, the MSA had a high probability of including an RCT component, leading to a somewhat longer time to roll-out or rejection. Assuming 100 new EVD cases per day, the MSA led to between 6% and 15% greater reductions in epidemic mortality over the first 100 d for highly effective treatments compared to the SRCT. Both the MSA and SRCT led to substantially fewer deaths than a conventional RCT if the tested interventions were either highly effective or harmful. In the proposed MSA, the major threat to the validity of the results of the non-randomised components is that referral patterns, standard of care, or the virus itself may change during the study period in ways that affect mortality. Adverse events are also harder to quantify without a concurrent control group. CONCLUSIONS: The MSA discards ineffective treatments quickly, while reliably providing evidence concerning effective treatments. The MSA is appropriate for the clinical evaluation of EVD treatments. Hide abstract
BACKGROUND: Predictive models to identify unknown methicillin-resistant Staphylococcus aureus (MRSA) carriage on admission may optimise targeted MRSA screening and efficient use of resources. However, common approaches to model selection can result in overconfident estimates and poor predictive performance. We aimed to compare the performance of various models to predict previously unknown MRSA carriage on admission to surgical wards. METHODS: The study analysed data collected during a prospective cohort study which enrolled consecutive adult patients admitted to 13 surgical wards in 4 European hospitals. The participating hospitals were located in Athens (Greece), Barcelona (Spain), Cremona (Italy) and Paris (France). Universal admission MRSA screening was performed in the surgical wards. Data regarding demographic characteristics and potential risk factors for MRSA carriage were prospectively collected during the study period. Four logistic regression models were used to predict probabilities of unknown MRSA carriage using risk factor data: "Stepwise" (variables selected by backward elimination); "Best BMA" (model with highest posterior probability using Bayesian model averaging which accounts for uncertainty in model choice); "BMA" (average of all models selected with BMA); and "Simple" (model including variables selected >50% of the time by both Stepwise and BMA approaches applied to repeated random sub-samples of 50% of the data). To assess model performance, cross-validation against data not used for model fitting was conducted and net reclassification improvement (NRI) was calculated. RESULTS: Of 2,901 patients enrolled, 111 (3.8%) were newly identified MRSA carriers. Recent hospitalisation and presence of a wound/ulcer were significantly associated with MRSA carriage in all models. While all models demonstrated limited predictive ability (mean c-statistics <0.7) the Simple model consistently detected more MRSA-positive individuals despite screening fewer patients than the Stepwise model. Moreover, the Simple model improved reclassification of patients into appropriate risk strata compared with the Stepwise model (NRI 6.6%, P = .07). CONCLUSIONS: Though commonly used, models developed using stepwise variable selection can have relatively poor predictive value. When developing MRSA risk indices, simpler models, which account for uncertainty in model selection, may better stratify patients' risk of unknown MRSA carriage. Hide abstract
Methicillin-resistant Staphylococcus aureus (MRSA) is a major cause of nosocomial infection. Whole-genome sequencing of MRSA has been used to define phylogeny and transmission in well-resourced healthcare settings, yet the greatest burden of nosocomial infection occurs in resource-restricted settings where barriers to transmission are lower. Here, we study the flux and genetic diversity of MRSA on ward and individual patient levels in a hospital where transmission was common. We repeatedly screened all patients on two intensive care units for MRSA carriage over a 3-mo period. All MRSA belonged to multilocus sequence type 239 (ST 239). We defined the population structure and charted the spread of MRSA by sequencing 79 isolates from 46 patients and five members of staff, including the first MRSA-positive screen isolates and up to two repeat isolates where available. Phylogenetic analysis identified a flux of distinct ST 239 clades over time in each intensive care unit. In total, five main clades were identified, which varied in the carriage of plasmids encoding antiseptic and antimicrobial resistance determinants. Sequence data confirmed intra- and interwards transmission events and identified individual patients who were colonized by more than one clade. One patient on each unit was the source of numerous transmission events, and deep sampling of one of these cases demonstrated colonization with a "cloud" of related MRSA variants. The application of whole-genome sequencing and analysis provides novel insights into the transmission of MRSA in under-resourced healthcare settings and has relevance to wider global health. Hide abstract
BMJ, 349 (aug21 5), pp. g5060. | Read more2014. Interpreting and comparing risks in the presence of competing events.
