How many patients will need ventilators tomorrow?
Dai, J. G.; Gluzman, M.; Janmohamed, A.; Xu, Y.
Show abstract
This paper develops an algorithm to predict the number of Covid-19 patients who will start to use ventilators tomorrow. This algorithm is intended to be utilized by a large hospital or a group of coordinated hospitals at the end of each day (e.g. 8pm) when the current number of non-ventilated Covid-19 patients and the predicated number of Covid-19 admissions for tomorrow are available. The predicted number of new admissions can be replaced by the numbers of Covid-19 admissions in the previous d days (including today) for some integer d [≥] 1 when such data is available. In our simulation model that is calibrated with New York Citys Covid-19 data, our predictions have consistently provided reliable estimates of the number of the ventilatorstarts next day. This algorithm has been implemented through a web interface at covidvent.github.io, which is available for public usage. Utilizing this algorithm, our paper also suggests a ventilator ordering and returning policy. The policy will dictate at the end of each day how many ventilators should be ordered tonight from a central stockpile so that they will arrive by tomorrow morning and how many ventilators should be returned tomorrow morning to the central stockpile. In 100 runs of operating our ventilator order and return policy, no patients were denied of ventilation and there was no excessive inventory of ventilators kept at hospitals.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- A Bayesian Susceptible-Infectious-Hospitalized-Ventilated-Recovered Model to Predict Demand for COVID-19 Inpatient Care in a Large Healthcare System 93%
- Computational simulation to assess patient safety of uncompensated COVID-19 two-patient ventilator sharing using the Pulse Physiology Engine 93%
- Estimating the impact of interventions against COVID-19: from lockdown to vaccination 93%
Similar papers in this journal
- Reinforcement learning derived chemotherapeutic schedules for robust patient-specific therapy 93%
- Identifying low-dimensional trajectories of mechanically-ventilated patient systems: Empirical phenotypes of joint patient+care processes to enhance temporal analysis in ARDS research 93%
- Developing Machine Learning Models for Predicting Intensive Care Unit Resource Use During the COVID-19 Pandemic 93%
Similar papers in this journal
- Deep reinforcement learning framework for controlling infectious disease outbreaks in the context of multi-jurisdictions 93%
- A reinforcement learning model to inform optimal decision paths for HIV elimination 1 92%
- A Machine Learning Approach to Differentiate Between COVID-19 and Influenza Infection Using Synthetic Infection and Immune Response Data 92%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.