Published core outcome sets
Twenty completed COS development studies were identified (table S6)
(17-20, 84-99) including: 13 (65%) in pregnancy and childbirth, 3
(15%) in newborn and neonatal health, 2 (10%) in benign gynaecology
and subfertility and 2 (10%) in oncology. Sixteen studies (80%) were
developed in an international context. All used the modified Delphi
method to identify core outcomes. Delphi survey participants included
healthcare professionals (range 34–154 participants), researchers
(range 0-53), and patients (range 0–244 participants). Total number of
participants included in the Delphi process ranged from 24 to 412.
Twelve (60%) studies discussed the limitation of attrition bias, 2
comparing the mean results of non/responders between rounds to determine
effect and 1 performed an attrition analysis to determine effect.
Seventeen (85%) arranged consensus development meetings to finalize the
core outcome set. Consensus meeting participants included healthcare
professionals (range 5–17 participants), researchers (range 2-10), and
patients (range 2–10 participants). All publications identified the
study as a COS development process; the COMET registration number
allowing researchers to identify the study within the database was
stated in 15 (75%). Fifteen (75%) clearly defined their scope,
including the health condition, intervention and setting for which the
core outcome set was applied. Thirteen (65%) adequately described their
methods identifying potential core outcomes; deviations from the
protocol were reported in six (30%). Eighteen (90%) involved patients
and or patient representatives as stakeholders in the consensus process.
Four (20%) used qualitative research methods (interviews or systematic
reviews) to explore the views of patients when generating the initial
list of outcomes for inclusion in the consensus process. The number of
outcomes entered into the Delphi process ranged from 15 to 263, with
final number of core outcomes ranging from six to 48. Overall four COS
(20%) fulfilled full criteria for COS‐STAD and three (15%) for
COS-STAR. (table S3)