Page 249 - 20211214_OHKF_Health_Finance_Research_Report_E (1)

APPENDIX N: ASSUMPTIONS LIST USED IN CREATING
ECONOMIC PROJECT MODEL

1. The modelling approach follows a similar structure between the three tabs/scenarios.
There are three main scenarios as follows:
a. Base Scenario: The Base Scenario will model the costs of diabetes treatment
over a thirty-year horizon, assuming that no population-wide programmes for
screening or risk management are implemented to establish a baseline against which
the impact of the programmes could be measured.
b. Scenario 1: Scenario 1 projects the costs and hence, savings over the same
period if a screening programme were implemented and adopted by the target
population over the said time horizon.
c. Scenario 2: Based on Scenario 1, Scenario 2 will model the costs and hence,
savings if a Risk and Assessment Management Programme–Diabetes Mellitus
(RAMP-DM) programme was implemented in conjunction with the screening
programme.
2. The model will use a 30-year horizon, beginning in 2022 and ending in 2051.
3. The target age demographic is those between 45–54 years of age at the start of
the programme. The model will utilise a closed prospective cohort, with the only “exit”
from the model being observed deaths. Based on population projection models, there will
be a total of 1.128 million individuals in the age group in year 2022.

Disease Prevalence & Incidence:

4. There is homogenous prevalence within each subdivided category of DM and all
complications, regardless of age, gender, predispositions/risk-factors etc.
a. Data is from 2014/15 Population Health Survey (CHP) based on a randomised sample
collected in 2013;
b. Data provides:
i. Prevalence of Diabetes (DM) from individuals with prior diagnosis (2.5%)

ii. Prevalence of DM from individuals without history of DM (sans diagnosis) (4.9%)
iii. Prevalence of prediabetes (IFG/GT) (1.2%)
5. The model will utilise a series of “flow” parameters, measuring the shifts in population
health within each scenario:
a. The cumulative incidence of prediabetes from a “healthy” population is 0.99%
(Quan et al., 2016);
b. Within the Base Scenario, the cumulative incidence of diabetes from a population
with pre-diabetes is 3.73%, based on the Japan Ningen Dock study (Okada et al.,
2017);
c. The prevalence of diabetes for Scenarios 1 and 2 will increase by an annual
cumulative incidence of 2.80% (Okada et al., 2017);
d. The model will also account for regression in all three scenarios from prediabetes
back to “healthy”, using a cumulative incidence of 2.82% (Paprott, 2018);
e. The model will also account for remission in Scenarios 1 and 2 from diabetes back to
pre-diabetes, using a cumulative incidence of 7.0% (Ried-Larsen, 2019).

247