Results

DRAFT: Not for Citation

Projected Undiscounted Outcomes with Screening

Undiscounted outcomes associated with the screening strategies are presented in Table 8A for the MISCAN model, Table 8B for the SimCRC model, and Table 8C for the CRC-SPIN model. Without screening we project that 53 to 60 out of every 1000 65-year old individuals will be diagnosed with CRC in their lifetimes. This induces approximately $3.0 to $4.0 million in lifetime direct medical costs ($57 to $71 thousand per CRC case). With screening and removal of adenomas that may have become cancer over time, many of these CRC cases can be prevented assuming 100% adherence to screening regiments; the reduction in the lifetime risk of CRC ranged from 32-49% with annual FOBT (Hemoccult II) screening to 53-85% with 10-year colonoscopy screening (reported ranges reflect differences in projections by model). Some of the benefit associated with the fecal-related tests is a result of the false-positive rate, which leads to individuals being placed on a colonoscopy schedule. In other words, some of the benefit of these tests can be attributed to the fact that a substantial number of individuals with false-positive test results subsequently undergo screening with 10-year colonoscopy. In the MISCAN model the combination of 5-yearly flexible sigmoidoscopy with an annual highly sensitive FOBT (Hemoccult SENSA or FIT) are the two most effective strategies in terms of life-years gained compared with no screening, saving 154 life-years per 1000 persons screened. In the SimCRC and CRC-SPIN models, 10-yearly colonoscopy is most effective, saving 171 and 185 life-years per 1000 persons screened, respectively. Five-yearly CT colonography with a 6mm referral threshold and the most optimistic test characteristics (i.e., DoD study) resulted in 2-7 fewer life-years gained per 1000 individuals compared with 10-yearly colonoscopy, with an increase in lifetime (undiscounted) costs of approximately $600,000-$700,000 per 1000.

Cost-Effectiveness Analysis from Payer Perspective

Table 9 shows the total discounted costs, discounted life-years gained, and the incremental cost-effectiveness ratios for a cohort of 65-year-olds by screening strategy, including no screening, for each model (results for a cohort of 50-year-olds are presented in Appendix 4). Note that the incremental cost-effectiveness ratios were calculated using each CT colonography strategy in turn as they are not competing options. The models varied somewhat as to which tests were on the efficient frontier (i.e., were not ruled out by simple or extended dominance). Strategies on the efficient frontier are those strategies with an associated incremental cost-effectiveness ratio and are potentially cost-effective depending on the societal willingness to pay for a life-year gained. All three models showed the CT colonography strategies to be the most costly options. Figure 2 shows the plots of the discounted life-years gained (compared with no screening), the discounted lifetime direct medical costs (from the Medicare perspective), and the cost-efficient frontier, where each non-dominated strategy is compared with the next least expensive strategy. Hemoccult II was cost-saving compared with no screening for all models. This was the only cost-saving strategy in the MISCAN model. For SimCRC and CRC-SPIN, however, all non-CT colonography strategies were cost-saving compared with no screening. That CT colonography strategies were the most costly can be easily seen from Figure 2 since for all three modes the CT colonography strategies lie to the far right of all screening strategies.

Threshold Analyses

At a cost per test of $488, none of the CT colonography strategies were on the efficient frontier (Figure 2). Table 10 shows the threshold CT colonography costs under the two base-case scenarios. Threshold analyses indicated that in order for the base-case 5-yearly CT colonography strategies with a 6mm referral threshold to be on the efficient frontier, a CT scan would need to cost between $108 and $205 (depending on the test characteristics and the simulation model used). The range of threshold costs required for CT colonography screening to be on the efficient frontier was wider when considering 10-yearly CTC strategies with a 6mm threshold, ranging from $103 to $371. Table 10 also presents threshold costs for CT colonography to reach the efficient frontier under different scenarios of the test characteristics for CT colonography (worst-case assumption and 2D reading from the DoD study). The threshold costs were much lower than the base-case values, while the 2D DoD analysis was more consistent with the base-case analysis, although the range was wider.

Table 10 also reports the secondary analyses where different criteria were used to calculate the CT scan cost thresholds. Note, that the primary analysis represents the theoretically correct analysis. The threshold costs tended to be slightly higher when compared with no screening and when compared with the strategy with the highest ACER. In order for the base-case CT colonography strategies (i.e., 5-yearly screening with a 6mm referral threshold) to have the same ACER compared with no screening as the colonoscopy strategy, a CT scan would have to cost between $179 and $237 (depending upon the CT colonography test characteristics and the model used). In only one case the threshold cost was greater than the base-case unit cost estimate of $488; this was the threshold cost that made 10-yearly CT colonography screening with a 6mm referral threshold cost-neutral compared with no screening and was true for only one model (Table 10). Figures 3-6 illustrate threshold cost values graphically.

