Advancing Optimal Care With Robust Clinical Practice Guidelines
The narration between entries paraphrases; it never reads an entry aloud before the entry plays. Every entry in this lecture plays exactly as it does on its own page — same file, same speed, nothing laid over it.
What you can do after this
- Name the six criteria a document must meet to be a clinical practice guideline
- Say what a grade combines, and what a weak grade asks of you
- Distinguish the strength of a recommendation from its priority
- Name the six AGREE domains and say which carries the most weight
- Say why a guideline is not a rule
What it covers
57 entries in 10 topics. Each number links to that entry on its own.
-
1. What a guideline is
-
2. Six criteria
-
3. Finding guidelines
-
4. Grades
-
5. Eight attributes
-
6. The questions you ask
-
7. AGREE
-
8. Writing your own
-
9. The three canonical errors
-
10. A guideline, appraised
Check yourself
What six criteria must a document meet to be a clinical practice guideline?
Systematically developed statements; institutional auspices; a systematic review underneath that can be identified; benefits, harms and alternatives all considered; public availability; and currency within roughly five years. The last is checkable in seconds and is the one most often failed.
What does a weak recommendation ask of you?
More conversation with the patient, not less. A weak grade means the panel judged the balance of benefits and harms to be close, or the evidence uncertain — so the decision depends more on this individual's values. It is not a lesser instruction.
How does the strength of a recommendation differ from its priority?
Strength is the panel's confidence in the balance of benefits and harms. Priority is how much it matters in your setting. A strongly graded recommendation about a rare complication may be lower priority on your unit than a weakly graded one about something seen daily.
Which AGREE domain carries the most weight, and why?
Rigor of development. It covers how the evidence was found, appraised and turned into recommendations — which is where a guideline either earns the word systematic or does not. The other five domains describe the document; this one describes how it was made.
Why is a guideline not a rule?
Because it is a recommendation developed for a population and applied by a clinician to an individual patient. Reading it as a rule removes the clinical judgement and the patient preference that are two of the three components — which would stop it being evidence-based practice at all.
Lecture transcript
The narration as recorded, in order. Where an entry plays, the transcript links to it rather than reprinting it — the entry plays from its own file, unaltered, and says the same thing on its own page.
This is the last unit of Part 1, and it closes something the course opened at the very beginning. Unit 1 listed a misconception — that guidelines are rules — and left it standing. This unit answers it. A guideline is a set of recommendations developed systematically on top of a synthesis. That word systematically is doing all the work: everything in this lecture is a way of checking whether the document in front of you actually earned it. By the end you can name the six criteria a document must meet to be a guideline at all; say what a grade combines and what a weak grade asks of you; separate the strength of a recommendation from its priority; name the six AGREE domains and which carries most weight; and say why a guideline is not a rule.
Eight videos on the object itself. The key point is that a guideline rests on a synthesis rather than on assembled opinion, and that everything after this is about checking whether that is true of the one in front of you.
The definition.
EBP 1051 What a Clinical Practice Guideline Is
A statement of recommendations for practice, developed systematically to help clinicians and patients decide.
What it rests on.
EBP 1052 Built on a Synthesis
A guideline should rest on a systematic review rather than on assembled opinion.
Who writes them.
EBP 1053 Developed by Panels
Written by panels rather than individuals, which is a strength and a source of its own problems.
What a panel weighs.
EBP 1054 What a Panel Considers
Evidence, benefits, harms, costs and values, together.
The judgement at the centre.
EBP 1055 Balancing Benefits and Harms
Every recommendation is a judgement about whether benefits outweigh harms for a population.
What they are for.
EBP 1056 Guidelines Reduce Variation
The purpose is to reduce unwarranted variation in care between clinicians and settings.
Why a synthesis and not a study.
EBP 1057 Why Synthesis Is Needed
A single study cannot support a recommendation for a population; a synthesis answers against the body.
Two forms it may rest on.
EBP 1058 Meta-Analysis and Integrative Review
The synthesis underneath may be a meta-analysis or an integrative review.
Six videos, one per criterion. The last one is the one you can check in about ten seconds and the one most often failed.
Six criteria begin.
