Experimental studies on performance of ventilators stored in the Strategic National Stockpile

Contributing to health system resilience during pandemics via purchasing and supply strategies: an exploratory system dynamics approach

BACKGROUND

Health systems worldwide struggled to obtain sufficient personal protective equipment (PPE) and ventilators during the COVID-19 pandemic due to global supply chain disruptions. Our study's aim was to create a proof-of-concept model that would simulate the effects of supply strategies under various scenarios, to ultimately help decision-makers decide on alternative supply strategies for future similar health system related crises.

METHODS

We developed a system dynamics model that linked a disease transmission model structure (susceptible, exposed, infectious, recovered (SEIR)) with a model for the availability of critical supplies in hospitals; thereby connecting care demand (patients' critical care in hospitals), with care supply (available critical equipment and supplies). To inform the model structure, we used data on critical decisions and events taking place surrounding purchase, supply, and availability of PPE and ventilators during the first phase of the COVID-19 pandemic within the English national health system. We used exploratory modelling and analysis to assess the effects of uncertainties on different supply strategies in the English health system under different scenarios. Strategies analysed were: (i) purchasing from the world market or (ii) through direct tender, (iii) stockpiling, (iv) domestic production, (v) supporting innovative supply strategies, or (vi) loaning ventilators from the private sector.

RESULTS

We found through our exploratory analysis that a long-lasting shortage in PPE and ventilators is likely to be apparent in various scenarios. When considering the worst-case scenario, our proof-of-concept model shows that purchasing PPE and ventilators from the world market or through direct tender have the greatest influence on reducing supply shortages, compared to producing domestically or through supporting innovative supply strategies. However, these supply strategies are affected most by delays in their shipment time or set-up.

CONCLUSION

We demonstrated that using a system dynamics and exploratory modelling approach can be helpful in identifying the purchasing and supply chain strategies that contribute to the preparedness and responsiveness of health systems during crises. Our results suggest that to improve health systems' resilience during pandemics or similar resource-constrained situations, purchasing and supply chain decision-makers can develop crisis frameworks that propose a plan of action and consequently accelerate and improve procurement processes and other governance processes during health-related crises; implement diverse supplier frameworks; and (re)consider stockpiling. This proof-of-concept model demonstrates the importance of including critical supply chain strategies as part of the preparedness and response activities to contribute to health system resilience.

Mortality prediction model for the triage of COVID-19, pneumonia, and mechanically ventilated ICU patients: A retrospective study

Rationale

Prediction of patients at risk for mortality can help triage patients and assist in resource allocation.

Objectives

Develop and evaluate a machine learning-based algorithm which accurately predicts mortality in COVID-19, pneumonia, and mechanically ventilated patients.

Methods

Retrospective study of 53,001 total ICU patients, including 9166 patients with pneumonia and 25,895 mechanically ventilated patients, performed on the MIMIC dataset. An additional retrospective analysis was performed on a community hospital dataset containing 114 patients positive for SARS-COV-2 by PCR test. The outcome of interest was in-hospital patient mortality.

Results

When trained and tested on the MIMIC dataset, the XGBoost predictor obtained area under the receiver operating characteristic (AUROC) values of 0.82, 0.81, 0.77, and 0.75 for mortality prediction on mechanically ventilated patients at 12-, 24-, 48-, and 72- hour windows, respectively, and AUROCs of 0.87, 0.78, 0.77, and 0.734 for mortality prediction on pneumonia patients at 12-, 24-, 48-, and 72- hour windows, respectively. The predictor outperformed the qSOFA, MEWS and CURB-65 risk scores at all prediction windows. When tested on the community hospital dataset, the predictor obtained AUROCs of 0.91, 0.90, 0.86, and 0.87 for mortality prediction on COVID-19 patients at 12-, 24-, 48-, and 72- hour windows, respectively, outperforming the qSOFA, MEWS and CURB-65 risk scores at all prediction windows.

Conclusions

This machine learning-based algorithm is a useful predictive tool for anticipating patient mortality at clinically useful timepoints, and is capable of accurate mortality prediction for mechanically ventilated patients as well as those diagnosed with pneumonia and COVID-19.

•

Mortality predictions have not previously been evaluated for COVID-19 patients.

•

Machine learning may be a useful predictive tool for anticipating patient mortality.

•

Prediction can be estimated at clinically useful windows up to 72 h in advance.