Thank you for the suggestions for improving the above mentioned manuscript. We have processed the comments of the BMJ and the reviewer as reported below.

Regarding the comments of the BMJ:
 

The denominator of the sample

The denominator of the sample has been added. This number includes all women consulting their family physicians consecutively during that period.
 

The reliability of the case finding system

In this practice there are no hand written records. The general practitioners only use computerised medical records, in which patients with specific risk factors and diseases can be marked and selected. This has been added to the text.
 

Compliance

This point has been changed in the way the reviewer proposed. We agree that general practitioners as well as women may have good reasons for not following genetic advice.
 

Small comments

• The design of the study has been added to the title.

• The word count is 595.

• Our study obtained ethical approval. This is mentioned in the paper.
 

Regarding the comments of the reviewer:

Professional non-compliance and patient non-compliance

The point of professional non-compliance has been added to this short report in the following way: It has become part of the title, and the aim of the study. Additionally, in the concluding sentence "uptake" has been changed into efficacy.

The number of general practitioners working in this practice (six) has been added to the paper. All six general practitioners were involved in this study and discussed the genetic advice with their own patients. This has been added more clearly to the short report. There were no significant differences between these six general practitioners regarding compliance and non-compliance. We agree with the reviewer that interviewing general practitioners about reasons for non-compliance might give a lot of information. However, in our opinion a number of six is too small for yielding scientifically reliable information.
 

The advice given the latest findings

The advice given was as follows: Risk estimates were communicated as relative risks (RR) and were divided into three groups: a normal or slightly increased familial breast cancer risk (RR < 2), a moderately increased familial breast cancer risk (2 £ RR < 3), a highly increased familial breast cancer risk (RR ³ 3). A normal or slightly increased familial breast cancer risk implied reassurance. A moderately increased familial breast cancer risk implied an advice for GP-surveillance (annual screening by mammography and palpation by the GP). A highly increased familial breast cancer risk implied a referral to a family cancer clinic. All women received an advice for monthly breast self-examination. A summary of this has been added to the Table. This advice given is in line with Dutch national guidelines as developed in 1999/2000 (currently in press). This has been added to the text.
 

Non-responders

The 14 women who consulted their general practitioner about their family history but did not consent to the study did not re-visit their general practitioner about their family history. For that reason, we do not have any further information on them related to this topic. The 18 who did not consent to the second interview were not statistically significant different from the 42 who did participate in the second interview, regarding relative risks or advice categories.
 

Abstract unclear and an annotated figure

The abstract has been excluded from the new version (short report). Due to lack of space we are not able to provide the extra figure the reviewer requests.

We hope the present manuscript is acceptable for publication in your journal.

Looking forward to hearing from you,

Yours sincerely,
Geertruida H de Bock, on behalf of the co-authors