Result card

  • ECO2: What amounts of resources are used when delivering FIT and its comparators gFOBT and no screening (resource use measurement)?
English

What amounts of resources are used when delivering FIT and its comparators gFOBT and no screening (resource use measurement)?

Authors: Principal Investigators: Anna-Theresa Renner, Ingrid Rosian-Schikuta, Investigators: Nika Berlic, Neill Booth, Valentina Prevolnik Rupel

Internal reviewers: Pseudo178 Pseudo178, Pseudo283 Pseudo283, Pseudo291 Pseudo291, Pseudo293 Pseudo293, Pseudo294 Pseudo294, Pseudo297 Pseudo297, Pseudo298 Pseudo298

Eighteen different articles from the systematic review of the literature outlined in the domain methodology section were used to identify estimates of the amounts of resources that are used when delivering FIT and its comparators gFOBT and no-screening.

The term “resources” refers to the natural units of health care services used. After  reviewing the studies it can be concluded that majority of studies mainly focused on the estimation of financial resources {10-12, 16-19, 21, 22, 25}. Nevertheless, there are some studies which also mentioned other types of resources used, namely: the number of screening tests {13-15, 30}, colonoscopies (COL) {13-15, 20, 23, 24, 30}, computed tomographic colonographies (CTC) {30}, polypectomies {13, 14, 20}, number of ultrasounds (TUS), number of persons receiving PET scan, MRI scan, CT scan, pre-operative radiotherapy and undergoing colorectal resection {30}.

By comparing estimates of financial resources that would be used when delivering FIT and its comparator gFOBT, it can be summarized that the amount of resources used are approximately at the same level. Studies {11, 13, 17, 20-22, 25} have shown that the amounts of resources, when delivering FIT, can be slightly higher in comparison with gFOBT. Some other studies reported similar costs – depending also on specificity of the test or the different types of thresholds (i.e. positivity threshold to determine the optimal cut-off) {10, 11, 15}.

Van Ballegooijen et al. (2003) indicated that FIT at 95% specificity level (462,794,391 $) is more expensive than gFOBT Hemoccult II (205,556,566 $) but less expensive in comparison to gFOBT Hemoccult – SENSA (775,643,892 $). FIT at 98% specificity level is the least expensive strategy; total costs are estimated to 83,110,600 $ (results per 1 million individuals, age 65-79 at the beginning of the screening program) {10}.

Results of Berchi et al.’s study (2010) indicates that screening with FIT at a 20 ng/ml (1,555,041 €) and 55ng/ml cut-off level (1,119,406 €) is more expensive than gFOBT (907,805 €) – if comparing only total costs. However, FIT can be less expensive if using thresholds at 93ng/ml (713,764 €) and 148ng/ml (192,702 €) – if comparing only total costs. Berchi et al. concluded that at a threshold of 93 ng/ml screening with FIT would cost € 94,041 less than one round of screening with gFOBT and would allow the detection of four additional advanced tumors. At a threshold of 75 ng/ml one round of screening with FIT would cost €6,282 less than one round of screening with gFOBT and would allow the detection of forty-two additional advanced tumors. For both thresholds, the positivity rate of FIT was lower than that of gFOBT. Consequently, FIT enabled the detection of a higher number of tumors without substantially increasing the risk associated with confirmatory colonoscopy. The threshold at which FIT and gFOBT positivity rates were almost identical was 67 ng/ ml. Using this threshold, screening with FIT resulted in more advanced cancers screened than screening with gFOBT, but was also more costly (€863 per advanced tumor screened) {11}.

In one of the previous studies Berchi et al. (2004) mentioned that the total annual cost of organizing the campaign, which had been assessed by the Social Security Department, amounts to 63,256 €. They also conclude that FIT can be slightly more expensive in comparison to gFOBT. Five biennial screening campaigns cost 230 € per targeted individual (including refusals) with FIT and 177 € per targeted person with gFOBT. With ten biennial screening campaigns, the costs per targeted individual were 316 € for FIT and 234 € for gFOBT {12}.