Sensitivity Analyses

The threshold costs associated with varying the test characteristics for CT colonography strategies with a 10 mm colonoscopy referral threshold are shown in Table 11. Threshold analyses indicated that in order for 5-yearly CT colonography with a 10mm referral threshold to be on the efficient frontier, a CT scan would need to cost in the range of $98 to $192 for primary 3D reads, $49 to $135 for mixed 2D and 3D reads, and $73 to $160 for primary 2D reads (depending on the test characteristics and the simulation model used). The ranges of threshold costs were wider when considering 10-yearly CT colonography strategies with a 10mm threshold, ranging from $71 to $238 for primary 3D reads, $3 to $167 for mixed 2D and 3D reads, and $72 to $175 for primary 2D reads. Using the secondary criteria to determine thresholds, the threshold costs tended to be slightly higher than the primary analysis (i.e., on the efficient frontier). In no case was the threshold cost greater than the base-case unit cost estimate of $488.

If individuals who would not be screened otherwise would get screened with CT colonography, its cost-effectiveness would improve. The threshold costs for the test to lie on the efficient frontier under varying adherence assumptions are shown in Table 12. With a 10% improvement in CT colonography screening adherence compared with other tests (i.e., 55% overall adherence), the CT colonography cost threshold for being on the efficient frontier increased to $293-$408. With a 25% improvement in CT colonography screening adherence compared with other tests (i.e., 62.5% overall adherence), the CT colonography cost threshold for being on the efficient frontier increased to $547-$694.

Table 13 contains the results of the threshold analysis from a modified societal perspective. From this perspective the threshold costs that result in a CT colonography strategy reaching the efficient frontier are $154-$336 for the 5-yearly testing with a 6 mm referral threshold and $166-$480 for 10-yearly testing with a 6 mm referral threshold. These thresholds costs are a bit higher than those from the payer perspective. The higher frequency of Hemoccult II and Hemoccult SENSA scenarios results in considerably higher additional time costs than with CT screening, allowing for higher per-test costs for the CT scan. The total threshold costs include co-payments and patient time costs. To obtain CMS reimbursement rates co-payments and patient time costs should be subtracted from the total threshold costs. Assuming no co-payments and patient time costs of $17 per hour yields CMS reimbursement rates of $26-$181 for 5-yearly CT colonography screening with a 6mm referral threshold and $11-$325 for 10-yearly CTC screening.

All analyses were conducted for the Medicare population aged 65 years and older assuming no prior CRC screening among this group. To assess the effect of this assumption, we evaluated the cost-effectiveness of the 15 screening strategies for a cohort of 50-year-olds, with screening starting at age 50. Results are presented in Appendix 4. The CT colonography strategies remained the most costly of the screening strategies considered. Threshold analyses indicated that in order for 5-yearly CT colonography with a 6mm referral threshold to be on the efficient frontier, a CT scan would need to cost between $72 and $179 (depending on the test characteristics and the simulation model used), which was lower than we found in the analysis of 65-year-old individuals. The range of threshold costs was wider when considering 10-yearly CT colonography strategies with a 6mm threshold, ranging from $15 to $220, which is also lower than the Medicare payer analysis.

Table 8A. Undiscounted costs by type, number of life-years gained, and number of cases of CRC per 1000 65-year-olds, by screening scenario—MISCAN

Scenario Costs ($) Outcomes
Screening Follow-Up Polyp Resection Surveillance Complications CRC Treatment Total Costs LYG SymDx CRC ScnDx CRC
No screening 0 0 0 0 0 4,030,647 4,030,647 0 57 0
HII 45,577 207,470 86,984 418,620 15,647 2,927,696 3,701,995 116.5 18 21
HS 31,762 370,237 125,488 693,037 26,573 2,501,443 3,748,541 142.8 12 20
FIT 178,116 318,912 116,129 614,068 23,317 2,573,214 3,823,757 141.0 12 21
SIGB 516,641 193,530 115,568 545,450 19,110 2,415,702 3,806,002 132.2 16 14
SIG 378,703 268,592 124,815 633,967 23,143 2,371,694 3,800,914 135.4 15 15
HII + SIGB 471,033 279,361 130,886 665,461 24,154 2,098,139 3,669,035 149.1 11 17
HII + SIG 355,281 333,025 136,711 730,181 26,790 2,275,248 3,857,236 149.9 11 17
HS + SIGB 344,285 398,694 145,073 819,404 30,834 2,016,539 3,754,829 154.1 10 17
HS + SIG 262,997 422,676 147,776 854,913 32,091 2,208,379 3,928,832 154.1 10 17
FIT + SIGB 507,549 356,996 140,678 765,688 28,504 2,229,174 4,028,589 154.3 10 18
FIT + SIG 402,045 391,252 144,355 811,232 30,469 2,219,036 3,998,390 154.3 10 18
COL 776,369 0 152,502 677,187 36,327 2,198,866 3,841,252 151.6 12 15
CTC DoD 3D 6mm 5y 1,007,280 354,666 135,665 748,110 27,561 2,264,920 4,538,212 149.5 11 17
CTC NCTC 2D/3D 6mm 5y 1,129,911 290,386 123,520 644,144 23,369 2,375,757 4,587,088 142.7 13 17

LYG = life-years gained compared with no screening; SymDx CRC = symptom-detected colorectal cancer; ScnDx CRC = screen-detected colorectal cancer.

