Colorectal Cancer Risk Assessment Tool

The Client

A non-profit health organization focused on cancer prevention. The project was developed in collaboration with leading oncologists and based on years of clinical research into colorectal cancer risk factors.

The tool was later promoted through a nationwide campaign in partnership with major health funds, reaching a wide public audience.

The Problem

Colorectal cancer is one of the most treatable cancers – if caught early. The challenge is that most people at risk don’t know they’re at risk until symptoms appear. By then, treatment outcomes drop significantly.

Before this tool existed:

• Risk assessment required a visit to a specialist
• Most people simply never got screened
• Family history data – one of the strongest risk indicators – was rarely collected in a structured way
• Preventive care only started after the disease had already progressed

The organization wanted to change this. They needed a tool that anyone could use from home, for free, without identifying themselves – and get a medically accurate risk assessment in minutes.

What We Built

A multi-step digital questionnaire that calculates colorectal cancer risk based on personal health data and family medical history. The logic was defined by oncologists and contains roughly 2,700 possible decision nodes – meaning the path through the form adapts to each person’s specific answers.

How It Works

1. Personal health questions – Age, gender, lifestyle factors, and personal medical history. Each answer can change which questions come next.
2. Family history mapping – Detailed questions about cancer history in the family. The system maps relationships and conditions to calculate inherited risk factors.
3. Real-time risk calculation – As the user fills in data, calculations run in parallel. Some happen in the browser (JavaScript), others through a dedicated Salesforce object with formula fields, Flows, and Apex.
4. Dynamic summary screen – At the end, a personalized results page shows the calculated CRC risk coefficient and specific recommendations based on that score.
5. Document generation – Based on risk level, the system generates: a letter for the user’s doctor, personal recommendations, and (when indicated) referrals for genetic testing. All sent to the user’s email.

The Privacy Challenge

This was a public health tool. People needed to answer sensitive medical questions about themselves and their families. The organization made a clear decision: complete anonymity.

No user accounts. No names stored in Salesforce. No way to trace results back to an individual.

To make this work technically:

• All form submissions run through a single shared Titan user – no individual Salesforce records are created per person
• Risk calculations that involve personal data happen client-side in JavaScript, never leaving the browser
• Server-side calculations use derived values (the CRC risk coefficient), not raw personal data
• Generated documents are sent directly to the user’s email and not stored on the server

This approach kept the medical accuracy intact while giving users complete confidence that their data stays private.

Technical Architecture

The complexity here was in the decision logic, not the infrastructure. 2,700 possible paths through a single form means every conditional branch had to be mapped, tested, and validated against the oncologists’ clinical model.

Titan Side

• Multi-step form – Dynamic sections that show or hide based on previous answers. The stepper pattern kept the experience manageable despite the depth of the questionnaire.
• Client-side JavaScript – Specific calculations ran directly in the browser to maintain anonymity and reduce server round-trips. This also made the form feel fast and responsive.
• Dynamic summary page – The results screen adapts completely to the individual’s risk profile. Different risk levels trigger different content blocks, recommendations, and document types.

Salesforce Side

• Dedicated calculation object – A custom Salesforce object built specifically for risk computation, with formula fields that process the derived data points.
• Apex logic – Complex calculations that couldn’t be handled by formulas alone, running the oncologists’ risk model against the input data.
• Salesforce Flows – Orchestrating the calculation sequence and document generation triggers.

Results

Technology Stack

• Titan Forms – Multi-step questionnaire with conditional branching across 2,700 paths
• Titan Web – Dynamic results page with personalized content
• JavaScript (client-side) – Privacy-preserving calculations in the browser
• Salesforce custom object – Risk calculation engine with formula fields
• Salesforce Apex – Complex risk model computations
• Salesforce Flow – Calculation orchestration and document triggers
• Document generation – Doctor letters, patient recommendations, genetic testing referrals

What Made This Special

We sat with oncologists and translated years of clinical research into a digital decision tree. That’s not typical implementation work. It required understanding the medical logic deeply enough to build something that produces the same conclusions a specialist would reach – but accessible to anyone with a phone.

The external QA firm that tested the tool ran it through every possible path. Thousands of test cases. Every single one returned the correct result.

And the technical decisions mattered. Running sensitive calculations in the browser instead of on the server wasn’t just a performance choice – it was a privacy architecture decision that made the whole project possible. People filled in the questionnaire because they trusted that no one would see their answers.

This is probably the project we’re most proud of. Not because of the technology – but because of what it did. Early detection saves lives, and this tool made early detection available to everyone.