A Systematic Review of Randomized Controlled Trials to Assess Outcomes of Genetic Counseling

Abstract

With the advancements in precision medicine and health care reform, it is critical that genetic counseling practice respond to emerging evidence to maximize client benefit. The objective of this review was to synthesize evidence on outcomes from randomized controlled trials (RCTs) of genetic counseling to inform clinical practice. Seven databases were searched in conducting this review. Studies were selected for inclusion if they were: (a) RCTs published from 1990 to 2015, and (b) assessed a direct outcome of genetic counseling. Extracted data included study population, aims, and outcomes. Risk of bias was evaluated using the Cochrane Handbook for Systematic Reviews of Interventions guidelines. A review of 1654 abstracts identified 58 publications of 54 unique RCTs that met inclusion criteria, the vast majority of which were conducted in cancer genetic counseling setting. Twenty-seven publications assessed ‘enhancements’ to genetic counseling, and 31 publications compared delivery modes. The methodological rigor varied considerably, highlighting the need for attention to quality criteria in RCT design. While most studies assessed several client outcomes hypothesized to be affected by genetic counseling (e.g., psychological wellbeing, knowledge, perceived risk, patient satisfaction), disparate validated and reliable scales and other assessments were often used to evaluate the same outcome(s). This limits opportunity to compare findings across studies. While RCTs of genetic counseling demonstrate enhanced client outcomes in a number of studies and pave the way to evidence-based practice, the heterogeneity of the research questions suggest an important need for more complementary studies with consistent outcome assessments.

Introduction

In the eras of precision and genomic medicine, it is critical that research on genetic counseling outcomes excels, and that client outcomes are used to inform evidence-based practice. Data from randomized controlled trials (RCTs) can inform design of interventions to improve clinical care. The sheer number of published RCTs is testament to advances in genetic counseling research quality. To enhance practice, recognition of existing evidence supporting effective interventions is critical, and championing consistent use of complementary research design and assessment is needed. Applying well-founded research findings to clinical practice can advance the practice of genetic counseling and inform development of advanced practice guidelines.

Randomized Controlled Trials

RCTs are designed to answer research questions that directly affect clinical practice, treatment interventions, and professional guidelines. The design randomly assigns participants to mutually exclusive groups to assess the efficacy of distinct interventions while controlling for variables across groups. This direct comparison with minimal effect of confounding variables can lead to empirically robust findings. However, not all RCTs are created equal, and methodological rigor still needs to be evaluated to interpret the strength of findings (Sibbald and Roland 1998). Ideally, these experiments should be replicated to affirm or dispute original findings and thus, more fully answer the original research question. Data from RCTs top of the hierarchy of evidence, second only to systematic literature reviews (Barton 2000). The high-quality evidence from RCTs can be used to design downstream intervention research and to inform clinical practice. Evidence-based practices are more effective, and associated with increased positive patient outcomes (Kitson et al. 1998; Youngblut and Brooten 2001).

Genetic Counseling Outcomes

Identifying clinically relevant and measurable genetic counseling client outcomes is essential to assessing the effectiveness of clinical practice (Biesecker 2001). Bernhardt et al. (2000) elicited genetic counseling goals from experienced genetic counselors and a patient sample. Counselors cited education and provision of psychosocial support as primary goals, whereas patients described attending genetic counseling with few preconceived goals: patients reported being pleasantly surprised by the counseling, and described the experience as a positive interpersonal interaction and connection (Bernhardt et al. 2000). Long-term outcomes valued by patients were anticipatory guidance about the condition in their family and improved family communication.

In an effort to develop an assessment tool, McAllister et al. (2011) queried genetics patients about their perceptions of service outcomes. Patient perceptions of benefits led to design of the Genetic Counseling Outcome Scale. While the scale is populated with items representing outcomes of medical genetics services, a number of items on this scale such as “I can make decisions about the condition that may change my child(ren)’s future / the future of any child(ren) I may have” and “I can see that good things have come from having this condition in my family” represent feasible outcomes of genetic counseling.

Recent efforts to catalogue key genetic counseling outcomes elicited from focus groups with genetic counselors reflect early evidence of the genetic counseling outcomes from the provider’s viewpoint (Redlinger-Grosse et al. 2016; Zierhut et al. 2016). Yet, these listings include both patient outcomes and professional competencies. Patient outcomes address benefits of genetic counseling: increasing understanding, enhancing coping and adaptation, facilitating informed choice, and mitigating health risks. Alongside them appear outcomes that address the expertise of genetic counselors: providing accurate information, ordering appropriate testing, interpreting risks accurately, recommending appropriate management, and reducing inappropriate use of services. Certainly, clients benefit from competent providers; however, health services outcomes are primarily used to ensure professional standards, justify services, and demonstrate cost effectiveness. Practice outcomes are of paramount importance in assessing patient benefit from genetic counseling, and RCTs offer a rigorous method to control for the effects of confounding variables on those outcomes. Our systematic literature review yields key patient outcomes for further investigation across clinical settings.

While direct queries from patients and providers are useful in generating outcomes, particularly when the outcomes differ between these groups, a more objective source of intended outcomes is evidence from experimental studies. Using well-defined and endorsed genetic counseling outcomes, studies can achieve greater consistency in measurement when assessing the effectiveness of genetic counseling. Use of standard patient outcomes further allows for cross-comparison of findings and ultimately meta-analyses. Such efforts can provide the evidence upon which practice can be improved to enhance patient-centered care. The primary objective of this systematic review is to summarize 25 years of published RCTs that highlight genetic counseling client outcomes by synthesizing a narrative of enhanced client outcomes, assessing the quality of data and relevance for clinical practice, and if feasible, conducting a meta-analysis of outcome data. Secondary goals are to promote professional consensus on desirable outcomes for genetic counseling clients (and valid, reliable methods for assessing them), and to identify key research questions to enhance client outcomes that require further study. This project represents the first major systematic review of RCTs of genetic counseling, and strives to inform clinical practice and future study design.

Methods

Search Strategy

Reports of original research studies, systematic and process reviews on outcomes of genetic counseling were reviewed. Publications in peer-reviewed journals from 1990 to 2015 were identified in a comprehensive search via PubMed, Embase, PsycINFO, CINAHL, Cochrane Library, CRD Database and National Research Register searches. Database searches were initially conducted using the keyword combinations including ‘genetic counseling outcomes,’ ‘genetic counseling goals,’ ‘genetic counseling evidence,’ and ‘genetic testing’ with related search terms. Upon further assessment, an updated search was conducted on 01/10/16 using the original keyword combinations as well as ‘information,’ ‘education,’ and related search terms.¹ For the purpose of this review, only randomized controlled trials using patient outcomes of genetic counseling were included.

Selection Criteria

After a preliminary review of RCTs assessing genetic counseling outcomes, we elected to employ broad selection criteria to maximize the literature reviewed for eligibility. Our inclusion criteria dictated English-language publications employing a randomized controlled trial with at least one primary study aim referencing an outcome of genetic counseling. Studies that were not peer-reviewed original research or were published outside of the last 25 years were excluded.

Study Selection

We reviewed and evaluated each abstract in the literature search using the selection criteria and disagreements were resolved by discussion. All authors were involved in the selection of appropriate publications. Each selected publication was then assessed for design, methodology and outcomes.

Data Extraction

Information regarding study population, aims, and outcomes were gleaned from each study. Many studies reported multiple outcomes among the following: psychological wellbeing, knowledge, perceived risk, satisfaction, decisional quality, information sharing, and informed choice. A Cochrane narrative synthesis of outcomes was conducted, where main findings were summarized.

Data Collection and Analysis

Risk of Bias

Risk of bias was assessed using the guidelines from two sets of criteria. First, methodologies were evaluated using the Cochrane Handbook for Systematic Reviews of Interventions (Higgins and Green 2011), in which a set of quality assessment aims were designated as “low risk” if the aim was fully met, “medium risk” if the aim was partially met, or “high risk” if the aim was not met. An aim was evaluated as unclear if there was not enough information provided for assessment. A study was considered to have an overall low risk of bias if “low risk” was assigned to four out of six of the categories. Second, methodologies were evaluated using the Updated Method Guidelines for Systematic Reviews in the Cochrane Collaboration Back Review Group (Van Tulder et al. 2003), by which articles were considered low risk if they adequately fulfilled six out of the 11 categories represented.

