Fully Bayesian hierarchical modelling in two stages, with application to meta-analysis

Effectiveness of population-based recovery actions for threatened southern mountain caribou

Habitat loss is affecting many species, including the southern mountain caribou (Rangifer tarandus caribou) population in western North America. Over the last half century, this threatened caribou population's range and abundance have dramatically contracted. An integrated population model was used to analyze 51 years (1973-2023) of demographic data from 40 southern mountain caribou subpopulations to assess the effectiveness of population-based recovery actions at increasing population growth. Reducing potential limiting factors on threatened caribou populations offered a rare opportunity to identify the causes of decline and assess methods of recovery. Southern mountain caribou abundance declined by 51% between 1991 and 2023, and 37% of subpopulations were functionally extirpated. Wolf reduction was the only recovery action that consistently increased population growth when applied in isolation, and combinations of wolf reductions with maternal penning or supplemental feeding provided rapid growth but were applied to only four subpopulations. As of 2023, recovery actions have increased the abundance of southern mountain caribou by 52%, compared to a simulation with no interventions. When predation pressure was reduced, rapid population growth was observed, even under contemporary climate change and high levels of habitat loss. Unless predation is reduced, caribou subpopulations will continue to be extirpated well before habitat conservation and restoration can become effective.

Screening for Asymptomatic Osteonecrosis of the Hip in Systemic Lupus Erythematous: A Systematic Review and Meta-Analysis of MRI-Based Prevalence

Objective. This paper aims to estimate asymptomatic hip osteonecrosis prevalence in SLE patients using MRI examination and to determine the prevalence among higher risk subpopulations. Materials and Methods. PubMed, Embase, Cochrane, and SCOPUS were searched from inception to May 9th, 2023. Studies on patients who were clinically diagnosed with systemic lupus erythematosus without reported symptoms attributable to hip osteonecrosis were included. Two independent reviewers extracted data and assessed the risk of bias. Data collected from each study include the study year, the number of hips screened, the number of hips with osteonecrosis, demographics, laboratory data, medications, follow-up time, radiological protocols, and MRI-based osteonecrosis detection and grading criteria. Results. Eleven eligible studies including 503 participants (15-35 years old; 74-100% female) with SLE were identified. Significant risk of bias was determined in one study. The overall prevalence of osteonecrosis of the hip was found to be 14% (184/1006 hip joints, 95% confidence interval: 7-22%, number needed to scan: 7.1). SLE patients who received corticosteroid treatment had a higher prevalence of asymptomatic hip osteonecrosis (18%) compared to non-corticosteroid users (0%, p-value < 0.01). Additionally, meta-regression results revealed that daily corticosteroid dose was associated with increased prevalence of asymptomatic osteonecrosis (0.5%/milligram, p-value < 0.01). Conclusions. The high prevalence of asymptomatic hip osteonecrosis in SLE patients raises concerns about the timeliness of interventions. The limitations of this study include a relatively low number of identified studies; and one study lacked full-text availability.

Solomon versus selective fetoscopic laser photocoagulation for twin-twin transfusion syndrome: A systematic review and meta-analysis

This meta-analysis aims to compare the perinatal outcome of twin-twin transfusion syndrome (TTTS) pregnancies undergoing selective versus vascular equator (Solomon) fetoscopic laser photocoagulation (FLP). We performed a systematic search in PubMed and Web of Science from inception up to 25 July 2021. Studies comparing the Solomon and selective techniques of FLP for treatment of TTTS pregnancies were eligible. Random-effects or fixed-effect models were used to pool standardized mean differences (SMD) and log odds ratio. Seven studies with a total of 1664 TTTS pregnancies (n = 671 undergoing Solomon and n = 993 selective techniques) were included. As compared to the selective FLP, Solomon was associated with a lower risk of recurrent TTTS compared to the selective technique (Log odds ratio [OR]: -1.167; 95% credible interval [CrI]: -2.01, -0.33; p = 0.021; I² : 67%). In addition, Solomon was significantly associated with a higher risk of placental abruption than the selective technique (Log [OR]: 1.44; 95% CrI: 0.45, 2.47; p = 0.012; I² : 0.0%). Furthermore, a trend for the higher risk of preterm premature rupture of membranes was observed among those undergoing Solomon (Log [OR]: 0.581; 95% CrI: -0.43, 1.49; p = 0.131; I² : 17%). As compared to selective FLP, the Solomon technique for TTTS pregnancies is associated with a significantly lower recurrence of TTTS; however, it significantly increases the risk of placental abruption.

