Perceived Understanding and Psychosocial Outcomes: Employees’ Responses to Learning Results of Workplace Genetic Testing

Abstract

Aims:

This study explored employees’ understanding of, and psychosocial responses to, workplace genetic testing (wGT) results.

Materials & Methods:

Employees of a US healthcare system who underwent wGT (hereditary cancer/heart disease risk, pharmacogenomics) and received results were surveyed. We ascertained pretest education engagement, test understanding, and psychosocial responses. Regression analyses identified predictors of scores on a modified Feelings About genomiC Test Results questionnaire (positive feelings, negative emotions, and uncertainty after wGT).

Results:

N=418 employees (mean age=44 years; 88.3% female; 80.6% white) completed the survey. Mean scores (out of 12; higher scores indicate a greater extent of each feeling) were 5.2 (SD=2.9) for positive feelings, 1.2 (SD=2.2) for negative emotions, and 2.0 (SD=2.5) for uncertainty. Identifying as non-Hispanic African American/Black and receiving increased risk (cancer/heart disease) wGT results were associated with lower positive feelings and higher negative emotions and uncertainty scores (all p<0.05). Open-ended responses indicated difficulty interpreting, recalling, and utilizing results.

Conclusions:

wGT was associated with low levels of measured psychosocial harm among participants. However, results suggested a greater likelihood of negative psychosocial responses among those with increased risk and non-Hispanic African American/Black employees. Future studies should explore strategies to ensure all employees undergoing wGT have educational and psychosocial support.

Introduction

There are long-standing concerns about the potential for genetic testing to cause psychological harm, especially when performed on healthy individuals outside of a clinical setting [1–3]. Distress and anxiety about the identification of pathogenic variants (e.g., BRCA1/2 associated with hereditary breast and ovarian cancer) through direct-to-consumer (DTC) genetic testing (without clinical counseling and guidance) have been of particular concern [2,3]. However, studies have demonstrated that individuals generally adapt well to genetic information, perceive their test results as useful, and experience minimal test-related distress in a DTC setting [4,5]. Genetic testing for actionable conditions (e.g., hereditary cancer, heart disease) has been associated with overall low levels of uncertainty, negative emotions, and decisional regret among individuals participating in population screening programs [6–9].

Psychosocial responses to genetic testing also depend on one’s understanding and interpretation of results [10]. Genetic counseling has been historically regarded as an important step in genetic testing [11–13], helping patients understand and adapt to their results [14]. As the expansion of genetic testing outpaces the genetic counseling workforce, traditional one-on-one pretest counseling may become less feasible [15]. This may not necessarily be problematic because (a) unselected individuals (who may be at low risk of genetic disease) are the target population of wGT, and (b) studies show low levels of decisional conflict after testing when alternative pretest counseling and education methods are used [16,17]. Regardless of the extent to which pretest counseling is incorporated, pretest education (discussing medical management implications of potential test results, the voluntary nature of testing, and implications for relatives, etc.) facilitates informed decision-making [18,19], which can help individuals cope with results. While studies have demonstrated overall low levels of psychosocial harm after DTC genetic testing [4,5], studies have also shown that individuals who undergo such testing have misconceptions or misinterpret their results (e.g., false reassurance based on a negative test result) [20–22]. Given the evidence suggesting that the general population has a limited understanding of genetics [23,24], these findings demonstrate a continued need for pretest educational resources and guidance to psychologically adapt to genetic test results.

Workplace genetic testing (wGT), offered by employers through workplace wellness programs [25,26], is a unique genetic testing format performed without the involvement of a clinician and with limited pretest education [27]. This raises questions about employees’ understanding of, and subsequent responses to, their results. Typically, wGT provides employees insight into their inherited predispositions to cancer, heart disease, and/or pharmacogenomic (PGx) drug response profiles [25,27,28]. There is a need to explore the potential psychosocial impacts of wGT to evaluate the model’s potential benefits and limitations. This analysis, which builds on prior work that examined employees’ health behaviors after wGT [27], characterized employees’ perceived understanding of, and psychosocial responses to, learning their wGT results.

