Guidelines vs Protocols vs Pathways

"It's in the guideline." "There's a protocol for that." "She's on the pathway."

Three sentences, used in most departments as though they were the same sentence. They are not. They carry different degrees of obligation, different amounts of room for clinical judgement, and — the part that matters when something goes wrong — different consequences when you depart from them.

Getting this straight is worth ten minutes, because it changes how you document a decision to do something different. It sits within our governance series.

The three, distinguished

A guideline informs judgement. A protocol replaces it. A pathway sequences it.

That is the whole distinction, and it explains why the same departure can be entirely unremarkable in one case and a serious incident in another.

Why protocols exist at all

It looks paternalistic to remove clinical discretion, and in most situations it would be. Protocols are reserved for circumstances where discretion reliably makes outcomes worse:

• Time-critical emergencies, where deliberation costs more than a suboptimal-but-immediate action — massive haemorrhage, cardiac arrest, sepsis.
• High-consequence, low-frequency events, where nobody has enough personal experience to reason well under pressure — malignant hyperthermia, local anaesthetic toxicity.
• Tasks where variation itself is the hazard — the WHO checklist, swab counts, site marking.

The common feature is that the cost of unwarranted variation exceeds the value of individual judgement. That is a genuinely narrow set of situations, and it is why most clinical documents are, correctly, guidelines rather than protocols.

The medico-legal difference, which is the practical point

This is where the distinction stops being semantic.

Departing from a guideline is legitimate. Guidelines are, by design, recommendations for the typical patient. Your patient may not be typical. Courts and regulators do not expect guideline adherence; they expect a reasonable decision, reasonably reasoned. A documented departure with a stated rationale is entirely defensible — and in some cases following a guideline blindly, where the patient plainly did not fit it, is the criticised decision.

Departing from a protocol is a different conversation. Protocols exist precisely because discretion is unhelpful in that setting. Departure is sometimes right, but it needs a clear reason, usually senior involvement, and always a contemporaneous record.

What this means in practice:

A guideline is not a legal standard, and it is not a substitute for one. The standard remains what a reasonable practitioner would do — see The Medico-Legal Duties of a Surgeon. But an unexplained departure from widely accepted guidance is difficult to defend, because the guideline is strong evidence of what reasonable practice looks like.

The asymmetry is worth internalising: the guideline does not bind you, but it does shift the burden. Follow it and little explanation is needed. Depart from it and the explanation is the whole of your defence — which is fine, provided you wrote it down at the time.

How to document a departure

Four elements, and it takes two sentences:

1. Name the guidance you are departing from.
2. State the patient-specific reason — what about this patient makes the recommendation inapplicable.
3. State what you did instead, and why it is reasonable.
4. Record any discussion — senior colleague, MDT, the patient.

For example: "NICE guidance would suggest X. Not appropriate here given her previous reaction to Y and her stated priority of avoiding Z. Discussed with Mr A and with the patient; proceeding with W, which achieves the same objective. Patient understands the reasoning."

That note converts a departure from a liability into a demonstration of judgement. Its absence converts the same clinical decision into an apparent oversight. See Defensible Clinical Documentation.

Strength of recommendation is not strength of evidence

A distinction that trips up trainees in vivas and consultants in practice.

Most modern guidelines separate the quality of the evidence from the strength of the recommendation, typically using GRADE or something like it. They are independent:

• A strong recommendation on low-quality evidence is common where harm is severe and alternatives are poor. Nobody will run the trial, but the recommendation is firm.
• A weak recommendation on high-quality evidence happens where the effect is real but small, or where it depends heavily on patient preference.

Weak or conditional recommendations are precisely where individual judgement and patient values are supposed to operate. Treating every recommendation as equally binding misreads the document — and treating a conditional recommendation as a rule is how shared decision-making gets skipped.

Where they come from, and why they conflict

Orthopaedics has a genuine problem here: multiple national bodies produce guidance on the same questions, and they do not always agree.

Where national guidance genuinely conflicts, that itself is evidence of legitimate professional disagreement — which usually means either approach is defensible, provided you can say which you followed and why. What is not defensible is being unaware that the disagreement exists.

Local policy deserves attention. It binds you contractually even where it is more restrictive than national guidance. If your local policy is out of date or wrong, the fix is to change it through the governance route, not to quietly ignore it.

Pathways, and the patients who leave them

Pathways — enhanced recovery, fractured neck of femur, day-case arthroplasty — sequence an episode of care over time. Their value is reliability: everyone knows what happens on day one, the physiotherapist is expecting the patient, discharge planning starts at admission rather than on day four.

Their characteristic failure is different from a guideline's. A pathway works well for patients who fit it, and the risk is that the patient who stops fitting it continues along it anyway because the pathway has its own momentum.

The safeguard is an explicit variance step: when a patient deviates, that is recorded, reviewed, and acted on rather than treated as a nuisance. A pathway without a variance process quietly converts individual patients into an average one.

Why so many documents go unused

Most departments have far more guidance than anyone reads. The common causes are consistent:

• Too long. A twelve-page document will not be consulted at 2am.
• Not where the work is. Guidance in a folder on the intranet does not exist at the bedside.
• Out of date, with no review date and no owner.
• Conflicting with another local document nobody has reconciled.
• Written for assurance, to demonstrate that a policy exists, rather than for use.
• No named owner, so nobody updates it.

The fixes are unglamorous: every document gets a named owner and a review date; anything intended for use in an emergency fits on one page; and the format matches the type — a guideline can be prose, but a protocol should be a numbered list you can follow while your hands are busy.

If you are writing one

• Decide which of the three it is before writing a word. Most documents fail because the author has not chosen.
• Match the language. "Consider" in a guideline; "must" in a protocol. Mixing them makes the obligation ambiguous, which is the worst outcome of all.
• State the evidence and its strength, and where a recommendation is conditional, say so.
• Name an owner and a review date on the front page.
• Say what to do when it does not apply — an explicit escalation route beats an implicit one.
• Keep an emergency protocol to one page, and test it in a simulation before publishing.
• Reconcile it with existing documents rather than adding to the pile.

The summary

A guideline informs your judgement; a protocol replaces it; a pathway sequences it. Departing from a guideline is legitimate and needs a documented, patient-specific reason. Departing from a protocol needs a reason, usually a senior discussion, and always a note.

Guidelines are not law, but they are strong evidence of what reasonable practice looks like — so the departure is defensible exactly to the extent that you wrote down why you made it.