Rapidrecommenda-ons,point-of-caredecisionaids,andshareddecision-making

June7,2018

HHS-MedicalGrandRounds

ThomasAgoritsas,MD,PhD

@ThomasAgoritsas

SeniorConsultantDivisionofInternalMedicine,UniversityHospitalsofGenevaSwitzerland

AssistantProfessor,DepartmentofHEI,McMasterUniversity

ReproducedfromcoverpageofJAMA,Users’GuidetotheMedicalLiterature,3rded.

I have no financial conflict of interest in relation to this presentation. My intellectual conflict of interests:

Ø  Member of the GRADE Working Group http://www.gradeworkinggroup.org

Ø  Deputy editor ACP journal club – McMaster PLUS Evidence Alerts

Ø  Member of the MAGIC organization http://magicproject.org

A non-for-profit initiative to improve the creation, dissemination,

and dynamic updating of guidelines, evidence summaries and decisions aids.

Ø  Co-founded the BMJ RapidRec http://www.bmj.com/rapid-recommendations

Disclosures

1.  IntroducingtheEvidenceEcosystem2.  Problemswithcurrentguidelines3.  PotenHalsoluHons:theBMJRapidRecs

–  Methodology–  Examples

4.  GuidelinesandSharedDecisionmaking5.  BacktotheenhancingtheEvidenceEcosystem

Plan for presentation

Evidence Ecosystem: challenges and opportunities

Evidencedisseminatorstoclinicians

Actorsanddataflow

Evidenceevaluators&improvers

Evidencedisseminatorstopa-ents

Evidenceimplementers

Evidenceproducers

Evidencesynthesizers

Brandt,Agoritsas,etal.[inpreparaNon]

6

Criticism on guidelines

@ThomasAgoritsas

Ø  3000-4000 publications / day Ø  75-100 RCT & 10-20 SR Ø  60% of clinical questions are not

informed by current best evidence

www.g-i-n.net/library/international-guidelines-library

>9400Recommanda-ons

Evidencesummary+RecommendaHon

Evidencesummary

GroupA GroupB

EvidencesummaryEvidencesummary+RecommendaHon

Scenario2

*Orderofscenariosalsoatrandom

Scenario1

Doclinicianswantrecommenda8ons:aRCT

NeumannI,Alonso-CoelloP,VandvikPO,AgoritsasT,etal.JCE2018.

StrongRecommenda-ons

1 2 3 4 5 6 7

Doclinicianswantrecommenda8ons:aRCT(2)

NeumannI,Alonso-CoelloP,VandvikPO,AgoritsasT,etal.JCE2018.

Institute of Medicine (IOM) – 2011 Trustworthiness standards (25 items)

1.  Establish transparent process 2.  Manage conflict of interest (COI) 3.  Panel composition: balanced,

multidisciplinary, including patients 4.  Based on SR for each question 5.  Clarify the “ingredients” for each

recommendation •  Summaries of benefits and harsm •  Quality of the evidence (or lack thereof) •  Role of values and preferences

6.  Articulation of the recommendation : •  Clarity, strength, rationale

7.  External review, patient involvement 8.  Updating strategy AGREE

Financial COI •  71% of guideline chairs •  91% of co-chairs

Patients included – 15%

Kung et al. Arch Intern Med. 2012 * Evaluation on 18 criteria (from 25) – N=130 guidelines

Lack of transparency

14

BMJ RapidRecs (WikiRecs with other journals)

hFp://www.bmj.com/rapid-recommenda-ons

MAGIC & RapidRec: A wonderful collaboration

Per Olav Vandvik

Head of MAGIC Reed Siemieniuk Gordon Guyatt Lyubov Lytvyn Linn Brandt Anja Fog Heen Thomas Agoritsas

Frankie Achille Deno Vichas Frankie Achille Annette Kristiansen Christopher Berntzen Romina Brignardello Alfonso Iorio

Fiona Godlee Helen Macdonald Elizabeth Loder Will Stahl-Timmins Duncan Jarvies

The BMJ

Sophie Cook

BMJ RapidRecs – 90 days objective

Gather the panel

Ø Patients

Ø Clinicians Ø General (GP, Family docs, pediatricians)

Ø Experts

Ø Methodologists Ø EBM Ø SDM

Management of conflict of interest

Ø Financial COI: excluded

Ø Intellectual COI: balanced

Ø Methods editor

Ø Approval by both RapidRec executive & BMJ executive

Patients (citizens) involvement

Ø No COI Ø Gets individual training and support

Ø Recruitment through: –  CiHzensUnitedforEvidence–  SocietyforParHcipatoryMedicine–  CochraneConsumers/TaskExchange–  RelevantorganizaHons(e.g.InternaHonalCommunityofWomenlivingwithHIV)–  Twi^er–  Referrals

Ø  EachRapidRecs:3-5pa-entsperpanelØ  Chairinvitesthemtotalkfirstateachround

Systematic review(s) teams

Ø Semi-independent Ø Panel input (incl. Patient) before start and in the end.

