Hospital Formulary Management and the P&T Committee

The Pharmacy and Therapeutics Committee needs enough clinical authority that its decisions are respected and enough procedural discipline that they are made on time. Typical composition includes senior physicians across major specialties, surgery and anaesthesia, intensive care, the chief pharmacist, nursing leadership, microbiology or infection control, and a finance representative who advises rather than votes. An executive sponsor matters, because some decisions will be unpopular with a department that has an established preference.

Meeting rhythm should be predictable — monthly or bi-monthly — with a published agenda deadline, a written dossier for each application circulated in advance, and a documented decision with reasons. Committees that decide by discussion in the room, without a dossier, reliably produce decisions that get relitigated. Committees that meet only when someone escalates produce a backlog that pushes clinicians toward the non-formulary route permanently.

Conflict of interest declarations should be routine rather than exceptional. Requiring every applicant and every voting member to declare industry relationships for the specific agent under review protects the committee's decisions from later challenge.

A structured application form does most of the work, because it forces the requester to answer the questions the committee would otherwise have to extract. The dossier should establish clinical need relative to existing formulary options, the evidence base, safety profile including interactions and monitoring requirements, the practical handling burden, and total cost of therapy rather than unit price. Total cost matters because a more expensive vial that avoids a second-line escalation or shortens an admission can be the cheaper choice.

The most useful single question is comparative: what does this agent do that the current formulary option does not, for which patients, and how many such patients does this hospital see in a year? An agent that is genuinely better for four patients a year may be better handled as a named-patient procurement than as a stocked item.

Decisions should have more options than yes and no. Conditional approval — restricted to a specialty, requiring a specialist countersignature, limited to a defined indication, or approved for a trial period with a scheduled utilisation review — resolves most contested applications better than a binary vote.

Almost no committee enjoys deletion, and almost every formulary needs it. The mechanism that works is a standing agenda item driven by data rather than opinion: each meeting, pharmacy tables the items with negligible consumption over the past twelve months, items superseded by a newer approval, items with recurring supply failures, and duplicate agents within a therapeutic class. The default for a zero-consumption item is deletion unless someone argues for retention.

Deletion needs a runway. Announce the removal, set a date, allow existing patients on the agent to complete therapy or be switched deliberately, and communicate the substitute. Silent deletion — where the item simply stops being stocked — produces the exact mid-treatment scramble that makes clinicians distrust the whole process.

Supply reliability is a legitimate deletion criterion that committees underuse. An agent that is nominally on the formulary but out of stock a third of the time is worse than an alternative that is always available, because it produces repeated last-minute substitution decisions made under time pressure by whoever is on duty.

Therapeutic interchange is the pharmacist's authority to dispense a formulary agent in place of a prescribed non-formulary one within a class, according to a protocol the committee approved in advance. It is not generic substitution, which swaps the identical molecule, and it is not an ad-hoc clinical decision. The distinction matters legally and clinically, and it should be written down in exactly those terms.

Each interchange protocol needs its own document: the classes and specific agents covered, the dose-equivalence table, explicit exclusions, the notification requirement to the prescriber, and the prescriber's right to override by marking the order as clinically necessary. Exclusions are the part that gets skipped and the part that causes harm — patients already stabilised on a specific agent, narrow therapeutic index drugs, paediatric and pregnancy cases, and transplant or oncology protocols usually belong outside automatic interchange.

Interchange works only if the substitution is visible in the record. The dispensed product, not the ordered one, must appear in the medication administration record and the discharge summary, or the next clinician reconciles against a medicine the patient never received.

Every formulary needs a legitimate exception path, and the design goal is that it is available, auditable, and slightly effortful. A request should capture the patient, the indication, why each formulary alternative is unsuitable for this patient, the intended duration, and the requesting consultant's name. Approval sits with a designated clinical authority — the chair, a duty pharmacist under delegated criteria, or a specialist reviewer — with a defined turnaround so urgent cases are not stranded.

The critical governance step is what happens to the data afterwards. Non-formulary requests should be tabled at every committee meeting in aggregate, by agent and by requester. Any agent appearing repeatedly is telling the committee that the formulary has a genuine gap, and the correct response is to reconsider the addition rather than to keep approving exceptions one at a time. HealUDoc dashboards can summarise non-formulary request volume by agent and specialty, which turns that standing agenda item into a two-minute review rather than a manual collation.

Set an explicit rule for out-of-hours requests, because that is where the process leaks. A duty-pharmacist approval with mandatory next-working-day committee notification is usually the right balance between patient need and control.

A formulary that lives in a PDF on the intranet is not enforced; it is merely published. Enforcement means the ordering screen defaults to formulary agents, marks non-formulary items clearly, and routes any attempt to order one into the request workflow rather than silently allowing it. A platform such as HealUDoc can carry formulary status and restriction rules on the medication master so that specialty limits and countersignature requirements apply at order entry rather than being discovered at dispensing.

Restriction logic should be expressed in the same terms the committee used. If the committee approved an antimicrobial for intensive care and infectious diseases only, the system should permit it for those prescribers and prompt everyone else into the exception path — not present the drug identically to everyone and rely on pharmacy to catch it later. Catching restrictions at verification works, but it costs a phone call and a delay every single time.

Multi-branch hospitals need one clinical formulary with branch-level stocking differences represented separately. Letting each site maintain its own formulary produces divergent standards of care and makes cross-branch utilisation analysis meaningless within about two years.

A formulary needs a scheduled full review — typically annual — in addition to the continuous stream of additions and deletions. The annual pass should check every therapeutic class for redundancy, confirm that restrictions still reflect current practice, retire trial approvals that were never reviewed, and align the formulary with national essential medicines guidance and the hospital's antimicrobial stewardship policy.

Publish the outcome in a form clinicians will actually consult: searchable, current, showing restriction status and the approved alternative for anything deleted. A dated version history matters more than it sounds, because prescribing decisions get reviewed retrospectively and the reviewer needs to know what the formulary said at the time.

Finally, measure the committee itself. Median days from application to decision, proportion of non-formulary requests concentrated in a handful of agents, and the count of approvals that were never reviewed at their scheduled date tell you whether the process is functioning or merely convening.