Hassan et al. (2011) displayed the use of resources for a cohort of 100,000 subjects that were invited to screening. (In order to project the simulation outcomes on the French population; a steady state for population size and age distribution was assumed, as represented by the year 2010 French census data. The model outputs reflected all persons aged 50–100 years of age at a given point in time in the steady state, as opposed to a cohort aging from 50 to 100 years over 50 years.) Besides financial resources they presented also the number of colonoscopies, total number of screening tests and number of polypectomies. They indicated that the number of colonoscopies (COL), which were performed as a screening strategy every 10 years, amounted 153,862. COL performed after annual FIT screening were estimated to 89,265, after biennial FIT screening 56,827, after annual gFOBT screening 38,219 and after biennial gFOBT 21.160. The number of total screening tests after annual FIT screening was estimated to 615,237, after biennial FIT screening 346.930, after annual gFOBT screening 772,361 and after biennial gFOBT screening 425,987. They indicated that 22,168 polypectomies have been performed after the colonoscopy, 22,639 number of polypectomies after annual FIT screening, 14,683 number of polypectomies after biennial FIT screening, 10,833 after annual gFOBT screening and 6,005 after biennial gFOBT screening. They estimated that the amount of resources that is required for FIT is higher than the amount that is necessary when screening with gFOBT. The estimated costs for delivering FIT biennially are 90,851,477 € (909 € per individual) and increase to 108,657,236 € (1087 € per individual), when delivering the screening with FIT annually. The presented data stands for a cohort of 100,000 French subjects. If the gFOBT screening method is used, the necessary amount of resources for biennially screening amounts to 79,359,152 € (794 € per individual) and for annually screening – 88,132,104 € (881 € per individual) {13}.

 

A Canadian report from 2009 (Heitman et al.) {14}mentioned number of performed colonoscopies and polypectomies and also number of FIT and gFOBT screening tests (i.e., test kits, as well as colonoscopies, polypectomies performed after the results of primary screening test). The results are indicated in the table below:

 

Table 1: Results of Heitman et al. (2009) {14}

10936.ECO-2 Table 1

hey came to the conclusion that FIT tests differ in methods (type of assay and collection) and test performance. Given the heterogeneity of the available FIT tests, they modeled three independent scenarios: one representing studies reporting “lower” test performance (FIT-low), one representing studies with “mid-range” test performance (FIT-mid), and one representing studies with “high” test performance (FIT-high). Their use of FIT-low, FIT-mid, and FIT-high represents a spectrum of FIT sensitivity and specificity, which in their opinion may be the result of differences in the testing kits and collection methods. The results were based on 100,000 people screened and indicated that the amount of resources (i.e. the costs of screening and CRC management) are equal among gFOBT and FIT-low (1,820 CAN$ in average per patient). This could be due to the fact that approximately the same number of tests (FIT and FOBT) were used and similar number of cancer was detected. If screening is delivered with FIT-mid (1,730 CAN$ in average per patient) the costs are lower, while with FIT-high (1,920 CAN$ in average per patient) the costs are higher. This could be due to smaller number of FIT’s used and due to further treatment (higher number of colonoscopies, polypectomies and colonoscopy complications). The data seem to suggest a better detection of advanced adenomas and CRC with two or three days of fecal sampling compared with one day {14}.

Heitman’s et al. study, published one year later (2010), indicated the number of screening tests required (during the lifetimes for hypothetical cohort of 100.000 average risk patients), number of colonoscopies and base case costs of screening and managing CRC. The study compared FIT-low, FIT-mid and FIT-high with gFOBT-low and gFOBT-high. Using base case estimates, over the lifetimes of a 100,000 patient cohort, the estimations showed that in general the FIT screening option is less expensive in comparison to the gFOBT strategy. Screening with FIT-low amounts to 2,005 CAN$ (on average per patient), with FIT-mid to 1,833 CAN$ (on average per patient) and with FIT-high to 2,004 CAN$ (on average per patient). On the other hand screening with gFOBT-low amounts to 2,195 CAN$ (on average per patient) and with gFOBT-high 2,084 CAN$ (on average per patient). The data on the base case costs as well as on number of screening tests required, are presented also in the following table {15}:

Table 2: Results of Heitman et al. (2010) {15}

10936.ECO-2 Table 2

Heresbach et al. (2010) were the only ones (compared to other studies), who outlined the comparison between screening costs and overall costs. The results of the study were based on a cohort of 100,000 individuals from 50-74 years, observing a time horizon of 30 years. In the study by Heresbach et al. total costs of the strategy that uses FIT are higher in comparison to the overall/total costs of strategy that uses gFOBT. They estimated that invitations to FOBT screenings structurally cost much more than those for CTC screening, the latter induced a much higher expenditure associated with the cost of the screening procedure itself. However, FIT generated much more colonoscopy costs. The increase in the total cost of COL-P and COL-S relative to no screening was 48,951,079 € with FIT and 28,348,797 € with CTC. Fewer cancers were prevented by FIT than by CTC, which induced more costs of CRC treatments. On the whole, gFOBT was the least costly competing strategy (13,583,934 €) and FIT the most expensive one (28,560,396 €). When comparing only the amount of resources that are necessary for screening (invitation and the screening procedure) it can be seen that   the invitation to FIT screening and FIT procedure amounted slightly less in comparison to gFOBT’s invitation and procedure. Invitation to FIT screening amounted 1,717,911 €, in addition the FIT screening procedure amounted 3,037,266 €, while the invitation to gFOBT screening amounted 1,767,118 € and the gFOBT screening procedure – 3,880,590 € {16}.

