Consider a Health Day story about an experimental breathalyzer for stomach cancer identification, which was said to be “85 percent accurate.”
To several people, a detection rate of 85 percent probably sounds great in complete medical diagnosis — as well as the test story seemed to foster a sense of accuracy. These check individuals for stomach cancer, a test could result in “earlier care and treatment, and stronger survival.”
But perhaps the experts at HealthNewsReview.org were also not almost as optimistic.
In their analysis of the press release on which the story was based, they pointed out that the exam, if widely supported, could lead of hundreds of fake-positive outcomes for each person identified with stomach cancer. Either in the press release and the story that regurgitated, those fake-positive results were not noted.
Sensitivity & Specificity
What else can be done?
Two important notions in medical testing might have enhanced both of the news releases and also the news story: sensitivity and sensitivity.
They are also the test world’s ying and yang and demonstrate crucial information on what a test wants and can inform us. Both are necessary to fully understand the strengths of a test including its weaknesses.
Sensitivity – Measure how well a test correctly produces a positive outcome for people with the disease being examined (also recognized as the “real positive” rate). An extremely sensitive test can flag nearly anyone with the disease and does not produce many false-negative outcomes. (Example: the 90 percent sensitivity test can correctly yield a positive outcome for 90 percent of people with diabetes but will yield a negative outcome— a false-negative — for 10 percent of those who have the disease or who should have come back positive.)
Specificity – Measures its ability of a test to properly generate a negative result to people who do not have the disease being checked (also referred to it as the “real negative” rate). An elevated-specificity test can correctly govern out nearly anyone who does not have the illness and will not produce many false-positive outcomes. (Instance: a90%-specific test can correctly come back a negative result to 90% of individuals who do not have the illness but will transfer a positive outcome— a fake-positive test— for 10% of individuals who do not have the illness or who should have come back negative.)
In this graphic shows how well these words apply with one of the most frequently used tests: the pregnancy sample.
Recognizing that sensitivity but also specificity occurs in a state of equilibrium is essential. Tactile sensitivity–its ability to accurately identify individuals with the illness–usually occurs at the cost of reduced particularity (implying more false negatives). Similarly, high specificity— whenever a test is doing a good job of governing out individuals without the disease — generally means that the exam has lower susceptibility (more fake negatives).
A good example of how these tradeoffs play out during practice is another daily example of passport control. Scanners at such a security checkpoint could also alarm for innocuous items such as epaulets, watches, and jewelry to make sure that genuinely dangerous products such as weapons can never be brought on deck an aircraft. Read here for more info!