An One-Year Static Cross-Sectional Population Model to Assess the Infection Burden Across the Ageing Population: An Application to the Flemish County
Author(s)
Standaert B1, Pham TH2, Topachevskyi O3, Postma MJ4
1Hasselt University, Antwerp, Belgium, 2University Medical Center Groningen, Department of Health Sciences, Groningen, GR, Netherlands, 3Expert Committee on Selection and Use of Essential Medicines, Kyiv, Ukraine, 4University Medical Center Groningen, Department of Health Sciences, Groningen, Netherlands
Presentation Documents
OBJECTIVES: Cohort models are often used to assess disease progression over different health states, for measuring total burden over time. That logic becomes complicated when the population composition is highly dynamic in function of age and sex, with changes in health condition, in decreases of immune responses, being relocated over different healthcare settings with different infection transmission risks. Our objective is to identify an alternative to cohort modelling, which also estimates adequately the infection burden, using a transparent evaluation method, that supports decision makers in their choices of applying vaccination of older adults.
METHODS: We compare the disease burden assessment using 2 different approaches. One is the cohort modelling following people until everyone is dead. The other is the cross-sectional 1-year population approach having the same infectious disease problem. The cohort modelling simulates progression to different infectious disease levels moving people to different settings (home, home care, nursing home, hospital), introducing many assumptions on the probabilities of event occurrence because of missing data. The cross-section evaluation makes the inventory of all infectious events happening at the different settings. It presents a snapshot of the problem based on real observations without patient flows.
RESULTS: Although the 2 approaches can be forced to reach a same accumulated burden result of health events, the cross-sectional approach needs less assumptions, is easier to verify and validate, and facilitates the impact measurement of preventative interventions. The cohort modelling gets complicated when more health states are to be considered. It then becomes driven by assumptions, rather than having real-world data presented.
CONCLUSIONS: When exposed to various dynamics in the population, like age, sex, health condition, place of living, infection’s types, severity, and mortality, having straightforward inventory figures over a fixed time period are more insightful for understanding the problem, rather than using a cohort modelling approach.
Conference/Value in Health Info
Value in Health, Volume 25, Issue 12S (December 2022)
Code
EPH83
Topic
Epidemiology & Public Health
Topic Subcategory
Public Health
Disease
STA: Vaccines