Search PubMed⌕ Search

Biomedical subjects

A E Ades

Publications and source records attributed to A E Ades.

At least 19 recordsLinked to original sources
PubMed2006

Evidence synthesis, parameter correlation and probabilistic sensitivity analysis.

Over the last decade or so, there have been many developments in methods to handle uncertainty in cost-effectiveness studies. In decision modelling, it is widely accepted that there needs to be an assessment of how sensitive the decision is to uncertainty in parameter values. The rationale for probabilistic sensitivity analysis (PSA) is primarily based on a consideration of the needs of decision makers in assessing the consequences of decision uncertainty. In this paper, we highlight some further compelling reasons for adopting probabilistic methods for decision modelling and sensitivity analysis, and specifically for adopting simulation from a Bayesian posterior distribution. Our reasoning is as follows. Firstly, cost-effectiveness analyses need to be based on all the available evidence, not a selected subset, and the uncertainties in the data need to be propagated through the model in order to provide a correct analysis of the uncertainties in the decision. In many--perhaps most--cases the evidence structure requires a statistical analysis that inevitably induces correlations between parameters. Deterministic sensitivity analysis requires that models are run with parameters fixed at 'extreme' values, but where parameter correlation exists it is not possible to identify sets of parameter values that can be considered 'extreme' in a meaningful sense. However, a correct probabilistic analysis can be readily achieved by Monte Carlo sampling from the joint posterior distribution of parameters. In this paper, we review some evidence structures commonly occurring in decision models, where analyses that correctly reflect the uncertainty in the data induce correlations between parameters. Frequently, this is because the evidence base includes information on functions of several parameters. It follows that, if health technology assessments are to be based on a correct analysis of all available data, then probabilistic methods must be used both for sensitivity analysis and for estimation of expected costs and benefits.

Bayes Theorem↗
PubMed2006

Bayesian methods for evidence synthesis in cost-effectiveness analysis.

Recently, health systems internationally have begun to use cost-effectiveness research as formal inputs into decisions about which interventions and programmes should be funded from collective resources. This process has raised some important methodological questions for this area of research. This paper considers one set of issues related to the synthesis of effectiveness evidence for use in decision-analytic cost-effectiveness (CE) models, namely the need for the synthesis of all sources of available evidence, although these may not 'fit neatly' into a CE model. Commonly encountered problems include the absence of head-to-head trial evidence comparing all options under comparison, the presence of multiple endpoints from trials and different follow-up periods. Full evidence synthesis for CE analysis also needs to consider treatment effects between patient subpopulations and the use of nonrandomised evidence. Bayesian statistical methods represent a valuable set of analytical tools to utilise indirect evidence and can make a powerful contribution to the decision-analytic approach to CE analysis. This paper provides a worked example and a general overview of these methods with particular emphasis on their use in economic evaluation.

Bayes Theorem↗
PubMed2004

Combination of direct and indirect evidence in mixed treatment comparisons.

Mixed treatment comparison (MTC) meta-analysis is a generalization of standard pairwise meta-analysis for A vs B trials, to data structures that include, for example, A vs B, B vs C, and A vs C trials. There are two roles for MTC: one is to strengthen inference concerning the relative efficacy of two treatments, by including both 'direct' and 'indirect' comparisons. The other is to facilitate simultaneous inference regarding all treatments, in order for example to select the best treatment. In this paper, we present a range of Bayesian hierarchical models using the Markov chain Monte Carlo software WinBUGS. These are multivariate random effects models that allow for variation in true treatment effects across trials. We consider models where the between-trials variance is homogeneous across treatment comparisons as well as heterogeneous variance models. We also compare models with fixed (unconstrained) baseline study effects with models with random baselines drawn from a common distribution. These models are applied to an illustrative data set and posterior parameter distributions are compared. We discuss model critique and model selection, illustrating the role of Bayesian deviance analysis, and node-based model criticism. The assumptions underlying the MTC models and their parameterization are also discussed.

Humans↗
PubMed2003

A chain of evidence with mixed comparisons: models for multi-parameter synthesis and consistency of evidence.

