Introduction

Despite 1.75 million cases of COVID-19 in the US and more than 60% of states saying they need emergency health care support that need, 1.4 million health care workers lost their job in April. U.S. hospitals and health systems are predicted to take a $200 billion hit by June. Since labor costs account for a large part operating expenses, hospitals and health systems have to layoff healthcare workers. With fewer providers to care for sicker patients, health care workers are demonstrating high rates of burnout. Burnout in turn has been associated increased medical errors and worse patient outcomes.

There is also a significant disparity of available staff and staff needs. Illustratively, in New York, although it has the second largest number of physicians across the USA, there is a shortage of providers in areas with large populations, with a distribution of staff not aligned with the patient population.

Purpose & Motivation

DeploySmart is focusing on deploying staff smartly to provide the best care during COVID-19 and beyond.

How this application works

DeploySmart benefits hospital systems by streamlining shifts, health care workers by simplifying scheduling, and ultimately patients by ensuring adequate staff to care for them, therefore saving lives. Our platform provides a predictive analytics-based decision making tool for hospital providers, to decide whether they need to source staffing or can outsource their own staffing to units or hospitals in need. The hospital needs are analyzed based on the COVID-19 condition & the hospital capacity.

Our machine learning algorithm works by first identifying hospitals within the same system that needs additional staffing. We deploy staff by calculating a utility score which is based on specialty, shift preference, language proficiency, and individual profiles.

There are 2 core use cases - allowing hospital administrators and CMOs at command centers to view and dynamically allocate staff, with our platform collecting the data together, applying our algorithm and providing a view to the administrators; and to automatically allocate staff based on administrators' preferences. This allows for adaptive resource allocation, learning as the parameters change and updating to provide the best allocation. Health care workers will then receive push notifications to their device if they are redeployed to another area based on their indicated preferences.

Based on our calculation, automation of scheduiling saves hospitals over 20% for scheduling per year, $3.600 per year in saving per staff and increase healthcare quality since it allows additional 200 hours time per staff for patient care.

How this application was developed

We are building the application using Angular and Javascript. Database was developed using SQL. We will be developing the ML algorithm. Currently, our prototype is publicly available in Figma.

Difficulties & Challenges you faced

Our challenge is to get representative data for our use cases, as hospital data is not widely available, Currently, we are using county data to show how our model works.

Go-to Market Strategy

The app will follow tiered-subscription model. We estimate to break-even operational cost in early Year 2 of the business. Our marketing channel would be industry conferences and social media.

In the future, DeploySmart is not only a solution for the health care sector as we move into the new normal way of working but for a wide range industries. Same as in the hospital sector it takes training to move people between insitution, therefore DeploySmart will provide a diversification and offer training to instantly place staffing across institutions.

Built With

Share this project:

Updates