Harold Sox, M.D.

Dartmouth Medical School

September 23, 1997

My name is Harold Sox. I am a specialist in internal medicine and chair of the department of medicine at Dartmouth Medical School. I am President-elect of the American College of Physicians, which issued guidelines on prostate cancer screening in 1997. I chaired the United States Preventive Services Task Force from 1990 to 1996. The Task Force issued its prostate cancer screening guidelines in 1995. The Task Force is a federally sponsored panel now administered by the Agency for Health Care Policy and Research. Most people consider the Task Force to be the definitive resource for evidence-based prevention guidelines, owing to its rigorous methods for evaluating the evidence and its reputation for impartiality. The Task Force's recommendations strongly influence coverage decisions in the private sector as well as the patient care quality standards of organizations, such as the National Council for Quality Assurance.

Mr. Chairman, this hearing and related efforts have a critical message to impart: that screening and other preventive services can be life-enhancing and, indeed, life-saving. Congress recognized the value of preventive care when it significantly expanded Medicare coverage in the Balanced Budget Act. With this legislation, Congress has made it possible for many people to receive important services. It is now up to the medical profession to decide the often very complex issues of whom to screen and how often, so that the American people receive good value for the support they provide to preventive services under Medicare.

The issue of whom to screen and how often is nowhere more complex than with prostate cancer, the subject of today's hearing. I wish that I had an easy answer for the committee. I wish that physicians knew enough about prostate cancer and its treatment to provide uniform advice to patients, as we do regarding breast cancer screening for some women. There is very high quality evidence that breast cancer screening reduces the death rate from breast cancer in women aged 50-69 years. There is broad agreement that this evidence is compelling and that physicians should encourage women in this age group to undergo mammography.

In comparison, the evidence for prostate cancer screening, like many other areas of medicine, is very weak, and there is no broad professional consensus that prostate cancer screening is effective. Absent firm scientific grounding, it is not possible or desirable, in my opinion, to promote a uniform policy on screening. As I will discuss, our uncertainty about the value of prostate cancer screening means that physicians make screening decisions on an individual basis. The best policy is shared, informed decision making, in which a physician treats each patient as an individual, teaching him about prostate cancer and helping him to decide. Prostate cancer screening is not for everyone.

Prostate cancer is, in many respects, a more complicated disease than breast cancer or colon cancer, for which we have generally agreed-upon screening policies. Prostate cancers vary in their rate of' growth; half of prostate cancer patients have very slow growing tumors. Prostate cancer is largely a disease of older men, many of whom have other serious diseases. Therefore, most men with prostate cancer die of something else. The side effects of prostate cancer treatment are more frequent, more long-lasting, and more serious than the treatment of cancer of the breast or colon.

The components of screening are the screening test, a test to verify the diagnosis when the screening test is abnormal, and treatments for proven prostate cancer. The two screening tests are digital rectal examination (DRE) and prostate specific antigen (PSA). Positive results are confirmed by a biopsy of the prostate. Patient with a positive biopsy may undergo staging tests, such as MRI and bone scan, to determine the extent of tumor spread. If these tests are negative, the cancer is "clinically localized," and the patient must decide whether to undergo potentially curative treatments, such as radical prostatectomy or radiation therapy, or to choose no treatment.

Why might some patients with prostate cancer decide against potentially curative treatment? Presumably, they feel that the harms of treatment outweigh the benefits in their case. I believe that everyone would agree that the balance between the harms and benefits of screening should determine a patient's decision. We know some of the harms, but crucial information about benefits is missing.


1. Treatment: In the U.S., we introduce new health care technology first and evaluate it only after it has become common practice. As a result, we know a great deal about the harms of radical prostatectomy but next to nothing about its benefits. The principal harms are urinary incontinence and sexual dysfunction. Urinary incontinence was a major problem in a study of Medicare patients who underwent prostate cancer surgery: 32% use pads or a penile clamp, 2% use an indwelling urinary catheter, and 6% require a surgical procedure for incontinence. Sexual dysfunction is widespread also: In the same study of Medicare patients, 91% had erections before surgery, but 61% had had no erection since surgery, and only 11% had erections firm enough for intercourse in the month preceding the interview. Finally, radical surgery for prostate cancer is the cause of death in 1% of patients.

2. The screening tests: Both PSA and digital rectal examination are inaccurate tests that frequently give misleading information. PSA, for example, detects only one-half of patients with clinically localized prostate cancer, so that a normal PSA can give false reassurance that cancer is not present. Because a PSA is abnormal in 10% of people who don't have prostate cancer, only one-third of men with an abnormal PSA have prostate cancer.


