Guardian Health

It is an integrated booking system to schedule clinic appointments and home visits for community health workers(CHW's) using optimisation techniques and it uses machine learning to refer patients.

Comment

Inspiration

While doing her masters research in data science focusing on health issues, Tsitsi found out that doctors are usually able to tell whether or not a fetus might have an abnormality only after the second scan which happens after 22 weeks. It is at this point that an expecting mother gets options including termination in certain cases. This can be very emotionally exhausting and also leaves less room to mitigate risk. This realisation coupled with Tino's dream to go back to medical school and his passion for medicine gave birth to Guardian Health. With the both us well versed in technology and passionate about health, we started on our journey to transform the health system for the better with solutions that benefit every citizen regardless of social status.

The first thing that we did is we built a system that can help health professionals with the early detection of fetal abnormalities and gives the risk of an expecting mother to get a caesarean. Our solution looked great, however we realized that having such a solution without solving some of the problems the health sector is facing might make the adoption a little difficult. After doing a lot of research, we found out that the public health system in South Africa is very overloaded as most people cannot afford to go to private hospitals and clinics. Currently, patients just go to the clinic on any day without surety that the service they need is provided at the clinic. The numbers are also very unpredictable for the clinics, they could get 10 times the number they can service a day putting them under pressure and also results in patients being turned away after spending the whole day at the clinic. This also results in patients choosing to go directly to the hospital even though their condition is suitable for a clinic resulting in overloading of hospitals. In fact, the department of health recently reported that new born babies were getting infected in wards due to overloading of rooms and also lack of resources.

We decided to create a booking system for the clinics and community health workers where we provide better resource planning for them. We use the first part of Guardian health which is the detection of fetal abnormalities and caesarean sections to refer patients to clinics and hospitals using data science in order to book patients optimally based on reason for visit, distance and load of clinic. This also solves the current problem where people prefer going to hospitals instead of following the current hierarchy of being referred to a hospital from a clinic. It makes community health workers work a lot more efficient by providing them with their appointments for the day and the optimal routes they can take to achieve their goals.

What it does

Our solution is a system that connects patients, community health workers and clinics.

  1. Patients are able to book appointments at a clinic, and they get routed to the best clinic based on factors such as distance, day, time, reason for clinic visit and also take into consideration the expected number of patients.
  2. Patients are able to request home based visit by community health workers allowing CHWs to do proper planning by getting necessary help and resources before hand.
  3. Community health workers get scheduled automatically using optimisation methods. The schedule takes in mind their appointments and how they can arrive on time to all those appointments and further more travelling routes to take in order to avoid exhaustion by minimising the travel route.
  4. CHWs get a dashboard for their resource planning showing appointment allocation for everyone and optimal routes.
  5. An app for the CHWs to get information and referrals more efficient on the go.
  6. Hospitals get a dashboard that shows their referrals, they get to be able to mark themselves as fully booked to allow them not to get referrals if there are other options in a certain period of time.
  7. Clinics get a dashboard where they can check their workload for the day to make sure they have enough resources at hand in the different departments based on information such as reasons for visit in the app.
  8. Chatbot to allow people to chat instead of using the app, and also get information that clinics would normally give as pamphlets. Let's save a tree while at it. Our chatbot already allows for voice chatting which is good for people with eye impairments.
  9. Patients get referral letters automatically once they have been referred to a hospital and also get an sms with CHW details that will be visiting them once allocation is done for safety reasons.
  10. Patient risk factors such as high risk of getting a caesarean section are used to refer them to a hospital.
  11. Early detection of fetal abnormalities is performed also allowing for the referral of high risk patients to hospital.

How we built it

  1. We built a micro app for the booking of clinics and home visits using Ayoba.
  2. We used optimisation methods for the allocation of patients to clinics and for scheduling optimal routes for community health care workers.
  3. We used mongoDB for our storing of data from the app and chatbot.
  4. We built dashboards for the clinics, CHWs and hospitals.
  5. For our chatbot, we used dialogflow and we also created a knowledge base in order to answer frequently asked questions using the bot e.g why should I vaccinate?
  6. We used machine learning algorithms for the predictions of caesarean risks and fetal abnormalities.

Challenges we ran into

Unfortunately, we were not able to integrate a native chatbot into Ayoba.

Accomplishments that we're proud of

  • Smart scheduling
  • Ability to bring together optimisation methods, natural language processing, analytics and AI into our solution.

What we learned

  • How to use the Ayoba app, how to integrate while we created our app.
  • Through research, we learnt how vulnerable the health sector is and how much of a difference solutions that can make it more efficient would make.

What's next for Guardian Health

  • Actively developing Guardian Health to launch into the private sector. -Continue reaching out to relevant stakeholders in the public sector so we can deploy our solution for the general public.

Built With

Share this project:

Updates