Table 8B. Undiscounted costs by type, number of life-years gained, and number of cases of CRC per 1000 65-year-olds, by screening scenario—SimCRC

Scenario Costs ($) Outcomes
Screening Follow-Up Polyp Resection Surveillance Complications CRC Treatment Total Costs LYG SymDx CRC ScnDx CRC
No screening 0 0 0 0 0 3,540,411 3,540,411 0 60 0
HII 74,558 189,224 63,882 251,236 11,119 2,213,526 2,803,544 113.9 14 21
HS 121,839 359,983 100,870 409,826 20,408 1,636,905 2,649,832 150.7 8 18
FIT 248,015 305,726 91,444 371,278 17,606 1,711,732 2,745,801 148.3 8 19
SIGB 458,414 129,774 153,495 302,136 11,130 1,795,444 2,850,392 120.6 19 10
SIG 452,330 218,999 82,962 355,829 15,267 1,684,643 2,810,029 128.0 16 10
HII + SIGB 522,284 251,218 168,972 239,952 13,014 1,446,187 2,641,626 157.7 7 15
HII + SIG 529,760 331,172 89,836 255,648 15,279 1,395,290 2,616,985 160.1 7 15
HS + SIGB 437,692 388,531 171,293 417,676 21,751 1,255,331 2,692,275 169.3 6 14
HS + SIG 444,054 442,437 114,584 431,707 23,361 1,231,886 2,688,030 170.2 5 13
FIT + SIGB 628,080 342,482 171,280 366,098 18,916 1,278,827 2,805,683 168.9 6 14
FIT + SIG 638,476 405,523 107,594 379,303 20,723 1,251,488 2,803,107 169.9 5 14
COL 783,430 0 137,876 598,884 32,857 1,124,529 2,677,576 171.3 6 11
CTC DoD 3D 6mm 5y 1,115,618 348,524 114,329 500,485 23,565 1,172,674 3,275,196 168.2 6 12
CTC NCTC 2D/3D 6mm 5y 1,213,047 280,882 101,516 441,470 19,842 1,288,954 3,345,711 160.2 7 12

LYG = life-years gained compared with no screening; SymDx CRC = symptom-detected colorectal cancer; ScnDx CRC = screen-detected colorectal cancer.

Table 8C. Undiscounted costs by type, number of life-years gained, and number of cases of CRC per 1000 65-year-olds, by screening scenario—CRC-SPIN

Scenario Costs ($) Outcomes
Screening Follow-Up Polyp Resection Surveillance Complications CRC Treatment Total Costs LYG SymDx CRC ScnDx CRC
No screening 0 0 0 0 0 2,999,824 2,999,824 0 53 0
HII 80,263 169,980 50,324 200,706 10,036 1,663,309 2,174,619 114.5 17 12
HS 135,166 353,732 83,847 337,414 19,782 1,057,232 1,987,173 155.1 7 11
FIT 267,328 293,055 74,803 302,324 16,660 1,160,290 2,114,460 150.4 8 11
SIGB 478,290 110,463 209,824 269,120 10,365 1,211,533 2,289,595 133.7 17 4
SIG 474,358 206,889 72,375 311,882 14,770 1,079,869 2,160,144 142.2 14 5
HII + SIGB 479,837 221,064 204,285 347,052 15,715 877,095 2,145,048 163.7 7 7
HII + SIG 476,977 289,511 86,877 373,491 18,922 813,753 2,059,531 166.7 7 7
HS + SIGB 420,636 374,095 189,459 415,934 22,787 692,561 2,115,471 175.9 5 7
HS + SIG 425,961 404,518 100,708 426,792 24,437 666,213 2,048,629 176.8 4 7
FIT + SIGB 581,132 320,807 194,795 394,441 20,268 729,944 2,241,386 174.4 5 7
FIT + SIG 567,998 364,345 96,403 411,602 22,497 694,657 2,157,501 175.8 5 7
COL 822,584 0 118,456 506,142 33,208 496,246 1,976,636 184.9 3 5
CTC DoD 3D 6mm 5y 1,202,218 329,204 92,468 398,610 21,994 610,307 2,654,802 177.7 5 5
CTC NCTC 2D/3D 6mm 5y 1,287,352 258,000 83,325 363,894 18,549 686,995 2,698,114 172.2 6 5

LYG = life-years gained compared with no screening; SymDx CRC = symptom-detected colorectal cancer; ScnDx CRC = screen-detected colorectal cancer.

Table 9. Discounted costs and life-years gained per 1000 65-year-olds without CRC screening and with 14 CRC screening strategies and associated incremental cost-effectiveness ratios

