A Bayesian hierarchical model for prediction of latent health states from multiple data sources with application to active surveillance of prostate cancer

Aggressive Prostate Cancer in Patients Treated with Active Surveillance

Active surveillance has emerged as a promising approach for managing low-risk and favorable intermediate-risk prostate cancer (PC), with the aim of minimizing overtreatment and maintaining the quality of life. However, concerns remain about identifying "aggressive prostate cancer" within the active surveillance cohort, which refers to cancers with a higher potential for progression. Previous studies are predictors of aggressive PC during active surveillance. To address this, a personalized risk-based follow-up approach that integrates clinical data, biomarkers, and genetic factors using risk calculators was proposed. This approach enables an efficient risk assessment and the early detection of disease progression, minimizes unnecessary interventions, and improves patient management and outcomes. As active surveillance indications expand, the importance of identifying aggressive PC through a personalized risk-based follow-up is expected to increase.

Active Surveillance for Prostate Cancer: Past, Current, and Future Trends

In response to the rising incidence of indolent, low-risk prostate cancer (PCa) due to increased prostate-specific antigen (PSA) screening in the 1990s, active surveillance (AS) emerged as a treatment modality to combat overtreatment by delaying or avoiding unnecessary definitive treatment and its associated morbidity. AS consists of regular monitoring of PSA levels, digital rectal exams, medical imaging, and prostate biopsies, so that definitive treatment is only offered when deemed necessary. This paper provides a narrative review of the evolution of AS since its inception and an overview of its current landscape and challenges. Although AS was initially only performed in a study setting, numerous studies have provided evidence for the safety and efficacy of AS which has led guidelines to recommend it as a treatment option for patients with low-risk PCa. For intermediate-risk disease, AS appears to be a viable option for those with favourable clinical characteristics. Over the years, the inclusion criteria, follow-up schedule and triggers for definitive treatment have evolved based on the results of various large AS cohorts. Given the burdensome nature of repeat biopsies, risk-based dynamic monitoring may further reduce overtreatment by avoiding repeat biopsies in selected patients.

Joint models for dynamic prediction in localised prostate cancer: a literature review

BACKGROUND

Prostate cancer is a very prevalent disease in men. Patients are monitored regularly during and after treatment with repeated assessment of prostate-specific antigen (PSA) levels. Prognosis of localised prostate cancer is generally good after treatment, and the risk of having a recurrence is usually estimated based on factors measured at diagnosis. Incorporating PSA measurements over time in a dynamic prediction joint model enables updates of patients' risk as new information becomes available. We review joint model strategies that have been applied to model time-dependent PSA trajectories to predict time-to-event outcomes in localised prostate cancer.

METHODS

We identify articles that developed joint models for prediction of localised prostate cancer recurrence over the last two decades. We report, compare, and summarise the methodological approaches and applications that use joint modelling accounting for two processes: the longitudinal model (PSA), and the time-to-event process (clinical failure). The methods explored differ in how they specify the association between these two processes.

RESULTS

Twelve relevant articles were identified. A range of methodological frameworks were found, and we describe in detail shared-parameter joint models (9 of 12, 75%) and joint latent class models (3 of 12, 25%). Within each framework, these articles presented model development, estimation of dynamic predictions and model validations.

CONCLUSIONS

Each framework has its unique principles with corresponding advantages and differing interpretations. Regardless of the framework used, dynamic prediction models enable real-time prediction of individual patient prognosis. They utilise all available longitudinal information, in addition to baseline prognostic risk factors, and are superior to traditional baseline-only prediction models.

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.

Informative presence and observation in routine health data: A review of methodology for clinical risk prediction

Objective

Informative presence (IP) is the phenomenon whereby the presence or absence of patient data is potentially informative with respect to their health condition, with informative observation (IO) being the longitudinal equivalent. These phenomena predominantly exist within routinely collected healthcare data, in which data collection is driven by the clinical requirements of patients and clinicians. The extent to which IP and IO are considered when using such data to develop clinical prediction models (CPMs) is unknown, as is the existing methodology aiming at handling these issues. This review aims to synthesize such existing methodology, thereby helping identify an agenda for future methodological work.

Materials and Methods

A systematic literature search was conducted by 2 independent reviewers using prespecified keywords.

