Background: This study evaluates predictors of BRCA1/2 testing among breast and ovarian cancer survivors who received genetic counseling as part of a randomized trial and evaluates moderators of counseling mode on testing uptake.

Methods: Predictors of BRCA1/2 testing within one year postcounseling were evaluated using multivariable logistic regression in a population-based sample of breast and ovarian cancer survivors at increased hereditary risk randomly assigned to in-person counseling (IPC; n = 379) versus telephone counseling (TC; n = 402). Variables that moderated the association between counseling mode and testing were identified by subgroup analysis.

Results: Testing uptake was associated with higher perceived comparative mutation risk [OR = 1.32; 95% confidence interval (CI), 1.11–1.57] in the adjusted analysis. Those without cost barriers had higher testing uptake (OR = 18.73; 95% CI, 7.09–49.46). Psychologic distress and perceived comparative mutation risk moderated the effect of counseling and testing. Uptake between IPC versus TC did not differ at low levels of distress and risk, but differed at high distress (26.3% TC vs. 44.3% IPC) and high perceived comparative risk (33.9% TC vs. 50.5% IPC).

Conclusions: Cost concerns are a strong determinant of testing. Differences in testing uptake by counseling mode may depend on precounseling distress and risk perceptions.

Impact: Cost concerns may contribute to low testing in population-based samples of at-risk cancer survivors. Precounseling psychosocial characteristics should be considered when offering in-person versus telephone counseling. Cancer Epidemiol Biomarkers Prev; 26(12); 1772–80. ©2017 AACR.

For breast and ovarian cancer survivors at increased risk for having BRCA1/2 mutations (1), genetic testing can personalize treatment, estimate the risk of new cancers, and inform risk management and discussions with family members. Population-based studies suggest that less than 60% of breast and ovarian cancer survivors at increased risk for a BRCA1/2 mutation undergo genetic testing (2–6). Much of our understanding of the factors associated with genetic testing decisions comes from studies of healthy at-risk individuals or cancer survivors self-referred to genetics clinics where testing uptake is high (2, 7). Existing population-based studies have focused on access and clinical factors (e.g., cost barriers, age of diagnosis; refs. 4, 8), yet the potential psychosocial (e.g., individual distress, social support, informed decision-making) and treatment implications (e.g., prophylactic surgery) of testing may also be salient concerns.

The psychosocial and treatment-related complexity of genetic testing is part of why genetic counseling is a standard of care for survivors at increased risk (1); however, only 40%–50% of survivors who meet the guidelines for a genetics referral are referred to or receive genetic counseling (4, 8). To address access barriers, genetic counselors are increasingly using alternative counseling delivery methods, such as telephone-based counseling (9). Our randomized controlled noninferiority trial of in-person versus telephone-based genetic counseling for breast and ovarian cancer survivors at increased hereditary cancer risk, the Risk Education and Assessment for Cancer Heredity (REACH) trial, found that telephone-based counseling was noninferior on key patient-reported outcomes such as distress and decisional satisfaction (5, 10). A secondary aim of the REACH trial was to examine factors associated with genetic testing decisions. However, genetic testing uptake in REACH was low overall (∼30%), with lower testing uptake among survivors receiving telephone-based counseling (5, 10), suggesting the need to not only identify predictors of testing decisions, but also to determine whether in-person versus telephone-based counseling yielded different uptake depending on patient characteristics or patient perceptions of counseling.

To address the secondary aim of the trial, we evaluated psychosocial, clinical, and access factors in BRCA1/2 testing decisions among breast and ovarian cancer survivors at increased hereditary cancer risk. We also evaluated moderators of the impact of in-person versus telephone-based genetic counseling on genetic testing uptake. This study addresses limitations of previous research by examining predictors from multiple domains in a population-based sample of cancer cases considered at increased risk for hereditary breast and ovarian cancer (HBOC) who received genetic counseling as part of a clinical trial and by evaluating patient perceptions of counseling in relation to testing decisions. Results are intended to inform the literature and care delivery on predictors of testing uptake in breast and ovarian cancer survivors at increased risk and advance research on who may be more or less likely to test if they receive in-person versus telephone-based counseling.

Hypotheses were guided by the Extended Parallel Process Model (EPPM; ref. 11) and the Ottawa Decision Support Framework (ODSF; ref. 12). When applied to genetic testing, the EPPM suggests that survivors who believe they are at increased risk for hereditary cancers may undergo genetic testing if they believe testing can help them manage risk. Therefore, we hypothesized that women with higher perceived risk of carrying a mutation, with more precounseling cancer-specific distress, and higher perceived personal control would be more likely to test, adjusting for other factors such as perceived cost barriers.

In the context of genetic counseling, the ODSF (12) suggests that patient-centered counseling helps patients assimilate complex information about testing, cancer risk, and medical management, addresses patient concerns, and increases personal control (13–17). We hypothesized that higher counselor patient-centeredness would predict higher testing uptake. We examined counseling-related variables and psychosocial variables as moderators of in-person versus telephone-based counseling based on prior literature suggesting differential responses to counseling methods (18–20).

Trial design and oversight

We describe the REACH trial in detail elsewhere (5, 10). Briefly, REACH was a family cluster randomized, two-group, equivalency/noninferiority trial comparing in-person to telephone counseling for women at high risk for BRCA1/2 mutations (NCT01346761). Random assignment began in August 2010 and ended in September 2012 with one-year outcome assessments completed in February 2014. This study was conducted in accordance with the ethical guidelines outlined in the Belmont Report. The University of Utah and University of New Mexico Institutional Review Boards approved the study. All participants provided informed consent.

Participants

We used the Utah Population Database (21) to identify breast and ovarian cancer patients who were then recruited through the Utah Cancer Registry. Eligible women were between the ages of 25 and 74 years, living in Utah with telephone access, had a personal or family history suggestive of HBOC meeting National Comprehensive Cancer Network criteria for a referral to genetic counseling at the time of the study and were willing to travel to in-person counseling at the nearest of 14 clinics from 8 rural and 6 urban areas. Women with prior genetic counseling and/or BRCA1/2 testing were excluded. Only women who completed pretest genetic counseling and had known testing information were included in this analysis to allow for tests of genetic counseling variables (see Fig. 1, CONSORT).

Figure 1.

Consolidated Standards of Reporting Trials (CONSORT) flowchart. GC, genetic counseling; HBOC, hereditary breast and ovarian cancer. CONSORT flow diagram depicting study recruitment and retention.

Figure 1.

Consolidated Standards of Reporting Trials (CONSORT) flowchart. GC, genetic counseling; HBOC, hereditary breast and ovarian cancer. CONSORT flow diagram depicting study recruitment and retention.

Close modal

The interventions were telephone versus in-person genetic counseling [see Kinney and colleagues (5) for details on counseling sessions]. Counseling was delivered by five board-certified genetic counselors using a semistructured protocol, counselor checklists, and standard visual aids. Visual aids were used during the counseling sessions in both study arms. Those who received telephone counseling were mailed visual aids to view during the session. Regardless of counseling mode, counselors explored the emotional aspects of testing, including the impact of a personal cancer history and anticipated reactions to test results [see Kinney and colleagues (5)]. Following telephone counseling, women who chose to have genetic testing were mailed a saliva genetic testing kit with instructions and were then offered posttest counseling with the same genetic counselor who conducted their initial counseling session. Women counseled in person who chose to have genetic testing were given the option of giving a sample immediately following in-person counseling or taking home a test kit to send a sample later. The cost of genetic testing was not covered by the study; however, if cost was identified as a barrier to testing during the counseling sessions, counselors discussed financial assistance programs. Women were instructed to contact counselors if they encountered financial barriers to testing. Pre- and posttest sessions were conducted by the same genetic counselor. Intervention fidelity was assessed with review of 10% of randomly selected audiotapes of counseling sessions.

Data collection

Assessments occurred at prerandomization (i.e., baseline), 1 week after genetic counseling, and 12 months after the last counseling session. Factors hypothesized to potentially influence how a participant perceived counseling were assessed at baseline. Factors that reflected the counseling process (e.g., patient-centeredness, education about the cost of testing) were assessed 1 week after counseling. Genetic testing that occurred within 12 months of counseling was chosen to allow for women who may have needed more time to make a decision.

Measures

Testing decision outcome.

Testing was defined as completion of testing within 12 months after the counseling session. The majority (92%) of the genetic testing was completed through the study with verified testing status through Myriad Laboratories.

Predictors.

Psychosocial factors included precounseling cancer-specific distress [Impact of Event Scale (IES); α = 0.90; ref. 22] and perceived risk of having a genetic mutation compared with other women their age (i.e., perceived comparative risk for a mutation; ref. 23). A score of 20 or above on the IES, which can range from 0–75, was considered clinically significant distress (22, 24). A risk score of 4 or 5 on the scale of 1–5 indicated that women perceived themselves to be at higher risk than other women their age.

Counseling factors were ascertained after the counseling session and included perceived personal control about the next steps, decisions, and information related to breast and/or ovarian cancer (Perceived Personal Control Questionnaire; α = 0.82; ref. 25), and patients' ratings of the counselor's patient-centered communication (an adapted 13-item scale reflecting perceptions of their counselor's informativeness, interpersonal sensitivity, and partnership building; α = 0.90; ref. 26).

Covariates.

Covariates were selected for evaluation based on literature review (7, 8, 19, 20).

Sociodemographic factors included self-reported age, education, income, employment, marital status, and race/ethnicity. Clinical factors included personal cancer history and number of relatives affected by breast and ovarian cancer. Accessibility factors included rural/urban residence (27) and cost barriers. Cost barriers to testing, including lack of insurance coverage for the test, lack of insurance, fear of losing employment, or fear of losing health or life insurance, were assessed one week after genetic counseling with two questions asking about the main and other reasons why they had not yet tested. Response options included: desire to not know BRCA status, concerns about ability to cope with cancer risk, lack of insurance coverage for the test, lack of insurance, fear of losing employment, and fear of losing health or life insurance. Participants also had the option of listing other barriers to testing. “Other” responses were double-coded (κ = 0.79; 95% CI, 0.72–0.86). Cost barriers were divided into short-term (e.g., “Insurance will not cover the cost of the test”) and long-term concerns (e.g., “Concerns about getting health or life insurance in the future”).

Psychosocial factors included precounseling state anxiety (Brief Symptom Inventory-18; α = 0.90; ref. 28), mental health (12-item SF-12v2TM Health Survey; α = 0.90; ref. 29), physical health (12-item SF-12v2TM Health Survey; α = 0.88; ref. 29), BRCA1/2 knowledge (10-item adapted index representing the number of items correctly answered; ref. 30), and post-counseling social support (Interpersonal Support Evaluation List; α = 0.76; ref. 31) and decisional conflict (Decisional Conflict Scale; α = 0.89; ref. 32).

Statistical analysis

We restricted analysis to participants with known testing status. Bivariate association between a variable and testing uptake was assessed using simple logistic regression model. A multivariable logistic regression model was fitted using backward variable selection for a parsimonious model of the independent predictors of genetic testing uptake. The cluster effect of family (randomization unit) was insignificant based on a generalized linear mixed effects model (P = 0.4). Therefore, all analyses were conducted without the random effect of cluster.

Variables had to have a significance of P < 0.10 to be retained in the model (33). We chose this P value to assure that we retained potentially significant variables and accounted for confounders (33). Collinearity concerns among variables were evaluated using variance inflation factors (VIF) approach and not identified (i.e., the VIFs ∼1). Bias-reducing penalized likelihood optimization was applied for logistic regression fitting (34). We included ORs per a half SD increase for continuous variables retained in the multivariable logistic model to aid their interpretation. For each putative moderator tested, an interaction of the counseling method (telephone vs. in-person) by the factor was added separately to the logistic regression model and tested. Moderators evaluated included cancer-specific distress, perceived risk of a genetic mutation compared to population risk, counselor patient-centeredness, cost barriers, decisional conflict, education, social support, and family history of cancer (18–20). Model goodness-of-fit was examined using the deviance test.

We conducted sensitivity analysis to compare results from those with known testing status (n = 781; per protocol) to those with known and unknown testing status (n = 877; intent-to-treat). We imputed 100 datasets to estimate genetic testing uptake among the 96 survivors with missing testing status using age, education, health care coverage status, breast cancer history, ovarian cancer history, baseline physical health, and baseline anxiety. All analyses were conducted using SAS version 9.4 (SAS Institute Inc.).

Participants had a personal history of breast cancer (n = 715; 91.5%), ovarian cancer (n = 63, 8.1%), or both cancers (n = 3, 0.4%). The average age was 56.6 years (SD = 7.7 years). The majority had health insurance (n = 761, 97.4%) and an established health care provider (n = 697, 89.2%). The majority of the sample was non-Hispanic white (n = 738, 94.5%). A total of 241 (30.9%) women obtained testing. Cost represented 38.9% (n = 142) of the primary barriers to testing cited one week after genetic counseling. No one cited concern about losing employment as a primary barrier to testing, although 12 people cited concerns about losing health or life insurance or future insurability for themselves or their family as a primary barrier. Long-term cost concerns (which included concerns related to future insurability) were less common than out-of-pocket cost concerns (e.g., lack of insurance coverage for the test) that were endorsed by 17.7% of survivors (Table 1). Long-term cost concerns did not differ by income level (χ2 = 2.30; df = 3; P = 0.51). Short-term cost concerns differed by income level (χ2 = 16.15; df = 3; P = 0.001). Survivors earning $50–69 K reported higher short-term cost concerns. Those earning between dollar;30–49 K and those earning between dollar;50–69 K had similar health insurance coverage (χ2 = 0.14; df = 1; P = 0.712). Survivors earning at least dollar;50 K were more likely to have a college degree (χ2 = 56.21; df = 3; P < 0.0001).

Table 1.

Descriptive statistics and bivariate analysis of genetic testing uptake

(N = 781)
Primary variables of interestMean (SD)/N (%)OR (95% CI)aP
Precounseling cancer distressb 15.18 (14.24) 1.01 (1.00–1.02) 0.051 
Precounseling perceived comparative riskc 2.99 (1.10) 1.42 (1.23–1.63) <0.0001 
Perceptions of patient centeredness 4.25 (0.48) 0.80 (0.58–1.11) 0.179 
Perceived personal control 1.55 (0.36) 1.14 (0.74–1.77) 0.542 
Covariates 
Decisional conflict 27.90 (13.39) 0.97 (0.96–0.98) <0.0001 
Social support 8.53 (1.91) 0.93 (0.86–1.00) 0.061 
Age 56.58 (7.71) 1.02 (1.00–1.04) 0.063 
Anxiety 2.65 (2.99) 1.00 (0.95–1.05) 0.920 
Physical health 49.79 (9.05) 0.99 (0.97–1.00) 0.106 
Mental health 51.25 (8.92) 0.99 (0.97–1.01) 0.307 
Knowledge 7.05 (1.44) 1.08 (0.97–1.21) 0.153 
Intervention 
 In-person (ref.) 379 (48.53)   
 Telephone 402 (51.47) 0.68 (0.50–0.92) 0.013 
Short-term cost concerns 
 Yes (ref.) 138 (17.67)   
 No 643 (82.33) 17.46 (6.71–45.43) <0.0001 
Long-term cost concerns 
 Yes (ref.) 27 (3.46)   
 No 754 (96.54) 8.26 (1.54–44.43) 0.014 
Relatives with breast or ovarian cancer 
 0 FDR/SDR (ref.) 411 (52.62)  0.0001 
 1 FDR/SDR 216 (27.66) 1.55 (1.08–2.22)  
 ≥2 FDR/SDR 154 (19.72) 2.27 (1.54–3.35)  
Residence    
 Urban (ref.) 658 (84.25)  0.019 
 Rural 123 (15.75) 1.61 (1.08–2.40)  
Marital status    
 Not married (ref.) 151 (19.33)  0.773 
 Married 630 (80.67) 1.06 (0.72–1.56)  
Education    
 High school or less (ref.) 166 (21.25)  0.939 
 Some college 287 (36.75) 1.00 (0.66–1.51)  
 Bachelor's or more 328 (42.00) 0.95 (0.63–1.42)  
Income (dollar;)    
 ≤29,999 (ref.) 86 (11.44)  0.153 
 30,000–49,999 148 (19.68) 0.68 (0.38–1.20)  
 50,000–69,999 145 (19.28) 0.57 (0.32–1.01)  
 ≥70,000 373 (49.60) 0.86 (0.53–1.41)  
(N = 781)
Primary variables of interestMean (SD)/N (%)OR (95% CI)aP
Precounseling cancer distressb 15.18 (14.24) 1.01 (1.00–1.02) 0.051 
Precounseling perceived comparative riskc 2.99 (1.10) 1.42 (1.23–1.63) <0.0001 
Perceptions of patient centeredness 4.25 (0.48) 0.80 (0.58–1.11) 0.179 
Perceived personal control 1.55 (0.36) 1.14 (0.74–1.77) 0.542 
Covariates 
Decisional conflict 27.90 (13.39) 0.97 (0.96–0.98) <0.0001 
Social support 8.53 (1.91) 0.93 (0.86–1.00) 0.061 
Age 56.58 (7.71) 1.02 (1.00–1.04) 0.063 
Anxiety 2.65 (2.99) 1.00 (0.95–1.05) 0.920 
Physical health 49.79 (9.05) 0.99 (0.97–1.00) 0.106 
Mental health 51.25 (8.92) 0.99 (0.97–1.01) 0.307 
Knowledge 7.05 (1.44) 1.08 (0.97–1.21) 0.153 
Intervention 
 In-person (ref.) 379 (48.53)   
 Telephone 402 (51.47) 0.68 (0.50–0.92) 0.013 
Short-term cost concerns 
 Yes (ref.) 138 (17.67)   
 No 643 (82.33) 17.46 (6.71–45.43) <0.0001 
Long-term cost concerns 
 Yes (ref.) 27 (3.46)   
 No 754 (96.54) 8.26 (1.54–44.43) 0.014 
Relatives with breast or ovarian cancer 
 0 FDR/SDR (ref.) 411 (52.62)  0.0001 
 1 FDR/SDR 216 (27.66) 1.55 (1.08–2.22)  
 ≥2 FDR/SDR 154 (19.72) 2.27 (1.54–3.35)  
Residence    
 Urban (ref.) 658 (84.25)  0.019 
 Rural 123 (15.75) 1.61 (1.08–2.40)  
Marital status    
 Not married (ref.) 151 (19.33)  0.773 
 Married 630 (80.67) 1.06 (0.72–1.56)  
Education    
 High school or less (ref.) 166 (21.25)  0.939 
 Some college 287 (36.75) 1.00 (0.66–1.51)  
 Bachelor's or more 328 (42.00) 0.95 (0.63–1.42)  
Income (dollar;)    
 ≤29,999 (ref.) 86 (11.44)  0.153 
 30,000–49,999 148 (19.68) 0.68 (0.38–1.20)  
 50,000–69,999 145 (19.28) 0.57 (0.32–1.01)  
 ≥70,000 373 (49.60) 0.86 (0.53–1.41)  

Abbreviations: FDR, first-degree relative; ref, reference group; SDR, second-degree relative.

aORs per one-point increase on continuous scales.

bObserved scores ranged 0–63 on scale of 0–75.

cScale ranged from 1–5; 3 = perceived “the same” risk as other women their age.

Independent focal predictors of testing uptake

After adjusting for other independent variables, testing uptake was associated with higher cancer-specific distress (OR = 1.01; 95% CI, 1.002–1.03; equivalent to OR0.5 SD = 1.11; 95% CI0.5 SD = 1.01–1.21) and perceived comparative risk (OR = 1.32; 95% CI, 1.11–1.57; equivalent to OR0.5 SD = 1.16; 95% CI0.5 SD = 1.06–1.28; Table 2). Cancer-specific distress and perceived comparative risk were not considerably linearly correlated (r = 0.08; 95% CI, 0.01–0.15).

Table 2.

Multivariable predictors of genetic testing uptake

(N = 781a)
Primary variables of interestMean (SD)/N (%)OR (95% CI)bOR0.5 SDc (95% CI0.5 SD)P
Precounseling cancer distress 15.18 (14.24) 1.01 (1.002–1.03) 1.11 (1.01–1.21) 0.026 
Precounseling perceived comparative risk 2.99 (1.10) 1.32 (1.11–1.57) 1.16 (1.06–1.28) 0.002 
Perceptions of patient-centeredness 4.25 (0.48) 0.45 (0.30–0.68) 0.83 (0.75–0.91) 0.0001 
Covariates 
Residence 
 Urban (ref.) 658 (84.25)   0.057 
 Rural 123 (15.75) 1.57 (0.99–2.49)   
Relatives with breast or ovarian cancer 
 0 FDR/SDR (ref.) 411 (52.62)   0.010 
 1 FDR/SDR 216 (27.66) 1.45 (0.96–2.21)   
 ≥2 FDR/SDR 154 (19.72) 2.09 (1.29–3.40)   
Intervention 
 In-person (ref.) 379 (48.53)    
 Telephone 402 (51.47) 0.53 (0.37–0.76)  0.0005 
Short-term cost concerns 
 Yes (ref.) 138 (17.67)    
 No 643 (82.33) 18.73 (7.09–49.46)  <0.0001 
Long-term cost concerns 
 Yes (ref.) 27 (3.46)    
 No 754 (96.54) 6.64 (1.15–38.41)  0.034 
Social support 8.53 (1.91) 0.90 (0.82–0.99) 0.91 (0.83–0.99) 0.035 
Decisional conflict 27.9 (13.39) 0.95 (0.94–0.97) 0.72 (0.65–0.80) <0.0001 
(N = 781a)
Primary variables of interestMean (SD)/N (%)OR (95% CI)bOR0.5 SDc (95% CI0.5 SD)P
Precounseling cancer distress 15.18 (14.24) 1.01 (1.002–1.03) 1.11 (1.01–1.21) 0.026 
Precounseling perceived comparative risk 2.99 (1.10) 1.32 (1.11–1.57) 1.16 (1.06–1.28) 0.002 
Perceptions of patient-centeredness 4.25 (0.48) 0.45 (0.30–0.68) 0.83 (0.75–0.91) 0.0001 
Covariates 
Residence 
 Urban (ref.) 658 (84.25)   0.057 
 Rural 123 (15.75) 1.57 (0.99–2.49)   
Relatives with breast or ovarian cancer 
 0 FDR/SDR (ref.) 411 (52.62)   0.010 
 1 FDR/SDR 216 (27.66) 1.45 (0.96–2.21)   
 ≥2 FDR/SDR 154 (19.72) 2.09 (1.29–3.40)   
Intervention 
 In-person (ref.) 379 (48.53)    
 Telephone 402 (51.47) 0.53 (0.37–0.76)  0.0005 
Short-term cost concerns 
 Yes (ref.) 138 (17.67)    
 No 643 (82.33) 18.73 (7.09–49.46)  <0.0001 
Long-term cost concerns 
 Yes (ref.) 27 (3.46)    
 No 754 (96.54) 6.64 (1.15–38.41)  0.034 
Social support 8.53 (1.91) 0.90 (0.82–0.99) 0.91 (0.83–0.99) 0.035 
Decisional conflict 27.9 (13.39) 0.95 (0.94–0.97) 0.72 (0.65–0.80) <0.0001 

Abbreviations: FDR, first-degree relative; ref., reference group; SDR, second-degree relative.

aModel shows estimates after adjustment for other variables.

bORs per one-point increase on the scale for continuous variables.

cEquivalent ORs and 95% CI per 0.50 SD increase in continuous variables; provided to aid clinical interpretation. Categorical ORs and 95% CI estimates remain the same as footnote “b”.

The goodness of fit of this model indicated adequate model fit (P = 0.43).

Testing was also independently associated with lower ratings of counselor patient-centeredness (OR = 0.45; 95% CI, 0.30–0.68; equivalent to OR0.5 SD = 0.83; 95% CI0.5 SD = 0.75–0.91). Cost barriers exhibited the largest effect on testing uptake. Testing uptake was independently associated with the absence of short-term cost barriers (OR = 18.73; 95% CI, 7.09–49.46) and long-term cost barriers (OR = 6.64; 95% CI, 1.15–38.41; Table 2).

Results with imputed data were consistent with results from known testing outcomes.

Moderators of counseling

Precounseling cancer distress and perceived comparative risk of an inherited gene mutation modified the effect of the type of counseling and completion of genetic testing (P = 0.003, P = 0.038, respectively, for the interaction effect). For clinical interpretation, precounseling cancer-specific distress was plotted at values indicative of clinically significant distress (20 or above) and perceived comparative risk of a genetic mutation was plotted at values indicating high versus average or lower perceived risk.

Among survivors at clinically significant levels of cancer distress, 26.3% of survivors who received telephone-counseling tested compared with 44.3% of those who received in-person counseling (OR = 0.45; 95% CI, 0.27–0.76; Fig. 2). In contrast, among those with lower levels of distress, 27.0% tested after telephone-counseling compared with 30% after in-person counseling (OR = 0.86; 95% CI, 0.59–1.26; Fig. 2). Among survivors who believed that they were at higher risk for a genetic mutation compared with other women their age, 33.9% of those who received telephone-counseling tested compared with 50.5% of those who received in-person counseling (OR = 0.50; 95% CI, 0.29–0.87; Fig. 3). In contrast, among survivors who believed they had the same or lower risk for a mutation compared with other women their age, 23.8% tested who received telephone counseling compared with 29.8% who received in-person counseling (OR = 0.74; 95% CI, 0.51–1.07; Fig. 3).

Figure 2.

Clinically significant precounseling cancer distress modifies counseling method's effect on testing uptake. *, statistically significant difference observed. A score of 20 or above was considered clinically significant. Among survivors at clinically significant levels of cancer distress, 26.3% of survivors who received telephone counseling tested compared with 44.3% of those who received in-person counseling (OR = 0.45; 95% CI, 0.27–0.76).

Figure 2.

Clinically significant precounseling cancer distress modifies counseling method's effect on testing uptake. *, statistically significant difference observed. A score of 20 or above was considered clinically significant. Among survivors at clinically significant levels of cancer distress, 26.3% of survivors who received telephone counseling tested compared with 44.3% of those who received in-person counseling (OR = 0.45; 95% CI, 0.27–0.76).

Close modal
Figure 3.

High baseline precounseling perceived comparative risk of inherited gene mutation modifies counseling method's effect on testing uptake. *, statistically significant difference observed. Among survivors who believed that they were at higher risk for a genetic mutation compared with other women their age, 33.9% of those who received telephone counseling tested compared with 50.5% of those who received in-person counseling (OR = 0.50; 95% CI, 0.29–0.87).

Figure 3.

High baseline precounseling perceived comparative risk of inherited gene mutation modifies counseling method's effect on testing uptake. *, statistically significant difference observed. Among survivors who believed that they were at higher risk for a genetic mutation compared with other women their age, 33.9% of those who received telephone counseling tested compared with 50.5% of those who received in-person counseling (OR = 0.50; 95% CI, 0.29–0.87).

Close modal

Education was of marginal significance as a moderator of counseling mode and genetic testing uptake (P = 0.055). Those with a high school education or lower appeared less likely to test if they received telephone-counseling compared with in-person counseling (OR = 0.23; 95% CI, 0.11–0.50), whereas testing uptake by counseling mode did not differ among those with some college (OR = 0.66; 95% CI, 0.37–1.17) or a bachelor's degree or more (OR = 0.68; 95% CI, 0.40–1.17). No other moderators were significant, including family cancer history which was moderately correlated (polyserial correlation = 0.45; P < 0.001) with precounseling perceived comparative risk for a genetic mutation.

This article examined predictors of genetic testing uptake among a population-based sample of breast and ovarian cancer survivors at increased risk for HBOC who received either telephone or in-person genetic counseling as part of a randomized controlled trial. Our data show that women without cost barriers to testing are more likely to complete genetic testing. Findings suggest that survivors entering counseling with high cancer distress or high perceived comparative risk for a mutation may respond differently to telephone versus in-person counseling, and indicate that more patient-centered counseling may yield lower testing uptake.

From a policy perspective, results highlight the importance of addressing cost barriers to testing. Long-term cost concerns did not differ by income. Short-term cost concerns were similar across incomes with the exception of those earning between dollar;50,000–dollar;69,000 a year who were more likely to report short-term cost concerns as a barrier. This income bracket and above was more likely to be college educated, but had similar insurance coverage as those earning dollar;30,000–dollar;49,000. They may have been an educated group of women more used to consuming health information compared with those earning less than them who were less likely to have a college degree. Thus, financial strain from testing and indicated procedures may have been a more salient barrier to testing compared to other factors such as its perceived lack of benefit for managing health.

At the time our study was conducted, the cost of BRCA sequencing was over dollar;3,000 (35). The cost of testing has dropped due to the Supreme Court's ruling on BRCA patenting (36). BRCA sequencing is now available for less than dollar;300 (37). Most insurance companies cover the cost of BRCA testing for populations indicated by the National Comprehensive Cancer Network guidelines (38). Many genetic testing laboratories provide financial assistance programs for uninsured and underinsured individuals. However, coverage varies for multi-gene panel testing (39) and for preventative services (e.g., prophylactic surgery, surveillance magnetic resonance imaging) following a positive test, which places some survivors at risk of being able to afford the cost of a test itself, but unable to afford and benefit from the interventions indicated by the test (40). These concerns are compounded by the uncertain future of the Affordable Care Act, the repeal of which would disparately affect subgroups of the population with known disparities accessing and utilizing genetic services and associated preventative services (16, 41). Yet a priority of the national Precision Medicine Initiative is to address insurance coverage to enable access to genomic discoveries and guideline-concordant genetic care delivery (42).

This study has implications for the practice of genetic counseling. Our moderation analysis results suggest that survivors who enter counseling with high levels of cancer-specific distress or high perceived comparative risk for a mutation will be more likely to test if they receive in-person counseling and less likely to test if they receive telephone counseling. Counselor response to nonverbal cues of distress exhibited by survivors may have magnified the effects of precounseling cancer-specific distress and perceived comparative risk for a mutation by prompting more discussion about the perceived threat of cancer and carrying a mutation. In contrast, unaddressed nonverbal cues of distress may have impeded information exchange or discussion of distress as a barrier to testing in telephone-counseling. Research should evaluate patient–counselor interactions [e.g., linguistic analysis for counselor and survivor affective word use (43), session length, survivor question frequency, etc.] to understand the function of distress and perceived comparative risk in these counseling delivery methods.

Moderation results suggest that survivors with a high school education or less may be less likely to test if they receive telephone counseling than if they receive in-person counseling. This result must be interpreted cautiously, as it was of only borderline statistical significance. The logistics of sending in a sample may have been more of a barrier for this subgroup. Alternatively, even though visual aids were used in both in-person and telephone-based counseling, those with less education in telephone counseling may have had more difficulty understanding material presented or had undetected questions about the information. Future research should examine whether those with lower educational attainment and health literacy show smaller knowledge gains after telephone counseling than in-person counseling. Additional materials (e.g., computer-based modules) to supplement telephone counseling or additional assistance overcoming logistical barriers to testing may be useful for those with less education.

The relation between higher patient-centeredness ratings and lower testing uptake in this study was unexpected. However, patient-centered care focused on the unique needs and preferences of a survivor given their family history, risk perception, knowledge, and stated concerns. Thus, a patient-centered approach may have yielded higher or lower testing uptake to the extent that counseling reflected a patient's readiness for testing. Survivors who were ambivalent about testing may have felt supported by their counselor to not rush a testing decision. In contrast, those who were further in the decision process may have received more directive counseling that addressed specific barriers. Similarly, a patient-centered focus might have allowed survivors at lower risk to feel comfortable declining testing. Research on the patient–counselor interaction is needed to better understand this counterintuitive result.

Limitations

Generalizability is limited by population characteristics of mostly non-Hispanic white women with relatively high income and education levels from one geographic area. Ethnic minorities and those with lower socioeconomic status face substantial disparities in genetic counseling and testing and may have additional factors that influence testing decisions (3). We did not assess the extent to which survivors discussed genetic counseling and testing with their family, friends, or health care providers, or the nature of those interactions. Finally, this study was conducted before the use of multigene panel testing. Certain psychosocial factors (e.g., concern about discovering unexpected positive results and secondary findings) may be more salient with adoption of multigene panel testing and whole genome sequencing.

Conclusion

The study's findings suggest that cost concerns are critical in testing decisions among survivors at risk for HBOC who receive genetic counseling. Our findings further suggest that telephone counseling, while improving access to qualified genetic counselors, may yield lower testing uptake among survivors who enter counseling with clinically significant cancer-specific distress or high perceived comparative risk for a mutation. In-person counseling may be indicated for survivors with elevated cancer distress or high perceived comparative risk for a mutation. Screening for psychosocial concerns, including distress, could help allocate in-person resources. Alternatively, counselors may consider the use of videotechnology or consider screening for psychosocial issues, including distress, prior to delivering telephone-based counseling to inform counseling (44). Research should examine if there are differential effects of video-enabled versus audio alone technologies among diverse populations and counselor–patient interactions during genetic counseling sessions to better understand the effect of cancer distress, perceived comparative risk for mutations, and counselor–patient centeredness during counseling.

A. Gammon is a consultant/advisory board member for Invitae. No potential conflicts of interests were disclosed by the other authors.

This content is solely the responsibility of the authors and does not necessarily reflect the opinions or views of the funding and supporting agencies, and the National Heart, Lung, and Blood Institute.

Conception and design: L.E. Steffen, W.K. Kohlmann, M.D. Schwartz, A.Y. Kinney

Development of methodology: A.M. Stroup, R.A. Campo, A.Y. Kinney

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Gammon, W.K. Kohlmann, S.S. Buys, A.M. Stroup, R.A. Campo, A.Y. Kinney

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L.E. Steffen, R. Du, J.S. Mandelblatt, W.K. Kohlmann, J.-H. Lee, A.M. Stroup, A.Y. Kinney

Writing, review, and/or revision of the manuscript: L.E. Steffen, R. Du, A. Gammon, J.S. Mandelblatt, W.K. Kohlmann, J.-H. Lee, S.S. Buys, A.M. Stroup, R.A. Campo, K.G. Flores, B. Vicuña, M.D. Schwartz, A.Y. Kinney

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.M. Stroup, K.G. Flores

Study supervision: A.M. Stroup, R.A. Campo, A.Y. Kinney

Other (principal investigator): A.Y. Kinney

This work was supported by grants from the National Cancer Institute at the NIH (R01CA129142; principal investigator: A.Y. Kinney; L.E. Steffen supported by R25CA122061; principal investigator: N.E. Avis) and the Huntsman Cancer Foundation. The project was also supported by the Shared Resources (P30 CA042014) at Huntsman Cancer Institute (Biostatistics and Research Design, Genetic Counseling, Research Informatics, and the Utah Population Database) and the Biostatistics Shared Resource and Developmental Funds, University of New Mexico Comprehensive Cancer Center (P30CA118100; principal investigator: C.L. Willman); the Utah Cancer Registry, which is funded by contract no. HHSN261201000026C from the National Cancer Institute's Surveillance, Epidemiology and End Results Program with additional support from the Utah State Department of Health and the University of Utah, the National Center for Research Resources, and the National Center for Advancing Translational Sciences, NIH, through grant 8UL1TR000105 (formerly UL1RR025764).

We would like to thank the REACH Project participants who made this research possible; Roger Edwards, BS, programmer/analyst; and Sandra Edwards, research associate. In addition, we would like to thank the study coordinators, Amy Rogers and Madison Briggs, as well as the staff and administrators at the community clinics and the Utah Cancer Registry.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Daly
MB
,
Pilarski
R
,
Axilbund
JE
,
Berry
M
,
Buys
SS
,
Crawford
B
, et al
Genetic/familial high-risk assessment: breast and ovarian, version 2.2015
.
J Natl Compr Canc Netw
2016
;
14
:
153
62
.
2.
Ropka
ME
,
Wenzel
J
,
Phillips
EK
,
Siadaty
M
,
Philbrick
JT
. 
Uptake rates for breast cancer genetic testing: a systematic review
.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
840
55
.
3.
Levy
DE
,
Byfield
SD
,
Comstock
CB
,
Garber
JE
,
Syngal
S
,
Crown
WH
, et al
Underutilization of BRCA1/2 testing to guide breast cancer treatment: Black and Hispanic women particularly at risk
.
Genet Med
2011
;
13
:
349
55
.
4.
Kurian
AW
,
Griffith
KA
,
Hamilton
AS
,
Ward
KC
,
Morrow
M
,
Katz
SJ
, et al
Genetic testing and counseling among patients with newly diagnosed breast cancer
.
JAMA
2017
;
317
:
531
4
.
5.
Kinney
AY
,
Butler
KM
,
Schwartz
MD
,
Mandelblatt
JS
,
Boucher
KM
,
Pappas
LM
, et al
Expanding access to BRCA1/2 genetic counseling with telephone delivery: A cluster randomized trial
.
J Natl Cancer Inst
2014
;
106
:
djv328
.
6.
Cragun
D
,
Weidner
A
,
Lewis
C
,
Bonner
D
,
Kim
J
,
Vadaparampil
ST
, et al
Racial disparities in BRCA testing and cancer risk management across a population-based sample of young breast cancer survivors
.
Cancer
2017
;
123
:
2497
505
.
7.
Sweeny
K
,
Ghane
A
,
Legg
AM
,
Huynh
HP
,
Andrews
SE
. 
Predictors of genetic testing decisions: a systematic review and critique of the literature
.
J Genet Couns
2014
;
23
:
263
88
.
8.
Cragun
D
,
Bonner
D
,
Kim
J
,
Akbari
MR
,
Narod
SA
,
Gomez-Fuego
A
, et al
Factors associated with genetic counseling and BRCA testing in a population-based sample of young Black women with breast cancer
.
Breast Cancer Res Treat
2015
;
151
:
169
76
.
9.
Cohen
SA
,
Marvin
ML
,
Riley
BD
,
Vig
HS
,
Rousseau
JA
,
Gustafson
SL
. 
Identification of genetic counseling service delivery models in practice: a report from the NSGC Service Delivery Model Task Force
.
J Genet Couns
2013
;
22
:
411
21
.
10.
Kinney
AY
,
Steffen
LE
,
Brumbach
BH
,
Kohlmann
W
,
Du
R
,
Lee
JH
, et al
Randomized noninferiority trial of telephone delivery of BRCA1/2 genetic counseling compared with in-person counseling: 1-Year follow-up
.
J Clin Oncol
2016
;
34
:
2914
24
.
11.
Maloney
EK
,
Lapinski
MK
,
Witte
K
. 
Fear appeals and persuasion: a review and update of the Extended Parallel Process Model
.
Social and Personality Psychology Compass
2011
;
5
:
206
19
.
12.
Durand
M-A
,
Stiel
M
,
Boivin
J
,
Elwyn
G
. 
Where is the theory? Evaluating the theoretical frameworks described in decision support technologies
.
Patient Educ Couns
2008
;
71
:
125
35
.
13.
O'Connor
AM
,
Jacobsen
MJ
,
Stacey
D
. 
An evidence-based approach to managing women's decisional conflict
.
J Obstet Gynecol Neonatal Nurs
2002
;
31
:
570
81
.
14.
Ormond
KE
. 
From genetic counseling to “genomic counseling.”
Mol Genet Genomic Med
2013
;
1
:
189
93
.
15.
Pieterse
AH
,
Ausems
MGEM
,
Van Dulmen
AM
,
Beemer
FA
,
Bensing
JM
. 
Initial cancer genetic counseling consultation: change in counselees' cognitions and anxiety, and association with addressing their needs and preferences
.
Am J Med Genet Assoc
2005
;
137A
:
27
35
.
16.
Cragun
D
,
Kinney
AY
,
Pal
T
. 
Care delivery considerations for widespread and equitable implementation of inherited cancer predisposition testing
.
Expert Rev Mol Diagn
2017
;
17
:
57
70
.
17.
Skirton
H
. 
The client's perspective of genetic counseling-a grounded theory study
.
J Genet Couns
2001
;
10
:
311
29
.
18.
Anderson
B
,
McLosky
J
,
Wasilevich
E
,
Lyon-Callo
S
,
Duquette
D
,
Copeland
G
. 
Barriers and facilitators for utilization of genetic counseling and risk assessment services in young female breast cancer survivors
.
J Cancer Epidemiol
2012
;
2012
:
11
.
19.
Butrick
M
,
Kelly
S
,
Peshkin
BN
,
Luta
G
,
Nusbaum
R
,
Hooker
GW
, et al
Disparities in uptake of BRCA1/2 genetic testing in a randomized trial of telephone counseling
.
Genet Med
2015
;
17
:
467
75
.
20.
Gooding
HC
,
Organista
K
,
Burack
J
,
Biesecker
BB
. 
Genetic susceptibility testing from a stress and coping perspective
.
Soc Sci Med
2006
;
62
:
1880
90
.
21.
