A simulation model of colorectal cancer surveillance and recurrence

Table 1 Estimates for non-calibrated parameters

Variable	Input parameter estimate	Range for sensitivity analysis	Source^a
Screening test performance
*Carcinoembryonic Antigen test (CEA)*
Sensitivity	.64	.49 - .79	[41]
Specificity	.90	.75 – 1.00	[41]
*Chest X ray*
Sensitivity	.76	.61 - .91	[39, 42, 43]
Specificity	.95	.80 – 1.00	[43]
*CT – Hepatic metastases*
Sensitivity	.83	.68 - .98	[44–47]
Specificity	.93	.78 – 1.00	[44, 47]
*CT – Other abdominal metastases*
Sensitivity	.46	.31 - .61	[44]
Specificity	.98	.83 – 1.00	[44]
*Hepatic ultrasound*
Sensitivity	.62	.47 - .77	[47–49]
Specificity	.85	.70 – 1.00	[47–49]
*Colonoscopy*
Sensitivity	.95	.80 – 1.00	[23, 50]
Specificity	1.00	.85 – 1.00	[23, 50]
*Clinical interview/examination* ^b
Sensitivity	.42	.27 - .57	[5]
Specificity	.95	.70 – 1.00	[5]
Life expectancy
After initial surgery, given no recurrence	20.7 years	--^c	[51]
After recurrence with curative salvage	21 months	15 – 27	[52]^d
After diagnosis of unresectable recurrence	8 months	4 – 12	[53]^e

^aWhen multiple sources are given, the parameter used in the model was estimated based on a sample-size-weighted mean.
^bIn the absence of laboratory or other diagnostic findings suggesting recurrence.
^cSince the time horizon for the model was 5 years, varying this parameter over all but the most extreme low bounds would have no effect on results.
^dEstimate based on incomplete reporting of survival among subjects undergoing curative salvage surgery in the Pietra study.
^eBased on reported survival of 13 months following recurrence of colon cancer during the era of 1986–1992, and assumption that 1/3 of those relapsing will undergo curative salvage surgery with survival of 21 months as described above. The result is that those recurring but not undergoing curative salvage surgery would be expected to survive approximately 8 months after diagnosis of recurrence.

ISSN: 1472-6947