EBP 1059 Criterion One — Systematically Developed Statements
The recommendations are developed by a stated method rather than assembled by discussion.
The second.
EBP 1060 Criterion Two — Institutional Auspices
Who stands behind it, which bears on independence and on what interests were in the room.
The third.
EBP 1061 Criterion Three — Based on a Systematic Review
There is a systematic review underneath, and it can be identified and checked.
The fourth.
EBP 1062 Criterion Four — Benefits, Harms and Alternatives
Benefits, harms and alternatives are all considered, not only the option being recommended.
The fifth.
EBP 1063 Criterion Five — Publicly Available
It can be read by the people expected to follow it.
The sixth, and the checkable one.
EBP 1064 Criterion Six — Current Within Five Years
Currency within roughly five years, because the evidence underneath moves.
Seven videos on where they live. Note that a web search returns guidelines without telling you what it missed — the same reason a search engine was not a source of evidence in unit 3.
Where they live.
EBP 1065 Guideline Databases
Guidelines live in their own repositories rather than in the bibliographic databases.
A useful search handle.
EBP 1066 Practice Guideline as a Publication Type
Guideline is a publication type you can limit a database search to.
Why searching the open web fails.
EBP 1067 Why a Web Search Fails Here
A web search returns guidelines without telling you what it missed, or how current any of them are.
The technique instead.
EBP 1068 Refine Inside the Source
Search inside the repository, where the coverage is stated and the filters mean something.
A shortcut that exists.
EBP 1069 Guideline Trust Scorecards
Some repositories score guidelines for trustworthiness, which is a starting point rather than a verdict.
A caution about infrastructure.
EBP 1070 Repositories Can Disappear
Guideline repositories have closed before, so the infrastructure is less permanent than it looks.
The alternative to writing one.
EBP 1071 Adapting an Existing Guideline
Adapting a good guideline to local conditions is usually better than writing a new one.
Six videos on what a grade means. The one to hold is that a weak grade demands more conversation with the patient rather than less — it is not a lesser instruction.
How recommendations are marked.
EBP 1072 Recommendations Carry a Grade
Each recommendation carries a grade expressing how strongly it is made.
What a grade actually is.
EBP 1073 Grades Combine Evidence and Judgement
A grade combines the strength of evidence with the panel's judgement about benefits and harms.
The first kind.
EBP 1074 A Strong Recommendation
Strong means most patients should receive this, and the panel was confident about the balance.
The second.
EBP 1075 A Weak Recommendation
Weak means the balance is close or the evidence uncertain, so the choice depends more on the individual.
What a weak grade asks of you.
EBP 1076 Weak Grades Demand More Conversation
A weak recommendation calls for more discussion with the patient, not less. It is not a lesser instruction.
A distinction easily missed.
EBP 1077 Strong Is Not the Same as High Priority
Strength is about confidence in the balance. Priority is about how much it matters here. Different axes.
Eight videos on what a good guideline has. These are properties of the document; the next topic is the questions you ask of it.
Eight attributes of a good guideline.
EBP 1078 Attribute One — Validity
Following it produces the outcomes it predicts.
The second.
EBP 1079 Attribute Two — Reliability and Reproducibility
Another panel with the same evidence would reach the same recommendations.
The third.
EBP 1080 Attribute Three — Clinical Applicability
The populations it applies to are stated plainly.
The fourth.
EBP 1081 Attribute Four — Clinical Flexibility
It says where exceptions apply, rather than pretending there are none.
The fifth.
EBP 1082 Attribute Five — Clarity
Unambiguous language, so two readers act the same way.
The sixth.
EBP 1083 Attribute Six — Documentation
The process, participants and evidence are recorded.
The seventh.
EBP 1084 Attribute Seven — Multidisciplinary Development
Everyone affected was represented in developing it, including patients.
The eighth.
EBP 1085 Attribute Eight — Plans for Review
There is a stated plan for when it will be revisited.
Nine videos. The first is asked before you read a single recommendation, and the last is the one that turns appraisal into a plan: how far this sits from what you do now.
The first question to ask.
EBP 1086 Who Developed It?
Ask who wrote it and who funded it before you read a single recommendation.
