Targeted learning in real‐world comparative effectiveness research with time‐varying interventions

In comparative effectiveness research (CER), often the aim is to contrast survival outcomes between exposure groups defined by time‐varying interventions. With observational data, standard regression analyses (e.g., Cox modeling) cannot account for time‐dependent confounders on causal pathways between exposures and outcome nor for time‐dependent selection bias that may arise from informative right censoring. Inverse probability weighting (IPW) estimation to fit marginal structural models (MSMs) has commonly been applied to properly adjust for these expected sources of bias in real‐world observational studies. We describe the application and performance of an alternate estimation approach in such a study. The approach is based on the recently proposed targeted learning methodology and consists in targeted minimum loss‐based estimation (TMLE) with super learning (SL) within a nonparametric MSM. The evaluation is based on the analysis of electronic health record data with both IPW estimation and TMLE to contrast cumulative risks under four more or less aggressive strategies for treatment intensification in adults with type 2 diabetes already on 2+ oral agents or basal insulin. Results from randomized experiments provide a surrogate gold standard to validate confounding and selection bias adjustment. Bootstrapping is used to validate analytic estimation of standard errors. This application does the following: (1) establishes the feasibility of TMLE in real‐world CER based on large healthcare databases; (2) provides evidence of proper confounding and selection bias adjustment with TMLE and SL; and (3) motivates their application for improving estimation efficiency. Claims are reinforced with a simulation study that also illustrates the double‐robustness property of TMLE. Copyright © 2014 John Wiley & Sons, Ltd.     

The use of artificial intelligence-based innovations in the health sector in Tanzania: A scoping review

BackgroundArtificial Intelligence (AI) has great potential to transform health systems to improve the quality of healthcare services. However, AI is still new in Tanzania, and there is limited knowledge about the application of AI technology in the Tanzanian health sector.ObjectivesThis study aims to explore the current status, challenges, and opportunities for AI application in the health system in Tanzania.MethodsA scoping review was conducted using the Preferred Reporting Items for Systematic Review and Meta-Analysis Extensions for Scoping Review (PRISMA-ScR). We searched different electronic databases such as PubMed, Embase, African Journal Online, and Google Scholar.ResultsEighteen (18) studies met the inclusion criteria out of 2,017 studies from different electronic databases and known AI-related project websites. Amongst AI-driven solutions, the studies mostly used machine learning (ML) and deep learning for various purposes, including prediction and diagnosis of diseases and vaccine stock optimisation. The most commonly used algorithms were conventional machine learning, including Random Forest and Neural network, Naive Bayes K-Nearest Neighbour and Logistic regression.ConclusionsThis review shows that AI-based innovations may have a role in improving health service delivery, including early outbreak prediction and detection, disease diagnosis and treatment, and efficient management of healthcare resources in Tanzania. Our results indicate the need for developing national AI policies and regulatory frameworks for adopting responsible and ethical AI solutions in the health sector in accordance with the World Health Organisation (WHO) guidance on ethics and governance of AI for health.     

High‐dimensional propensity score algorithm in comparative effectiveness research with time‐varying interventions

The high‐dimensional propensity score (hdPS) algorithm was proposed for automation of confounding adjustment in problems involving large healthcare databases. It has been evaluated in comparative effectiveness research (CER) with point treatments to handle baseline confounding through matching or covariance adjustment on the hdPS. In observational studies with time‐varying interventions, such hdPS approaches are often inadequate to handle time‐dependent confounding and selection bias. Inverse probability weighting (IPW) estimation to fit marginal structural models can adequately handle these biases under the fundamental assumption of no unmeasured confounders. Upholding of this assumption relies on the selection of an adequate set of covariates for bias adjustment. We describe the application and performance of the hdPS algorithm to improve covariate selection in CER with time‐varying interventions based on IPW estimation and explore stabilization of the resulting estimates using Super Learning. The evaluation is based on both the analysis of electronic health records data in a real‐world CER study of adults with type 2 diabetes and a simulation study. This report (i) establishes the feasibility of IPW estimation with the hdPS algorithm based on large electronic health records databases, (ii) demonstrates little impact on inferences when supplementing the set of expert‐selected covariates using the hdPS algorithm in a setting with extensive background knowledge, (iii) supports the application of the hdPS algorithm in discovery settings with little background knowledge or limited data availability, and (iv) motivates the application of Super Learning to stabilize effect estimates based on the hdPS algorithm. Copyright © 2014 John Wiley & Sons, Ltd.     

