Background and aims
Colorectal cancer costs the NHS > £1B each year.41 Preventing or diagnosing this cancer earlier will reduce the costs associated with treatment and complications. Since the introduction of the national BCSP in 2006 there has been a marked increase in demand for NHS endoscopy services. Surveillance colonoscopies constitute a sizeable proportion of all colonoscopies carried out in the NHS. A national colonoscopy audit from 2011 estimated that surveillance colonoscopies accounted for approximately 20% of all colonoscopies performed in the UK over a 2-week period.21 Therefore, the implications of changing surveillance for NHS resource use and costs are likely to be substantial.
The aim of this economic evaluation was to undertake a costing and cost-effectiveness analysis, comparing costs and outcomes of patients who underwent surveillance colonoscopy following baseline colonoscopy with those who did not. We follow the clinical analysis presented in this report by comparing a strategy of surveillance with no surveillance within the three main risk groups defined in the 2002 UK-ASG,7 each of which we additionally stratified into lower- and higher-risk subgroups on the basis of baseline CRC risk factors (as described in Chapter 3). However, unlike the clinical analysis, we make no comparison with the general population. (For an economic analysis this would require detailed information on resource use, as well as incidence rates of CRC.)
Methods
Form of evaluation
The economic analysis consists of both a within-study analysis using individual patient-level data recorded on the study database and a lifetime analysis using a Markov model. The study database provided resource use information on the number and type of colonic examinations. Both analyses included all 28,972 patients included in the clinical analysis (i.e. 14,401 low-risk patients, 11,852 intermediate-risk patients and 2719 high-risk patients).
For the within-study analysis an annual rate of total cost per person-year was calculated for each risk subgroup across a median follow-up of 9.3 years. As the study database did not include any information on QoL, it was not possible to estimate QALYs directly from the study data. The within-study analysis used the diagnosis of CRC as the main outcome measure, with cost-effectiveness expressed in terms of the incremental cost per CRC prevented.
An extrapolation model was then used to estimate the cost-effectiveness of each surveillance strategy over the lifetime of the cohort. Surveillance costs and transition probabilities for each risk subgroup were estimated using patient-level data on the study database. For this analysis to measure outcomes using QALYs, QoL data were obtained from the PHE PROMs survey of CRC patients known to the National Cancer Registration and Analysis Service37 and combined with survival data from the Northern and Yorkshire Cancer Registry and Information Service.42
Estimation of costs
shows the unit costs used in our analyses and the sources of these costs. Costs were applied to the resource use associated with both adenoma surveillance and cancer treatment. The analysis was undertaken from the perspective of the NHS, with costs reported in GBP using 2017/18 prices. Unit costs for surveillance procedures were taken from the NHS national schedule of reference costs for 2017–18.33 Estimates for the lifetime cost of CRC treatment were taken from published estimates from a whole disease model of CRC.44 These costs were inflated from 2012/13 prices to 2017/18 prices using the gross domestic product deflator.
The definition of surveillance visits used in previous chapters is maintained in this chapter. Visits were costed by applying unit costs to each colonic examination. Costs were different for colonoscopies, flexible sigmoidoscopies and rigid sigmoidoscopies, and varied depending on whether the examination was diagnostic or therapeutic. The distinction between a diagnostic and a therapeutic examination was made using data on whether or not a polyp was removed or whether there was an associated biopsy or pathology report.
Within-study analysis
The within-study analysis compared patients who underwent surveillance with those who did not for each of the six risk subgroups (i.e. for the lower- and higher-risk subgroups within each of the low-, intermediate- and high-risk groups). Cost-effectiveness was measured in terms of the incremental cost per CRC prevented by adopting surveillance compared with no surveillance. This follows the approach of a recent cost-effectiveness study of post-polypectomy surveillance.45
Total annual costs and CRC incidence rates were calculated for each of the three main risk groups and for the lower- and higher-risk subgroups within each risk group. Poisson models were used to estimate annual CRC incidence rates for each risk group across varying exposure time. Differences in CRC incidence rates were compared using the Wald test, with SEs being combined using the delta method. We calculated ICERs as the ratio between the difference in costs and the difference in CRC incidence rates.