Results

One thousand six hundred fifty-four citations were returned by our set of search queries, 1563 of which were excluded based upon the information contained in the abstract. Ultimately, 58 publications describing 54 unique RCTs met inclusion criteria; the selection process is depicted in Fig. 1. The studies included in this review vary widely, investigating a broad range of questions in genetic counseling. Due to dissimilarity in setting, inclusion criteria, and importantly, outcome characterization and assessment, a meta-analysis was not feasible. However, comparisons among similar studies provide initial evidence to guide practice, specifically those investigating delivery modes. Studies included in the review are summarized in Table 1 (RCTs comparing modes of delivery) and Table 2 (RCTs comparing counseling strategies) in terms of sample characteristics, objectives, and major findings.

Fig. 1.

A visual representation of the process of selecting the 58 publications (reporting on 54 unique RCTs) included in this review from the 1654 abstracts returned by the literature search

Table 1.

Summary of studies focused on a comparison of modes of delivery

	Title	Samplea: N, Age, Characteristics	Aim	Finding Highlights (Selected Outcomes: Psychological Wellbeing [A], Knowledge [B], Perceived Risk [C], Satisfaction [D], Genetic Testing/Screening Intentions [E] Genetic Testing Uptake [F], Decision Quality [G], Medical Management / Health Behavior [H], Sharing of Information [I], & Informed Choice [J]
Temme et at (2015)	1. Assessment of parental understanding of positive newborn screening results and carrier sta tus for Cystic Fibrosis with the use of a short educational video	96 participants (mean 29.4y) who have a child with a positive newborn screening for Cystic Fibrosis (CF) and one identified CFTR mutation, but have no other children with CF	Evaluate the impact of augmenting genetic counseling with a short educational video in terms of parents’ knowledge about the genetics of CF and their child’s carrier status.	Participants in the video group had significantly higher post-counseling knowledge scores compared to control, but both groups achieved higher knowledge scores compared to baseline. [B] Post-counseling knowledge was maintained 6 weeks following the appointment. [B] Participants who sought out resources prior to counseling had higher pre-counseling knowledge scores. [B]
Butrick et at (2015)	2. Disparities in uptake of BRCA1/2 genetic testing in a randomized trial of telephone counseling	669 self-referred or physician-referred women (mean 48y) seeking BRCA1/2 genetic counseling	Compare in-person to telephone-based counseling by a genetic counselor in terms of genetic testing uptake among women at high risk for hereditary breast/ovarian cancer.	Completion of counseling and testing was associated with race/ethnicity, perceived stress, and knowledge. [B, F] Participants randomized to telephone counseling were significantly less likely to be tested. [F] Minority women receiving telephone counseling were least likely to complete testing. [F]
Buchanan et at (2015)	3. Randomized trial of telegenetics vs. in-person cancer genetic counseling: Cost, patient satisfaction and attendance	162 participants (mean 51.3y) referred to cancer genetic counseling in four rural oncology clinics	Compare telegenetics to in-person genetic counseling in terms of cost, satisfaction, and attendance.	There were no differences between groups in terms of knowledge of breast cancer genetics, patient satisfaction, or testing uptake. [B, D, F] Total costs were $106 per telegenetics patient and $244 per in-person patient. In-person patients were significantly more likely to attend cancer genetic counseling than telegenetics patients (89% vs. 79%).
Yee et al. (2014)	4. A randomized trial of a prenatal genetic testing interactive computerized information aid	150 women (mean 26.6y) with a gestational age 6–26wk who had not yet received any prenatal testing	Evaluate the impact of augmenting provider-based counseling with an interactive computer program on patients’ knowledge about genetic screening and diagnostic concepts.	Women in the intervention group had a significantly higher knowledge score compared to the control group on the initial evaluation, as well as on the second questionnaire 2 to 4 weeks later. [B] All women who experienced it found the education tool helpful. A majority reporting feeling the tool was clear (96%).
Wevers et at (2014)	5. Impact of rapid genetic counselling and testing on the decision to undergo immediate or delayed prophylactic mastectomy in newly diagnosed breast cancer patients: Findings from a randomised controlled trial	265 women newly diagnosed with breast cancer (mean age at diagnosis 44.9y) and ≥10% BRCA1/2 mutation risk in the Netherlands	Compare rapid genetic counselling and testing (RGCT; result disclosure within 4 weeks) to routine testing (result disclosure after 4 months or more) provided by a clinical geneticist in terms of choice of surgery.	Intention-to-treat analyses indicated no significant differences in surgery uptake between the RGCT and control groups. [H] However, per-protocol analysis indicated that patients receiving genetic test results before surgery were significantly more likely to undergo bilateral mastectomy than patients who received usual care. [H]
Schwartz et al (2014)	6. Randomized noninferiority trial of telephone versus in-person genetic counseling for hereditaiy breast and ovarian cancer	669 women (mean 48y) with a minimum 10% risk for BRCA1/2 mutation and without a newly diagnosed or metastatic cancer	Compare telephone delivery to in-person delivery of BRCA1/2 counseling by a genetic counselor.	Telephone counseling was noninferior to in-person delivery in terms of knowledge, perceived stress, and satisfaction at 2 weeks and 3 months post-counseling. [A, B, D] Noninferiority of telephone counseling was not established in terms of cancer distress, BRCA1/2 testing uptake, or decision conflict. [A, F, G] The cost of telephone counseling was $114.40 less per patient.
Lowery et al. (2014)	7. A randomized trial to increase colonoscopy screening in members of high-risk families in the colorectal cancer family registry and cancer genetics network	632 men and women (≥2 ly) unaffected but at high risk for colorectal cancer (CRC) due to family history	Compare a tailored telephone-based counseling intervention delivered by trained interviewers to a mailed packet in terms of adherence to colonoscopy screening.	Baseline factors significantly associated with greater adherence included having had a colonoscopy, screening intentions, appropriate CRC risk perception. [H] The telephone counseling intervention was associated with a 32% increase in screening adherence; no significant change was observed in the group receiving the mailed intervention. [H] Many of the participants (47%) were adherent to CRC screening at baseline, and the majority reported intentions to receive a colonoscopy within the next 2 years. [E]
Kinney et al. (2014)	8. Expanding access to BRCA1/2 genetic counseling with telephone delivery: A cluster randomized trial	901 women (mean 56. ly) who were breast or ovarian cancer patients and at-risk females of relatives positive for BRCA1/2 mutations	Compare telephone to in-person counseling by a genetic counselor in terms of testing uptake and changes in psychosocial measures.	Telephone counseling was found to be noninferior to in-person counseling in terms of anxiety, cancer-related distress, knowledge, and decision quality. [A, B, G] A smaller proportion of women in the telephone counseling group (21.8%) than in the in-person counseling group (31.8%) elected to have genetic testing. [F] Testing uptake was higher for rural than urban dwellers, but the differences were not statistically significant. [F]
Hag a et al (2014)	9. Impact of delivery models on understanding genomic risk for type 2 diabetes	300 non-diabetic participants (44% aged 18–29y) recruited from Durham, NC	Compare an online delivery model to in-person counseling by a genetic counselor in terms of understanding genomic risk and adopting healthy behaviors.	92% of participants expressed no regret after receiving results, and participants expressed very low levels of distress. [A] Participants who received results in-person were more likely than those who received results online to correctly interpret and report their genomic risk at both the 1 -week and 3-month follow-ups. [C] There were no differences in correct reporting of overall lifetime risk between groups at either 1-week or 3-month follow-ups. [C] Genetic knowledge did not impact comprehension when comparing delivery methods.
