Med Caddy

Household - centered platform for chronic disease medication management. Making it easier for family members of all ages to participate and share the burden while also having a bit of fun!

Comment

The stress and burden associated with providing home care to elderly family members in multi-generational homes often goes unnoticed in society despite it being the most common living arrangement around the world. Historically, female household members have taken the bulk of the load of ensuring adequate support for the family member - including the critical task of administering and monitoring clinical drug adherence. Last year one of our team members, Kirusha and her family were in this situation. Kirusha's grandmother suffered a severe complication as a result of a chronic condition. Despite being in a separate household, hours away from her grandmother - she and her mum found themselves a part of the center of her grandmother's care. Where there was once calm and comfort there was a whirlwind of chaos - a hospital room at home - in midst covid. Pill bottles everwhere, long distance video calls trying to monitor and respond to the situation.

Hedgy Med Caddy came together through a mutual interest in improving the less than glamorous parts of patient care - home care. Family caregivers need support yet they are often the least positioned to get it - with limited training, competing life priorities and the grief associated with taking care of someone with aging with deteriorating health. The team came together at HopHacks -a collaboration between students from both JHU and Boston University. The team members shared a common vision of aiming to develop a project using a user centered design methodology - adopted from the medical device industry.

We quickly broke the ice despite only 2 of knowing one another prior to the start of the hackathon. Everyone was open minded and supportive while also direct and critical. We began in front of a chalk board in room 313 of Hodgson - where we were for the majority of the duration of the event.

We wanted to follow the general flow of medical device product development but in and accelerated form. From experience our team members knew that for a fully commercialized product specific application, product and regulatory standards would be needed - for instance IEC 62304 which covers medical device software lifecycle as well as SaMD (Software as Medical Device). These are standards that one team members works with on a day to day basis. Since the intent of the hackathon is to generate a solution, concepts and prototypes we focused on trying to align our methods to what we knew would be needed to commercialize but not focus on the strict adherence of product development standards.

We did however follow the general methodology that is used in the medical device industry. Starting with very high level pain point identification. Sine our team's selected topic was patient safety three areas were initially interesting to us:

1) Minimizing Errors during Shift Change of Clinical Staff at Hospitals 2) Improving drug adherence in vulnerable populations 3) Improving EMS communications with hospital staff 4) Enhanced scheduling of clinical staff

In order to better differentiate the potential for a HopHacks project in each of these areas we quickly tried to do some sensemaking to evaluate key factors such 1) compatibility with the issue and technology solutions that our team could feasibly develop and deploy 2) total addressable market (TAM) 3) Overall interest of team members.

After some insightful discussion - it became clear that the TAM of improving drug adherence in vulnerable populations was staggering. We also really thought that this was really an issue that is worth tacking as many people in the Baltimore Community are facing issues with chronic disease drug adherence - and we wanted to think global but starting with acting locally.

We then identified different patient types that could need support for chronic disease drug adherence. That is how stumbled upon the fact that a large part of chronic disease care happens at home and that it is the behavior and routine of the people at home with the patient could have the most positive impact on them taking their medication - the correct dosage at the correct time. This really struck a chord with us as we realized that while people often say "it takes a village to raise a child" they often don't think about the fact that as people age "it takes a "family" to age gracefully." We immediately realized in multigenerational households - people of all ages could be part of the caregiving and we wanted to examine what we could do to make it easier for kids, teens and adults alike to enhance patient health while cutting household stress and anxiety.

We moved to brainstorming the barriers and issues that patients with chronic disease have in their lives that could serve as a barrier to them taking their medication. This is in one of our blackboard mind maps. Some takeaways were: low vision, mistrust, side-effects, forgetfulness, busyness, embarassment, movement disorders.

We then moved onto identifying an exemplar user - which gave us some help with visualizing and ideation for the next steps of solution design and hazard analysis.

Given our team's background in med device - we knew it was important for us to determine both the system requirements not only thinking about the system but also from a hazard analysis and a DFMEA (design failure modes and effects analysis perspective). We started listing out possible hazardous situations that a software of hardware solution that aimed at improving drug adherence in vulnerable populations could create. We also drafted risk mitigations in the form of requirements (informal alignment to ISO 14971 - thinking about both severity and likelihood. This involved a lot of discussion and assumptions being generated.

When the dust settled we realized that we wanted to take a household focus and that each member of a household (including the patient) had a unique position in the care circle. We wanted to take a platform approach centered around trying to help build routine surrounding taking medication - the right dosage at the right time with positive feedback for compliance. We recognized that the patient themselves may suffer from different comorbidities that could impact their independence in administering their medication. We realized that other household members likely do not have the same issues and could fill in in different ways - with varying levels of medical knowledge, familiarity and confidence when it comes to supporting someone taking pills or injections. We figured that we could start with creating a web based solution to help household members verify they have the right dosage of the right meds at the right time using a user profile based prompting system. So we started there - and then we also realized that we could leverage drug databases and image processing algorithms to further bolster error checking. Separately we realized that we wanted to create a more positive patient experience for self administered medication by making a hardware station that tied into our software to further simplify the user experience and also serve storage and transport purposes for meds.

The resulting work is a combined hardware and software effort. The team has put together several demos:

  1. Generation of The Main User Workflow 1) Point Of Care Kiosk Version for those who will benefit from the medication organizer, scale and pill count and type verification - with supporting accessibility features 2) Super User Interface for those wanting quick dosage checking 3) Hardware prototype development and manufacturing for the MedCaddy Kiosk -designed entirely from scratch in Blendr on premises - queued for printing remotely at FFU but FFU was closed so we had to print at DMC 4) End user Notification and Reminder Engine using Twilio 5) Patient Registration portal including basic error checking of medication leveraging FDA drug database. 6) Preliminary Computer Vision Pipeline Development for pill identification

Major Challenges

  • Learning TaiPy
  • Time Constraint
  • Limited time to read over library documentation
  • Getting the Computer Vision Libraries to work in our development environment

We overcame these challenges by supporting one another, clear communication and expectation management. We opted to focus more heavily on showing different aspects of the solution instead of investing in perfection.

What's next for Med Caddy?

The Med Caddy team is very interested in continuing the project after the Hackathon is over. The team members truly believe in the utility that a more refined, integrated platform could look like and believe that this is achievable by the project team with their existing skills and knowledge. The team would like to make a more refined hardware prototype and complete the image processing pipeline for the pill detection functions.

The major items left in prototyping/concept phase

  • preliminary usability based on FDA guidance
  • development of representative user group to support refinement of system requirements and formative testing
  • prototype iteration
  • Formal system requirements, prototype hazard analysis and DFMEA
  • Regulatory strategy and External standards List development
  • Pricing and business value proposition

In the product development phase we will take the necessary steps to ensure that we use a 13485 certified contract manufacturer, comply with IEC 62304 for software lifcycle development 60601-1 3.2 edition for medical electrical safety. During V&V we willensure that our summative safety testing is done with a wide range of age group users.

Share this project:

Updates