Research ArticleOriginal ResearchA

Cumulative Incidence of False-Positive Results in Repeated, Multimodal Cancer Screening

Jennifer Miller Croswell, Barnett S. Kramer, Aimee R. Kreimer, Phil C. Prorok, Jian-Lun Xu, Stuart G. Baker, Richard Fagerstrom, Thomas L. Riley, Jonathan D. Clapp, Christine D. Berg, John K. Gohagan, Gerald L. Andriole, David Chia, Timothy R. Church, E. David Crawford, Mona N. Fouad, Edward P. Gelmann, Lois Lamerato, Douglas J. Reding and Robert E. Schoen
The Annals of Family Medicine May 2009, 7 (3) 212-222; DOI: https://doi.org/10.1370/afm.942

Article Figures & Data

Figures

Tables

  • Table 1.

    Criteria for a Positive Screening Examination and Classification of Diagnostic Follow-Up Procedures

    Diagnostic Follow-Up Procedures
    Screening ModalityDefinition of a Positive Screening ResultMinimally Invasive ProceduresModerately Invasive ProceduresMajor Surgical Procedures
    CA-125 = cancer antigen 125; CXR = chest radiograph; DRE = digital rectal examination; FSG = flexible sigmoidoscopy; PSA = prostate-specific antigen; TVU = transvaginal ultrasonography.
    a Assuming a perfect sphere and using the prolate ellipsoid formula, this volume correlates to an approximate diameter of 2.67 cm. A 10-cc cutoff was chosen, as it corresponds to 2 standard deviations above mean ovarian volume for postmenopausal women.24
    b We define a true-positive FSG as both overt cancer and advanced adenomas (lesions with villous histologic findings, severe cellular dysplasia, and/or ≥1 cm in diameter).
    PSA>4 ng/mLEndoscopyBiopsyProstatectomy
    Cystourethroscopy
    Cystoscopy
    Cystopanendoscopy
    Lymphadenectomy
    Ureterogram
    Transurethral resection of the prostate (TURP)
    DREOne or more of the following findings:As for PSAAs for PSAAs for PSA
        Nodularity
        Induration
        Asymmetry
        Loss of anatomic landmarks
    CA-125>35 U/mLColonoscopyBiopsyHysterectomy
    CuldocentesisLaparotomy
    HysteroscopyOmentectomy
    Intra-abdominal washingsOophorectomy
    LaparoscopyOvarian lymphadenectomy (with laparotomy)
    Paracentesis
    TVUOne or more of the following findings:As for CA-125As for CA-125As for CA-125
        Ovary or cyst >10 cca
        Solid area or papillary projection extending into the cavity of a cystic ovarian tumor of any size
        Mixed (solid/cystic) component within a cystic ovarian tumor
    CXROne or more of the following findings:BronchoscopyBiopsyLung resection
        NoduleLymphadenectomyThoracotomy
        MassMediastinoscopy
        Hilar or mediastinal lymph node enlargementThoracoscopy
        Major atelectasis or lobar collapseTransbronchial aspiration
        Infiltrate, consolidation, or alveolar opacityTransthoracic aspiration
    FSGOne or more of the following findingsb:Colonoscopy (with- out biopsy)BiopsyColon resection
        Rectal nodule(s)CystoscopyHemicolectomy
        Rectal and/or colon mass(es)EndoscopyLymphadenectomyLaparotomy
        Colon polyp(s)Rigid sigmoidoscopy
  • Table 2.

    Proportion of Participants With 1 or More False-Positive Results

    False-Positive ResultParticipants No. (%)
    118,394 (26.9)
    26,043 (8.8)
    32,531 (3.7)
    41,535 (2.2)
    5642 (0.9)
    6228 (0.3)
    778 (0.1)
    853 (0.1)
    911 (0.0)
    102 (0.0)
    110 (0.0)
    120 (0.0)
    130 (0.0)
    140 (0.0)
  • Table 3.