Lejeune et al. (2010), similar to the studies mentioned above, concluded in their economic modeling study that costs for screening with FIT over a time period of 20 years (in a cohort of 100,000 persons over 50) are higher in comparison to gFOBT screening. The costs for FIT screening amounted to 78,579,147 € in comparison to gFOBT, which amounted to 74,608,067 € {17}.

Sharp et al. (2012) defined the use of resources on financial ones as well as on the screening-related endoscopic procedures. They indicated that FIT screening is a little bit more expensive than gFOBT screening. But the costs differ only slightly. Costs of FIT screening and CRC management per person was 1,114 € (age 55-74), in comparison to gFOBT, where the costs of screening and CRC management per person amounted to 1,107 € (age 55-74). They presented that in case of gFOBT at 55-74 years (over the entire lifetime of cohort, i.e. 10 screening rounds, rated per 100.000 population) 3.386 colonoscopies have been required, while for FIT screening 34.632 colonoscopies have been required. As regards polypectomy they have indicated that 1.215 polypectomies have been required after the gFOBT screening and 9.486 polypectomies after FIT screening {20}.

In one of the later studies Sharp et al. (2013) estimated the screening-related resource use for biennial gFOBT (at 55-74 years) and biennial FIT (at 55-74 years) as it is presented in the table below. They made a comparison between the first year of screening and the 10th year of screening {30}.

Table 3: Results of Sharp et al. (2013) {30}

10936.ECO-2 Table 3

  It can be concluded from the table above that after FIT screening scenario higher number of colonoscopies and colonographies has been implemented. The numbers are higher for FIT screening also within CRC work-up and treatment. As regards the number of screening tests in year 1 the number of kits sent out and processed was the same for gFOBT and FIT screening scenario, while in year 10 the number of kits sent out and processed was a little lower within FIT screening scenario {30}.

Sobhani et al. (2011) indicated the expected costs for an individual at age 50, using FIT screening, as to 694 €, in comparison to gFOBT, where the expected costs are slightly lower – 584 € per individual. The expected costs for an individual at age 50 after 24 years (3-sample Oc-Sensor) for a cut-off level of 50ng/ml are 1,141 € and for a cut-off level of 100ng/ml 1,593 € gFOBT was estimated to be the cheapest screening test, with expected costs for an individual after 24 years of 1,120 € {21}.

Two US studies {22, 25} also indicated that the costs of FIT screening is comparable to gFOBT screening, but that FIT is still slightly more expensive. Telford et al. have estimated that the costs of FIT, performed annually (over the lifetime of 100,000 individuals, who commence screening at age 50 years) amount to 65,429,821 CAN$, whereas gFOBT, also performed annually (over the lifetime of 100,000 individuals, who commence screening at age 50 years) amounts 63,139,823 CAN$. They have estimated that the strategy of screening with FIT amounts to 1,437 CAN$ per person (in 2007) and with gFOBT to 1,415 CAN$ (in 2007) {22}.

Zauber et al. estimated that FIT screening costs 2,688,092 US$ (per 1,000 50 year olds), whereas gFOBT screening costs 2,369,426 US$ (per 1,000 50 year olds) if Hemoccult II is used and 2,615,292 US$ (per 1,000 50 year olds) if Hemoccult-SENSA is used {25}.

Wilschut et al. (2011) also presented the amount of screening resources, but only for FIT screening. Information on screening resources can be evident from the following table {24}:

