TY - JOUR
T1 - Linking clinical and population-based data in older patients with cancer in Belgium
T2 - Feasibility and clinical outcomes
AU - Depoorter, Victoria
AU - Vanschoenbeek, Katrijn
AU - Decoster, Lore
AU - De Schutter, Harlinde
AU - Debruyne, Philip R
AU - De Groof, Inge
AU - Bron, Dominique
AU - Cornélis, Frank
AU - Luce, Sylvie
AU - Focan, Christian
AU - Verschaeve, Vincent
AU - Debugne, Gwenaëlle
AU - Langenaeken, Christine
AU - Van Den Bulck, Heidi
AU - Goeminne, Jean-Charles
AU - Teurfs, Wesley
AU - Jerusalem, Guy
AU - Schrijvers, Dirk
AU - Petit, Bénédicte
AU - Rasschaert, Marika
AU - Praet, Jean-Philippe
AU - Vandenborre, Katherine
AU - Milisen, Koen
AU - Flamaing, Johan
AU - Kenis, Cindy
AU - Verdoodt, Freija
AU - Wildiers, Hans
N1 - Copyright © 2023 Elsevier Ltd. All rights reserved.
PY - 2023/3
Y1 - 2023/3
N2 - INTRODUCTION: Geriatric screening and geriatric assessment (GS/GA) have proven their benefits in the care for older patients with cancer. However, less is known about the predictive value of GS/GA for outcomes. To research this, clinical data on GS/GA can be enriched with population-based data. In this article we describe the methods and feasibility of data linkage, and first clinical outcomes (GS/GA results and overall survival).MATERIALS AND METHODS: A large cohort study consisting of patients aged ≥70 years with a new cancer diagnosis was established using linked data from clinical and population-based databases. Clinical data were derived from a previous prospective study where older patients with cancer were screened with G8, followed by GA in case of an abnormal result (GS/GA study; 2009-2015). These data were linked to cancer registration data from the Belgian Cancer Registry (BCR), reimbursement data of the health insurance companies (InterMutualistic Agency, IMA), and hospital discharge data (Technical Cell, TCT). Cox regression analyses were conducted to evaluate the prognostic value of the G8 geriatric screening tool.RESULTS: Of the 8067 eligible patients with a new cancer diagnosis, linkage of data from the GS/GA study and data from the BCR was successful for 93.7%, resulting in a cohort of 7556 patients available for the current analysis. Further linkage with the IMA and TCT database resulted in a cohort of 7314 patients (96.8%). Based on G8 geriatric screening, 67.9% of the patients had a geriatric risk profile. Malnutrition and functional dependence were the most common GA-identified risk factors. An abnormal baseline G8 score (≤14/17) was associated with lower overall survival (adjusted HR [aHR] = 1.62 [1.50-1.75], p < 0.001).DISCUSSION: Linking clinical and population-based databases for older patients with cancer has shown to be feasible. The GS/GA results at cancer diagnosis demonstrate the vulnerability of this population and the G8 score showed prognostic value for overall survival. The established cohort of almost 8000 patients with long-term follow-up will serve as a basis in the future for detailed analyses on long-term outcomes beyond survival.
AB - INTRODUCTION: Geriatric screening and geriatric assessment (GS/GA) have proven their benefits in the care for older patients with cancer. However, less is known about the predictive value of GS/GA for outcomes. To research this, clinical data on GS/GA can be enriched with population-based data. In this article we describe the methods and feasibility of data linkage, and first clinical outcomes (GS/GA results and overall survival).MATERIALS AND METHODS: A large cohort study consisting of patients aged ≥70 years with a new cancer diagnosis was established using linked data from clinical and population-based databases. Clinical data were derived from a previous prospective study where older patients with cancer were screened with G8, followed by GA in case of an abnormal result (GS/GA study; 2009-2015). These data were linked to cancer registration data from the Belgian Cancer Registry (BCR), reimbursement data of the health insurance companies (InterMutualistic Agency, IMA), and hospital discharge data (Technical Cell, TCT). Cox regression analyses were conducted to evaluate the prognostic value of the G8 geriatric screening tool.RESULTS: Of the 8067 eligible patients with a new cancer diagnosis, linkage of data from the GS/GA study and data from the BCR was successful for 93.7%, resulting in a cohort of 7556 patients available for the current analysis. Further linkage with the IMA and TCT database resulted in a cohort of 7314 patients (96.8%). Based on G8 geriatric screening, 67.9% of the patients had a geriatric risk profile. Malnutrition and functional dependence were the most common GA-identified risk factors. An abnormal baseline G8 score (≤14/17) was associated with lower overall survival (adjusted HR [aHR] = 1.62 [1.50-1.75], p < 0.001).DISCUSSION: Linking clinical and population-based databases for older patients with cancer has shown to be feasible. The GS/GA results at cancer diagnosis demonstrate the vulnerability of this population and the G8 score showed prognostic value for overall survival. The established cohort of almost 8000 patients with long-term follow-up will serve as a basis in the future for detailed analyses on long-term outcomes beyond survival.
KW - Aged
KW - Humans
KW - Belgium/epidemiology
KW - Cohort Studies
KW - Feasibility Studies
KW - Neoplasms/epidemiology
KW - Prospective Studies
KW - Geriatric Assessment/methods
UR - http://www.scopus.com/inward/record.url?scp=85148680090&partnerID=8YFLogxK
U2 - 10.1016/j.jgo.2023.101428
DO - 10.1016/j.jgo.2023.101428
M3 - Article
C2 - 36804333
VL - 14
JO - Journal of Geriatric Oncology
JF - Journal of Geriatric Oncology
SN - 1879-4068
IS - 2
M1 - 101428
ER -