Multi-parameter evidence synthesis is a generalization of meta-analysis in which several parameters are estimated jointly. Here this approach is applied to a common form of data structure in which mixed pairwise comparisons are made between treatments in a 'chain of evidence' structure. The data set investigated here, which first appeared in the confidence profile method literature, features three types of data relating to thrombolytic treatment following acute myocardial infarction. There is information on reperfusion (coronary patency) following treatment, information on survival in reperfused and non-reperfused patients (conditional survival), and independent information on the effect of treatment on survival without specifying reperfusion state (overall survival). The objective of this study is to explore models for combining these three types of evidence within a single model, and some of the evidence consistency issues that arise. Bayesian Markov chain Monte Carlo methods are used to fit two models. The first, proposed in the confidence profile literature, assumes there are no differences between studies in baseline effects (equal study effects). The second model assumes fixed treatment effects but allows for study differences in baseline reperfusion and conditional survival rates by assuming random study effects. The equal study effects model fits the data poorly, and cross-validation shows that the overall survival evidence is not consistent with the reperfusion and conditional survival evidence under this model. The evidence appears to be consistent within the more loosely structured random study effects model, but results are sensitive to prior assumptions about the between-study variance. The inconsistency between the three evidence types relates to overall survival, rather than to the effect of treatment on survival. Multi-parameter evidence synthesis, with suitable checks for evidence consistency, can be used to help determine whether chains of evidence 'add up' epidemiologically within an overall model, and to construct parameter distributions for an entire model simultaneously, consistent with all the available data.

Bayes Theorem↗
PubMed2003

Comparison of methods for analysing cluster randomized trials: an example involving a factorial design.

BACKGROUND: Studies involving clustering effects are common, but there is little consistency in their analysis. Various analytical methods were compared for a factorial cluster randomized trial (CRT) of two primary care-based interventions designed to increase breast screening attendance. METHODS: Three cluster-level and five individual-level options were compared in respect of log odds ratios of attendance and their standard errors (SE), for the two intervention effects and their interaction. Cluster-level analyses comprised: (C1) unweighted regression of practice log odds; (C2) regression of log odds weighted by their inverse variance; (C3) random-effects meta-regression of log odds with practice as a random effect. Individual-level analyses comprised: (I1) standard logistic regression ignoring clustering; (I2) robust SE; (I3) generalized estimating equations; (I4) random-effects logistic regression; (I5) Bayesian random-effects logistic regression. Adjustments for stratification and baseline variables were investigated. RESULTS: As expected, method I1 was highly anti-conservative. The other, valid, methods exhibited considerable differences in parameter estimates and standard errors, even between the various random-effects methods based on the same statistical model. Method I4 was particularly sensitive to between-cluster variation and was computationally stable only after controlling for baseline uptake. CONCLUSIONS: Commonly used methods for the analysis of CRT can give divergent results. Simulation studies are needed to compare results from different methods in situations typical of cluster trials but when the true model parameters are known.

Breast Neoplasms↗
PubMed2003

Correlations between parameters in risk models: estimation and propagation of uncertainty by Markov Chain Monte Carlo.

Monte Carlo simulation has become the accepted method for propagating parameter uncertainty through risk models. It is widely appreciated, however, that correlations between input variables must be taken into account if models are to deliver correct assessments of uncertainty in risk. Various two-stage methods have been proposed that first estimate a correlation structure and then generate Monte Carlo simulations, which incorporate this structure while leaving marginal distributions of parameters unchanged. Here we propose a one-stage alternative, in which the correlation structure is estimated from the data directly by Bayesian Markov Chain Monte Carlo methods. Samples from the posterior distribution of the outputs then correctly reflect the correlation between parameters, given the data and the model. Besides its computational simplicity, this approach utilizes the available evidence from a wide variety of structures, including incomplete data and correlated and uncorrelated repeat observations. The major advantage of a Bayesian approach is that, rather than assuming the correlation structure is fixed and known, it captures the joint uncertainty induced by the data in all parameters, including variances and covariances, and correctly propagates this through the decision or risk model. These features are illustrated with examples on emissions of dioxin congeners from solid waste incinerators.

Journal Article↗
PubMed2001

Screening for Down's syndrome: effects, safety, and cost effectiveness of first and second trimester strategies.

