Pete Philipson

Evaluating the quality of research into a single prognostic biomarker: a systematic review and meta-analysis of 83 studies of C-reactive protein in stable coronary artery disease.

In a systematic review and meta-analysis of 83 prognostic studies of C-reactive protein in coronary disease, Hemingway and colleagues find substantial biases, preventing them from drawing clear conclusions relating to the use of this marker in clinical practice.

Joint modelling of longitudinal and competing risks data.

Available methods for joint modelling of longitudinal and survival data typically have only one failure type for the time to event outcome. We extend the methodology to allow for competing risks data. We fit a cause-specific hazards sub-model to allow for competing risks, with a separate latent association between longitudinal measurements and each cause of failure.The method is applied to data from the SANAD trial of anti-epileptic drugs (AEDs), as a means of investigating the effect of drug titration on the relative effects of lamotrigine (LTG) and carbamazepine (CBZ) on treatment failure. Concern had been expressed that differential titration rates may have been to the disadvantage of CBZ. The beneficial effect of LTG on unacceptable adverse events leading to drug withdrawal did not lessen and indeed increased slightly when a calibrated dose was accounted for in the joint model. Adjustment for the titration rate of LTG relative to CBZ resulted in an unchanged effect of the former on drug withdrawals due to inadequate seizure control. LTG remains the AED of choice from this analysis.

South Asians and coronary disease: is there discordance between effects on incidence and prognosis?

Objective To determine whether the effect of South Asian ethnicity differs between studies of incidence and prognosis of coronary disease. Design Systematic literature review and meta-analysis, and cohort analysis from a national acute coronary syndrome (ACS) registry linked to mortality (National Institute of Cardiovascular Outcomes Research/Myocardial Infarction National Audit Project). Setting International for the review, and England and Wales for the cohort analysis. Patients The numbers of South Asians included in the meta-analysis were 111 555 (incidence) and 14 531 (prognosis) of whom 8251 were from the ACS cohort. Main outcome measures Incidence studies: non-fatal myocardial infarction or fatal coronary heart disease; prognostic studies: mortality; HRs for 1-year all-cause death in ACS cohort. Results South Asians had higher incidence of coronary disease compared with white subjects (HR 1.35 95% CI 1.30 to 1.40) based on meta-analysis of nine studies. Among 10 studies on prognosis, South Asians had better prognosis compared with white subjects (HR 0.78 95% CI 0.74 to 0.82). In the ACS cohort, the impact of diabetes (42.4% of South Asians, 16.9% of white subjects) on 1-year mortality was stronger in South Asians than white subjects (age-adjusted HR 1.83 95% CI 1.59 to 2.11 vs 1.53 95% CI 1.49 to 1.57). However, prognosis was better in South Asians even among diabetics, older people and those living in areas of the highest social deprivation. Conclusions South Asian ethnicity is associated with higher incidence of coronary disease, but lower mortality once coronary disease is manifest. The dissociation between effects on incidence and prognosis suggests that public health initiatives to reduce inequalities in mortality between South Asian and white populations should focus on primary prevention. This is a CALIBER study with ClinicalTrials.gov Identifier: NCT01163513.

Comparative review of methods for handling drop‐out in longitudinal studies

Longitudinal data analysis is frequently complicated by drop-out. In this paper we consider several methods for dealing with drop-out afflicted data. Along with a general comparison, particular attention is paid to the consequences of model misspecification. The purpose of our approach is two-fold. We first deliberate the form of the drop-out model and compare two alternatives. Furthermore, the extent to which each method is dependent on its core assumptions is assessed through scenarios where one or more such assumptions are compromised. Second, the extent to which we can identify adequacy of model fit is investigated via recently developed diagnostics. These twin targets are pursued via simulation scenarios and application to a schizophrenia trial of over 500 patients with near 50 per cent drop-out.