Demographic responses of nearly extirpated endangered mountain caribou to recovery actions in Central British Columbia

Recovering endangered species is a difficult and often controversial task that challenges status quo land uses. Southern Mountain caribou are a threatened ecotype of caribou that historically ranged in southwestern Canada and northwestern USA and epitomize the tension between resource extraction, biodiversity conservation, and Indigenous Peoples' treaty rights. Human-induced habitat alteration is considered the ultimate cause of caribou population declines, whereby an increased abundance of primary prey-such as moose and deer-elevates predator populations and creates unsustainable caribou mortality. Here we focus on the Klinse-Za and Quintette subpopulations, part of the endangered Central Group of Southern Mountain caribou in British Columbia. These subpopulations were trending toward immediate extirpation until a collaborative group initiated recovery by implementing two short-term recovery actions. We test the effectiveness of these recovery actions-maternity penning of adult females and their calves, and the reduction of a primary predator, wolves-in increasing vital rates and population growth. Klinse-Za received both recovery actions, whereas Quintette only received wolf reductions, providing an opportunity to test efficacy between recovery actions. Between 1995 and 2021, we followed 162 collared female caribou for 414 animal-years to estimate survival and used aerial counts to estimate population abundance and calf recruitment. We combined these data in an integrated population model to estimate female population growth, total population abundance, and recovery action effectiveness. Results suggest that the subpopulations were declining rapidly (λ = 0.90-0.93) before interventions and would have been functionally extirpated (<10 animals) within 10-15 years. Wolf reduction increased population growth rates by ~0.12 for each subpopulation. Wolf reduction halted the decline of Quintette caribou and allowed them to increase (λ = 1.05), but alone would have only stabilized the Klinse-Za (λ = 1.02). However, maternity penning in the Klinse-Za increased population growth by a further ~0.06, which when combined with wolf reductions, allowed populations to grow (λ = 1.08). Taken together, the recovery actions in these subpopulations increased adult female survival, calf recruitment, and overall population growth, more than doubling abundance. Our results suggest that maternity penning and wolf reductions can be effective at increasing caribou numbers in the short term, while long-term commitments to habitat protection and restoration are made.

Estimation Parameters of Dependence Meta-Analytic Model: New Techniques for the Hierarchical Bayesian Model

Dependence in meta-analytic models can happen due to the same collected data or from the same researchers. The hierarchical Bayesian linear model in a meta-analysis that allows dependence in effect sizes is investigated in this paper. The interested parameters on the hierarchical Bayesian linear dependence (HBLD) model which was developed using the Bayesian techniques will then be estimated. The joint posterior distribution of all parameters for the hierarchical Bayesian linear dependence (HBLD) model is obtained by applying the Gibbs sampling algorithm. Furthermore, in order to measure the robustness of the HBLD model, the sensitivity analysis is conducted using a different prior distribution on the model. This is carried out by applying the Metropolis within the Gibbs algorithm. The simulation study is performed for the estimation of all parameters in the model. The results show that the obtained estimated parameters are close to the true parameters, indicating the consistency of the parameters for the model. The model is also not sensitive because of the changing prior distribution which shows the robustness of the model. A case study, to assess the effects of native-language vocabulary aids on second language reading, is conducted successfully in testing the parameters of the models.

Interoperability of statistical models in pandemic preparedness: principles and reality

We present interoperability as a guiding framework for statistical modelling to assist policy makers asking multiple questions using diverse datasets in the face of an evolving pandemic response. Interoperability provides an important set of principles for future pandemic preparedness, through the joint design and deployment of adaptable systems of statistical models for disease surveillance using probabilistic reasoning. We illustrate this through case studies for inferring and characterising spatial-temporal prevalence and reproduction numbers of SARS-CoV-2 infections in England.

A numerically stable algorithm for integrating Bayesian models using Markov melding

When statistical analyses consider multiple data sources, Markov melding provides a method for combining the source-specific Bayesian models. Markov melding joins together submodels that have a common quantity. One challenge is that the prior for this quantity can be implicit, and its prior density must be estimated. We show that error in this density estimate makes the two-stage Markov chain Monte Carlo sampler employed by Markov melding unstable and unreliable. We propose a robust two-stage algorithm that estimates the required prior marginal self-density ratios using weighted samples, dramatically improving accuracy in the tails of the distribution. The stabilised version of the algorithm is pragmatic and provides reliable inference. We demonstrate our approach using an evidence synthesis for inferring HIV prevalence, and an evidence synthesis of A/H1N1 influenza.

Individual heterogeneity influences the effects of translocation on urban dispersal of an invasive reptile

BACKGROUND

Invasive reptiles pose a serious threat to global biodiversity, but early detection of individuals in an incipient population is often hindered by their cryptic nature, sporadic movements, and variation among individuals. Little is known about the mechanisms that affect the movement of these species, which limits our understanding of their dispersal. Our aim was to determine whether translocation or small-scale landscape features affect movement patterns of brown treesnakes (Boiga irregularis), a destructive invasive predator on the island of Guam.