Cannon-Albright
LA
,
Thomas
A
,
Goldgar
DE
,
Gholami
K
,
Rowe
K
,
Jacobsen
M
, et al
Familiality of cancer in Utah
.
Cancer Res
1994
;
54
:
2378
85
.
22.
Horowitz
M
,
Wilner
N
,
Alvarez
W
. 
Impact of Event Scale: a measure of subjective stress
.
Psychosom Med
1979
;
41
:
209
18
.
23.
Gurmankin Levy
A
,
Shea
J
,
Williams
SV
,
Quistberg
A
,
Armstrong
K
. 
Measuring perceptions of breast cancer risk
.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
1893
8
.
24.
Bleiker
EMA
,
Pouwer
F
,
van der Ploeg
HM
,
Leer
J-WH
,
Adèr
HJ
. 
Psychological distress two years after diagnosis of breast cancer: frequency and prediction
.
Patient Educ Couns
2000
;
40
:
209
17
.
25.
Smets
EM
,
Pieterse
AH
,
Aalfs
CM
,
Ausems
MG
,
van Dulmen
AM
. 
The Perceived Personal Control (PPC) questionnaire as an outcome of genetic counseling: reliability and validity of the instrument
.
Am J Med Genet A
2006
;
140
:
843
50
.
26.
Epstein
RM
,
Street
RL Jr
. 
Patient-centered communication in cancer care: Promoting healing and reducing suffering
;
National Cancer Institute, NIH Publication No. 07-6225.
2007
.
Available from: https://healthcaredelivery.cancer.gov/pcc/monograph.html
.
27.
University of Washington Rural Health Research Center
. 
Rural-urban commuting area codes (Version 2.0)
; 
2006
.
Available from:
http://depts.washington.edu/uwruca/.
28.
Derogatis
LR
.
Brief Symptom Inventory 18: administration, scoring, and procedures manual
.
Minneapolis, MN
:
NCS Pearson, Inc.
; 
2000
.
29.
Ware
J
 Jr
,
Kosinski
M
,
Keller
SD
. 
A 12-Item Short-Form Health Survey: Construction of scales and preliminary tests of reliability and validity
.
Med Care
1996
;
34
:
220
33
.
30.
Wang
C
,
Gonzalez
R
,
Milliron
KJ
,
Strecher
VJ
,
Merajver
SD
. 
Genetic counseling for BRCA1/2: a randomized controlled trial of two strategies to facilitate the education and counseling process
.
Am J Med Genet A
2005
;
134a
:
66
73
.
31.
Cohen
S
,
Mermelstein
R
,
Kamarck
T
,
Hoberman
HM
. 
Measuring the functional components of social support
.
In:
Sarason
IG
,
Sarason
BR
,
editors
.
Social support: theory, research and applications
.
Dordrecht, the Netherlands
:
Springer
; 
1985
.
p.
73
94
.
32.
O'Connor
AM
. 
Validation of a decisional conflict scale
.
Med Decis Making
1995
;
15
:
25
30
.
33.
Bursac
Z
,
Gauss
CH
,
Williams
DK
,
Hosmer
DW
. 
Purposeful selection of variables in logistic regression
.
Source Code Biol Med
2008
;
3
:
17
.
34.
Firth
D
. 
Bias reduction of maximum likelihood estimates
.
Biometrika
1993
;
80
:
27
38
.
35.
Chang
Y
,
Near
AM
,
Butler
KM
,
Hoeffken
A
,
Edwards
SL
,
Stroup
AM
, et al
ReCAP: Economic evaluation alongside a clinical trial of telephone versus in-person genetic counseling for BRCA1/2 mutations in geographically underserved areas
.
J Oncol Pract
2016
;
12
:
59
.
36.
Cook-Deegan
R
,
Niehaus
A
. 
After myriad: genetic testing in the wake of recent Supreme Court decisions about gene patents
.
Curr Genet Med Rep
2014
;
2
:
223
41
.
37.
Color Test
.
Available from:
https://getcolor.com/kit/how-it-works.
38.
Clain
E
,
Trosman
JR
,
Douglas
MP
,
Weldon
CB
,
Phillips
KA
. 
Availability and payer coverage of BRCA1/2 tests and gene panels
.
Nat Biotechnol
2015
;
33
:
900
2
.
39.
Curnutte
MA
,
Frumovitz
KL
,
Bollinger
JM
,
McGuire
AL
,
Kaufman
DJ
. 
Development of the clinical next-generation sequencing industry in a shifting policy climate
.
Nat Biotechnol
2014
;
32
:
980
2
.
40.
Prince
AER
. 
Prevention for those who can pay: insurance reimbursement of genetic-based preventive interventions in the liminal state between health and disease
.
J Law Biosci
2015
;
2
:
365
95
.
41.
Karaca-Mandic
P
,
Jena
AB
,
Ross
JS
. 
Health and health care use among individuals at risk to lose health insurance with repeal of the Affordable Care Act
.
JAMA Intern Med
2017
;
177
:
590
3
.
42.
Trosman
JR
,
Weldon
CB
,
Douglas
MP
,
Kurian
AW
,
Kelley
RK
,
Deverka
PA
, et al
Payer coverage for hereditary cancer panels: barriers, opportunities, and implications for the Precision Medicine Initiative
.
J Natl Compr Canc Netw
2017
;
15
:
219
28
.
43.
Kelly
KM
,
Ellington
L
,
Schoenberg
N
,
Jackson
T
,
Dickinson
S
,
Porter
K
, et al
Genetic counseling content: how does it impact health behavior?
J Behav Med
2015
;
38
:
766
76
.
44.
Eijzenga
W
,
Bleiker
EMA
,
Hahn
DEE
,
Kluijt
I
,
Sidharta
GN
,
Gundy
C
, et al
Psychosocial aspects of hereditary cancer (PAHC) questionnaire: development and testing of a screening questionnaire for use in clinical cancer genetics
.
Psychooncology
2014
;
23
:
862
9
.