Neoplasms: 7

This is the seventh blog in a series of blogs on neoplasia.

In the past few blogs, I've been trying to explain the disconnect between cancer survival data and cancer death rate data. The cancer survival data seems to indicate that we're making enormous improvements in cancer treatment. The cancer death rate indicates that Americans are dying from cancer at about the same rate as they had been a half-century ago.

Several days ago, I listed over a dozen biases in cancer survival data that contribute to an overly optimistic sense of medical progress.

In this and the next few blogs, I thought I'd review some of these biases. The purpose of this exercise is to explain that the interpretation of survival data is enormously complex and that survival data is probably not the best way to gauge progress in the field of cancer research.

Today, let's look at diagnosis bias.

Diagnosis bias occurs when a group of people who have a disease or lesion, that is different from the disease or lesion carried by the previously studied group, is used in a repeat clinical trial. This distorts the survival outcome after treatment.

For example, if patients were accrued into a study based on erroneous diagnoses rendered by an incompetent pathologist, a group of patients with breast cancer might become diluted with patients who have benign breast disease. It is possible to generate an inflated survival rate, compared with earlier studies that did not include erroneous diagnoses, by including healthy patients in the treatment group.

Most modern clinical trials impose strict pathology reviews in an effort to reduce the effect of diagnosis bias.

More commonly, treatment groups are tainted when a new subset of patients is added to the original group, through enhanced detection and diagnosis of a biologically distinctive variant of disease. For example, there has been a many-fold increase in the detection and diagnosis of ductal carcinoma in situ (DCIS) of breast, over the past few decades, thanks to mass mammographic screenings. DCIS has a very good prognosis. Only a small percentage of patients with DCIS progress to invasive ductal carcinoma.

As the proportion of breast cancer patients with diagnosed DCIS increases, the survival of the group that includes DCIS cases also increases. By increasing the number of patients with a good prognosis breast cancer, an improved survival can be achieved, without an improvement in therapeutic protocol.

The remedy for diagnosis bias is to subdivide your study groups carefully. This remedy has its own problems. If the first study did not carefully separate patients by their prognostic subtypes, how can you compare prior results with future results (where the subtypes were separated). Also, if you break groups by their tumor subtypes, will each group contain a sufficient number of patients to produce statistically useful results?

We'll continue discussing the biases in survival data in the next blog.

-Copyright (C) 2008 Jules J. Berman

key words: cancer, tumor, tumour, carcinogen, neoplasia, neoplastic development, classification, biomedical informatics, tumor development, precancer, benign tumor, ontology, classification, developmental lineage classification and taxonomy of neoplasms