Unlike diagnosis efforts prompted by symptoms and medical signs, cancer screening involves efforts to detect cancer after it has formed, but before any noticeable symptoms appear.^[67] This may involve physical examination, blood or urine tests, or medical imaging.^[67]

Cancer screening is not currently^[when?] possible for some types of cancers, and even when tests are available, they are not recommended to everyone. Universal screening or mass screening involves screening everyone.^[68] Selective screening identifies people who are known to be at higher risk of developing cancer, such as people with a family history of cancer.^[68]

Several factors are considered to determine whether the benefits of screening outweigh the risks and the costs of screening.^[67] These factors include:

• Possible harms from the screening test: Some types of screening tests, such as X-ray images, expose the body to potentially harmful ionizing radiation. There is a small chance that the radiation in the test could cause a new cancer in a healthy person. Screening mammography, used to detect breast cancer, is not recommended to men or to young women because they are more likely to be harmed by the test than to benefit from it. Other tests, such as a skin check for skin cancer, have no significant risk of harm to the patient. A test that has high potential harms is only recommended when the benefits are also high.
• The likelihood of the test correctly identifying cancer: If the test is not sensitive, then it may miss cancers. If the test is not specific, then it may wrongly indicate cancer in a healthy person. All cancer screening tests produce both false positives and false negatives, and most produce more false positives. Experts consider the rate of errors when making recommendations about which test, if any, to use. A test may work better in some populations than others. The positive predictive value is a calculation of the likelihood that a positive test result actually represents cancer in a given individual, based on the results of people with similar risk factors.
• The likelihood of cancer being present: Screening is not normally useful for rare cancers. It is rarely done for young people, since cancer is largely a disease found in people over the age of 50. Countries often focus their screening recommendations on the major forms of treatable cancer found in their population. For example, the United States recommends universal screening for colon cancer, which is common in the US, but not for stomach cancer, which is less common; by contrast, Japan recommends screening for stomach cancer, but not colon cancer, which is rarer in Japan. Screening recommendations depend on the individual's risk, with high-risk people receiving earlier and more frequent screening than low-risk people.
• Possible harms from follow-up procedures: If the screening test is positive, further diagnostic testing is normally done, such as a biopsy of the tissue. If the test produces many false positives, then many people will undergo needless medical procedures, some of which may be dangerous.
• Whether suitable treatment is available and appropriate: Screening is discouraged if no effective treatment is available.^[68] When effective and suitable treatment is not available, then diagnosis of a fatal disease produces significant mental and emotional harms. For example, routine screening for cancer is typically not appropriate in a very frail elderly person, because the treatment for any cancer that is detected might kill the patient.
• Whether early detection improves treatment outcomes: Even when treatment is available, sometimes early detection does not improve the outcome. If the treatment result is the same as if the screening had not been done, then the only screening program does is increase the length of time the person lived with the knowledge that he had cancer. This phenomenon is called lead-time bias. A useful screening program reduces the number of years of potential life lost (longer lives) and disability-adjusted life years lost (longer healthy lives).
• Whether the cancer will ever need treatment: Diagnosis of a cancer in a person who will never be harmed by the cancer is called overdiagnosis. Overdiagnosis is most common among older people with slow-growing cancers. Concerns about overdiagnosis are common for breast and prostate cancer.
• Whether the test is acceptable to the patients:If a screening test is too burdensome, such as requiring too much time, too much pain, or culturally unacceptable behaviors, then people will refuse to participate.^[68]
• Cost of the test: Some expert bodies, such as the U.S. Preventive Services Task Force, completely ignore the question of money. Most, however, include a cost-effectiveness analysis that, all else being equal, favors less expensive tests over more expensive tests, and attempt to balance the cost of the screening program against the benefits of using those funds for other health programs. These analyses usually include the total cost of the screening program to the healthcare system, such as ordering the test, performing the test, reporting the results, and biopsies for suspicious results, but not usually the costs to the individual, such as for time taken away from employment.