OBJECTIVE: To compare the effects, safety, and cost effectiveness of antenatal screening strategies for Down's syndrome. DESIGN: Analysis of incremental cost effectiveness. SETTING: United Kingdom. MAIN OUTCOME MEASURES: Number of liveborn babies with Down's syndrome, miscarriages due to chorionic villus sampling or amniocentesis, health care costs of screening programme, and additional costs and additional miscarriages per additional affected live birth prevented by adopting a more effective strategy. RESULTS: Compared with no screening, the additional cost per additional liveborn baby with Down's syndrome prevented was 22 000 pound sterling for measurement of nuchal translucency. The cost of the integrated test was 51 000 pound sterling compared with measurement of nuchal translucency. All other strategies were more costly and less effective, or cost more per additional affected baby prevented. Depending on the cost of the screening test, the first trimester combined test and the quadruple test would also be cost effective options. CONCLUSIONS: The choice of screening strategy should be between the integrated test, first trimester combined test, quadruple test, or nuchal translucency measurement depending on how much service providers are willing to pay, the total budget available, and values on safety. Screening based on maternal age, the second trimester double test, and the first trimester serum test was less effective, less safe, and more costly than these four options.

Abortion, Spontaneous↗
PubMed2001

Effect of prenatal treatment on mother to child transmission of Toxoplasma gondii: retrospective cohort study of 554 mother-child pairs in Lyon, France.

BACKGROUND: The aim of prenatal serological screening for toxoplasmosis is to identify and treat maternal infection as soon as possible in order to prevent transmission of the parasite to the fetus. However, despite widespread provision of prenatal toxoplasma screening across Europe, the effectiveness of prenatal treatment is uncertain. The study aimed to determine the effect of the timing and type of prenatal treatment on mother to child transmission of Toxoplasma gondii. METHOD: A cohort of 554 infected pregnant women were identified in Lyon, France between 1987 and 1995 and their children were followed to determine congenital infection status. We determined the effect of prenatal treatment on transmission by examining the effect of the delay between maternal seroconversion and start of treatment. We also compared the effect of the type of treatment and no treatment on the risk of mother to child transmission. Analyses were adjusted for gestation at maternal seroconversion. RESULTS: Compared to treatment within 4 weeks from seroconversion, the adjusted odds ratios (OR) for mother to child transmission after a treatment delay of 4-7 weeks was 1.29 (95% CI : 0.61, 2.73) and after more than 8 weeks, 1.44 (95% CI : 0.60, 3.31). The adjusted OR associated with spiramycin alone compared with pyrimethamine-sulfadiazine treatment was 0.91 (95% CI : 0.45, 1.84) and the OR for no treatment compared with pyrimethamine-sulfadiazine treatment was 1.06 (95% CI : 0.37, 3.03). CONCLUSIONS: The authors hypothesize that the absence of an effect of prenatal treatment is due to transmission before the start of treatment.

Adult↗
PubMed2001

Effect of prenatal treatment on the risk of intracranial and ocular lesions in children with congenital toxoplasmosis.

BACKGROUND: Hydrocephalus, intracranial calcification and retinochoroiditis are the most common manifestations of tissue damage due to congenital toxoplasmosis, but the effect of prenatal treatment on these outcomes is unclear. We aimed to determine the effect of prenatal treatment for toxoplasmosis on the risk of intracranial and ocular lesions in congenitally infected children at 3 years of age. METHODS: A cohort of mothers identified during pregnancy with toxoplasma infection and their 181 liveborn children with confirmed congenital toxoplasmosis was retrospectively analysed to determine the presence of intracranial and ocular lesions. As few women are not treated, we compared the effects of the treatment potency (pyrimethamine-sulfadiazine versus spiramycin or no treatment), and the timing of treatment, on the risks of intracranial lesions, time to detection of ocular lesions, and detection of any lesions (intracranial or ocular) by 3 years of age. Analyses took account of the gestation at maternal seroconversion. RESULTS: There was no evidence for an effect of pyrimethamine-sulfadiazine on intracranial, ocular or any lesions by 3 years: odds ratio (OR) for any lesions 0.89 (95% CI : 0.41, 1.88). There was no evidence of an effect of delayed treatment on ocular lesions (hazard ratio = 0.69, 95% CI : 0.28, 1.68) or any lesions by 3 years of age (OR = 0.44, 95% CI : 0.16, 1.19). CONCLUSIONS: Our study failed to detect a beneficial effect of early or more potent anti toxoplasma treatment on the risks of intracranial or ocular lesions in children with congenital toxoplasmosis. However, larger, prospective studies, which determine the effect of prenatal treatment on long-term developmental outcomes are required to justify changes in clinical practice.

Anti-Bacterial Agents↗
PubMed2001

Coverage of neonatal screening: failure of coverage or failure of information system.

