Do you speak Statistics ?

There is an age old expression that “a little knowledge is a dangerous thing.” This is definitely true in the world of medicine. There are those who argue that studies, even those being done today, are seriously lacking in terms of the data they include. For example, now that there is at least an appreciation for the effect of the microbiome [the collection of bacteria that are supposed to be in our bowels], it almost seems as if every previous drug study has fundamental flaws.

How so? It seems to be that the microbiome determines, to a great extent, how anything we ingest [and maybe intake even by intravenous or other routes] affects our bodies.. There is a very recent study that has come out of Israel that strongly demonstrates a fundamental difference in an individual’s response to food. It seems that even something like ice cream can be processed very differently from one person to another, so much so that it may cause weight gain in one person yet leave a second person unaffected.

One of the major topics in medicine these days is individualization of care. Imagine two people with the same general class of cancer. They each receive the approved list of medications specifically intended to kill off the cancer [but not the patient]. Nevertheless, one person may be cured while the second perishes from the disease. This is a classic situation where doctors will quote statistics, i.e., the percentage chance of success versus failure. So a doctor may say to a patient that a specific treatment for a specific cancer has a 70% success rate. These statistics don’t refer to an individual, i.e., that the individual will be rid of 70% of the cancer and will then need to take other medications to eliminate the last 30%. What these numbers mean is that 70% of the patients will be cured of the disease while the other 30% will not. Basically, you can’t be half pregnant.

Unfortunately, even such basic statistics fly over the head of many patients. This has nothing to do with the level of intelligence of the patients. I have seen professors of various topics of study, have difficulty calculating the change they will get from a $20 bill, after purchasing a given item. Basic statistics are actually very straightforward and logical. But if you have never been exposed to them, you won’t know what they mean. Much of medicine depends on statistics to decide which treatments to use. This leads to a fundamental disconnect between many patients and their [not so patient] physicians.

There has been a great deal of discussion about breast cancer in the last couple of years. A number of papers have come out that have appeared to question the basic value of doing any screening mammograms. Mammograms are a special kind of x-ray that many women do each year and are intended to identify breast cancer. The concept is very straightforward — just like you would do a chest x-ray to look for findings consistent with a pneumonia, you do a breast x-ray to look for abnormalities consistent with cancer. Once a mammogram is deemed suspicious or worse, highly consistent with cancer, then a decision has to be made how to further evaluate the finding, in order to determine if treatment, and what treatment, is necessary.

Breast screening is considered life-critical by many. Breast screening is the process by which you do mammograms on all women fitting a higher risk category, even though there are no signs of disease, yet. The idea is to cast a wide net, to catch the relatively small number of women who will be found to have cancer. All of the otherwise healthy women, with negative mammograms, will repeat the process the following year or two years. The amount of time to wait between mammograms is based on certain protocols that have been developed by teams of top experts in the field.

If you do mammograms on every woman over a given age, every one to two years, the hope is that any developing cancers will be caught before they do any real damage. In this way, a woman could be freed of her growing cancer before it spreads, and can also be spared far more extreme treatments that are needed once the cancer has grown and spread beyond a certain degree.

When explaining the concept of screening in order to capture breast cancer while it is still small, most people understand this intuitively. But, intuition can be a problematic creature. There are things that are inherently understandable to most people, via the connections between all of the nerves in our brains. If dropping a 1 pound weight on your foot hurts, it’s intuitive that dropping a 5 pound weight on your foot will hurt more and could even do significant damage. And this way of thinking continues. If you drop a 500 pound weight on your foot, you can expect to need advanced medical care.

Using the same intuitive type of thinking, it makes absolutely perfect sense that if you identify cancer when it is small, then your success rates with treatment will be much higher. “Clearly”, a half centimeter breast cancer is less dangerous than a 2 cm breast cancer. This type of logic, by the way, applies almost everywhere in medicine. If you have a 2 cm cancer on a kidney, most people [and doctors] tend to think that this is far less dangerous than a 5 cm cancer on a kidney.

The problem is that intuition is not fortune-telling. Evolution has not given us the inherent ability to understand  how our own bodies function and malfunction. We don’t inherently/intuitively understand the difference between the different sized breast cancers I mentioned above. This lack of intuitive understanding is so dramatic that there is literally a backlash on the part of many doctors, and definitely the public, when any paper is published that argues that mammography does not improve the outcomes from managing breast cancer.

Let’s say this a different way. A growing number of papers are arguing that whether you identify a breast cancer when it is smaller versus when it is large enough to actually be felt on physical exam of the breast, makes no difference. It seems that present day treatments for breast cancer are just as successful on larger cancers than they are on smaller ones. Literally waiting until a cancer grows to such a point that one can feel it through the woman’s skin, does not lead to more suffering or death from breast cancer.

In fact, and this is the kicker, treating breast cancers when they are smaller and only identifiable by mammograms, ultrasounds and even MRIs, leads to many women receiving treatment for lumps that would otherwise not cause any harm. This is yet another case where the general public and even many doctors fail to appreciate the danger in treating diseases that do not need to be treated. If a woman who is otherwise healthy, but has a non-cancerous nodule in her breast, gets surgery, chemotherapy and/or radiation therapy, she may end up suffering severe complications from these unnecessary treatments. In some cases, patients can even die from the treatment.

The back-and-forth in the medical literature about whether it is worthwhile to do mammograms at all, is infuriating. As a physician, I can read the papers and understand their content, but still not be able to advise my own wife on the best way to move forward for herself. Thank G-d, she has never had a problem with breast cancer. But I cannot, in good faith, recommend that she have  yearly mammograms that may end up causing her more harm than good. At the same time, she has an excellent primary care physician who makes sure to do a full and proper breast examination. So it is not as if we are ignoring the issue. Rather, we are dancing between the raindrops trying to do enough to identify a potentially lethal disease, but not too much so as to cause suffering from unnecessary treatment.

I won’t extend this conversation by talking about specific genes that are strong indicators for dangerous breast cancer. The famous example of Angelina Jolie being so proactive so as to have treatment based on her genetics, was an incredibly brave thing to do. Ms. Jolie’s actions may very well have saved the lives of many women who have never been assessed for their own risk of breast cancer. Using genetics to classify people as higher versus lower risk is entering a new and very exciting period in history. Within a few years [5? 20? Pick a number] it will be possible to quickly and inexpensively evaluate an individual’s genetic code and determine what that person truly needs to have further evaluated. Combining the genetic codes of millions and even billions of people with clinical information about these people [i.e., what happened to these people from a medical perspective over the course of their lives] will fundamentally change the practice of medicine. Rather than treating each person based on studies that generated very course statistics [like the “70%” example I mentioned above], we will be able to tell a given person their unique and specific risk for a whole variety of diseases. This will totally change the way we interpret studies such as mammograms.

As best said in the movie “The Life of Brian”, “We are all individuals”. It seems that medicine and tools like statistics are finally embracing this concept. In not too much time, literally the whole world will benefit from new and far more specific and accurate diagnostic tools.

Thanks for listening.

About the Author
Dr. Nahum Kovalski received his bachelor's of science in computer science and his medical degree in Canada. He came to Israel in 1991 and married his wife of 22 years in 1992. He has 3 amazing children and has lived in Jerusalem since making Aliyah. Dr. Kovalski was with TEREM Emergency Medical Services for 21 years until June of 2014, and is now a private consultant on medicine and technology.