Product Hack: Patient Base

How we hacked Patientbase, an “Excel killer” for the NHS and what we learned from it.

The Problem

Within the NHS there are lots of “legacy digital systems”. They are generally important, broadly useful and often large enough to have their own ecosystem of processes that ensure they are hard to remove. One such system is Excel, which provides the basis for a huge amount of data processing and analysis for clinician and managers alike.

Excel is great, it’s flexible and powerful enough to even create pseudo-products from. However there are areas where it is not great, such as version control, speed and simple UX. Over our time working with the NHS we saw hundreds of “we use this spreadsheet for…” scenarios. Unfortunately the pattern that emerged was:

• Lack of data validation
• Lack of versioning, user table, access logs or audit trail
• Potential crashes and data-loss from shared access
• Slow load times, with limited archiving
• Horrible user interfaces, that created errors

The Solution

We rapidly developed a wireframe prototype using Bootstrap, JSON data and some simple JavaScript libraries. The prototype was designed to offer clear value over Excel.

In particular:

• A clear and mobile responsive front end interface, with a modular design to allow for iterative refinement
• Controlled access, with an audit trail for each users input, automated version control and back ups
• Validation of all data input, through a reusable and systematic set of input logic
• An extensible data schema, that could be aligned with standards such as SnoMed or FHIR
• A customisable “reporting” layer, that allowed for completely bespoke algorithmic analysis
• An excel export feature for specific data sets, so further analysis could be covered with familiar tools

What we Learned

• We can quickly identify and generate solutions for problems experienced by clinicians. Which is a really rewarding experience. However…
• It’s really tough to move people away from familiar and embedded systems at both an organisational and individual level. Getting the budget released, engaging the team, filling the role that something like Excel fills, is all hard and required more resource than we had to spend.
• There is also a degree of resistance within the NHS to “outside help”. This manifests as fear of commercial models, persistance and repeated piloting, resistance to any data reaching “the cloud”.
• Because of this, there are very few pathways for solutions that don’t require scaled sales/marketing investment. This is the same for clinicians operating internally, as well as companies like us, externally.