More on p-values
Government heavily regulates healthcare industry. It is interesting that the research that the healthcare is based on is hardly regulated at all. In fact, peer reviewed medical journals only have [...]
P-values
Medical literature is full of p-values. It is amazing that this test is used on small data sets and people try to draw conclusions about significance of their results based on p-values they get from n=10 to 30, or so. Statistics 101 taught us that power of the study and consequently appropriateness of the statistical test results largely depends on the number of subjects. Every time, I sit down with a statistician, the “n” that we come up with to have reasonable power of the study is on the order of 1oo or more.
Another issue is that researchers constantly say things like n=75, and than calculate p-values for smaller subgroups in their study. What good is a p-value on a sample of 9? In my opinion, if you think that n=9 is enough of a power for meaningful statitistics, you are dealing with a question that is crystal clear already, or you have no clue what the differences between the groups are and you are just guessing what the power should be.
Large general medical journals are pretty good in concentrating on large studies with “n” sometimes in the thousands. As soon as we look at subspecialty journals, the numbers quickly drop into teens even for big name journals in the subspecialties.
How do we solve this? Collaboration between centers and facilities would be nice. Current research based grants for each individual research does allow for meaningful collaboration in many research projects. But, that’s a topic for another blog…
mriDoc
Click here to go back to mriDoc.com
Health Care Reforms
So, what is it exactly that we are trying to reform? One can listen to the news to find out answers to that question. The number of uninsured is too great. The insurance companies deny coverage based on pre-existent conditions. The health care cost is too high. Ok, so those are the major problems. What about solutions?
Before we talk about possible solutions, lets first analyze how we got here.
The number of uninsured is too great:
Is lack of insurance such a bad thing? What happened before people thought of health insurance and health maintenance organizations? If someone got sick, they went to a doctor. Doctor gave them medicine or admitted them to infirmary. The patient or his family paid doctor’s bill and for medications with money or with a sack of potatoes. Not bad. Everyone seems happy. Then, twentieth century rolls around. Now we have drugs that take up to twenty years to develop, surgeries done by robots, and regulations from federal government, compliance with which is not free for providers. All of a sudden, if one gets fever and wants a medicine for it, the true cost of writing a prescrition and cost of treatment itself are astronomical. No one person can afford getting slightly sick more than couple of times. So, lets “spread the risk.” Insurance companies come about. The shareholder and staff need to be paid and computers need to be bought. Health maintenance organizations come about – not sure if overall there is cost saving from those or more cost after patients give up and find alternative arrangements. Government keeps adding regulations – more cost to everyone. Insurance get way too costly. Many people can not afford it.
Insurance companies denying coverage based on pre-existent conditions:
Ok, lets say that everyone is insured and there is perfect and fair spread of risk. Is insurance company that covers inner city population at the same risk of loss as the insurance company that covers suburbs or congressman. Conditions are different, so the risk is different. No perfect world out there. So, insurance companies minimize their risk and only cover someone who is not yet very sick or impose waiting periods for coverage.
The cost is too high:
That’s a big one. If someone things of panacea for something, that person has every right to develop it. The development costs are getting higher like costs for anything else. In general, the American public wants solutions right at the moment of a problem. So, it is usually unacceptaable to wait for the routine person in the morning to show up. Lets get on-call specialists and techs to come in the middle of the night. Takes resources to attract the specialist and the tech into the hospital at 2 am. Then, there is the issue of practicing defensive medicine. And the list goes on and on.