Steve Jobs was very lucky. They detected his pancreatic cancer early. It happened incidentally. He was being screened for something else and this came to light. Pancreatic cancer is a very aggressive form of cancer and it is very important to detect it early. This almost never happens as by the time symptoms appear the cancer has advanced quite a bit. So what did Steve Jobs do? Sadly he decided not to treat it and relied on home remedies.
Just as you never bring a knife to a gunfight, never bring carrot juice to a fight against pancreatic cancer as your sole weapon. Later Steve Jobs was blunt in his assessment of his actions, “I was stupid,” he said.
There is tremendous distrust of medical science. Many second-guess their doctors and go against doctor’s recommendation based on what they find on Google. Sometimes we rely on obscure web sites and blogs rather than trust doctors. This can be very risky as Steve Jobs case illustrates.
But it is also true that there is a risk of being over treated if one relies solely on the advice of some doctors. Doctors find themselves in a difficult situation. They are at the risk of being sued so their entire practice is based on defensive medicine, which sometimes causes them to recommend useless tests and procedures. Doctors are now more inclined to push decisions to their patients, explaining pros and cons of the various choices. Rightly or wrongly lay people find themselves to have research and read medical literature to make sense of their options. This can be confusing and tricky. You may find two different studies reach exactly opposing conclusions. This article seeks to provide some guidance on how to navigate through the maze:
1. Who paid?
It’s important to understand who paid for the study. If a pharmaceutical company or a medical device company funded the study then you want to be cautious in accepting its conclusions. You want to look for government funded studies such as NIH or studies done by nonprofits. Even here you may want to research the background of the lead researchers. Are they having links to companies that have a vested interest in the conclusion?
2. Size matters
The size of the study matters. A study done on 50 people is more prone to errors than one done on 5000. This is because in a smaller study, the way one or two people turn out, can have a dramatic impact on the conclusions. In a larger study such variations are smoothened out. Sometimes all you have is study done on animals or cell cultures. Be careful not to conclude too much from such studies.
3. Direct or indirect conclusions?
Be cautious of indirect conclusions. Is the benefit of the procedure or pill expressed indirectly? So if pill X reduces blood pressure or cholesterol that is an indirect benefit. What is more important is that did it reduce “all cause mortality”? A pill may reduce cholesterol but may not have any benefit in terms of overall mortality as it may be damaging the liver and causing other issues. If the benefit you are looking for is better eyesight then that is the conclusion you should be looking for in the study. Not that it reduced the pressure within the eye, which is an indirect benefit.
4. For preventive stuff be wary of relative risks for specific cause
If you suffer from breast cancer and are evaluating the benefit of a procedure that reduces risk of dying, that’s not the same as evaluating a procedure like a mammogram when you are healthy.
If you are healthy and young your chances of dying from all causes over the next year is pretty slim. For US women in the age 45 and 64 the number is less than 0.5%. (For age 15 to 44 the number is even smaller.) The probability of dying from a specific cause like cancer is even smaller. And if you drill down to a specific cancer the number becomes very, very tiny. So even for a large study of healthy people, involving say 5000 people and being carried over a period of say 5 years the number of people dying from a specific cause, like say breast cancer, is very small. This is why preventive studies of healthy people are different from studies of treatment options of people who are already ill and have to be reviewed carefully.
Most studies are divided into two equal groups and one group is treated as a “control” group. The control group is not given the medicine or treatment and the result is compared with the treated group. So the question asked in a typical study will be like: Did less number of women die from breast cancer in the group that was given regular mammograms? Usually the numbers are pretty small and the difference may boil down to a few people in a study of thousands.
Further to make the results of the study appear to be significant the number is expressed in relative terms. This may be expressed as something like “35% reduction due to death from breast cancer by doing mammogram for women aged 50 or older”. (In absolute terms it boils down to 1 less death in 10 years from Breast Cancer out of 2000 women who undergo mammography. More than 200 will go through psychological distress and false scares and most of these will be overtreated. Notice how when absolute numbers are used instead of percentages the benefit seems pretty miniscule compared to the risks.)
When you look at relative numbers like 35% reduction in death from breast cancer it may sound impressive but be cautious of such claims. One reason is that it does not talk about death from all causes, just reduction of death from a specific cause like breast cancer. What you and I expect is that if we are going to do some preventive procedure our life expectancy is going to increase. Right? But that is not being promised. All you are being told is that your chance of dying from a very specific cause is being reduced. What if there are other risks associated with the procedures and increase risks from dying due to other causes? Sometimes it is not possible to accurately ascertain the “cause of death” and researchers may bias the results by creatively interpreting the ambiguity. As we have seen even a small bias can result in a big swing in the conclusions of the study as the difference between two groups in a preventive study is usually very small.
If you are out researching please continue to dig deeper till you find the answer in terms of “all cause mortality”. This is what we found for Mammograms in a NY times article by Dr. David Newman:
“It may be surprising, therefore, to learn that numerous trials of mammography have indeed randomly assigned nearly 600,000 women to undergo either regular mammography screening or no screening. The results of more than a decade of follow-up on such studies, published more than 10 years ago, show that women in the mammogram group were just as likely to die as women in the no-mammogram group. The women having mammograms were, however, more likely to be treated for cancer and have surgeries like a mastectomy.”
After a lot of digging we finally find something that talks of dying from all causes, not just breast cancer. And when this happens the purported benefits of mammography simply vanish.
We have used mammography as an illustration. The point being that if you dig deeper and ask the right questions the answers will eventually surface. Even experts get confused between “all cause mortality” and mortality from specific cause and medical literature reflects this confusion, so be cautious and continue to dig till you find your answer.
5. Check Cochrane and NY Times
Where to do your research? We recommend Cochrane and NY Times.
Cochrane.org is a group of more than 31,000 dedicated people, mostly doctors and scientists, from over 120 countries who work together to create “study of studies”. They review all the scientific publications on a given subject and publish it as a “Cochrane Review” after weeding out studies that do not measure up to standards. To date there are more than 5000 such reviews that are available at Cochrane.org. Another good source is the NY times website. Use the search feature.
In both Cochrane and NY times web site you will have to dig deeper till you find the answer you are looking for. Do not settle for the first link you find. Do not get carried away by anecdotal stuff published on blogs. The articles must be based on scientific evidence published in peer reviewed journals.
6. Don’t let the doctor off the hook
Researching on your own can make your head spin and not something everybody can do. Relying on doctors advice is always a great idea. So don’t let your doctor off the hook. Here are some pointers:
- Get a good doctor. Do this when you are well. Take the time to “shop around” till you find a doctor who relies on sound medical evidence and is skeptical of “oversold” studies.
- Take notes. Do not hesitate to stop the doctor to write down stuff. Ask for spellings and make sure you understand all the issues at hand and are able to write down both the pros and cons that the doctor is telling you. This will give you material to research on. Once you do your research, write down all your questions and doubts and consult your doctor till you have clarity. Asking for second opinion is a great idea.
- Ask the most important question: What would you do if you were in my shoes? Don’t let the doctor off the hook. Make sure you understand your doctor’s recommendations not just the choices she presented to you.
Advance made by medicine are spectacular and can both help improve the quality of life and the life expectancy. However there is a lot of money that is being injected into medicine that is creating research of questionable quality. It is important not to use this as a pretext for “throwing the baby out with the bath water”. The article provides you with ideas and resources to make the most of medicine and medical literature.
You may also like: The Biology Of Yoga
Credits:This has been written by Raj Shah and edited by Ketna Shah.