Developing a Set of Core Outcomes for Trials in Hemodialysis: An International Delphi Survey.
Summary (3 min read)
INTRODUCTION
- The enormous investment in biomedical research, particularly in randomized trials, may not have led to the improvements in health that were hoped for.
- It has been estimated that 85% of the worldwide US $240 billion invested in research annually is wasted.
- In nephrology, there has been substantial research investment into hemodialysis (HD), yet survival rates have not improved correspondingly over the past 40 years and quality of life remains poor even compared with patients with many cancers.
- This may be partly attributable to what outcomes are selected and reported in trials – a challenge well-recognized across medical specialties.
- 9,11,12 In HD, biochemical markers such as serum phosphorus, calcium, and parathyroid hormone, are commonly reported but are not strongly and consistently associated with mortality, cardiovascular disease (CVD), or quality of life.
Study design
- The Delphi method is a technique for achieving consensus among a panel of experts.
- This process involves sequential surveys, typically conducted over three rounds, answered anonymously and gives equal influence to all who participate.
- It was first developed by the RAND Corporation in the 1950’s 25 and has since been increasingly used as a valid approach to develop consensus-based core outcomes for clinical trials in various medical specialty areas.
Participant selection and recruitment
- Stakeholders including patients, caregivers/family members, nephrologists, surgeons, nurses, social workers, psychologists, dieticians, pharmacists, policy makers, researchers and industry, with experience or interest in HD were invited to join the Delphi Panel.
- Participants worldwide were eligible if they were aged over 18 years and able to complete an online survey in English-language.
- Using an opt-in, snowballing sampling frame, the authors recruited patients/caregivers through participating hospitals, patient/consumer organizations, and social media listed in Supplementary File S1.
- Participants registered their email on www.songinitiative.org prior to the survey launch.
Data collection
- The 34 outcome domains for the three-round Delphi survey were identified from a systematic review of outcomes reported in trials in HD, stakeholder interviews, and nominal group technique conducted with patients on HD and caregivers.
- The Delphi survey was completed online via LimeSurvey between September and November, 2015.
- Outcomes with a mean and median of less than 7 for patients/ caregivers and health professionals were not included in round 2.
- Participants reviewed the group scores and their own score for each outcome and re-rated the 29 outcomes using the same 9-point Likert scale, also known as Round 2.
- For each outcome, an optional comments box allowed participants to explain reasons for their rating.
Data analysis
- The authors used SPSS (IBM; Version 22.0) to calculate descriptive statistics.
- The authors calculated the median, mean, and proportion of participants (rating 7-9) for each outcome.
- The scores were calculated separately for patients/caregivers and health professionals, with the difference in means considered significant at P<0.05 based on the t-test.
- Any analysis of the total sample was weighted equally between patients/caregivers and health professionals.
- Consensus was defined a priori based on the OMERACT definition using proportion scores.
Definition of consensus
- Outcomes from round 1 with a mean and median score greater than or equal to 7 for patients/caregivers and health professionals were included in round 2.
- This was validated against the proportion of critically important scores (rated 7-9) for each outcome to ensure that important outcomes were not excluded.
- Outcomes from round 2 with a mean and median score greater than 7 for patients/caregivers and health professionals were included in round 3.
- This was validated against the proportion of critically important scores (rated 7-9) for each outcome.
Delphi scores
- The means, medians, and proportion of participants rating the outcome 7 – 9 (critical importance) for each of the 34 outcomes are shown in Supplementary Table S1, also known as Round 1.
- The following outcomes were excluded from round 2 as they had a mean or median score of less than 7 (not of critical importance) among both patient/caregiver and health professional groups in round 1: nausea/vomiting, sexual function, restless legs syndrome, itching and cramps.
- The differences in ratings between stakeholder groups are shown in Figure 3.
DISCUSSION
- The highest-priority outcomes shared among patients/caregivers and health professionals were vascular access problems, dialysis adequacy, fatigue, CVD, and mortality.
- The SONG-HD Delphi survey had a large sample size (five times greater than most similar Delphi surveys to establish core outcomes)23,38,39 and achieved a broad engagement of a range of stakeholders including patients/caregivers, multidisciplinary healthcare providers, policy makers and industry, across 73 countries, and retained a high response rate of 71% by round 3.