Green et al (2014)	10. A randomized noninferiority trial of condensed protocols for genetic risk disclosure of Alzheimer’s disease	343 asymptomatic adults (mean 58.3y) with a first-degree relative affected by Alzheimer’s disease and unaffected by severe anxiety or depression	Compare standard protocols (SP-GC) with condensed protocols delivered by genetic counselors (CP-GC) and physicians (CP-MD) for disclosing APOE genotype and concordant risk of Alzheimer’s disease.	CP-GC was noninferior to SP-GC in regards to anxiety and depression at all time points, and to test-related distress at 6 and 12 months. [A] CP-MD was noninferior to SP-GC in regards to anxiety at all time points, and to depression at 12 months. [A] Results were similar between European Americans and African Americans.
Hanprasertpong et al (2013)	11. Compa rison of the effectiveness of different counseling methods before second trimester genetic amniocentesis in Thailand	321 pregnant women (mean 37.7y) scheduled for their second trimester genetic amniocentesis because of maternal age > 35y in Thailand	Compare computer-assisted instruction (CAI) to leaflet self-reading (LSR) before individual counseling with a senior midwife in a developing country.	Anxiety decreased and satisfaction increased in both CAI and LSR groups, but no difference was found between the two groups. [A, D] Knowledge increased significantly after CAI and LSR, and increased further after individual counseling. [B] A significantly greater improvement in overall and basic knowledge scores was observed in LSR compared to CAI after both stages. [B] Method of counseling did not affect testing uptake. [F] Pain was decreased significantly after CAI and LSR, with the CAI group showing a larger decrease.
Platten et al (2012)	12. The use of telephone in genetic counseling versus in-person counseling: A randomized study on counselees’ outcome	215 patients (mean 45y) seeking oncogenetic counseling in Stockholm County, Sweden	Compare oncogenetic nurse counseling conducted via telephone to traditional in-person counseling.	No differences in cancer worry, patient satisfaction, or health-related quality of life were observed between groups. [A, D] A high overall satisfaction rate was demonstrated in both groups, but a considerable number of participants reported dissatisfaction with information on surveillance and prevention. [D]
Hooker et al (2011)	13. Longitudinal changes in patient distress following interactive decision aid use among BRCA1/2 earners: A randomized trial	214 women (25–75y) with a positive BRCA1/2 mutation test but no metastatic breast or ovarian cancer	Evaluate the impact of augmenting breast cancer risk management counseling by a genetic counselor with subsequent provision of a computer-based, interactive decision aid.	The usual care (UC) group reported decreased cancer-specific distress at the 1-month follow-up; both groups reported similar decreases in cancer-specific distress at 12 months. [A] Participants who actually viewed the decision aid had lower genetic testing-specific distress than the UC group at 12 months. [A] The decision aid had a significant longitudinal impact (which varied over time) on cancer-specific distress and genetic testing-specific distress. [A]
Castellani et al (2011)	14. An interactive computer program can effectively educate potential users of cystic fibrosis carrier tests	44 participants (mean 37.9y) from infertile couples having CTFR mutation analysis as part of an assisted reproduction protocol in Italy	Compare a computer-based, educational tool with traditional counseling sessions by a trained geneticist for CF carrier testing in terms of knowledge and understanding.	There were significant increases in correct answers and decreases in unanswered questions in both the computer program and traditional counseling groups. [B] No participants changed their intention to undergo testing after the education sessions. [E]
Hwa et al (2010)	15. Informed consent for antenatal serum screening for Down syndrome	193 women (mean 31y) at 15–2 lwk gestation in Taiwan	Evaluate the impact of providing a trained research counselor with results from a questionnaire assessing patients’ understanding of Down syndrome.	There were no significant differences between groups in anxiety or depression. [A] Members of the group whose practitioner received results from the questionnaire before the genetic counseling session had higher levels of understanding of the disease nature of Down syndrome and the testing itself. [B]
Bowen and Powers (2010)	16. Effects of a mail and telephone intervention on breast health behaviors	1510 women (18–74y) not previously diagnosed with cancer	Evaluate the impact of a mail and telephone intervention to improve breast health behavior while maintaining quality of life for women with a family history (delivered by a genetic counselor) and without (delivered by a health counselor).	Quality of life increased in the intervention group compared to the delayed treatment control group. [A] Mammography/breast self-examination use increased and interest in genetic testing decreased in the intervention group to a significantly greater degree than in the control group. [H]
Schwartz et al (2009)	17. Randomized trial of a decision aid for BRCA1/BRCA2 mutation carriers: Impact on measures of decision making and satisfaction	214 women (mean 44y) with a positive BRCA1/2 mutation test result	Evaluate the impact of augmenting standard genetic counseling with a computer-based interactive decision aid in helping women with a BRCA1/2 mutation consider their options for risk management and maximize satisfaction.	The decision aid did not affect the rate of risk-reducing mastectomy, but it did affect the timing: women in the usual care (UC) group were more likely to have obtained a risk-reducing mastectomy earlier. [H] The decision aid group experienced less decisional conflict and more decisional satisfaction than the UC group for participants who were undecided at the time of randomization, but not for those who had previously decided. [D, G] The decision aid group had an easier time coming to a final management decision than the UC group.
Jenkins et al (2007)	18. Randomized comparison of phone versus in-person BRCA1/2 predisposition genetic test result disclosure counseling	111 men and women (median 45y) with BRCA1/2 mutation or a breast/ovarian cancer diagnosis	Compare phone to in-person disclosure of results by advance-practice nurses for BRCA1/2 predisposition genetic testing.	The in-person group had a greater decrease in state anxiety than the telephone group 1 week after disclosure, but not 3 months later. [A] There were no differences in change in knowledge or in general satisfaction between the groups. [B, D] 90% of those who received results by phone indicated that they preferred this method of disclosure, while only 4% indicated that they preferred in-person disclosure.
Helmes et al (2006)	19. Results of a randomized study of telephone versus in-person breast cancer risk counseling	340 women (mean 41y) with no personal history of breast/ovarian cancer	Compare in-person counseling and telephone counseling by a genetic counselor to control (no counseling).	Risk perception, cancer worry, and interest in genetic testing decreased in the experimental groups. [A, C, E] There were no notable differences in measured outcomes between the two types of counseling in the experimental groups.
Bowen et al (2006)	20. Effects of counseling Ashkenazi Jewish women about breast cancer risk	221 Ashkenazi Jewish women (mean 47y) with no personal history of breast cancer	Compare individual genetic counseling and psychosocial group counseling to control in terms of breast cancer worry, risk perception, and interest in genetic testing.	Women in experimental arms had more accurate perceived risk, lower cancer worry, and less interest in receiving genetic testing. [A, C, E] No significant differences were observed between groups receiving different methods of counseling.