OBJECTIVES: To evaluate neonatal screening coverage using data routinely collected on the laboratory computer. SUBJECTS: 90 850 births in 14 North East Thames community provider districts over a 21 month period. METHODS: Births notified to local child health computers are electronically copied to the neonatal laboratory computer system, and incoming Guthrie cards are matched against these birth records before testing. The computer records for the study period were processed to estimate the coverage of the screening programme. RESULTS: Out of an estimated 90 850 births notified to child health computers, all but 746 (0.82%) appeared to have been screened or could be otherwise accounted for (0.14% in non-metropolitan districts, 0.39% in suburban districts, and 1.68% in inner city districts). A further 893 resident infants had been tested, but could not be matched to the list of notified resident births. The calculated programme coverage already exceeds the 99.5% National Audit Programme standard in 7/14 districts. Elsewhere it is not clear whether it is coverage or recording of coverage that is low. CONCLUSION: Previous reports of low coverage may have been exaggerated. High coverage can be shown using routine information systems. Design of information systems that deliver accurate measures of coverage would be more useful than comparison of inadequately measured coverage with a national standard. The new NHS number project will create an opportunity to achieve this.

Birth Certificates↗
PubMed2000

Human T cell leukaemia/lymphoma virus infection in pregnant women in the United Kingdom: population study.

OBJECTIVE: To assess the prevalence of human T cell leukaemia/lymphoma virus (HTLV) infection in pregnant women in the United Kingdom. DESIGN: Population study. SUBJECTS: Guthrie card samples from babies born in 1997-8. Samples were linked to data on mother's age and ethnic status and parents' country of birth and then anonymised. SETTING: North Thames Regional Health Authority. MAIN OUTCOME MEASURES: Presence of antibodies against HTLV in eluates tested by gelatin particle agglutination assay and results confirmed by immunoblot. RESULTS: Of 126 010 samples tested, 67 had confirmed antibodies to HTLV (59 HTLV-I, 2 HTLV-II, 6 untyped) and six had indeterminate results. Seroprevalence was 17.0 per 1000 (95% confidence interval 9.2 to 28.3) in infants whose mothers were born in the Caribbean, 3.2/1000 (1.5 to 5.9) with mothers born in west and central Africa, and 6.8/1000 (3.1 to 12.9) in infants of black Caribbean mothers born in non-endemic regions. In infants with no known risk (both parents born in non-endemic regions and mother not black Caribbean) seroprevalence was 0.06-0.12 per 1000. Mother's country of birth, father's country of birth, and mother's ethnic status were all independently associated with neonatal seroprevalence. An estimated 223 (95% confidence interval 110 to 350) of the 720 000 pregnant women each year in the United Kingdom are infected with HTLV. CONCLUSIONS: The prevalence of HTLV and HIV infections in pregnant women in the United Kingdom are comparable. The cost effectiveness of antenatal HTLV screening should be evaluated, and screening of blood donations should be considered.

Adult↗
PubMed2000

HCV prevalence in pregnant women in the UK.

The objective of the study was to assess the prevalence and epidemiology of hepatitis C virus (HCV) infection in pregnant women in the North Thames region, and in the UK in general. Demographic data were linked to neonatal samples prior to anonymization and testing by anti-HCV EIA, and with RIBA 3 confirmation. Risk factors for maternal infection were explored. Area-specific seroprevalence rates were multiplied into population sizes to estimate HCV prevalence in pregnant women in the UK. A total of 241/126,009 samples were confirmed anti-HCV positive, and a further 40 were indeterminate, representing a seroprevalence of 0.19-0.22%; 51% of maternal HCV infections were in UK-born women (71% of the population), and 22% in women from continental Europe (5% of the population). Among European-born women, HCV prevalence was associated with birth in continental Europe, partner not being notified at birth registration, partner born in a different region to the mother, and inner city residence. Four of the 241 anti-HCV positive samples (1.7%) were also anti-HIV-1 positive. It was estimated that each year an estimated 1150 out of 730,000 pregnancies in the UK would involve a woman infected with HCV (uncertainty range 660-1850), a prevalence of 0.16% (0.09-0.25%). On the basis of reported rates of mother-to-child transmission of HCV, this would represent approximately 70 paediatric HCV infections per year.

Adolescent↗
PubMed2000

The public health applications of unlinked anonymous seroprevalence monitoring for HIV in the United Kingdom.