A scoping review to understand the effectiveness of linking schemes from healthcare providers to community resources to improve the health and well‐being of people with long‐term conditions

The prevalence of people living with long‐term conditions is increasing, accompanied by an increased expectation that patients will become more involved in self‐management. Long‐term conditions are associated with increased social isolation and poor physical and mental health. But there remains a gap in health provision between providing medical treatment and effectively addressing psychosocial well‐being. One potential way of addressing this gap is by utilising social interventions which link patients from health services to community‐based sources of support. However, the mechanisms involved in the delivery of interventions providing that link and their effectiveness remain unclear. This review adopted the methodological framework for conducting scoping studies, searching for both academic and grey literature on social interventions which link people from healthcare settings to a range of community and voluntary sector organisations. A literature search between May and June 2013, involving five electronic databases, hand searching of two journals and the use of Google search engine, identified seven studies relevant to the review question. In terms of key characteristics and mechanisms of the interventions, mental health conditions and social isolation were the most common reasons for referral to the interventions, and referrals were usually made through general practices. Almost all the interventions were facilitator‐led, whereby the facilitator worked to identify and link participants to appropriate community‐based resources. In regard to health and social outcomes and their cost‐effectiveness, studies reported improvement to participants' psychological and social well‐being as well as their decreased use of health services, although there were limited measures of participants' physical health outcomes. Interventions for linking patients from healthcare setting to community‐based resources target and address psychosocial needs of participants. The review identified involvement of health professionals in aiding the referral of patients to the intervention and the role of the intervention facilitators as key components of the interventions.     

High‐dimensional variable selection and prediction under competing risks with application to SEER‐Medicare linked data

Competing risk analysis considers event times due to multiple causes or of more than one event types. Commonly used regression models for such data include (1) cause‐specific hazards model, which focuses on modeling one type of event while acknowledging other event types simultaneously, and (2) subdistribution hazards model, which links the covariate effects directly to the cumulative incidence function. Their use in the presence of high‐dimensional predictors are largely unexplored. Motivated by an analysis using the linked SEER‐Medicare database for the purposes of predicting cancer versus noncancer mortality for patients with prostate cancer, we study the accuracy of prediction and variable selection of existing machine learning methods under both models using extensive simulation experiments, including different approaches to choosing penalty parameters in each method. We then apply the optimal approaches to the analysis of the SEER‐Medicare data.     

Alternative evaluation metrics for risk adjustment methods

Risk adjustment is instituted to counter risk selection by accurately equating payments with expected expenditures. Traditional risk‐adjustment methods are designed to estimate accurate payments at the group level. However, this generates residual risks at the individual level, especially for high‐expenditure individuals, thereby inducing health plans to avoid those with high residual risks. To identify an optimal risk‐adjustment method, we perform a comprehensive comparison of prediction accuracies at the group level, at the tail distributions, and at the individual level across 19 estimators: 9 parametric regression, 7 machine learning, and 3 distributional estimators. Using the 2013–2014 MarketScan database, we find that no one estimator performs best in all prediction accuracies. Generally, machine learning and distribution‐based estimators achieve higher group‐level prediction accuracy than parametric regression estimators. However, parametric regression estimators show higher tail distribution prediction accuracy and individual‐level prediction accuracy, especially at the tails of the distribution. This suggests that there is a trade‐off in selecting an appropriate risk‐adjustment method between estimating accurate payments at the group level and lower residual risks at the individual level. Our results indicate that an optimal method cannot be determined solely on the basis of statistical metrics but rather needs to account for simulating plans' risk selective behaviors.     

Using web‐based simulation learning to deal with shoulder dystocia

Simulation‐based learning can provide education and training to individuals in a variety of industries. The military, aviation and nuclear power industries have seen demonstrable improvement in critical event handling through the use of this type of learning when coupled with crew resource management techniques. Simulation training is also emerging in the healthcare field as an effective patient safety and risk management tool. In the area of obstetrics, healthcare risk managers can have an impact on injury and claim reduction rates related to shoulder dystocia by advocating for and implementing simulation training for obstetrical providers. This article will describe a shoulder dystocia simulation training program that uses a high‐fidelity mannequin, and illustrate the impact that the healthcare risk manager can have in reducing injuries due to shoulder dystocia.     

