A scoping review of studies using observational data to optimise dynamic treatment regimens

Optimizing warfarin dosing for patients with atrial fibrillation using machine learning

While novel oral anticoagulants are increasingly used to reduce risk of stroke in patients with atrial fibrillation, vitamin K antagonists such as warfarin continue to be used extensively for stroke prevention across the world. While effective in reducing the risk of strokes, the complex pharmacodynamics of warfarin make it difficult to use clinically, with many patients experiencing under- and/or over- anticoagulation. In this study we employed a novel implementation of deep reinforcement learning to provide clinical decision support to optimize time in therapeutic International Normalized Ratio (INR) range. We used a novel semi-Markov decision process formulation of the Batch-Constrained deep Q-learning algorithm to develop a reinforcement learning model to dynamically recommend optimal warfarin dosing to achieve INR of 2.0-3.0 for patients with atrial fibrillation. The model was developed using data from 22,502 patients in the warfarin treated groups of the pivotal randomized clinical trials of edoxaban (ENGAGE AF-TIMI 48), apixaban (ARISTOTLE) and rivaroxaban (ROCKET AF). The model was externally validated on data from 5730 warfarin-treated patients in a fourth trial of dabigatran (RE-LY) using multilevel regression models to estimate the relationship between center-level algorithm consistent dosing, time in therapeutic INR range (TTR), and a composite clinical outcome of stroke, systemic embolism or major hemorrhage. External validation showed a positive association between center-level algorithm-consistent dosing and TTR (R2 = 0.56). Each 10% increase in algorithm-consistent dosing at the center level independently predicted a 6.78% improvement in TTR (95% CI 6.29, 7.28; p < 0.001) and a 11% decrease in the composite clinical outcome (HR 0.89; 95% CI 0.81, 1.00; p = 0.015). These results were comparable to those of a rules-based clinical algorithm used for benchmarking, for which each 10% increase in algorithm-consistent dosing independently predicted a 6.10% increase in TTR (95% CI 5.67, 6.54, p < 0.001) and a 10% decrease in the composite outcome (HR 0.90; 95% CI 0.83, 0.98, p = 0.018). Our findings suggest that a deep reinforcement learning algorithm can optimize time in therapeutic range for patients taking warfarin. A digital clinical decision support system to promote algorithm-consistent warfarin dosing could optimize time in therapeutic range and improve clinical outcomes in atrial fibrillation globally.

Imputation-based Q-learning for optimizing dynamic treatment regimes with right-censored survival outcome

Q-learning has been one of the most commonly used methods for optimizing dynamic treatment regimes (DTRs) in multistage decision-making. Right-censored survival outcome poses a significant challenge to Q-Learning due to its reliance on parametric models for counterfactual estimation which are subject to misspecification and sensitive to missing covariates. In this paper, we propose an imputation-based Q-learning (IQ-learning) where flexible nonparametric or semiparametric models are employed to estimate optimal treatment rules for each stage and then weighted hot-deck multiple imputation (MI) and direct-draw MI are used to predict optimal potential survival times. Missing data are handled using inverse probability weighting and MI, and the nonrandom treatment assignment among the observed is accounted for using a propensity-score approach. We investigate the performance of IQ-learning via extensive simulations and show that it is more robust to model misspecification than existing Q-Learning methods, imputes only plausible potential survival times contrary to parametric models and provides more flexibility in terms of baseline hazard shape. Using IQ-learning, we developed an optimal DTR for leukemia treatment based on a randomized trial with observational follow-up that motivated this study.

Research Reviews: Advances in methods for evaluating child and adolescent mental health interventions

BACKROUND

The evidence base for interventions for child mental health and neurodevelopment is weak and the current capacity for rigorous evaluation limited. We describe some of the challenges that make this field particularly difficult and expensive for evaluation studies.

METHODS

We describe and review the use of novel study designs and analysis methodology for their potential to improve this situation.

RESULTS

While several novel designs appeared ill-suited to our field, systematic review found others that offered potential but had yet to be widely adopted, some not at all.

