Part III: Cost-effectiveness of Population-based Screening for PD89

for PD

5.3.1 Simulation for the effect of screening policy

Table 5-3-1 showed the simulated results based on estimated transition rates in Table

5-2-4 to a hypothetical cohort of 9829 elderly people in 12-year period. The results

showed that without screening, there would be 35% Parkinson diseases were H-Y stage III

or more severe at diagnosis. However, annual screen would bring down the figure to alyyyyysisisisisisssss ofofofoffffffjjjjjjjjjoioioioioioioioiointntntntntntntntnt

b b b by

by11111110 00 0000 0yeyeyeyeyeyeyeyearararararararars sss s ss shahahahahahahahadd d dd dd d

r traaaaansnsnsnsnsn itititititioioioionn frfrrrromomomomomo

90

10.2%, referring to 71% (95% CI: 64-77%) severe stage cases reduction. When the

screening intervals were 2-yearly, 3-yearly, 4-year, or 6-yearly, the severe cases reduction

compared to no screen was by 54% (95% CI: 45-62%), 43% (95% CI: 32-52%), 35%

(95% CI: 23-45%), and 25% (95% CI: 12-36%), respectively.

5.3.2 Results of deterministic cost-effectiveness and cost-utility analysis

The results from deterministic Markov decision analysis for the cost-effectiveness

and cost-utility analysis given a hypothetical cohort of 10,000 peoples aged 60 with 20

years follow-up are presented in Table 5-3.2. The incremental cost-effectiveness ratios

(ICER) of PD screening with different inter-screening intervals compared to no screen

ranged from $1169 to $1804 per life-year gained. The incremental cost-utility ratio ranged

from $1715 to $2606 per quality-adjusted life-year gained. Lower attendance rate (60%)

resulted in slightly small ICER and ICUR but the trend with changing inter-screening

intervals had the same trend. For higher attendance rate (100%), the absolute vales of

ICER and ICUR elevated a bit and still remained the same trend with different

inter-screening intervals.

Table 5-3-3 showed the distribution of cost and effectiveness, and net monetary

benefit from Monte Carlo simulation. The cost was $1050 for no screen, and increased

with more frequent inter-screening intervals, ranging from $1075 for triennial screen to

$1115 for annual screen. For life-year gained and QALY gained (QALYG), no screen had n. WWWWWheheheheheeeenn n n nn nnnthththththththththeeeee e ee

e e e e

ererererecccccccccasasasasasasasasesesesesesesesesrrrrrrrededededededededucucucucucucucuctitititititititioonononononoon n

32-55555222222%)%)%)%)%), , ,,,3355%%%%%%%

91

the least gains, and this figure increased with more intensive inter-screening interval.

Considering the net monetary benefit with cost-effectiveness analysis given WTP at

$20,000, annual screen had the greatest NMB ($280,687) in terms of life-year gained,

followed by biennial ($280,511), triennial ($280,416) screen, and no screen ($280,113).

Under the cost-utility analysis, the trend was still the same, the most net monetary benefit

was for annual screen ($275,620), followed by biennial ($275,484), triennial ($275,438)

screen, and no screen ($275,272).

5.3.3 Results of probabilistic cost-effectiveness and cost-utility analysis

The probabilistic cost-effectiveness analysis was conducted to compare the results

in different screening intervals and attendance rate for Parkinson disease screening

strategies. Given 10000 first-order trials and 1000 second-order parameter samples in

Monte Carlo simulation, the scattered incremental cost-effectiveness analysis for

population-based screening program are shown in Figure 5-3-1~5-3-12. The results

demonstrated the 65-78% simulations were cost-effective. Taking the attendance rate with

triangular distribution, under the willing-to-pay of $20,000, the probability of

cost-effectiveness were 78.4%, 71.4%, and 67.3% for 1-year, 2-year, and 3-year respectively

when compared with no screening policy. To monitor the robustness of CEA by screening

attendance rate, the 100% and 60% attendance rate were applied to simulate compared

with no screening. Given attendance rate of 100%, the probability of being cost-eninininininggggg ininnnnnnnnteteteteteteteteervrvrvrvrvrvrvrvalalalalalalalalal.

g g g g

givivivivivennnnnnnnnWWWWWWWWTPTPTPTPTPTPTPTPaaaaaaaatttttt t t

fe-yeyeyeyeyeeararararar gggggaaiaaneneed,d,d,d,ddd,

92

effectiveness was 79.2%, 75.8%, and 69.0% for screening interval with 1-year, 2-year, and

3-year compared with no screening, respectively (Figure 5-3-5~5-3-7). The acceptability

curve of cost-effectiveness analysis is shown in Figure 5-3-8. Given of attendance rate of

60%, the probability of being cost-effectiveness was 74.4%, 64.9%, and 64.4% for

different screening intervals (Figure 5-3-9~5-3-11).

The scattered incremental cost-utility analysis for different screening intervals for

population-based Parkinson disease screening compared with no screening were

demonstrated in Figure 5-3-13~5-3-24. Regardless the different screening intervals

chosen, approximate 59-69% simulations were cost-effective. Given the willing-to-pay of

$20,000, the probability of being cost-utility was 68.8%, 62.6%, and 59.0% for

inter-screening interval with 1-year, 2-year, and 3-year compared with no inter-screening,

respectively. The acceptability curves by different screening intervals were demonstrated

in Figure 5-3-16.

To examine the cost-utility difference by screening attendance rate, we simulated

based on 100% and 60% with different screening interval scenarios to estimate the CUA.

Given screening attendance rate at 100%, it was about 63-70% simulations were

cost-effective. Given the willing-to-pay of $20,000, the probability of being cost- effectiveness

was 70.2%, 66.3%, and 62.6% for inter-screening interval with 1-year, 2-year, and 3-year

compared with no screening, respectively (Figure 5-3-17~5-3-20). Furthermore, given on h 1---yeyyeyeyeararrrrrrrr,,,, ,222222222--yeyeyeyeyeyeyeyearearararararararar, , , , , , ,anannnnd d d d d

T T T Th

Thee eeeeeeeaaaaaaaaccccccccccccccccepepepepepeptapptatatatatatatabibibibibibibibililililililililitytytytytytytyty

atteeeeendndndndndn anananana cececc ratatatatatate ee e e ofofofofofof

93

attendance rate with 60%, the probability of being cost- effectiveness was 62.6%, 61.0%,

and 58.2% (Figure 5-3-21~5-3-24). The acceptability curve of cost-utility analysis was

shown in Figure 5-3-24. The probability of being cost-effectiveness given 100%

attendance rate was greater than 60%.

waasssss6262626262.6.66666666%%,%,%,%,%%,%,%,6666666661.11.1111.1.10%0%0%0%0%0000 ,

it it it i

ity yyyaananananananananalalalalalalalalysysysysysysysysisisisisisisisiswwwwwwwwasasasasasasasas

ven 101010101000%0%0%0%0%

94