Let us now pick up our discussion regarding progressive disease states in a Markovian framework.
In part I, we discussed a number of issues many of which are fairly intuitive:
- Preventing a disease is the best-case scenario (state 1)
- Intervening early in a disease process is much better than intervening late
- As we jump left to right in a series of disease states, each subsequent jump costs more than the previous jump.
Also, it was emphasized how most of us, absent sufficient inducement or education, frequently put critical health screenings on the back burner for any number of reasons. Let me discuss one representative case.
Over my career, I have had the honor of treating many medical colleagues. A few years ago, an internist (a physician who, like myself, treats adults) presented as a new patient for an initial physical exam. He was in his late 60s and had not had any colon cancer screening. He had many years to protect so I suggested undergoing screening but he declined for no logical reason (medical Russian roulette, I have noticed, is unintuitively common in the medically sophisticated population, even physicians). A few years later, I was called by an ER physician because the patient had a so-called “acute abdomen”. The general surgeon who operated on him found that he had a colonic adenocarcinoma which had perforated the colon wall. Please note that the discomfort from a perforated colon is a character-building experience. He was extremely sick for a number of days. In spite of the moderately advanced colon cancer, the patient made his 5-year anniversary with no evidence of residual cancer. He may be (hopefully) a long-term survivor.
Using the Markovian framework, perforating colon cancer should be considered the last state in that disease process (please note that this terminology is distinct from the pathologic staging of disease). At the outset of my career, colon cancer screening was not reimbursed for asymptomatic patients and thus was virtually never done. However, insurance companies eventually capitulated to pressure and began to pay for asymptomatic screening of colon cancer beginning age 50. These days, the initial screening has been pushed down to age 45 in part because the screening was so successful and has saved so many lives and much morbidity but also because colon cancer lately seems to arise in a younger and younger population.
Also using the Markovian framework from a cost perspective, if he had undergone a screening colonoscopy the first year he was eligible for screening, he would’ve likely saved himself a great deal of discomfort and the system circa $100,000 for just himself (elective colonoscopy would likely cost approximately $1000 for screening an asymptomatic patient). For a patient with colonic adenocarcinoma perforating the bowel wall, an ER visit, emergency surgery, perioperative care, subsequent critical care followed by inpatient care, and ultimately chemotherapy are extremely expensive. Roughly speaking, the United States population has a lifetime risk of slightly over 4% of developing colorectal cancer. Currently, that leads to approximately 150,000 new cases/year. Although many a forest has been felled to describe the best population screening for colorectal cancer, it is extremely effective at saving money.
As a penultimate thought, I think it is important to discuss, once again, alignment of incentives. Physicians and clinicians recommend to their patients, all the time, testing and screening. To make these recommendations, they usually use guidelines of national societies, or, absent that, use the training they have received on the topic. Free will in the patient allows one to accept or decline medical guidance. It is also a central principle of law. However, I now sincerely believe that certain recommendations, if declined, should shift financial burden onto the patient. Because insurance is a pool, one person’s decision affects all others’ insurance costs. I believe that there is strong evidence that the patient population, if given the right information and incentives, will make better choices. In the above example, the patient was no less than a physician himself. If he agreed to waive obligation of the insurance company to pay for his consequences if he developed colon cancer, it would seem fairer to the other members of insurance pool. It also may have compelled him to screen which would’ve saved him income, pain, and suffering. Obviously, allocation of this risk would have to be done very carefully but, if one thinks about it, it is fair.
So, in summary, prevention is best. Early intervention is the next best. I assert that return on investment by shifting medical attention upstream (to earlier states) in the Markovian model would pay for itself handsomely. The next question becomes how do we do this. Development of these thoughts will be shortly forthcoming.