Raising concerns about family history of breast cancer in primary care consultations: prospective, population based study

Following the availability of genetic tests for the genes for breast cancer BRCA1 and 2, genetic centres have reported increasing referral, often of women who are at low risk of breast cancer but who are concerned about their chances of inheriting it, and they have called for better management in primary care.¹ To inform appropriate management strategies we counted consultations in primary care in which a family history of breast cancer was mentioned. We obtained ethical approval from the Cambridge local research ethics committee.

Population, method, and results

Nineteen of the 36 partnerships with four or more partners in one health authority were recruited by letter and visit (mean list size 8904 (SD 2231); 74% training practices). A total of 240 clinicians participated: 152 doctors and 88 nurses, including locums and those working part time.

Each practice collected data over four weeks between August 1997 and July 1998. After all consultations with women aged 16 or older, clinicians recorded the patient's reference number, birth date, mention of a family history of breast cancer or other cancers, breast symptoms, risk of breast cancer, and who first mentioned any of these topics. Consultation data were checked against records of attendance at the practice. Agreement between the patient and clinician on who first mentioned a family history of breast cancer was assessed in a selected subsample of women. These women were invited to participate in a telephone interview by letter (no reminders). Respondents included 39 of 107 women classified as originating discussion of a family history of breast cancer and 33 of a 10% sample of those classified as not originating such discussions (total 681). Data were double-entered and analysed using STATA 5.0 (Statacorp, College Station, TX).

Eighteen of 19 practices participated, and 20 614 of 24 269 consultations (85%) were usable. A sensitivity analysis that assumed that all missing consultations came from the practice with the highest or lowest rate of reporting for a family history of breast cancer gave results within the confidence intervals of the main analysis. No differences in frequency of mentions of family history of breast cancer by clinicians were found over time.

Of the topics recorded, breast symptoms were mentioned in consultation most often, and family history of cancers other than breast cancer least often (table). Mention of a family history of breast cancer was recorded in 3.7% of consultations. Clinicians were 6.6 times more likely to raise the issue of family history of breast cancer than were women and were more likely to raise the issue in all topics counted.

Fifty of 72 women interviewed recalled mention of family history of breast cancer during their consultation; 42/50 agreed with the reporting clinician as to who raised the issue. Those disagreeing reported that the clinician, and not themselves, had done so. Five women reported consulting with specific concerns about family history of breast cancer.

Comment

In consultations with their general practitioners women raised the issue of a family history of breast cancer relatively infrequently, in only 5/1000 consultations—an average of 0.6 per clinician per month. This is consistent with the few other data available.²,³ In contrast, consultations in which women initiated discussion of breast symptoms were four times more common.

Applying list sizes and rates of consultation from the study practices allows a rough comparison with the morbidity statistics for general practice. For each 1000 women (⩾16 years) on the list, about 15 a year will raise the issue of a family history of breast cancer. Almost 10 times that number (141) consult for contraceptive advice, and three times that number consult for menstrual disorders.⁴

Interviews with women suggested that only a minority consult with specific concerns about family history. Primary care teams might manage these women most appropriately by training a team member in assessment and management techniques, possibly with computer support.⁵

Much has been made of the potential of the media for raising women's concerns about familial risk of breast cancer. These data draw attention to the potential of the primary care team itself to influence women's views through repeated inquiry about family history in the consultation.

Table.

Consultations (n=20 614^*) in which topics relating to breast cancer were mentioned. Values are number (%; 95% CI)

Topic area	All mentions†	Mentioned by clinician‡	Mentioned by patient	Clinician: patient ratio
Breast symptoms	1284 (6.1; 5.3 to 7.0)	830 (3.9; 3.1 to 4.8)	454 (2.2; 1.9 to 2.5)	1.82
Risk of breast cancer	849 (4.0; 3.3 to 4.9)	678 (3.3; 2.6 to 4.2)	171 (0.8; 0.7 to 1.0)	4.13
Family history of breast cancer	788 (3.7; 3.0 to 4.6)	681 (3.3; 2.6 to 4.2)	107 (0.5; 0.4 to 0.6)	6.60
Family history of other cancer	381 (1.8; 1.3 to 2.4)	283 (1.4; 1.0 to 1.9)	98 (0.5; 0.4 to 0.6)	2.80

Different areas were commonly mentioned together, such that family history of breast cancer was mentioned alone in only 0.5% of consultations. 

^*Missing values for some numbers. 

^†Occasionally the same patient mentioned a topic at more than one consultation. 

^‡General practitioner or practice nurse.