“Caregiving is a full‐time job” impacting stroke caregivers' health and well‐being: A qualitative meta‐synthesis

Family caregivers contribute to the sustainability of healthcare systems. Stroke is a leading cause of adult disability and many people with stroke rely on caregiver support to return home and remain in the community. Research has demonstrated the importance of caregivers, but suggests that caregiving can have adverse consequences. Despite the body of qualitative stroke literature, there is little clarity about how to incorporate these findings into clinical practice. This review aimed to characterise stroke caregivers' experiences and the impact of these experiences on their health and well‐being. We conducted a qualitative meta‐synthesis. Four electronic databases were searched to identify original qualitative research examining stroke caregivers' experiences. In total, 4,481 citations were found, with 39 studies remaining after removing duplicates and applying inclusion and exclusions criteria. Articles were appraised for quality using the Critical Appraisal Skills Programme (CASP), coded using NVivo software, and analysed through thematic synthesis. One overarching theme, ‘caregiving is a full‐time job’ was identified, encompassing four sub‐themes: (a) restructured life, (b) altered relationships, (c) physical challenges, and (d) psychosocial challenges. Community and institution‐based clinicians should be aware of the physical and psychosocial consequences of caregiving and provide appropriate supports, such as education and respite, to optimise caregiver health and well‐being. Future research may build upon this study to identify caregivers in most need of support and the types of support needed across a broad range of health conditions.     

Comparing methods for estimation of heterogeneous treatment effects using observational data from health care databases

There is growing interest in using routinely collected data from health care databases to study the safety and effectiveness of therapies in “real‐world” conditions, as it can provide complementary evidence to that of randomized controlled trials. Causal inference from health care databases is challenging because the data are typically noisy, high dimensional, and most importantly, observational. It requires methods that can estimate heterogeneous treatment effects while controlling for confounding in high dimensions. Bayesian additive regression trees, causal forests, causal boosting, and causal multivariate adaptive regression splines are off‐the‐shelf methods that have shown good performance for estimation of heterogeneous treatment effects in observational studies of continuous outcomes. However, it is not clear how these methods would perform in health care database studies where outcomes are often binary and rare and data structures are complex. In this study, we evaluate these methods in simulation studies that recapitulate key characteristics of comparative effectiveness studies. We focus on the conditional average effect of a binary treatment on a binary outcome using the conditional risk difference as an estimand. To emulate health care database studies, we propose a simulation design where real covariate and treatment assignment data are used and only outcomes are simulated based on nonparametric models of the real outcomes. We apply this design to 4 published observational studies that used records from 2 major health care databases in the United States. Our results suggest that Bayesian additive regression trees and causal boosting consistently provide low bias in conditional risk difference estimates in the context of health care database studies.     

Health‐related quality of life and activities of daily living in 85‐year‐olds in Sweden

Few studies have examined health‐related quality of life (HRQoL) with respect to daily living and health factors for relatively healthy elderly individuals. To this end, this study examines 85‐year‐olds' reported HRQoL in relation to social support, perceived health, chronic diseases, healthcare use and instrumental activities of daily living (IADL). Data were collected from 360 participants (55% response rate) between March 2007 and March 2008 using a postal questionnaire and a home visit interview. HRQoL was assessed using the EQ‐5D‐3L. For the items in the EQ‐5D‐3L, more problems were related to lower HRQoL. Restricted mobility and occurrence of pain/discomfort was common. Lower HRQoL was associated with increased risk for depression, increased use of medication, increased number of chronic diseases and more problems with IADL. Healthcare use and healthcare costs were correlated with lower HRQoL. HRQoL is of importance to healthcare providers and must be considered together with IADL in the elderly population when planning interventions. These should take into account the specific needs and resources of the older individuals.     

Mixture drug‐count response model for the high‐dimensional drug combinatory effect on myopathy

Drug‐drug interactions (DDIs) are a common cause of adverse drug events (ADEs). The electronic medical record (EMR) database and the FDA's adverse event reporting system (FAERS) database are the major data sources for mining and testing the ADE associated DDI signals. Most DDI data mining methods focus on pair‐wise drug interactions, and methods to detect high‐dimensional DDIs in medical databases are lacking. In this paper, we propose 2 novel mixture drug‐count response models for detecting high‐dimensional drug combinations that induce myopathy. The “count” indicates the number of drugs in a combination. One model is called fixed probability mixture drug‐count response model with a maximum risk threshold (FMDRM‐MRT). The other model is called count‐dependent probability mixture drug‐count response model with a maximum risk threshold (CMDRM‐MRT), in which the mixture probability is count dependent. Compared with the previous mixture drug‐count response model (MDRM) developed by our group, these 2 new models show a better likelihood in detecting high‐dimensional drug combinatory effects on myopathy. CMDRM‐MRT identified and validated (54; 374; 637; 442; 131) 2‐way to 6‐way drug interactions, respectively, which induce myopathy in both EMR and FAERS databases. We further demonstrate FAERS data capture much higher maximum myopathy risk than EMR data do. The consistency of 2 mixture models' parameters and local false discovery rate estimates are evaluated through statistical simulation studies.     