Missing data were infrequent in the data set, with the exception of CRC staging information, which was missing for 32% of CRC cases. Logit models indicated that these missing data were significantly positively associated with age at baseline and negatively associated with the date of the surveillance visit (data not shown). Missing CRC staging data were handled by means of multiple imputation using an ordered logit model. The procedure was repeated to produce 40 imputed data sets, with Rubin’s rule used to summarise across imputations.46 In the main analysis, total costs were then estimated based on the imputed data. Ten patients in the data were untraceable. These patients were treated as having had no surveillance visits and having not developed CRC. Both costs and number of CRC cases were discounted at a rate of 3.5% per year. The analysis was conducted in Stata®/SE 14 (StataCorp LP, College Station, TX, USA).
Extrapolation model
We designed a multistate Markov model to extrapolate the results from the within-study analysis over a lifetime horizon. shows the structure of the model. The model consisted of eight states: (1) no surveillance visits, (2) a positive count of surveillance visits, (3) Dukes’ stage A CRC, (4) Dukes’ stage B CRC, (5) Dukes’ stage C CRC, (6) Dukes’ stage D CRC, (7) death from CRC and (8) death from other causes. Time-homogeneous transition probabilities were estimated for each of the three main risk groups using state and time data from the main study database and using the msm package in R (The R Foundation for Statistical Computing, Vienna, Austria). SEs for each set of transition probabilities were calculated using an assumed multivariate normal distribution of the maximum likelihood estimates and covariance matrix.
Markov model for our lifetime economic analysis.
All patients begin the model in the ‘no visits’ state, having attended a baseline colonoscopy but having had no surveillance visits. In each cycle patients have a probability of attending a surveillance visit that is dependent on which risk group they are in. The patients also face a probability of developing CRC and having it diagnosed at Dukes’ stage A, B, C or D. A transition to the ‘visits’ state reduces the patient’s risk of CRC and this benefit is assumed to last for the remaining length of the model. The transition probabilities between the ‘no visits’ state and the ‘visits’ state, as well as the transitions to each of the cancer states, were estimated separately for each risk group.
The numbers of surveillance visits and intervals between them as observed within each risk subgroup in the study data were used to estimate a constant annual probability for each subgroup. The model does not explicitly model the impact of the number of surveillance visits as the surveillance strategies being compared are defined in terms of whether or not surveillance occurred. However, the impact of the number of surveillance visits on CRC cases diagnosed is captured in the probability of transitioning from the ‘visits’ state to each of the four cancer states.
Missing CRC staging data were handled using the imputed data from the within-study analysis. Transition probabilities were calculated for each of the 40 imputed data sets, with SEs estimated by combining the estimates using Rubin’s rule.46 The sensitivity of the results to the use of the imputed CRC data were assessed by collapsing the model to a single cancer state for each of the risk subgroups.
A full set of transition probabilities and SEs could not be estimated for the higher-risk subgroup of the high-risk group because of a failure of model convergence. Therefore, a simplified version of the model was used for this group, in which the four cancer states were collapsed to a single state. QoL estimates, CRC treatment costs and CRC survival rates, which are separately estimated for each cancer state, were collapsed to mean values for this subgroup.
The probability of death from any cause, QoL estimates and costs associated with CRC are dependent on the age of the cohort. These parameters required an initial age to be specified for the model. The initial age used for the base-case analysis was 60 years, as this was the mean age of patients in the hospital cohort.
The main clinical study was not powered to identify differences in CRC-related mortality by stage and there were relatively few CRC deaths in each category when stratifying by stage, which would make extrapolation based on these data unreliable. Therefore, the probabilities of survival of CRC were sourced from the 2013 National Bowel Cancer Audit Annual Report.35 The probability of death from any cause was taken from national lifetables published by the Office for National Statistics.36
Resource use and costs
The extrapolation model includes costs for colonoscopy and lifetime costs associated with the diagnosis and treatment of CRC. A separate cost for adenoma surveillance was entered into the model for each risk subgroup. This cost was derived by estimating the mean annual cost of surveillance per patient during the within-study period and then applying that cost over the modelled duration of surveillance, from a starting age of 60 years through to 75 years, the upper age limit in the 2002 UK-ASG.7 The lifetime cost of CRC was applied only to the first cycle in which CRC was diagnosed.
Quality of life
Quality-of-life data were not collected in the main clinical study. Therefore, the primary outcome measure in the within-study analysis was the incidence of CRC. The extrapolation model used estimated mean EQ-5D scores from a one-off study of CRC patients by a PHE PROMs survey.37 This study collected QoL data from 21,802 patients who were diagnosed with CRC in England from 2010 to 2011 and were alive between 12 and 36 months after diagnosis, when they were sent a questionnaire including the EuroQol-5 Dimensions, five-level version. Responses were then scored using the ‘crosswalk’ mapping function developed by van Hout et al.47 This allowed for comparison with the utility estimates derived from the EuroQol-5 Dimensions, three-level version (EQ-5D-3L), which was used for the non-cancer health states. The EQ-5D-3L scores for the non-cancer health states came from Ara and Brazier,38 who reported EQ-5D-3L scores from pooled responses in the Health Survey for England from people without cancer.