Strategy MISCAN SimCRC CRC-SPIN
Discounted Costs ($) Discounted LYG ICER ($) Discounted Costs ($) Discounted LYG ICER ($) Discounted Costs ($) Discounted LYG ICER ($)
No Screening 2,714,556 0 d 2,367,514 0 d 1,976,803 0 d
HII 2,631,879 65.7 --- 2,082,788 59.9 d 1,536,474 64.0 d
HS 2,715,683 81.1 5,455 2,042,708 81.1 --- 1,482,449 87.3 ---
FIT 2,777,228 80.1 d 2,116,618 79.8 d 1,574,679 84.7 d
SIGB 2,823,217 75.0 d 2,168,782 65.2 d 1,716,321 75.8 d
SIG 2,810,249 76.7 d 2,151,925 69.1 d 1,626,360 80.4 d
HII + SIGB 2,790,651 84.9 19,381 2,085,889 85.7 d 1,656,317 92.9 d
HII + SIG 2,839,118 85.4 d 2,072,929 87.0 5,147 1,590,434 94.5 d
HS + SIGB 2,859,815 88.0 22,940 2,151,806 92.5 d 1,666,766 99.9 d
HS + SIG 2,907,440 87.9 d 2,150,786 93.0 12,938 1,611,331 100.5 d
FIT + SIGB 3,022,139 88.1 d 2,244,313 92.3 d 1,768,508 99.2 d
FIT + SIG 2,990,860 88.1 988,660 2,244,650 92.8 d 1,699,373 99.9 d
COL 2,906,228 86.7 d 2,173,712 93.8 27,737 1,600,155 105.5 6,465
CTC DoD 3D 6mm 5ya 3,469,661 85.3 d 2,674,721 92.0 d 2,156,740 101.2 d
CTC NCTC 2D/3D 6mm 5ya 3,489,238 81.3 d 2,706,113 87.2 d 2,172,677 98.0 d

--- indicates default strategy (i.e., the least costly and least effective non-dominated strategy).
LYG = life-years gained vs. no screening; ICER = incremental cost-effectiveness ratio; d = dominated.
a. The two CTC strategies are not competing options; they represent a range of estimates of CTC test characteristics. They are shown here together for comparison purposes only. The ICERs are assessed separately using each CTC strategy in turn.

Figure 2, Panel A. Discounted costs and discounted life-years gained per 1000 65-year-olds for 14 CRC screening strategiesa and the efficient frontier connecting the efficient strategies—MISCAN

Figure 2A  is a graph of the life-years gained per 1000 65-year olds compared to no screening compared to the total costs of screening per 1000 65-year-olds  The data for these figures are in Table 9.  The point of this figure is to assess what strategies provide the most life years gained for a given level of costs required to complete the screening strategies. Both life years and costs are discounted. The strategies considered are Hemoccult II (annually), Hemoccult SENSA (annually), fecal immunochemical test (annually), sigmoidoscopy with biopsy every 5 years, sigmoidoscopy without biopsy every 5 years, Hemoccult II plus sigmoidoscopy with biopsy, Hemoccult II plus sigmoidoscopy without biopsy, Hemoccult SENSA plus sigmoidoscopy with biopsy, Hemoccult SENSA plus sigmoidoscopy without biopsy, fecal immunochemical test plus sigmoidoscopy with biopsy, fecal immunochemical test plus sigmoidoscopy without biopsy, colonoscopy every 10 years, CT colonography with the Department of Defense (DoD) test characteristics, and CT colonography with the National CT Colonography Trial (NCTC) test characteristics.  The strategies that provide the most life years gained for a given level of costs are called  non-dominated strategies and represents the efficient frontier. For MISCAN the strategies on the efficient frontier are Hemoccult II, Hemoccult SENSA, Hemoccult II plus sigmoidoscopy with biopsy, Hemoccult SENSA plus sigmoidoscopy with biopsy, and fecal immunochemical test plus sigmoidoscopy without biopsy.  With the exception of the sigmoidoscopy alone strategies (i.e., without an annual FOBT) and the CT colonography strategies, all other strategies are quite close to the efficient frontier.  The CT colonography strategies (particularly with the test characteristics from the DoD study) are close to the colonoscopy strategy for life-years gained but the costs are significantly higher than the other.

a.The two CTC strategies are not competing options; they represent a range of estimates of CTC test characteristics. They are shown here together for comparison purposes only. The ICERs are assessed separately using each CTC strategy in turn.

Figure 2, Panel B. Discounted costs and discounted life-years gained per 1000 65-year-olds for 14 CRC screening strategiesa and the efficient frontier connecting the efficient strategies—SimCRC

Figure 2B  is a graph of the life-years gained per 1000 65-year olds compared to no screening compared to the total costs of screening per 1000 65-year-olds  The data for these figures are in Table 9.  The point of this figure is to assess what strategies provide the most life years gained for a given level of costs required to complete the screening strategies. Both life years and costs are discounted. The strategies considered are Hemoccult II (annually), Hemoccult SENSA (annually), fecal immunochemical test (annually), sigmoidoscopy with biopsy every 5 years, sigmoidoscopy without biopsy every 5 years, Hemoccult II plus sigmoidoscopy with biopsy, Hemoccult II plus sigmoidoscopy without biopsy, Hemoccult SENSA plus sigmoidoscopy with biopsy, Hemoccult SENSA plus sigmoidoscopy without biopsy, fecal immunochemical test plus sigmoidoscopy with biopsy, fecal immunochemical test plus sigmoidoscopy without biopsy, colonoscopy every 10 years, CT colonography with the Department of Defense (DoD) test characteristics, and CT colonography with the National CT Colonography Trial (NCTC) test characteristics.  The strategies that provide the most life years gained for a given level of costs are called  non-dominated strategies and represents the efficient frontier. For SimCRC the strategies on the efficient frontier are Hemoccult SENSA, Hemoccult II plus sigmoidoscopy without biopsy, Hemoccult SENSA plus sigmoidoscopy without biopsy, and colonoscopy.  Strategies that are quite close to efficient frontier include Hemoccult II plus sigmoidoscopy with biopsy, Hemoccult SENSA plus sigmoidoscopy with biopsy, fecal immunochemical test plus sigmoidoscopy with biopsy, and fecal immunochemical test plus sigmoidoscopy without biopsy.  The CTC strategies (particularly the strategy using the DoD test characteristics) are close to those of the colonoscopy strategy for life-years gained but the costs are significantly higher than the other strategies considered and are far from the efficient frontier.

a.The two CTC strategies are not competing options; they represent a range of estimates of CTC test characteristics. They are shown here together for comparison purposes only. The ICERs are assessed separately using each CTC strategy in turn.

