Designing risk prediction models for ambulatory no-shows across different specialties and clinics

Abstract

ObjectiveAs available data increases, so does the opportunity to develop risk scores on more refined patient populations. In this paper we assessed the ability to derive a risk score for a patient no-showing to a clinic visit.MethodsUsing data from 2 264 235 outpatient appointments we assessed the performance of models built across 14 different specialties and 55 clinics. We used regularized logistic regression models to fit and assess models built on the health system, specialty, and clinic levels. We evaluated fits based on their discrimination and calibration.ResultsOverall, the results suggest that a relatively robust risk score for patient no-shows could be derived with an average C-statistic of 0.83 across clinic level models and strong calibration. Moreover, the clinic specific models, even with lower training set sizes, often performed better than the more general models. Examination of the individual models showed that risk factors had different degrees of predictability across the different specialties. Implementation of optimal modeling strategies would lead to capturing an additional 4819 no-shows per-year.ConclusionOverall, this work highlights both the opportunity for and the importance of leveraging the available electronic health record data to develop more refined risk models.

Modeling the No-Show of Patients to Exam Appointments of Computed Tomography

“…For that, it imposes a penalty on the model β coe cients to induce dispersion, such that coe cients of less relevant predictors are reduced to zero [17]. This process results in biased coe cient estimates that can no longer be interpreted as odds ratios, as in a logistic regression, but rather as a weight or relative importance of the predictor variable [18]. The best value for the penalty parameter λ is the one that leads to the smallest mean square error (MSE) in the cross validation.…”

“…No-show rates may vary from 3-80% in healthcare systems, depending on the service being provided and patients' demographic characteristics [4]. Studies report different no-show rates; e.g., 18.80% in a study carried out in a medical center offering consultation on 10 different ambulatory specialties [4], 21.90% in bariatric surgery appointments in a specialized clinic -considering only consultations before and after surgery [2], and 6.50% in a radiology department considering all exams performed [3]. In the Brazilian public health network, patient no-show has become a chronic problem, similar to what is observed in other countries.…”

A metaheuristic-based stacking model for predicting the risk of patient no-show and late cancellation for neurology appointments

“…They also proposed a stacking model by using logistic regression as the meta-classifiers and three algorithms of random forests, artificial neural networks, and stochastic gradient boosting as the base-learners. Ding et al (2018) used regularized logistic regression to build predictive models using datasets from different specialties and clinics. They concluded that, in almost all specialties, the appointments rescheduled by the provider were the most important predictor of patient no-shows.…”

“…Kalb et al (2012) and Mohammadi et al (2018) considered a late cancellation as a no-show if the patient canceled the appointment within 24 hours before the appointment date. Ding et al (2018) considered cancellations on the day of the appointment as no-shows. Topuz et al (2018) and Srinivas and Ravindran (2018) considered a late cancellation as a missed appointment if the patient canceled the appointment within eight hours and 72 hours before the appointment date, respectively.…”

“…There is not a unique definition for a late cancellation, and every clinic and department may have different approaches to consider a canceled appointment as a late cancellation. Primarily, the definition of late cancellation varies based on the ease of rescheduling the appointments (Ding et al, 2018). In this article, we will define late cancellation as a cancellation notified by the patient within seven days of the appointment.…”