Tom E Robinson1, C Raina Elley2, Tim Kenealy2, Paul L Drury3. 1. School of Population Health, University of Auckland, New Zealand. Electronic address: trob006@aucklanduni.ac.nz. 2. School of Population Health, University of Auckland, New Zealand. 3. Auckland Diabetes Centre, Auckland District Health Board, New Zealand.
Abstract
AIMS: Type 2 diabetes is common and is associated with an approximate 80% increase in the rate of mortality. Management decisions may be assisted by an estimate of the patient's absolute risk of adverse outcomes, including death. This study aimed to derive a predictive risk model for all-cause mortality in type 2 diabetes. METHODS: We used primary care data from a large national multi-ethnic cohort of patients with type 2 diabetes in New Zealand and linked mortality records to develop a predictive risk model for 5-year risk of mortality. We then validated this model using information from a separate cohort of patients with type 2 diabetes. RESULTS: 26,864 people were included in the development cohort with a median follow up time of 9.1 years. We developed three models initially using demographic information and then progressively more clinical detail. The final model, which also included markers of renal disease, proved to give best prediction of all-cause mortality with a C-statistic of 0.80 in the development cohort and 0.79 in the validation cohort (7610 people) and was well calibrated. Ethnicity was a major factor with hazard ratios of 1.37 for indigenous Maori, 0.41 for East Asian and 0.55 for Indo Asian compared with European (P<0.001). CONCLUSIONS: We have developed a model using information usually available in primary care that provides good assessment of patient's risk of death. Results are similar to models previously published from smaller cohorts in other countries and apply to a wider range of patient ethnic groups.
AIMS: Type 2 diabetes is common and is associated with an approximate 80% increase in the rate of mortality. Management decisions may be assisted by an estimate of the patient's absolute risk of adverse outcomes, including death. This study aimed to derive a predictive risk model for all-cause mortality in type 2 diabetes. METHODS: We used primary care data from a large national multi-ethnic cohort of patients with type 2 diabetes in New Zealand and linked mortality records to develop a predictive risk model for 5-year risk of mortality. We then validated this model using information from a separate cohort of patients with type 2 diabetes. RESULTS: 26,864 people were included in the development cohort with a median follow up time of 9.1 years. We developed three models initially using demographic information and then progressively more clinical detail. The final model, which also included markers of renal disease, proved to give best prediction of all-cause mortality with a C-statistic of 0.80 in the development cohort and 0.79 in the validation cohort (7610 people) and was well calibrated. Ethnicity was a major factor with hazard ratios of 1.37 for indigenous Maori, 0.41 for East Asian and 0.55 for Indo Asian compared with European (P<0.001). CONCLUSIONS: We have developed a model using information usually available in primary care that provides good assessment of patient's risk of death. Results are similar to models previously published from smaller cohorts in other countries and apply to a wider range of patient ethnic groups.
Authors: D Martinez-Laguna; X Nogues; B Abrahamsen; C Reyes; C Carbonell-Abella; A Diez-Perez; D Prieto-Alhambra Journal: Osteoporos Int Date: 2017-07-25 Impact factor: 4.507
Authors: Rohan Khera; Martijn J Schuemie; Yuan Lu; Anna Ostropolets; RuiJun Chen; George Hripcsak; Patrick B Ryan; Harlan M Krumholz; Marc A Suchard Journal: BMJ Open Date: 2022-06-09 Impact factor: 3.006
Authors: Ali Aminian; Alexander Zajichek; David E Arterburn; Kathy E Wolski; Stacy A Brethauer; Philip R Schauer; Steven E Nissen; Michael W Kattan Journal: Diabetes Care Date: 2020-02-06 Impact factor: 17.152