The Ara and Brazier38 scores and the estimates from the PROMs survey37 were combined with estimated survival from the lifetime model to calculate QALYs. The use of the PROMs survey for QoL data differs from recent studies examining the cost-effectiveness of CRC screening or surveillance strategies,23,43 which have mainly used estimates reported in Ara and Brazier.38 The estimates from the PROMs survey37 were preferred over those from Ara and Brazier38 as the former is the largest UK survey of QoL for CRC patients currently available, and provides estimates by both CRC stage and age. Whyte et al.48 estimated EQ-5D-3L scores by age and estimated CRC stage; however, these estimates are based on all cancer patients included in the Health Survey for England rather than only CRC patients. Therefore, the PROMs data were deemed more representative of the patients in our study cohort. The effect of using the estimates from Whyte et al.48 on our model results was assessed in a sensitivity analysis.48
Future QALYs and costs were discounted to present values at an annual rate of 3.5%. ICERs were then calculated as the ratio between the mean difference in QALYs and the mean difference in costs. Cost-effectiveness was evaluated assuming a willingness-to-pay threshold of £20,000 per QALY.49
Sensitivity analyses
Uncertainty in the model estimates was characterised using a deterministic sensitivity analysis (DSA) and probabilistic sensitivity analysis (PSA). The DSA assessed the sensitivity of the model estimates to variation in individual parameters, whereas the PSA aimed to estimate the joint effect of uncertainty in all the parameters. For the DSA each parameter was both increased and decreased by 25% of its baseline value. The impact of these changes was measured in terms of the effect on the estimated ICER for each surveillance strategy. A sensitivity analysis was also used to assess the impact of using different strategies to handle the missing CRC staging data.
In the PSA all transition probabilities, costs and QoL estimates were varied, except for the survival probabilities. The PSA was carried out by sampling 2000 sets of the model parameters drawn at random from appropriate statistical distributions. Uncertainty around each cost estimate was characterised using a gamma distribution, whereas beta distributions were fitted to both the transition probabilities and QoL estimates. These replications were used to plot the cost-effectiveness plane50 and to construct cost-effectiveness acceptability curves that show the likelihood that the intervention is cost-effective as the willingness-to-pay changes.51
shows the estimated transition probabilities, costs and QoL estimates used in the model. The transition probabilities between the ‘no visit’ and ‘visit’ states differ between risk groups, with the lowest probability estimated for the lower-risk subgroup of the low-risk group and the highest probability estimated for the higher-risk subgroup of the high-risk group. For the whole low-risk group and the higher-risk subgroup of intermediate-risk patients the transition probabilities to the different CRC states were mostly lower with surveillance than without. In contrast, for the lower-risk subgroup of intermediate-risk patients and the lower-risk subgroup of high-risk patients the transition probabilities to CRC were mostly higher with surveillance than without. Mean EQ-5D scores were higher for patients with CRC than for people in the same age group without cancer.
Transition probabilities, costs and QoL estimates used in our extrapolation model
The below list summarises the main modelling assumptions made in the lifetime analysis:
The level of surveillance recorded in the study data set reflects practice under the 2002 UK-ASG.
7The level of surveillance offered to each risk group is independent of the level of surveillance offered to the other risk groups.
Transition probabilities are time homogeneous.
Patients who did not attend surveillance are representative of patients who attended surveillance if their surveillance was withdrawn.
There is no age cut-off point for surveillance in either of the considered surveillance strategies.
There were no changes to the BCSP during the modelled period.
The probability of a bleed occurring during a surveillance procedure is independent of polypectomy being performed.
The probability of a bleed or bowel perforation occurring during a surveillance procedure is independent of risk group.
The costs and QoL estimates associated with surveillance can be captured by explicitly modelling the growth of polyps.