Results

Thirty-six articles were included. We categorized the methods presented within as derived predictors (including some representation of the measurement process as a predictor in the model), modeling under IP, and latent structures. Including missing indicators or summary measures as predictors is the most commonly presented approach amongst the included studies (24 of 36 articles).

Discussion

This is the first review to collate the literature in this area under a prediction framework. A considerable body relevant of literature exists, and we present ways in which the described methods could be developed further. Guidance is required for specifying the conditions under which each method should be used to enable applied prediction modelers to use these methods.

Conclusions

A growing recognition of IP and IO exists within the literature, and methodology is increasingly becoming available to leverage these phenomena for prediction purposes. IP and IO should be approached differently in a prediction context than when the primary goal is explanation. The work included in this review has demonstrated theoretical and empirical benefits of incorporating IP and IO, and therefore we recommend that applied health researchers consider incorporating these methods in their work.

Machine learning for predictive data analytics in medicine: A review illustrated by cardiovascular and nuclear medicine examples

The evidence-based medicine allows the physician to evaluate the risk-benefit ratio of a treatment through setting and data. Risk-based choices can be done by the doctor using different information. With the emergence of new technologies, a large amount of data is recorded offering interesting perspectives with machine learning for predictive data analytics. Machine learning is an ensemble of methods that process data to model a learning problem. Supervised machine learning algorithms consist in using annotated data to construct the model. This category allows to solve prediction data analytics problems. In this paper, we detail the use of supervised machine learning algorithms for predictive data analytics problems in medicine. In the medical field, data can be split into two categories: medical images and other data. For brevity, our review deals with any kind of medical data excluding images. In this article, we offer a discussion around four supervised machine learning approaches: information-based, similarity-based, probability-based and error-based approaches. Each method is illustrated with detailed cardiovascular and nuclear medicine examples. Our review shows that model ensemble (ME) and support vector machine (SVM) methods are the most popular. SVM, ME and artificial neural networks often lead to better results than those given by other algorithms. In the coming years, more studies, more data, more tools and more methods will, for sure, be proposed.

Comparison of biopsy under-sampling and annual progression using hidden markov models to learn from prostate cancer active surveillance studies

This study aimed to estimate the rates of biopsy undersampling and progression for four prostate cancer (PCa) active surveillance (AS) cohorts within the Movember Foundation's Global Action Plan Prostate Cancer Active Surveillance (GAP3) consortium. We used a hidden Markov model (HMM) to estimate factors that define PCa dynamics for men on AS including biopsy under-sampling and progression that are implied by longitudinal data in four large cohorts included in the GAP3 database. The HMM was subsequently used as the basis for a simulation model to evaluate the biopsy strategies previously proposed for each of these cohorts. For the four AS cohorts, the estimated annual progression rate was between 6%-13%. The estimated probability of a biopsy successfully sampling undiagnosed non-favorable risk cancer (biopsy sensitivity) was between 71% and 80%. In the simulation study of patients diagnosed with favorable risk cancer at age 50, the mean number of biopsies performed before age 75 was between 4.11 and 12.60, depending on the biopsy strategy. The mean delay time to detection of non-favorable risk cancer was between 0.38 and 2.17 years. Biopsy undersampling and progression varied considerably across study cohorts. There was no single best biopsy protocol that is optimal for all cohorts, because of the variation in biopsy under-sampling error and annual progression rates across cohorts. All strategies demonstrated diminishing benefits from additional biopsies.

The Movember Prostate Cancer Landscape Analysis: an assessment of unmet research needs

Prostate cancer is a heterogeneous cancer with widely varying levels of morbidity and mortality. Approaches to prostate cancer screening, diagnosis, surveillance, treatment and management differ around the world. To identify the highest priority research needs across the prostate cancer biomedical research domain, Movember conducted a landscape analysis with the aim of maximizing the effect of future research investment through global collaborative efforts and partnerships. A global Landscape Analysis Committee (LAC) was established to act as an independent group of experts across urology, medical oncology, radiation oncology, radiology, pathology, translational research, health economics and patient advocacy. Men with prostate cancer and thought leaders from a variety of disciplines provided a range of key insights through a range of interviews. Insights were prioritized against predetermined criteria to understand the areas of greatest unmet need. From these efforts, 17 research needs in prostate cancer were agreed on and prioritized, and 3 received the maximum prioritization score by the LAC: first, to establish more sensitive and specific tests to improve disease screening and diagnosis; second, to develop indicators to better stratify low-risk prostate cancer for determining which men should go on active surveillance; and third, to integrate companion diagnostics into randomized clinical trials to enable prediction of treatment response. On the basis of the findings from the landscape analysis, Movember will now have an increased focus on addressing the specific research needs that have been identified, with particular investment in research efforts that reduce disease progression and lead to improved therapies for advanced prostate cancer.

