Van Gool & Bridges
TABLE 13: SENSITIVITY ANALYSIS: ADVERSE SELECTION AND LENGTH OF
INTERVENTION (WITH 70% TAKE-UP RATE AMONGST GPs)
|
SCENARIO 1 50% |
SCENARIO 2 |
SCENARIO 3 | |
|
Years of effect |
4 |
53 |
7^ |
|
Deaths prevented |
71 |
135 |
221 |
|
Total costs |
$5,936,000 |
$5,936,000 |
$5,936,000 |
|
Cost-effectiveness |
$84,144 |
$43,873 |
$26,811 |
From the results in Table 13 it can be seen that the joint effects of adverse selection and
length of intervention (i.e. a minimum and maximum range of $84,144 to $26,811)
show the GP based intervention to be a high-risk strategy.
Summary of results
Incorporating the extended sensitivity analysis for the GP intervention (Table 13) along
with the length of effectiveness sensitivity analysis of the remaining interventions
(Table 10) the results are reported in table 14.
TABLE 14: SUMMARY OF CEA RESULTS
|
INTERVENTION |
CEA RATIO SENSITIVITY ANALYSIS BOUNDS |
|
GP Based |
$43,873 ($26,811, $84,144) |
|
Community Mass Media |
$30,151 ($21,338, $53,346) $22,727 ($15,152, $45,455) |
|
HospitaVRehab |
$23,054 ($16,138, $40,345) |
The least ‘risky’ intervention is the HospitalZRehab program closely followed by the
Mass Media and Community interventions. The GP intervention is perceived as the
most risky, which reflects the added uncertainty in GP participation rates. The next
section will provide more details on how these ‘risks’ can be interpreted and also how
they can be used.
33
Chere Project Report 11- November 1999