Wang et al (2005)	21. Genetic counseling for BRCA1/2: A randomized controlled trial of two strategies to facilitate the education and counseling process	197 women (mean 44.5y) from the Breast and Ovarian Cancer Risk Evaluation Program (BOCREP)	Evaluate the impact of augmenting standard genetic counseling with two strategies: a CD-ROM computer program for patients and a feedback checklist for the genetic counselor.	Women who viewed the CD-ROM experienced slightly greater declines in worry compared to those who did not, though the change was not significantly different between groups. [A] Feedback to the genetic counselor resulted in greater gains in knowledge of genetics and breast cancer. [B] Overall, women who viewed the CD-ROM were less likely to undergo genetic testing, and spent less time with the genetic counselor than those who did not view the CD-ROM. [F]
Hunter et al. (2005)	22. A randomized trial comparing alternative approaches to prenatal diagnosis counseling in advanced maternal age patients	352 women (mean 36.8y) of advanced maternal age with no family history of genetic disorders in Canada	Compare individual genetic counseling, group genetic counseling, and a decision aid (explained by a genetic counselor) regarding prenatal diagnosis for women of advanced maternal age.	Anxiety levels did not change from pre- to post-counseling. [A] Patient knowledge was highest in those who were counseled in a group of 2–3 women and their partners. [B] All patients reported high levels of satisfaction, but it was highest in those who were counseled individually. [D] Decisional conflict was lowest in those who received counseling via an audiotape-booklet decision aid. [G]
Green et al (2005)	23. Use of an educational computer program before genetic counseling for breast cancer susceptibility: Effects on duration and content of counseling sessions	211 women (mean 44y) referred to a genetic counselor for evaluation of breast cancer risk	Evaluate the impact of augmenting standard genetic counseling with a computer-based decision aid for educating women about BRCA1 and BRCA2 genetic testing.	Access of the computer-based decision aid resulted in significantly shorter counseling sessions among women at low risk for carrying BRCA1/2 mutations. Counselors indicated that the clients’ use of the computer- based decision aid affected the way they used time, shifting focus from basic education toward personal risk and decision-making.
Calzone et al (2005)	24. Randomized comparison of group versus individual genetic education and counseling for familial breast and/or ovarian cancer	142 men and women (25–68y) with a known BRCA1/2 mutation or a breast/ovarian cancer diagnosis	Compare pretest education and counseling for breast cancer by advanced-practice nurses in a group setting to that provided on an individual basis by the same nurses.	No significant differences were found between groups in psychological distress or patient satisfaction. [A, D] Knowledge increased significantly from baseline in both groups, but no differences were found between groups. [B] 96% of patients opted to undergo genetic testing after counseling, with no differences between groups. [F]
Green et al (2004)	25. Effect of a computer-based decision aid on knowledge, perceptions, and intentions about genetic testing for breast cancer susceptibility	211 women (mean 44y) referred for genetic counseling due to personal or family history of breast cancer	Evaluate the impact of augmenting standard genetic counseling with a computer-based decision aid for educating women about BRCA1 and BRCA2 genetic testing.	Anxiety decreased in the standard care group among low- and high-risk participants, but did not change in the computer group. [A] Knowledge increased in both groups, but the increase in knowledge was greater in the computer group than the standard care group among low-risk women. [B] Participants’ perception of their absolute risk of developing breast cancer decreased significantly in both groups. [C] There was no significant difference in decisional conflict between groups, and participants rated both interventions highly. [D, G] The percentage of women reporting testing intentions declined among low- but not high-risk women in both groups. [E]
Bowen et al (2004)	26. Breast cancer risk counseling improves women’s functioning	354 women (mean 42y) with a family history of breast cancer	Compare individual genetic counseling (a 2-h session) and psychosocial group counseling (4 2-h sessions with 4–6 others) to control.	Cancer-specific worry and perceived risk were decreased in both counseling arms compared to the control (delayed intervention) at follow-up; there were no differences between groups. [A, C] Baseline anxiety scores were a significant predictor of change in perceived risk from baseline to follow-up. [A, C]
Bowen et al (2002)	27. Effects of risk counseling on interest in breast cancer genetic testing for lower risk women	317 women (mean 42.4y) with some familial history of breast cancer but no indications of autosomal dominant genetic mutation	Compare individual genetic counseling and psychosocial group counseling to a control condition in terms of interest in genetic testing.	76% of the sample expressed interest in genetic testing at baseline. [E] By the end of the survey, 71 % of women in the counseling group were still interested in testing while only 68% of controls were interested. [E] Changes in beliefs about genetic testing altered the effects of counseling interventions.
Schwartz et al (2001)	28. Impact of educational print materials on knowledge, attitudes, and interest in BRCA1/BRCA2 testing among Ashkenazi Jewish women	391 Ashkenazi Jewish women (18–83y) without a family history suggestive of hereditary breast- ovarian carcinoma	Compare a brief educational booklet on BRCA1/BRCA2 testing to one on signs and symptoms of breast carcinoma in terms of knowledge, attitudes, and interest in testing.	The genetic testing education (GTE) intervention led to significant increases in knowledge and perceived importance of the cons of undergoing genetic testing. [B] Contrary to the hypotheses, the GTE intervention did not lead to significant decreases in perceived importance of the pros of undergoing genetic testing nor in testing intentions. [E]
Green et al (2001a)	29. Education about genetic testing for breast cancer susceptibility: Patient preferences for a computer program or genetic counselor	72 women (mean 40y) with a first-degree relative diagnosed with breast cancer	Compare face-to-face education and counseling by a genetic counselor to education by an interactive computer program.	The majority of participants indicated they’d prefer a genetic counselor to a computer program, though they also indicated that they still liked the computer for nearly all the measured tasks. [D]
Green et al (2001b)	30. An interactive computer program can effectively educate patients about genetic testing for breast cancer susceptibility	72 women (mean 44y) with a first-degree relative diagnosed with breast cancer	Compare face-to-face education and counseling by a genetic counselor with education by an interactive computer program.	Knowledge scores were greater for those who participated in educational sessions than in a control group, with no significant difference between the computer and counselor groups. [B] Testing intentions were reduced for participants after attending educational sessions. [E]
Miedzybrodzka et al (1995)	31. Antenatal screening for carriers of cystic fibrosis: Randomized trial of stepwise v couple screening	2002 women (mean 27.7y) with no family history of cystic fibrosis recruited from an antenatal clinic in Scotland	Compare stepwise to couple approaches in antenatal carrier screening for cystic fibrosis (CF) by a clinical geneticist.	There was no difference in anxiety between groups after delivery, and anxiety levels were lower than at recruitment. Most women and their partners remembered some facts about CF and carrier status. [A, B] There was no significant difference in the perception of the baby’s risk of cystic fibrosis between the two methods. [C] Uptake of screening was 90% for both approaches. [F] All 35 carriers in the step-wise arm who completed the questionnaires had communicated their test result to at least 1 relative. [I]