The Movember global Landscape Analysis Committee (LAC) was established to act as an independent group of experts across urology, medical oncology, radiation oncology, radiology, pathology, translational research, health economics and patient advocacy to identify the highest priority research needs across the prostate cancer biomedical research domain. Findings from the landscape analysis illustrate the research priorities in prostate cancer and will enable Movember to focus on specific needs, with particular investment in research to reduce disease progression and improve therapies for advanced prostate cancer.

An Individualized, Data-Driven Digital Approach for Precision Behavior Change

Chronic disease now affects approximately half of the US population, causes 7 in 10 deaths, and accounts for roughly 80% of US health care expenditure. Because the root causes of chronic diseases are largely behavioral, effective therapies require frequent, individualized interventions that extend beyond the hospital and clinic to reach patients in their day-to-day lives. However, a mismatch currently exists between what the health care system is equipped to provide and the interventions necessary to effectively address the chronic disease burden. To remedy this health crisis, we present an individualized, data-driven digital approach for chronic disease management and prevention through precision behavior change. The rapid growth of information, biological, and communication technologies makes this an opportune time to develop digital tools that deliver precision interventions for health behavior change to address the chronic disease crisis. Building on this rapid growth, we propose a framework that includes the precise targeting of risk-producing behaviors using real-time sensing technology, machine learning data analysis to identify the most effective intervention, and delivery of that intervention with health-reinforcing feedback to provide real-time, individualized support to empower sustainable health behavior change.

Harnessing repeated measurements of predictor variables for clinical risk prediction: a review of existing methods

BACKGROUND

Clinical prediction models (CPMs) predict the risk of health outcomes for individual patients. The majority of existing CPMs only harness cross-sectional patient information. Incorporating repeated measurements, such as those stored in electronic health records, into CPMs may provide an opportunity to enhance their performance. However, the number and complexity of methodological approaches available could make it difficult for researchers to explore this opportunity. Our objective was to review the literature and summarise existing approaches for harnessing repeated measurements of predictor variables in CPMs, primarily to make this field more accessible for applied researchers.

METHODS

MEDLINE, Embase and Web of Science were searched for articles reporting the development of a multivariable CPM for individual-level prediction of future binary or time-to-event outcomes and modelling repeated measurements of at least one predictor. Information was extracted on the following: the methodology used, its specific aim, reported advantages and limitations, and software available to apply the method.

RESULTS

The search revealed 217 relevant articles. Seven methodological frameworks were identified: time-dependent covariate modelling, generalised estimating equations, landmark analysis, two-stage modelling, joint-modelling, trajectory classification and machine learning. Each of these frameworks satisfies at least one of three aims: to better represent the predictor-outcome relationship over time, to infer a covariate value at a pre-specified time and to account for the effect of covariate change.

CONCLUSIONS

The applicability of identified methods depends on the motivation for including longitudinal information and the method's compatibility with the clinical context and available patient data, for both model development and risk estimation in practice.

Combination possibility and deep learning model as clinical decision-aided approach for prostate cancer

This study aims to introduce as proof of concept a combination model for classification of prostate cancer using deep learning approaches. We utilized patients with prostate cancer who underwent surgical treatment representing the various conditions of disease progression. All possible combinations of significant variables from logistic regression and correlation analyses were determined from study data sets. The combination possibility and deep learning model was developed to predict these combinations that represented clinically meaningful patient's subgroups. The observed relative frequencies of different tumor stages and Gleason score Gls changes from biopsy to prostatectomy were available for each group. Deep learning models and seven machine learning approaches were compared for the classification performance of Gleason score changes and pT2 stage. Deep models achieved the highest F1 scores by pT2 tumors (0.849) and Gls change (0.574). Combination possibility and deep learning model is a useful decision-aided tool for prostate cancer and to group patients with prostate cancer into clinically meaningful groups.