Cure of prostate cancer: Screening will have benefits only if it reduces the death rate from prostate cancer by identifying men who can benefit from an effective treatment. But we do not know the effectiveness of treating clinically localized prostate cancer in reducing the mortality rate from prostate cancer. The ideal way to find out is to randomly allocate selected prostate cancer patients to treatment or watchful waiting and measure the death rate from prostate cancer. Randomized clinical trials of screening have been completed for breast cancer, colon cancer, and lung cancer but not for prostate cancer treatment. Several studies are now underway in this country and in Europe. Recently, the American Urological Association Prostate Cancer Clinical Guidelines Panel (December 1995) reviewed all of the evidence and stated their findings as follows: "the panel found the outcomes data inadequate for valid comparisons of treatments." In other words, specialists in prostate cancer treatment agree that there is no proof that radical prostatectomy prolongs life from what it would be with no treatment.

We know that radical prostatectomy is not a perfect treatment because 28% of Medicare patients had undergone treatments for metastatic prostate cancer within 4 years of a radical prostatectomy, presumably because their cancers had recurred. Although many men are cured of a prostate cancer that would have caused them suffering and death, other men suffer the side effects of surgery for a slow-growing cancer that would not have caused either suffering or death. We don't know the balance between benefit and harm in prostate cancer surgery. We would have to tell a patient considering screening "if you have prostate cancer, I won't be able to tell you whether treatment is better than watchful waiting."

The role of patient counseling: The ethical principle of informed consent requires that patients with clinically localized prostate cancer learn about these known harms and potential benefits. Many believe that the same principle applies to the decision to undergo screening. The recently released guidelines of the American College of Physicians stated, "Rather than screening all men for prostate cancer as a matter of routine, physicians should describe the potential benefits and known harms of screening, diagnosis, and treatment, listen to the patient's concerns, and then individualize the decision to screen." The underlying assumption is that the potential harms and benefits will differ from patient to patient, and therefore the balance of harms and benefits will also differ.

Furthermore, patients are likely to differ in how they value a benefit or harm and how they take account of the uncertainty about the benefits of radical treatment of prostate cancer.

Patients can understand complex information that may affect their health and use it to make decisions. In work done at Dartmouth, men at a veterans hospital were randomly assigned to learn about the benefits and harms of prostate cancer screening by watching a videotaped description or to a control group. Eighty percent of the patients who saw the videotape said that they would prefer no treatment if they had clinically localized prostate cancer. Only 40% of those who did not see the videotape said that they would prefer no treatment. This result shows that many patients have considerable ambivalence about prostate cancer treatment. Furthermore, learning more about the known harms and unknown benefits of treatment reduced the number who wanted treatment. This research argues strongly against a uniform policy of screening and strongly for informed, individualized decision making prior to screening.


What harm will a screening test do? Why not screen and then deal with the decision to accept treatment when the patient is fully informed about his own status?" In reply to these questions, I hold to the principle that one should not do something to a patient if it could not alter subsequent steps in evaluating him. Many fully informed patients will say, as many of my patients have, "I would not place myself at that much risk of incontinence or impotence unless I was more confident that I could benefit from surgery. There's no point in doing the PSA." The average man on the street believes that a PSA test is as innocuous as any other blood test. In fact, screening places the patient on a slippery slope, in which test results tend to propel the process of evaluation and treatment as physician and patient become caught up in the need to know. The patient should know about the nature of that slippery slope before venturing out on it.

A number of expert panels have considered the question of prostate cancer screening. The US Preventive Services Task Force, the American College of Physicians, and the Canadian Task Force on the Periodic Health Examination have exhaustively reviewed the evidence and made recommendations. The US and Canadian Task Forces recommended against routine screening. The governments of England, Sweden, Australia, and the Netherlands have reached the same conclusion. As noted earlier, the American College of Physicians also recommended against routine screening but stated that patients need full information in order to make an individualized choice that takes into account their risks and preferences. The American Cancer Society recently changed its recommendation for routine annual screening staffing at age 50 years. The current statement says that "screening should be offered starting at age 50" and that patient should be fully informed before deciding. The American Urological Association recommends routine screening starting at age 50 years.

This Congress has enacted legislation authorizing the Medicare program to cover prostate cancer screening. It is now up to the medical profession to use this enhanced coverage wisely, so that it benefits our patients. We will need all the help that we can get. The Health Care Financing Administration can help by framing their regulations for prostate cancer screening so that they acknowledge the uncertain balance of harms and benefits and strongly caution physicians to avoid routine screening. The Congress can help in several ways. Physicians enjoy talking with patients about difficult decisions, but the current health care environment does not reward the time required for counseling. We need to rethink the evaluation and management codes so that they encourage counseling about screening. We need support for research on what constitutes effective counseling and how it affects patients' decisions about prostate cancer screening. Finally, we should evaluate health care organizations on their success in informing patients about screening procedures, such as prostate cancer, in which the balance of harms and benefits is uncertain and may vary from patient to patient.