Developing a core outcome set (COS) for Dementia with Lewy bodies (DLB)

Background: Dementia with Lewy bodies (DLB) is an important cause of dementia with a range of clinical manifestations, including motor, neuropsychiatric, and autonomic symptoms. Compared with more common forms of dementia such as Alzheimer’s disease, DLB has been the focus of significantly fewer treatment studies, often with diverse outcome measures, making comparison and clinical implementation difficult. A core outcome set (COS) can address this by ensuring that data are comparable, relevant, useful, and usable for making the best healthcare decisions. Methods: Using a multi-stage approach, development of the DLB-COS will include the following stages: (1) A systematic review, following PRISMA guidelines to create an initial long list of outcomes; (2) A two-round online Delphi including clinicians, scientists, policymakers, and individuals with lived experience of DLB and their representatives; (3) An online consensus meeting to agree on the final core list of outcomes (the final DLB-COS) for use in research and clinical practice; (4) A systematic review to identify appropriate measurement instruments for the DLB-COS outcomes; (5) A final consensus meeting of the professional stakeholders who attended the online consensus meeting to agree on the instruments that should be used to measure the outcomes in the DLB-COS; and (6) Global dissemination. Discussion: This is a multi-stage project to develop a COS to be used in treatment trials for DLB. A DLB-COS will ensure the selection of relevant outcomes and will identify the instruments to be used to measure DLB globally. Keywords: Dementia, Dementia with Lewy Bodies, Core Outcome Set, Delphi, Systematic Review, Ageing, Cognition, Memory

outcomes (the final DLB-COS) for use in research and clinical practice; (4) A literature search to identify appropriate measurement instruments for the DLB-COS outcomes; (5) A final consensus meeting of the professional stakeholders who attended the online consensus meeting to agree on the instruments that should be used to measure the outcomes in the DLB-COS; and (6) Global dissemination.Discussion: This is a multi-stage project to develop a COS to be used in treatment trials for DLB.A DLB-COS will ensure the selection of relevant outcomes and will identify the instruments to be used to measure DLB globally.

Introduction
Dementia with Lewy Bodies (DLB) and Parkinson's disease dementia (PDD) together constitute 10-15% of cases, the second most common dementia worldwide 1 .DLB is characterized by clinical features including cognitive impairment, parkinsonism, visual hallucinations, REM sleep behaviour disorder (RBD), and fluctuations in cognition 2 .As the global population ages, the prevalence and incidence of DLB is rising, as are the associated healthcare costs 3 .
Considerably fewer studies are conducted in DLB than in Alzheimer's disease (AD) and Parkinson's disease (PD), despite the close biological relationship between these disorders 4 ; fewer specifically investigate interventions or treatments.Comparison of the treatments examined in existing research is complicated by the wide range of clinical outcomes reported 5 .Digital, mobile, and wearable technology for outcome data collection add to this complex picture, creating methodological heterogeneity in how outcomes are measured.This complexifies evidence synthesis of DLB trial data, precludes rigorous meta-analysis, and weakens translation of evidence into clinical care.A core outcome set (COS) is an agreed standardized set of outcomes that should be measured and reported in all clinical trials.Standardization of clinical trial outcomes supports consistent measurement of patient symptoms, decreases healthcare costs, and minimizes bias.This project aims to address the methodological heterogeneity in future DLB clinical trials through development and dissemination of a COS for DLB.In developing this, we will consider the number and type of outcomes measured, and the existence of any standardized data collection methods.We will also ascertain the most appropriate outcome rating tool to represent each outcome in the final COS; however, if more than one candidate tool is available, inclusion will be agreed at the level of the consensus meeting.

Scope
The DLB-COS and the identified measurement instruments are aimed to be used in future research in randomized and non-randomized pharmacological and non-pharmacological intervention studies and in clinical practice.The target population are individuals diagnosed with DLB.

Design
This study will follow methodological principles developed by COMET and COSMIN and with modifications where necessary [6][7][8][9] .The DLB-COS was registered with COMET and is publicly available (https://www.comet-initiative.org/Studies/Details/1963).The study Working Group (WG) will include researchers, clinicians, health economists, and methodologists who will steer and monitor the progress of the study.Any significant changes to the protocol will be communicated to the ethics committee, the journal, and the founders.
This study will involve six distinct stages: 1. Identifying outcomes from a systematic review, and developing a preliminary list 2. Reaching consensus on a preliminary DLB-COS from the perspective of professional and lay stakeholders via two rounds of Delphi surveys; 3. Building on the preliminary list of outcomes (Objective 2), develop a final DLB-COS for use in future DLB research and clinical practice, via an online consensus meeting with professional and lay stakeholders; 4. Identifying and reaching consensus on the most appropriate instruments to measure outcomes in the final DLB-COS, via an online consensus meeting attended by professional stakeholders; 5. Agreeing on the instruments that should be used to measure the outcomes in the DLB-COS via a final consensus meeting of the professional stakeholders who attended the online consensus meeting; 6. Disseminating and promoting the implementation of the DLB-COS globally.