Materials & Methods

Study Setting

Employees of an Eastern US healthcare system were offered wGT that analyzed genes associated with cancer and heart disease risk, as well as medication response (PGx), via a third-party CLIA/CAP-certified (Clinical Laboratory Improvement Amendments/College of American Pathologists) laboratory. Eligible employees (i.e., those 18 years or older and employed for 60 days or more) received an email from their employer about this wellness benefit. Approximately 8,000 out of roughly 30,000 employees had submitted a sample at the time of our study (Figure 1) [27]. At the time of our study, the healthcare system’s employee population was approximately 3:1 female to male, and the majority was white. Additionally, of those who requested a sample collection kit (Figure 1), roughly 80% were female, and the majority were white. Positive results (i.e., pathogenic variant in a cancer or heart disease-related gene) were disclosed to employees through post-test genetic counseling with a laboratory genetic counselor (GC). All PGx and negative cancer/heart disease results were returned to employees via email [27].

Figure 1.

Study Population Overview and Survey Items Utilized for this Publication.*

*Adapted with permission from: Charnysh E, Pal S, Reader JM, et al. Health Care Utilization and Behavior Changes After Workplace Genetic Testing at a Large US Health Care System. Genet Med. 2024;26(8):101160.

Abbreviations: N (Total Number)

Study Design & Procedures

We recruited wGT-eligible employees to participate in an electronic survey developed by an interdisciplinary team based on a literature review, broad collaborator input, and adaptation of survey items and validated measures from previous studies (e.g., the PGen Study [29]). Survey administration occurred from December 2021 to May 2022 via a secure web-based platform (SoundRocket, Ann Arbor, MI). A detailed description of the study design, recruitment, and data collection procedures has been previously reported [27]. Before beginning the survey, participants read the information sheet, which implied consent through the submission of a survey response. This study was reviewed by the Institutional Review Board of The Jackson Laboratory and determined to be exempt.

Survey Measures

We ascertained self-reported demographic characteristics and personal/family histories of cancer and heart disease through multiple-response (i.e., “select all that apply”) questions. General medical and mental health status (1=Excellent, 5=Poor), perceived test result utility (1=Not at all useful, 4=Very useful), and self-reported ease of test result understanding (1=Very difficult, 5=Very easy) were obtained using Likert-scale questions. We collected employees’ self-reported wGT results (i.e., increased risk for cancer and/or heart disease, informative for the use of prescription medications) that included open-ended questions for participants to expand upon what they learned about their results (Figure 1). If participants received more than one increased risk or informative test result, they would have been prompted to answer up to three open-ended questions (Figure 1). We describe self-reported results for cancer and/or heart disease as “increased risk” (IR) and “no increased risk” (NIR). Likewise, responses regarding whether wGT results might inform the use of prescription medications are described as “informative PGx” results or “uninformative PGx” results. Results were not confirmed with employees’ medical records. The Feelings About genomiC Test Results (FACToR) [30] questionnaire, validated to measure the psychosocial impact of receiving genetic test results, was modified specifically for wGT (Supplemental Table 1; negative emotions, positive feelings, and uncertainty out of 12; higher scores indicate a greater extent of each feeling).

In addition to the above measures, this analysis included items from a multiple-response question examining employee engagement with pretest education and activities (See Charnysh et al., 2024 Supplemental Table S1). Employees indicated which activities they engaged in before deciding whether to pursue wGT, including talking with their primary care physician (PCP) or other clinicians about wGT, attending an information session, or engaging with material from the genetic testing laboratory (i.e., watching the video included with the email invitation from the testing laboratory, or reading information provided by the testing laboratory).

Data Analysis

Quantitative Analysis

Descriptive statistics were used to characterize participants who completed wGT and reported receiving their results. Subscales adapted from the FACToR questionnaire [30] were scored by summing the individual items. Three of the four subscales from the modified FACToR questionnaire were examined for this analysis (negative emotions, positive feelings, and uncertainty; all ranging from 0–12; not including the privacy concerns subscale) (Supplemental Table 1). These subscales demonstrated good internal consistency with Cronbach’s alpha coefficients of 0.85 (negative emotions: 95% CI: 0.82, 0.87), 0.72 (positive feelings: 95% CI: 0.67, 0.77), and 0.80 (uncertainty: 95% CI: 0.76, 0.83) [27]. For this analysis, higher scores on all subscales indicated a greater extent of the corresponding feeling. Analyses used R version 4.2.3 and RStudio Version 2023.09.1+494.

