In my ongoing attempt to better understand the inner workings of the US healthcare industry, I've read yet another book: Robert Aronowitz's Risky Medicine (2015). Within, the author argues that medical practice has undergone a consistent shift away from treating disease and towards 'managing risk' via an ever expanding array of preventative screenings and medications. To summarize the message: Risk reducing drugs are incredibly appealing to pharma companies as their markets are vast, advocacy groups often blindly support screenings irrespective of their efficacy because they help manage fear, and regulation generally has a much lower bar for risk-prevention than actual disease treatment.
It can take a long time for epidemiologists tease out these factors, and the true value of preventative services often take decades to evaluate. By then, they can become entrenched within the culture of advocacy groups, who will often trot out versions of the above, naïve statistics in order to argue for continued screening. Furthermore, the above example (and many advocacy groups ) focus entirely on sensitivity of screenings to the exclusion of specificity: no test is completely without false-positives, which can offset the value of the test if the condition that it's trying to detect is sufficiently rare (see positive/negative predictive values).
Another major argument made by the book is that the standards for approving risk-reducing drugs are too low, thus ever increasing the pool of preventive treatments. Disease risk factors identified during the course of clinical research are rarely as predictive at the whole population-level as they are in study groups. Furthermore, statistically significant association says nothing about the magnitude of the effect of reducing risk factors. Nevertheless, when pharma companies develop a drug that reduces a risk factor for a disease (say blood pressure), they need only show safe efficacy in targeting the risk factor, and not that it actually lowers incidence of the ultimate disease. Therefore, as above, drugs can become entrenched in medical practice long before we realize that their efficacy is marginal or non-existent.
While thought-provoking concepts, Risky Medicine never gave me a sense of the actual magnitudes of these issues. Most of the discussion is theoretical, calling for increased skepticism in the face of new preventative strategies. While some examples of unnecessary focus on risk are discussed, the case-studies given chapter-length treatment are quite-complex, spanning a large range of issues that muddy the main message and make it difficult to form an opinion without more knowledge. This likely explains why ~45% of the book consists of detailed footnotes, which aren't citations so much as additional background necessary to understand the circumstances being discussed .
Controversies such as the efficacy of the prostate-specific antigen test or hormone-replacement therapy in post-menopausal women indicate that both the definition of 'risk' as well as how we determine what risk factors are worth addressing are worthy of continued consideration. Furthermore, as the Aronowitz indicates, we haven't comprehensively addressed the quality-of-life effect of moving millions of people out of what was classically regarded as a state of 'good health' to an endless maze of varying levels of risk. Hopefully, someday I'll read a book that discusses these topics more cogently, with explanations of actual research.
 In reality, values are rarely this large and obvious, and it's often more complicated. For example, according to Aronowitz, the tendency in the medical community has been to broaden the definition of diseases over time, generally increasing the number of diagnoses regardless of treatment efficacy and therefore further reducing the apparent mortality.
 According to the author, patients rarely complain about false-positive diagnoses, even when it's clear that such a case occurred. For example, if a test falsely suggests that a patient has cancer and leads to painful biopsies that rule out disease, it's more likely that the patient will feel elated for having 'dodged a bullet' than ask why the original test was positive.
 Only at the very end of the book is it indicated that its various chapters are collected from previously published essays, chapters, and papers. Unfortunately, it doesn't make for a very clear presentation of the argument and leads to a lot of repeated examples and redundancy.