^aUnless otherwise noted in this Sample column, the study was conducted in the United States of America

Table 2.

Summary of studies focused on a comparison of counseling strategies

	Title	Samplea: N, Age, Characteristics	Aim	Finding Highlights (Selected Outcomes: Psychological Wellbeing [A], Knowledge [B], Perceived Risk [C], Satisfaction [D], Genetic Testing/Screening Intentions [E] Genetic Testing Uptake [F], Decision Quality [G], Medical Management / Health Behavior [H], Sharing of Information [I], & Informed Choice [J]
Hodgson et al (2015)	32. Outcomes of a randomised controlled trial of a complex genetic counselling intervention to improve family communication	95 probands or parents of probands (mean 47.6y) recruited from genetics clinics in six public hospitals in Victoria, Australia	Evaluate the impact of augmenting standard care with three phone calls from a genetic counselor in terms of family communication about a new genetic diagnosis.	25.6% of intervention group relatives compared with 20.9% of control group relatives made contact with genetic services. [E] Although no major difference was detected overall between the intervention and control groups, there was more contact in the intervention group where the genetic condition conferred a high risk to offspring. [I]
Eijzenga et al (2015)	33. Routine assessment of psychosocial problems after cancer genetic counseling: Results from a randomized controlled trial	118 individuals (mean 49.5y) who underwent diagnostic or pre-symptomatic DNA testing and had their final counseling session prior to 12/15/12 in the Netherlands	Evaluate the impact of augmenting genetic counseling with provision of information about the counselees’ psychosocial problems to genetic counselors as part of a telephone session held 1 month after cancer genetic counseling and testing (CGCT) procedure.	There were no significant differences between groups in regards to cancer worry or distress, problem management, prevalence of CGCT-related problems over time, satisfaction levels, or frequency with which psychosocial issues were discussed. [A, D] Both the intervention group (93%) and the control group (90%) were highly satisfied with the overall genetic counseling process. [D] Counselors’ awareness of problems was higher in the intervention group, but only significantly with “practical issues.” One-third of participants would recommend that all counselees have a session, while 55% would recommend it for only those who express an interest.
Albada et al (2015)	34. Follow-up effects of a tailored pre-counseling website with question prompt in breast cancer genetic counseling	197 women (mean 41.35y) who were the first in their families to seek breast cancer genetic counseling in the Netherlands	Evaluate the immediate and long-term effects of a pre-counseling website with question prompts on patients receiving cancer genetic counseling.	No significant effects were found in regards to recall, knowledge, risk perception, anxiety, cancer worry, or adherence to breast surveillance advice. [A, B, C, H, E] Counselees in the intervention group were significantly more satisfied, had more positive experiences with counseling, and had higher perceived personal control 1 year post-counseling. [A, D]
Nishigaki et al (2014)	35. The effect of genetic counseling for adult offspring of patients with type 2 diabetes on attitudes toward diabetes and its heredity: A randomized controlled trial	216 individuals (mean 45.9y) with a first-degree relative with type 2 diabetes but no personal diagnosis in Tokyo, Japan	Compare a lifestyle intervention with counseling by a genetic counselor (GC + LI) and without (LI) to control (no intervention) in terms of attitudes toward diabetes and its prevention.	More than half of participants indicated feeling anxiety about their potential for developing diabetes, but there were no significant differences between intervention groups. [A] Approximately 40% of participants perceived that their diabetes risk was higher than normal. [C] Significantly more subjects perceived that diabetes onset is controllable in the GC + LI group than in the LI group at both 1-week and 1-year post counseling.
Kuppermann et al (2014)	36. Effect of enhanced information, values clarification, and removal of financial barriers on use of prenatal genetic testing: A randomized clinical trial	710 women (mean 29.25y) who were ≤20wk gestation and had not undergone any prenatal testing for fetal aneuploidy	Evaluate the impact of a decision- support guide and elimination of financial barriers to testing on decision making and prenatal genetic testing uptake.	Knowledge scores were higher in the intervention group. [B] Age-adjusted risk for carrying a fetus with trisomy 21 and risk of miscarriage via amniocentesis were more likely to be correctly reported in the intervention group. [C] Significantly fewer women in the intervention group underwent invasive prenatal diagnostic testing. [F] There were no significant differences found in regards to decisional conflict and decisional regret. [G]
Eijzenga et al (2014)	37. Effect of routine assessment on specific psychosocial problems on personalized communication, counselors’ awareness, and distress levels in cancer genetic counseling practice: A randomized controlled trial	246 participants (mean 48y) referred to genetic counseling at two family cancer clinics in the Netherlands	Evaluate the impact of augmenting genetic counseling with provision of results from the Psychosocial Aspects of Hereditary Cancer (PAHC) as part of a telephone session held 1 month after cancer genetic counseling and testing procedure.	The intervention group had significantly lower levels of cancer worries and distress after counseling. [A] No significant difference was observed in patient satisfaction. [D] Self-reported use of psychosocial services was not statistically significant between groups. The counselors’ awareness of psychosocial problems was significantly increased in the intervention group, as was the frequency with which psychosocial problems were discussed. Counselors who received the patient questionnaire initiated more discussion of psychosocial problems, while the length of the session remained unaffected.
Montgomery et al (2013)	38. Preparing individuals to communicate genetic test results to their relatives: Report of a randomized control trial	249 women (mean 48.5y) who had BRCA1/2 testing with at least one first-degree relative with whom to share results	Compare the efficacy of a communication skills-building intervention to wellness education in preparing individuals to communicate results to family members.	There were no differences in distress about sharing results. [A] While there were no significant differences between groups in percentage disclosing results, probands were more likely to share with children, females, and relatives expected to have a favorable opinion about learning results. [I] Sense of control was positively associated with disclosure. [I]
Henneman et al (2013)	39. The effectiveness of a graphical presentation in addition to a frequency format in the context of familial breast cancer risk communication: A multicenter controlled trial	410 unaffected women (mean 41y) with breast cancer family history attending the cancer clinic for the first time in the Netherlands	Evaluate the impact of augmenting the standard frequency format of lifetime breast cancer risk with a graphical presentation (10 × 10 human icons).	Risk estimation accuracy increased and cancer worry decreased in both groups, but no significant differences were found between groups. [A, C] For women with no or slightly increased risk of breast cancer, there were no differences in terms of cancer-prevention intentions (e.g., breast cancer screening or mammography) between the 2 groups. [E]
Voonvinden et al (2012)	40. The introduction of a choice to learn pre-symptomatic DNA test results for BRCA or Lynch syndrome either face-to-face or by letter	198 asymptomatic counselees (mean 44.6y) from families in the Netherlands with a known BRCA1/2 or Lynch syndrome mutation in the Netherlands	Compare traditional test result disclosure to a procedure in which counselees are offered a choice regarding how they receive their results.	The disclosure procedure had no influence on counselees’ emotional problems, cancer worry, knowledge, or risk perception. [A, B, C] The counselees with a favorable test result in the intervention group were the most satisfied. [D] The majority of cases chose for disclosure by letter.
Roberts et al (2012)	41. Effectiveness of a condensed protocol for disclosing APOE genotype and providing risk education for Alzheimer disease	264 participants (mean 58.ly) with only one first-degree relative affected with Alzheimer’s disease (onset ≥60y)	Compare a brief educational model to the standard extended protocol for describing Alzheimer’s disease risk related to APOE allele status.	Alzheimer’s disease knowledge increased from baseline in both groups. [B] No significant differences were observed between groups in patient recall of APOE status and disease risk, or in disease knowledge. [B,] The brief model took, on average, about half as much time as the extended model.
Roussi et al (2010)	42. Enhanced counseling for women undergoing BRCA1/2 testing: Impact on knowledge and psychological distress - Results from a randomized clinical trial	134 women (≥21y) with a family history suggestive of a hereditary pattern of breast and/or ovarian cancer	Evaluate the impact of augmenting standard genetic counseling with a 45-min, enhanced counseling intervention in role-play format from a health educator.	For women who tested positive, the intervention group had significantly lower distress levels 1 week post-disclosure. [A] Women who received the enhanced intervention were significantly more knowledgeable about the heritability of breast and ovarian cancer than those in the control group. [B]
Halbert et al (2010)	43. Effect of genetic counseling and testing for BRCA1 and BRCA2 mutations in African American women: A randomized trial	139 African American and/or black women (59% ≤50y) with ≥5% risk of BRCA1/2 mutation	Compare culturally tailored counseling to standard counseling (both provided by a genetic counselor) across a cohort of randomized and non-randomized women.	Test result acceptors had a greater likelihood of reporting decreased cancer worry. [A] Perceived risk decreased, though there was no difference in perceived risk between the two different types of counseling. [C] BRCA1/2 testing decisions did not have a significant effect on changes in risk perception. [C, F]
Graves et al (2010)	44. Randomized controlled trial of a psychosocial telephone counseling intervention in BRCA1 and BRCA2 mutation carriers	128 women (≥18y) who were BRCA1/2 mutation carriers without metastatic disease or recurrent ovarian cancer in Canada and the USA	Evaluate the impact of augmenting standard genetic counseling with a psychosocial counseling intervention by master’s-level mental health counselors consisting of 5 weekly telephone sessions and a mailed booklet.	Women in the intervention group reported greater reductions in anxiety, depression, and genetic testing distress at the 6-month follow-up compared to women receiving standard genetic counseling, but there were no differences between groups at the 12-month follow-up. [A]