Plasmids are important drivers of bacterial evolution, but it is challenging to understand how plasmids persist over the long term because plasmid carriage is costly. Classical models predict that horizontal transfer is necessary for plasmid persistence, but recent work shows that almost half of plasmids are non-transmissible. Here we use a combination of mathematical modelling and experimental evolution to investigate how a costly, non-transmissible plasmid, pNUK73, can be maintained in populations of Pseudomonas aeruginosa. Compensatory adaptation increases plasmid stability by eliminating the cost of plasmid carriage. However, positive selection for plasmid-encoded antibiotic resistance is required to maintain the plasmid by offsetting reductions in plasmid frequency due to segregational loss. Crucially, we show that compensatory adaptation and positive selection reinforce each other's effects. Our study provides a new understanding of how plasmids persist in bacterial populations, and it helps to explain why resistance can be maintained after antibiotic use is stopped. Hide abstract
BACKGROUND: Little is known about the epidemiology of nosocomial bloodstream infections in public hospitals in developing countries. We evaluated trends in incidence of hospital-acquired bacteremia (HAB) and healthcare-associated bacteremia (HCAB) and associated mortality in a developing country using routinely available databases. METHODS: Information from the microbiology and hospital databases of 10 provincial hospitals in northeast Thailand was linked with the national death registry for 2004-2010. Bacteremia was considered hospital-acquired if detected after the first two days of hospital admission, and healthcare-associated if detected within two days of hospital admission with a prior inpatient episode in the preceding 30 days. RESULTS: A total of 3,424 patients out of 1,069,443 at risk developed HAB and 2,184 out of 119,286 at risk had HCAB. Of these 1,559 (45.5%) and 913 (41.8%) died within 30 days, respectively. Between 2004 and 2010, the incidence rate of HAB increased from 0.6 to 0.8 per 1,000 patient-days at risk (p<0.001), and the cumulative incidence of HCAB increased from 1.2 to 2.0 per 100 readmissions (p<0.001). The most common causes of HAB were Acinetobacter spp. (16.2%), Klebsiella pneumoniae (13.9%), and Staphylococcus aureus (13.9%), while those of HCAB were Escherichia coli (26.3%), S. aureus (14.0%), and K. pneumoniae (9.7%). There was an overall increase over time in the proportions of ESBL-producing E. coli causing HAB and HCAB. CONCLUSIONS: This study demonstrates a high and increasing incidence of HAB and HCAB in provincial hospitals in northeast Thailand, increasing proportions of ESBL-producing isolates, and very high associated mortality. Hide abstract
In nested case-control studies, incidence density sampling is the time-dependent matching procedure to approximate hazard ratios. The cumulative incidence function can also be estimated if information from the full cohort is used. In the presence of competing events, however, the cumulative incidence function depends on the hazard of the disease of interest and on the competing events hazard. Using hospital-acquired infection as an example (full cohort), we propose a sampling method for nested case-control studies to estimate subdistribution hazard ratios. With further information on the full cohort, the cumulative incidence function for the event of interest can then be estimated as well. Hide abstract
Background: Intensive care units (ICUs) are high-risk areas for transmission of antimicrobial-resistant bacteria, but no controlled study has tested the effect of rapid screening and isolation of carriers on transmission in settings with best-standard precautions. We assessed interventions to reduce colonisation and transmission of antimicrobial-resistant bacteria in European ICUs. Methods: We did this study in three phases at 13 ICUs. After a 6 month baseline period (phase 1), we did an interrupted time series study of universal chlorhexidine body-washing combined with hand hygiene improvement for 6 months (phase 2), followed by a 12-15 month cluster randomised trial (phase 3). ICUs were randomly assigned by computer generated randomisation schedule to either conventional screening (chromogenic screening for meticillin-resistant Staphylococcus aureus [MRSA] and vancomycin-resistant enterococci [VRE]) or rapid screening (PCR testing for MRSA and VRE and chromogenic screening for highly resistant Enterobacteriaceae [HRE]); with contact precautions for identified carriers. The primary outcome was acquisition of resistant bacteria per 100 patient-days at risk, for which we calculated step changes and changes in trends after the introduction of each intervention. We assessed acquisition by microbiological surveillance and analysed it with a multilevel Poisson