Can people on patient-initiated follow-up (PIFU) pathways be identified in OpenSAFELY?

[alschaffer]

Patient initiated follow-up (PIFU) was introduced in 2021. A PIFU pathway is when a patient or their carer can initiate an outpatient appointment when they need one, based on their symptoms and individual circumstances, rather than on a set schedule.

Can patients on PIFU pathways be identified using current OpenSAFELY data holdings, or are additional data columns needed?

[alschaffer]

Some background:

• Patients must first be identified as eligible for PIFU, and patients must identify that they are on a PIFU pathway when booking an appointment.
• Providers are meant to record when a patient is moved to a PIFU pathway.
• The question is whether PIFU can be identified either with current OpenSAFELY data holdings, or would require additional data columns. It is also not clear whether providers are actually recording when patients are on a PIFU pathway as they should.

How is PIFU captured in HES Outpatient (OPA) data?

PIFU does not appear to be captured in any standard coding system (SNOMED, ICD-10, CTV3). CTV3 does have several codes for “patient-initiated encounter”, but CTV3 is no longer in widespread use and it’s unclear whether these codes would capture PIFU regardless.

In 2022/23 (CDS V6-3), new categories were introduced in HES to the variable OUT-PATIENT ATTENDANCE OUTCOME. This identifies patients discharged to or moved to a personalised out-patient follow-up pathway. The categories are:

• 1 = PATIENT discharged from the care of the CARE PROFESSIONAL without Personalised Out-Patient Follow Up
• 2 = PATIENT given a Timed Out-Patient Follow Up Appointment while at the out-patient attendance without Personalised Out-Patient Follow Up
• 3 = PATIENT to be given a Timed Out-Patient Follow Up Appointment at a later date without Personalised Out-Patient Follow Up
• 4 = PATIENT moved to a Personalised Out-Patient Follow Up Pathway
• 5 = PATIENT discharged to a Personalised Out-Patient Follow Up Pathway
• 9 = Not known

Importantly, a “personalised out-patient follow-up pathway” includes both PIFU, and remote monitoring. In contrast to PIFU, “remote monitoring” is when the clinical team initiates an appointment based on clinical information obtained through monitoring the patient’s condition. This might be through using wearable technologies, apps, clinical questionnaires or from test results. A patient can be both on PIFU and remote monitoring.

Also in 2022/23, several new data items were introduced to HES related to personalised out-patient follow-up, specifically:

• APPOINTMENT BOOKED REASON: whether an appointment is planned (1), part of PIFU (2), or remote monitoring (3);
• PATIENT ON PATIENT INITIATED OUT-PATIENT FOLLOW UP PATHWAY INDICATOR AT CDS ACTIVITY DATE whether a patient is on a PIFU pathway (Y/N);
• PATIENT SUBJECT TO REMOTE MONITORING INDICATOR AT CDS ACTIVITY DATE: whether a patient is on a remote monitoring pathway (Y/N);
• PERSONALISED OUT-PATIENT FOLLOW UP PATHWAY REVIEW DATE: when the personalised follow-up pathway needs to be reviewed;
• PERSONALISED OUT-PATIENT FOLLOW UP PATHWAY EXPIRY DATE: when the personalised follow-up pathway will expire.

There are some data suggesting that certain trusts are not recording patients being moved to a PIFU pathway, and this may differ by patient administration system (PAS) in use.

Can PIFU be identified in OpenSAFELY?

The list of available columns in current OpenSAFELY HES OPA holdings is available here. Getting new columns is not currently straightforward.

Within OpenSAFELY, we currently have a column called OUTCOME_OF_ATTENDANCE. It is likely that this is the same as OUT-PATIENT ATTENDANCE OUTCOME, but data curation work will be needed to confirm. If it is the same, then this could be used to narrow down the population of people on PIFU pathways.

While the new categories in OUTCOME_OF_ATTENDANCE may identify people on personalised outpatient follow-up pathways, without the new PIFU flag variables, it does not appear possible to distinguish people on PIFU pathways vs remote monitoring pathways. Clinical input is needed to determine how much this matters.

Next steps

Data curation work is needed to find out the following:

• Are the new categories in OpenSAFELY being used? If so, how has their use changed over time?
• How does their use compare with publicly available aggregate PIFU data?
• How can we estimate what proportion of personalised outpatient follow-up patients are attributable to PIFU vs remote monitoring? Does this differ by specialty?

[alschaffer]

FYI @brianmackenna