CONCLUSIONS

While funding is inevitably a constraint, we argue that improvements in the evidence base of both current and new treatments will only be achieved by the adoption of a number of these new technologies and study designs, the consistent application of rigorous constructive but demanding standards, and the engagement of the public, patients, clinical and research services to build a design, recruitment, and analysis infrastructure.

Personalising airway clearance in chronic suppurative lung diseases: a scoping review

Background

Personalised airway clearance techniques are commonly recommended to augment mucus clearance in chronic suppurative lung diseases. It is unclear what current literature tells us about how airway clearance regimens should be personalised. This scoping review explores current research on airway clearance technique in chronic suppurative lung diseases, to establish the extent and type of guidance in this area, identify knowledge gaps and determine the factors which physiotherapists should consider when personalising airway clearance regimens.

Methods

Systematic searching of online databases (MEDLINE, EMBASE, CINAHL, PEDro, Cochrane, Web of Science) was used to identify full-text publications in the last 25 years that described methods of personalising airway clearance techniques in chronic suppurative lung diseases. Items from the TIDieR framework provided a priori categories which were modified based on the initial data to develop a "Best-fit" framework for data charting. The findings were subsequently transformed into a personalisation model.

Results

A broad range of publications were identified, most commonly general review papers (44%). The items identified were grouped into seven personalisation factors: physical, psychosocial, airway clearance technique (ACT) type, procedures, dosage, response and provider. As only two divergent models of ACT personalisation were found, the personalisation factors identified were then used to develop a model for physiotherapists.

Conclusions

The personalisation of airway clearance regimens is widely discussed in the current literature, which provides a range of factors that should be considered. This review summarises the current literature, organising findings into a proposed airway clearance personalisation model, to provide clarity in this field.

Real-world clinical outcomes and cost estimates of metastatic castration-resistant prostate cancer treatment: does sequencing of taxanes and androgen receptor-targeted agents matter?

INTRODUCTION

Health economic outcomes of real-world treatment sequencing of androgen receptor-targeted agents (ARTA) and docetaxel (DOC) remain unclear.

MATERIAL AND METHODS

Data from the electronic Castration-resistant Prostate cancer Australian Database (ePAD) were analyzed including median overall survival (mOS) and median time-to-treatment failure (mTTF). Mean total costs (mTC) and incremental cost-effectiveness ratios (ICER) of treatment sequences were estimated using the average sample method and Zhao and Tian estimator.

RESULTS

Of 752 men, 441 received ARTA, 194 DOC, and 175 both sequentially. Of participants treated with both, first-line DOC followed by ARTA was the more common sequence (n = 125, 71%). mOS for first-line ARTA was 8.38 years (95% CI: 3.48, not-estimated) vs. 3.29 years (95% CI: 2.92, 4.02) for DOC. mTTF was 15.7 months (95% CI: 14.2, 23.7) for the ARTA-DOC sequence and 18.2 months (95% CI: 16.2, 23.2) for DOC-ARTA. In first-line, ARTA cost an additional $13,244 per mTTF month compared to DOC. In second-line, ARTA cost $6726 per mTTF month. The DOC-ARTA sequence saved $2139 per mTTF compared to ARTA-DOC, though not statistically significant.

CONCLUSION

ICERs show ARTA had improved clinical benefit compared to DOC but at higher cost. There were no significant cost differences between combined sequences.

Doubly robust estimation of optimal dynamic treatment regimes with multicategory treatments and survival outcomes

Patients with chronic diseases, such as cancer or epilepsy, are often followed through multiple stages of clinical interventions. Dynamic treatment regimes (DTRs) are sequences of decision rules that assign treatments at each stage based on measured covariates for each patient. A DTR is said to be optimal if the expectation of the desirable clinical benefit reaches a maximum when applied to a population. When there are three or more options for treatments at each decision point and the clinical outcome of interest is a time-to-event variable, estimating an optimal DTR can be complicated. We propose a doubly robust method to estimate optimal DTRs with multicategory treatments and survival outcomes. A novel blip function is defined to measure the difference in expected outcomes among treatments, and a doubly robust weighted least squares algorithm is designed for parameter estimation. Simulations using various weight functions and scenarios support the advantages of the proposed method in estimating optimal DTRs over existing approaches. We further illustrate the practical value of our method by applying it to data from the Standard and New Antiepileptic Drugs study. In this analysis, the proposed method supports the use of the new drug lamotrigine over the standard option carbamazepine. When the actual treatments match the estimated optimal treatments, survival outcomes tend to be better. The newly developed method provides a practical approach for clinicians that is not limited to cases of binary treatment options.

