David Atkins, MD, MPH; David Ross, MD, PhD, MBI; Michael Kelley, MD
Disclaimer: The opinions represented here are those of the authors and do not reflect official policy of the Department of Veterans Affairs or the federal government.
Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M14-1344.
Requests for Single Reprints: David Atkins, MD, MPH, Director, Health Services Research and Development, Office of Research and Development, Department of Veterans Affairs, 810 Vermont Avenue NW (10P9H), Washington, DC 20420.
This article was published online first at www.annals.org on 17 June 2014.
Current Author Addresses: Dr. Atkins: Director, Health Services Research and Development, Office of Research and Development, Department of Veterans Affairs, 810 Vermont Avenue NW (10P9H), Washington, DC 20420.
Dr. Ross: Director, HIV, HCV, and Public Health Pathogens Programs, Office of Public Health, Department of Veterans Affairs, 810 Vermont Avenue NW (10P3B), Washington, DC 20420.
Dr. Kelley: National Program Director for Oncology, Durham VA Medical Center, 508 Fulton Street, Hematology/Oncology (111G), Durham, NC 27705.
Atkins D., Ross D., Kelley M.; Acting in the Face of Uncertainty. Ann Intern Med. 2014;161:300-301. doi: 10.7326/M14-1344
Download citation file:
Published: Ann Intern Med. 2014;161(4):300-301.
The Achilles' heel of evidence-based medicine is that the state of evidence for many common clinical interventions is not sufficient to either confirm or refute their benefits. Screening for hepatocellular carcinoma (HCC) in high-risk patients with cirrhosis is recommended by guidelines from the American Association for the Study of Liver Diseases (1) and is practiced in many parts of our health care system (the Veterans Health Administration [VHA]). Data from the VHA indicate that 56% of the approximately 25 000 patients with cirrhosis and chronic hepatitis C under VHA care in 2013 received at least 1 imaging test within the past 12 months. Yet, as noted in the review by Kansagara and colleagues in this issue (2), the evidence of the mortality benefit of screening remains insufficient to make a strong recommendation for or against screening. We oversee different areas relevant to this clinical puzzle at the VHA: health services research, care for high-risk patients with viral hepatitis, and oncology services. Although our individual perspectives undoubtedly shape our opinions about what to do in the face of uncertainty, we agree on much of the evidence, what research might be useful, and the policy options we need to consider while waiting for better evidence.
Hepatocellular carcinoma may seem to be a natural candidate for screening. In high-risk groups in the United States, the risk for HCC may range from 10% to 65% (3) over 10 years; HCC incidence and mortality are increasing in Vietnam-era veterans and others among whom hepatitis C is prevalent; and 5-year survival with HCC in the absence of screening is dismal (<10%). Ultrasonography can detect HCC at an early preclinical stage and can increase the proportion of patients who receive potentially curative therapy (resection or transplant). All of this, however, is necessary but not sufficient to support cancer screening. As noted by Kansagara and colleagues, the evidence that screening reduces cancer mortality has major limitations. One randomized screening trial in China among patients with chronic hepatitis B showed a significant mortality benefit (hazard ratio, 0.63) but had many methodological flaws; a smaller, better-designed Chinese trial showed no benefit but may have used an insufficiently sensitive screening protocol (4). A study examining different screening intervals showed that improving early detection does not necessarily improve survival. Moreover, because hepatitis B virus is associated with HCC in the absence of cirrhosis, it is not clear whether these studies can be generalized to at-risk patients in the United States, most of whom have cirrhosis due to chronic hepatitis C or alcohol use.
The rest of the available data come from observational studies comparing outcomes of screened and unscreened patients with HCC, and such studies are prone to well-known biases when applied to cancer screening: lead-time, length-time (including overdiagnosis), and selection biases (5). Kansagara and colleagues' review demonstrates nicely how these biases could produce the benefits attributed to screening in nonrandomized studies.
