The price of admission to the final stages of life is the increase in disease testing. Those near 50 dread taking a colonoscopy, which is considered costly, painful, and invasive. Many of my friends approaching 50 have continued to skip their appointments for years afterwards. Given my interest in statistical tests and testing convenience, I recently took the 2nd Gen FIT test, and this article is about my experience with this type of testing. All so that you can make informed decisions in your own life.
making the decision
During my annual physical with my doctor, I noticed along with my doctor that I didn’t have any family history of related medical problems. I am also very healthy and physically fit, documenting my swimming on social media regularly. Could I responsibly choose with a less precise test, and what does a less precise test actually entail?
Armed with research, I approached my doctor, who conceded in my case it would be ok to take an at-home test. But this conclusion was driven based on my conversation, otherwise the boilerplate recommendation would be to go to the hospital for a regular colonoscopy. With some back and forth, we settled on the take-home test, but even then was encouraged to purchase it from this office at a mark-up and also a type that would have to return for their lab examination of results. That’s a hybrid result, which I declined. I simply purchased my product direct from the manufacturer, and it was comfortably delivered to my home. About $30 total.
The test was easy, and as the results show above I didn’t have worries afterwards. But what does this result even mean? And this is where one needs to know the quality of a negative result like mine.
general statistics
The product has an advertised 98% sensitivity and 95% specificity, and those two terms we explore here as they are somewhat associated with terms like false positive and false negative: terms that may be slightly more familiar with people. Kaiser independently researched this product and found the sensitivity in their experiments were more moderate and closer to 80% for cancer, which reinforces what we learned even in coronavirus era, which is to provide an error around these estimates which at times can themselves be quite large depending on who is sampled and what definitions or caveats are applied to the sample. What sample are you: a question always worth raising.
In a screening test you want a product that has a low false negative, or a test that is less likely to falsely conclude you are fine. Sometimes one is willing to pay the price with a higher false positive as a result, or a test that has a higher likelihood therefore of triggering a false alarm for a disease you in fact don’t have. But you are not going to die from a false alarm, which is better than being falsely unaware you have the disease.
colonoscopy long-term trade-offs
For my analysis I assumed the product’s sensitivity of say 85%-90%, or a blend between what’s marketed and the Kaiser study. This implies that if 85% of those who have the disease would be correctly identified for cancer. However for other polyps or other early-stage cancers the sensitivity plummets to 30%, all while the regular in-hospital colonoscopy is considered the gold standard at 90%+ for both.
And the 95% specificity implies 95% of people without cancer who are fine would be correctly identified. So one can’t simply aim for a low false negative and a high sensitivity, if it comes at an unproductive cost of a high false positive and a low specificity.
Now the lower sensitivity [~85% and 30% for other polyps] versus specificity [~95%] for the 2nd Gen FIT test was one of my concerns at these levels, but recall this is for all people taking the test. For someone young with no historical issues such as myself, and willing to take the at-home test more regularly, I was willing to make this trade-off on testing set-up. As the key is: the sensitivity estimate would be more error prone for me.
math summary
Note a regular colonoscopy at the hospital has a diagnostic sensitivity for cancer and advanced adenomas that is far higher, but not by that much to justify the invasiveness to me. And the results apparently come with a 10-year shelf life, but similar to other drug research effectiveness it is questionable in terms of retaining that accuracy over the entire horizon. The test I took was so benign I wouldn’t mind taking it multiple times a decade, for higher assurance against polyps and to avoid the discomfort of a “sore” hospital procedure.
And while the medical community pushes the "Gold Standard," they often ignore the "Compliance Standard": a test is 0% effective if the patient refuses to take it. By opting for a high-frequency, low-friction FIT test, I am essentially "smoothing the curve" of my risk profile. Instead of one high-stakes invasive procedure every decade, I am opting for a series of high-probability snapshots. This frequency helps mitigate the lower sensitivity for polyps by ensuring that if something is missed this year, it is caught in the next, long before it reaches a critical stage.
Other benefits to me in doing this, including my own pre-research before going to the doctor’s office were many. And perhaps my experience will help in your own decisions. I avoided the cost, invasiveness, anesthesia risk, as well as prep burden of a full colonoscopy. And I gained a convenient, rapid, low-friction screening. With instant at-home results. And I always retain the free optionality of a full colonoscopy afterwards, if my test turns positive or symptoms appear, instead of simply rushing towards that level of test from the start.
flattening the risk tail
Your own health decisions are sometimes a continuous and complex problem. But it’s worth exploring particularly for deadly diseases. This $30 test cut-off the convex copula discussed in my recent The Actuary print article by IFoA, and a longer run longevity and health outcome concern in your retirement.
No comments:
Post a Comment