CovidDB

International Covid-19 Database

Comment 3

Inspiration

We, the team of ClarData, are determined to make a contribution to the fight against the global threat of the Covid-19 pandemic. Managing a crisis is not all about start-ups but also about well established medical IT companies. We felt, that it is our duty to use our experience in this fight.

What it does

With all our experience and know-how and in cooperation with participating clinics, we developed an application that provides highly relevant clinical data. Our application CovidDB (www.covid-db.com) aims to give clinics a simple and productive way to document data of Covid-19 patients. The Covid-19 database primarily documents the stationary care of patients, an extension for the outpatients could of course also be implemented. The central data are risk factors, performed therapies, types of ventilations and the clinical and laboratory course. At the push of a button the application provides meaningful real time statistics regarding relevant parameters, such as length of hospital stay or clinical/laboratory evolution. Benchmarking statistics allow each clinic to compare itself with the parameters from all clinics. Similar to the tumour boards from oncology, the application facilitates case presentations with other experienced doctors.

The benefits of the application and the outputs generated from it are not only evident for the clinics but it is also of great benefit for science, research and politics to understand the course of the disease and quickly identify the positive experiences in treatment in order to update guidelines for the therapy and use of medication. Thus, adding clinical studies from other scientists is also possible.

How we built it

As we have 15 years of experience in creating applications for oncology, we took one of our applications and have transformed it into a documentation and reporting system for Covid patients.

As I am also a specialist doctor for anesthesia and intensive care, it was easy for me to define the fields which are needed for documentation.

Technical notes The application architecture is a classic n-tier structure. A repository, built around Entity Framework, handles the data layer. The magic happens in the business logic layer. This layer takes care of all computations, data aggregations, statistics and exports. The external communication interface is provided by a Web API2 layer. The presentation layer is build with angular 8, and Devextreme components.

Challenges we ran into

The most important challenge we ran into was that actually hospitals are not willing to document, they only want results. Unfortunately, results and evidence based medicine are not possible without documentation.

Technical notes Going with a dynamic form, this choosing configuration over code, had advantages and disadvantages. It allowed us to quickly add new fields or modify existing fields on the fly. But it has the disadvantage of increasing code complexity and decreasing the UI performance. We had to find a good balance between configuration and code and also take into account the DOM overloading. As the number of fields increased we had to resort to different optimization features to decrease the page components load, the DB calls and to reduce the amount of data loaded brought from the back end.

Also, handling the data for 100.000 patients might constitute a performance challenge, especially for statistics, where calculations cannot be handled dynamically in every case. But a caching system and some query and DB optimizations should take care of this.

Accomplishments that we're proud of

We are very proud that by gathering all our forces, we managed to have a fully functioning application only in 2 weeks.

Technical notes We’re happy that we build a configurable documentation system. Thus, we can provide flexible and extendable documentation and easy maintenance for the required data fields.

What we learned

We’ve learned that for this application to be useful, it needed to be build together with doctors and specialists, practically the people who are going to use this in their day to day work. We learned what they needed, how to communicate with them and how to integrate them in our team. And based on their feedback we are better at being flexible and at being fast.

The problem our project solves

Due to our 15 years of experience in gathering oncology data for 200.000 new patients a year, we know how to collect structured and validated information from thousands of different sources.

Anybody can make a database, but filling it with meaningful input needs certain skills. What we do is to get a structure in the unstructured wilderness of the individual hospital documentation, so that the outcome of different hospitals is comparable. The primary goal of the application is to quickly find a treatment for Covid patients.

The solution we bring to the table (including technical details, architecture, tools used)

What we offer is a novelty because we:

  • Gather information directly into our database or by import from other sources with same data set, as we do not want to monopolise the data input methods
  • Include new clinical trials for hospitals without data platforms
  • Managed a modularity in the documentation
  • We have a real time statistic analysis and maps, for the center alone or in comparison with all others
  • We export patient timelines, case presentations and release documents
  • We have a learning platform in realtime
  • We manage anonymous consultations with experienced doctors
  • We gather follow-up data after release by smartphone apps
  • We make research and predictions on quality big data

Technical notes The most important part of the application is represented by the documentation form. This is implemented with the help of a dynamic form, meaning every field is built from a series of configurable properties and attributes. Given the high number of fields needed, each field value change are saved after each input.

What we have done during the weekend

During the weekend we have worked on different types of exports from our application, styling of real time charts and on a new feature: the audit trail. This feature is required for clinical studies always to know which documentation changes were done by who and when. We also worked on the pitch video and made some new contacts in slack who can help us recruit more clinics and establish new clinical trials.

The solution’s impact to the crisis

The world needs now quality data and comparable data, not just any type of data.

What's next for CovidDB

We already have recruited many clinics in Germany and Romania, now we are expanding more internationally. As our CovidDB solution has not only future research objectives, but is also a real time database and learning platform, we imperatively need the collaboration of many hospitals as soon as possible. We cannot deal anymore with press articles of doctors, each saying that they have the cure or that they have the understanding of what is happening with Covid, without having the proof of those statements.

Built With

+ 12 more
Share this project:

Updates