Lisa M. Schwartz, MD, MS
Steven Woloshin, MD, MS
H. Gilbert Welch, MD, MPH

Authors' response

Thank you for the opportunity to revise our manuscript "Screening Mammography: Women's attitudes towards false positive results and the detection of ductal carcinoma in-situ" and to respond to the suggestions of the editorial committee and statistical referee. We appreciate the careful review that the manuscript received.
 

Response to editorial committee comments

1. Explain current US practices for breast screening, in particular how the women taking part are warned about the dangers of false positives.

We have revised the introduction to highlight the prominent role of screening mammography in US practice and describe specific recommendations for screening. In addition, we also state in the introduction that it is unknown what women are currently being told about false positives.
 

2. Weighted column in table 1 was hard to understand - explain the weighting for demographic data.

The "weighted" results reflected the age and racial distribution of adult women in the United States. Given the statistical referee's concerns about these weighted results, we now present only crude data throughout the paper. We still mention weights (briefly in the methods) because we used stratified random sampling to select the study sample; we simply mention that weighting responses according to the probability of selection makes almost no difference to the results, and for simplicity we present only crude (i.e., unweighted) data.
 

3. In view of the recent Lancet article, should "correct answer" for the reduction in breast cancer screening attributable to screening be altered.

We have edited this sentence so that correct answer is now in quotations and have included a statement (and citation) that a recent study has raised questions about the correct answer.
 

Response to statistical reviewer's comments

1. Helpful to have details of the screening system in the US. In the UK, screening does not begin until 50 years old.

As stated above, we now provide this information in the introduction.
 

2. It may be better to present the crude results and comment on the sample (and its weaknesses). The weighted analysis adds little.

We agree with you that weighted results do not add very much. Thus, we have dropped the weighted analyses from the tables and discussion. We have left the data from the US Census in Table 1 to help readers understand the differences between our sample and the US population and comment on the weaknesses of the sample in the limitations section of the discussion.
 

3. Problem with last sentence on page 6. Suggest that the error of 4-6% be given but not the explanation, as there is little point when the sample is not representative of the population as a whole anyway.

We agree with your concern and have altered the sentence to read: "Based on our sample size, we estimate the margin of error of the results to be + 4-6%."

We look forward to hearing your thoughts on the revised manuscript and hope that you find it acceptable for publication.
 

Sincerely yours,
Lisa M. Schwartz, MD, MS
Steven Woloshin, MD, MS
H. Gilbert Welch, MD, MPH