In Northern Kenya, a pregnant mother is 100 times more likely to die of untreated complications than if she was here in Canada. She has access to at most 4 visits to a doctor before delivering her baby, as compared to 29 visits if she was in Canada. Further, she is more likely to give birth at home than in the hospital, increasing the risks of birth complications such as haemorrhage, hypertension, or sepsis.
We wanted to build a system that would bridge the gap in number of visits available to expectant and new mothers, by providing them with a web platform that effectively monitors their health via SMS. It is a proactive system, that probes users with questions to monitor their health, instead of waiting on patients (that already face barriers in regularly visiting the clinic) to seek medical help/advise in person.
What it does
This SMS/web app provides a monitoring and diagnostic tool available to patients remotely. Through consistent probing, the system monitors health of patient via their response to questions. The answers are parsed into 3 categories: no-risk, low-medium risk, and high-risk conditions.
For no-risk response, an automated response is generated confirming the absence of risk, and informing them of a customized reminder/tip based on their condition.
For low-medium response, the condition is monitored for a few days via appropriate follow-up questions to determine whether condition is improving or deteriorating. In the latter case, an "urgent alert" is triggered and a doctor is notified to take appropriate action. For high-risk condition, an "urgent alert" is triggered and doctor is notified to take necessary action. An appropriate text message is generated informing user of further instructions.
Additionally, the system provides powerful reporting and data logging tools. The doctor is able to see communication history between patient and system, providing him/her with data to make more effective and efficient diagnosis and treatment, as necessary.
How I built it
Built on node.js, using Twilio API for SMS communication. MongoDB is used for database, and jade for the front-end.
Challenges I ran into
Keeping track of who ate the caffeinated chocolates. On a more serious note, some of use were challenged with learning curve of learning node, jade and mongo. Twilio gave us issues here and there, and a bit of getting used to the mongoDB.
Accomplishments that I'm proud of
It works! Proud to see it come together in its various components, with the team collaborating with their expertise to provide different functionalities. My team members are proud of themselves for being great teachers (for teaching me a lot, that is!). And we had fun along the way after all, while accomplishing our vision.
What I learned
How to structure databases, how to build a web app that provides significantly more functionalities than I have ever worked on. (It was personally my first time working in node and jade, so that was a steep learning curve). We were able to dive deeper into jade, node and mongoDB, broadening our skill set along the way.
What's next for Mommy Monitor
Incorporating this app with the IBM Watson Developer tools, such as Watson for Oncology, will provide this system with more powerful diagnostic and treatment capabilities. We will also be able to expand to other types of patients with various types of conditions/situations, and not pregnant/new mothers. We also hope to provide access to our database and diagnostic tools to non-registered patients without access to a clinic. This is possible by a registration system via sms.