- The process provided a transparent and systematic way to gain consensus on the importance of outcomes to consider for a core outcome set in HD.
- The authors recognize that these outcomes may not be shared by those who did not participate in the study, or by individuals or groups within the study.
- To effectively implement a core outcome set in chronic kidney disease, patients/caregivers and health professional participants suggested that outcome measures should be feasible and valid, and take patient priorities into account.
Contributions
- Research idea and study design: all authors; Data acquisition: NE, AT, BM, BH, DCW, PT, SC, TH, WvB, WCW, HT, SY, SM, DCH, JCC; Data analysis/interpretation: all authors.
- Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.
- NE takes responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.
Financial Disclosure
- David C Harris received fees from Roche Myanmar to give a lecture in Myanmar.
- David W Johnson has received consultancy fees, research grants, travel sponsorships and speaker’s honoraria from Baxter Healthcare and Fresenius Medical Care.
- Sreedhar Mandayam has received research funding from Keryx and is on the Advisory Board/consulting for Corp Medix.
- He has received lecture fees from Fresenius, Amgen, Janssen, ZS Pharma and Vifor Fresenius.
- All other authors have no financial interests to disclose.
TABLES AND FIGURES
- Definitions of high rating outcome domains Table 3.
- Mean scores for round 2 and 3 are not available for outcomes that were excluded in those rounds.
- Figure 3. Difference in mean scores between patients/caregivers and health professionals (rounds 1-3) Note: Ordered by difference in mean scores between patients/caregivers and health professionals.
Outcome domain Definition
- Vascular access problems Problems with fistula, graft, or catheter required for dialysis e.g. access infections, bleeding, bruising, pain, discomfort, clotting Death/Mortality Number of people on hemodialysis who have died, risk of death, how long the patient will live Cardiovascular disease Disease of the heart and blood vessels e.g. heart attack, stroke, blockage of blood vessels Dialysis adequacy.
- How well the dialysis cleans the blood, clearance, Kt/V Fatigue/Energy Feeling tired with no energy for weeks, for most of the time Ability to travel Inclusion criteria for the core outcome set based on median, mean, proportion (7-9) and median rank scores Median score ≥.
Did you find this useful? Give us your feedback
Citations
213 citations
Cites background from "Developing a Set of Core Outcomes f..."
...Patients and clinicians can have divergent and sometimes conflicting goals for hemodialysis treatment,with clinicians focused onoutcomes such as mortality and biochemical markers and patients prioritizing their well-being and lifestyle.(9) For example, some patients on homehemodialysis have reported awillingness to trademonths of survival for ability to travel....
[...]
212 citations
113 citations
107 citations
95 citations
Cites background from "Developing a Set of Core Outcomes f..."
...For patients, the ‘umbilical cord’ keeping them alive can be a constant source of stressful experiences [1]....
[...]
References
4,856 citations
1,535 citations
1,069 citations
"Developing a Set of Core Outcomes f..." refers background in this paper
...This may be partly attributable to what outcomes are selected and reported in trials – a challenge well-recognized across medical specialties.(1,8-10)...
[...]
796 citations
646 citations
"Developing a Set of Core Outcomes f..." refers background or methods in this paper
...It was first developed by the RAND Corporation in the 1950’s 25 and has since been increasingly used as a valid approach to develop consensus-based core outcomes for clinical trials in various medical specialty areas.(10,22,26-28) The SONG-HD Delphi process is shown in Supplementary Figure S1....
[...]
...Initiatives to develop core outcomes are seen in rheumatology and oncology, and have demonstrated improvements in consistent reporting of relevant outcomes.(10,23,24) As part of the international Standardized Outcomes in NephrologyHemodialysis (SONG-HD) initiative, this study aimed to generate a consensus-based prioritized list...
[...]
...This has been done successfully by the Outcome Measures in Rheumatology (OMERACT) initiative, which has improved the reporting and relevance of outcomes in rheumatology trials.(10,24) Recent analyses have shown that a higher percentage of trials are now incorporating the OMERACT core outcome set in rheumatology trials....
[...]