Joint modelling of time-to-event and multivariate longitudinal outcomes: recent developments and issues

BackgroundAvailable methods for the joint modelling of longitudinal and time-to-event outcomes have typically only allowed for a single longitudinal outcome and a solitary event time. In practice, clinical studies are likely to record multiple longitudinal outcomes. Incorporating all sources of data will improve the predictive capability of any model and lead to more informative inferences for the purpose of medical decision-making.MethodsWe reviewed current methodologies of joint modelling for time-to-event data and multivariate longitudinal data including the distributional and modelling assumptions, the association structures, estimation approaches, software tools for implementation and clinical applications of the methodologies.ResultsWe found that a large number of different models have recently been proposed. Most considered jointly modelling linear mixed models with proportional hazard models, with correlation between multiple longitudinal outcomes accounted for through multivariate normally distributed random effects. So-called current value and random effects parameterisations are commonly used to link the models. Despite developments, software is still lacking, which has translated into limited uptake by medical researchers.ConclusionAlthough, in an era of personalized medicine, the value of multivariate joint modelling has been established, researchers are currently limited in their ability to fit these models routinely. We make a series of recommendations for future research needs.

Measuring the Impact of Research: Lessons from the UK’s Research Excellence Framework 2014

Impactful academic research plays a stellar role in society, pressing to ask the question of how one measures the impact created by different areas of academic research. Measuring the societal, cultural, economic and scientific impact of research is currently the priority of the National Science Foundation, European Commission and several research funding agencies. The recently concluded United Kingdom’s national research quality exercise, the Research Excellence Framework (REF) 2014, which piloted impact assessment as part of the overall evaluation offers a lens to view how impact of research in different disciplines can be measured. Overall research quality was assessed through quality of outputs, ‘impact’ and research environment. We performed two studies using the REF 2014 as a case study. The first study on 363 Impact Case Studies (ICSs) submitted in 5 research areas (UoAs) reveals that, in general, the impact scores were constructed upon a combination of factors i.e. quantity of quartile-one (Q1) publications, quantity and value of grants/income, number of researchers stated in the ICSs, spin-offs created, discoveries/patents and presentation of esteem data, informing researchers/ academics of the factors to consider in order to achieve a better impact score in research impact assessments. However, there were differences among disciplines in terms of the role played by the factors in achieving their overall scores for the ICSs. The outcome of this study is thus a set of impact indicators, and their relationship with the overall score of impact of research in different disciplines as determined in REF2014, which would in the first instance provide some answers to impact measures that would be useful for researchers in different disciplines. The second study extracts the general themes of impact reported by universities by performing a word frequency analysis in all the ICSs submitted in the five chosen research areas, which were substantially varied owing to their fields.

Translational phases of evidence in a prognostic biomarker: a systematic review and meta-analysis of natriuretic peptides and the prognosis of stable coronary disease

Context Translational phases of study are important in evaluating whether a prognostic biomarker is likely to have impact on clinical practice but systematic evaluations of such evidence are lacking. Objective To systematically evaluate the clinical usefulness of the published literature on the association of natriuretic peptides (NP) and prognosis in stable coronary disease. Data sources MEDLINE and EMBASE until the end of July 2009, without restrictions. Study selection Prospective studies measuring NP in people with stable coronary disease who were followed-up for all cause mortality, coronary or cardiovascular events. Data extraction Two independent reviewers categorised studies according to the American Heart Association phase of study, and extracted data according to the study reporting guidelines from the American Heart Association and REMARK. Results Systematic review of 19 studies found 17 which were phase 2, reporting an association between NP and events, two phase 3 studies, statistically examining the incremental prognostic value of NP, but no studies assessing whether NP predicted risk sufficiently to change management (phase 4), improve clinical outcomes (phase 5) or cost effectiveness (phase 6). No study referred to a statistical analytic protocol. Meta-analysis of 14 studies, reporting 18 841 patients and 1655 outcome events, found an RR for events of 3.28 (95% CI 2.45 to 4.38) comparing top versus bottom third of NP. This effect was 26% lower among the five studies which adjusted for a priori confounders (age, sex, renal function and left ventricular function) and 38% lower when adjusting for publication bias (Eggers p=0.001). Conclusion The unbiased strength of association of NP with prognosis in stable coronary disease is unclear, and there is a lack of reports of clinically useful measures of prediction and discrimination or studies relating NP levels to clinical decision making. The available literature is confined to early phases and is of limited clinical usefulness.