Impact of COVID‐19 on access to healthcare in low‐ and middle‐income countries: Current evidence and future recommendations

The COVID‐19 pandemic continues to be a major public health threat globally and low‐ and middle‐income countries (LMICs) are not an exception. The impact of the COVID‐19 pandemic is far‐reaching on many areas including but not limited to global health security, economic and healthcare delivery with a potential impact on access to healthcare in LMICs. We evaluate the impact of the COVID‐19 pandemic on access to healthcare in LMICs, as well as plausible strategies that can be put in place to ensure that the delivery of healthcare is not halted. In order to mitigate the devastating effect of the COVID‐19 pandemic on the already weak health systems in LMICs, it is much necessary to reinforce and scale up interventions and proactive measures that will ensure that access to healthcare is not disrupted even in course of the pandemic.     

Developing points‐based risk‐scoring systems in the presence of competing risks

Predicting the occurrence of an adverse event over time is an important issue in clinical medicine. Clinical prediction models and associated points‐based risk‐scoring systems are popular statistical methods for summarizing the relationship between a multivariable set of patient risk factors and the risk of the occurrence of an adverse event. Points‐based risk‐scoring systems are popular amongst physicians as they permit a rapid assessment of patient risk without the use of computers or other electronic devices. The use of such points‐based risk‐scoring systems facilitates evidence‐based clinical decision making. There is a growing interest in cause‐specific mortality and in non‐fatal outcomes. However, when considering these types of outcomes, one must account for competing risks whose occurrence precludes the occurrence of the event of interest. We describe how points‐based risk‐scoring systems can be developed in the presence of competing events. We illustrate the application of these methods by developing risk‐scoring systems for predicting cardiovascular mortality in patients hospitalized with acute myocardial infarction. Code in the R statistical programming language is provided for the implementation of the described methods. © 2016 The Authors. Statistics in Medicine published by John Wiley & Sons Ltd.     

Internet health scams—Developing a taxonomy and risk‐of‐deception assessment tool

The prevalence of health scams in Canada is increasing, facilitated by the rise of the Internet. However, little is known about the nature of this phenomena. This study sought to methodically identify and categorise Internet‐based Health Scams (IHS) currently active in Canada, creating an initial taxonomy based on systematic Internet searches. A five‐step Delphi approach, comprised of a multidisciplinary panel of health professionals from the University of British Columbia, in Vancouver, Canada, was used to establish consensus. The resulting taxonomy is the first to characterise the nature of IHS in North America. Five core areas of activity were identified: body image products, medical products, alternative health services, healthy lifestyle products, and diagnostic testing services. IHS purveyors relied on social expectations and psychological persuasion techniques to target consumers. Persuasion techniques included social engagement, claims of miraculous effects, scarcity, and the use of pseudoscientific language. These techniques exploited personality traits of sensation seeking, needing self‐control, openness to taking risks, and the preference for uniqueness. The data gathered from the taxonomy allowed the Delphi panel to develop and pilot a simple risk‐of‐deception tool. This tool is intended to help healthcare professionals educate the public about IHS. It is suggested that, where relevant, healthcare professionals include a general discussion of IHS risks and marketing techniques with clients as a part of health promotion activities.     

‘Genuine doctor care’: Perspectives on general practice and community‐based care of Australian men experiencing homelessness

People with complex health and social needs, including tri‐morbidity and homelessness, are challenging for modern healthcare systems. These clients have poor health and social outcomes. They tend to use available health resources inefficiently, with fragmented, uncoordinated use of multiple health and social care services. Increasing access for these clients to well‐supported general practice care may be an effective response to these challenges. The aim of this study was to explore client experiences of, and attitudes to, community‐based healthcare, and general practice in particular, to identify opportunities to improve healthcare provision. Five focus groups with a total of 20 men currently experiencing homelessness were facilitated by the corresponding author in an inner‐city homeless hostel. Discussions were transcribed, coded and analysed thematically. The analysis was informed by earlier focus group discussions with community‐based homeless healthcare providers. Participants reported reluctance to engage with healthcare providers outside times of perceived crisis, and experiences of stigma and dismissive care. Some participants were sceptical of the motivations of health and social care providers, including general practitioners. Presentations with physical and psychological pain featured prominently in participant accounts. Three key themes identified important aspects of client experiences of community‐based healthcare which indicate potential areas for improvement. These themes were as follows: the relative invisibility and low salience of general practice compared to hospital‐based emergency and inpatient services; discontinuity within community‐based healthcare and across transitions between community‐based and other healthcare; and inconsistent and unsatisfactory general practitioner responses to physical and psychological pain. These responses included apparent over‐prescribing, under‐prescribing and short‐term ‘band‐aid’ responses. Generalist medical expertise was valued in general practitioners, but not consistently experienced. A number of challenges and opportunities exist, at both individual and system levels, for general practice to realise its potential to deliver effective, compassionate and efficient care to clients experiencing homelessness.     