Although we assumed that the probability of a bleed occurring during a surveillance procedure is independent of polypectomy being performed, polypectomy does in fact increase the risk of a bleed.52 However, a simplification of the model is that it does not distinguish between a polypectomy being performed or not. Therefore, when applying the probability of a bleed it was necessary that the probability was independent of a polypectomy being performed. To achieve this we used an average across all colonoscopies performed sourced from Rutter et al.52
Results
Within-study analysis
Of the 525 CRCs in the study, 167 (32%) were missing CRC staging data. The proportion of missing CRC staging data varied between the risk groups. The highest proportion was 35% in the higher-risk subgroup of low-risk patients, whereas the lowest proportion was 24% in the lower-risk subgroup of intermediate-risk patients. However, the risk group was not a significant predictor of missing CRC staging data (data not shown).
The low-risk group
shows the mean surveillance resources used and their associated cost in the lower- and higher-risk subgroups of the low-risk group over the full follow-up period.
Surveillance resource use and costs in the low-risk group
Among those attending surveillance, the mean total number of surveillance examinations performed in the lower- and higher-risk subgroups was 1.78 and 2.04, respectively. The mean total cost of surveillance was £1041 in the lower-risk subgroup and £1231 in the higher-risk subgroup (see ).
In the lower-risk subgroup of low-risk patients, the mean discounted total cost for patients attending surveillance was £133,612 per 1000 person-years, whereas the equivalent figure for patients not attending surveillance was £1906 per 1000 person-years. Therefore, in this subgroup, the total incremental cost for those with surveillance compared with those with no surveillance was £131,706 (see ).
In the lower-risk subgroup of low-risk patients, more CRCs were diagnosed among those with no surveillance than among those with surveillance, at 0.94 and 0.65 per 1000 person-years, respectively. When we combined the difference in costs and CRC cases, the incremental cost per CRC prevented was £453,221 (see ).
In the higher-risk subgroup of low-risk patients, the total incremental cost for those with surveillance compared with those with no surveillance was £137,081 per 1000 person-years, similar to that in the lower-risk subgroup. However, the difference in the CRC incidence rate among those with no surveillance and those with surveillance was greater than in the lower-risk subgroup, at 2.17 and 1.10 per 1000 person-years, respectively. Therefore, when we combined the difference in costs and CRC cases the incremental cost per CRC prevented was £127,945, which is far lower than in the lower-risk subgroup (see ).
The intermediate-risk group
In the intermediate-risk group surveillance resource use was similar in the lower- and higher-risk subgroups (). The total incremental cost of surveillance per 1000 person years was £140,780 in the lower-risk subgroup and £153,409 in the higher-risk subgroup. As the incidence of CRC was lower in the lower- than higher-risk subgroup, the incremental cost per CRC prevented was higher in the lower-risk subgroup (i.e. £2,587,860) than in the higher-risk subgroup (i.e. £145,729) (see ).
Surveillance resource use and costs in the intermediate-risk group
The high-risk group
The high-risk group had the greatest level of surveillance resource use and associated costs of any of the three risk groups (). The total incremental cost of surveillance per 1000 person-years was similar in the lower- and higher-risk subgroups of high-risk patients, at £196,436 and £186,212, respectively. However, because the subgroups had different CRC incidence rates, the incremental costs per CRC prevented differed significantly, being £568,719 in the lower-risk subgroup and £36,636 in the higher-risk subgroup (see ).
Surveillance resource use and costs in the high-risk group
Appendix 3, Tables 43–45, report total costs for each baseline risk group when using only complete CRC staging data. These tables apply mean costs by age group to known CRC cases with unknown staging data. This change did not alter the main results to a large degree, as the lifetime cost of CRC care made up a relatively small proportion of the total cost.
Lifetime analysis
A comparison of observed and predicted percentage in each state from a single imputed data set showed that the model fitted the data reasonably well (see Appendix 3, –). The results from our lifetime model are shown in . For each of the three main risk groups ICERs were lower in the higher- than in the lower-risk subgroup.
Estimates of cost-effectiveness from our extrapolation model
In the case of the low-risk group, the incremental cost of surveillance per patient was only slightly higher in the higher-risk subgroup than in the lower-risk subgroup. However, as the QALY benefit from surveillance was much larger in the higher-risk subgroup than in the lower-risk subgroup, the ICER was approximately five times lower in the higher- than in the lower-risk subgroup (see ).
A similar pattern was found in the intermediate-risk group. In both the higher- and lower-risk subgroups of intermediate-risk patients, the incremental cost of surveillance was approximately £1000 per patient. However, as the higher-risk subgroup had a higher QALY gain from surveillance than the lower-risk subgroup, the ICER in the higher-risk subgroup was approximately £47,000, whereas surveillance in the lower-risk subgroup was dominated (i.e. more costly and less effective) (see ).
