Edmond S. W. Ng

Randomised comparison of three methods of administering a screening questionnaire to elderly people: findings from the MRC trial of the assessment and management of older people in the community

Abstract Objective: To compare three different methods of administering a brief screening questionnaire to elderly people: post, interview by lay interviewer, and interview by nurse. Design: Randomised comparison of methods within a cluster randomised trial. Setting: 106 general practices in the United Kingdom. Participants: 32 990 people aged 75 years or over registered with participating practices. Main outcome measures: Response rates, proportion of missing values, prevalence of self reported morbidity, and sensitivity and specificity of self reported measures by method of administration of questionnaire for four domains. Results: The response rate was higher for the postal questionnaire than for the two interview methods combined (83.5% v 74.9%; difference 8.5%, 95% confidence interval 4.4% to 12.7%, P<0.001). The proportion of missing or invalid responses was low overall (mean 2.1%) but was greater for the postal method than for the interview methods combined (4.1% v 0.9%; difference 3.2%, 2.7% to 3.6%, P<0.001). With a few exceptions, levels of self reported morbidity were lower in the interview groups, particularly for interviews by nurses. The sensitivity of the self reported measures was lower in the nurse interview group for three out of four domains, but 95% confidence intervals for the estimates overlapped. Specificity of the self reported measures varied little by method of administration. Conclusions: Postal questionnaires were associated with higher response rates but also higher proportions of missing values than were interview methods. Lower estimates of self reported morbidity were obtained with the nurse interview method and to a lesser extent with the lay interview method than with postal questionnaires. What is already known on this topic The optimum method of administering a brief multidimensional screening assessment to elderly people is not known What this study adds Postal questionnaires produce a higher response rate than interviews by nurses or lay interviewers but also higher proportions of missing data Interview by nurses and to a lesser degree by lay interviewers is associated with lower levels of self reported morbidity than are postal questionnaires

Developing Appropriate Methods for Cost-Effectiveness Analysis of Cluster Randomized Trials

Aim. Cost-effectiveness analyses (CEAs) may use data from cluster randomized trials (CRTs), where the unit of randomization is the cluster, not the individual. However, most studies use analytical methods that ignore clustering. This article compares alternative statistical methods for accommodating clustering in CEAs of CRTs. Methods. Our simulation study compared the performance of statistical methods for CEAs of CRTs with 2 treatment arms. The study considered a method that ignored clustering—seemingly unrelated regression (SUR) without a robust standard error (SE)—and 4 methods that recognized clustering—SUR and generalized estimating equations (GEEs), both with robust SE, a “2-stage” nonparametric bootstrap (TSB) with shrinkage correction, and a multilevel model (MLM). The base case assumed CRTs with moderate numbers of balanced clusters (20 per arm) and normally distributed costs. Other scenarios included CRTs with few clusters, imbalanced cluster sizes, and skewed costs. Performance was reported as bias, root mean squared error (rMSE), and confidence interval (CI) coverage for estimating incremental net benefits (INBs). We also compared the methods in a case study. Results. Each method reported low levels of bias. Without the robust SE, SUR gave poor CI coverage (base case: 0.89 v. nominal level: 0.95). The MLM and TSB performed well in each scenario (CI coverage, 0.92–0.95). With few clusters, the GEE and SUR (with robust SE) had coverage below 0.90. In the case study, the mean INBs were similar across all methods, but ignoring clustering underestimated statistical uncertainty and the value of further research. Conclusions. MLMs and the TSB are appropriate analytical methods for CEAs of CRTs with the characteristics described. SUR and GEE are not recommended for studies with few clusters.