Phase 1: Stage 1: Protocol design and evidence synthesis through systematic review
We conducted a systematic review, building on the narrative review on outcome measures in DLB trials, led by Rodriguez-Porcel et al. 10 .
Our systematic review examined and synthesized evidence from qualitative and health economics studies, as well as that reported by clinical trials.Eligibility criteria for the systematic review of outcomes used in the DLB literature can

Amendments from Version 1
We have updated the protocol paper to address the comments of reviewers.The changes we made were as follows: 1) we changed "literature search" to "systematic review" throughout the paper; 2) we elaborated on ascertaining outcome measurement tools; 3) we corrected the missing stage in the design section; 4) we changed future tenses to past tenses when referring to the systematic review, and added in the PROSPERO link; 5) we got rid of grey literature as an exclusion criteria; 6) we clarified that the number of COS items will not be prespecified, but rather we will adopt a pragmatic approach to ensure that the COS will be clinically useful in a clinical and research context; 7) we clarified that we will pilot the survey among lay stakeholders as well as working group members; 8) we clarified that, when creating the final COS, outcomes falling under the same domain will be grouped into single outcomes; 9) we changed the statement that the consensus meeting will aim to have "proportional representation from the two stakeholder groups" to "all respondents from the two stakeholder groups"; 10) we made amendments to the quality assessment and recommendations section; 11) instead of selecting "only one" outcome measurement instrument per outcome, we will select "the most appropriate" measurement instrument per outcome; and 12) we elaborated the discussion to acknowledge limitations.
Any further responses from the reviewers can be found at the end of the article • Intervention trials with pre-/post assessments regardless of randomization, case reports • Observational studies • Reviews, meta analysis (for further trials)

Interventions
Any pharmacological, surgical Or non-pharmacological approach for treating motor and non-motor symptoms in Individuals with DLB at any disease stage and in any setting (e.g., outpatients as well as inpatients)

Comparators
Any: • None • Placebo/Passive control groups/Wait-list • Qualitative approaches to evaluate intervention success across all domains (e.g., outcomes obtained through patient interviews, focus groups) • Economic evaluation outcomes (cost analysis, cost-effectiveness analysis, cost-utility analysis, cost-benefit analysis) be found in Table 1.The review will be prospectively registered and posted at the International Prospective Register of Systematic Reviews (PROSPERO), and can be found at the following link: https://www.crd.york.ac.uk/prospero/display_ record.php?RecordID=346808.

Search strategy
The search identified studies through bibliographic databases, trial registers and the grey literature.Bibliographic Databases and trial registers included the following: Medline Ovid (1946-present); EMBASE (1974-present); PsycInfo (1806-present); CINAHL (1981-present); CENTRAL; Web of Science and specific economic databases including NHS EED and EconLit.
Studies were identified using an elaborated search string including keywords regarding the population and the types of studies that will be covered by the COS.

Study selection and data extraction
Study selection and data extraction will be performed according to PRISMA reporting guidelines.

Reporting the outcomes
As recommended by Williamson et al. 6 , the outcome matrix recommended by the ORBIT project 11 will be used to display the outcomes reported in eligible studies.This can demonstrate the inconsistency of outcomes measured to date and identify potential outcome reporting bias.

