Page 195 - 20211214_OHKF_Health_Finance_Research_Report_E (1)

iv) Costs of care for prediabetic patients
Finally, our model projects that the size of the population with prediabetes will grow
significantly if screening and management for DM were widely implemented. Under the
current Hong Kong Reference Framework for Diabetes Care guide, individuals with
prediabetes do not receive care or management for their prediabetes beyond more regular
DM rescreening. However, given that prediabetes is a high-risk state for DM, we also
explored the impact on cost should an additional management component for prediabetes
be added. To do so, we factored an additional HKD 350, or the average cost of one doctor
visit per the GOPC-PPP subsidy, for each patient with prediabetes. We found that cost-
savings would be reduced to 24.73%. While this marks a decrease in the expected cost-
savings of the Scheme, it is possible that this addition may further reduce the rate of
transition from prediabetes towards DM, and therefore, the Government
should consider subsidising prediabetes management services within the
Scheme.

Limitations

Our model does not include initial set-up costs for the Scheme for DM or the RAMP-DM
Scheme within the private sector. We acknowledge this limitation in the projection and
recognise that the cost projections may therefore be insufficiently representative of the cost
in the initial years of implementation. However, insufficient data on private sector participation
in PPPs of this capacity renders it difficult to estimate set-up costs. As such, we recommend
that the Government complete an assessment of private sector doctor,
nurse, and allied health professional capacity to partake in the Scheme.
Additional Government incentives for private sector participation may add additional
uncertainty to our projections.

The BIA model utilises the most timely and relevant data within the parameters and relies on
available literature on the progression of DM to model the expected costs and impact of the
Scheme. Nonetheless, given that the model attempts to capture epidemiological changes
over a 30-year horizon with no stochasticity (i.e., we do not account for randomness of event
occurrence), we acknowledge that certain values are likely to be insufficiently representative
of the target demographic and the larger Hong Kong population.

4.3.7 CONCLUSION

Our BIA findings suggest that implementing a chronic disease screening voucher and
management Scheme for DM in Hong Kong contributes to reduced health system
expenditures in the long-term, and additionally helps to reduce preventable mortalities
attributable to chronic diseases. The Government should consider the adoption of a
screening voucher and management scheme for chronic conditions, such as DM, and further
explore the potential for inclusion of other chronic conditions, such as hypertension and
hyperlipidaemia. Greater cost-savings and benefits to the health of the target population are
more likely to be accrued if strategic policies are directed towards prevention and
management in early stages of disease prognosis.

193