Estimation in generalised linear mixed models with binary outcomes by simulated maximum likelihood

Fitting multilevel models to discrete outcome data is problematic because the discrete distribution of the response variable implies an analytically intractable log-likelihood function. Among a number of approximate methods proposed, second-order penalised quasi-likelihood (PQL) is commonly used and is one of the most accurate. Unfortunately, even the second-order PQL approximation has been shown to produce estimates biased toward zero in certain circumstances. This bias can be marked especially when the data are sparse. One option to reduce this bias is to use Monte-Carlo simulation. A bootstrap bias correction method proposed by Kuk has been implemented in MLwiN. However, a similar technique based on the Robbins-Monro (RM) algorithm is potentially more efficient. An alternative is to use simulated maximum likelihood (SML), either alone or to refine estimates identified by other methods. In this article, we first compare bias correction using the RM algorithm, Kuk’s method and SML. We find that SML performs as efficiently as the other two methods and also yields standard errors of the bias-corrected parameter estimates and an estimate of the log-likelihood at the maximum, with which nested models can be compared. Secondly, using simulated and real data examples, we compare SML, second-order Laplace approximation (as implemented in HLM), Markov Chain Monte-Carlo (MCMC) (in MLwiN) and numerical integration using adaptive quadrature methods (in Stata’s GLLAMM and in SAS’s proc NLMIXED). We find that when the data are sparse, the second-order Laplace approximation produces markedly lower parameter estimates, whereas the MCMC method produces estimates that are noticeably higher than those from the SML and quadrature methods. Although proc NLMIXED is much faster than GLLAMM, it is not designed to fit models of more than two levels. SML produces parameter estimates and log-likelihoods very similar to those from quadrature methods. Further our SML approach extends to handle other link functions, discrete data distributions, non-normal random effects and higher-level models.

The efficiency of cardiovascular risk assessment: do the right patients get statin treatment?

Objective To evaluate targeting of statin prescribing for primary prevention to those with high cardiovascular disease (CVD) risk. Design Two cohort studies including the general population and initiators of statins aged 35–74 years. Setting UK primary care records in the Clinical Practice Research Datalink. Patients 3.8 million general population patients and 300 914 statin users. Intervention Statin prescribing. Main outcome measures Statin prescribing by CVD risk; observed 5-year CVD risks; variability between practices. Results Statin prescribing increased substantially over time to patients with high 10-year CVD risk (≥20%): 7.0% of these received a statin prior to 2007, and 30.4% in 2007 onwards. Prescribing to patients with low risk (<15%) also increased (from 1.9% to 5.0%). Only about half the patients initiating statin treatment were high risk according to CVD risk score. The 5-year CVD risks, as observed during statin treatment, reduced over calendar time (from 17.0% to 7.1%). There was a large variation between general practices in the percentage of high-risk patients prescribed a statin in 2007 onwards, ranging from 8.2% to 61.5%. For low-risk patients, these varied from 2.1% to 29.1%. Conclusions There appeared to be substantive overuse in low CVD risk and underuse in high CVD risk (600 000 and 850 000 patients, respectively, in the UK since 2007). There is wide variation between practices in statin prescribing to patients at high CVD risk. There is a clear need for randomised trials for the best strategy to target statin treatment and manage CVD risk for primary prevention.

Prediction of Cardiovascular Risk Using Framingham, ASSIGN and QRISK2: How Well Do They Predict Individual Rather than Population Risk?

Background The objective of this study was to evaluate the performance of risk scores (Framingham, Assign and QRISK2) in predicting high cardiovascular disease (CVD) risk in individuals rather than populations. Methods and findings This study included 1.8 million persons without CVD and prior statin prescribing using the Clinical Practice Research Datalink. This contains electronic medical records of the general population registered with a UK general practice. Individual CVD risks were estimated using competing risk regression models. Individual differences in the 10-year CVD risks as predicted by risk scores and competing risk models were estimated; the population was divided into 20 subgroups based on predicted risk. CVD outcomes occurred in 69,870 persons. In the subgroup with lowest risks, risk predictions by QRISK2 were similar to individual risks predicted using our competing risk model (99.9% of people had differences of less than 2%); in the subgroup with highest risks, risk predictions varied greatly (only 13.3% of people had differences of less than 2%). Larger deviations between QRISK2 and our individual predicted risks occurred with calendar year, different ethnicities, diabetes mellitus and number of records for medical events in the electronic health records in the year before the index date. A QRISK2 estimate of low 10-year CVD risk (<15%) was confirmed by Framingham, ASSIGN and our individual predicted risks in 89.8% while an estimate of high 10-year CVD risk (≥20%) was confirmed in only 48.6% of people. The majority of cases occurred in people who had predicted 10-year CVD risk of less than 20%. Conclusions Application of existing CVD risk scores may result in considerable misclassification of high risk status. Current practice to use a constant threshold level for intervention for all patients, together with the use of different scoring methods, may inadvertently create an arbitrary classification of high CVD risk.