Regression models were used to determine whether key independent variables, including demographics (age, sex, race and ethnicity, education), personal and family history (heart disease and/or cancer, general/mental health status), and testing perceptions and experiences (pretest education, wGT results, ease of understanding test results, usefulness of test results), were associated with higher scores on each of the positive feelings, negative emotions, and uncertainty subscales of the modified FACToR questionnaire. Analysis of the distribution of scores for each subscale determined the appropriate types of regressions to perform. Based on the normal distribution of data for the positive feelings subscale, we utilized linear regression. Due to non-normal data distributions for the negative emotions and uncertainty scales, logistic regression models were used. A Fisher test was performed to determine Cramer’s V [31] for perceived usefulness and ease of understanding. P-values <0.05 were considered statistically significant.

Qualitative Analysis of Open-Ended Responses

A GC researcher (EC) and a public health researcher (SM) reviewed open-ended responses to questions that asked participants about what they learned from their wGT results (Figure 1). A preliminary codebook was inductively developed and applied to all responses through independent double coding. The codebook was reviewed by a qualitative researcher with expertise in the ethical, legal, and social implications (ELSI) of PGx (KH) and revised for clarity. Once the codebook (Supplemental Table 2) was finalized, responses were again double-coded (EC, SM). Coding disagreements were discussed and resolved as needed. Coded data were categorized by result type (cancer, heart disease, PGx), and thematic findings were summarized within each category, highlighted by representative quotes.

Results

Participant Demographic & Health Characteristics

A total of 776 eligible employees participated in the survey, of whom 418 (53.9%) indicated they underwent wGT and received their results (Figure 1). This sample of n=418 analyzed in this paper will hereafter be referred to as “participants.” Participants had a mean age of 44 years and were majority female (88.3%), college-educated (72.0%), and self-identified as non-Hispanic white (80.6%). Most reported good to excellent mental (87.7%) and general (94.1%) health (Table 1). While few reported a personal history of cancer or heart disease, many reported a family history of these conditions (Table 1). Additional descriptive analyses indicated the sub-groups analyzed for the regression had demographic and health characteristics consistent with the larger sample.

Table 1.

Participant Demographic and Health Characteristics (N=418).

Participant Characteristics	N (%)
Age: (mean=44 years; SD=12.8)
20–29	59 (14.1%)
30–39	34 (32.1%)
40–49	71 (17.0%)
50–59	88 (21.1%)
60–69	63 (15.1%)
70–74	3 (0.7%)
Sex:
Female	369 (88.3%)
Male	49 (11.7%)
Race & Ethnicity:
Non-Hispanic African American/Black	24 (5.7%)
American Indian/Alaska Native	2 (0.5%)
Asian American	27 (6.5%)
Latino/Hispanic	12 (2.9%)
Middle Eastern/Northern African	1 (0.2%)
Non-Hispanic White	337 (80.6%)
More than one race	13 (3.1%)
Race not listed	2 (0.5%)
Education:
Master’s or Professional Degree	151 (36.1%)
Bachelor’s Degree	150 (35.9%)
Some College, Technical School, or Two-Year Degree	102 (24.4%)
High School	5 (3.6%)
Health System Occupation:
Clinical Care Provider	43 (34.2%)
Clinical Support	63 (15.1%)
Corporate Services	39 (9.3%)
Clinical Administrative Staff	36 (8.6%)
Clinical Leader/Administrator/Manager	32 (7.7%)
Other	83 (19.9%)
More than one occupation selected	22 (5.3%)
General Health Status:
Excellent	70 (17.8%)
Very Good	89 (48.1%)
Good	11 (28.2%)
Fair	21 (5.3%)
Poor	2 (0.5%)
Missing	25
Mental Health Status:
Excellent	88 (22.4%)
Very Good	39 (35.5%)
Good	17 (29.8%)
Fair	36 (9.2%)
Poor	12 (3.1%)
Missing	26
Personal History of Heart Disease:
Heart Disease History	64 (16.4%)
No Heart Disease History	327 (83.6%)
Missing	27
Personal History of Cancer:
Cancer History	46 (11.8%)
No Cancer History	343 (88.2%)
Missing	29
Family History of Heart Disease:
None	71 (18.4%)
One Relative	97 (25.1%)
Two or More Relatives	92 (49.7%)
Don’t Know	26 (6.7%)
Missing	32
Family History of Cancer:
None	76 (19.5%)
One Relative	85 (21.9%)
Two or More Relatives	210 (54.0%)
Don’t Know	18 (4.6%)
Missing	29

Abbreviations: SD (Standard Deviation); n (Total Number).

