# The platform didn't know what the hospital knew

> An RNSH audit of 1,532 patients who missed their digital pre-op questionnaire found 21 invited while admitted, 9 after death and 23 after surgery, while 4.9% had been transferred and 5.6% had declined or been removed. Every fact was in the medical record. None stopped the invitation, because the questionnaire platform was not listening to the record's events.

*Source: https://getpathways.ai/articles/the-platform-didnt-know*

When RNSH audited 1,532 patients who didn't complete their digital
pre-operative health questionnaire in 2025, some of the most striking findings
had nothing to do with the patients:

- 21 were invited while acutely unwell or already admitted
- 9 were invited after death
- 23 were invited after their surgery had already happened
- 4.9% had been transferred to another site
- 5.6% had asked to be removed or had declined surgery

In all, 541 non-completers (35.3%) never proceeded to surgery. Every one of these
facts was recorded in the electronic medical record. None of them stopped the
invitation going out.

## The cost nobody counts

The most obvious cost is human. A family receiving a reminder to complete a
health questionnaire for someone who has just died will remember it. So will a
patient lying in a ward bed being asked about their exercise tolerance.

The less visible cost is labour. The poster lists the staffing factors behind
the digital pathway: uploading patients into the system, troubleshooting, and
follow-up. Every invalid invitation creates follow-up work for a team that is
already stretched. It also inflates the non-completion figures, which makes the
platform look less effective and the patients look less engaged than they are.

And in 24.4% of cases, there was no documented reason for non-completion at all.
Without a feedback loop between the platform and the record, nobody can tell a
disengaged patient from an invitation that should never have been sent.

## What the EMR knows, and what the platform should do with it

The authors' diagnosis is weak integration between the digital platform and the
EMR. In practice that means the platform acts on a list, while the EMR holds a
stream of events that should change what happens next.

| EMR event | Audit finding | What the platform should do |
| --- | --- | --- |
| Death recorded | 9 invited after death | Stop all outbound contact immediately |
| Admitted or acutely unwell | 21 invited | Pause; re-time after discharge |
| Surgery date has passed | 23 invited after surgery | Never send |
| Transferred to another site | 4.9% of non-completers | Cancel and hand over to the receiving site |
| Removed from list or declined | 5.6% of non-completers | Cancel |
| Surgery rescheduled | Cited as a driver | Re-time to the new date |

Most Australian hospitals already publish admission, discharge and transfer
events through HL7 interfaces, and newer integrations increasingly use FHIR. The
data exists. What's missing is a platform that listens and acts.

## What good integration looks like

1. **A validity check before every send.** Is the patient alive, not currently
   admitted, still on the waiting list, still booked at this site, with a surgery
   date in the future?
2. **Event-driven updates.** Admission, discharge, transfer, death, cancellation
   and rescheduling events pause, re-time or cancel invitations automatically.
3. **Status written back to the record.** Clinicians should see in the EMR
   whether a questionnaire was sent, opened, completed, or stopped, and why.
4. **Clean metrics.** Completion rates should exclude invitations that were
   invalid when sent, so the numbers measure patient engagement rather than data
   quality.

Integration can sound like a back-office concern. This audit shows it decides
whether the digital pathway reaches the right patient at the right time, and
whether it treats them with basic dignity when it does.

*Source: Zhang D, Marks A, Doane M, et al. Barriers to current perioperative
digital platforms – how can innovation solve these issues? Poster, ASA CSC.*