Dia iabetes care and dementia among old lder adults: a nati tionwide 3-year lo longitudinal stu tudy Wargny M, Gallini A, Hanaire H, Nourhashemi F, Andrieu S, Gardette V UMR 1027 Epidemiology and analyses in public health EUGMS, Nice September 20/22 2017
Conflict of interest disclosure • I have no potential conflict of interest to report
Diabetes and dementia • Co-occurrence diabetes - dementia • 6-39% of patients with dementia suffer from diabetes • Challenges for care • Dementia: a “discordant” comorbidity towards diabetes (Piette and Kerr 2006) • Lack of guidelines for clinicians • Impact of dementia on diabetes care • Few studies with discordant results (Zhang 2010, Thorpe 2012, Connolly 2013, Müther 2010) • No data available with a longitudinal perspective
Aim • To assess whether incident dementia affects the frequency of • diabetes-related health service use and • diabetes-related hospitalizations in French older adults during the year preceding dementia identification and the 2 following years
Design and population • Longitudinal cohort study • Data source: FRA-DEM cohort study in national health system database (SNIIRAM) • all cases of incident dementia in France in 2011-2012 • Based on 1 of 3 criteria: long-term disease registration (ICD-10 F00-F03, G30, G31) , hospitalisation (ICD-10 F00-F03, G30, G31) , anti-dementia drug (cholinesterase inhibitor, memantine) • 5 year-window free of dementia criteria • a random sample of subjects without dementia matched (1:1) on age, sex, residence area and insurance scheme • ≥ 65 years old • Diabetes mellitus • long-term disease registration (ICD-10 E10-E14) with start date > 2 years before dementia identification
Outcomes • Health resource use related to diabetes control and prevention of complications • Biological monitoring: HbA1c, lipid profile, microalbuminuria • Eye examination • Diabetes-related hospitalisations • All diabetes-related hospitalisations • Diabetes-related coma
Dementia identification (in 2011- 2012) Analysis • Yearly incidences for 3 years Y -1 Y 0 Y 1 • Standardized incidence ratios (SIR) between subjects with and without dementia with 95% confidence intervals • Indirect standardisation on age (5-year classes) • Sensitivity analyses • On the uncensored population • Excluding subjects with dementia identified through hospitalisation
Results • 87 816 subjects (40 117 with dementia) • Mean age ± SD : 81.8 ± 6.7 • 57% female • 86% type 2 diabetes • 50% with cardiovascular disease
At least 2 HbA1c tests in a given year % 80 74,7 73,6 73,4 70 60 50 40 30 20 10 0 Y-1 Y0 Y1 With dementia Without dementia
At least 2 HbA1c tests in a given year % 80 74,7 73,6 73,4 70 65,5 58,8 60 SIR (CI 95% ) 54,6 50 40 30 20 10 0 Y-1 Y0 Y1 Y -1 Y 0 Y 1 With dementia Without dementia
LDL-cholesterol Microalbuminuria % % 80 80 70 70 63,2 62,3 62,0 60 60 56,0 50 50 45,4 43,0 40 40 30 30 21,8 21,5 21,2 20 16,1 20 11,6 10,6 10 10 0 0 Y-1 Y0 Y1 Y-1 Y0 Y1 With dementia Without dementia With dementia Without dementia SIR 0.68 0.72 0.48 0.52 0.89 0.72 CI95% 0.88-0.90 0.67-0.69 0.71-0.74 0.47-0.50 0.50-0.53 0.70-0.74
Eye examination % 80 70 60 49,9 47,7 50 46,5 38,4 40 32,4 29,5 30 20 10 0 Y-1 Y0 Y1 With dementia Without dementia SIR 0.67 0.63 0.77 CI95% 0.75-0.78 0.66-0.68 0.61-0.64
Diabetes-related hospitalizations Incidence / 1000 person- years SIR 3.14 1.67 2.04 CI95% 1.97-2.12 3.04-3.24 1.59-1.76
Hospitalizations for diabetes-related coma Incidence / 1000 person- years 10 9 Similar trends were seen for 8 7 hospitalizations for 6 5,5 hypoglycemia, falls or hip 5 fractures 4 2,6 3 2,0 2 0,7 1 0,6 0,6 0 Y-1 Y0 Y1 With dementia Without dementia SIR 9.30 3.06 3.84 CI95% 3.14-4.65 8.08-10.64 2.38-3.88
Discussion – conclusion • Limitations • Diagnoses are not clinically ascertained and no information about severity • Inpatient monitoring is only partially captured in the database • For all outcomes and years, subjects with dementia were less likely to receive basic minimal diabetes monitoring • Gap worsened over years • Reasons (patients, physicians, caregivers) need to be investigated • 2- to 9-fold increase in risk of hospitalization for diabetes complication • Despite the low burden of monitoring and lack of financial barriers
Dia iabetes care and dementia among old lder adults: a nati tionwide 3-year lo longitudinal stu tudy Wargny M, Gallini A, Hanaire H, Nourhashemi F, Andrieu S, Gardette V adeline.gallini@univ-tlse3.fr
≥ 2 HbA1c dans l’année % SIR (IC 95% ) A -1 A 0 A 1 A -1 A 0 A 1 17
Comparaison des taux de réalisation des examens biologiques entre les 2 groupes ≥ 2 HbA1c dans l’année ≥ 1 Triglycérides dans l’année SIR (IC 95% ) SIR (IC 95% ) A 0 A -1 A 1 A 0 A -1 A 1 ≥ 1 microalbuminurie dans l’année ≥ 1 créatininémie dans l’année SIR (IC 95% ) SIR (IC 95% ) A 0 A 0 A -1 A 1 A -1 A 1 18
Analyse de sensibilité Analyse principale - Population non censurée à A 1 SIR SIR 19
Fréquence des recours aux hospitalisations Fréquence des recours aux hospitalisations Analyse principale Analyse principale SIR SIR 11 11 9 9 8,4 Liée au diabète 7 7 Liée au diabète 5 Acidocétose ou coma 5 4,13 diabétique 3 3 2,88 2,04 2,04 3,14 3,14 1,67 1,67 1 1 A-1 A-1 A0 A0 A1 A1
FRA-DEM MASA NON-EXPOSES Temps N = 352 595 paires FRA-DEM FRA-DEM ALD diabète (A -1 ) 13,5 % DIA-FRA-DEM 11,4 % Date NON EXPOSES MASA N = 40 117 N = 47 699 index DIA-FRA-DEM DIA-FRA-DEM A 0 3907 censurés (8,2 %) 36 323 PA 46 256 PA 8542 censurés (21,3 %) MASA NON EXPOSES N = 31 575 N = 43 792 à 1 an à 1 an 5131 censurés (11,7 %) A 1 5737 censurés (18,2 %) 28 583 PA 41 131 PA MASA NON EXPOSES Au terme N = 25 838 N = 38 661 à 2 ans à 2 ans de A 1 21
Comparaison des délivrances médicamenteuses entre les 2 groupes ≥ 2 délivrances dans l’année (%) Metformine Sulfamides hypoglycémiants Insuline 22
≥ 12 consultations chez un médecin généralistes/an % SIR (IC 95% )
Recommend
More recommend