    Initial Entry Round (Year 1) Through Last Entry Round (Year 4) of Screening: False-Positive and Resulting Diagnostic Follow-Up Rates

    Physician Response
    Repeated ScreeningImagingMinimally Invasive ProcedureaModerately Invasive ProcedurebMajor Surgical Procedurec
    SubgroupEver False Positive No. (%)No. (%)d% of FPseNo. (%)d% of FPseNo. (%)d% of FPseNo. (%)d% of FPseNo. (%)d% of FPse
    CA-125 = cancer antigen 125; CXR = chest radiograph; DRE = digital rectal examination; FP = false positive; FSG = flexible sigmoidoscopy; PSA = prostate-specific antigen; TVU=transvaginal ultrasonography.
    Note: These percentages reflect only the subset of diagnostic procedures directly caused by a false-positive result; they do not include procedures associated with a true-positive or negative.
    a Example: simple endoscopy with conscious sedation (see Table 1 for the complete list of procedure classifications).
    b Example: tissue removal, more involved instrumentation, or general anesthesia, including laparoscopy.
    c Example: prostatectomy, or laparotomy with colectomy or oophorectomy.
    d Percentage among participants in the intervention arm that took at least 1 screening examination.
    e Percentage among participants in the intervention arm that received at least 1 false-positive examination (change in denominator). This percentage (and the accompanying crude numbers) does not include procedures that resulted in a true positive (diagnosis of cancer).
    f In each case, the denominator is the total number of participants taking examinations for a given screening modality; this number varied depending on differences in compliance with different tests.
    Overall rates29,517 (43.1)13,576 (19.8)46.07,306 (10.7)24.82,049 (3.0)6.910,822 (15.8)36.71,062 (1.6)3.6
        Men17,432 (50.9)9,251 (27.0)53.14,056 (11.9)23.31,088 (3.2)6.26,972 (20.4)40.047 (0.1)0.3
        Women12,085 (35.3)4,325 (12.6)35.83,250 (9.5)26.9961 (2.8)8.03,850 (11.3)31.91,015 (3.0)8.4
    By screening modality: menf
        PSA3,388 (10.4)2,884 (8.9)85.11,494 (4.6)44.11 (0.0)0.01,491 (4.6)44.06 (0.0)0.2
        DRE4,882 (15.0)3,881 (12.0)79.51,326 (4.1)27.20 (0.0)0.01,259 (3.9)25.81 (0.0)0.0
        CXR6,320 (18.6)3,216 (9.5)50.91,466 (4.3)23.252 (0.2)0.877 (0.2)1.235 (0.1)0.6
        FSG8,186 (26.8)645 (2.1)7.9157 (0.5)1.91,036 (3.4)12.74,821 (15.8)58.95 (0.0)0.1
    By screening modality: womenf
        CA-125888 (3.0)567 (1.9)63.9349 (1.2)39.30 (0.0)0.0103 (0.4)11.6125 (0.4)14.1
        TVU2,310 (7.8)745 (2.5)32.31,394 (4.7)60.41 (0.0)0.0677 (2.3)29.3874 (3.0)37.8
        CXR5,531 (16.3)2,907 (8.6)52.61,498 (4.4)27.156 (0.2)1.093 (0.3)1.740 (0.1)0.7
        FSG5,239 (17.2)321 (1.1)6.1151 (0.5)2.9906 (3.0)17.33,072 (10.1)58.67 (0.0)0.1

Additional Files

  • Supplemental Tables 1-4

    Supplemental Table 1. Estimators and 95% Confidence Intervals of the Cumulative Risk of Receiving at Least 1 False-Positive Result in 14 Tests; Supplemental Table 2. Estimators and 95% Confidence Intervals of the Cumulative Risk of Receiving at Least 1 False-Positive Result, by Screening Modality; Supplemental Table 3. Estimators and 95% Confidence Intervals of the Cumulative Risk of Receiving at Least 1 Invasive Diagnostic Procedure as a Result of a False-Positive Screening Test; Supplemental Table 4. Estimators and 95% Confidence Intervals of the Cumulative Risk of Receiving at Least 1 Invasive Diagnostic Procedure as a Result of a False-Positive Screening Test, by Screening Modality

    Files in this Data Supplement:

  • The Article in Brief

    Cumulative Incidence of False-Positive Results in Repeated, Multimodal Cancer Screening

    Jennifer Miller Croswell , and colleagues

    Background Although cancer screening tests play an important role in early detection of the disease, they can produce false-positive results (incorrect results that mistakenly identify cancer). This study examines the costs, in resources and human terms, of false-positive test results over time.