BACKGROUND: In order to monitor the epidemiology of human immunodeficiency virus (HIV), integrated national programmes of unlinked anonymous (blinded) HIV sero-surveys have taken place in the UK since 1990. METHODS: The programmes comprise multi-centre surveys primarily using specimens gathered routinely for screening groups of patients. All specimens are irreversibly unlinked from patient identifiers before being tested. RESULTS: The surveys have met their prime aim of providing at low cost minimally biased estimates of current HIV prevalence and trends in sentinel populations. The surveys have remained acceptable to professionals and the public, being successfully implemented without breech of their founding principles. The findings have had major public health applications, have influenced HIV policy and funding, been used for monitoring the spread of HIV, for targeting and evaluating health promotion and improving projections of severe HIV disease. The surveys have detected substantial prevalence rises and under-diagnosis of HIV which would otherwise have been unrecognised. The programmes' value is being increased by sub-typing HIV-1 isolates, capturing additional demographic information to detect spread among minority groups. The same specimens are used for monitoring other infections (initially hepatitis A, B and C). CONCLUSIONS: Monitoring HIV and other infections through unlinked anonymous HIV surveillance has become an integral essential part of national HIV and AIDS surveillance. Although it has unique applications the value of unlinked anonymous surveillance is maximized when used in conjunction with behavioural data, information from HIV and AIDS reporting, and behavioural data and surveillance for other sexually transmitted infections.

Confidentiality↗
PubMed2000

The effects of neonatal screening for sickle cell disorders on lifetime treatment costs and early deaths avoided: a modelling approach.

BACKGROUND: The aim of the study was to calculate the cost to the UK National Health Service of providing treatment services for patients with sickle cell disorders. The rates of differential morbidity and mortality, in the first 10 years of life, between screen-detected early diagnosed and clinically presenting late diagnosed cohorts of sickle cell disorder patients are also estimated. METHOD: A cost model was developed, based on predictions of survival and the incidence of sickle cell disorder-related events. Direct data from the NHS are lacking, so data were incorporated from disparate sources. Patients with sickle cell disorders were divided into two categories: those with sickle cell anaemia and those with sickle HbC disease. RESULTS: Differentiating between sickle cell anaemia and sickle HbC disorder patients, the results show that the undiscounted (discounted at 6 per cent) lifetime treatment costs range from pound sterling 92323 (pound sterling 24917) to pound sterling 185614 (pound sterling 53861). The number of early deaths avoided per 100 births, as a result of early diagnosis through screening, ranges from 0.57 to 1.25. CONCLUSIONS: The resulting estimates may act as a guide to those involved in the planning of health care provision with regard to the resources required to treat sickle cell disorder patients. Such information may also be incorporated into the evaluation of both antenatal and neonatal screening programmes for sickle cell disorders.

Adolescent↗
PubMed2000

Economic issues in the prevention of vertical transmission of HIV.

In the absence of interventions, 20% of infants born to women infected with HIV acquire infection from their mother at or before delivery. A further 15% are infected through breast feeding. Prenatal testing for HIV allows infected women to be reliably identified so that they can receive antiretroviral therapy and, in countries with safe water supplies, be advised not to breast feed. These and other interventions can reduce the risk of transmission to 5% or less. Economic evaluations of prenatal testing for HIV are reviewed and compared in this article, and future research priorities outlined. These studies set the costs of testing and intervention against the averted lifetime costs of paediatric infection, and generate estimates of the HIV prevalence threshold above which there would be a net cost saving, or calculate the cost per life-year saved given a particular prevalence. In the developed world, prenatal testing has been adopted in many countries, and recent economic analyses broadly support this. Future research is likely to focus on the incremental benefits of different antiretroviral regimens in lowering transmission rates still further, with or without elective caesarean section, and the possibility that some may lead to adverse effects in uninfected infants exposed to them in utero. Some earlier assessments in resource-poor settings concluded that prenatal testing was unaffordable or of doubtful cost effectiveness. This negative conclusion appears to be the result of very low estimates of the lifetime costs of paediatric HIV infection, together with developed world conceptions of pre-test counselling. The demonstration that nevirapine reduces transmission risk at a low cost has transformed the outlook, and there is hope that antiretrovirals can act prophylactically to prevent infection of the breast-fed child. However, to achieve a sustained reduction in vertical transmission there may be a need to evaluate the need for a strengthened infrastructure to deliver prenatal HIV testing and treatment, as well as programmes to reduce HIV incidence in adults.

HIV Infections↗