Dataset Information

Bayesian strategy selection identifies optimal solutions to complex problems using an example from GP prescribing.

ABSTRACT: Complex health problems require multi-strategy, multi-target interventions. We present a method that uses machine learning techniques to choose optimal interventions from a set of possible interventions within a case study aiming to increase General Practitioner (GP) discussions of physical activity (PA) with their patients. Interventions were developed based on a causal loop diagram with 26 GPs across 13 clinics in Geelong, Australia. GPs prioritised eight from more than 80 potential interventions to increase GP discussion of PA with patients. Following a 2-week baseline, a multi-arm bandit algorithm was used to assign optimal strategies to GP clinics with the target outcome being GP PA discussion rates. The algorithm was updated weekly and the process iterated until the more promising strategies emerged (a duration of seven weeks). The top three performing strategies were continued for 3 weeks to improve the power of the hypothesis test of effectiveness for each strategy compared to baseline. GPs recorded a total of 11,176 conversations about PA. GPs identified 15 factors affecting GP PA discussion rates with patients including GP skills and awareness, fragmentation of care and fear of adverse outcomes. The two most effective strategies were correctly identified within seven weeks of the algorithm-based assignment of strategies. These were clinic reception staff providing PA information to patients at check in and PA screening questionnaires completed in the waiting room. This study demonstrates an efficient way to test and identify optimal strategies from multiple possible solutions.

SUBMITTER: Allender S

PROVIDER: S-EPMC6971230 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Bayesian strategy selection identifies optimal solutions to complex problems using an example from GP prescribing.

Allender S S Hayward J J Gupta S S Sanigorski A A Rana S S Seward H H Jacobs S S Venkatesh S S

NPJ digital medicine 20200120

Complex health problems require multi-strategy, multi-target interventions. We present a method that uses machine learning techniques to choose optimal interventions from a set of possible interventions within a case study aiming to increase General Practitioner (GP) discussions of physical activity (PA) with their patients. Interventions were developed based on a causal loop diagram with 26 GPs across 13 clinics in Geelong, Australia. GPs prioritised eight from more than 80 potential interventi ...[more]

PMID: 31993505

Similar Datasets

Project description:BACKGROUND:Stroke causes death, disability and increases the use of healthcare resources worldwide. The outcome of intravenous thrombolysis and mechanical endovascular thrombectomy highly depends on the delay from symptom onset to initiation of definitive treatment. The purpose of this study was to compare the various patient transportation strategies to minimize pre-hospital delays. METHODS:Emergency medical services (EMS) mission locations and ambulance response times in Finland with urgent stroke-suspected dispatch codes were collected from Emergency Response Centre (ERC) records between 1 January 2016 and 31 December 2016. Four transport scenarios were simulated for each mission, comparing ground and helicopter transportation to hospital with different treatment capabilities. RESULTS:In 2016, a total of 20,513 urgent stroke-suspected missions occurred in Finland. Of these, we were able to locate and calculate a route to scenario-based hospitals in 98.7% (20,240) of the missions. For ground transport, the estimated median pre-hospital time to a thrombolysis-capable and thrombectomy-capable hospital were 54.5?min (95% confidence interval (CI), 31.7-111.4) and 94.4?min (95% CI, 33.3-195.8), respectively. Should patients be transported on the ground to thrombectomy-capable hospitals only, the pre-hospital time would increase in 11,003 (54.4%) of missions, most of which were in rural areas. With the fastest possible transportation method, the estimated mean transport time to a thrombectomy-capable hospital was 80.84?min (median, 80.80?min; 95% CI, 33.3-143.1). Helicopter transportation was the fastest method in 68.8% (13,921) of missions, and the time saved was greater than 30?min in 27.1% (5475) of missions. In rural areas, helicopter transportation was the fastest option in nearly all missions if dispatched simultaneously with ground ambulance. CONCLUSION:Helicopter transportation may significantly decrease pre-hospital delays for stroke patients, especially in rural areas, but the selection of an optimal transportation method or chain of methods should be determined case-by-case.

Project description:BackgroundHow foragers move across the landscape to search for resources and obtain energy is a central issue in ecology. Direct energetic quantification of animal movements allows for testing optimal foraging theory predictions which assumes that animals forage so as to maximise net energy gain. Thanks to biologging advances, we coupled instantaneous energy-budget models and behavioural mode analysis to test optimal foraging theory predictions on wandering albatross Diomedea exulans during the brooding period. Specifically, the instantaneous energy-budget model considered the energetic balance (i.e., the difference between empirical energy gain data and modelled energy expenditure via heart rate values) along the trajectory of a given individual. Four stereotypic instantaneous behavioural modes were identified based on trajectory properties (e.g., speed and turning angle) by applying a new algorithm called Expectation Maximization Binary Clustering. Previous studies on this species have shown that foraging-in-flight is the optimal foraging strategy during the incubation period when albatrosses undertake long-distance movements but no specific foraging strategy has been determined for shorter foraging movements (e.g., brooding period).ResultsThe output of our energy-budget model (measured as net energy gain) highlighted the potential optimality of alternative search strategies (e.g., sit-and-wait) during brooding, when birds may be subjected to specific energetic trade-offs and have to adapt their foraging strategies accordingly. However, not all birds showed this pattern, revealing the importance of considering individual variability in foraging strategies, as well as any switching among strategies, before drawing population-level generalizations. Finally, our study unveils the importance of considering fine scale activities to make realistic estimates of trip energy expenditure for flying birds at sea.ConclusionsThe up-scaling of accurately measured fine-scale energy patterns is essential to quantify energy balances, and their fluctuations by season of different activities among individuals or populations. In particular, we offer new insights for the energetic quantification of the effect of changing oceanic winds on the biology of pelagic predators in the southern oceans.

Dataset Information

Bayesian strategy selection identifies optimal solutions to complex problems using an example from GP prescribing.

Publications

Bayesian strategy selection identifies optimal solutions to complex problems using an example from GP prescribing.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets