A 35-year-old woman presents for an initial visit and during the medical interview mentions that her mother and grandmother had breast cancer. She reports that her mother was diagnosed with cancer at age 42 years, and she believes that her grandmother, on her mother's side, was diagnosed in her 30s. Because of her family history, she is concerned about her risk of developing breast cancer. Despite having no symptoms of breast cancer, she wonders at what age she should start having mammograms and whether she should have genetic testing.
Why Is It Important to Record an Accurate Family History of Cancer?
Individuals with a family history for certain kinds of cancers can have an increased risk of developing cancer themselves.1, 2 Two meta-analyses found relative risks of 2.1 (95% confidence interval [CI], 2.0 2.2) for breast cancer3 and 3.1 (95% CI, 2.6 3.7) for ovarian cancer4 in women with affected first-degree relatives. Similar higher risks have been observed for endometrial cancer,5-7 colon cancer,7-10 and prostate cancer.11-13
Accurate reporting of family history helps risk-stratify patients, which in turn determines screening and prevention interventions. Individuals with family histories that are suggestive of a hereditary cancer syndrome (Box 20-1)14-17 are typically considered at high or very high risk of developing cancer. Individuals with family histories for cancers not recognized as hereditary are generally at a moderately increased risk of developing cancer compared with the general population. Several organizations have recommended initiating screening earlier, more frequently, or both in patients at moderately increased risk of developing cancer according to their family history.18-27 Guidelines have also been published regarding the management of individuals who are at high cancer risk.26, 28-30 A family history of malignancy can not only influence cancer screening initiation and frequency but also affect treatment strategies. Family histories affect decisions about cancer chemoprevention,31, 32 and those individuals identified as being at very high risk may also be considered for risk-reducing surgeries.33 Likewise, algorithms that predict individuals who might be candidates for genetic testing rely almost exclusively on family history information.14, 16
Box 20-1Examples of Clinical Diagnostic Criteria for 2 Familial Cancer Syndromes and Recommendations Regarding Genetic Testing for Cancer Susceptibility |Favorite Table|Download (.pdf) Box 20-1 Examples of Clinical Diagnostic Criteria for 2 Familial Cancer Syndromes and Recommendations Regarding Genetic Testing for Cancer Susceptibility
|Hereditary Nonpolyposis Colon Cancer (HNPCC)14, 15 |
|All of the following criteria should be present: |
|At least 3 relatives must have a cancer associated with HNPCC (colon, endometrial, ovarian, stomach, small bowel, hepatobiliary, ureter, renal pelvis, brain). |
|One should be a first-degree relative of the other 2. |
|At least 2 successive generations should be affected. |
|At least 1 of the relatives with cancer associated with HNPCC should have received the diagnosis before aged 50 years. |
|Hereditary Breast/Ovarian Cancer16a |
|Any of the following criteria should be present: |
|Two breast cancers in a first- or second-degree relative and mean age at diagnosis of 40 years. |
|One breast cancer and 1 ovarian cancer in a first- or second-degree relative and mean age at diagnosis of 41 to 50 yearsa |
|Two or more breast cancers and 1 ovarian cancer in a first- or second-degree relative. |
|Ovarian cancer in 2 relatives. |
|American Society of Clinical Oncologists Policy Statement on Genetic Testing17 |
|Indications for genetic testing: |
|The individual has personal or family history features suggestive of a genetic cancer susceptibility condition. |
|The test can be adequately interpreted. |
|The results will aid in diagnosis or influence the medical or surgical treatment of the patient or family members at hereditary risk of cancer. |
Because many screening and prevention strategies for cancer rely on self-reported family history information, inaccurate information could result in inappropriate care. A false-negative family cancer history results in an underestimation of cancer risk and missed opportunities for cancer screening. Failure to collect adequate family history information and appropriately manage a patient's cancer risk may result in substandard care and in some cases has resulted in malpractice litigation.34
Conversely, a patient's false belief in a positive family cancer history can cause stress35 and, when compounded by the physician's overestimation of risk, may lead to unnecessary procedures or surgeries.36 The overestimation of cancer risk based on pedigree data creates unneeded referrals for genetic testing or cancer risk counseling.37, 38 The increased availability and demand for genetic services require an even more important role for primary care physicians in recording an accurate family cancer history39, 40; however, many physicians lack adequate training in genetics to accurately identify and refer appropriate candidates for genetic services.41, 42 In addition, with direct-to-consumer advertising for genetic testing now a reality,43 accurate family history collection and cancer risk assessment by primary care physicians might help decrease the likelihood of inappropriate referrals for genetic counseling and testing.
Few data exist describing how often inaccurate risk assessments are made according to faulty pedigree data. In one retrospective study35 that examined patients referred to 2 cancer genetic clinics, patient treatment was changed in 23 (11%) of 213 patients after their previously reported family history information was found to be inaccurate. In 15 of these patients, screening was thought to be unnecessary, although in 8 patients cancer risk was determined to be greater than initially believed. Further studies have supported these findings, with one study44 determining that 6 (5%) of 120 patients referred to a cancer clinic had changes in treatment after confirmation of the family cancer history revealed discrepancies. In most of these patients, the cancer risk had been overestimated.
Prevalence of a Positive Family History of Specific Familial Cancers
The prevalence of a family history of cancer varies, depending on the cancer type. The prevalence of a family history of breast cancer has been estimated to range from 5% to 22%45-48; colon cancer, 2.0% to 9.4%8, 45, 46; ovarian cancer, 1.1% to 3.5%45, 46, 48; endometrial cancer, 0.5% to 1.4%45, 46; and prostate cancer, 4.6% to 9.5%.11, 13, 46, 49 Most of this variation is based on methodology and study population. Some studies included distant relatives in the definition of a positive family cancer history, whereas other studies have focused only on first-degree relatives. Variability in rates also occurs when the results are derived from the general population as opposed to patients referred to cancer or genetic centers, which have higher prevalence rates.
How to Elicit a Family Cancer History
Family medical history information is important for risk assessment in numerous chronic medical conditions in addition to cancer, such as diabetes mellitus and cardiovascular disease; therefore, eliciting a family cancer history can serve as a model for collecting family history information for other disorders. Typically, family history information is collected directly from the patient or from screening questionnaires filled out by the patient. Alternatively, the patient's parent or another family member may provide the information.
Screening questionnaires are often either a list of relatives, with space to provide information on overall health, age, and cause of death, or a list of adult-onset diseases with space to list the affected relatives. Disease history should be collected on first-degree relatives (mother, father, sisters, brothers, and children) and second-degree relatives (maternal and paternal grandparents, aunts, uncles, nieces, and nephews). It is important to inquire about various types of cancers because certain hereditary cancer syndromes can be identified by specific cancers that cluster within families (Box 20-1), such as endometrial with colon50 and breast with ovarian.51, 52 If the initial screening interview or questionnaire reveals a potential familial predisposition to a particular disease, the family history should be expanded.
Although establishing the numbers of both affected and unaffected relatives is important for determining penetrance and predicting the likelihood of gene mutations, this information for primary care physicians would seldom influence cancer screening decisions. For affected relatives, documenting the age at cancer diagnosis is important because patients developing cancer at ages significantly earlier than typically expected increases the possibility of a hereditary cancer syndrome. Inaccurate reporting of ages at diagnosis for breast cancer can have a considerable influence on risk prediction in families with fewer than 4 affected relatives.53 A 3-generation pedigree, displayed graphically in Figure 20-1, offers a convenient symbolic method of summarizing information. Because of previous inconsistencies in pedigree symbol usage, the Pedigree Standardization Task Force, organized through the National Society of Genetic Counselors, has proposed recommendations for a standardized pedigree nomenclature.54
Hypothetical Pedigree for a Consultand With a Family History Suggestive of a Hereditary Nonpolyposis Colon Cancer
A consultand is an individual under evaluation for predicting his or her own future risk or the risk of his or her offspring. The arrow identifies the consultand. The 2 letter abbreviation with number (eg, Co 53) represents the diagnosis an individual received, followed by the age at the diagnosis. This pedigree meets Amsterdam II criteria,14 which include 3 relatives with a hereditary nonpolyposis colorectal cancer–associated tumor, such as colon cancer, endometrial cancer, ureteral cancer, cancer of the renal pelvis, ovarian cancer, stomach cancer, or small bowel cancer; 1 relative who is a first-degree relative of the other 2; cancers affecting at least 2 generations; and 1 or more cases diagnosed before aged 50 years.
When recording a pedigree, particularly for breast and gynecologic cancers, it is important to inquire about disease in both maternal and paternal lineages because mutations can be transmitted through either parent. When collecting information on second-degree relatives, it is important to note the lineage to which the relative belongs (such as paternal vs maternal grandparents) because the degree of risk might vary if affected relatives do not belong to the same lineage. A brief reference for physicians on the family medical history has been prepared by the American Medical Association (http://www.ama-assn.org/ama/pub/category/2380.html; accessed May 29, 2008). Many other electronic sources and texts55 are available.
Family medical history information can be collected during a patient care visit or outside of the clinical encounter. Methods of collecting family history information outside of the clinical encounter can include paper questionnaires,56 computer questionnaires in kiosks within a clinic waiting area,37 Web-based electronic collection, and interviews by health care professionals. The optimal means of collection has not been determined.
Two of the authors (H.J.M. and D.R.S.) performed independent searches of the MEDLINE database for English-language articles dated 1966 to June 2004 from the PubMed search engine. The following Medical Subject Headings were used: “family,” “genetic predisposition to disease,” “medical history taking,” “neoplasm,” and “reproducibility of results.” We also searched using the following textwords: “accuracy,” “sensitivity,” “specificity,” and “family history,” combined with the conditions “breast cancer,” “colon cancer,” “ovarian cancer,” “prostate cancer,” “endometrial cancer,” or “uterine cancer.” We specifically included cancers that were likely to be commonly encountered by primary care physicians and whose management might be altered according to family history information. The reviewers evaluated article abstracts and chose studies for full-text review according to the abstract. We searched the bibliographies of all retrieved articles to identify additional sources.
Articles were included if they were original articles describing the accuracy of the site-specific family history for the prespecified cancers and contained a criterion standard. Studies presenting aggregate data (all cancer types combined into a single measure) for self-reported family cancer history information were excluded. For purposes of this study, the criterion standard for a positive family history of cancer required verification from the identified relative's medical record, physician, or death certificate or verification within a population cancer registry. For studies to be included in our analysis, verification of a negative family history for cancer had to have been performed. Thus, if a study participant reported that a relative had no history of breast cancer, the relative's medical records, death certificate if applicable, or local cancer registry was examined for verification of this report.
The completeness of case findings within tumor registries varied, with 83% to 99% of cancers identified through medical record reviews and patient interviews also being present within the registry.57-61 Specific cancer sites are correctly recorded within the registry in 93% to 97% of cases. Forty nine percent of discrepancies within tumor registries result from changes in an initial diagnosis with a failure to update registry information.61 For breast cancer, the sensitivity and specificity of tumor registries are high.62 Other tumors listed within registries have similar high sensitivities.63, 64 The National Program of Cancer Registries of the Centers for Disease Control and Prevention has created a system that provides the rationale for accepting these data in studies that attempt validation of the patient's family cancer history.65 Although death certificates probably lack the accuracy of tumor registries, the poorer performance of death certificates is more likely attributed to poor sensitivity (the death certificates do not record the information when in fact the decedent had cancer).66 According to autopsy studies, the death certificate is estimated to have a sensitivity of 87% for identifying cancer.67
We identified 22 studies from our search, using the listed criteria.44, 68-88 Of these, only 7 provided information on both the test characteristics of a positive and negative report of a family cancer history.71, 73, 74, 76, 78-80 One study specifically assessed pedigrees suggestive of hereditary nonpolyposis colorectal cancer (HNPCC)76 and was included within the analysis. Sensitivity and specificity were determined for the family history interview for HNPCC, but this information was not combined with other colon cancer studies. We used techniques described from previous Rational Clinical Examination articles to determine study quality, and all 7 evaluated studies were assigned a quality score of C,89 which reflects a study with an independent blind comparison of sign or symptom and a criterion standard of diagnosis among nonconsecutive patients suspected of having the target condition.
Because the population studied could influence reporting accuracy, test characteristics were calculated separately for individuals with a personal history of cancer, as well as individuals without a personal history of cancer. Sensitivity and specificity of patient self-report of a family history of cancer and likelihood ratios (LRs) of a positive or negative report were calculated according to raw data supplied by the original articles that met our search criteria. CIs for LRs were computed with previously described methods.90 We used random-effects summary measures for combining the data because this provided broader CIs that display the uncertainty around the point estimates. The summary measures described this uncertainty better than the simple range of possible data from the original studies. For colon cancer, one study76 was not included within the summary LRs because it specifically evaluated the family history for HNPCC rather than colon cancer in general.
Precision reflects the reproducibility of a measurement. Assessing the precision of the family history interview is difficult because it can be influenced by both patient and physician factors. Although we were unable to identify any studies assessing the reliability of the physician's family cancer history assessment, one study91 in breast cancer examined the reliability of patient self-report. In this nested case-control study,91 comparisons were made between self-reported family history information in women before the development of the disease and after the development of the disease. Follow-up surveys were completed 2 years after the initial survey. Women who had developed breast cancer, as well as those who had not developed breast cancer, were surveyed. The agreement for maternal history of breast cancer was κ = 0.92 and κ = 1.0 for cases and controls, respectively; and for a history of breast cancer in a sister, κ = 0.65 and κ = 0.88, respectively. Although the study did not assess whether a real change in family history might have occurred during the study period, these results suggest that self-reported family breast cancer history is probably only slightly influenced by recall bias. Patient precision regarding the family history interview for other cancers has not been reported.
Accuracy represents how well a particular test measures the value it is intending to measure. Seven studies concerning family cancer history were ultimately included in this analysis (Table 20-1).71, 73, 74, 76, 78-80 Three studies collected family history from personal patient interviews,73, 74, 79 whereas the other 4 relied on a self-completed survey.71, 76, 78, 80 Four studies solely relied on cancer registry data as their criterion standard 73, 74, 79, 80; 2 studies used a combination of medical records and death certificates,71, 78 whereas the remaining study used all 3 sources as its criterion standard.76 Only information for first-degree relatives was extracted.
Table 20-1Characteristics of Included Studies of Patient Report of a Family History of Cancer in a First-Degree Relativea |Favorite Table|Download (.pdf) Table 20-1 Characteristics of Included Studies of Patient Report of a Family History of Cancer in a First-Degree Relativea
|Source, y ||Cancer Site ||Method of Family History Information Collection ||Criterion Standard |
| ||Affected Individuals ||Unaffected Individuals || || |
|Love et al,68 1985 ||Colon, breast || ||Personal interview ||Medical records, death certificate |
|Breuer et al,69 1993 ||Breast || ||Self-completed survey ||Medical records |
|Theis et al,70 1994 ||Colon, prostate, breast, ovarian || ||Personal interview and self-completed survey ||Medical records, death certificate, cancer registry |
|Aitken et al,71 1995 || ||Colon ||Self-completed survey ||Medical records, death certificate |
|Parent et al,72 1995 ||Breast ||Breast ||Personal interview ||Medical records |
|Anton-Culver et al,73 1996 ||Breast || ||Personal interview ||Cancer registry |
|Kerber and Slattery,74 1997 ||Colon, prostate, breast, endometrial, ovarian ||Colon, prostate, breast, endometrial, ovarian ||Personal interview ||Cancer registry |
|Sijmons et al,44 2000 ||Colon, breast, ovarian || ||Personal interview and self-completed survey ||Medical records, death certificate |
|Eerola et al,75 2000 ||Breast || ||Personal interview and self-completed survey ||Medical records, cancer registry |
|Katballe et al,76 2001 ||Colon || ||Self-completed survey ||Medical records, death certificate, cancer registry |
|King et al,77 2002 ||Colon, prostate, breast, endometrial, ovarian || ||Personal interview ||Medical records, death certificate |
|Ziogas and Anton-Culver,78 2003 ||Colon, prostate, breast, endometrial, ovarian || ||Self-completed survey ||Medical records, death certificate |
|Mitchell et al,79 2004 ||Colon ||Colon ||Personal interview ||Cancer registry |
|Verkooijen et al,80 2004 ||Breast, ovarian || ||Self-completed survey ||Cancer registry |
For individuals with cancer (Table 20-2), the positive likelihood ratio (LR+) and negative likelihood ratio (LR–) of a self-reported family cancer history in a first-degree relative were 23 (95% CI, 8.1 64) and 0.29 (95% CI, 0.13 0.67) for colon, 41 (95% CI, 23 75) and 0.07 (95% CI, 0.03 0.13) for breast, 20 (95% CI, 4.3 89) and 0.55 (95% CI, 0.35 0.86) for endometrial, 44 (95% CI, 15 132) and 0.21 (95% CI, 0.12 0.37) for ovarian, and 24 (95% CI, 2.3 262) and 0.25 (95% CI, 0.16 0.39) for prostate cancers, respectively. For patients without a personal history of cancer (Table 20-3), the LR+ and LR– of a family history for the following cancers in a first degree relative were 23 (95% CI, 6.4 81) and 0.25 (95% CI, 0.10 0.63) for colon, 8.9 (95% CI, 5.4 15) and 0.20 (95% CI, 0.08 0.49) for breast, 14 (95% CI, 2.2 83) and 0.68 (95% CI, 0.31 1.5) for endometrial, 34 (95% CI, 5.7 202) and 0.51 (95% CI, 0.13 2.1) for ovarian, and 12 (95% CI, 6.5 24) and 0.32 (95% CI, 0.18 0.55) for prostate cancers, respectively. The estimates for sensitivity, specificity, and LRs for unaffected individuals for prostate, breast, endometrial, and ovarian cancers are based on data from a single study by Kerber and Slattery.74
Table 20-2Studies Evaluating Both Sensitivity and Specificity of Patient Report of a Family History of Cancer in a First-Degree Relative in Individuals With Cancer |Favorite Table|Download (.pdf) Table 20-2 Studies Evaluating Both Sensitivity and Specificity of Patient Report of a Family History of Cancer in a First-Degree Relative in Individuals With Cancer
|Source, y ||Cancer Type ||No. of Patients/Total (%) ||LR (95% CI) |
| || ||Sensitivity ||Specificity ||Positive ||Negative |
|Kerber and Slattery,74 1997 ||Colon ||11/17 (65) ||98/108 (91) ||6.9 (3.5-13) ||0.39 (0.20-0.74) |
|Katballe et al,76 2001a ||Colon ||11/18 (61) ||66/69 (96) ||14 (4.4-45) ||0.41 (0.23-0.73) |
|Ziogas and Anton-Culver,78 2003 ||Colon ||174/194 (90) ||1454/1498 (97) ||31 (23-41) ||0.11 (0.07-0.16) |
|Mitchell et al,79 2004 ||Colon ||30/53 (57) ||1256/1269 (99) ||55 (31-100) ||0.44 (0.32-0.60) |
|Summaryb || || || ||23 (8.1-64) ||0.29 (0.13-0.67) |
|Kerber and Slattery,74 1997 ||Prostate ||11/16 (69) ||101/109 (93) ||9.4 (4.5-20) ||0.34 (0.16-0.70) |
|Ziogas and Anton-Culver,78 2003 ||Prostate ||46/58 (79) ||557/564 (99) ||64 (30-135) ||0.21 (0.13-0.35) |
|Summary || || || ||24 (2.3-262) ||0.25 (0.16-0.39) |
|Anton-Culver et al,73 1996 ||Breast ||54/59 (92) ||364/370 (98) ||56 (25-125) ||0.09 (0.04-0.20) |
|Kerber and Slattery,74 1997 ||Breast ||11/13 (85) ||107/112 (96) ||19 (7.8-46) ||0.16 (0.05-0.58) |
|Verkooijen et al,80 2004 ||Breast ||60/61 (98) ||247/249 (99) ||122 (31-487) ||0.02 (0-0.12) |
|Ziogas and Anton-Culver,78 2003 ||Breast ||188/197 (95) ||850/873 (97) ||36 (24-54) ||0.05 (0.03-0.09) |
|Summary || || || ||41 (23-75) ||0.07 (0.03-0.13) |
|Kerber and Slattery,74 1997 ||Endometrial ||2/7 (29) ||114/118 (97) ||8.4 (1.9-38) ||0.74 (0.46-1.2) |
|Ziogas and Anton-Culver,78 2003 ||Endometrial ||10/18 (56) ||1035/1052 (98) ||34 (18-64) ||0.45 (0.27-0.76) |
|Summary || || || ||20 (4.3-89) ||0.55 (0.35-0.86) |
|Kerber and Slattery,74 1997 ||Ovarian ||2/3 (67) ||117/122 (96) ||16 (5.0-53) ||0.35 (0.07-1.7) |
|Verkooijen et al,80 2004 ||Ovarian ||4/6 (67) ||168/170 (99) ||57 (13-251) ||0.34 (0.11-1.0) |
|Ziogas and Anton-Culver,78 2003 ||Ovarian ||35/42 (83) ||1017/1028 (99) ||78 (43-142) ||0.17 (0.09-0.33) |
|Summary || || || ||44 (15-132) ||0.21 (0.12-0.37) |
Table 20-3Studies Evaluating Both Sensitivity and Specificity of Patient Report of a Family History of Cancer in a First-Degree Relative in Healthy Individuals |Favorite Table|Download (.pdf) Table 20-3 Studies Evaluating Both Sensitivity and Specificity of Patient Report of a Family History of Cancer in a First-Degree Relative in Healthy Individuals
|Source, y ||Cancer Type ||No. of Patients/Total (%) ||LR (95% CI) |
| || ||Sensitivity ||Specificity ||Positive ||Negative |
|Aitken et al,71 1995 ||Colon ||70/81 (86) ||219/239 (92) ||10 (6.7-16) ||0.15 (0.09-0.26) |
|Kerber and Slattery,74 1997 ||Colon ||13/16 (81) ||178/190 (94) ||13 (7.1-23) ||0.20 (0.07-0.56) |
|Mitchell et al,79 2004 ||Colon ||9/17 (53) ||1015/1020 (99) ||108 (40-288) ||0.47 (0.29-0.78) |
|Summary || || || ||23 (6.4-81) ||0.25 (0.10-0.63) |
|Kerber and Slattery,74 1997 ||Prostate ||21/30 (70) ||166/176 (94) ||12 (6.5-24) ||0.32 (0.18-0.55) |
|Kerber and Slattery,74 1997 ||Breast ||18/22 (82) ||167/184 (91) ||8.9 (5.4-15) ||0.20 (0.08-0.49) |
|Kerber and Slattery,74 1997 ||Endometrial ||1/3 (33) ||198/203 (98) ||14 (2.2-83) ||0.68 (0.31-1.5) |
|Kerber and Slattery,74 1997 ||Ovarian ||1/2 (50) ||201/204 (99) ||34 (5.7-202) ||0.51 (0.13-2.1) |
Of the remaining 15 studies, we excluded 8 studies81-88 because the tumor data were presented in aggregate (ie, family history of any cancer) or were unclear; therefore, we were unable to extrapolate site-specific numbers. Seven studies evaluated only the positive predictive value of self-reported family history information for breast, colon, ovarian, prostate, and endometrial cancers (Table 20-4). Positive predictive values tended to be better in articles concerning first-degree relatives compared with second-degree relatives. Individuals with personal histories of cancer tended to report family histories with a greater positive predictive value, although the number of studies evaluating unaffected individuals was limited.
Table 20-4Predictive Value of a Positive Report of a Family History of Cancer in a First-Degree or Second-Degree Relative |Favorite Table|Download (.pdf) Table 20-4 Predictive Value of a Positive Report of a Family History of Cancer in a First-Degree or Second-Degree Relative
|Source, y ||Cancer Type || ||Positive Predictive Value, No. of Patients/Total (%) |
| || ||Cancer Cases ||Healthy Controls |
| || ||First-Degree Relative ||Second-Degree Relative ||First-Degree Relative ||Second-Degree Relative |
|Kerber and Slattery,74 1997 ||Colon ||11/21 (52) || ||13/25 (52) || |
|Mitchell et al,79 2004 ||Colon ||33/43 (70) ||13/22 (62) ||9/14 (63) ||10/16 (63) |
|King et al,77 2002 ||Colon ||22/24 (92) || || || |
|Love et al,68 1985 ||Colon ||39/42 (93) ||31/37 (84) || || |
|Sijmons et al,44 2000 ||Colon ||30/33 (91) ||15/15 (100) || || |
|Theis et al,70 1994 ||Colon ||13/14 (93) ||21/29 (72) || || |
|Ziogas and Anton-Culver,78 2003 ||Colon ||174/218 (80) ||52/70 (74) || || |
|Aitken et al,71 1995 ||Colon || || ||70/90 (78) || |
|Summarya || ||81 (77-85) ||77 (70-83) ||71 (63-78) ||63 (39-82) |
|Kerber and Slattery,74 1997 ||Prostate ||11/19 (58) || ||21/31 (68) || |
|Ziogas and Anton-Culver,78 2003 ||Prostate ||46/53 (87) ||30/40 (75) || || |
|Theis et al,70 1994 ||Prostate ||11/13 (85) ||11/11 (100) || || |
|King et al,77 2002 ||Prostate ||25/29 (86) || || || |
|Summarya || ||85 (78-90) ||80 (67-89) ||68 (50-82) || |
|Kerber and Slattery,74 1997 ||Breast ||11/16 (69) || ||18/35 (51) || |
|Parent et al,72 1995 ||Breast ||67/74 (91) || ||33/34 (97) || |
|Ziogas and Anton-Culver,78 2003 ||Breast ||188/211 (89) ||103/115 (90) || || |
|Eerola et al,75 2000 ||Breast ||94/99 (95) ||109/114 (96) || || |
|Sijmons et al,44 2000 ||Breast ||65/69 (94) ||28/31 (90) || || |
|Theis et al,70 1994 ||Breast ||166/167 (99) ||33/39 (85) || || |
|Love et al,68 1985 ||Breast ||78/83 (94) ||65/74 (88) || || |
|Anton-Culver et al,73 1996 ||Breast ||54/60 (90) || || || |
|Verkooijen et al,80 2004 ||Breast ||60/62 (97) || || || |
|King et al,77 2002 ||Breast ||38/40 (95) || || || |
|Breuer et al,69 1993 ||Breast ||84/94 (89) || || || |
|Summarya || ||93 (91-94) ||91 (88-94) ||74 (63-83) || |
|Kerber and Slattery,74 1997 ||Endometrial ||2/6 (33) || ||1/6 (17) || |
|Ziogas and Anton-Culver,78 2003 ||Endometrial ||10/27 (37) ||3/14 (21) || || |
|King et al,77 2002 ||Endometrial ||2/5 (40) || || || |
|Summarya || ||37 (24-53) ||21 (7-47) ||17 (3-57) || |
|Kerber and Slattery,74 1997 ||Ovarian ||2/7 (28) || ||1/4 (25) || |
|Ziogas and Anton-Culver,78 2003 ||Ovarian ||35/46 (76) ||15/24 (63) || || |
|Sijmons et al,44 2000 ||Ovarian ||10/15 (67) || || || |
|Verkooijen et al,80 2004 ||Ovarian ||4/6 (67) || || || |
|Theis et al,70 1994 ||Ovarian ||2/2 (100) || || || |
|King et al,77 2002 ||Ovarian ||2/4 (50) || || || |
|Summarya || ||69 (58-78) ||63 (43-79) ||25 (5-70) || |
Common Reasons for False-Positive or False-Negative Reports
In cancers in which patients are likely to be accurate in their report, such as breast cancer, case reports have indicated that false-positive reports are associated with malingering, problems with patient-physician communication, or history of benign breast disease being reported as malignant.36 Other common reasons for false-positive reports of family cancer history result from confusion based on primary vs metastatic disease.68, 92 This confusion has been described with false reports of primary liver cancer, as well as central nervous system cancers. Cancers that are frequently overreported include melanoma, which is incorrectly reported in almost half the reports,93 and noncolonic gastrointestinal malignancies.35
Several factors relate to a false-negative report of a family history of cancer. In one study,94 older patients and nonwhite respondents were more likely to underreport a family history of cancer. Another study74 demonstrated that older patients were more likely to falsely report a negative family history of cancer, whereas patient sex and education level have little effect on the accuracy of reporting. Specific cancers with high rates of false-negative reporting include central nervous system tumors and hematologic malignancies.94
Other Means for Collecting Family History Information and Ways to Improve Family History Data Collection
Several barriers exist for the collection of family history information. Patient-specific factors that might result in poor pedigree collection include poor family communication, family myths, or individual spiritual beliefs. For physicians, probably the most significant barrier is time. Although a comprehensive family history assessment can take 15 to 30 minutes,95 the average primary care visit lasts only 16 minutes.96 Several alternative methods that involve collecting this information outside the context of the clinical visit may facilitate the collection of family history information. These other methods include self-completed patient paper surveys, computer-based tools, and personal visits arranged solely for pedigree collection.
Family history questionnaires offered outside of a clinical visit confer several theoretic advantages to visit-based pedigree assessment.97 Besides saving clinic time, patients can consult with family members to check the accuracy of the information, which can then be reviewed and integrated into a clinic appointment when relevant. The data from a questionnaire developed in Switzerland compared with information found within 2 population-based cancer registries exhibited sensitivities of 74% and 85% and specificities of 97%.56 Family history assessment tools (Box 20-2) have also been developed to assist physicians in determining which individuals might be candidates for genetic testing.98
Box 20-2Selected Web Sites for Cancer Risk Calculatorsa
Computerized genograms can also be effective and convenient tools for both patients and physicians.99-101 These tools offer the benefits of paper-based systems and, through clinical decision support, educate patients and offer guidance to physicians.102-104 Sweet et al37 compared family history information obtained by physicians at a comprehensive cancer clinic with those directly entered by patients into a computer program. Patients were then determined to be “high risk” for cancer according to pedigree information collected from either the computer program or information recorded within the medical record. Of 362 computer entries, 69% had some form of family history information recorded within their medical record. A total of 101 patients were considered high risk according to their pedigree information collected from the computer program, but only 69 of these patients had information recorded within their medical record to confirm this high risk.
Special visits outside of the clinical encounter have also been evaluated as a means to obtain family history information. In one study,105 patients observed at a single primary care practice were invited to a special visit designed to collect detailed family history information. Ten percent of patients observed in the pedigree clinic had a family history of cancer (breast, colon, melanoma, or thyroid) and some patients were referred for further care according to their pedigree. Patients were less anxious about their family history after the special visit, but this effect was not sustained beyond 12 weeks. A major limitation of the study was the poor attendance to the special clinic; only 16% of invited patients attended.
This patient's pretest probability of having a positive family history of breast cancer before her clinical interview can be estimated at 22%.48 After consideration of her self-reported pedigree, her posttest probability for a family history of breast cancer increased to 71% (95% CI, 60% 81%) (Table 20-5). According to the family history information presented, it is likely that this woman does in fact have a positive family history for breast cancer and that seeking confirmatory evidence is unlikely to offer any additional gain. This patient has 2 relatives affected with breast cancer, both diagnosed before 50 years and one a first-degree relative. According to online mutation prevalence tables (Box 20-2), this woman's risk estimate of harboring a deleterious mutation for either BRCA1 or BRCA2 was 10%, and a referral was recommended to a specialized cancer risk assessment program for counseling on genetic testing.106 During pretesting counseling, she learned of the potential limitations of testing on a presymptomatic individual without a known deleterious mutation. The genetic counselor suggested testing one of her affected relatives first to produce a more informative test. The patient said that her maternal grandmother is no longer alive, and she is not particularly close to her mother and unsure whether her mother would be willing to undergo testing. After pretesting genetic counseling, the patient decided not to be tested for a BRCA1 or BRCA2 mutation and planned to broach the subject of testing with her mother. According to this patient's family history, clinical breast examinations every 6 months are prescribed and the patient is shown how to perform monthly self–breast examinations. Mammography screening is initiated and an appointment is scheduled.
Table 20-5Posttest Probabilities of Having a Family History of Cancera
|Favorite Table|Download (.pdf)
Posttest Probabilities of Having a Family History of Cancera
||Estimated Cancer Family History Prevalence (Pretest Probability), %
||Personal History of Cancer
||Posttest Probability of Having a Family Cancer History in a First-Degree Relative, LR (95% CI)
Family history assessment is taking on greater importance as high risk individuals are being offered earlier screening interventions and risk reducing therapies. Cancer family histories acquired on first degree relatives for breast and colon cancer are likely to represent true positives and true negatives for the disease and may not require further evaluation to substantiate. However, other cancers with a familial disposition are less accurately reported.
The following disclosures were reported at the time this original article was first published in JAMA.
Funding/Support: This study was supported in part by grant K07CA08453 from the National Institutes of Health (Dr Syngal).
Role of the Sponsor: The National Institutes of Health did not participate in the design and conduct of the study, in the collection, analysis, and interpretation of the data, or in the preparation, review, or approval of the manuscript.
H. Familial cancer: consequences for the oncological practice. Cancer Detect Prev. 1986;9(5-6):455–458.
et al.. Risk factors
among young women with endometrial cancer: a Danish case-control study
. Am J Obstet Gynecol.
2000;182(1 pt 1):23–29.
WD; Cancer and Steroid Hormone Study Group. A population-based study of endometrial cancer and familial risk in younger women. Cancer Epidemiol Biomarkers Prev.
LH. Familial clustering of colon, breast, uterine, and ovarian cancers as assessed by family history. Genet Epidemiol.
WC. A prospective study
of family history and the risk of colorectal cancer. N Engl J Med.
ES. Cancer risk in relatives of patients with common colorectal cancer. Ann Intern Med. 1993;118(10):785–790.
W. Cancer control by family history. Anticancer Res.
JB. Family history and the risk of prostate cancer. Urology.
S. Family history and prostate cancer risk. Am J Epidemiol.
et al.. Family history and prostate cancer risk in a population-based cohort
of Iowa men. Cancer Epidemiol Biomarkers Prev.
HT. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome
) proposed by the International Collaborative group on HNPCC. Gastroenterology.
HT, de la Chapelle
A. Hereditary colorectal cancer. N Engl J Med.
American Society of Clinical Oncology. American Society of Clinical Oncology policy statement update: genetic testing for cancer susceptibility. J Clin Oncol. 2003;21(12):2397–2406.
US Preventive Services Task Force. Screening
for breast cancer: recommendations and rationale. Ann Intern Med.
2002;137(5 pt 1):344–346.
US Preventive Services Task Force. Screening
for prostate cancer: recommendation and rationale. Ann Intern Med.
US Preventive Services Task Force. Screening
for colorectal cancer: recommendation and rationale. Ann Intern Med.
RA, von Eschenbach
et al.. American Cancer Society guidelines for the early detection of cancer. CA Cancer J Clin.
National Institutes of Health Consensus Development Panel. National Institutes of Health Consensus Development Conference Statement: breast cancer screening
for women ages 40 49, January 21 23, 1997. J Natl Cancer Inst.
mammography for breast cancer: American College of Preventive Medicine practice policy statement. Am J Prev Med.
SJ. Risk and surveillance of individuals with heritable factors for colorectal cancer: WHO Collaborating Centre for the Prevention of Colorectal Cancer. Bull World Health Organ.
et al.. Colorectal cancer screening
and surveillance: clinical guidelines and rationale-update based on new evidence. Gastroenterology.
GP. Revision in American Cancer Society recommendations for the early detection of colorectal cancer. CA Cancer J Clin.
National Institutes of Health Consensus Development Conference Statement. Ovarian cancer: screening
, treatment, and follow-up
. Gynecol Oncol.
1994;55(3 pt 2):S4–S14.
et al.. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant
Breast and Bowel Project P 1 Study. J Natl Cancer Inst.
et al.. Continued breast cancer risk reduction in postmenopausal women treated with raloxifene: 4-year results from the MORE Trial. Breast Cancer Res Treat.
et al.. Efficacy of bilateral prophylactic mastectomy in women with a family history of breast cancer. N Engl J Med.
MJ. Hereditary cancer litigation: a status report. Oncology (Williston Park).
SA. The accuracy of diagnoses as reported in families with cancer: a retrospective study. J Med Genet.
et al.. False family history of breast cancer in the family cancer clinic. Eur J Surg Oncol.
JA. Identification and referral of families at high risk for cancer susceptibility. J Clin Oncol.
et al.. GPs’ views on their role in cancer genetics services and current practice. Fam Pract.
P. Expanding the role of the family history in primary care [editorial]. Br J Gen Pract.
H. Genetics in primary care: report on workshop of EC Concerted Action on Genetics Services in Europe (CAGSE) in association with the Royal College of GP Spring Meeting, Blackpool, UK, 28 April 1995. J Med Genet.
A. A systematic review
of the literature exploring the role of primary care in genetic services. Fam Pract.
L. Physician knowledge and attitudes towards molecular genetic (DNA) testing of their patients. Clin Genet.
et al.. Accuracy of family history of cancer: clinical genetic implications. Eur J Hum Genet.
SV. The prevalence
of a family history of cancer in general practice. Fam Pract.
JI. Family history: a comprehensive genetic risk assessment method for the chronic conditions of adulthood. Am J Med Genet.
et al.. Prevalence
of family history of breast and ovarian cancer in a single primary care practice using a self-administered questionnaire. Breast J.
et al.. Family history and prostate cancer screening
with prostate-specific antigen. J Clin Oncol.
et al.. Hereditary nonpolyposis colorectal cancer (Lynch syndromes I and II), I: clinical description of resource. Cancer.