    What This Study Found A large study that screened for cancers of the prostate, lung, colon and ovary found that the risk of a false-positive result increases with the number of screening tests. Specifically, by the fourth test, the risk of having at least one false-positive is about 37 percent for men and 26 percent for women. By the 14th test, the risk is approximately 60 percent for men and 49 percent for women. The risk of undergoing an invasive diagnostic procedure because of false-positive results is about 17 percent for men and 12 percent for women after four tests and 29 percent for men and 22 percent for women after 14 screening tests.

    Implications

    • This study provides a clearer picture of the burdens and risks associated with multiple cancer screening programs.
    • Physicians and patients should discuss the likelihood of false-positive results in cancer screening, and they should examine the balance of risks and benefits in deciding on a patient's best course of action.

  • Annals Journal Club Selection:

    May/Jun 2009

    The Annals of Family Medicine encourages readers to develop the learning community of those seeking to improve health care and health through enhanced primary care. You can participate by conducting a RADICAL journal club, and sharing the results of your discussions in the Annals online discussion for the featured articles. RADICAL is an acronym for: Read, Ask, Discuss, Inquire, Collaborate, Act, and Learn. The word radical also indicates the need to engage diverse participants in thinking critically about important issues affecting primary care, and then acting on those discussions.1

    How it Works

    In each issue, the Annals selects an article or articles and provides discussion tips and questions. We encourage you to take a RADICAL approach to these materials and to post a summary of your conversation in our online discussion. (Open the article online and click on "TRACK Comments: Submit a response.") You can find discussion questions and more information online at: http://www.AnnFamMed.org/AJC/.

    Article for Discussion

    • Croswell JM, Kramer BS, Kreimer AR, et al. Cumulative incidence of false-positive results in repeated, multimodal cancer screening. Ann Fam Med. 2009;7(3):212-222.

    Discussion Tips

    This analysis of a large cancer screening trial2 shows a negative aspect of cancer screening: the risk of a false-positive test and resulting further invasive testing. As a background for reviewing this article, it might be helpful to think of a range of patients: patients who have experienced a false-positive test, patients who did or did not do screening and were found to have cancer, patients who want every test. You also may wish to use discussion of this article to prime participants to look for further publications of results from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial.

    Discussion Questions

    • What question is addressed by the article? How does the question fit with what already is known on this topic?
    • How strong is the study design for answering the question?
    • What is the degree to which can the findings be accounted for by:
    1. How participants were selected? The exclusion criteria and loss to follow-up? Are any biases likely to be important?
    2. How outcomes were measured?
    3. Confounding (false attribution of causality because two variables discovered to be associated actually are associated with a 3rd factor)?
    4. Chance?
  • What are the main findings? How large is the effect for individual tests? What is the cumulative effect of repeated screening tests?
  • What is the downside of screening tests in terms of follow-up invasive testing? What are possible other unintended consequences (both negative and positive)?
  • What are the implications for how we talk with patients about screening tests?
  • How might the forthcoming results of other outcomes of the PLCO trial affect how we interpret and use these findings?
  • How transportable are the findings to your clinical setting? To other types of screening tests? What factors might affect this transportability?
  • What are some next steps for applying the findings or answering other questions that this study raises?

    • References

    1. Stange KC, Miller WL, McLellan LA, et al. Annals journal club: It�s time to get RADICAL. Ann Fam Med. 2006;4(3):196-197. http://annfammed.org/cgi/content/full/4/3/196.
    2. Early Detection Research Group (EDRG). Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial. http://prevention.cancer.gov/programs-resources/groups/ed/programs/plco. Accessed Feb 1, 2009.
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Get Permissions

Jump to section

Related Articles

Cited By...

More in this TOC Section

Similar Articles

Subjects