Table 4: Results of Wilschut et al. (2011) {24}

10936.ECO-2 Table 4

Their results showed (as also indicated in the table) that FIT screening (cut-off at 50ng/ml) amounts to 493€ per 1,000 individuals aged 45–80 years during the year 2005. For an unlimited capacity, it was most beneficial to screen intensively with the lowest FIT hemoglobin cutoff level for referral to colonoscopy set at 50 ng hemoglobin per mL for those aged 45–80 years with an annual screening interval and offering colonoscopy surveillance to all individuals with adenomas. The colonoscopy demand with this strategy was 49 colonoscopies per 1000 individuals. To optimally adapt screening when capacity was limited to 40 colonoscopies per 1000 individuals, individuals with a FIT hemoglobin measurement between 50 and 75 ng hemoglobin per mL were no longer referred to colonoscopy and individuals between ages 45 and 50 years were no longer invited. This decreased the demand to 36 colonoscopies per 1000 individuals. If capacity was limited to 20 colonoscopies per 1000 individuals, the next step was to further increase the FIT hemoglobin cutoff to 200 ng/mL and to stop screening 5 years earlier at age 75. Also surveillance colonoscopies in individuals with only one or two non-advanced adenomas were cancelled. If colonoscopy demand had to decrease even further, it became efficient to greatly reduce the number of screening rounds by first narrowing the age range to 60–80 years and lengthening the screening interval to 2 years (11 rounds) to reach a demand of 10 colonoscopies per 1000 individuals, and then to narrow the age range to 60–69 years every 3 years (four rounds) for a final capacity of five colonoscopies per 1000 individuals {24}.

 

Parekh et al. (2008) estimated that the costs of FIT screening amounts to 2,428 US$, per person, whereas the costs of gFOBT screening amounts to 2,683 US$ per person. The results are presenting CRC cases and costs per 100,000 persons from age 50 – 100 years {18}.

 

Van Rossum et al. (2011) concluded that over a period of 10 years, an average person aged between 50 and 75 years would cost the healthcare system, €327 with G-FOBT and €301 with FIT screening. These costs included CRC-related costs {19}.

 

The last study, which found FIT screening to be less expensive than gFOBT screening, is the study of Whyte et al. (2012). They concluded that FIT screening amounts to 530 £ per person (age 60-74) over a lifetime, whereas gFOBT screening amounts to 558 £ per person. They also indicated the endoscopy resource use requirements for a cohort of 649,400 individuals – all 50 years-old, monitoring the data for year 2010. The results showed that after biennial gFOBT screening strategy (population of 60-74 year-olds) 38.242 screening colonoscopies and 21.030 surveillance colonoscopies have been performed. In case of FIT screening strategy (population of 60-74 year-olds) 117,681 screening colonoscopies and 43,254 surveillance colonoscopies have been performed {23}.

 

The review made on amount of resources revealed that the majority of studies mentioned mainly financial resources, although some studies specified also other types of resources used, namely: the number of screening tests, colonoscopies, colonographies, polypectomies, number of ultrasounds, number of receiving PET scan, MRI scan, CT scan, pre-operative radiotherapy and undergoing colorectal resection.

By comparing the amount of other (i.e. non-financial) types of resources it can be concluded that in studies that mentioned those type of resources the numbers of screening tests used were higher within gFOBT screening strategy in comparison to FIT screening strategy {13-15, 30}. In addition, all studies indicated that the number of colonoscopies and polypectomies performed after the primary screening test were higher after FIT screening strategy in comparison to gFOBT screening strategy {13-15, 20, 30}.

 

As regards the financial resources, the majority of studies revealed that FIT screening can be slightly more expensive than gFOBT screening {11-13, 17, 20-22, 25, 27; 10 and 14 – depends on the test performance and specificity). However, a few studies presented later revealed just the opposite {15, 16, 18, 19, 23}, thus that FIT screening is less expensive than gFOBT screening {15, 16}. The reason for this hasn’t been directly highlighted in the studies. Nevertheless it could be concluded that the costs vary according to the cut-off values and test sensitivities of FOBTs (i.e. FIT and gFOBT). Some studies, for example, indicated that FIT should be used at higher hemoglobin cut-off levels when colonoscopy capacity is limited. This means, if FIT is used at a low cut-off value, a higher number of colonoscopies will be required which consequently results in higher total costs of screening (and vice versa). Van Rossum et al. {19}, for example, indicated that colonoscopy costs have a relatively high impact on the total screening costs because the relative screening costs of gFOBT screening compared to FIT screening is lower in terms of the costs of follow-up colonoscopies.

 

Summary table of type of resources and estimated or assumed amounts of resources used when delivering FIT and its comparator gFOBT:

pdf10936.ECO-2 Table 5

All costs presented in the results use a discount rate of 3% {10, 13, 16-19, 21, 24, 25}, 3.5% {23}, 4% {20} or 5% {12, 22} in their base case analysis; further details can be found under ECO 6. 

Critical
Not
Renner P et al. Result Card ECO2 In: Renner P et al. Costs and economic evaluation In: Jefferson T, Cerbo M, Vicari N [eds.]. Fecal Immunochemical Test (FIT ) versus guaiac-based fecal occult blood test (FOBT) for colorectal cancer screening [Core HTA], Agenas - Agenzia nazionale per i servizi sanitari regionali; 2014. [cited 21 June 2021]. Available from: http://corehta.info/ViewCover.aspx?id=206

References