Weighing Benefits and Harms Decision Making About Screening

Weighing the benefits and harms of screening programs is difficult. First, one must determine the presence and magnitude of the benefits and harms. Benefits seem intuitive, but closer inspection shows that they are easy to overestimate. Evidence about the accuracy or yield of screening tests by itself is inadequate, as is evidence about the effectiveness of treatment in people detected clinically. Etiologic data about trends in mortality over time are open to multiple interpretations. To avoid the strong biases involved, the ideal study design is a randomized controlled trial (RCT) of screening. Even this design, however, is open to criticism, as has been shown by the controversy over the breast cancer screening trials.16,17 In the end, the determination of the benefits of screening depends on examining evidence from many sources with different designs, and then considering the relevance of the studies to the community setting.18 One should consider not only whether there are benefits, but how many people benefit and by how much. Judgment is involved in this complex process.

The same process is involved in determining the harms of screening. This part of the equation is often forgotten: a screening program is justified if, and only if, the benefits outweigh the harms of the program. Even RCTs of screening often do not report on the adverse effects caused by screening.

After determining the magnitude of benefits and harms, decision makers need to consider whether one outweighs the other. A problem is that the benefits and harms are usually in different metrics. Benefits should be stated in terms of the estimated number of lives extended in 1,000 people screened over a given time period. Some have claimed reassurance from negative screening tests as a benefit, but careful examination shows that such reassurance is based on questionable assumptions. The actual reduction in the probability of cancer after a negative cancer screening test is very small. Harms should include the number of people in 1,000 screened with false-positive tests, people who are referred for workups and may suffer psychologic distress. Ideally, one should also estimate the number of people who were overtreated and the consequences of overtreatment.

Weighing the (usually) small number of lives extended against the larger number of people with various types of harms requires a value judgment. What value is placed on the experience of the people suffering harms and what value is placed on the people whose lives are extended? Currently, our culture seems to have decided that even a small number of lives extended by cancer screening outweighs a larger number of people having problems with false positives and overtreat-ment. It is not clear whether this concept will change as the public becomes more aware of the magnitude of the benefits and harms from cancer screening.

For many cancer screening decisions, some have suggested that the people undergoing the screening should be informed of the potential harms as well as the potential benefits and should be involved in the decision.19 This approach is called shared decision making (SDM). Research is currently exploring ways of making SDM more feasible and effective in the clinical setting, using decision aids and other decision support resources.

The benefits and harms of cancer screening usually vary by age. The incidence and mortality of most cancers usually associated with screening increase with age. If screening leads to a constant relative reduction in mortality risk, then the absolute reduction in risk (i.e., mortality benefit) increases with age. Thus, younger people with a low probability of having cancer may receive few benefits and expose themselves to important harms by having screening. For example, women in their forties probably receive some benefit from mammography screening for breast cancer, but the magnitude of benefit is small. These women do, however, have a higher probability of having a false-positive mammogram. Given this information, some women will choose to have mam-mography in their forties and some will not. The age at which the benefits of screening overcome the harms depends on personal values. In general, however, screening is not offered to people at a younger age who have a very low probability of having cancer.

In considering an upper age limit for screening, it is important to remember that any life extension from screening does not occur immediately after screening but rather some years in the future. Thus, to realize the benefit from screening one must live a certain period into the future. As people age, however, the risk of dying of a cause other than cancer increases. Some older people, then, do not live long enough to benefit from screening. Clearly, people with a limited life expectancy have little to gain from screening.

The harms of screening may or may not vary by age. In some cases (e.g., mammography), the screening test may yield fewer false positives in older people, thus decreasing the harms of this finding. In other cases (e.g., PSA for prostate cancer), the screening test may yield more false positives and thus potentially increase the probability of harms. Harms should be carefully considered in every screening decision; the weight they are given by the patient may vary by age. Interventions to reduce psychologic harms by educating people about false positives before screening may be useful but need more research.

An important decision for people who have decided to be screened is the frequency of repeat screening. It is unusual to have RCT evidence about the relative benefits and harms of various screening intervals. More commonly, we reason about this issue with indirect evidence, including our understanding of the natural history of the cancer and measurements of cancer incidence in people who have waited different times to be rescreened.

In general, cancer detection is greatest with the first screening round, the so-called prevalence screen, because there are more asymptomatic cancers to be detected initially than on later screening rounds. Cancers detected in a previous round of screening have been removed from the pool of remaining asymptomatic cancers. Thus, a short screening interval will likely find fewer cancers than a longer interval. The longer the interval, the closer cancer detection will revert to the initial round.

The balance of benefits and harms from various screening intervals is more complex. A shorter screening interval usually means increased sensitivity, but specificity may be reduced. In deciding the most appropriate screening interval, one must consider the trade-offs between finding all appropriate cancers (i.e., increased sensitivity) and increasing the rate of harms related to false positives.

The value of cancer screening after previously negative screening tests is uncertain and needs further study. It is possible that various results from previous screening tests (e.g., men with very low PSA values or women with benign types of parenchymal findings on a mammogram) may be markers of people at decreased risk of developing cancer. These markers may help us define a group of people who do not need further screening, thus allowing us to target screening to people who have the greatest possibility of benefit. If by defining a low-risk population (rather than a high-risk population) we could reduce the number of people requiring screening (thus reducing costs and potential false positives and overtreatment), the balance between the benefits and harms of screening could be improved.

How To Add Ten Years To Your Life

How To Add Ten Years To Your Life

When over eighty years of age, the poet Bryant said that he had added more than ten years to his life by taking a simple exercise while dressing in the morning. Those who knew Bryant and the facts of his life never doubted the truth of this statement.

Get My Free Ebook

Post a comment