Challenges and Opportunities for Using Big Health Care Data to Advance Medical Science and Public Health

Methodological advancements in epidemiology, biostatistics, and data science have strengthened the research world's ability to use data captured from electronic health records (EHRs) to address pressing medical questions, but gaps remain. We describe methods investments that are needed to curate EHR data toward research quality and to integrate complementary data sources when EHR data alone are insufficient for research goals. We highlight new methods and directions for improving the integrity of medical evidence generated from pragmatic trials, observational studies, and predictive modeling. We also discuss needed methods contributions to further ease data sharing across multisite EHR data networks. Throughout, we identify opportunities for training and for bolstering collaboration among subject matter experts, methodologists, practicing clinicians, and health system leaders to help ensure that methods problems are identified and resulting advances are translated into mainstream research practice more quickly.

A Bayesian hierarchical logistic regression model of multiple informant family health histories

Background

Family health history (FHH) inherently involves collecting proxy reports of health statuses of related family members. Traditionally, such information has been collected from a single informant. More recently, research has suggested that a multiple informant approach to collecting FHH results in improved individual risk assessments. Likewise, recent work has emphasized the importance of incorporating health-related behaviors into FHH-based risk calculations. Integrating both multiple accounts of FHH with behavioral information on family members represents a significant methodological challenge as such FHH data is hierarchical in nature and arises from potentially error-prone processes.

Methods

In this paper, we introduce a statistical model that addresses these challenges using informative priors for background variation in disease prevalence and the effect of other, potentially correlated, variables while accounting for the nested structure of these data. Our empirical example is drawn from previously published data on families with a history of diabetes.

Results

The results of the comparative model assessment suggest that simply accounting for the structured nature of multiple informant FHH data improves classification accuracy over the baseline and that incorporating family member health-related behavioral information into the model is preferred over alternative specifications.

Conclusions

The proposed modelling framework is a flexible solution to integrate multiple informant FHH for risk prediction purposes.

Electronic supplementary material

The online version of this article (10.1186/s12874-019-0700-5) contains supplementary material, which is available to authorized users.

Estimating autoantibody signatures to detect autoimmune disease patient subsets

Autoimmune diseases are characterized by highly specific immune responses against molecules in self-tissues. Different autoimmune diseases are characterized by distinct immune responses, making autoantibodies useful for diagnosis and prediction. In many diseases, the targets of autoantibodies are incompletely defined. Although the technologies for autoantibody discovery have advanced dramatically over the past decade, each of these techniques generates hundreds of possibilities, which are onerous and expensive to validate. We set out to establish a method to greatly simplify autoantibody discovery, using a pre-filtering step to define subgroups with similar specificities based on migration of radiolabeled, immunoprecipitated proteins on sodium dodecyl sulfate (SDS) gels and autoradiography [Gel Electrophoresis and band detection on Autoradiograms (GEA)]. Human recognition of patterns is not optimal when the patterns are complex or scattered across many samples. Multiple sources of errors-including irrelevant intensity differences and warping of gels-have challenged automation of pattern discovery from autoradiograms.In this article, we address these limitations using a Bayesian hierarchical model with shrinkage priors for pattern alignment and spatial dewarping. The Bayesian model combines information from multiple gel sets and corrects spatial warping for coherent estimation of autoantibody signatures defined by presence or absence of a grid of landmark proteins. We show the pre-processing creates more clearly separated clusters and improves the accuracy of autoantibody subset detection via hierarchical clustering. Finally, we demonstrate the utility of the proposed methods with GEA data from scleroderma patients.

A Bayesian latent class approach for EHR-based phenotyping

Phenotyping, ie, identification of patients possessing a characteristic of interest, is a fundamental task for research conducted using electronic health records. However, challenges to this task include imperfect sensitivity and specificity of clinical codes and inconsistent availability of more detailed data such as laboratory test results. Despite these challenges, most existing electronic health records-derived phenotypes are rule-based, consisting of a series of Boolean arguments informed by expert knowledge of the disease of interest and its coding. The objective of this paper is to introduce a Bayesian latent phenotyping approach that accounts for imperfect data elements and missing not at random missingness patterns that can be used when no gold-standard data are available. We conducted simulation studies to compare alternative phenotyping methods under different patterns of missingness and applied these approaches to a cohort of 68 265 children at elevated risk for type 2 diabetes mellitus (T2DM). In simulation studies, the latent class approach had similar sensitivity to a rule-based approach (95.9% vs 91.9%) while substantially improving specificity (99.7% vs 90.8%). In the PEDSnet cohort, we found that biomarkers and clinical codes were strongly associated with latent T2DM status. The latent T2DM class was also strongly predictive of missingness in biomarkers. Glucose was missing in 83.4% of patients (odds ratio for latent T2DM status = 0.52) while hemoglobin A1c was missing in 91.2% (odds ratio for latent T2DM status = 0.03 ), suggesting missing not at random missingness. The latent phenotype approach may substantially improve on rule-based phenotyping.