Figure 2, Panel C. Discounted costs and discounted life-years gained per 1000 65-year-olds for 14 CRC screening strategiesa and the efficient frontier connecting the efficient strategies—CRC-SPIN

Figure 2C  is a graph of the life-years gained per 1000 65-year olds compared to no screening compared to the total costs of screening per 1000 65-year-olds  The data for these figures are in Table 9.  The point of this figure is to assess what strategies provide the most life years gained for a given level of costs required to complete the screening strategies. Both life years and costs are discounted. The strategies considered are Hemoccult II (annually), Hemoccult SENSA (annually), fecal immunochemical test (annually), sigmoidoscopy with biopsy every 5 years, sigmoidoscopy without biopsy every 5 years, Hemoccult II plus sigmoidoscopy with biopsy, Hemoccult II plus sigmoidoscopy without biopsy, Hemoccult SENSA plus sigmoidoscopy with biopsy, Hemoccult SENSA plus sigmoidoscopy without biopsy, fecal immunochemical test plus sigmoidoscopy with biopsy, fecal immunochemical test plus sigmoidoscopy without biopsy, colonoscopy every 10 years, CT colonography with the Department of Defense (DoD) test characteristics, and CT colonography with the National CT Colonography Trial (NCTC) test characteristics.  The strategies that provide the most life years gained for a given level of costs are called  non-dominated strategies and represents the efficient frontier. For CRC-SPIN  the strategies on the efficient frontier are Hemoccult SENSA and colonoscopy.  Strategies that are closest to the efficient frontier include Hemoccult II plus sigmoidoscopy without biopsy and Hemoccult SENSA plus sigmoidoscopy without biopsy. As with the MISCAN and SimCRC models, the CRC-SPIN model finds that the CT colonography strategies, and the strategy using the DoD parameters in particular, are close to the colonoscopy strategy in terms of the number of life-years gained but are considerably more costly than all of the other screening strategies and far from the efficient frontier.

a. The two CTC strategies are not competing options; they represent a range of estimates of CTC test characteristics. They are shown here together for comparison purposes only. The ICERs are assessed separately using each CTC strategy in turn.

Table 10. Threshold analysis on CT colonography test characteristics for scenarios with a 6mm colonoscopy referral threshold: unit cost of CT colonography screening test resulting in equal outcomes compared to other recommended CRC screening strategies for different estimates of CT colonography test characteristicsa

CTC outcome Base cases Sensitivity analysisb
CTC DoD 3D 6mm CTC NCTC 2D/3D 6mm CTC WC 2D/3D 6mm CTC DoD 2D 6mm
5-yearly CTC screening
On efficient frontier 122, 196, 199 108, 183, 205 25, 83c, 173 98, 163c, 246
Cost-neutral vs. no screening 76, 323, 398 105, 324, 398 38, 251, 336 112, 308, 393
Equal to highest ACER 238, 258, 294 245, 268, 304 179, 197, 233 232, 261, 303
Equal to colonoscopy ACER 179, 210, 221 194, 227, 237 127, 150, 167 188, 231, 235
10-yearly CTC screening
On efficient frontier 103, 266, 352 108, 241c, 371 9, 115c, 123c 89, 211c, 249c
Cost-neutral vs. no screening 114, 482, 599 143, 473, 599 68, 351, 472 147, 435, 582
Equal to highest ACER 320, 396, 450 325, 398, 455 237, 285, 325 303, 372, 442
Equal to colonoscopy ACER 244, 330, 348 258, 339, 356 175, 206, 248 246, 328, 337

ACER = average cost-effectiveness ratio compared with no screening (calculated using discounted costs and life-years gained).
a. MISCAN values in plain text; SimCRC values in italics; CRC-SPIN values in bold.
b. Go to Table 7 for the test characteristics used in these scenarios.
c. CTC strategy is on the frontier as the least effective and least costly non-dominated strategy if the cost is at most this amount.