The same pattern was observed in the high-risk group. Surveillance was dominated in the lower-risk subgroup, whereas the ICER in the higher-risk subgroup was £7821. This shows that surveillance was cost-effective in the higher-risk subgroup of high-risk patients when using a cost-effectiveness threshold of £20,000 per QALY gained (see ).
Sensitivity analyses
The results of the DSA for the low-risk group are shown in Appendix 3, . In both the lower- and higher-risk subgroups the results were most sensitive to variation in the cost of surveillance. In the case of the higher-risk subgroup, a 25% reduction in the cost of surveillance reduced the estimated ICER to below the £20,000 per QALY threshold. ICERs in the lower-risk subgroup did not fall below this threshold. Estimated ICERs for both subgroups were insensitive to changes in costs associated with CRC treatment.
The results of the DSA for the intermediate-risk group are shown in Appendix 3, . The ICER for the lower-risk subgroup was most sensitive to variation in the transition probabilities between the ‘visits’ state and the Dukes’ stage D state, and between the ‘no visits’ state and the Dukes’ stage C state. The ICER in the higher-risk subgroup was most sensitive to variation in the QoL estimates for the non-cancer states. Variation in the cost of surveillance produced the largest increase in the ICER in the lower-risk subgroup and the largest decrease in the ICER in the higher-risk subgroup. In the case of the higher-risk subgroup, a 25% reduction in the cost of surveillance reduced the ICER to < £30,000 per QALY.
The results of the DSA for the high-risk group are shown in Appendix 3, . For the lower-risk subgroup changes in the transition probability between the ‘visits’ state and the Dukes’ stage D state resulted in large changes in the ICER. In the case of the higher-risk subgroup, no variation in the analysis resulted in an ICER above the £20,000 per QALY threshold. The results for each risk group were relatively insensitive to changes in the estimates of CRC incidence.
The results from the PSA are shown in and . shows the cost-effectiveness plane for each of the three risk groups. The plane shows a similar spread of costs for each risk group and a great variability in QALYs. shows the cost-effectiveness acceptability curves for each risk group. In each risk group surveillance had a higher probability of being cost-effective in the higher-risk subgroup than in the lower-risk subgroup. For example, in the low-risk group the probabilities of surveillance being cost-effective were 11% and 58% in the lower- and higher-risk subgroups, respectively. In the intermediate-risk group the equivalent probabilities were 4% and 34% in the lower- and higher-risk subgroups, respectively. In the high-risk group surveillance had a 10% probability of being cost-effective in the lower-risk subgroup compared with an 86% probability in the higher-risk subgroup (see ).
Cost-effectiveness plane showing the incremental costs and QALYs for surveillance compared with no surveillance from our PSA. (a) Low-risk group; (b) intermediate-risk group; and (c) high-risk group. The dashed line indicates a cost-effectiveness threshold (more...)
Cost-effectiveness acceptability curves showing the probabilities of surveillance being cost-effective at different willingness-to-pay thresholds. (a) Low-risk group; (b) intermediate-risk group; and (c) high-risk group.
shows the expected value of perfect information (EVPI) for each risk group. The greatest value was attached to eliminating uncertainty around the estimates for the higher-risk subgroup of the low-risk group. The EVPI for this subgroup was £2,451,967 at a threshold of £20,000 per QALY, compared with an EVPI of just £18 in the lower-risk subgroup of the low-risk group.
The EVPI for different willingness-to-pay thresholds. (a) Low-risk group; (b) intermediate-risk group; and (c) high-risk group.
In additional sensitivity analyses we found that the results from the extrapolation model were robust to the use of different QoL estimates, from Whyte et al.48 (see Appendix 3, Table 46), and to the use of a simplified model structure with the four cancer states collapsed to a single state (see Appendix 3, Table 47).
Discussion
The economic analysis presented in this chapter examined the cost-effectiveness of surveillance compared with no surveillance in patients in whom adenomas were detected and removed at baseline colonoscopy, considering the three risk groups defined in the 2002 UK-ASG.7 To our knowledge, this is the first analysis of its kind.
The within-study analysis found a high degree of heterogeneity between the three risk groups. In each risk group costs per CRC diagnosis were lower in the higher-risk subgroup than in the lower-risk subgroup. The lack of QoL data for patients in the study meant that a full within-study cost–utility analysis was not feasible. We therefore developed an extrapolation model to estimate the cost-effectiveness of surveillance over a lifetime horizon. The model results showed that for each risk group surveillance was more cost-effective for patients in the higher-risk subgroup than for those in the lower-risk subgroup. However, the higher-risk subgroup of high-risk patients was the only subgroup to have an ICER below the cost-effectiveness threshold of £20,000 per QALY. The ICER in this subgroup was £7821.