Initiating SGLT2 inhibitor therapy to improve renal outcomes for persons with diabetes eligible for an intensified glucose-lowering regimen: hypothetical intervention using parametric g-formula modeling

INTRODUCTION

Sodium-glucose cotransporter 2 (SGLT2) inhibitors are now recommended in guidelines for persons with type 2 diabetes mellitus (T2DM) and at risk of advanced kidney disease as part of the glucose-lowering regimen.

RESEARCH DESIGN AND METHODS

To explore the optimal threshold at which to initiate SGLT2 inhibitor therapy, we conducted an observational study analyzed under a counterfactual framework. This study used the electronic healthcare database in Japan, comprising data from approximately 20 million patients at approximately 160 medical institutions. Persons with T2DM with an estimated glomerular filtration rate (eGFR) ≥ 30 mL/min/1.73 m² in April 2014 were eligible. The primary end point was the composite of renal deterioration (>40% decline in eGFR) and the development of eGFR<30 mL/min/1.73 m². We estimated the risk of the composite end point occurring over 77 months in different scenarios, such as early or delayed intervention with SGLT2 inhibitors for uncontrolled diabetes at different hemoglobin A1c (HbA_1c) thresholds. The parametric g-formula was used to estimate the risk of the composite end point, adjusting for time-fixed and time-varying confounders.

RESULTS

We analyzed data from 36 237 persons (149 346 person-years observation), of whom 4679 started SGLT2 inhibitor therapy (9470 person-years observation). Overall, initiating SGLT2 inhibitor therapy was associated with a 77-month risk reduction in the end point by 1.3-3.7%. The largest risk reduction was observed within 3 months of initiation once the HbA_1c level exceeded 6.5% (risk reduction of 3.7% (95% CI 1.6% to 6.7%)) compared with a threshold of 7.0% or higher.

CONCLUSIONS

Our analyses favored early intervention with SGLT2 inhibitors to reduce the renal end point, even for persons with moderately controlled HbA_1c levels. Our findings also suggest caution against clinical inertia in the care of diabetes.

Pharmacological treatment patterns in patients with juvenile idiopathic arthritis in the Netherlands: a real-world data analysis

OBJECTIVE

To investigate medication prescription patterns among children with JIA, including duration, sequence and reasons for medication discontinuation.

METHODS

This study is a single-centre, retrospective analysis of prospective data from the electronic medical records of JIA patients receiving systemic therapy aged 0-18 years between 1 April 2011 and 31 March 2019. Patient characteristics (age, gender, JIA subtype) and medication prescriptions were extracted and analysed using descriptive statistics, Sankey diagrams and Kaplan-Meier survival methods.

RESULTS

Over a median of 4.2 years follow-up, the 20 different medicines analysed were prescribed as monotherapy (n = 15) or combination therapy (n = 48 unique combinations) among 236 patients. In non-systemic JIA, synthetic DMARDs were prescribed to almost all patients (99.5%), and always included MTX. In contrast, 43.9% of non-systemic JIA patients received a biologic DMARD (mostly adalimumab or etanercept), ranging from 30.9% for oligoarticular persistent ANA-positive JIA, to 90.9% for polyarticular RF-positive JIA. Among systemic JIA, 91.7% received a biologic DMARD (always including anakinra). When analysing medication prescriptions according to their class, 32.6% involved combination therapy. In 56.8% of patients, subsequent treatment lines were initiated after unsuccessful first-line treatment, resulting in 68 unique sequences. Remission was the most common reason for DMARD discontinuation (44.7%), followed by adverse events (28.9%) and ineffectiveness (22.1%).

CONCLUSION

This paper reveals the complexity of pharmacological treatment in JIA, as indicated by: the variety of mono- and combination therapies prescribed, substantial variation in medication prescriptions between subtypes, most patients receiving two or more treatment lines, and the large number of unique treatment sequences.