Phase 2: Stage 2: A 2-round online Delphi survey Delphi Technique and Design
An e-Delphi survey will be used to reach consensus for the final list of COS for DLB.The outcomes will likely include commonly reported outcomes addressing the key domains affected by DLB: functioning and quality of life, motor and non-motor parkinsonisms, cognitive ability and fluctuations, health economic outcomes, and psychiatric and sleep-related symptoms 10 .The number of COS items will not be prespecified, but we will adopt a pragmatic approach to ensure that the COS will be clinically useful in a clinical and research context.The survey will be completed by professional and lay stakeholders.Delphi technique is a widely used approach applied to elicit consensus from domain experts regarding real-world knowledge and defined clinical issues for which no previous consensus existed [12][13][14] .This process gathers information from multiple stakeholders while maintaining anonymity and minimizing the challenges of group dynamics among experts 6,9,[15][16][17] .The administration of e-Delphi usually involves at least two series of questionnaires (referred to as "rounds"), after which structured feedback is provided to all participants.Then, an online or face-to-face meeting takes place to reach consensus 6,11 .Even though this component was not a part of the original Delphi process 11 , it was adopted from the modified e-Delphi procedure, which allowed for experts' interaction to reach a final consensus 18,19 .This method has proven to be effective 17,20,21 .
The e-Delphi will be administered through the DelphiManager software (http://www.comet-initiative.org/delphimanager/), with each round lasting three weeks and reminders sent at 14 and 18 days.The data from each round will be analyzed and presented to all participants in the subsequent e-Delphi round.
Prior to initiation of the first e-Delphi round, the questionnaire will be piloted among work group members as well as lay stakeholders to assess its validity and clarity.

Delphi panel participant composition and selection
There is no consensus on the recommended sample size for a Delphi study 7,[22][23][24] and it is common practice to use the existing literature as a guiding example [23][24][25][26] .We will recruit different groups of lay and professional stakeholders representing the DLB field.We will include enough participants so at least two representatives from each subgroup can attend the consensus meeting 7,23,24 .

Eligibility
(1) Professional respondents will include geriatric psychiatrists, neurologists, geriatricians, general practitioners, nurses, psychologists, occupational therapists, health economists, researchers, neural engineers, pharmaceutical industry representatives, and representatives of drug regulatory authorities.Should they know other DLB experts, they can nominate them 27 .
(2) Lay stakeholders will include people with DLB and their care partners or supporters.They will be recruited through relevant civic society organizations.

Delphi rounds and consensus criteria
Following previous Delphi studies 17,23,24,28,29 , survey respondents will assess the importance and meaningfulness of the outcomes.Moreover, following the design adopted by the studies that included people with lived experience of the disease 22,26,[30][31][32] .At the end of the first e-Delphi round, the participants will have the opportunity to suggest additional outcome domains to be included in the second e-Delphi round (open text option) 6,9,22,24,26 .

Rating scale
In line with the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) 33 and with the RAND appropriateness method 34 , each outcome will be scored on a 9-point Likert scale, where 1 designates the lowest and 9 the highest score.Overall, scores from 1 to 3 will be defined as 'not important', 4 to 6 will be defined as 'important but not critical', and scores from 7 to 9 will be defined as 'critical for inclusion'.
The criteria dictating inclusion of outcomes for both versions will follow the '70/15' consensus approach 6,23,24,28,29,35 .This is defined by at least 70% of stakeholders scoring an outcome between 7-9 and less than 15% scoring it between 1-3.Outcomes which receive at least 70% of scores between 1-3 and less than 15% of scores between 7-9 will not be included in the second e-Delphi round.These thresholds are based on the common agreement that the outcomes constituting the final COS are regarded as critical for inclusion, with a clear minority of stakeholders deeming them 'not important' 6,35 .Note, outcomes falling under the same domain will be grouped into single outcomes.

Stage 3: Consensus meeting finalizing LBD-COS
Some participants who complete both e-Delphi rounds will be invited via e-mail to an online consensus meeting, hosted on Zoom.The meeting will aim to have all respondents from the two stakeholder groups.They will discuss, vote, and agree on the final outcomes.Each outcome will be rated according to four additional criteria 36 : • Frequency of the outcome in people with DLB; • Impact of the outcome on people with DLB; • Preventability/treatability of the outcome; • Feasibility to address the outcome in clinical practice and research intervention studies.
Consensus will include at least one lay stakeholder voting for inclusion of the outcome, as established by Wuytack et al. 24 .

Phase 3: Stage 4: Systematic review identifying measurement instruments to be used in the core outcome set
We will undertake a systematic review for the instruments measuring specific outcomes, following COSMIN guidelines 37 .We will use the following databases: Medline, PubMed, EMBASE, PsycInfo, CINAHL, and CENTRAL.We will use the key sources of measurement in DLB, such as the DIAMOND Lewy toolkit 38 and the Movement Disorders Society Recommendations for measurement tools (https://www.movementdisorders.org/).

Quality assessment and recommendations
In developing our COS protocol, we will follow the steps outlined by the COSMIN checklist to ensure quality of the procedure 39 .Recommendations for outcome measurement instruments will be selected based on the preliminary DLB-COS, according to the quality of evidence and feasibility of the outcomes to be used in research and clinical practice.