The PSA found a relatively high degree of uncertainty at a willingness-to-pay threshold of £20,000 per QALY. This resulted in an EVPI of close to £2.5M over the length of the model. This high figure suggests that future research aimed at reducing this uncertainty is likely to be cost-effective. However, the DSA indicated that much of this variation results from the high degree of variability in the estimated costs of surveillance, which reflects heterogeneity in the treatment of individual patients rather than sampling uncertainty.
To our knowledge, this is the first study to estimate the cost-effectiveness of the surveillance recommendations in the 2002 UK-ASG.7 The previous NIHR Health Technology Assessment report23 using this study database found 3-yearly surveillance with no age cut-off point to be highly cost-effective and the most cost-effective strategy for the intermediate-risk group. However, the results from this previous study address a different research question and are not directly comparable to the results reported here. The results are likely to differ because of differences in model structure. The previous report did not model CRC treatment by stage, whereas the model in this chapter applied EQ-5D scores and costs to CRC states by both stage and age. The analysis in this chapter also used a different source for the EQ-5D scores for the different CRC stages. These scores were higher than those used in the previous report.
The present analysis benefited from the high-quality data of the study database, drawn from 17 UK hospitals. There were few missing data and the follow-up period was long. The estimated transition probabilities for the lifetime model were therefore based on high-quality data.
The study data also allowed model parameters to be estimated separately for each baseline risk subgroup. This allowed the model to capture the heterogeneity between these groups, which have been collapsed in previous studies.23,43,48 However, the model does not account for heterogeneity within these subgroups. For example, the cost of surveillance may be lower for a patient attending a single surveillance visit than for one attending five or more visits, but both patients may experience a similar reduction in CRC risk. The difficulty in modelling this is simultaneously estimating the probability of attending surveillance and the reduction in CRC risk, together with the uncertainty around these estimates. The lifetime analysis stratified patients by baseline risk subgroup and by CRC stage within each subgroup. Stratifying by further covariates, such as age, resulted in the model failing to converge. Even without stratifying further, there were insufficient data to estimate all model parameters for the higher-risk subgroup of high-risk patients. Instead, we estimated a simplified model for this subgroup. There is clear evidence from the clinical data and the simplified extrapolation model that surveillance in this subgroup is both highly clinically effective and cost-effective.
The analysis in this chapter has several limitations. First, both the within-study analysis and the lifetime model assumed that patients who did not attend surveillance were representative in terms of characteristics and outcomes of patients who attended surveillance if their surveillance was withdrawn. However, the findings from the main clinical study show that these groups are not well matched on several baseline characteristics. For example, non-attenders tended to be older than attenders. Age is a positive predictor of CRC risk and, therefore, unadjusted estimates of the difference between the groups may be upwardly biased. Adjusting estimates for this source of bias is not straightforward because the number of surveillance visits attended is a time-variant covariate. For the within-study analysis we examined this source of bias by estimating ICERs for a representative patient at the mean age for each risk subgroup (see Appendix 3, Table 48). These estimates show the results for the higher-risk subgroups to be relatively robust whereas the ICER for the lower-risk subgroups increased. For the lifetime model it is possible to estimate the effect of explanatory variables in a multistate model using a proportional intensities model. However, attempts at estimating age-dependent transition probabilities resulted in the model failing to converge.
Second, the PROMs survey is known to have a relatively high degree of non-response, which varies by both age group and cancer stage. However, the DSA found the results of the extrapolation model to be relatively insensitive to changes in the mean EQ-5D scores associated with the different cancer stages.
An additional limitation is that CRC staging data were missing for 32% of patients. We addressed this using multiple imputation in the main analysis and by using a simplified model structure in a sensitivity analysis, collapsing the four cancer states to a single state. The sensitivity analysis showed that the model results were robust to the use of the simplified model.
Further limitations include the fact that estimated costs of lifetime CRC treatment came from a source that is nearly 10 years old.44 Other studies examining the cost-effectiveness of CRC screening and surveillance strategies have also used this source;23,43,48 however, given the age of the estimates, there is a need for more recent high-quality data on the cost of CRC treatment. This analysis could be extended by considering how a change in the UK surveillance guidelines could affect the BCSP, as it is likely that patients who are no longer offered surveillance would instead be offered a faecal immunochemical test (FIT) as part of routine screening.
