Caprini Risk Assessment Model (Caprini RAM)

Inspiration

We started from a very simple reality: venous thromboembolism is a major problem, but it is often recognized too late. VTE is not rare, not minor, and not something that happens only in “unlucky” cases. In the United States alone, it is estimated to affect around 900,000 people each year and cause roughly 60,000 to 100,000 deaths annually. The CDC also identifies it as one of the leading preventable causes of hospital death.

In the preoperative setting, this risk does not always present itself clearly, yet missing it can lead to serious consequences. The problem in real clinical practice is not that physicians are unaware of how dangerous VTE can be. The problem is that they have to review too many risk factors in too little time. When this process is done entirely by hand, inconsistency and omission are almost inevitable.

We wanted to build a tool that is simple enough to use immediately, clear enough to understand at a glance, and structured enough to make risk assessment a truly useful part of preoperative workflow.

What it does

Caprini Pre-Op Assistant is a web-based tool that helps clinicians assess preoperative venous thromboembolism risk using the Caprini Risk Assessment Model.

The application allows clinicians to enter case information, automatically calculate BMI, automatically link age and BMI into the score, and then review a Caprini checklist organized by point group. As risk factors are selected, the system updates the score instantly and generates a results screen showing the total Caprini score, risk category, and structured decision support based on clinical context, including bleeding risk, renal function, surgical type, and the availability of mechanical prophylaxis.

What matters most is that this is not just a score calculator. It organizes information in a way that is actually useful in practice: clinicians can immediately see which risk group the patient falls into, which factors are driving the score, and what kind of prophylaxis discussion should follow. The app also supports exporting results as PDF, Word, or image files, making it easier to document and use within real preoperative workflows.

How we built it

We designed the product as a mobile-first experience because in real clinical settings, physicians need something they can use quickly on a phone or tablet, not a heavy system that slows them down.

The workflow was reduced to a few clear steps: choose a language, enter case information, complete the clinical decision-support fields, review the Caprini checklist, and see the result.

From a logic standpoint, we separated the system into two layers. The first layer handles Caprini scoring, where risk factors are translated into a total score. The second layer handles prophylaxis decision support, where the system looks beyond the score itself and considers bleeding risk, renal function, procedure type, heparin contraindications, and mechanical prophylaxis to make the results screen more clinically meaningful.

In terms of development speed, MeDo made a real difference for us. It helped us move from idea to working product much faster. Instead of spending too much time building the basic app structure from scratch, we were able to create an early working version, test the flow, identify what felt unclear, and iterate quickly. For a project like this, where the interface and clinical logic both needed repeated refinement, that speed was extremely valuable. It allowed us to spend more time on what actually mattered: clinical logic, user experience, and clarity of output.

Challenges we ran into

The biggest challenge was finding the balance between a simple interface and the real complexity of clinical decision-making.

Scoring with Caprini is only one part of the problem. Moving from a score to a meaningful prophylaxis recommendation requires additional context, such as bleeding risk, renal function, cancer surgery, orthopedic surgery, or neurosurgical considerations. In other words, the hardest part was not generating a number. The hardest part was making sure that number was placed in the right context.

Another major challenge was writing content that was clear enough for fast clinical use without making the application feel overly rigid or misleadingly automated. We had to repeatedly revise screen order, question grouping, the automatic handling of age and BMI, and the wording of the results so that the most important message would be obvious immediately.

In a medical product, even a small wording problem can increase the risk of misunderstanding. That is why we went through many iterations before arriving at a structure that felt both usable and clinically responsible.

Accomplishments that we're proud of

What we are most proud of is that we turned a tool that could easily feel dry and mechanical into an experience that is clear, coherent, and much closer to real clinical workflow.

The app does not simply calculate a score. It presents the result in a way clinicians can use immediately: total score, risk category, selected factors, and the next layer of clinical support all appear in one place.

We are also proud that we built the product in a way that remains flexible for future expansion, rather than creating a rigid one-off version that would be difficult to adapt later. And from a product-building perspective, we are proud that we used MeDo in the way it is most powerful: not just to build quickly, but to iterate quickly. In a hackathon setting, that is not just convenient. It is a real competitive advantage.

Most importantly, the project has kept its human purpose at the center: recognizing risk earlier so that patients can be protected before complications happen.

What we learned

We learned that in healthcare technology, a good interface is not just about aesthetics. It has to reduce cognitive error.

A line of text, a button label, or the way questions are grouped can all affect how users interpret a situation and make decisions. A tool can be logically correct and still be hard to use in real life.

We also learned that risk models like Caprini become truly valuable only when they are embedded into a workflow that fits clinical reality. If a tool makes clinicians slower, more confused, or more hesitant, then even a correct model loses practical value. But if it helps them recognize risk faster, more consistently, and with fewer omissions, then the model becomes genuinely useful in practice.

We also learned that the right development platform can change the pace of the entire project. With MeDo, we were able to spend more time refining clinical logic, user flow, and output clarity instead of being dragged down by the basic mechanics of app construction. For a project that required constant iteration, that made a major difference.

What's next for Caprini Risk Assessment Model (Caprini RAM)

The next step is to turn the application into a more deployment-ready clinical tool.

We want to refine the configuration layer so that healthcare institutions can adapt displayed content, prophylaxis logic, and export templates to fit their own workflows. That matters because in real practice, the same risk model may still need to be implemented differently depending on the clinical setting.

After that, we want to expand multilingual support, improve usability across more device environments, and move toward integration with electronic medical record workflows where appropriate.

In the long term, our goal is not just to digitize the Caprini score. Our goal is to turn it into an early protective layer that helps clinicians recognize risk faster, act more consistently, and intervene before complications occur.

Built With

• medo