Stage 5: Final consensus meeting finalizing the choice of instruments measuring the core outcome set
A final consensus meeting attended by professional stakeholders will take place to discuss, vote, and agree on the instruments measuring the selected DLB-COS.We aim to select the most appropriate outcome measurement instrument per outcome.Consensus will be defined as at least 70% of participants voting for inclusion of a measurement instrument 24 .

Stage 6: Reporting and dissemination
The development of this DLB-COS will be reported based on the Core Outcome Set-Standards for Reporting (COS-STAR) guidelines 8 .
Our dissemination plan will leverage the wider professional, civic society, and lived experience networks in this area.Professional dissemination will include participation in international Lewy Body conferences and dissemination through the International Alzheimer Association ISTAART PIA group.Dissemination through civic society organisations will include Lewy Body Society (LBS), Lewy Body Ireland (LBI), and Lewy Body Dementia Association (LBDA).They will help us by sharing information on their websites and social media, using accessible language, and highlighting the contribution of PPI and the third sector partners.
The systematic review is registered and posted at the International Prospective Register of Systematic Reviews, (PROSPERO 2022 CRD42022346808).
The development of this DLB-COS will be reported based on the Core Outcome Set-Standards for Reporting (COS-STAR) guidelines 8 .

Ethical approval
Since we will be working with sensitive information and including people with lived experience in our work, we will apply for ethical approval from the TCD Faculty of Health Sciences Research Ethics Committee.We will also complete a Data Protection Impact Assessment (DPIA), following General Data Protection Regulation (GDPR) guidelines.Professional and lay participants will receive information about the study via email.If they choose to participate, prior to the first e-Delphi round, they will receive and sign a consent form.The e-Delphi responses will be confidential, and participants will have the right to withdraw at any point prior to data analysis.

Discussion
Currently, no COS for DLB exists.Since there are very few DLB intervention and clinical research studies, a well-developed and globally disseminated DLB-COS for research and clinical use is required.It is imperative that the measured outcomes have relevance to people with DLB and their care partners.This COS will include the views of a wide range of stakeholders.By developing a standardized COS and ensuring that outcomes are measured with appropriate instruments we aim to increase trial efficiency, improve evidence synthesis, reduce research waste, and improve the development of interventions for people with DLB.
However, we acknowledge that the online nature of the survey is a potential limitation in our methods as this may disadvantage people with limited digital access.Nonetheless, increasingly, online Delphi surveys are acceptable and reflect the experience of the majority of stakeholders.We have had to balance the practical elements of completing the surveys and obtaining transnational representation with the need to include under-served groups.

Study status
At present, our systematic review has been completed and this has been used to develop a long list of outcomes to be rated in our e-Delphi survey.The e-Delphi survey has been composed and is ready to go live once ethical approval is received.

Data availability Underlying data
Underlying data which is relevant to this protocol include the following studies; -Rodriguez-Porcel, F., Wyman appropriate for the research question, though there is a lack of detail in the reporting in parts.

Systematic review
The systematic review to identify the long-list of outcomes is appropriately registered with PROSPERO.The authors report that the review has now been fully completed therefore it is not helpful to comment on the methods as these cannot be changed.I have reservations about whether it is appropriate to describe this review in the future tense in the paper given it is fully completed.A much briefer account with a link through to the PROSPERO registration may be more appropriate -I will leave this for editorial direction.If the review is retained then my comments are as follows: overall the review is appropriate and a relevant range of databases are searched.
○ more detail needs to be added on what the plans were for synthesis of data, in particular the qualitative data -how was it planned to extract this data and what synthesis method was planned to generate from the qualitative data.This could be moved across from the PROSPERO registration only as it would not be approriate to add information post-hoc.
○ Table 1 says grey literature was excluded but the search strategy says studies were identified through grey literature.
○ excluding study protocols for ongoing studies seems a missed opportunity in terms of currency of outcome included.

○
PRISMA is a reporting guideline not a methods guideline so avoid statements such as saying the review methods followed PRISMA.

Other aspects of design
It is good to see the study registered with COMET.The link needs to be checked as I could not get it to work -though I did find the registration by a search of COMET.
○ How will outcomes be prepared for presentation in the delphi -is it just a raw list or will there be any consolidation or grouping?

○
It is good to see a plan to pilot the delphi survey with the Working Group before circulating it.However there does not appear to be any patients or carers on the Working Group.It is important to pilot the delphi with patients and carers to ensure that the terminology is understandable.

○
It is good to see a pre-specified definition of consensus for the survey and offering the opportunity to suggest new outcomes at round 1.However, there does not appear to be a plan to account for different priorities amongst patients and carers compared to healthcare professionals during the two survey rounds or to report on this.This could be clarified and justification added for the specific approach used.