METHODS

We conducted a field experiment to compare the movements of resident (control) snakes to those of snakes translocated from forests and urban areas into new urban habitats. We developed a Bayesian hierarchical model to analyze snake movement mechanisms and account for attributes unique to invasive reptiles by incorporating multiple behavioral states and individual heterogeneity in movement parameters.

RESULTS

We did not observe strong differences in mechanistic movement parameters (turning angle or step length) among experimental treatment groups. We found some evidence that translocated snakes from both forests and urban areas made longer movements than resident snakes, but variation among individuals within treatment groups weakened this effect. Snakes translocated from forests moved more frequently from pavement than those translocated from urban areas. Snakes translocated from urban areas moved less frequently from buildings than resident snakes. Resident snakes had high individual heterogeneity in movement probability.

CONCLUSIONS

Our approach to modeling movement improved our understanding of invasive reptile dispersal by allowing us to examine the mechanisms that influence their movement. We also demonstrated the importance of accounting for individual heterogeneity in population-level analyses, especially when management goals involve eradication of an invasive species.

Combining chains of Bayesian models with Markov melding

A challenge for practitioners of Bayesian inference is specifying a model that incorporates multiple relevant, heterogeneous data sets. It may be easier to instead specify distinct submodels for each source of data, then join the submodels together. We consider chains of submodels, where submodels directly relate to their neighbours via common quantities which may be parameters or deterministic functions thereof. We propose chained Markov melding, an extension of Markov melding, a generic method to combine chains of submodels into a joint model. One challenge we address is appropriately capturing the prior dependence between common quantities within a submodel, whilst also reconciling differences in priors for the same common quantity between two adjacent submodels. Estimating the posterior of the resulting overall joint model is also challenging, so we describe a sampler that uses the chain structure to incorporate information contained in the submodels in multiple stages, possibly in parallel. We demonstrate our methodology using two examples. The first example considers an ecological integrated population model, where multiple data sets are required to accurately estimate population immigration and reproduction rates. We also consider a joint longitudinal and time-to-event model with uncertain, submodel-derived event times. Chained Markov melding is a conceptually appealing approach to integrating submodels in these settings.

Conceptual and methodological advances in habitat-selection modeling: guidelines for ecology and evolution

Habitat selection is a fundamental animal behavior that shapes a wide range of ecological processes, including animal movement, nutrient transfer, trophic dynamics and population distribution. Although habitat selection has been a focus of ecological studies for decades, technological, conceptual and methodological advances over the last 20 yr have led to a surge in studies addressing this process. Despite the substantial literature focused on quantifying the habitat-selection patterns of animals, there is a marked lack of guidance on best analytical practices. The conceptual foundations of the most commonly applied modeling frameworks can be confusing even to those well versed in their application. Furthermore, there has yet to be a synthesis of the advances made over the last 20 yr. Therefore, there is a need for both synthesis of the current state of knowledge on habitat selection, and guidance for those seeking to study this process. Here, we provide an approachable overview and synthesis of the literature on habitat-selection analyses (HSAs) conducted using selection functions, which are by far the most applied modeling framework for understanding the habitat-selection process. This review is purposefully non-technical and focused on understanding without heavy mathematical and statistical notation, which can confuse many practitioners. We offer an overview and history of HSAs, describing the tortuous conceptual path to our current understanding. Through this overview, we also aim to address the areas of greatest confusion in the literature. We synthesize the literature outlining the most exciting conceptual advances in the field of habitat-selection modeling, discussing the substantial ecological and evolutionary inference that can be made using contemporary techniques. We aim for this paper to provide clarity for those navigating the complex literature on HSAs while acting as a reference and best practices guide for practitioners.

A continued learning approach for model-informed precision dosing: updating models in clinical practice

Model-informed precision dosing (MIPD) is a quantitative dosing framework that combines prior knowledge on the drug-disease-patient system with patient data from therapeutic drug/biomarker monitoring (TDM) to support individualized dosing in ongoing treatment. Structural models and prior parameter distributions used in MIPD approaches typically build on prior clinical trials that involve only a limited number of patients selected according to some exclusion/inclusion criteria. Compared to the prior clinical trial population, the patient population in clinical practice can be expected to include also altered behavior and/or increased interindividual variability, the extent of which, however, is typically unknown. Here, we address the question of how to adapt and refine models on the level of the model parameters to better reflect this real-world diversity. We propose an approach for continued learning across patients during MIPD using a sequential hierarchical Bayesian framework. The approach builds on two stages to separate the update of the individual patient parameters from updating the population parameters. Consequently, it enables continued learning across hospitals or study centers, since only summary patient data (on the level of model parameters) need to be shared, but no individual TDM data. We illustrate this continued learning approach with neutrophil-guided dosing of paclitaxel. The present study constitutes an important step towards building confidence in MIPD and eventually establishing MIPD increasingly in everyday therapeutic use.

