Thursday, September 30, 2010

Mammography: An Analysis of Effectiveness?

In a recent well publicized NEJM article the authors stated that 2/3 s of the gain in drop of mortality from mammography is from improved procedures and only 1/3 from screening. We have examined this contention and in our view the remark is a bit weak, especially give the results given.

They conclude:

The availability of screening mammography was associated with a reduction in the rate of death from breast cancer, but the screening itself accounted for only about a third of the total reduction.

The analysis went as follows:

1. They chose three age groups, 20-49, 50-69, 70-84.

2. They then chose two periods, 1986-1995 and 11995-2005.

3. They then selected those who were screened and those who were not.

From the report we show death rates in three categories is presented below:

1. For 20-49 we have shown below the results. First note that things got worse not better with screening. Second mortality did drip from the first period to the second.

2. For the target age group, 50-69 we have the result as shown below:

Here we have better for screening and better for recent period.

3. Now for the older group we show below another rather confusing result, it gets better with no screening and it was better in the earlier period.

Now the results presented by the authors is based upon taking chart 2 above and presenting it as follows;

Here they first show a reduction of 18% in those not screened and then when they compare those screened there is an added 10% or a 1/3rd of the gain from screening.

Here is my concerns:

1. Why reject the data refuting this argument from the other two age groups.

2. What about other factors including the increase in obesity during that period and increased personal awareness etc. They have not been factored in.

3. This is for a specific country registry and the results may not be projectable.

This is an interesting study but it fails in my opinion to factor in many of the complex but critical side issues.