Less is more: Guidelines - PDF Free Download

ESIM Summer School June 20 Friday, 2014 Less is more: Guidelines Primiano Iannone, MD Head of Emergency Department Ospedali del Tigullio, Lavagna (GE) Italy

layout What is a clinical guideline Why do we need clinical guidelines How guidelines are (and should be) produced Quality of current guidelines How to decide whether a guideline is trustworthy How to use a clinical guideline The future of clinical guidelines

2011

How clinical guidelines are percieved Farquhar CM, et al. Med J Aus 2002

What is a clinical guideline Why do we need clinical guidelines How guidelines are (and should be) produced Quality of current guidelines How to decide whether a guideline is trustworthy How to use a clinical guideline The future of clinical guidelines

Physicians take decisions about (with) their patients We take clinical decisions every day, several times per day. Often, these decisions are semiautomatic choices dictated by consolidated practices, previous professional experiences and knowledge background, local clinical habits and policies, with an heuristic approach typical of type 1 thinking («fast thinking») according to Kahneman. Often we consider also what our patients say and think about their illness. Sometimes we don t.

However, we face often serious uncertainty about the quality of evidences on which to base our decisions, as well as to what extent individual patients conditions related to age, gender, morbidity, personal preferences and beliefs could modify the picture. In these cases a sound, slow and complex rational approach («type 2 thinking», referring again to Kahneman s terminology) is required.

So, we need searching, appraising and staying up-to-date with the best evidence, integrating it with our personal knowledge and experiences, as well as with cost considerations, weighting risks and benefits carefeully, patients preferences, with a clever clinical reasoning but can we do this efficiently?

the exponential growth of randomized controlled trials

we need leaner and more efficient methods of staying up-to-date with the evidence. Using current methods, the Cochrane Collaboration has not been able to keep even half of its reviews up-todate

RCTs indexed on PubMED 1978-2013 355.272 1978: 1787 RCTs 2013: 16944 RCTs Heart failure [MeSH] RCTs 2010-2013: 1104

So, what do we need? Raise the right questions in an answerable manner (PICO) Search for evidences efficiently Appraising critically evidences and rating them Integrating evidence with our experiences and previous knowledge Adapting evidences and deciding whehter it is worth applying them to individual patients

JAMA, 1992; 268: 2420-25

Evidence based medicine: what it is and what it isn't Evidence based medicine is the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients. The practice of evidence based medicine means integrating individual clinical expertise with the best available external clinical evidence from systematic research. David L Sackett, William M C Rosenberg, J A Muir Gray, R Brian Haynes, W Scott Richardson BMJ 1996;312:71-72 (13 January)

Evidence based medicine Evidence Based Clinical Guidelines

Multidisciplinary development Studies have shown that the balance of disciplines within a guideline development group has considerable influence on the guideline recommendations Systematic review of literature Guidelines based on a consensus of expert opinion or on unsystematic literature surveys have been widely criticised as not reflecting current medical knowledge and being liable to bias. Graded recommendations Guideline recommendations are graded to differentiate between those based on strong evidence and those based on weak evidence Miller J, Petrie J. Development of a practice guideline. Lancet 2000; 355:82 3.

What does it mean systematic review of literature? To minimise potential sources of bias in the guideline recommendations, the literature should be identified according to an explicit search strategy, selected according to defined inclusion criteria,and assessed against consistent methodological standards

Certainty (Level of evidence) Strenght of recommendations high strong low weak Graded recommendations

Many guidelines derive(d) level of evidence almost exclusively from study type Moreover classification of level of evidences with letters, numbers, or symbols was chaotic

ESC/AHA

SIGN

ERC 2010 guidelines

Type of study Quality of evidence Strength of recommendation

And so, no RCT, no strong recommendation?

Sometimes trials are unethical or impossible

Sometimes trials are unethical or impossible yet some treatments are quite effective DC shock for ventricular fibrillation Insulin for diabetic coma Blood trasfusion for haemorrhagic shock

Type of study Quality of evidence Strenght of recommendation Other factors?

1.Relevance of outcomes Importance of the outcome that treatment prevents Deep vein thrombosis : Postflebitic syndrome vs death Embolism from Pulmonary Atrial Fibrillation: Palpitations Vs stroke

2. Magnitude of treatment effect the lower the NNT (=1/ARR), more effective the treatment is

Relative risk reduction overestimates effect of treatment

3.Risk of Bias Systematic error leading to overestimate or underestimate of true treatment effect Also RCTs may be affected by several biases that weaken their quality Selection bias Detection bias Attrition bias Reporting bias..