Mechanistic movement models identify continuously updated autumn migration cues in Arctic caribou

BACKGROUND

Migrations in temperate systems typically have two migratory phases, spring and autumn, and many migratory ungulates track the pulse of spring vegetation growth during a synchronized spring migration. In contrast, autumn migrations are generally less synchronous and the cues driving them remain understudied. Our goal was to identify the cues that migrants use in deciding when to initiate migration and how this is updated while en route.

METHODS

We analyzed autumn migrations of Arctic barren-ground caribou (Rangifer tarandus) as a series of persistent and directional movements and assessed the influence of a suite of environmental factors. We fitted a dynamic-parameter movement model at the individual-level and estimated annual population-level parameters for weather covariates on 389 individual-seasons across 9 years.

RESULTS

Our results revealed strong, consistent effects of decreasing temperature and increasing snow depth on migratory movements, indicating that caribou continuously update their migratory decision based on dynamic environmental conditions. This suggests that individuals pace migration along gradients of these environmental variables. Whereas temperature and snow appeared to be the most consistent cues for migration, we also found interannual variability in the effect of wind, NDVI, and barometric pressure. The dispersed distribution of individuals in autumn resulted in diverse environmental conditions experienced by individual caribou and thus pronounced variability in migratory patterns.

CONCLUSIONS

By analyzing autumn migration as a continuous process across the entire migration period, we found that caribou migration was largely related to temperature and snow conditions experienced throughout the journey. This mechanism of pacing autumn migration based on indicators of the approaching winter is analogous to the more widely researched mechanism of spring migration, when many migrants pace migration with a resource wave. Such a similarity in mechanisms highlights the different environmental stimuli to which migrants have adapted their movements throughout their annual cycle. These insights have implications for how long-distance migratory patterns may change as the Arctic climate continues to warm.

Particulate Matter and Premature Mortality: A Bayesian Meta-Analysis

BACKGROUND

We present a systematic review of studies assessing the association between ambient particulate matter (PM) and premature mortality and the results of a Bayesian hierarchical meta-analysis while accounting for population differences of the included studies.

METHODS

The review protocol was registered in the PROSPERO systematic review registry. Medline, CINAHL and Global Health databases were systematically searched. Bayesian hierarchical meta-analysis was conducted using a non-informative prior to assess whether the regression coefficients differed across observations due to the heterogeneity among studies.

RESULTS

We identified 3248 records for title and abstract review, of which 309 underwent full text screening. Thirty-six studies were included, based on the inclusion criteria. Most of the studies were from China (n = 14), India (n = 6) and the USA (n = 3). PM_2.5 was the most frequently reported pollutant. PM was estimated using modelling techniques (22 studies), satellite-based measures (four studies) and direct measurements (ten studies). Mortality data were sourced from country-specific mortality statistics for 17 studies, Global Burden of Disease data for 16 studies, WHO data for two studies and life tables for one study. Sixteen studies were included in the Bayesian hierarchical meta-analysis. The meta-analysis revealed that the annual estimate of premature mortality attributed to PM_2.5 was 253 per 1,000,000 population (95% CI: 90, 643) and 587 per 1,000,000 population (95% CI: 1, 39,746) for PM₁₀.

CONCLUSION

253 premature deaths per million population are associated with exposure to ambient PM_2.5. We observed an unstable estimate for PM₁₀, most likely due to heterogeneity among the studies. Future research efforts should focus on the effects of ambient PM₁₀ and premature mortality, as well as include populations outside Asia. Key messages: Ambient PM_2.5 is associated with premature mortality. Given that rapid urbanization may increase this burden in the coming decades, our study highlights the urgency of implementing air pollution mitigation strategies to reduce the risk to population and planetary health.

On weakly informative prior distributions for the heterogeneity parameter in Bayesian random-effects meta-analysis

The normal-normal hierarchical model (NNHM) constitutes a simple and widely used framework for meta-analysis. In the common case of only few studies contributing to the meta-analysis, standard approaches to inference tend to perform poorly, and Bayesian meta-analysis has been suggested as a potential solution. The Bayesian approach, however, requires the sensible specification of prior distributions. While noninformative priors are commonly used for the overall mean effect, the use of weakly informative priors has been suggested for the heterogeneity parameter, in particular in the setting of (very) few studies. To date, however, a consensus on how to generally specify a weakly informative heterogeneity prior is lacking. Here we investigate the problem more closely and provide some guidance on prior specification.

