Endocrine Abstracts

JOINT1464

Introduction: The completeness of data within a registry can be directly linked to the quality and reliability of data within a registry and this can directly impact the accuracy of research findings generated from data within that registry. The project focus is to evaluate the completeness of the dataset within the GloBE-Reg registry (https://globe-reg.net/).

Methods: Core data variables were categorised into ‘mandatory’ (n = 15), and ‘non-mandatory’ fields (n = 12). Completeness was assessed as a median score of completeness per patient and an overall median percentage completeness for that category. To accurately identify whether or not a variable had been completed, any field left blank or ‘unknown’ was taken as incomplete.

Results: The analysis examined 2,435 cases that were enrolled in GloBE-Reg registry between October 2022 and December 2024. These cases were from 28 centres in 18 countries with a median number of cases per centre of 28 (10^th and 90^th centile, 5, 145). All core data fields that were mandatory had a median completeness of 100% (100,100). When the completeness rate for non-mandatory data were categorised, of the 12 non-mandatory fields, 7 (58%) had a completion rate of >75% with a median completeness of 100% (0,100) and the remaining 5 (42%) had a completion rate of <25% with a median completeness of 0% (0,100). The fields which fell into the highest quartile were, Patient follow-up status, Country of birth, Country of usual residence, Data shared for research purposes, Participation in other registries, Name of other registries and registry ID(s). The fields which fell into the lowest quartile were, Can be contacted for patient reported outcomes (PRO), Consent for newsletters, Biobank sample availability, Biobank details, Biobank patient ID. Of the 28 centres, 26 (92%) achieved a median completeness of 100% (0,100) for the non-mandatory data. The median number of cases at these 26 centres was 28 (5, 147). The remaining 2 centres had a median completeness of 50% (0,100), with 34 and 20 cases in these two centres. There was no association between the number of cases and percentage completeness at the centre.

Conclusions: Despite some variability there is generally a high level of completeness of mandatory and non-mandatory fields within GloBE-Reg core data. Future analysis will extend to all variables in the minimum data set within the registry, to provide a comprehensive assessment of registry completeness.