The path to longer and healthier lives for all Africans by 2030: the Lancet Commission on the future of health in sub-Saharan Africa.

Sub-Saharan Africa’s health challenges are numerous and wide-ranging. Most sub-Saharan African countries face a double burden of traditional, persisting health challenges, such as infectious diseases, malnutrition, and child and maternal mortality, and emerging challenges from an increasing prevalence of chronic conditions, mental health disorders, injuries, and health problems related to climate change and environmental degradation. Although there has been real progress on many health indicators, life expectancy and most population health indicators remain behind most low-income and middle-income countries in other parts of the world. Our Commission was prompted by sub-Saharan Africa’s potential to improve health on its own terms, and largely with its own resources. The spirit of this Commission is one of evidence-based optimism, with caution. We recognise that major health inequities exist and that health outcomes are worst in fragile countries, rural areas, urban slums, and conflict zones, and among the poor, disabled, and marginalised. Moreover, sub- Saharan Africa is facing the challenges and opportunities of the largest cohort of young people in history, with the youth population aged under 25 years predicted to almost double from 230 million to 450 million by 2050. The future of health in Africa is bright, but only if no one is left behind.

Multilevel models for cost-effectiveness analyses that use cluster randomised trial data: An approach to model choice.

Multilevel models provide a flexible modelling framework for cost-effectiveness analyses that use cluster randomised trial data. However, there is a lack of guidance on how to choose the most appropriate multilevel models. This paper illustrates an approach for deciding what level of model complexity is warranted; in particular how best to accommodate complex variance–covariance structures, right-skewed costs and missing data. Our proposed models differ according to whether or not they allow individual-level variances and correlations to differ across treatment arms or clusters and by the assumed cost distribution (Normal, Gamma, Inverse Gaussian). The models are fitted by Markov chain Monte Carlo methods. Our approach to model choice is based on four main criteria: the characteristics of the data, model pre-specification informed by the previous literature, diagnostic plots and assessment of model appropriateness. This is illustrated by re-analysing a previous cost-effectiveness analysis that uses data from a cluster randomised trial. We find that the most useful criterion for model choice was the deviance information criterion, which distinguishes amongst models with alternative variance–covariance structures, as well as between those with different cost distributions. This strategy for model choice can help cost-effectiveness analyses provide reliable inferences for policy-making when using cluster trials, including those with missing data.

Local, national, and regional viral haemorrhagic fever pandemic potential in Africa: a multistage analysis