A Bayesian approach to assessing small-study effects in meta-analysis of a binary outcome with controlled false positive rate

Publication bias threatens meta-analysis validity. It is often assessed via the funnel plot; an asymmetric plot implies small-study effects, and publication bias is one cause of the asymmetry. Egger's regression test is a widely used tool to quantitatively assess such asymmetry. It examines the association between the observed effect sizes and their sample SEs; a strong association indicates small-study effects. However, its false positive rates may be inflated if such an association intrinsically exists even if no small-study effects appear, particularly in meta-analyses of odds ratios (ORs). Various alternatives are available to address this problem. They usually replace Egger's regression predictor or response with different measures; consequently, they are powerful only in specific cases. We propose a Bayesian approach to assessing small-study effects in meta-analyses of ORs. It controls false positive rates by using latent "true" SEs, rather than sample SEs, in the Egger-type regression to avoid the intrinsic association between ORs and their SEs. Although "true" SEs are unknown in practice, they can be modeled under the Bayesian framework. We use simulated and real data to compare various methods. When ORs are away from 1, the proposed method may have high powers with controlled false positive rates, while Egger's test has seriously inflated false positive rates; nevertheless, in other situations, some other methods may be superior. In general, the proposed method may serve as an alternative to rule out potential confounding effects caused by the intrinsic association between ORs and their SEs in the assessment of small-study effects.

Association of microRNA biosynthesis genes XPO5 and RAN polymorphisms with cancer susceptibility: Bayesian hierarchical meta-analysis

XPO5/RAN-GTP complex mediates the nuclear transport of pre-miRNAs in the miRNA processing system, its altered expression is indicated to be correlated with cancer risk. Several studies have inspected the association between XPO5 or RAN polymorphisms and the risk of various cancers, but the findings remain controversial. A Bayesian hierarchical meta-analysis was carried out to review and analyze the effect of XPO5 and RAN polymorphisms on cancer risk. The association was estimated by calculating the logarithm of odds ratio (Log OR) and 95% credible interval (95% CrI). The expression quantitative trait loci (eQTL) analysis was used for in silico functional validation of the identified significant susceptibility loci. Consequently, 38 case-control studies (from 27 citations) with 27,459 cancer cases and 25,151controls were included in the meta-analysis of the five most prevalent SNPs (rs11077 A/C, rs2257082 G/A, rs3803012 A/G, rs14035 C/T, rs3809142 C/T). In the XPO5 gene rs11077 SNP, the minor C allele significantly increased the risk of cancer (Log OR = 0.120, 95% CrI = 0.013, 0.241), and a strong association between rs11077 SNP and cancer risk was also found in the dominant model (CC + AC vs. AA: Log OR = 0.132, 95% CrI = 0.009, 0.275). In addition, the minor GG genotype allele of the RAN gene rs3803012 SNP significantly increased the cancer risk (Log OR = 0.707, 95% CrI = 0.059, 1.385). Statistically significant associations between rs3803012 SNP and cancer risk were also observed in the recessive model (GG vs. AG + AA: Log OR = 0.708, 95% CrI = 0.059, 1.359). Furthermore, the eQTL analysis revealed that rs11077 SNP was significantly correlated with XPO5 mRNA expression, which provided additional biological basis for the observed positive association. Our results suggest that XPO5 rs11077 may be a possible functional susceptibility locus for cancer risk.

Synergistic effects of climate and land-use change influence broad-scale avian population declines

Climate and land-use changes are expected to be the primary drivers of future global biodiversity loss. Although theory suggests that these factors impact species synergistically, past studies have either focused on only one in isolation or have substituted space for time, which often results in confounding between drivers. Tests of synergistic effects require congruent time series on animal populations, climate change and land-use change replicated across landscapes that span the gradient of correlations between the drivers of change. Using a unique time series of high-resolution climate (measured as temperature and precipitation) and land-use change (measured as forest change) data, we show that these drivers of global change act synergistically to influence forest bird population declines over 29 years in the Pacific Northwest of the United States. Nearly half of the species examined had declined over this time. Populations declined most in response to loss of early seral and mature forest, with responses to loss of early seral forest amplified in landscapes that had warmed over time. In addition, birds declined more in response to loss of mature forest in areas that had dried over time. Climate change did not appear to impact populations in landscapes with limited habitat loss, except when those landscapes were initially warmer than the average landscape. Our results provide some of the first empirical evidence of synergistic effects of climate and land-use change on animal population dynamics, suggesting accelerated loss of biodiversity in areas under pressure from multiple global change drivers. Furthermore, our findings suggest strong spatial variability in the impacts of climate change and highlight the need for future studies to evaluate multiple drivers simultaneously to avoid potential misattribution of effects.