4. Precision Reliable measurement of the effect size of the treatment 95 %Confidence intervals ASA vs Placebo for stroke prevention in Atrial Fibrillation has wider 95% CI than in Transient Ischemic Attacks

5. inconsistency Conflicting results across trials

6. directness differences between studied and target population as regard of Interventions Patients (applicability) Outcomes (hard vs surrogate) Absence of head to head comparisons

a more complex approach is needed type of study directness precision Balance of all favorable /unfavorable outcomes Quality of evidence Strenght of recommendation relevance consistency resources? magnitude of effect risk of bias patients values & preferences

According to GRADE Quality of evidence must be summarized in a table Iannone et Al, JAMA Intern Med, 2014

a more complex approach

http://www.gradeworkinggroup.org/

There are good guidelines

However many medical specialty societies haven t adopted GRADE yet

Low quality of early guidelines Grilli et al : Lancet, 2000

We have also another problem

Conflict of interests Is not a source of a random error COI generates BIAS Bias almost always results in an overestimation of benefit and an underestimation of harm

A COI is a set of conditions in which professional judgment concerning a primary interest (such as the health and well being of a patient or the validity of research), is unduly influenced by a secondary interest - The secondary interests may be financial or nonfinancial. Thompson DF (1993) Understanding financial conflicts of interest. NEJM 329: 573 576.

Managing COI within a guideline panel is of paramount importance to warrant trustworthy recommendations

Lenzer et Al, BMJ 2013

How to decide whether a guideline is trustworthy Traditional approach Evaluation frameworks (AGREE, GIN, IOM standards) Concordance between guidelines

2011 1. Establishing Transparency 2. Management of Conflict of Interest (COI) 3. Guideline Development Group Composition 4. Clinical Practice Guideline Systematic Review Intersection 5. Establishing Evidence Foundations for and Rating Strength of Recommendations 6. Articulation of Recommendations 7. External Review 8. Updating

Evaluation frameworks explore the quality of producing and reporting guidelines NOT the trustworthiness of their recommendations

Concordance of recommendations between (among) different guidelines Adherence to quality standards how many guidelines for a disease? Proxy of trustiworthiness Or not?

A case study Three renowned medical specialty societies Three guidelines on the same disease (why?) Same evidence base about a given drug X One of three guideline declares to comply with GRADE Full disclosures of conflict of interests Another guideline declares to comply with AGREE criteria Substantial agreement among them about the effectiveness of drug X Who could doubt it?

Can dronedarone be recommended for preventing recurrences of Atrial Fibrillation? Three renowned medical specialty societies (AHA, ESC, CCS) Three guidelines on the same disease (why?) Same evidence base (6 RCTs) about dronedarone One guideline declared to comply with GRADE Full disclosures of conflict of interests Another guideline declared to comply with AGREE criteria Substantial agreement among them about the effectiveness of dronedarone

However applying GRADE methods to the same evidence base considered by these three guidelines. We didn t find relevant favorable outcomes, we found unexplained heterogeneity of results, and we could not exclude an unfavorable effect of dronedarone on mortality, with an excess of 13 (95%CI, 15 to 61) deaths per 1000 patients treated with it

Iannone et Al, JAMA Internal Medicine, 2014

Can dronedarone be recommended for Atrial Fibrillation? Three renowned medical specialty societies (AHA, ESC, CCS) Three guidelines on the NO same disease (why?) Same evidence base (6 RCTs) about dronedarone One guideline declares to comply with GRADE Full disclosures of conflict of interests Another guideline declares to comply with AGREE criteria Substantial agreement among them about the effectiveness of dronedarone

in presence of flawed methods (no GRADE guidelines) uncontrolled conflict of interests restricted panel compositions Concordance of recommendations between guidelines and declared adherence to quality standards do not warrant their trustworthiness

A roadmap I would suggest (a very modest & weak recommendation )

Have You a clinical problem? IOM criteria helpful PICO conceptualisation Search whether a guideline addressing relevant outcomes does exsist YES No/Negligible conflict of Interest? YES Sound methodology? (GRADE fully exploited) YES Multidisciplinary involvement? YES Low risk of untrustworthiness NO NO NO NO Search for other evidences Evaluate primary evidences carefully in case of any doubt

Follow GRADE conceptualisation Overall quality of evidences Relevance of outcomes Type of studies Precision Consistency Directness Risk of bias Modifiers Balance across all favourable and unfavourable outcomes Patients values and preferences Resources use

assessing their trustworthiness, too.

If a recommendation is trustworthy, the main issue is deciding whether It can be applied to our patient(s) It is a matter of clinical judgement considering differences beween ideal study conditions where evidences were produced and real life settings in terms of Patients Type of intervention Outcomes considered absence or presence of head-to-head comparisons There is often some uncertainty about this

If a recommendation is untrustworthy It should be openly and widely presented and discussed to avoid unnecessary harm to the patients and resources wasting Reasons of untrustworthiness should be clarified and addressed An in depth GRADE based, multidisciplinar, unconflicted reassessment of flawed recommendations should be urgently carried out to produce more firm guidelines

When guidelines highlight the absence of firm evidences clinical research agenda should be prioritized to fulfill these gaps, if relevant for our patients problem driven research vs curiosity driven research EBM helps ethical integrity of biomedical research

Clinical guidelines at their crossroad Evolution or extinction

Their transformation into More trustworthy, more evidence based, unconflicted, balanced tools to inform wise clinical decisions and manage uncertainty

I didn t mean to confuse You But Evidence Based Medicine is an eminently creative methodology which emphasizes critical reasoning and not the robotic application of rules and recommendations