JM. Family history of gynaecological cancers: relationships to the incidence
of breast cancer prior to age 55. Int J Epidemiol.
F, La Vecchia
L. Family history of breast, ovarian and endometrial cancer and risk of breast cancer. Int J Epidemiol.
J. Breast cancer risk assessment: use of complete pedigree
information and the effect of misspecified ages at diagnosis of affected relatives. Hum Genet.
et al.. Recommendations for standardized human pedigree
nomenclature. Am J Hum Genet.
RL. The Practical Guide to the Genetic Family History. New York, NY: Wiley-Liss; 1999.
J. Taking a family history in cancer patients with a simple questionnaire. Anticancer Res. 1998;18(4B):2811–2814.
BF. Completeness of hospital cancer case reporting from the SEER Program of the National Cancer Institute. Cancer.
AJ. Completeness of cancer and death follow-up
obtained through the National Health Service Central Register for England and Wales. Br J Cancer.
et al.. Using hospital discharge files to enhance cancer surveillance. Am J Epidemiol.
et al.. The Finnish Cancer Registry as follow-up
source of a large trial cohort
: accuracy and delay. Acta Oncol.
et al.. How complete and accurate are cancer registrations notified by the National Health Service Central Register for England and Wales? J Epidemiol Community Health.
MR. Determining the quality of breast cancer care: do tumor registries measure up? Ann Intern Med. 2000;132(9):705–710.
et al.. Validity
of self-reported cancers in a prospective cohort study
in comparison with data from state cancer registries. Am J Epidemiol.
C. The completeness and accuracy of health authority and cancer registry records according to a study of ovarian neoplasms. Public Health.
VJ. The potential and limitations of data from population-based state cancer registries. Am J Public Health.
. Information on death certificates: cause for concern? J Clin Pathol.
H. The autopsy as a measure of accuracy of the death certificate. N Engl J Med.
DM. The accuracy of patient reports of a family history of cancer. J Chronic Dis.
DG. Reporting bilaterality status in first-degree relatives with breast cancer: a validity
study. Genet Epidemiol.
D. Accuracy of family cancer history in breast cancer patients. Eur J Cancer Prev.
R. How accurate is self-reported family history of colorectal cancer? Am J Epidemiol.
C. Accuracy of reports of familial breast cancer in a case-control series. Epidemiology.
D. Validation of family history of breast cancer and identification of the BRCA1 and other syndromes using a population-based cancer registry. Genet Epidemiol.
ML. Comparison of self-reported and database-linked family history of cancer data in a case-control study
. Am J Epidemiol.
H. Familial breast cancer in southern Finland: how prevalent are breast cancer families and can we trust the family history reported by patients? Eur J Cancer.
S. Patient accuracy of reporting on hereditary non-polyposis colorectal cancer-related malignancy in family members. Br J Surg.
CI. Accuracy of family history of cancer as reported by men with prostate cancer. Urology.
H. Validation of family history data in cancer family registries. Am J Prev Med.
et al.. Accuracy of reporting of family history of colorectal cancer. Gut.
et al.. Set-up of a population-based familial breast cancer registry in Geneva, Switzerland: validation of first results. Ann Oncol.
et al.. Evaluation of the family history collection process and the accuracy of cancer reporting among a series of women with endometrial cancer. Clin Cancer Res.
GB. Problems in establishing accurate family history in patients with ovarian cancer of epithelial origin. Cancer Detect Prev. 1987;10(3-4):279–283.
et al.. The accuracy of diagnosis as reported in families with cancer: a retrospective study. J Med Genet.
AB. The accuracy of siblings’ family history reports of thyroid and other cancers. Am J Public Health.
MA. Validation of family history of cancer in deceased members. J Natl Cancer Inst.
AT. Validation and accuracy of cancer reporting in first-degree relatives [abstract]. Am J Epidemiol. 1992;136(8):1019.
et al.. Family history of cancer in patients with glioma: a validation study of accuracy. J Natl Cancer Inst.
LC. Accuracy of family history of cancer obtained through interviews with relatives of patients with childhood sarcoma. J Clin Epidemiol.
DB. Likelihood ratios with confidence: sample size estimation for diagnostic test studies. J Clin Epidemiol.
E. Changes in self reported family history of breast cancer with change in case-control status. Eur J Epidemiol.
MA. Validation of family history of cancer in deceased family members. J Natl Cancer Inst.
in the assessment of family history of melanoma and its association with dysplastic nevi in a case-control study
. J Clin Epidemiol.
of self-reported cancer history: a comparison of health interview
data and cancer registry records. Am J Epidemiol.
LJ. The family history component of cancer genetic risk counseling. Cancer Nurs.
et al.. The duration of ambulatory visits to physicians. J Fam Pract.
. Genetic family history questionnaire. J Med Genet.
AG. A preliminary validation of a family history assessment form to select women at risk for breast or ovarian cancer for referral to a genetics center. Clin Genet.
M. The computerized genogram. Prim Care.
CJ. Development and evaluation of a computer genogram. J Fam Pract.
T. Efficacy of a touchscreen computer based family cancer history questionnaire and subsequent cancer risk assessment. J Med Genet.
J. Computer support for genetic advice in primary care. Br J Gen Pract.
N. An interactive computer program for educating and counseling patients about genetic susceptibility to breast cancer. J Cancer Educ.
N. Who should provide genetic education prior to gene testing? computers and other methods for improving patient understanding. Genet Test.
SM. Family history taking and genetic counselling in primary care. Fam Pract.
JG. Clinical practice: mammographic screening
for breast cancer. N Engl J Med.