segmented regression model. We compared screening groups with a likelihood ratio test that combined step changes and changes to trend. This study is registered with ClinicalTrials.gov, number NCT00976638. Findings: Seven ICUs were assigned to rapid screening and six to conventional screening. Mean hand hygiene compliance improved from 52% in phase 1 to 69% in phase 2, and 77% in phase 3. Median proportions of patients receiving chlorhexidine body-washing increased from 0% to 100% at the start of phase 2. For trends in acquisition of antimicrobial-resistant bacteria, weekly incidence rate ratio (IRR) was 0·976 (0·954-0·999) for phase 2 and 1·015 (0·998-1·032) for phase 3. For step changes, weekly IRR was 0·955 (0·676-1·348) for phase 2 and 0·634 (0·349-1·153) for phase 3. The decrease in trend in phase 2 was largely caused by changes in acquisition of MRSA (weekly IRR 0·925, 95% CI 0·890-0·962). Acquisition was lower in the conventional screening group than in the rapid screening group, but did not differ significantly (p=0·06). Interpretation: Improved hand hygiene plus unit-wide chlorhexidine body-washing reduced acquisition of antimicrobial-resistant bacteria, particularly MRSA. In the context of a sustained high level of compliance to hand hygiene and chlorhexidine bathings, screening and isolation of carriers do not reduce acquisition rates of multidrug-resistant bacteria, whether or not screening is done with rapid testing or conventional testing. Funding: European Commission. © 2014 Derde et al. Open Access article distributed under the terms of CC BY-NC-SA. Hide abstract
INTRODUCTION: Economic evaluations of interventions in the hospital setting often rely on the estimated long-term impact on patient survival. Estimates of mortality rates and long-term outcomes among patients discharged alive from the intensive care unit (ICU) are lacking from lower- and middle-income countries. This study aimed to assess the long-term survival and life expectancy (LE) amongst post-ICU patients in Thailand, a middle-income country. METHODS: In this retrospective cohort study, data from a regional tertiary hospital in northeast Thailand and the regional death registry were linked and used to assess patient survival time after ICU discharge. Adult ICU patients aged at least 15 years who had been discharged alive from an ICU between 1 January 2004 and 31 December 2005 were included in the study, and the death registry was used to determine deaths occurring in this cohort up to 31st December 2010. These data were used in conjunction with standard mortality life tables to estimate annual mortality and life expectancy. RESULTS: This analysis included 10,321 ICU patients. During ICU admission, 3,251 patients (31.5%) died. Of 7,070 patients discharged alive, 2,527 (35.7%) were known to have died within the five-year follow-up period, a mortality rate 2.5 times higher than that in the Thai general population (age and sex matched). The mean LE was estimated as 18.3 years compared with 25.2 years in the general population. CONCLUSIONS: Post-ICU patients experienced much higher rates of mortality than members of the general population over the five-year follow-up period, particularly in the first year after discharge. Further work assessing Health Related Quality of Life (HRQOL) in both post-ICU patients and in the general population in developing countries is needed. Hide abstract
BACKGROUND: There are frequent reports of intensive care unit (ICU) outbreaks due to transmission of particular antibiotic-resistant bacteria. Less is known about the burden of outbreaks of resistance due to horizontal transfer of mobile genetic elements between species. Moreover, the potential of existing statistical software as a preliminary means for detecting such events has never been assessed. This study uses a software package to determine the burden of species and resistance outbreaks in 2 adjacent ICUs and to look for evidence of clustering of resistance outbreaks consistent with interspecies transmission of resistance elements. METHODS: A retrospective analysis of data from 2 adjacent 15-bed adult ICUs between 2002 and 2009 was undertaken. Detection of bacterial species-groups and resistance outbreaks was conducted using SaTScan and WHONet-SaTScan software. Resampling and permutation methods were applied to investigate temporal clustering of outbreaks. RESULTS: Outbreaks occurred for 69% of bacterial species-groups (18/26), and resistance outbreaks were detected against 63% of antibiotics (10/16). Resistance outbreaks against 7 of 10 antibiotics were observed in multiple species-groups simultaneously and there was evidence of inter-species-group dependence for 4 of 7 antibiotics; background temporal changes in resistance did not explain the temporal aggregation of outbreaks in 3 of 7 antibiotics. CONCLUSIONS: Species outbreaks occurred for the majority of bacteria commonly identified in the ICU. There was evidence for frequent temporal clustering of resistance outbreaks consistent with interspecies transmission of resistance elements. Wider application of outbreak detection software combined with targeted sequencing of bacterial genomes is needed to understand the contribution of interspecies gene transfer to resistance emergence. Hide abstract
