Authors' recommendations: In summary, for advanced neoplasms, systematic reviews of direct comparison between FIT and standard or sensitive GT did not find sufficient evidence to state whether one FOBT is preferable to another; however for CRC, FIT had higher sensitivity and specificity than standard or sensitive GT. HSAC identified only one RCT, but this study reported similar figures for sensitivity, specificity and PPV between OC-Sensor (FIT) and Hemoccult Sensa (sensitive GT). HSAC identified three observational epidemiological studies with three different brands of FIT (Flexsure, Insure and OC-Micro) found that while FIT might have lower sensitivity compared to sensitive GT (Hemoccult Sensa or Hemoccult II Sensa), all three studies found that at commonly used threshold values, FIT had higher specificity, and higher PPV compared to sensitive GT. When the data of all the tests were pooled together, FIT as a class of test had higher PPV and therefore lower NND compared to sensitive or standard GT for the diagnosis and screening of advanced neoplasms.These suggest that although use of FIT might lead to a higher percentage of individuals being identified as positive for advanced neoplasms, because this family of tests also had lower NND compared to the family of sensitive GT tests in the primary observational studies, this suggests that FIT – when considered only on the basis of screening parameters – might result in more efficient use of colonoscopy to confirm cases of advanced neoplasms.However, the findings of this review need to be interpreted in the light of several limitations and restrictions of the approach taken by this review. First of all, this review is based on very few studies to justify the conclusions. Second, this review has limited itself to consider only screening parameters while conducting head to head comparisons between standard & sensitive GT and FITs. These include sensitivity, specificity, positivity, positive predictive value, and number of colonoscopies needed to confirm one case of advanced neoplasm (NND or NNS). None of these measures by itself can provide any guidance as to the superiority of a screening test for a given diagnosis. A sensitive GT is more sensitive than standard GT but is also less specific. Comparison of diagnostic accuracy between immunochemical and guaiac based faecal occult blood tests for colorectal cancer detection This indicates while a sensitive GT is less likely to miss true positive cases, being low in specificity, it leads to higher false positive individuals who nevertheless need to be followed up by colonoscopy. Thus, a test that is both highly sensitive and highly specific might lead to a balance between increased positivity, increased utilization of colonoscopy services as well as fewer colonoscopies to detect one true case of CRC or advanced neoplasm, leading to better utilization of these services and therefore desirable. Since the thresholds of detection of FIT can be varied to adjust the optimum sensitivity and specificity is advantageous. Third, other factors regarding choice of screening tests must be considered before their deployment but these were not addressed as part of this review. These include relative costs of the screening tests, difficulty of procedures and options that might be acceptable to the population, compliance of, uptake or adoption by people to whom the screening options areoffered, and in the case of CRC, issues around preparation of the screening population and interpretation of the results. Fourth, as stated above, this review has attempted to evaluate two rival screening options (guaiac based tests versus faecal immunochemical tests) in terms of their screening performances only. It did not identify studies that had compared sensitive guaiac versus standard guaiac, nor did it identify studies that had compared different types of FIT tests. Only one study in this review had data about comparison of different threshold levels of FIT (OC-Micro)with each other and with GT (Hemoccult II Sensa). In this study by Rozen and colleagues in Israel (2008), OC- Micro was found to have lower test sensitivity at levels of 125 mg/ 100 g of faeces (equivalent of ng/ml), but even at these thresholds, the specificity of OC-Micro (FIT) was better than that of Hemoccult II Sensa, as was the corresponding PPV, indicating that even when the thresholds of detection are set at higher levels than usual, OC-Micro could still be better in terms of specificity, PPV and NND. Therefore, while this review found some evidence in support of superior performance of FIT tests as a family of screening options over sensitive guaiac tests (and some indirect evidence of superiority of sensitive guaiac tests over standard guaiac tests because of their designed increased sensitivity to identify pseudoperoxidase activity in blood at lower concentrations), the results of the study cannot provide any indication whether one form of FIT is superior to another. Other, newer forms of FOBT and other types of screening tests were not considered as part of this review. Finally, the time frame for this review was between 2004-2008. This time frame was selected because comprehensive systematic reviews based on research on CRC screening prior to 2004 were already available and were included in thisreview. This review found some support from observational epidemiological studies that different forms of FIT (Flexsure, OC-Micro, and Insure) were superior to GT for screening of CRC.In conclusion, this review has evaluated the relative screening properties of standard and sensitive guaiac versus faecal immunochemical tests as possible screening options for advanced neoplasms, in order to inform a process of selection of the better of the two tools for screening of CRC. While the systematic reviews and RCTs were inconclusive whether FIT or GT were likely to be better alternatives, observational epidemiological studies found that even with low sensitivity, all three brands of FIT (Flexsure, Insure and OC-Micro) were superior in PPV and NND when compared to Hemoccult Sensa or Hemoccult II Sensa. This indicates that performance wise, as a single test alternative, FIT might be deemed to make better use of colonoscopy resources, even though they may result in higher overall positivity.

Project Status: Completed

URL for project: http://www.healthsac.net/downloads/publications/HSAC16A%20FOBT-Part2%20220609%20FINAL.pdf

Year Published: 2009

English language abstract: An English language summary is available

Publication Type: Not Assigned

Country: New Zealand

Organisation Name: Health Services Assessment Collaboration

Contact Address: University of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand

Contact Name: hsac@canterbury.ac.nz

Contact Email: hsac@canterbury.ac.nz

Copyright: Health Services Assessment Collaboration (HSAC)