Figure 3. CT colonography unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 6mm colonoscopy referral threshold are efficient screening options compared to other recommended CRC screening strategiesa

Panel A: 5-yearly CTC screening with a 6mm referral threshold

Figure 3A has bar graphs depicting the unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 6 mm colonoscopy referral threshold are efficient screening options compared to other recommended CRC screening strategies.  The costs are given for MISCAN, SimCRC and CRC-SPIN models. The costs are given for the DoD and NCTC base-cases  and worst case (WC) and DoD 2D sensitivity analyses. In figure 3A strategies with 5-yearly CT colonography (and a 6mm colonoscopy referral threshold) are based on the DoD and NCTC base cases as well as for the worst case sensitivity analysis.  The data plotted come from Table 10 and correspond with the row "On efficient frontier" under 5-yearly screening.  For the DoD 3D base-case parameters the threshold cost is $122 for MISCAN, $199 for SimCRC, and $196 CRC-SPIN and for the NCTC base-case parameters the threshold cost is $108 for MISCAN, $183 for SimCRC, and $205 for CRC-SPIN.  For the worst case sensitivity analysis parameters the threshold cost is $25 for MISCAN, $83 for SimCRC, and $173 for CRC-SPIN.  For the DoD sensitivity analysis parameters with primary 2D reading, the threshold cost is $98 for MISCAN, $163 for SimCRC, and $246 for CRC-SPIN. All values are below the $488 base case considered for CTC.

Panel B: 10-yearly CTC screening with a 6mm referral threshold

Figure 3B has bar graphs depicting the unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with 10-yearly CT colonography (and a 6mm colonoscopy referral threshold) are efficient screening options compared to other recommended CRC screening strategies. The data plotted come from Table 10 and correspond with the row "On efficient frontier" under 10-yearly screening.  For the DoD 3D base-case parameters the threshold cost is $103 for MISCAN, $266 for SimCRC, and $352 for CRC-SPIN and for the NCTC base-case parameters the threshold cost is $108 for MISCAN, $241 for SimCRC, and $371 for CRC-SPIN.  For the worst case sensitivity analysis parameters the threshold cost is $9 for MISCAN, $115 for SimCRC, and $123 for CRC-SPIN.  Finally, for the DoD 2D sensitivity analysis parameters the threshold cost is $89 for MISCAN, $211 for SimCRC, and $249 for CRC-SPIN. All values are below the $488 base case considered for CTC.

DoD = Department of Defense Study (Pickhardt 2003, 2007a); NCTC = National CT Colonography study (Johnson 2008); WC = hypothetical worst-case scenario.
a. CTC strategy is on the frontier as the least effective and least costly non-dominated strategy if the cost is at most this amount.

Figure 4. CT colonography unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 10mm colonoscopy referral threshold are efficient screening options compared to other recommended CRC screening strategiesa

Panel A: 5-yearly CTC screening with a 10mm referral threshold

Figure 4A has bar graphs depicting the unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 10 mm colonoscopy referral threshold (two using primary 3D reads, two using a mix of 2D and 3D reads and two using primary 2D reads) are efficient screening options compared to other recommended CRC screening strategies.  The results are given for the MISCAN, SimCRC, and CRC-SPIN models.  Figure 4A is for strategies with 5-yearly CT colonography (and a 10mm colonoscopy referral threshold).  The data plotted come from Table 11 and correspond with the row "On efficient frontier" under 5-yearly screening.  For the DoD 3D parameters the threshold cost is $98 for MISCAN, $192 for SimCRC, and $132 CRC-SPIN and for the Johnson 3D parameters the threshold cost is $71 for MISCAN, $153 for SimCRC, and $105 for CRC-SPIN.  For the NCTC parameters the threshold cost is $49 for MISCAN, $135 for SimCRC, and $90 for CRC-SPIN and for the worst case parameters the threshold cost is $10 for MISCAN, $81 for SimCRC, and $43 for CRC-SPIN.  Finally, for the primary 2D sensitivity analysis scenarios, the threshold cost for the DoD study parameters is $75 for MISCAN, $160 for SimCRC, and $105 for CRC-SPIN and the threshold for the Johnson study is $73 for MISCAN, $154 for SimCRC, and $105 for CRC-SPIN.

Panel B: 10-yearly CTC screening with a 10mm referral threshold

Figure 4B is for strategies with 10-yearly CT colonography (and a 10mm colonoscopy referral threshold). The data plotted come from Table 11 and correspond with the row "On efficient frontier" under 10-yearly screening.  For the DoD 3D parameters the threshold cost is $82 for MISCAN, $238 for SimCRC, and $163 CRC-SPIN and for the Johnson 3D strategy the threshold cost is $71 for MISCAN, $166 for SimCRC, and $99 for CRC-SPIN.  For the NCTC parameters the threshold cost is $30 for MISCAN, $167 for SimCRC, and $104 for CRC-SPIN and for the worst case parameters the threshold cost is $3 for MISCAN, $84 for SimCRC, and $20 for CRC-SPIN.  Finally, for the primary 2D sensitivity analysis scenarios, the threshold cost for the DoD study parameters is $75 for MISCAN, $175 for SimCRC, and $108 for CRC-SPIN and the threshold for the Johnson study is $72 for MISCAN, $164 for SimCRC, and $96 for CRC-SPIN.

DoD = Department of Defense Study (Pickhardt 2003, 2007a); J = Johnson study (Johnson 2007); NCTC = National CT Colonography study (Johnson 2008); WC = hypothetical worst-case scenario.
a. CTC strategy is on the frontier as the least effective and least costly non-dominated strategy if the cost is at most this amount.

