There are potential harms(harms greater than benefits) associated with interventions based on testing & imaging because of false positive and false negative results which are not uncommon in medical practice

Pitfalls in Medical Test Interpretation

The COVID-19 pandemic in 2020 was a once-in-a-century global pandemic with over 2 million cases in the USA and over 1 million deaths in the USA. Tedros Adhanom, Director-General of the World Health Organization, and US political and health leaders talked a lot about testing and contact tracing of cases to try and contain the spread of the pandemic. Testing, testing, testing was indeed the repetitive mantra by experts, politicians, and health officials. As a result of widespread publicity, there was a lot of discussion in the news media and medical journals. Whereas epidemiologists and scientists understood the nuances in the interpretation of tests, there was a great deal of confusion and lack of understanding amongst the public, journalists, and politicians.  Even physicians and healthcare providers fared badly and poorly when tested for their understanding of test interpretations. Experts felt that communications about laboratory testing including nuances in test interpretations were necessary to keep everyone informed in the fight against the next epidemic.

Understanding the variety of reasons for inaccurate test results is crucial, as it empowers us to make informed decisions. Some reasons are easily understandable, such as the quality of the nose swabs—they must be deep enough up the nose for a good specimen. However, there are also scientific reasons for the existence of false positive and false negative readings, which we must comprehend to navigate the complexities of medical testing. 
Scientists and epidemiologists talk extensively about the sensitivity and specificity of a test. The sensitivity of a test identifies people who have the disease (or condition) in a given population. So if a test is positive and has 100% sensitivity (e.g., specimen from young women who have missed a period and have early pregnancy symptoms), the actual or true positive rate ( confirmation of pregnancy ) is 100%, and the false positive rate is zero. The specificity of a test identifies people who do not have the disease( or condition). So if the test is negative and has 100% specificity like a negative pregnancy test (e.g., specimen from men ), the true negative is 100, and the false negative is zero. In real life situations tests with 100% sensitivity and 100% specificity are nonexistent.

As mentioned in the above examples, there is never a black or white in natural world medicine. We always talk about the probability of having a disease or the probability of not having a disease. When interpreting a test, understanding the base rate, which reflects how common or prevalent a disease is in a given population, is crucial. This knowledge prepares us to calculate the pre-test probability of disease, a concept that is the same as prevalence or base rate of a disease.
Epidemiologists have adopted the concepts of sensitivity and specificity of tests based on Bayes’ theorem, which can be confusing for most practitioners. However, it is more practical for physicians, as well as patients, to think in terms of false positives and false negatives, making these concepts more accessible and less intimidating.

An article in JAMA on April 5, 2021, titled “Accuracy of Practitioner Estimates of Probability of Diagnosis Before and After Testing,” posed this question to primary care providers.

The prevalence of the disease is 1 in 1000. A test’s sensitivity is 100%, and its specificity is 95%.

Q1. What is the chance that a person found to have a positive result has the disease?
The median answer was 95%, a significant deviation from the correct answer, which is 2%.!! This correct answer is crucial in understanding the probability of diagnosis based on the above information.
Let’s break it down: With 100% sensitivity, the True Positive (TP) is 1. (Remember, the prevalence of the disease is 1 in 1000 persons.) With a specificity of 95%, the False Positive rate is 5%, so 50 people out of 1000 people will test falsely positive. It’s simple math that is easy to grasp. 
One true positive divided by 50 false positive tests is 2%. This starkly contrasts with the median estimate of 95%. The vast discrepancy in the estimate of a probable diagnosis of a disease is of great concern.

To emphazise the principles involved, if a disease prevalence is low in a given population or community(in other words very few people are afflicted with a disease in a given community), a positive test is much more likely to be a false positive than a true positive, as illustrated above.
Below are two real-life examples of misconceptions:
The chance that a positive PSA test for prostate cancer represents a true positive, for example, is only about thirty percent. (There are calculators available online to estimate the probability of a true diagnosis when the calculator is filled with false positive and false negative percentages). To put it a simpler way a patient has a 1 in 3 chance of having prostate cancer. Numerous unnecessary prostate biopsies have been done based on faulty interpretation of PSA results.
Similarly, a study in which the Ca-125 biological blood marker was measured to detect ovarian cancer early in average-risk female patients found that a positive result for the marker only indicated a 1 in 4 chance of having the disease and a 3 in 4 chance of not having the disease.
It’s important to note that the USPSTF ( US Preventive Services Task Force) and many other organizations do not recommend PSA screening or CA-125 screening in average-risk patients. However, many patients are eager to be “proactive” in catching early cancer and request screening tests. This underscores the need for patient interaction to ensure they understand the potential harms associated with testing.

Footnote: I wish to clarify that oncologists often do a CA-125 test to follow patients who have been treated with ovarian cancer, and that is valid appropriate testing to give a patient crucial prognostic information. This testing has a different purpose than the testing done to screen a patient for early ovarian cancer or early prostate cancer, as mentioned in the above paragraph.