Page 261 - 20211214_OHKF_Health_Finance_Research_Report_E (1)

Separately, in conjunction with the five-year cost effectiveness study on the implementation
of RAMP-DM, there was also a parallel five-year effectiveness study on the impact of
RAMP-DM in relation to the development of complications (Wan et al., 2018). While the two
studies use similar research cohorts, there were notable differences in results. Namely, the
Wan et al. study found that the cumulative incidence of complications in the control ‘usual
care’ group was 43.6% (n = 26,718 usual care patients); additionally, the cumulative
incidence of complications in the RAMP-DM group was 23.2% (n = 26,718 RAMP-DM care
patients) (Wan et al., 2018). Using these figures in place of the figures derived from the cost-
effectiveness study by Jiao et al., we substituted the rate of complication development for
Base Scenario and Scenario 1 of 6.20% with a new complication rate of 8.72% (derived from
43.6% divided over 5 years for an annual cumulative incidence rate) and similarly substituted
the complication rate in Scenario 2 with a new rate of 4.64% (derived from 23.2%, split over 5
years) from the prior 3.81%. We can observe from Table 9 that substituting these figures
does not change the conclusion that the implementing a screening voucher and
management Scheme would be ultimately cost-saving. On top of the expected cost-savings,
we note that there are higher cost-savings in both Scenario 1 and Scenario 2 relative to the
results derived from the original parameters, with the cost-savings from Scenario 2
surpassing the cost-savings in Scenario 1. This is a likely consequence of the higher
complication rates, which helps to further elicit the benefits of the screening and RAMP-DM
Scheme.

Table 9
Shifts in costs and savings due to different complication rates

(HKD million) Base Scenario Scenario 1 Scenario 2
Costs over 30 years 51,766.97 34,284.67 34,133.84

Savings over 30 years 17,482.31 17,633.13
% Savings 33.77% 34.06%

Scenario analysis 3b: Multivariate scenario analysis of complication
rates and costs per patient
In conjunction with the univariate scenario analyses of the complication rates, we also
simultaneously conducted multivariate scenario analyses of complication rates and costs of
complications per each patient with one or more complications. While data and figures from
the Wan et al. and Jiao et al. research papers on RAMP-DM implementation in Hong Kong
are the most up-to-date research articles on the costs of healthcare utilisation and of disease
progression within usual Hospital Authority services and RAMP-DM services, we also
acknowledge that there may be more recent and updated figures on diabetes care that are
not publicly available (Jiao et al., 2018; Wan et al., 2018). In lieu of substituting international
figures or data from publicly available research prior to the Wan et al. and Jiao et al. articles,
we conducted a multivariate scenario analysis to approximate potential impact to the
percentage cost savings within the full screening and management Scheme (Scenario 2). For
the rate of complication development, we selected a range of 2.0% to 10.0% with incremental
increases of 1.0%; for cost per patient with one or more complications, we selected a range
of HKD 20,000 to HKD 200,000 at increments of HKD 20,000 (Table 10). The lower range
more closely approximates the average cost per diabetes patient from the Jiao et al. study
on cost effectiveness (Jiao et al., 2018). We believe that these ranges will allow us to reach
sufficient saturation of possible combinations. The possible percentages of cost-savings in
Scenario 2 are presented in the following table (Table 10).

259