Infection control for hospital pathogens such as methicillin-resistant Staphylococcus aureus (MRSA) often takes the form of a package of interventions, including the use of patient isolation and decolonization treatment. Such interventions, though widely used, have generated controversy because of their significant resource implications and the lack of robust evidence with regard to their effectiveness at reducing transmission. The aim of this study was to estimate the effectiveness of isolation and decolonization measures in reducing MRSA transmission in hospital general wards. Prospectively collected MRSA surveillance data from 10 general wards at Guy's and St. Thomas' hospitals, London, United Kingdom, in 2006-2007 were used, comprising 14,035 patient episodes. Data were analyzed with a Markov chain Monte Carlo algorithm to model transmission dynamics. The combined effect of isolation and decolonization was estimated to reduce transmission by 64% (95% confidence interval: 37, 79). Undetected MRSA-positive patients were estimated to be the source of 75% (95% confidence interval: 67, 86) of total transmission events. Isolation measures combined with decolonization treatment were strongly associated with a reduction in MRSA transmission in hospital general wards. These findings provide support for active methods of MRSA control, but further research is needed to determine the relative importance of isolation and decolonization in preventing transmission. Hide abstract
An important determinant of a pathogen's success is the rate at which it is transmitted from infected to susceptible hosts. Although there are anecdotal reports that methicillin-resistant Staphylococcus aureus (MRSA) clones vary in their transmissibility in hospital settings, attempts to quantify such variation are lacking for common subtypes, as are methods for addressing this question using routinely-collected MRSA screening data in endemic settings. Here we present a method to quantify the time-varying transmissibility of different subtypes of common bacterial nosocomial pathogens using routine surveillance data. The method adapts approaches for estimating reproduction numbers based on the probabilistic reconstruction of epidemic trees, but uses relative hazards rather than serial intervals to assign probabilities to different sources for observed transmission events. The method is applied to data collected as part of a retrospective observational study of a concurrent MRSA outbreak in the United Kingdom with dominant endemic MRSA clones (ST22 and ST36) and an Asian ST239 MRSA strain (ST239-TW) in two linked adult intensive care units, and compared with an approach based on a fully parametric transmission model. The results provide support for the hypothesis that the clones responded differently to an infection control measure based on the use of topical antiseptics, which was more effective at reducing transmission of endemic clones. They also suggest that in one of the two ICUs patients colonized or infected with the ST239-TW MRSA clone had consistently higher risks of transmitting MRSA to patients free of MRSA. These findings represent some of the first quantitative evidence of enhanced transmissibility of a pandemic MRSA lineage, and highlight the potential value of tailoring hospital infection control measures to specific pathogen subtypes. Hide abstract
The tracking and projection of emerging epidemics is hindered by the disconnect between apparent epidemic dynamics, discernible from noisy and incomplete surveillance data, and the underlying, imperfectly observed, system. Behavior changes compound this, altering both true dynamics and reporting patterns, particularly for diseases with nonspecific symptoms, such as influenza. We disentangle these effects to unravel the hidden dynamics of the 2009 influenza A/H1N1pdm pandemic in London, where surveillance suggests an unusual dominant peak in the summer. We embed an age-structured model into a bayesian synthesis of multiple evidence sources to reveal substantial changes in contact patterns and health-seeking behavior throughout the epidemic, uncovering two similar infection waves, despite large differences in the reported levels of disease. We show how this approach, which allows for real-time learning about model parameters as the epidemic progresses, is also able to provide a sequence of nested projections that are capable of accurately reflecting the epidemic evolution. Hide abstract
