Five high-risk drug monitoring gaps we find in almost every practice

High-risk drug monitoring is one of those areas where almost every practice believes it is doing reasonably well, and almost every practice has gaps when you look properly. These are the five we find most consistently, in practices of every size and system, together with the fix for each.

1. The register is not accurate

Everything depends on knowing exactly who is on a high-risk medicine. In practice, registers are built from searches that miss patients: a drug prescribed under a different brand name, a patient coded under an old read code, or a medicine issued by a hospital clinic and never added to the practice record. If the register is wrong, every downstream check is wrong too.

The fix: rebuild the register from prescribing data rather than problem codes, run it for each high-risk medicine individually, and reconcile it against hospital correspondence. Repeat the search with common misspellings and brand names. Treat the register as a living list that is re-run on a schedule, not a one-off exercise.

2. Overdue bloods with no follow-up

Most practices can produce a list of patients overdue for monitoring. Fewer can show what happened next. A patient who is three months overdue for a methotrexate FBC and has simply been sent one invitation is a known risk sitting in the system.

The fix: agree an escalation pathway for non-attendance — invitation, second invitation, direct phone contact, then a clinical decision about whether the next prescription is issued. The key step is linking monitoring to the repeat prescription: no current bloods, no automatic reissue, with a clinician review instead.

3. Shared-care protocol drift

Shared-care arrangements drift quietly. The specialist clinic changes its monitoring schedule, the practice keeps the old one. A patient moves area and the new shared-care letter never arrives. Responsibility for a test ends up assumed by both sides and done by neither.

The fix: keep a shared-care register separate from the general high-risk register, with the agreed protocol and the date of the last specialist review recorded for each patient. Chase missing clinic letters annually, and where a shared-care agreement cannot be confirmed, treat the medicine as practice-monitored until it is.

4. Results filed but not actioned

Abnormal results that are filed without action are the most dangerous gap because they look like work completed. A falling eGFR in a patient on metformin or a rising potassium on an ACE inhibitor only matters if someone acts on it.

The fix: route results for patients on high-risk medicines to a named clinician for review rather than general filing, and use a simple rule set — for example, any result outside the agreed range triggers a task, not just a comment. Audit a sample of filed results quarterly to confirm the rule is being applied.

5. No named owner

The underlying cause of the other four is usually this: monitoring is everyone's job and therefore no one's. Reception send invitations, GPs file results, the pharmacist does reviews — but nobody owns the system end to end.

The fix: name one person — usually a clinical pharmacist or prescribing lead — as the owner of high-risk drug monitoring. Their job is not to do every test but to run the register, chair the monthly exception review, and report gaps to the practice. Ownership turns five separate gaps into one managed workstream.

Where to start

Pick one medicine — methotrexate is a good choice because the monitoring requirements are unambiguous — and run the full cycle: accurate register, overdue list, escalation pathway, actioned results, named owner. Once the system works for one medicine, extending it to the rest of the high-risk list is repetition rather than reinvention.