The second.
EBP 1087 Were All Options and Outcomes Considered?
Whether alternatives were weighed or only the recommended option examined.
The third.
EBP 1088 Is Each Recommendation Tagged and Linked?
Whether each recommendation can be traced to the evidence behind it.
The fourth.
EBP 1089 Are the Recommendations Explicit?
Whether they say what to do precisely enough to act on.
The fifth.
EBP 1090 Peer Review and Testing
Whether it was reviewed externally and tested before release.
The sixth.
EBP 1091 Is the Intent of Use Stated?
Whether it says who it is for and in what setting.
The seventh.
EBP 1092 Are the Recommendations Relevant and Helpful?
Whether the recommendations address decisions clinicians actually face.
The eighth.
EBP 1093 Are They Practical and Feasible?
Whether they can be carried out with the resources available.
The last, and the practical one.
EBP 1094 How Far From Current Practice?
How far the recommendation is from what you do now. That distance is the implementation work.
Seven videos on the named instrument and its six domains. Rigor of development is the domain that carries the most weight.
The instrument by name.
EBP 1095 The AGREE Instrument
A structured instrument for appraising guideline quality across six domains.
The first domain.
EBP 1096 AGREE Domain — Scope and Purpose
What the guideline covers and who it is for.
The second.
EBP 1097 AGREE Domain — Stakeholder Involvement
Whether the right people, including patients, were involved.
The third, and the heaviest.
EBP 1098 AGREE Domain — Rigor of Development
How the evidence was found, appraised and turned into recommendations. The domain that carries the most weight.
The fourth.
EBP 1099 AGREE Domain — Clarity of Presentation
Whether the recommendations are unambiguous and easy to find.
The fifth.
EBP 1100 AGREE Domain — Applicability
Whether barriers, costs and implementation are addressed.
The sixth.
EBP 1101 AGREE Domain — Editorial Independence
Whether funders influenced the recommendations, and whether interests are declared.
Two videos for when no suitable guideline exists.
When you have to write one.
EBP 1102 Developing a Local Guideline
When no suitable guideline exists, a local one is developed by the same standards.
How to do that properly.
EBP 1103 Prefer Formal and Explicit Methods
Formal, explicit methods beat informal consensus, even locally.
Three videos, and the first closes something opened in unit 1 — the very first misconception this course raised, answered here at the end of Part 1.
The first error, and the one seeded in unit 1.
EBP 1104 Misconception: Guidelines Are Rules
A guideline is a recommendation for a population, applied by a clinician to a patient. Read as a rule, it removes the judgement that makes it evidence-based practice.
The second.
EBP 1105 Misconception: A Guideline Is Current Because It Is Published
Publication says nothing about currency. Check the date and what has appeared since.
The third.
EBP 1106 Misconception: A Strong Grade Settles It
A strong grade settles the evidence, not the decision. The patient in front of you is still a variable.
One procedure taking a guideline to an adoption decision.
The strand assembled.
EBP 1107 Appraising a Clinical Practice Guideline
One guideline taken through its development, its grades and the AGREE domains to an adoption decision.
That is guideline appraisal. Against the outcomes. Six criteria: systematically developed statements, institutional auspices, a systematic review underneath, benefits and harms and alternatives all considered, public availability, and currency within about five years. The last is checkable in ten seconds and is the one most often failed. A grade combines the strength of the evidence with the panel's judgement about benefits and harms. Strong means most patients should receive this. Weak means the balance is close — and a weak recommendation demands more conversation with the patient, not less. Strength and priority are different axes. A strongly supported recommendation about something rare may matter less on your unit than a weakly supported one about something you see every shift. The six AGREE domains are scope and purpose, stakeholder involvement, rigor of development, clarity of presentation, applicability, and editorial independence. Rigor of development carries the most weight, because it is where the evidence became recommendations. And a guideline is not a rule. It is a recommendation built for a population, applied by a clinician to a patient — which is exactly the judgement unit 5 said no formula can replace. Read as a rule, it removes the thing that makes this evidence-based practice. Part B takes the appraised guideline and asks how it actually reaches the bedside.
This is free, and stays free. If it helped, here is what keeps it running.