BRITISH MEDICAL JOURNAL, 343 (oct05 3), pp. d5694-d5694. | Read more2011. Screening, isolation, and decolonisation strategies in the control of meticillin resistant Staphylococcus aureus in intensive care units: cost effectiveness evaluation
BACKGROUND: Surface-active antiseptics, such as chlorhexidine, are increasingly being used as part of intervention programs to prevent methicillin-resistant Staphylococcus aureus (MRSA) transmission, despite limited evidence and potential for resistance. We report on the effect of an antiseptic protocol on acquisition of both endemic MRSA and an outbreak strain of MRSA sequence type 239 (designated TW). METHODS: Interrupted time-series data on MRSA acquisitions in two 15-bed intensive care units were analyzed using segmented regression models to estimate the effects of sequential introduction of an educational campaign, cohorting, and a chlorhexidine-based antiseptic protocol on transmission of TW and non-TW MRSA strains. Representative TW and non-TW MRSA strains were assessed for carriage of qacA/B genes and antiseptic susceptibility. RESULTS: The antiseptic protocol was associated with a highly significant, immediate 70% reduction in acquisition of non-TW MRSA strains (estimated model-averaged incidence rate ratio, 0.3; 95% confidence interval, 0.19-0.47) and an increase in acquisition of TW MRSA strains (estimated model-averaged incidence rate ratio, 3.85; 95% confidence interval, 0.80-18.59). There was only weak evidence of an effect of other interventions on MRSA transmission. All TW MRSA strains (21 of 21 isolates) and <5% (1 of 21 isolates) of non-TW MRSA strains tested carried the chlorhexidine resistance loci qacA/B. In vitro chlorhexidine minimum bactericidal concentrations of TW strains were 3-fold higher than those of non-TW MRSA strains, and in vivo, only patients with non-TW MRSA demonstrated a reduction in the number of colonization sites in response to chlorhexidine treatment. CONCLUSION: A chlorhexidine-based surface antiseptic protocol can interrupt transmission of MRSA in the intensive care unit, but strains carrying qacA/B genes may be unaffected or potentially spread more rapidly. Hide abstract
BACKGROUND: Screening and isolation are central components of hospital methicillin-resistant Staphylococcus aureus (MRSA) control policies. Their prevention of patient-to-patient spread depends on minimizing undetected and unisolated MRSA-positive patient days. Estimating these MRSA-positive patient days and the reduction in transmission due to isolation presents a major methodological challenge, but is essential for assessing both the value of existing control policies and the potential benefit of new rapid MRSA detection technologies. Recent methodological developments have made it possible to estimate these quantities using routine surveillance data. METHODS: Colonization data from admission and weekly nares cultures were collected from eight single-bed adult intensive care units (ICUs) over 17 months. Detected MRSA-positive patients were isolated using single rooms and barrier precautions. Data were analyzed using stochastic transmission models and model fitting was performed within a Bayesian framework using a Markov chain Monte Carlo algorithm, imputing unobserved MRSA carriage events. RESULTS: Models estimated the mean percent of colonized-patient-days attributed to undetected carriers as 14.1% (95% CI (11.7, 16.5)) averaged across ICUs. The percent of colonized-patient-days attributed to patients awaiting results averaged 7.8% (6.2, 9.2). Overall, the ratio of estimated transmission rates from unisolated MRSA-positive patients and those under barrier precautions was 1.34 (0.45, 3.97), but varied widely across ICUs. CONCLUSIONS: Screening consistently detected >80% of colonized-patient-days. Estimates of the effectiveness of barrier precautions showed considerable uncertainty, but in all units except burns/general surgery and one cardiac surgery ICU, the best estimates were consistent with reductions in transmission associated with barrier precautions. Hide abstract
New England Journal of Medicine, 360 (1), pp. 20-31. | Read more2009. Decontamination of the Digestive Tract and Oropharynx in ICU Patients