Adapted with permission from: Charnysh E, Pal S, Reader JM, et al. Health Care Utilization and Behavior Changes After Workplace Genetic Testing at a Large US Health Care System. Genet Med. 2024;26(8): 101160

Pretest Education & Activities

Half of the participants (n=207/414) indicated they read the information provided by the genetic testing laboratory (e.g., brochure, website) before testing. Roughly one-quarter (27.3%, n=113/414) of participants reported watching the video sent with the email invitation from the laboratory. Thirteen percent of participants (n=54/414) reported talking with a PCP or another clinician before testing. A small number of participants (2.7%, n=11/414) attended a virtual information session provided by the laboratory. Twenty percent of employees (n=84/414) reported no engagement in the pretest education activities. Of note, participants could select more than one pretest education activity.

Results of wGT

Figure 2 summarizes participants’ self-reported wGT results (n=418). A small number reported receiving test results indicating IR for cancer (12.0%) or heart disease (9.5%), while nearly one-third (31.4%) reported receiving informative PGx results. Four percent (n=14/371) of participants reported they received IR results for both cancer and heart disease in addition to an informative PGx result, and 4.9% (n=18/371) reported receiving at least two IR and/or informative results. Of note, some individuals reported they did not know their wGT results related to cancer (7.0%), heart disease (8.8%), or PGx (20.6%).

Figure 2.

Self-Reported Workplace Genetic Test (wGT) Results.*

*Adapted with permission from: Charnysh E, Pal S, Reader JM, et al. Health Care Utilization and Behavior Changes After Workplace Genetic Testing at a Large US Health Care System. Genet Med. 2024;26(8): 101160.

Note: Results (N=399 for cancer, N=398 for heart disease, and N=398 for pharmacogenomics (PGx)).

Abbreviations: PGx (Pharmacogenomics); wGT (Workplace Genetic Testing).

Testing included analysis of 30 genes associated with hereditary cancer (breast, ovarian, colon, uterine, skin, and other cancers), 30 genes associated with inherited cardiovascular conditions (cardiomyopathies, arrhythmias, and Familial Hypercholesterolemia), and analysis of 14 genes associated with pharmacogenomics (PGx) for medication response (e.g., non-response, adverse effects).

Perceived Test Utility & Understanding

Most participants found their wGT results either a little (20.7%, n=82/397), somewhat (37.8%, n=150/397), or very useful (36.3%, n= 144/397), while 5.3% (n=21/397) indicated their results were not at all useful. Furthermore, most participants found their results somewhat (42.2%, n=168/398) or very (35.2%, n=140/398) easy to understand, while 12.6% (n=50/398) found results neither difficult nor easy to understand or somewhat (9.5%, n=38/398) to very (0.5%, n=2/398) difficult to understand.

Feelings About wGT Results

Overall, positive feelings subscale scores were relatively evenly distributed among participants, while scores on the negative emotions and uncertainty scales had a non-normal distribution (e.g., skewed toward lower scores) (Figure 3). On a 12-item scale, mean scores were 5.2 (SD=2.9) for positive feelings, 1.2 (SD=2.2) for negative emotions, and 2.04 (SD=2.5) for uncertainty (Figure 3).

Figure 3.

Distributions of Participants’ FACToR Questionnaire* Scores in Response to Workplace Genetic Testing (wGT).

Note: N=384 for positive feelings, N=391 for negative emotions, and N=395 for uncertainty.

*The Feelings About genomiC Testing Results (FACToR) questionnaire (Li et al., 2018) was adapted for our survey to fit the wGT scenario. For this analysis, we included positive feelings (3 items), negative emotions (2 items), and uncertainty subscales (3 items). Subscales adapted from the FACToR questionnaire were scored by summing the individual items, all ranging from 0–12 (Supplemental Table 1). Higher scores on all scales indicated a greater extent of the corresponding feeling.

Abbreviations: wGT (Workplace Genetic Testing); FACToR (Feelings About genomiC Testing Results); n (Total Number).

Table 2 summarizes the results of linear (positive feelings) and logistic (negative emotions, uncertainty) regressions exploring predictors of psychosocial responses to wGT. Results demonstrated that self-identifying as non-Hispanic African American/Black (5.7%) (vs. non-Hispanic white, 80.6%) was associated with scoring lower on the positive feelings subscale (β=−2.1; 95% CI: −3.8, −0.48; p=0.012) and scoring higher on the negative emotions (aOR=8.68; 95% CI: 1.97, 61.6; p=0.010) and uncertainty subscales (aOR=5.92; 95% CI: 1.37, 41.4; p=0.032). Finding one’s test results easy to understand (β=1.6; 95% CI: 0.42, 2.8; p=0.008) and useful (β=3.1; 95% CI: 1.5, 4.8; p<0.001) were also significantly associated with higher scores on the positive feelings subscale. Of note, perceived usefulness and ease of understanding were weakly associated (Cramer’s V=0.17). Younger age (aOR=0.95; 95% CI: 0.93, 0.98; p<0.001) and speaking with a PCP or other clinician before undergoing wGT (aOR=3.83; 95% CI: 1.15, 14.0; p=0.033) were both significantly associated with a higher likelihood of experiencing negative emotions. Of note, those who spoke with a PCP or other clinician before undergoing wGT did not differ from the remainder of the sample with regards to demographic characteristics, general mental health status, or personal/family history of disease. Finally, IR results were significantly associated with more negative emotions (aOR=3.49; 95% CI: 1.63, 7.72; p=0.002), greater uncertainty (aOR=3.03; 95% CI: 1.39, 6.96; p=0.006), and fewer positive feelings (β=−1.0; 95% CI: −2.0, −0.09; p=0.033) than NIR results. Personal and family histories of cancer/heart disease, other forms of pretest education, general and mental health, and PGx results were not significant predictors of positive feelings, negative emotions, or uncertainty about wGT among participants (Table 2).

Table 2.

Regressions: Predictors of Higher Scores on FACToR Questionnaire in Response to Workplace Genetic Testing (wGT).