Whether a randomized screening trial with HCC mortality outcomes can be conducted today in the United States has been debated. Reviewers of one such proposal to the Department of Veterans Affairs Cooperative Studies Program in 2012 raised significant concerns about feasibility and clinical equipoise (Huang G. Personal communication.). The review by Kansagara and colleagues may reopen considerations of a trial, but without a randomized, controlled trial, what research might address the most important uncertainty? First, an inception cohort study collecting baseline clinical variables and long-term follow-up for all patients eligible for screening could produce less biased estimates than studies following only those diagnosed with cancer. The VHA's National Hepatitis C Clinical Case Registry, an electronic database that follows all veterans receiving VHA care who have been diagnosed with chronic hepatitis C, represents one possible resource (6). Instrumental variable or propensity scoring methods could control for other factors that might influence who gets screened. However, given a likely modest benefit of screening, such studies may not provide strong evidence of benefit.
The benefits of early detection are further limited by the fact that all screened patients have underlying liver disease, which may complicate treatment or shorten survival even if cancer is cured. Kansagara and colleagues' review noted that the high 5-year survival rate in some cohorts of patients treated with surgery or transplant stands in contrast to the low 10-year survival in a large screening study. Transplantation offers treatment of both the cancer and the underlying liver disease, but screening may simply add to the more than 15 000 patients waiting for a transplant in the United States.
Overdiagnosis (an extreme form of length-time bias) occurs when screening detects indolent tumors that meet diagnostic criteria for cancer but would not otherwise have come to clinical attention during the lifetime of the individual. It is best appreciated in prostate cancer and is recognized as a more generalizable problem among types of cancer detected by screening, accounting for up to 20% of screen-detected breast and lung cancer (5). It artificially inflates the case survival rate, but “overdiagnosed” patients obtain no benefit while being subjected to all of the harms of treatment. Although studies reporting that even small hepatocellular tumors grow rapidly and have a poor prognosis may seem to argue against overdiagnosis (7), additional studies in screened U.S. populations could clarify this debate.
Finally, using more recent data to update models of cost-effectiveness of screening may be worthwhile. Assuming a mortality benefit of screening, early studies suggested that ultrasonographic screening may be cost-effective (8), but these models lacked current representative data on costs, adherence to screening and treatment, and long-term survival. Sensitivity analysis could identify the necessary mortality benefit to make screening cost-effective. At a time when health systems are struggling with the costs of expensive new drugs to treat hepatitis C and prevent progression to cirrhosis and HCC, we should examine whether screening is likely to be the best place to invest scarce resources for patients with chronic hepatitis.
What is the appropriate practice and policy response while we wait for better evidence? Some would argue that we should halt all screening in the absence of stronger evidence; this approach would be cost-conscious and would adhere to the Hippocratic principle of “First, do no harm.” Although we agree that current screening should not be expanded and new screening programs should not be initiated, the range of uncertainty includes a clinically important benefit of screening. Screening has a much greater potential to produce benefits exceeding harms in the highest-risk patients, such as those with hepatitis C and cirrhosis, than in the general population. It is appropriate to allow clinicians caring for these patients to continue to offer screening, but offers should be targeted to patients who are good candidates for treatment and should include a shared decision-making approach that explicitly acknowledges the limitations of the evidence.
Most important, we should pair any screening with efforts to collect better data, including baseline characteristics and long-term outcomes in screened and unscreened patients, which may help reduce our uncertainty. In the VHA and integrated systems with electronic records and cancer registries, most of these data can be collected as part of routine clinical care. Given the high mortality from HCC and the high costs of delivering a truly effective program of early detection and treatment, better evidence is imperative. We hope that the VHA and other health systems can collaborate to develop such evidence so patients and their clinicians can be confident they are getting (and delivering) the best care available.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Gastroenterology/Hepatology, Hematology/Oncology, Infectious Disease, Cancer Screening/Prevention, Viral Hepatitis.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only