The analysis of nosocomial infection data for communicable pathogens is complicated by two facts. First, typical pathogens more commonly cause asymptomatic colonization than overt disease, so transmission can be only imperfectly observed through a sequence of surveillance swabs, which themselves have imperfect sensitivity. Any given set of swab results can therefore be consistent with many different patterns of transmission. Second, data are often highly dependent: the colonization status of one patient affects the risk for others, and, in some wards, repeated admissions are common. Here, the authors present a method for analyzing typical nosocomial infection data consisting of results from arbitrarily timed screening swabs that overcomes these problems and enables simultaneous estimation of transmission and importation parameters, duration of colonization, swab sensitivity, and ward- and patient-level covariates. The method accounts for dependencies by using a mechanistic stochastic transmission model, and it allows for uncertainty in the data by imputing the imperfectly observed colonization status of patients over repeated admissions. The approach uses a Markov chain Monte Carlo algorithm, allowing inference within a Bayesian framework. The method is applied to illustrative data from an interrupted time-series study of vancomycin-resistant enterococci transmission in a hematology ward. Hide abstract
Assessments of the importance of different routes of HIV-1 (HIV) transmission are vital for prioritization of control efforts. Lack of consistent direct data and large uncertainty in the risk of HIV transmission from HIV-contaminated injections has made quantifying the proportion of transmission caused by contaminated injections in sub-Saharan Africa difficult and unavoidably subjective. Depending on the risk assumed, estimates have ranged from 2.5% to 30% or more. We present a method based on an age-structured transmission model that allows the relative contribution of HIV-contaminated injections, and other routes of HIV transmission, to be robustly estimated, both fully quantifying and substantially reducing the associated uncertainty. To do this, we adopt a Bayesian perspective, and show how prior beliefs regarding the safety of injections and the proportion of HIV incidence due to contaminated injections should, in many cases, be substantially modified in light of age-stratified incidence and injection data, resulting in improved (posterior) estimates. Applying the method to data from rural southwest Uganda, we show that the highest estimates of the proportion of incidence due to injections are reduced from 15.5% (95% credible interval) (0.7%, 44.9%) to 5.2% (0.5%, 17.0%) if random mixing is assumed, and from 14.6% (0.7%, 42.5%) to 11.8% (1.2%, 32.5%) under assortative mixing. Lower, and more widely accepted, estimates remain largely unchanged, between 1% and 3% (0.1-6.3%). Although important uncertainty remains, our analysis shows that in rural Uganda, contaminated injections are unlikely to account for a large proportion of HIV incidence. This result is likely to be generalizable to many other populations in sub-Saharan Africa. Hide abstract
Proc Natl Acad Sci U S A, 103 (33), pp. 12221-12222. | Read more2006. Poxy models and rash decisions.
BACKGROUND: The recent emergence of hypervirulent subtypes of avian influenza has underlined the potentially devastating effects of pandemic influenza. Were such a virus to acquire the ability to spread efficiently between humans, control would almost certainly be hampered by limited vaccine supplies unless global spread could be substantially delayed. Moreover, the large increases that have occurred in international air travel might be expected to lead to more rapid global dissemination than in previous pandemics. METHODS AND FINDINGS: To evaluate the potential of local control measures and travel restrictions to impede global dissemination, we developed stochastic models of the international spread of influenza based on extensions of coupled epidemic transmission models. These models have been shown to be capable of accurately forecasting local and global spread of epidemic and pandemic influenza. We show that under most scenarios restrictions on air travel are likely to be of surprisingly little value in delaying epidemics, unless almost all travel ceases very soon after epidemics are detected. CONCLUSIONS: Interventions to reduce local transmission of influenza are likely to be more effective at reducing the rate of global spread and less vulnerable to implementation delays than air travel restrictions. Nevertheless, under the most plausible scenarios, achievable delays are small compared with the time needed to accumulate substantial vaccine stocks. Hide abstract
Methicillin-resistant Staphylococcus aureus (MRSA) represents a serious threat to the health of hospitalized patients. Attempts to reduce the spread of MRSA have largely depended on hospital hygiene and patient isolation. These measures have met with mixed success: although some countries have almost eliminated MRSA or remained largely free of the organism, others have seen substantial increases despite rigorous control policies. We use a mathematical model to show how these increases can be explained by considering both hospital and community reservoirs of MRSA colonization. We show how the timing of the intervention, the level of resource provision, and chance combine to determine whether control measures succeed or fail. We find that even control measures able to repeatedly prevent sustained outbreaks in the short-term can result in long-term control failure resulting from gradual increases in the community reservoir. If resources do not scale with MRSA prevalence, isolation policies can fail "catastrophically." Hide abstract