Direct and indirect costs of home healthcare in Japan: A cross‐sectional study

To tackle the rising healthcare expenditure in an ageing society in Japan, home healthcare has been promoted over the past several years. However, there is a dearth of literature on total costs incurring for home healthcare. In this study, we conducted a cross‐sectional study among patients, who received home healthcare in the month of May, 2018. Direct healthcare costs and patients’ clinical characteristics were collected from medical records and long‐term care databases (n = 166). Indirect costs were estimated using a questionnaire survey which obtained information on job absenteeism and care time from the caregiver. A total of 112 patients responded to the survey. The median age was 82 years (interquartile range: 74–88). Total per‐person per month home‐care costs averaged USD 6,163 with direct costs (USD 2,547) and indirect costs (USD 3,596) accounted for 41.3% and 58.3% of the total costs, respectively. The largest components of direct costs were long‐term care costs (48%) and medical costs (47%). Multivariable adjusted model showed that those with heavy healthcare were more likely to incur higher total as well as direct and indirect home healthcare cost (p<.05 for each). Patients aged >75 years (p = .041) were less likely and those who used oxygen at home were more likely to incur direct home healthcare cost (p = .001) than their counterpart. Our study findings show that indirect cost is a major contributor to total home healthcare costs in Japan. Also for patients who need heavy healthcare, both direct and indirect costs are large burden.     

Martingale residual‐based method to control for confounders measured only in a validation sample in time‐to‐event analysis

Unmeasured confounding remains an important problem in observational studies, including pharmacoepidemiological studies of large administrative databases. Several recently developed methods utilize smaller validation samples, with information on additional confounders, to control for confounders unmeasured in the main, larger database. However, up‐to‐date applications of these methods to survival analyses seem to be limited to propensity score calibration, which relies on a strong surrogacy assumption. We propose a new method, specifically designed for time‐to‐event analyses, which uses martingale residuals, in addition to measured covariates, to enhance imputation of the unmeasured confounders in the main database. The method is applicable for analyses with both time‐invariant data and time‐varying exposure/confounders. In simulations, our method consistently eliminated bias because of unmeasured confounding, regardless of surrogacy violation and other relevant design parameters, and almost always yielded lower mean squared errors than other methods applicable for survival analyses, outperforming propensity score calibration in several scenarios. We apply the method to a real‐life pharmacoepidemiological database study of the association between glucocorticoid therapy and risk of type II diabetes mellitus in patients with rheumatoid arthritis, with additional potential confounders available in an external validation sample. Compared with conventional analyses, which adjust only for confounders measured in the main database, our estimates suggest a considerably weaker association. Copyright © 2016 John Wiley & Sons, Ltd.     

Families' and healthcare professionals' perceptions of healthcare services for children and young people with medically unexplained symptoms: a narrative review of the literature

Children and young people frequently report physical complaints that have no observable physical pathology known as medically unexplained symptoms (MUS). Research suggests that MUS are associated with substantial physical and psychological impairments and may have a negative impact on children's and young people's functional status and well‐being in the long term. Due to the potentially complex needs of this group, children and young people with MUS may require timely access to suitable health and social care services to effectively manage symptoms and achieve their academic, social and personal potential. Families and professionals can offer important insights into the availability and appropriateness of current community and specialist health and social care services. This review is the first critical evaluation and synthesis of research that has examined families' and healthcare professionals' (HCP) perceptions of healthcare services for children and young people with MUS. A systematic search of electronic databases and manual searches of key journals and reference lists identified 17 papers from 15 studies for inclusion in the review. The review highlights the paucity of rigorously conducted research on this topic. Studies have been narrowly focused on the views of a homogeneous group of mothers and young people attending single centres. There has been some attempt to examine doctors' views, but the perceptions of children, fathers and health and social care professionals are absent or under‐represented, and multi‐site and longitudinal studies are lacking. Thematic analysis of the results from the included studies suggests that knowledge, communication, health beliefs and healthcare settings are factors that influence families' and HCPs' perceptions of services. Families report dissatisfaction with some HCPs' approach to managing MUS. The findings suggest that children and young people with MUS are at risk of receiving suboptimal care and support because there is insufficient research to inform high‐quality, evidence‐based practice.