TY - JOUR
T1 - Evaluating the harmonisation potential of diverse cohort datasets
AU - Bauermeister, Sarah
AU - Phatak, Mukta
AU - Sparks, Kelly
AU - Sargent, Lana
AU - Griswold, Michael
AU - McHugh, Caitlin
AU - Nalls, Mike
AU - Young, Simon
AU - Bauermeister, Joshua
AU - Elliott, Paul
AU - Steptoe, Andrew
AU - Porteous, David
AU - Dufouil, Carole
AU - Gallacher, John
N1 - Publisher Copyright:
© 2023, The Author(s).
PY - 2023/6
Y1 - 2023/6
N2 - Data discovery, the ability to find datasets relevant to an analysis, increases scientific opportunity, improves rigour and accelerates activity. Rapid growth in the depth, breadth, quantity and availability of data provides unprecedented opportunities and challenges for data discovery. A potential tool for increasing the efficiency of data discovery, particularly across multiple datasets is data harmonisation.A set of 124 variables, identified as being of broad interest to neurodegeneration, were harmonised using the C-Surv data model. Harmonisation strategies used were simple calibration, algorithmic transformation and standardisation to the Z-distribution. Widely used data conventions, optimised for inclusiveness rather than aetiological precision, were used as harmonisation rules. The harmonisation scheme was applied to data from four diverse population cohorts.Of the 120 variables that were found in the datasets, correspondence between the harmonised data schema and cohort-specific data models was complete or close for 111 (93%). For the remainder, harmonisation was possible with a marginal a loss of granularity.Although harmonisation is not an exact science, sufficient comparability across datasets was achieved to enable data discovery with relatively little loss of informativeness. This provides a basis for further work extending harmonisation to a larger variable list, applying the harmonisation to further datasets, and incentivising the development of data discovery tools.
AB - Data discovery, the ability to find datasets relevant to an analysis, increases scientific opportunity, improves rigour and accelerates activity. Rapid growth in the depth, breadth, quantity and availability of data provides unprecedented opportunities and challenges for data discovery. A potential tool for increasing the efficiency of data discovery, particularly across multiple datasets is data harmonisation.A set of 124 variables, identified as being of broad interest to neurodegeneration, were harmonised using the C-Surv data model. Harmonisation strategies used were simple calibration, algorithmic transformation and standardisation to the Z-distribution. Widely used data conventions, optimised for inclusiveness rather than aetiological precision, were used as harmonisation rules. The harmonisation scheme was applied to data from four diverse population cohorts.Of the 120 variables that were found in the datasets, correspondence between the harmonised data schema and cohort-specific data models was complete or close for 111 (93%). For the remainder, harmonisation was possible with a marginal a loss of granularity.Although harmonisation is not an exact science, sufficient comparability across datasets was achieved to enable data discovery with relatively little loss of informativeness. This provides a basis for further work extending harmonisation to a larger variable list, applying the harmonisation to further datasets, and incentivising the development of data discovery tools.
KW - C-surv data model
KW - Cohort
KW - Data discovery
KW - Data harmonisation
KW - Data visualisation
KW - Datasets
UR - http://www.scopus.com/inward/record.url?scp=85153611238&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85153611238&partnerID=8YFLogxK
U2 - 10.1007/s10654-023-00997-3
DO - 10.1007/s10654-023-00997-3
M3 - Article
C2 - 37099244
AN - SCOPUS:85153611238
SN - 0393-2990
VL - 38
SP - 605
EP - 615
JO - European Journal of Epidemiology
JF - European Journal of Epidemiology
IS - 6
ER -