Figure 5. CT colonography unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 6mm colonoscopy referral threshold have an average cost effectiveness ratio (ACER) equal to that of colonoscopy screening

Panel A: 5-yearly CTC screening with a 6mm referral threshold

Figure 5A has bar graphs depicting the unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 6 mm colonoscopy referral threshold (DoD and NCTC base-cases in the first grouping and worst case (WC) and DoD 2D sensitivity analyses in the second grouping) have average cost-effectiveness ratios (ACERs) equal to that of 10-yearly colonoscopy screening.  The results are given for the MISCAN, SimCRC, and CRC-SPIN models.  Figure 5A is for strategies with 5-yearly CT colonography (and a 6mm colonoscopy referral threshold).  The data plotted come from Table 10 and correspond with the row "Equal to colonoscopy ACER" under 5-yearly screening.  For the DoD 3D base-case parameters the threshold cost is $179 for MISCAN, $221 for SimCRC, and $210 CRC-SPIN and for the NCTC base-case parameters the threshold cost is $194 for MISCAN, $237 for SimCRC, and $227 for CRC-SPIN.  For the worst case sensitivity analysis parameters the threshold cost is $127 for MISCAN, $167 for SimCRC, and $150 for CRC-SPIN.  For the DoD sensitivity analysis parameters with primary 2D reading, the threshold cost is $188 for MISCAN, $235 for SimCRC, and $231 for CRC-SPIN.

Panel B: 10-yearly CTC screening with a 6mm referral threshold

Figure 5B is for strategies with 10-yearly CT colonography (and a 6mm colonoscopy referral threshold). The data plotted come from Table 10 and correspond with the row "Equal to colonoscopy ACER" under 10-yearly screening.  For the DoD 3D base-case parameters the threshold cost is $244 for MISCAN, $348 for SimCRC, and $330 for CRC-SPIN and for the NCTC base-case parameters the threshold cost is $258 for MISCAN, $356 for SimCRC, and $339 for CRC-SPIN.  For the worst case sensitivity analysis parameters the threshold cost is $175 for MISCAN, $248 for SimCRC, and $206 for CRC-SPIN.  Finally, for the DoD 2D sensitivity analysis parameters the threshold cost is $246 for MISCAN, $337 for SimCRC, and $328 for CRC-SPIN.

DoD = Department of Defense Study (Pickhardt 2003, 2007a); NCTC = National CT Colonography study (Johnson 2008); WC = hypothetical worst-case scenario.

Figure 6. CT colonography unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 10mm colonoscopy referral threshold have an average cost effectiveness ratio (ACER) equal to that of colonoscopy screening

Panel A: 5-yearly CTC screening with a 10mm referral threshold

Figure 6A has bar graphs depicting the unit cost thresholds (in 2007 US dollars) at which CT colonography strategies with a 10 mm colonoscopy referral threshold (two using primary 3D reads, two using a mix of 2D and 3D reads and two using primary 2D reads) have average cost-effectiveness ratios (ACERs) equal to that of 10-yearly colonoscopy screening.  There results are given for the  MISCAN, SimCRC, and CRC-SPIN models. Figure 6A is for strategies with 5-yearly CT colonography (and a 10mm colonoscopy referral threshold).  The data plotted come from Table 11 and correspond with the row "Equal to colonoscopy ACER" under 5-yearly screening.  For the DoD 3D parameters the threshold cost is $187 for MISCAN, $259 for SimCRC, and $178 for CRC-SPIN and for the Johnson 3D parameters the threshold cost is $167 for MISCAN, $228 for SimCRC, and $151 for CRC-SPIN.  For the NCTC parameters the threshold cost is $145 for MISCAN, $210 for SimCRC, and $142 for CRC-SPIN and for the worst case parameters the threshold cost is $115 for MISCAN, $166 for SimCRC, and $96 for CRC-SPIN.  Finally, for the primary 2D sensitivity analysis scenarios, the threshold cost for the DoD study parameters is $170 for MISCAN, $233 for SimCRC, and $155 for CRC-SPIN and the threshold for the Johnson study is $167 for MISCAN, $229 for SimCRC, and $150 for CRC-SPIN.

Panel B: 10-yearly CTC screening with a 10mm referral threshold

Figure 6B is for strategies with 10-yearly CT colonography (and a 10mm colonoscopy referral threshold). The data plotted come from Table 11 and correspond with the row "Equal to colonoscopy ACER" under 10-yearly screening.  For the DoD 3D parameters the threshold cost is $231 for MISCAN, $364 for SimCRC, and $240 CRC-SPIN and for the Johnson 3D strategy the threshold cost is $207 for MISCAN, $306 for SimCRC, and $176 for CRC-SPIN.  For the NCTC parameters the threshold cost is $185 for MISCAN, $298 for SimCRC, and $184 for CRC-SPIN and for the worst case parameters the threshold cost is $149 for MISCAN, $229 for SimCRC, and $100 for CRC-SPIN.  Finally, for the primary 2D sensitivity analysis scenarios, the threshold cost for the DoD study parameters is $211 for MISCAN, $314 for SimCRC, and $184 for CRC-SPIN and the threshold for the Johnson study is $206 for MISCAN, $306 for SimCRC, and $173 for CRC-SPIN.

DoD = Department of Defense Study (Pickhardt 2003, 2007a); J = Johnson study (Johnson 2007); NCTC = National CT Colonography study (Johnson 2008); WC = hypothetical worst-case scenario.