○
Consensus meeting -what is meant by 'proportional representation from the two stakeholder groups'?How will this be determined and what will be done to ensure the patient and carer perspective is adequately represented?Some thought has been given to this in that, at the consensus meeting, consensus will require at least one lay stakeholder ○ voting for inclusion of an outcome.However, this seems quite late on in the process, arguably it is a low bar, and is there a risk of patients important to patients being voted out in the earlier stages?Again justification and rationale for methods used would strengthen the paper.
'feasibility to treat the outcome in clinical practice' -how is this being defined and does it include outcomes that could be addressed in non-clinical settings such as social care?
○ Phase 3 Stage 4 -this is described as a literature search in the Abstract and in the section heading in the methods.Yet, in the text it says 'we will undertake a systematic review'.These are not the same thing and this needs clarified.If it is a systematic review I would expect much more detail on the methods.It is not enough just to say that COSMIN guidelines will be followed as these are not entirely prescriptive.

○
The statement 'We will conduct a quality assessment of our COS development process using the COSMIN checklist' is unclear.

○
The Design section refers to 'six distinct stages' but 5 are listed  The design of this study, in which a questionnaire is formed from the current literature and then placed through two rounds of consensus is an efficient and effective version of the Delphi method for obtaining a comprehensive set of the most impactful symptoms in DLB.

Parichita Choudhury
Starting with an openended questionnaire, a common form of the Delphi method, can lead to superfluous and irrelevant topics, thus prolonging the process.One of the major flaws of the Delphi method is retaining participants through the multiple rounds, and so by taking on the research by reviewing the literature themselves, the authors will be effectively reducing one round from the Delphi process that may yield a loss of participants, as well as ensuring that the first round of questionnaire is of high impact.Additionally, the two rounds of consensus, with a consensus percentage of 70%, will not only include researchers, but also patients, practitioners, and care partners, so all involved parties will have a sufficient voice when approaching decisions regarding what is most important in the disease trajectory, therefore leading to outcomes that truly matter to all parties.
There are some limitations to this method.One limitation is the omission of biomarkers.Biomarkers are becoming an essential aspect of diagnosing neurocognitive disorders including DLB, but this is an acceptable exclusion as time is a limitation, and the inclusion of biomarkers may further complicate the process.Another limitation is regarding the selection of the outcomes themselves for the questionnaire.There was no mention of how many outcomes are intended to be surveyed once placed through the ORBIT matrix.If there is no cap, there may need to be some flexibility to increase the number of rounds needed during the Delphi surveys if question volumes are high.Finally, the implementation methodology (an online Delphi) disadvantages stakeholders from under-represented and socio-economically underprivileged backgrounds.
The development of core outcome sets for DLB is an important undertaking.The five stages presented to obtain these core outcome sets are concise and deliberate.The protocol in its current form is, therefore, publishable, despite some limitations.
Is the rationale for, and objectives of, the study clearly described?Yes

Is the study design appropriate for the research question? Yes
Are sufficient details of the methods provided to allow replication by others?Partly

Are the datasets clearly presented in a useable and accessible format? Yes
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Neurodegenerative disease, Dementia with Lewy bodies, Alzheimer's Disease We confirm that we have read this submission and believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
Banner Sun Health Research Institute, Sun City, AZ, USA Andrew Ho Banner University Medical Center Phoenix, Phoenix, AZ, USA A number of clinical trials for Dementia with Lewy bodies (DLB) have been completed and more are coming for both disease-modifying therapies and symptomatic therapies.Currently, there are no outcome measures specific to DLB.Previous trials have borrowed from the Alzheimer's disease or Parkinson's disease outcome measures.Hence, developing a core outcome set for Dementia with Lewy Body is a high-yield undertaking and is essential for patients who live with this disease, as well as their practitioners and care partners who support them.In this article, Grycuk et al. provide an overview of a protocol aimed at developing core outcome measures, which encompasses the varied phenotypes and presentations of DLB.

the rationale for, and objectives of, the study clearly described? Yes Is the study design appropriate for the research question? Yes Are sufficient details of the methods provided to allow replication by others? Partly Are the datasets clearly presented in a useable and accessible format? Not applicable Competing Interests: No
○Is competing interests were disclosed.Reviewer Expertise: Applied health research, systematic reviews, RCTs, outcomes I

confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
No competing interests were disclosed.

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
https://doi.org/10.21956/hrbopenres.14846.r32799© 2022 Choudhury P et al.This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.