Are genetic variations in IL-21-IL-23R-IL-17A cytokine axis involved in a pathogenic pathway of rheumatoid arthritis? Bayesian hierarchical meta-analysis

Inflammatory cytokines have been established to be involved in the pathogenesis of rheumatoid arthritis (RA). The genetic polymorphisms in the interleukin (IL) 23 receptor (IL23R), IL21, and IL17 have been associated with RA risk. However, there is no conclusive understanding of the genes encoding the immunoinflammatory IL-21-IL-23R-IL-17A pathway in RA aetiopathogenesis. This meta-analysis was conducted to attain this goal. A comprehensive literature search was conducted in Scopus and PubMed to look for the relevant case-control studies up until 2018. A Bayesian hierarchical meta-analysis was carried out to assess the association between the polymorphisms and the risk of RA. The association was estimated by calculating the logarithm of odds ratio (Log OR) and 95% credible interval (95% CI). In this meta-analysis, 37 case-control studies comprising 23,506 RA patients and 25,984 healthy individuals were found for analyzing the IL23R, IL21, and IL1A gene polymorphism and risk of RA. In the IL23R gene rs1343151 SNP, the minor A allele significantly increased the risk of RA (Log OR = 0.085, 95% CI = 0.008, 0.156). Moreover, the minor AA genotype was significantly associated with increased RA risk (Log OR = 0.176, 95% CI = 0.028, 0.321). In addition, the C allele of the IL23R gene rs2201841 SNP significantly decreased the disease risk (Log OR = -0.544, 95% CI = -1.0, -0.065). Since Bayesian meta-analysis is a powerful strategy to pool the data, it can be mentioned that genetic polymorphisms of IL23R, but not IL21 and IL17A, are involved in susceptibility to RA.

Does depression after stroke negatively influence physical disability? A systematic review and meta-analysis of longitudinal studies

BACKGROUND

Depression after stroke is common and has been proposed to negatively affect disability by preventing optimal physical rehabilitation and recovery. However, the nature of this influence remains poorly understood. Here, we synthesise longitudinal studies to examine the hypotheses that depression after stroke (i) hampers physical rehabilitation, (ii) prevents functional improvement during recovery, and (iii) is associated with poor functional outcomes.

METHODS

A systematic literature search was conducted using the databases PubMed and Web of Science. A total of 5672 studies were screened; 28 met criteria for inclusion. The quality of included studies was assessed using the Newcastle-Ottawa Scale.

RESULTS

Individual studies showed no consistent effects of depression post-stroke on (i) the effectiveness of physical rehabilitation and (ii) functional improvements during recovery. In contrast, random-effects models revealed that (iii) depression after stroke was associated with an increased risk for poor long-term disability (OR: 2.16, 95% CI 1.70-2.77). Overall, the quality of studies was moderate and there was evidence for publication bias.

LIMITATIONS

The number of included studies was small. There was considerable methodological heterogeneity between studies, prohibiting meta-analyses for all effects of interest. Few studies examined the influence of antidepressants.

CONCLUSIONS

Depressed stroke patients are generally more disabled. However, depressed mood might not restrict improvements in physical disability during rehabilitation and recovery, although it seems to be linked to a delayed increase in the risk of poor functional outcome. High-quality evidence from longitudinal studies is needed to clarify the precise mechanisms and temporal dynamics underlying these associations.

Joining and splitting models with Markov melding

Analysing multiple evidence sources is often feasible only via a modular approach, with separate submodels specified for smaller components of the available evidence. Here we introduce a generic framework that enables fully Bayesian analysis in this setting. We propose a generic method for forming a suitable joint model when joining submodels, and a convenient computational algorithm for fitting this joint model in stages, rather than as a single, monolithic model. The approach also enables splitting of large joint models into smaller submodels, allowing inference for the original joint model to be conducted via our multi-stage algorithm. We motivate and demonstrate our approach through two examples: joining components of an evidence synthesis of A/H1N1 influenza, and splitting a large ecology model.

Pneumococcal Disease and the Effectiveness of the PPV23 Vaccine in Adults: A Two-Stage Bayesian Meta-Analysis of Observational and RCT Reports

The efficacy of PPV-23 vaccine on outcomes of pneumococcal disease in adults still remains controversial due mainly to the lack of consistency between the results obtained from observational studies(OSs) and those obtained from randomized controlled trials(RCTs). As a consequence, the complexity in the structure of evidence available, in turn, generates a challenge for combining disparate pieces of evidence quantitatively. In this regard, we used a hierarchical Bayesian inference-based evidence synthesis of RCTs and observational data using a two-stage approach (in addition to a traditional random-effects meta-analysis) to examine the effectiveness of PPV-23 in adults. To this end, 21 studies were included involving 826109 adult participants. By a two-stage Bayesian meta-analysis, which was directly used for combining studies of different designs, the overall log OR (95% credible interval) for IPDs was -0.1048 (-0.3920,-0.0250), indicating a significant protective effect of the vaccination against IPDs. No significant effect of PPV-23 was found on all-cause pneumonia, pneumococcal pneumonia, and death from pneumonia, which confirmed the results obtained by a traditional method followed by stratified and sensitivity analyses. The estimated overall log OR (95% credible interval) was -0.0002 (-0.0241,0.0142), -0.0002 (-0.0110,0.0122), and -6.3912 × 10^-5 (-0.0219,0.0131), respectively. The PPV-23 vaccine might be effective in preventing the most severe invasive forms of pneumococcal diseases, but not effective in preventing other clinical outcomes, in the adult population of 18 years and older.