Roshanai et al (2009)	45. Does enhanced information at cancer genetic counseling improve counselees’ knowledge, risk perception, satisfaction and negotiation of information to at-risk relatives? - A randomized study	147 men and women (median 56y) who did not suffer from any mental illnesses and read, wrote, and spoke Swedish	Evaluate the impact of augmenting genetic counseling with a subsequent session in which a specialist nurse offered extended cancer genetic information, help identifying important relatives, and an videotape from the counseling session.	Depression and anxiety decreased significantly over time, though there were no differences between groups at any time-point. [A] Knowledge increased significantly at 2-week and 8-month follow-ups, but no significant difference was found between groups. [B] The only significant difference in perceived risk between groups was observed immediately after counseling: the control group gave a more accurate estimation. [C] The only significant difference in satisfaction between groups was the intervention group’s greater satisfaction with the content. [D] Most participants (73%) had shared information with their all their relatives at the 8-month follow-up. [I]
Wakefield et al (2008)	46. A randomized trial of a breast/ovarian cancer genetic testing decision aid used as a communication aid during genetic counseling	148 women (mean 49y) who approached a familial cancer clinic within a two-year period in Australia	Evaluate the impact of a decision aid given during genetic counseling for women considering genetic testing for breast/ovarian cancer risk.	There were no significant differences between groups in anxiety, depression, or psychological distress. [A] Women who received the decision aid had higher levels of knowledge, and felt more informed about genetic testing. [B] No differences were observed in testing uptake, decision quality, or sharing information with family members. [F, G, I] The effect of the decision aid on the measure of informed choice was not significant. [J]
Glanz et al (2007)	47. Effects of colon cancer risk counseling for first-degree relatives	176 siblings and children (mean 54.4y) of colorectal cancer patients living in Hawaii	Compare culturally sensitive Colon Cancer Risk Counseling (CCRC) to General Health Counseling (GHC) by trained health or nurse educators for relatives of colorectal cancer patients.	Knowledge and perceived risk showed change related to treatment group, but were not significant mediators of the treatment effect on adherence. [B, C] The CCRC intervention resulted in a 13% greater increase in screening adherence than GHC at 4 months, but the treatment effect was only 11 % greater at 12 months. [H]
Bennett et al (2007)	48. A randomized controlled trial of a brief self-help coping intervention designed to reduce distress when awaiting genetic risk information	162 women (most >40y) referred to a specific breast and ovarian cancer service in Wales	Evaluate the impact of a distraction-based coping leaflet in women undergoing genetic risk assessment for breast/ovarian cancer.	Among women highly distressed at baseline, only those in the intervention group reported significant reductions in distress. [A] While 95% of participants reported having read the leaflet, significantly fewer reporting having used the distraction strategies exemplified within (but may have used other distraction strategies).
Matloff et al (2006)	49. Healthy women with a family history of breast cancer: Impact of a tailored genetic counseling intervention on risk perception, knowledge, and menopausal therapy decision making	64 women (mean 49y) with at least one first-degree relative with breast cancer	Evaluate the impact of a personalized risk assessment and intervention by a genetic counselor in a cohort of low- and moderate-risk women.	Worry about breast cancer, heart disease, and osteoporosis was correlated with perceived risk, but not actual risk. [A, C] Women in the intervention group gained more knowledge about menopause and menopausal therapy. [B] Participants in the intervention group had lower perceived risk, and reported being very satisfied with the counseling. [C, D] There were no differences between groups in terms of medication usage (hormone therapy, tamoxifen, or raloxifene). [H] Perceived effective decision making was higher for the intervention group at the 6-month follow-up. [G]
Charles et al (2006)	50. Satisfaction with genetic counseling for BRCA1 and BRCA2 mutations among African American women	54 African American women (mean 46y) with moderate to high risk for BRCA1/2 mutation	Compare culturally tailored genetic counseling (CTGC) to standard genetic counseling (SGC), both provided by a genetic counselor.	Women who reported annual household incomes less than $35 K were significantly more likely to report lessened worries than higher income women, and those receiving CTGC were significantly more likely than those receiving SGC to report lessened worry. [A] 96% reported that they were very satisfied with the counseling. [D]
Miller et al. (2005)	51. Enhanced counseling for women undergoing BRCA1/2 testing: Impact on subsequent decision making about risk reduction behaviors	77 high-risk women (most ≤50y) undergoing BRCA1/2 testing	Evaluate the impact of an enhanced counseling intervention by a health educator designed to promote well-informed decision making for risk reduction options for ovarian cancer.	One week after disclosure, women in the enhanced counseling group experienced a greater reduction in avoidant ideation. [A] At the 6-month follow-up, women in the intervention group reported seeking out more information about prophylactic oophorectomy, and were more likely to have actually undergone preventive surgery. [H]
Braithwaite et al (2005)	52. Development of a risk assessment tool for women with a family history of breast cancer	72 women (≥18y) with a first- or second-degree relative with breast cancer in the UK	Compare the genetic risk assessment in the clinical environment (GRACE) prototype to standard risk counseling from a clinical nurse specialist.	There were no significant differences between groups in risk perception or cancer-related worry. [A, C] Patient satisfaction was higher in the group receiving nurse counseling than in the GRACE product group. [D]
Brain et al (2005)	53. An exploratory comparison of genetic counselling protocols for HNPCC predictive testing	14 women and 12 men (mean 41y) currently unaffected individuals from families with an HNPCC-related mutation in Wales	Compare an extended genetic nurse specialist counseling protocol (2 sessions of education and reflection) to shortened protocol (1 session of education).	The effect of counseling protocol on knowledge and psychological distress was not significant. [A, B] No differences were found in satisfaction between groups. [D] All participants reported intentions to pursue testing prior to counseling, which did not change after counseling in either group. [E]
McInerney-Leo et al (2004)	54. BRCA1/2 testing in hereditary breast and ovarian cancer families: Effectiveness of problem-solving training as a counseling intervention	212 men and women (≥18y) with a BRCA1/2 mutation	Compare a problem-solving training intervention to a client-centered intervention, including assessment of changes at follow-up 6–9 months later.	A greater decrease in breast cancer worry was observed in those who chose testing than those who did not. [A, F] Depressive symptoms decreased more in the problem-solving training group than the client-centered counseling group. [A] Cancer worry was higher in participants with a personal history of breast cancer, regardless of decision to test. [A]
Brain et al (2000)	55. A randomized trial of specialist genetic assessment: Psychological impact on women at different levels of familial breast cancer risk	740 women (mean 42.6y) referred for a family history of breast cancer in Wales	Evaluate the impact of augmenting a surgical consultation with a multidisciplinary specialist risk assessment by a clinical geneticist.	Breast cancer worry and perceived risk decreased from baseline in low- and moderate-risk women, but not in high-risk women. [A, C] Women at high risk reported significantly lower instrumental satisfaction than women at low or moderate risk. [D] The effect of risk information on interest in genetic testing was not significant. [E]
Lerman et al (1999)	56. Racial differences in testing motivation and psychological distress following pretest education for BRCA1 gene testing	228 Caucasian and 70 African American women (18–75y) with at least one first-degree relative affected by breast or ovarian cancer	Compare a standard education model (E) to an education plus counseling model (E + C) in terms of racial differences.	All groups evidenced a reduction in distress. This decrease, although not significantly different, was smallest among African American women who received E + C. [A] In African American women E + C led to greater increases than E in intentions to be tested and provision of a blood sample. In Caucasian women there were no differential effects of the interventions on these outcomes. [E]
Watson et aL (1998)	57. Family history of breast cancer: What do women understand and recall about their genetic risk?	115 women (28–7ly) with a family history of breast cancer referred for genetic counseling in London	Evaluate the impact of augmenting standard clinical geneticist consultation with provision of an audiotape of the encounter.	Cancer worry was reduced in participants who received a tape. [A] Recall of the risk figure was more accurate when the clinical geneticist had given this to the woman as an odds ratio rather than in other formats. However, it was not affected by provision of a tape. [C] Overall, participants were satisfied with the service offered. [D] Increases in clinical breast examinations and mammograms were seen between 1- and 6-month follow-ups. [H]
Lerman et al (1997)	58. Controlled trial of pretest education approaches to enhance informed decision-making for BRCA1 gene testing	400 women (18–75y) with one first-degree relative affected by breast or ovarian cancer	Compare education only and education plus counseling to control (waiting list) for decision-making regarding BRCA1 testing.	Participants in both the educationaland counseling interventions showed increases in knowledge, while participants in the control group showed decreases. [B] The education alone, but not education plus counseling, resulted in significantly greater decreases in perceived risk of a BRCA1 mutation compared to controls. [C] Neither approach led to changes in intentions to pursue testing. [E]