Reducing inequalities in care for patients with non-malignant diseases: Insights from a realist evaluation of an integrated palliative care pathway

Background: The need for palliative care is growing internationally with an increasing prevalence of non-malignant diseases. The integrated care pathway was implemented in primary care by multidisciplinary teams from 2009 in a locality in the North East of England. Fourteen general practitioner practices provided data for the study. Aim: To find whether, how, and under what circumstances palliative care registrations are made for patients with non-malignant diseases in primary care. Design: General practitioner practice data were analysed statistically and qualitative data were collected from health care professionals and members of relevant organisations. Findings: A mixed-effects logistic model indicated a significant difference beyond the 0.1% level (p < 0.001) in registrations between the malignant and non-malignant groups in 2011, with an odds ratio of 0.09 (=exp(−2.4266)), indicating that patients in the non-malignant group are around 11 times (1/0.09) less likely to be registered than patients in the malignant group. However, patients with non-malignant diseases were significantly more likely to be registered in 2012 than in 2011 with an odds ratio of 1.46, significant beyond the 1% level. Qualitative analyses indicate that health care professionals find registering patients with non-malignant diseases stressful, yet feel that their confidence in treating this population is increasing. Conclusion: The integrated care pathway began to enable the reduction in inequalities in care by identifying, registering and managing an increasing number of palliative patients with non-malignant diseases. Consensual and inclusive definitions of palliative care were developed in order to legitimise the registration of such patients.

Joint Models of Longitudinal and Time-to-Event Data with More Than One Event Time Outcome: A Review.

Abstract Methodological development and clinical application of joint models of longitudinal and time-to-event outcomes have grown substantially over the past two decades. However, much of this research has concentrated on a single longitudinal outcome and a single event time outcome. In clinical and public health research, patients who are followed up over time may often experience multiple, recurrent, or a succession of clinical events. Models that utilise such multivariate event time outcomes are quite valuable in clinical decision-making. We comprehensively review the literature for implementation of joint models involving more than a single event time per subject. We consider the distributional and modelling assumptions, including the association structure, estimation approaches, software implementations, and clinical applications. Research into this area is proving highly promising, but to-date remains in its infancy.

joineRML: A joint model and software package for time-to-event and multivariate longitudinal outcomes

BackgroundJoint modelling of longitudinal and time-to-event outcomes has received considerable attention over recent years. Commensurate with this has been a rise in statistical software options for fitting these models. However, these tools have generally been limited to a single longitudinal outcome. Here, we describe the classical joint model to the case of multiple longitudinal outcomes, propose a practical algorithm for fitting the models, and demonstrate how to fit the models using a new package for the statistical software platform R, joineRML.ResultsA multivariate linear mixed sub-model is specified for the longitudinal outcomes, and a Cox proportional hazards regression model with time-varying covariates is specified for the event time sub-model. The association between models is captured through a zero-mean multivariate latent Gaussian process. The models are fitted using a Monte Carlo Expectation-Maximisation algorithm, and inferences are based on approximate standard errors from the empirical profile information matrix, which are contrasted to an alternative bootstrap estimation approach. We illustrate the model and software on a real data example for patients with primary biliary cirrhosis with three repeatedly measured biomarkers.ConclusionsAn open-source software package capable of fitting multivariate joint models is available. The underlying algorithm and source code makes use of several methods to increase computational speed.