Ø  1 to 3 reviews per RapidRec, on Ø  Treatment benefits & harms

Ø  Baseline risk

Ø  Values en preferences

Ø  Minimally important difference

1.  Close balance Ø  Close call between benefits

and risks/hassle/cost

Ø  Therefore more preference-sensitive

2.  Lower certainty in estimates

3.  Patients values & preferences: Ø  choice varies appreciably

(or is very uncertain)

1.  Clear balance Ø  benefits clearly outweigh risks/

hassle/cost

Ø  risk/hassle/cost clearly outweighs benefits

2.  Sufficient certainty in estimates (high or moderate)

3.  Patients values & preferences: Ø  almost all same choice

Strong recommendations

Just do it Shared

decision making

Weak recommendations

1.  Close balance Ø  Close call between benefits

and risks/hassle/cost

Ø  Therefore more preference-sensitive

2.  Lower certainty in estimates

3.  Patients values & preferences: Ø  choice varies appreciably

(or is very uncertain)

1.  Clear balance Ø  benefits clearly outweigh risks/

hassle/cost

Ø  risk/hassle/cost clearly outweighs benefits

2.  Sufficient certainty in estimates (high or moderate)

3.  Patients values & preferences: Ø  almost all same choice

Strong recommendations Weak recommendations

24

•  Just retired •  Hyperlipidemia, treated •  New shortness of breath •  Fainted twice during the last month

while exercising à Severe aortic stenosis

Jonathan, 67 years old

TAVI Transcatheter Aortic Valve Insertion

Treatment options

SAVR

Surgival Aortic Valve Replacement

27

28

Istheevidenceapplicabletoourpa-ent?

31

32

Weak recommendation in favour of surgery instead of TAVI? What would you do in his place?

What should Jonathan do?

Clinicians Patients

Evidence

Evidence dissemination to patients

Consulta-onDecisionAids

Pa-entDecisionAids

@ThomasAgoritsas

SharedDecisionMakingapa-entandaclinician

worktogether,

haveaconversa-on,

partnerwitheachother

toidenHfythebestcourseofac-on,thebesttreatmentortest

atthispointinHme.

ItisaaboutsharingwhatmaFersCliniciansshareinformaHonaboutthealternaHves,benefits,harms

PaHentssharepriorexperience,goals,expectaHons,values.Victor Montori

is a process by which

What one wants to see is…

37

PrintableDecisionAidstotakehome

Guidelines & SDM together Is harmony possible?

GlobalEvidenceSummit2017–CapeTown

Promoteevidence-informedchoiceWhatevertheevidence,valueandpreferencejudgmentsarecentralineverydecisionResultinCAREthateachpaHentvalues

EBHC

SDM

@ThomasAgoritsas

Strength of the recommendations The example of UpToDate (n=9451)

Agoritsas,Merglen,Heenetal.UpToDateadherencetoGRADEcriteriaforstrongrecommendaNons:ananalyNcalsurvey.BMJOpen.2017

1/32/3

Barnett et al. Lancet 2012; 380: 37–43

Back to the Evidence Ecosystem

Evidencedisseminatorstoclinicians

Actorsanddataflow

Evidenceevaluators&improvers

Evidencedisseminatorstopa-ents

Evidenceimplementers

Evidenceproducers

Evidencesynthesizers

Brandt,Agoritsas,etal.[inpreparaNon]

Conclusion

Ø 7 RapidRecs produced in 2 ans (april 2016)

Ø 6 more RapidRec in preparation en 2018

Ø Positive feedback & collaborations

Ø Future Ø Evalutate process & Standardize methodology

Ø Sustainability & Scaling up

Ø  Integration into Evidence Ecosystem & Implementation

Contact: thomas.agoritsas@gmail.com

Thankyou!

@ThomasAgoritsas