Surveillance data for communicable nosocomial pathogens usually consist of short time series of low-numbered counts of infected patients. These often show overdispersion and autocorrelation. To date, almost all analyses of such data have ignored the communicable nature of the organisms and have used methods appropriate only for independent outcomes. Inferences that depend on such analyses cannot be considered reliable when patient-to-patient transmission is important. We propose a new method for analysing these data based on a mechanistic model of the epidemic process. Since important nosocomial pathogens are often carried asymptomatically with overt infection developing in only a proportion of patients, the epidemic process is usually only partially observed by routine surveillance data. We therefore develop a 'structured' hidden Markov model where the underlying Markov chain is generated by a simple transmission model. We apply both structured and standard (unstructured) hidden Markov models to time series for three important pathogens. We find that both methods can offer marked improvements over currently used approaches when nosocomial spread is important. Compared to the standard hidden Markov model, the new approach is more parsimonious, is more biologically plausible, and allows key epidemiological parameters to be estimated. Hide abstract
OBJECTIVE: To evaluate the evidence for the effectiveness of isolation measures in reducing the incidence of methicillin resistant Staphylococcus aureus (MRSA) colonisation and infection in hospital inpatients. DESIGN: Systematic review of published articles. DATA SOURCES: Medline, Embase, CINAHL, Cochrane Library, System for Information on Grey Literature in Europe (SIGLE), and citation lists (1966-2000). REVIEW METHODS: Articles reporting MRSA related outcomes and describing an isolation policy were selected. No quality restrictions were imposed on studies using isolation wards or nurse cohorting. Other studies were included if they were prospective or employed planned comparisons of retrospective data. RESULTS: 46 studies were accepted; 18 used isolation wards, nine used nurse cohorting, and 19 used other isolation policies. Most were interrupted time series, with few planned formal prospective studies. All but one reported multiple interventions. Consideration of potential confounders, measures to prevent bias, and appropriate statistical analysis were mostly lacking. No conclusions could be drawn in a third of studies. Most others provided evidence consistent with a reduction of MRSA acquisition. Six long interrupted time series provided the strongest evidence. Four of these provided evidence that intensive control measures including patient isolation were effective in controlling MRSA. In two others, isolation wards failed to prevent endemic MRSA. CONCLUSION: Major methodological weaknesses and inadequate reporting in published research mean that many plausible alternative explanations for reductions in MRSA acquisition associated with interventions cannot be excluded. No well designed studies exist that allow the role of isolation measures alone to be assessed. None the less, there is evidence that concerted efforts that include isolation can reduce MRSA even in endemic settings. Current isolation measures recommended in national guidelines should continue to be applied until further research establishes otherwise. Hide abstract
Severe acute respiratory syndrome (SARS) is a recently described illness of humans that has spread widely over the past 6 months. With the use of detailed epidemiologic data from Singapore and epidemic curves from other settings, we estimated the reproductive number for SARS in the absence of interventions and in the presence of control efforts. We estimate that a single infectious case of SARS will infect about three secondary cases in a population that has not yet instituted control measures. Public-health efforts to reduce transmission are expected to have a substantial impact on reducing the size of the epidemic. Hide abstract
Mathematical modelling of the spread of antibiotic-resistant bacteria in Southeast Asia
Antimicrobial resistance is one of the most serious public health threats in the world today. Multiply antibiotic-resistant bacteria are endemic in hospitals in South and Southeast Asia and high prevalences of community carriage of important antibiotic-resistant pathogens have also been reported. However, there are major gaps in our understanding of how such drug-resistance spreads, how this spread is affected by antibiotic use, and the ecological interactions that lead to the co-existence of ...