^aUnless otherwise noted in this Sample column, the study was conducted in the United States of America

Study Design and Specialty Setting

Concordant with this review’s inclusion criteria, all study designs were experimental in nature. All studies used a randomized controlled design to assess the impact of a counseling intervention or delivery mode on genetic counseling patient outcomes. The majority of publications reported on studies conducted in the cancer genetics setting (n = 45). Other genetic counseling specialties were, prenatal (n = 7), adult (n = 5), and pediatrics (n = 1). RCTs of genetic counseling in the prenatal setting focused on Down syndrome (n = 1), Cystic Fibrosis carrier testing (n = 2), and prenatal diagnostic options. Studies in the adult setting centered upon Alzheimer’s disease (n = 2) and type 2 diabetes mellitus (n = 2). Lastly, the one pediatric study focused on child carrier status for Cystic Fibrosis. Refer to Table 3 for specialty distribution.

Table 3.

Frequency of specialty settings and outcomes

Specialty settings	n (%a)
Cancer	45 (76)
Prenatal	7 (12)
Cystic fibrosis	2 (3)
Down syndrome	1 (2)
Adult	5 (9)
Type 2 diabetes mellitus	2 (3)
Alzheimer’s disease	2 (3)
Pediatrics	1 (2)
Outcomes	n (%a)
Psychological wellbeing [A]	40 (69)
Knowledge [B]	29 (50)
Perceived risk [C]	23 (40)
Satisfaction [D]	20 (35)
Intentions to pursue genetic testing / Screening [E]	15 (26)
Genetic testing uptake [F]	11 (19)
Decision quality [G]	10 (17)
Medical management / Health behavior [H]	9 (16)
Sharing information [I]	5 (9)
Informed choice [J]	2 (5)

^aThese values correspond to the percentage of publications, and thus use 58 as the denominator

Sample Characteristics

The majority of studies were conducted in the United States (n = 37), while six studies each were conducted in the United Kingdom and the Netherlands. The number of participants involved in each study widely varied, ranging from 26 to 2002. On average, studies consisted of 304 participants. Collectively, 16,417 genetic counseling clients are represented by these studies. Two-thirds of studies (n = 39) enrolled only women, while no studies enrolled only men. Participant age ranged from 18 years to “greater than 50 years,” with most participants in their forties. The majority of publications (n = 45) reported on studies conducted in the cancer genetics setting where inclusion criteria varied from having a personal and/or family history of cancer to being at risk for or testing positive for a BRCA1/2 or hereditary nonpolyposis colorectal cancer (HNPCC)-related mutation. A number of studies sampled from healthy populations, requiring an absence of a cancer diagnosis or significant cancer family history. Few studies’ inclusion criteria specified ethnicity, with the exception of three studies that recruited African American women and two that recruited Jewish women. For studies of prenatal genetic counseling, advanced maternal age, infertility, and gestational age were the inclusion criteria across studies. Adult genetic counseling studies defined their inclusion criteria by family history of Alzheimer’s disease or increased risk for type II diabetes mellitus. In a study of pediatric genetic counseling, parents were eligible to enroll after their child screened positive on newborn screening for Cystic Fibrosis and one CFTR mutation was identified.

Providers and Outcomes

Eleven studies (reported in 14 publications) specified that the provider was a “certified” genetic counselor, though most described the providers as genetic counselors. Some studies involved providers other than genetic counselors including, for example, nurse counselors (n = 9) and medical geneticists (n = 7). The background of the providers was not specified in five studies. In addition to heterogeneity in provider training and expertise, participant outcome assessment was measured using a variety of scales.

The majority of RCTs assessed multiple patient outcomes following genetic counseling. Psychological wellbeing was the most frequently assessed outcome, and was reported in 69% of publications (n = 40). Psychological wellbeing was measured as feelings of anxiety, depressive symptoms, (lack of) intrusive thoughts, worries (often cancer-specific), and quality of life. Nearly all were measured using published validated and reliable scales. Depression, for example, was measured using the Brief Symptom Inventory (BSI), Center for Epidemiologic Studies Depression Scale (CES-D), 12-item General Health Questionnaire (GHQ-12), and Hospital Anxiety and Depression Scale (HADS). Other frequently assessed outcomes included knowledge (n = 29), perceived risk (n = 23), and satisfaction (n = 20), and intentions to pursue testing or screening (or interest in undergoing testing; n = 15). Information sharing with relatives and informed choice (having sufficient relevant knowledge and making a decision in accordance with one’s attitudes toward testing) were assessed least often (n = 5 and n = 2, respectively). Varied single-item or study-specific novel scales were often used to measure risk perception and knowledge. Table 3 lists the frequencies of the top ten genetic counseling outcomes identified by this systematic literature review.

Quality Assessment

Nine of the 58 publications were evaluated as having a low risk of bias per both sets of criteria (Albada et al. 2015; Bowen et al. 2004; Brain et al. 2005; Butrick et al. 2015; Hooker et al. 2011; Kinney et al. 2014; Kuppermann et al. 2014; Montgomery et al. 2013; Roussi et al. 2010), and accordingly considered to have a strong methodology. Notably, 24 other publications satisfied sufficient categories only in the second set of criteria (Bennett et al. 2007; Bowen et al. 2002; Calzone et al. 2005; Eijzenga et al. 2014; Eijzenga et al. 2015; R. Green et al. 2014; Halbert et al. 2010; Hanprasertpong et al. 2013; Helmes et al. 2006; Henneman et al. 2013; Hodgson et al. 2015; Hunter et al. 2005; Lowery et al. 2014; Miedzybrodzka et al. 1995; Schwartz et al. 2001; Schwartz et al. 2009; Schwartz et al. 2014; Roberts et al. 2012; Roshanai et al. 2009; Wakefield et al. 2008; Wang et al. 2005; Watson et al. 1998; Wevers et al. 2014; Yee et al. 2014), while none passed only the set of criteria from the Cochrane Handbook for Systematic Reviews of Interventions. Additional details regarding the evaluations of each of the publications are given in Online Resource 1.

Quality Criteria

All but three publications explicitly stated having obtained institutional ethics board approval to protect human subjects (Hwa et al. 2010; Roberts et al. 2012; Schwartz et al. 2001). Fifty-three publications specified that participants had given informed consent before enrolling in the studies; five did not (Bowen et al. 2002, 2004; Haga et al. 2014; McInerney-Leo et al. 2004; Platten et al. 2012).

Sensitivity Analysis

Randomization and allocation concealment were successfully implemented together in only nine publications (Albada et al. 2015; Bowen et al. 2004; Brain et al. 2005; Butrick et al. 2015; Hodgson et al. 2015; Hooker et al. 2011; Kinney et al. 2014; Kuppermann et al. 2014; Roussi et al. 2010). Sixteen publications reported randomization without proper allocation concealment (Bennett et al. 2007; Bowen et al. 2002; Bowen and Powers 2010; Brain et al. 2000; Buchanan et al. 2015; Eijzenga et al. 2014; Eijzenga et al. 2015; M. Green et al. 2004; Halbert et al. 2010; Lerman et al. 1999; Montgomery et al. 2013; Schwartz et al. 2001; Schwartz et al. 2009; Schwartz et al. 2014; Wevers et al. 2014; Yee et al. 2014), and five publications adequately concealed allocations without ensuring proper randomization (Hanprasertpong et al. 2013; Hunter et al. 2005; Miedzybrodzka et al. 1995; Wakefield et al. 2008; Watson et al. 1998).

Missing Data

Participation rates were noted, and 31% of the publications had a follow-up rate that did not exceed 80% (Bennett et al. 2007; Braithwaite et al. 2005; Eijzenga et al. 2014; Eijzenga et al. 2015; Graves et al. 2010; M. Green et al. 2004; Halbert et al. 2010; Henneman et al. 2013; Hodgson et al. 2015; Lerman et al. 1997; Lerman et al. 1999; Matloff et al. 2006; Miller et al. 2005; Montgomery et al. 2013; Nishigaki et al. 2014; Platten et al. 2012; Roussi et al. 2010; Wakefield et al. 2008). Seventeen publications analyzed the data with an intention-to-treat analysis (Butrick et al. 2015; Eijzenga et al. 2014; Eijzenga et al. 2015; Graves et al. 2010; R. Green et al. 2014; Helmes et al. 2006; Hodgson et al. 2015; Hooker et al. 2011; Kinney et al. 2014; Kuppermann et al. 2014; Lowery et al. 2014; Montgomery et al. 2013; Roberts et al. 2012; Schwartz et al. 2009; Schwartz et al. 2014; Watson et al. 1998; Wevers et al. 2014). The majority of publications reported follow-up rates that did not vary greatly between the intervention and control arms, and thus attrition was not considered a risk to analytical rigor.

Overall Synthesis of Evidence

The RCTs meeting inclusion criteria could be broadly categorized into two groups: comparisons between modes of genetic counseling service delivery (such as telephone vs. in-person counseling) and comparisons between genetic counseling strategies (such as person-centered vs. problem-based counseling). These two categories of RCT differ in purpose. Comparison between delivery modes yields data on alternative approaches to genetic counseling service provision. Comparison between strategies of counseling yields data on alternative ways to engage clients to achieve desired outcomes. We review the comparative outcomes according to these two general study aims. In both categories, genetic counseling was often provided in conjunction with the offer of genetic testing. While many of the studies were inadequately designed to determine whether differences in outcomes resulted from the counseling, the testing process and outcome, or a combination of the two, those assessing counseling interventions were largely designed to capture coutcomes separate from receipt of a test result.