Clinical Evaluation of an Individualized Risk Prediction Tool for Men on Active Surveillance for Prostate Cancer

OBJECTIVE

To determine whether providing individualized predictions of health outcomes to men on active surveillance (AS) alleviates cancer-related anxiety and improves risk understanding.

MATERIALS AND METHODS

We consecutively recruited men from our large, institutional AS program before (n = 36) and after (n = 31) implementation of a risk prediction tool. Men in both groups were surveyed before and after their regular visits to assess their perceived cancer control, biopsy-specific anxiety, and burden from cancer-related information. We compared pre-/post-visit differences between men who were and were not shown the tool using two-sample t-tests. Satisfaction with and understanding of the predictions were elicited from men in the intervention period.

RESULTS

Men reported a relatively high level of cancer control at baseline. Men who were not shown the tool saw a 6.3 point increase (scaled from 0 to 100) in their perceived cancer control from before to after their visit whereas men who were shown the tool saw a 12.8 point increase, indicating a statistically significant difference between groups (p = .04). Biopsy-specific anxiety and burden from cancer information were not significantly different between groups. Men were satisfied with the tool and demonstrated moderate understanding.

CONCLUSION

Providing individualized predictions to men on AS helps them better understand their cancer risk and should be considered at other clinical sites.

Active surveillance: a review of risk-based, dynamic monitoring

Active surveillance (AS) is an important treatment modality aiming to reduce the overtreatment of patients with prostate cancer (PCa) who have a low risk of disease reclassification. After enrolling in AS patients are actively monitored using different diagnostic tests (e.g., prostate specific-antigen, digital rectal exams (DREs), medical imaging, and prostate biopsies). Biopsy is the most burdensome test. We aimed to review schedules for monitoring men on AS. We compare fixed versus risk based dynamic monitoring, where biopsies are scheduled during follow-up based on dynamic risk predictions. Several prediction models and scheduling techniques have been published. All proposed risk prediction models need further external validation. We conclude that risk based, dynamic monitoring is a promising new strategy to further reduce overtreatment in PCa patients.

Predicting high risk births with contraceptive prevalence and contraceptive method-mix in an ecologic analysis

Background

Increased contraceptive use has been associated with a decrease in high parity births, births that occur close together in time, and births to very young or to older women. These types of births are also associated with high risk of under-five mortality. Previous studies have looked at the change in the level of contraception use and the average change in these types of high-risk births. We aim to predict the distribution of births in a specific country when there is a change in the level and method of modern contraception.

Methods

We used data from full birth histories and modern contraceptive use from 207 nationally representative Demographic and Health Surveys covering 71 countries to describe the distribution of births in each survey based on birth order, preceding birth space, and mother’s age at birth. We estimated the ecologic associations between the prevalence and method-mix of modern contraceptives and the proportion of births in each category. Hierarchical modelling was applied to these aggregated cross sectional proportions, so that random effects were estimated for countries with multiple surveys. We use these results to predict the change in type of births associated with scaling up modern contraception in three different scenarios.

Results

We observed marked differences between regions, in the absolute rates of contraception, the types of contraceptives in use, and in the distribution of type of birth. Contraceptive method-mix was a significant determinant of proportion of high-risk births, especially for birth spacing, but also for mother’s age and parity. Increased use of modern contraceptives is especially predictive of reduced parity and more births with longer preceding space. However, increased contraception alone is not associated with fewer births to women younger than 18 years or a decrease in short-spaced births.

Conclusions

Both the level and the type of contraception are important factors in determining the effects of family planning on changes in distribution of high-risk births. The best predictions for how birth risk changes with increased modern contraception and for different contraception methods allow for more nuanced predictions specific to each country and can aid better planning for the scaling up of modern contraception.

Electronic supplementary material

The online version of this article (10.1186/s12889-017-4741-6) contains supplementary material, which is available to authorized users.