Abstract
Failure to attend scheduled hospital appointments disrupts clinical management and consumes resource estimated at £1 billion annually in the United Kingdom National Health Service alone. Accurate stratification of absence risk can maximize the yield of preventative interventions. The wide multiplicity of potential causes, and the poor performance of systems based on simple, linear, low-dimensional models, suggests complex predictive models of attendance are needed. Here, we quantify the effect of using complex, non-linear, high-dimensional models enabled by machine learning. Models systematically varying in complexity based on logistic regression, support vector machines, random forests, AdaBoost, or gradient boosting machines were trained and evaluated on an unselected set of 22,318 consecutive scheduled magnetic resonance imaging appointments at two UCL hospitals. High-dimensional Gradient Boosting Machine-based models achieved the best performance reported in the literature, exhibiting an area under the receiver operating characteristic curve of 0.852 and average precision of 0.511. Optimal predictive performance required 81 variables. Simulations showed net potential benefit across a wide range of attendance characteristics, peaking at £3.15 per appointment at current prevalence and call efficiency. Optimal attendance prediction requires more complex models than have hitherto been applied in the field, reflecting the complex interplay of patient, environmental, and operational causal factors. Far from an exotic luxury, high-dimensional models based on machine learning are likely essential to optimal scheduling amongst other operational aspects of hospital care. High predictive performance is achievable with data from a single institution, obviating the need for aggregating large-scale sensitive data across governance boundaries.
Highlights
Failure to attend hospital appointments needlessly delays clinical care and consumes resource better spent on improving its quality.[1]
Of the eight studies to quantify out-of-sample attendance prediction performance identified in a systematic review of the literature, only three used non-linear models, and none included more than 49 variables (Table 1)
Studying a large sample of magnetic resonance imaging (MRI) appointments across two large UK hospitals, we sought to answer two related questions: what is the relationship between the complexity of predictive models of attendance and their predictive performance, and can sufficient predictive performance be achieved to render targeting costeffective? If complex models are convincingly shown to be required for optimal performance, a reorientation of hospital scheduling analytics to machine learning-based modelling would be indicated; if there is no difference between simple and complex approaches, other avenues for improving scheduling ought to be pursued
Summary
Failure to attend hospital appointments needlessly delays clinical care and consumes resource better spent on improving its quality.[1]. Chosen for their intelligibility and generalizability, are ill-suited to predicting individual events where the causal field is wide Combining machine learning with large-scale data allows us to create rich, complex, high-dimensional models able to operate within wider causal fields. If such models perform and generalize better than simpler variants their one defect—lack of easy intelligibility—is far outweighed. By capturing individual variability better, they may be used to infer systemic, modifiable hospital causes of non-attendance currently obscured by the many other factors in play Complex models both potentially enhance existing interventions and open the way to implementing categorially new ones. Capacity limitations distribute non-urgent initial secondary care appointments across a wide interval—18 weeks in the UK National Health Service (NHS)— where patients have varying freedom over the choice of an appointment slot; subsequent appointments are distributed even
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