Meta-analysis of Gaussian individual patient data: Two-stage or not two-stage?

Quantitative evidence synthesis through meta-analysis is central to evidence-based medicine. For well-documented reasons, the meta-analysis of individual patient data is held in higher regard than aggregate data. With access to individual patient data, the analysis is not restricted to a "two-stage" approach (combining estimates and standard errors) but can estimate parameters of interest by fitting a single model to all of the data, a so-called "one-stage" analysis. There has been debate about the merits of one- and two-stage analysis. Arguments for one-stage analysis have typically noted that a wider range of models can be fitted and overall estimates may be more precise. The two-stage side has emphasised that the models that can be fitted in two stages are sufficient to answer the relevant questions, with less scope for mistakes because there are fewer modelling choices to be made in the two-stage approach. For Gaussian data, we consider the statistical arguments for flexibility and precision in small-sample settings. Regarding flexibility, several of the models that can be fitted only in one stage may not be of serious interest to most meta-analysis practitioners. Regarding precision, we consider fixed- and random-effects meta-analysis and see that, for a model making certain assumptions, the number of stages used to fit this model is irrelevant; the precision will be approximately equal. Meta-analysts should choose modelling assumptions carefully. Sometimes relevant models can only be fitted in one stage. Otherwise, meta-analysts are free to use whichever procedure is most convenient to fit the identified model.

A model for meta-analysis of correlated binary outcomes: The case of split-body interventions

In several areas of clinical research, it is common for trials to assign different sites of the participants’ bodies to different interventions. For example, a randomized controlled trial comparing surgical techniques for correcting myopia may randomize each eye of a participant to a different operation. Under such bilateral (‘split-body’) interventions, the observations from each participant are correlated. It is challenging to account for these correlations at the meta-analysis level, especially when the outcome is rare. Here, we present a meta-analysis model based on the bivariate binomial distribution. Our model can synthesize studies on patients who received one intervention at one body site, patients who received two interventions at different sites or a mixture of these two groups. The model can analyse studies with zero events in one or both treatment arms and can handle the case of incomplete data reporting. We use simulations to assess the performance of our model and to compare it with the bivariate beta-binomial model. In the case of bilateral interventions, our model performed well and outperformed the bivariate beta-binomial model in all scenarios explored. We illustrate our methods using two previously published meta-analyses from the fields of orthopaedics and ophthalmology. We conclude that our model constitutes a useful new tool for the meta-analysis of binary outcomes in the presence of split-body interventions.

Comparative Effectiveness of Adding Alogliptin to Metformin Plus Sulfonylurea with Other DPP-4 Inhibitors in Type 2 Diabetes: A Systematic Review and Network Meta-Analysis

INTRODUCTION

Alogliptin is an oral antihyperglycemic agent that is a selective inhibitor of the enzyme dipeptidyl peptidase-4 (DPP-4), approved for the treatment of type 2 diabetes mellitus (T2DM). There currently exists no comparative data to support the use of alogliptin in combination with metformin (met) and sulfonylurea (SU). A decision-focused network meta-analysis (NMA) was performed to compare the relative efficacy and safety of alogliptin 25 mg once daily to other DPP-4 inhibitors as part of a triple therapy regimen for patients inadequately controlled on metformin and SU dual therapy.

METHODS

A systematic literature review was conducted to identify published papers of randomized controlled trials (RCTs) that compared alogliptin with other DPP-4 inhibitors (linagliptin, saxagliptin, sitagliptin, and vildagliptin) at their Summary of Product Characteristics (SmPC) recommended daily doses, added on to metformin and SU. Comprehensive comparative analysis involving frequentist meta-analysis and Bayesian NMA compared alogliptin to each DPP-4 inhibitor separately and collectively as a group. Quasi-random effect models were introduced when random effect models could not be estimated.

RESULTS

The review identified 2186 articles, and 94 full-text articles were assessed for eligibility. Eight RCTs contained appropriate data for inclusion in the NMA. All analyses over all trial population sets produced very similar results, and show that alogliptin 25 mg is as least as effective (as measured by change in HbA1c from baseline, but supported by other outcome measures: change in body weight and FPG from baseline) and safe (as measured by incidence of hypoglycemia and adverse events leading to study discontinuation) as all the other DPP-4 inhibitors in triple therapy.

CONCLUSION

This decision-focused systematic review and NMA demonstrated alogliptin 25 mg daily to have similar efficacy and safety compared to other DPP-4 inhibitors, for the treatment of T2DM in adults inadequately controlled on metformin and SU. (Funded by Takeda Development Centre Americas; EXAMINE ClinicalTrials.gov number, NCT00968708).

Bayesian hierarchical modeling of longitudinal glaucomatous visual fields using a two-stage approach

The Bayesian approach has become increasingly popular because it allows to fit quite complex models to data via Markov chain Monte Carlo sampling. However, it is also recognized nowadays that Markov chain Monte Carlo sampling can become computationally prohibitive when applied to a large data set. We encountered serious computational difficulties when fitting an hierarchical model to longitudinal glaucoma data of patients who participate in an ongoing Dutch study. To overcome this problem, we applied and extended a recently proposed two-stage approach to model these data. Glaucoma is one of the leading causes of blindness in the world. In order to detect deterioration at an early stage, a model for predicting visual fields (VFs) in time is needed. Hence, the true underlying VF progression can be determined, and treatment strategies can then be optimized to prevent further VF loss. Because we were unable to fit these data with the classical one-stage approach upon which the current popular Bayesian software is based, we made use of the two-stage Bayesian approach. The considered hierarchical longitudinal model involves estimating a large number of random effects and deals with censoring and high measurement variability. In addition, we extended the approach with tools for model evaluation. Copyright © 2017 John Wiley & Sons, Ltd.

Artemether-Lumefantrine and Dihydroartemisinin-Piperaquine Exert Inverse Selective Pressure on Plasmodium Falciparum Drug Sensitivity-Associated Haplotypes in Uganda

BACKGROUND

Altered sensitivity to multiple antimalarial drugs is mediated by polymorphisms in pfmdr1, which encodes the Plasmodium falciparum multidrug resistance transporter. In Africa the N86Y and D1246Y polymorphisms have been shown to be selected by treatment, with artemether-lumefantrine (AL) and dihydroartemisinin-piperaquine (DP) selecting for wild-type and mutant alleles, respectively. However, there has been little study of pfmdr1 haplotypes, in part because haplotype analyses are complicated by multiclonal infections.

METHODS

We fit a haplotype frequency estimation model, which accounts for multiclonal infections, to the polymorphic pfmdr1 N86Y, Y184F, and D1246Y alleles in samples from a longitudinal trial comparing AL and DP to treat uncomplicated P falciparum malaria in Tororo, Uganda from 2007 to 2012. We regressed estimates onto covariates of trial arm and selective drug pressure.

RESULTS

Yearly trends showed increasing frequency estimates for haplotypes with wild type pfmdr1 N86 and D1246 alleles and decreasing frequency estimates for haplotypes with the mutant pfmdr1 86Y allele. Considering days since prior therapy, we saw evidence suggestive of selection by AL for haplotypes with N86 combined with 184F, D1246, or both, and against all haplotypes with 86Y, and evidence suggestive of selection by DP for 86Y only when combined with Y184 and 1246Y (haplotype YYY) and against haplotypes NFD and NYY.

CONCLUSIONS

Based on our model, AL selected several haplotypes containing N86, whereas DP selection was haplotype specific, demonstrating the importance of haplotype analyses. Inverse selective pressure of AL and DP on the complementary haplotypes NFD and YYY suggests that rotating artemisinin-based antimalarial combination regimens may be the best treatment option to prevent resistance selection.

Rapid model exploration for complex hierarchical data: application to pharmacokinetics of insulin aspart

We consider situations, which are common in medical statistics, where we have a number of sets of response data, from different individuals, say, potentially under different conditions. A parametric model is defined for each set of data, giving rise to a set of random effects. Our goal here is to efficiently explore a range of possible ‘population’ models for the random effects, to select the most appropriate model. The range of possible models is potentially vast, because the random effects may depend on observed covariates, and there may be multiple credible ways of partitioning their variability. Here, we consider pharmacokinetic (PK) data on insulin aspart, a fast acting insulin analogue used in the treatment of diabetes. PK models are typically nonlinear (in their parameters), often complex and sometimes only available as a set of differential equations, with no closed‐form solution. Fitting such a model for just a single individual can be a challenging task. Fitting a joint model for all individuals can be even harder, even without the complication of an overarching model selection objective. We describe a two‐stage approach that decouples the population model for the random effects from the PK model applied to the response data but nevertheless fits the full, joint, hierarchical model, accounting fully for uncertainty. This allows us to repeatedly reuse results from a single analysis of the response data to explore various population models for the random effects. This greatly expedites not only model exploration but also cross‐validation for the purposes of model criticism. © 2015 The Authors. Statistics in Medicine published by John Wiley & Sons Ltd.

Four key challenges in infectious disease modelling using data from multiple sources

Public health-related decision-making on policies aimed at controlling epidemics is increasingly evidence-based, exploiting multiple sources of data. Policy makers rely on complex models that are required to be robust, realistically approximating epidemics and consistent with all relevant data. Meeting these requirements in a statistically rigorous and defendable manner poses a number of challenging problems. How to weight evidence from different datasets and handle dependence between them, efficiently estimate and critically assess complex models are key challenges that we expound in this paper, using examples from influenza modelling.