Summary Background Predicting when and where pathogens will emerge is difficult, yet, as shown by the recent Ebola and Zika epidemics, effective and timely responses are key. It is therefore crucial to transition from reactive to proactive responses for these pathogens. To better identify priorities for outbreak mitigation and prevention, we developed a cohesive framework combining disparate methods and data sources, and assessed subnational pandemic potential for four viral haemorrhagic fevers in Africa, Crimean–Congo haemorrhagic fever, Ebola virus disease, Lassa fever, and Marburg virus disease. Methods In this multistage analysis, we quantified three stages underlying the potential of widespread viral haemorrhagic fever epidemics. Environmental suitability maps were used to define stage 1, index-case potential, which assesses populations at risk of infection due to spillover from zoonotic hosts or vectors, identifying where index cases could present. Stage 2, outbreak potential, iterates upon an existing framework, the Index for Risk Management, to measure potential for secondary spread in people within specific communities. For stage 3, epidemic potential, we combined local and international scale connectivity assessments with stage 2 to evaluate possible spread of local outbreaks nationally, regionally, and internationally. Findings We found epidemic potential to vary within Africa, with regions where viral haemorrhagic fever outbreaks have previously occurred (eg, western Africa) and areas currently considered non-endemic (eg, Cameroon and Ethiopia) both ranking highly. Tracking transitions between stages showed how an index case can escalate into a widespread epidemic in the absence of intervention (eg, Nigeria and Guinea). Our analysis showed Chad, Somalia, and South Sudan to be highly susceptible to any outbreak at subnational levels. Interpretation Our analysis provides a unified assessment of potential epidemic trajectories, with the aim of allowing national and international agencies to pre-emptively evaluate needs and target resources. Within each country, our framework identifies at-risk subnational locations in which to improve surveillance, diagnostic capabilities, and health systems in parallel with the design of policies for optimal responses at each stage. In conjunction with pandemic preparedness activities, assessments such as ours can identify regions where needs and provisions do not align, and thus should be targeted for future strengthening and support. Funding Paul G Allen Family Foundation, Bill & Melinda Gates Foundation, Wellcome Trust, UK Department for International Development.

Risk patterns in drug safety study using relative times by accelerated failure time models when proportional hazards assumption is questionable : an illustrative case study of cancer risk of patients on glucose-lowering therapies

Observational drug safety studies may be susceptible to confounding or protopathic bias. This bias may cause a spurious relationship between drug exposure and adverse side effect when none exists and may lead to unwarranted safety alerts. The spurious relationship may manifest itself through substantially different risk levels between exposure groups at the start of follow-up when exposure is deemed too short to have any plausible biological effect of the drug. The restrictive proportional hazards assumption with its arbitrary choice of baseline hazard function renders the commonly used Cox proportional hazards model of limited use for revealing such potential bias. We demonstrate a fully parametric approach using accelerated failure time models with an illustrative safety study of glucose-lowering therapies and show that its results are comparable against other methods that allow time-varying exposure effects. Our approach includes a wide variety of models that are based on the flexible generalized gamma distribution and allows direct comparisons of estimated hazard functions following different exposure-specific distributions of survival times. This approach lends itself to two alternative metrics, namely relative times and difference in times to event, allowing physicians more ways to communicate patients prognosis without invoking the concept of risks, which some may find hard to grasp. In our illustrative case study, substantial differences in cancer risks at drug initiation followed by a gradual reduction towards null were found. This evidence is compatible with the presence of protopathic bias, in which undiagnosed symptoms of cancer lead to switches in diabetes medication.

Effectiveness of trauma-focused psychological therapies compared to usual postnatal care for treating post-traumatic stress symptoms in women following traumatic birth: A systematic review protocol

Introduction Maternal mental health has been largely neglected in the literature. Women, however, may be vulnerable to developing post-traumatic stress symptoms or post-traumatic stress disorder (PTSD), following traumatic birth. In turn, this may affect their capacity for child rearing and ability to form a secure bond with their baby and impact on the wider family. Trauma-focused psychological therapies (TFPT) are widely regarded as effective and acceptable interventions for PTSD in general and clinical populations. Relatively little is known about the effectiveness of TFPT for women postpartum who have post-traumatic stress symptoms. Methods and analysis We will conduct a review to assess the effectiveness of TFPT, compared with usual postpartum care, as a treatment for post-traumatic stress symptoms or PTSD for women following traumatic birth. Using a priori search criteria, we will search for randomised controlled trials (RCT) in four databases: Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, PsycINFO and OpenGrey. We will use search terms that relate to the population, TFPT and comparators. Screening of search results and data extraction will be undertaken by two reviewers, independently. Risk of bias will be assessed in RCTs which meet the review criteria. Data will be analysed using the following methods, as appropriate: narrative synthesis; meta-analysis; subgroup analysis and meta-regression. Dissemination and ethics As this work comprises a synthesis of existing studies, ethical approvals are not required. Results will be disseminated at conferences and in publications.