Patient flow in Oncology - The most complex market of them all?

Oncology is quite probably the most complex health care market, and modeling it has become an essential need for most players.



Oncology is quite probably the most complex health care market, and modeling it has become an essential need for most players.

Oncology is many things - it is the largest sector for clinical trials, one of the major causes of death in most of the world, a significant sink-hole for health resources - not to mention being one of the most emotionally destructive diagnoses anyone can be given. To most executives in the pharma and biotech world oncology has become the unavoidable 800lb gorilla. Just ask yourself - how many companies in this industry do not have some direct interest in oncology? The developing trends of research in oncology have led to the most incredible increases in scientific developments, and as forecasters we must now develop new ways to model this market.

So why (you may ask) do we need new ways to model the market? The answer is simple - complexity. Let's explore some of the issues:

1. The regulatory and reimbursement environments for oncology is centered on the approval of specific treatments for cancers of a specific part of the body - usually an organ. In recent times the sub-divisions have increased but still within a particular organ. Lung cancer became small-cell and non-small-cell, and so on. This regulatory approach has led to the proliferation of approvals for new products, where a single new product may be launched for one organ type but the company will frequently be conducting trials in 5 or 6 other organs to expand the label. In addition, while these trials are being conducted oncologists are frequently using the new product off-label. Modeling this environment requires that our models have a classical patient type description that is based principally on defining cancer in terms of the affected organs, but the number of organs that must be modeled for a new drug is multiplying all the time.

2. The scientific community is increasingly using mode of action at the cellular level to discover new treatments. This is, of course, very exciting, but it creates an environment in which science is evolving at a level that is independent of the organ that is affected by the cancer cell (given that most of the cells that become cancerous do not live exclusively in a single organ). To effectively model this environment we need models that are capable of exploring the impact of treatments around the entire body - almost regardless of where in the body they occur.

3. Treatments in oncology are protocol driven. No patient gets diagnosed with cancer, receives a prescription for a drug and goes away to refill the drug forever. Thousands of complex protocols have evolved and they use combinations of radiation, chemo-therapy, and maintenance therapies, all shifting by the stage of the disease, impacted by side-effects, lifestyle issues, reimbursement factors - and a host of other complicating factors. To model this environment we need models that are able to define a patient in terms of their stage on the pathway of therapy based on the protocols they have been given, allowing for shifts in protocol at any time as their disease develops - frequently adding organs to those affected (increasing severity of the original diagnosis and spreading to other organs are the two most common forms of progression of disease in oncology).

So, the people in your company working on early stage developments need a system based on science, the late stage and post-launch people need a system based on organs, and anyone involved in product management need a system based on treatment pathways and protocols. This sounds to me like a perfect recipe for having multiple different forecasts that are apparently trying to model the same market but actually come up with totally different numbers - and this leaves senior management in a perfect mess trying to base different decisions on different numbers.

My contention is that we need a single system that will allow companies to view oncology from all of these different directions. Such a modeling system will be based on comprehensive analysis of patient flow dynamics, and I'll start filling in those blanks in my next blog entry.....