Table 11. Threshold analysis on CT colonography test characteristics for scenarios with a 10mm colonoscopy referral threshold: unit cost of CT colonography screening test resulting in equal outcomes compared to other recommended CRC screening strategies for different estimates of CT colonography test characteristicsa

CTC outcome Sensitivity analysis scenarios with 10mm colonoscopy referral thresholds
Primary 3D reads 2D/3D reads Primary 2D reads
CTC DoD 3D 10mm CTC J 3D 10mm CTC NCTC 2D/3D 10mm CTC WC 2D/3D 10mm CTC DoD 2D 10mm CTC J 2D 10mm
5-yearly CTC screening
On efficient frontier 98, 132b, 192b 71, 105b, 153b 49, 90b, 135b 10, 43b, 81b 75, 110b, 160b 73, 105b, 154b
Cost-neutral vs. no screening 118, 327, 329 106, 284, 297 68, 284, 309 43, 232, 265 110, 290, 301 107, 284, 296
Equal to highest ACER 227, 246, 284 202, 216, 248 190, 216, 237 157, 172, 189 206, 221, 254 201, 215, 248
Equal to colonoscopy ACER 178, 187, 259 151, 167, 228 142, 145, 210 96, 115, 166 155, 170, 233 150, 167, 229
10-yearly CTC screening
On efficient frontier 82, 163b, 238b 71, 99b, 166b 30, 104b, 167b 3, 20b, 84b 75, 108b, 175b 72, 96b, 164b
Cost-neutral vs. no screening 139, 457, 487 127, 382, 420 88, 393, 440 61, 311, 356 131, 391, 428 128, 380, 417
Equal to highest ACER 285, 350, 397 253, 285, 333 242, 299, 332 200, 215, 259 258, 293, 342 252, 282, 332
Equal to colonoscopy ACER 231, 240, 364 176, 207, 306 184, 185, 298 100, 149, 229 184, 211, 314 173, 206, 306

ACER = average cost-effectiveness ratio compared with no screening (calculated using discounted costs and life-years gained).
a. MISCAN values in plain text; SimCRC values in italics; CRC-SPIN values in bold.
b. CTC strategy is on the frontier as the least effective and least costly non-dominated strategy if the cost is at most this amount.

Table 12. Threshold analysis on CT colonography adherence: unit cost of CT colonography screening test resulting in equal outcomes compared to other recommended CRC screening strategies for different levels of adherence with CT colonography screeninga

CTC outcome Base case
(CTC DoD 3D 6mm 5y)
Sensitivity Analysis on CTC Adherenceb
Adherence 50% for all strategies CTC adherence 55% CTC adherence 62.5%
On efficient frontier 122, 196, 199 293c, 360c, 408c 547c, 668c, 694c
Cost-neutral vs. no screening 76, 323, 398 76, 323, 398 76, 323, 398
Equal to highest ACER 238, 258, 294 238, 258, 294 238, 258, 294
Equal to colonoscopy ACER 179, 210, 221 179, 210, 221 179, 210, 221

ACER = average cost-effectiveness ratio compared with no screening (calculated using discounted costs and life-years gained)
a. MISCAN values in plain text; SimCRC values in italics; CRC-SPIN values in bold.
b. Strategies other than CTC remain at 50% adherence.
c. CTC strategy is on the frontier with an incremental cost-effectiveness ratio (ICER) of $50,000 if the cost is at least this amount.

Table 13. Threshold analysis from modified societal perspective: unit costs for CT colonography screening test resulting in equal outcomes compared to other recommended CRC screening strategies for modified societal perspectivea

CTC outcome Total threshold costs (includes co-payments and patient time costs) CMS reimbursement rates (excludes co-payments and patient time costs)
CTC DoD 3D 6mm CTC NCTC 2D/3D 6mm CTC DoD 3D 6mm CTC NCTC 2D/3D 6mm
5-yearly CTC screening
On efficient frontier 181, 318, 332 154, 324, 336 26, 163, 177 NT, 169, 181
Cost-neutral vs. no screening NT, 288, 406 12, 321, 432 NT, 133, 250 NT, 166, 277
Equal to highest ACER 294, 433,476 303, 445, 496 139, 278, 321 148, 290, 341
Equal to colonoscopy ACER 215, 340, 347 234, 371, 372 60, 185, 191 79, 216, 217
10-yearly CTC screening
On efficient frontier 166, 476, 480 176, 428, 474 11, 321, 325 21, 272, 318
Cost-neutral vs. no screening NT, 471, 646 28, 494, 671 NT, 315, 491 NT, 339, 515
Equal to highest ACER 398, 662, 747 405, 661, 768 243, 507, 591 250, 506, 613
Equal to colonoscopy ACER 298, 548, 552 316, 562, 580 143, 393, 397 161, 406, 425

ACER = average cost-effectiveness ratio compared with no screening (calculated using discounted costs and life-years gained); NT = no threshold found (i.e., negative CTC test cost).
a. MISCAN values in plain text; SimCRC values in italics; CRC-SPIN values in bold.
b. CTC strategy is on the frontier as the least effective and least costly non-dominated strategy if the cost is at most this amount.

Return to Contents
Proceed to Next Section