Studies Comparing Service Delivery Modes

Twenty-seven studies described in thirty-one publications compared ways to deliver genetic counseling services (See Table 1: same study reported in publications 2 & 6; 23 & 25; 26 & 27; and 29 & 30). Thirteen studies compared inperson counseling to either telephone counseling or interactive online platforms. These RCTs assessed the effectiveness of alternative delivery modes that consume fewer resources than in-person genetic counseling. Outcomes were most often found to be equivalent, non-inferior, or not statistically different between delivery modes; this suggests that telephone counseling or interactive online platforms may be sufficient for teaching key information about test results, specifically cancer risk status and screening and prevention strategies. Most of these RCTs (74%) assessed psychological wellbeing to ensure that these alternative, cost-effective interventions did not lead to greater distress than in-person counseling, and found they did not (Calzone et al. 2005; Graves et al. 2010; M. Green et al. 2001a; R. Green et al. 2014; Hooker et al. 2011; Hwa et al. 2010; Kinney et al. 2014; Schwartz et al. 2014).

Four RCTs found differences in testing uptake or intentions to undergo testing between study arms (M. Green et al. 2001b; Kinney et al. 2014; Schwartz et al. 2014; Wang et al. 2005), but their implications remain unclear. In certain contexts such a finding may indicate that one delivery mode is more successful in facilitating an informed choice about whether to undergo testing. In other contexts eligible candidates may be less likely to pursue testing after telephone counseling because it requires an office visit. Thus, differences in testing uptake or intentions may not reflect differential efficacy between the delivery modes.

Further, while the majority of studies found no differences in satisfaction, two studies reported that telephone counseling yielded higher satisfaction than in-person counseling (Buchanan et al. 2015; Jenkins et al. 2007), and one found that a computer-based interactive decision aid led to higher satisfaction than usual care (Schwartz et al. 2009). In another RCT, patients preferred in-person counseling to an interactive computer platform largely based on the opportunity to ask specific questions of the counselor (M. Green et al. 2001b).

Studies Comparing Counseling Interventions

Twenty-seven studies compared two different counseling strategies or interventions (See Table 1). Of these RCTs, 17 evaluated the impact of augmenting in-person genetic counseling (or usual care) with a variety of interventions: a decision aid (Kuppermann et al. 2014; Miller et al. 2005; Wakefield et al. 2008), personalized risk assessment (Brain et al. 2000; Braithwaite et al. 2005; Matloff et al. 2006), culturally sensitive genetic counseling (Charles et al. 2006; Glanz et al. 2007; Halbert et al. 2010), and counseling approaches described as psychological (Bowen et al. 2002) and problem solving (McInerney-Leo et al. 2004). A number of these interventions were informed by health behavior or stress and coping theory. Increases in knowledge followed “enhanced” genetic counseling (Roussi et al. 2010), a coping intervention (Bennett et al. 2007), and one decision aid (Wakefield et al. 2008). Reductions in perceived cancer risk were found following psychological counseling (Bowen et al. 2002), education without counseling (Lerman et al. 1997) and risk assessment (Brain et al. 2000). Cancer worries were reduced following a culturally sensitive intervention (Charles et al. 2006), psychological counseling (Bowen et al. 2002), and risk assessment (Brain et al. 2000). Intrusive thoughts (Bennett et al. 2007) and depressive symptoms (McInerney-Leo et al. 2004) were lower in a coping and a problem-solving intervention arm, respectively. Although these significant differences offer preliminary evidence for the efficacy of these interventions to improve patient outcomes, the lack of consistency in the intervention targets, theoretical underpinnings and the outcomes assessed preclude distillation of clear evidence on the most efficatious interventions to guide genetic counseling practice.

Discussion

This systematic literature review is the largest to date summarizing 25 years of RCTs assessing outcomes of genetic counseling. The sheer number of studies meeting eligibility criteria (N = 54) reflects significant advancement in genetic counseling research, amassing evidence for improved client outcomes grounded in rigorous comparisons of both delivery modes and counseling interventions. Notable is the prevalence of assessment of psychological wellbeing in 40 of 58 publications, reflecting some consensus that it is a sought-after goal of genetic counseling. Moreover, assessment of wellbeing was achieved using valid and reliable scales from health psychology research including a lack of anxiety or depressive symptoms, low worries, few intrusive thoughts and high quality of life.

The data assembled in this review offer initial evidence that alternative delivery modes may be as effective as in-person genetic counseling for women participants at risk for hereditary cancer, as reflected by the majority of the RCTs. While there was substantial variation in types of alternative counseling interventions, there is also preliminary evidence among these studies for the benefits of graphical representation of risk information, problem-based counseling, ‘enhanced’ counseling interventions and condensed counseling sessions. Outcomes of the intervention studies can be used to generate hypotheses for future trials among sub-specialties, though larger and more diverse populations of genetic counseling clients and patients are needed before conclusions can be drawn about their effectiveness and generalizability.

Lacking from this impressive array of studies is consistency in how outcomes (e.g., psychological wellbeing) are assessed, precluding meta-analysis. Further research exploring outcomes sought by genetic counseling clients and patients across sub-specialties is greatly needed. Although progress has been made in identifying outcomes genetic counselors seek to achieve on behalf of their clients (Redlinger-Grosse et al. 2016; Zierhut et al. 2016), insufficient evidence has been solicited from end-users (Bernhardt et al. 2000). The limited data identifying outcomes from genetic counseling valued by patients highlights a critical need for advancing evidence-based practice in the era of patient-centered care.

In addition to the need for distinction of beneficial client outcomes, there is a need to identify elements of genetic counseling that may be amenable to intervention. Counseling psychology, health behavior and health communication theories should be used to inform study design and to identify appropriate constructs likely to affect outcomes. For example, self-regulation theory may be used to frame studies of decisions about use of genomic information (Cameron et al. 2017). There are communication interventions that can be used to enhance understanding of complex genetic information, such as using simple language with low literacy words and teach back for the clients to state back to the genetic counselors what they have heard. Also, to enhance understanding of risk information, ten evidence-based interventions have been suggested by Fagerlin et al. (2011) that can be assessed in RCTs of genetic counseling. More accurate risk perception may lead to more informed decision making about genomic testing and better adherence to screening recommendations, both of which could be construed as beneficial outcomes of genetic counseling. In addition to the educational goals, there is a need for attention to the relational elements of genetic counseling. For example, to convey understanding of the client’s experiences, worries, concerns, and counseling needs, genetic counselors may use reflections such as paraphrasing and summarization, and empathic statements about the client’s feelings. Interventions to enhance use of these reflections can be compared to usual care.

A further limitation of the RCTs in this review is the overall low quality ratings according to the guidelines from the Cochrane Handbook for Systematic Reviews of Interventions. Only nine studies were rated low in risk of bias with strong methodology, eight of which were successful in implementing randomization and allocation concealment together. As new RCTs of genetic counseling are designed, attention to rigorous methods is essential to enhancing the reliability of the evidence generated.

A highlight of this review is the innovation in genetic counseling research as evidenced by the RCT conducted by Eijzenga et al. (2014). In this RCT clients referred for cancer genetic counseling completed a pre-counseling questionnaire that was randomly shared with half of the genetic counselors. Access to the questionnaire led the genetic counselors to more frequently initiate discussion of topics important to the client and to clients having lower levels of cancer worries and psychological distress four weeks later. The intervention allowed the counselors to reflect on client needs prior to the counseling sessions and use the information to initiate a discussion that resulted in greater psychological wellbeing for the clients. The novelty of a small intervention resulting in significant client benefit suggests that other small systematic changes to genetic counseling interactions may also result in benefit for clients without extending the length of sessions. This study paves the way for future RCTs aimed at introducing theoretically based interventions to enhance practice.

Conclusion

Overall, RCTs from the past 25 years investigating genetic counseling outcomes provide emerging evidence for the effective use of telephone counseling and online resources in the cancer setting. They illustrate a number of counseling interventions that need further evaluation in additional contexts, but hold promise for enhancing the effectiveness of genetic counseling in meeting clients’ needs. Specifically, improvements in theoretical research design, attention to risk of bias and quality parameters, and consistent use of valid and reliable measures of outcomes such as psychological wellbeing, risk assessment, and knowledge, and further use of patient outcomes such as effective coping, and adaptation will promote evidence-based advancements in genetic counseling practice.