Bear Flag Republic Radio Weblog

< 5:58:16 PM        >

After reading the above, you would think that U.S. political leaders would be concerned about the massive loss of not just U.S. jobs, but high paying technology jobs. The very jobs that U.S. politicians have told our young people to train for - to train as engineers as scientists. And who's large tax payments help fund California's government. Until the tech sector crashed, and with it, the State's revenue picture. The U.S. government has made a decision that the U.S. will no longer hold a technological superiority in world commerce. Fortunately, we will be the leader in frivolous lawsuits, CEO manipulation of the accounting records, and corporate holding companies that manage workers in developing nations. Our competitive advantage is in being Dilbert's boss. [Edward Mitchell: Common Sense Technology]

comments < 8:36:07 AM        >

News reporting on medical research is always amusing. The latest is a report on a research project that concluded that placing automatic defibrilators in public places doubled the survivor rate of cardiac arrest patients. Doubling the survivor rates sounds cool until your realize that the likelihood of surviving went from 5% to 9.9% who survived. Everyone else died. But the conclusion is highly misleading for two important reasons:
1) the study trained 20,000 people who worked at various places, in CPR and AED usage. That is not typical of the real world.
2) How many patients died within one week of surviving the initial cardiac arrest? How many died within six months?

---

Another study in today's news says that properly treating pain amongst workers would save U.S. employers $60 billion per year. The study was based on interviews with workers who estimated their own lost productivity, which is a highly subjective estimate. Thus, a confidence interval on that $60 billion per year loss should be about one mile wide. Further, the study was funded by a company that, amongst many things it does, runs pharmacy services that sell pain medication. There was absolutely no conflict of interest in their choice to fund a study on the effects of pain in the work place, of course. It must be worth spending at least $59.999 billion of other people's money to drug more people into productivity improvements.

---

Both of the above studies are typical of medical research, which is to say, studies that have interesting findings, but are typically useful only for soliciting a grant for future studies. When I took some graduate courses in medical informatics, I learned of one study(!) that had reviewed all of the research articles published in two peer reviewed journals for five years. This review concluded that 90+% of the research studies were essentially useless due to poor experimental design or insufficient sample sizes to draw statistically significant conclusions. They suggested that expending so much energy on studies that could not draw useful conclusions was unethical - since it meant that less research funding was available for studies that would provide definitive answers, and may have subjected patients to inappropriate treatments. In the end, it seemed that the purpose of many studies is merely to find something interesting enough to justify a grant for further studies. Which is not the same as actually coming up with a useful answer that benefits the public.

Another thing I learned is that virtually no studies in health care consider costs versus benefits. Instead, they focus on one treatment modality versus another, with seldom any attention towards measuring value. And why should they? Health care for most people is paid for by someone else. We will always choose the "best" (which is a pretty fuzzy concept), and ignore the cost, since cost is something we are usually oblivious too when it comes to health care. [Edward Mitchell: Common Sense Technology]

comments < 8:34:15 AM        >

Circuit City is outsourcing Human Resources to India. Hey, its only fair that the human resources department itself get outsourced. Pretty funny. [Edward Mitchell: Common Sense Technology]
comments < 8:33:43 AM        >