Integrations & data · Practice guide

Understand what an integration test can and cannot prove

Illustrative situation

A connection works in a simplified demonstration but fails when the receiving system applies its actual rules for accepting information. A demonstration environment may simplify rules that apply to the live service. Ask what the test setup represents so acceptance does not rely on conditions the real system never allows. Keep uncovered cases visible.

Healthcare booking calendars on two connected screens with blue and orange appointment cards.
Make the handover between tools clear for the team using them.

What this means for your practice

Ask the supplier what its test setup represents and which real situations it cannot reproduce. Record those limits in the agreed checks, rather than assuming a successful demonstration proves how the live service will behave.

An example to discuss with your team

Submit an incomplete fictional record that should be rejected by the receiving rules.

Healthcare data integration connecting two applications through a field-mapping board.
Agree what each field means and which system owns it.
Healthcare data validation separating matched records from an exception tray.
Make rejected or uncertain records visible for review.

Questions to take to your supplier

Submit an incomplete fictional record that the receiving rules should reject. How does the test environment respond, and what differs from live use? Agree authorised checks for important gaps. The supplier should explain the remaining uncertainty rather than treating a single successful sample as complete proof.

Agree the working process

Ask the supplier to explain which parts of the live service the test environment represents faithfully and which are simplified or unavailable. The business owner needs that limitation in plain language, especially where it affects permissions, validation or the receiving workflow. Choose examples that exercise actual requirements rather than only the vendor's easiest sample record. Include missing information, unsupported values and any relationship between records important to the task. Check the error or review route when the receiving system rejects an example.

Keep uncovered cases in the acceptance plan with a named owner and an authorised way to investigate them. A later live check should be planned proportionately, not improvised using real people or uncontrolled submissions. Record the differences that remain when deciding readiness. The useful question is what confidence the evidence provides for your specific journey. A successful demonstration can be valuable without being presented as proof of behaviours the available environment could not test.

How to check the result

Count acceptance assumptions not exercised by available testing.

Turn reading into a next step

Your action checklist

Work through these checks with your team or supplier. Tick the ones you have resolved and leave unknowns open.

Checks to discuss

Record what you know, what is still missing and the answer you need from your team or supplier.

Use project decisions only, without personal, patient or confidential details. Entries stay in this page and are not submitted to Kay & Co. Copy or download before leaving; this page does not save your notes.

Further reading

These sources provide background for the topic. The practice examples and checklist are illustrative planning suggestions from Kay & Co.

Related guides

Need help with this?

Tell us what is getting in the way. Kay & Co. can help you understand the options and turn the next step into something that works for your practice.

Explore integrations & data services or discuss your project.

Try the free Healthcare Digital Planner to find your starting priority.