Cost-effectiveness analysis of proton beam therapy for treatment decision making in paranasal sinus and nasal cavity cancers in China

Table 1 Model information of the base case used and probabilistic sensitivity analysis of the parameters

Parameters	Input Information	90% CI in PSA^a	Distribution^b	Source
Base case
Age (year-old)^c	47
Tumor stage	T3N1M0, Stage III
Evaluated radiotherapy modes	IMPT vs. IMRT
Cost ($)
IMPT	50,000	37,258.3–62,775.4	Normal	SPHIC
IMRT	12,000	10,723.2 - 13,287.5	Normal	SYSUCC
Concurrent chemotherapy	5000	3710.4 - 6284.7	Normal	SYSUCC
Follow-up / year	1000	872.1–1127.8	Normal	SYSUCC
Palliative therapy / year	5000	3724.0 - 6269.1	Normal	SYSUCC
Utilities (QALYs)
No cancer	0.94	0.822–1	Beta	Ward MC et al. [13]
Alive with cancer	0.47	0.343–0.598	Beta	Noel CW et al. [14]
Death	0
Transition probabilities
IMPT eradicating cancer	0.9	0.833–0.957	Beta	Samir H Patel et al. [7]
IMRT eradicating cancer	0.73	0.665–0.792	Beta	Dulguerov P et al. [4], Samir H Patel et al. [7]
“no cancer” - “alive with cancer”	0.1(2nd-3rd year); 0.05(4th–5th year); 0.01(6th -10th year); 0(6th -10th year)			Dulguerov P et al. [4], Samir H Patel et al. [7]
“alive with cancer” - “death”	0.3	0.237–0.365	Beta	Dulguerov P et al. [4], Samir H Patel et al. [7]
Risk of non-cancer death	2016 Life Tables			Social Security Administration [15]
Model set-up
Cycle length	1-year
No. of cycles^d	30
Discount rate / year	3%

CI confidence interval, PSA probabilistic sensitivity analysis, IMPT intensity modulated proton radiation therapy, IMRT intensity modulated photon-radiation therapy, $ US dollars, QALY quality adjusted life year, SPHIC Shanghai proton and heavy Ion Center, SYSUCC Sun Yat-sen university cancer center
^aProbabilistic sensitivity analysis was applied to determine 90% CI for model parameters by running over 50,000 iteration trials
^bThe utilities and probabilities were tested with beta distributions and the costs were tested with uniform distributions
^cMedian age of paranasal sinus and nasal cavity cancer in China [12]
^dMarkov models were to be cycled 30 times to evaluate the treatment outcomes over a 30-year time period for the base case until 77-year-old (Chinese life expectancy)

ISSN: 1471-2407