	Positive Feelings (N=384) Linear Regression			Negative Emotions (N=393) Logistic Regression			Uncertainty (N=395) Logistic Regression
Independent Variables	β	95% CI	P	aOR	95% CI	P	aOR	95% CI	P
Age:	−0.01	−0.04, 0.02	0.7	0.95	0.93, 0.98	0.001	0.98	0.96, 1.01	0.15
Sex:
Male	−0.55	−1.6, 0.49	0.3	0.98	0.40, 2.31	>0.9	1.50	0.66, 3.49	0.3
Female	-	-	-	-	-	-	-	-	-
Race & Ethnicity:
Non-Hispanic African American/Black	−2.1	−3.8, −0.48	0.012	8.68	1.97, 61.6	0.010	5.92	1.37, 41.4	0.032
Asian American	1.4	−0.22, 3.1	0.09	1.13	0.30, 4.14	0.9	1.82	0.49, 7.76	0.4
Latino/Hispanic	0.89	−1.6, 3.3	0.5	0.40	0.02, 3.05	0.4	1.34	0.18, 11.8	0.8
More than one race and/or ethnicity	−0.06	−1.9, 1.8	>0.9	0.14	0.01, 1.03	0.093	0.84	0.18, 3.83	0.8
Non-Hispanic White	-	-	-	-	-	-	-	-	-
Education:
High School	−1.8	−3.9, 0.33	0.10	0.18	0.01, 1.47	0.2	−5.91	0.83, 1.21	0.12
Some College, Technical School, or Two-Year Degree	1.00	0.08, 1.9	0.03	1.01	0.46, 2.19	>0.9	0.55	0.26, 1.15	0.12
Master’s or Professional Degree	−0.49	−1.3, 30.30	0.2	0.78	0.40, 1.52	0.5	0.77	0.41, 1.45	0.4
Bachelor’s Degree	-	-	-	-	-	-	-	-	-
Personal History:
Cancer:
Yes	0.45	−0.63, 1.5	0.4	1.13	0.46, 2.72	0.8	0.62	0.26, 1.46	0.3
No	-	-	-	-	-	-	-	-	-
Heart Disease:
Yes	−0.97	−2.0, 0.03	0.059	1.50	0.65, 3.44	0.3	1.38	0.63, 3.11	0.4
No	-	-	-	-	-	-	-	-	-
Family History of Cancer or Heart Disease:
One Relative	0.42	−1.1, 1.9	0.6	0.75	0.21, 2.72	0.6	0.72	0.20, 2.47	0.6
Two or More Relatives	0.96	−0.37, 2.3	0.2	1.16	0.37, 3.76	0.8	0.83	0.26, 2.57	0.7
No	-	-	-	-	-	-	-	-	-
Health Status:
General Health:
Good to Excellent	0.15	−1.5, 1.8	>0.9	1.13	0.31, 4.47	0.9	0.32	0.06, 1.30	0.14
Fair to Poor	-	-	-	-	-	-	-	-	-
Mental Health:
Good to Excellent	−0.02	−1.4, 1.4	>0.9	0.82	0.25, 2.67	0.7	0.94	0.30, 2.97	>0.9
Fair to Poor	-	-	-	-	-	-	-	-	-
Pretest Education: *
Talked with a PCP or other clinician	0.56	−0.87, 2.0	0.4	3.83	1.14, 14.0	0.033	1.33	0.41, 4.52	0.6
Engaged with material from lab**	0.95	−0.02, 1.9	0.055	0.88	0.39, 2.00	0.8	0.86	0.39, 1.87	0.7
More than One of the Above	0.81	−0.17, 1.8	0.11	1.16	0.52, 2.63	0.7	1.16	0.52, 2.56	0.7
None of the Above	-	-	-	-	-	-	-	-	-
wGT Results:
Cancer/Heart Disease:
IR: Cancer and/or Heart Disease (n=67)	−1.0	−2.0, −0.09	0.033	3.49	1.63, 7.72	0.002	3.03	1.39, 6.96	0.006
Don’t Know (n=24)	−1.8	−3.8, 0.23	0.085	1.47	0.25, 8.02	0.7	0.89	0.16, 4.99	0.9
NIR: Cancer and Heart Disease (n=298)	-	-	-	-	-	-	-	-	-
PGx:
Informative PGx (n=125)	−0.37	−1.1, 0.39	0.3	0.78	0.40, 1.48	0.4	0.72	0.39, 1.32	0.3
Don’t Know (n=82)	0.04	−1.1, 0.99	>0.9	1.93	0.80, 4.69	0.14	1.92	0.82, 4.70	0.14
Uninformative PGx (n=191)	-	-	-	-	-	-	-	-	-
Ease of Understanding wGT Results:
Easy	1.6	0.42, 2.8	0.008	1.10	0.43, 2.89	0.8	0.38	0.13, 0.99	0.057
Difficult	-	-	-	-	-	-	-	-	-
Usefulness of wGT Results:
Useful	3.1	1.5, 4.8	0.001	0.49	0.12, 2.08	0.3	0.96	0.22, 3.95	>0.9
Not Useful	-	-	-	-	-	-	-	-	-

^*Participants were also asked if they engaged in other pretest activities such as talking with family/loved ones, someone from Human Resources, a supervisor/manager, other employees, friends/neighbors, and performing an internet search, which was not analyzed for this paper or included in the regression model.

^**Includes watching the video sent with the email invitation from the laboratory and reading information provided by the laboratory (e.g., brochure, website). See Charnysh et al., 2024 27.

Note: Bolded values are statistically significant (p<0.05).

Abbreviations: N (Total Number); CI (Confidence Interval); aOR (Adjusted Odds Ratio); PCP (Primary Care Physician); wGT (Workplace Genetic Testing); NIR (No Increased Risk); IR (Increased Risk); PGx (Pharmacogenomics).

Qualitative Analysis of Open-Ended Responses

A total of 158 participants reported receiving at least one IR or informative wGT result (IR for cancer: n=48; IR for heart disease: n=38; informative PGx: n=125). Of these 158 participants, we received 164 open-ended responses (IR for cancer: n=37/48; IR for heart disease: n=27/38; informative PGx: n=100/125) when participants were asked what they learned from their results (Figure 1).

For all three test categories, participants generally discussed educational information gained through testing, such as how various factors influence disease risk: “...several health conditions, your lifestyle,...your age, and family history can increase your risk” (Female, 26 years, IR for heart disease) and how the body metabolizes medications: “Genetic changes in genes can influence how fast the body breaks down a medication, how well the body absorbs a medication, or how quickly a medication gets to where it needs to work” (Female, 31 years, informative PGx). Apparent discrepancies in participants’ initial risk perceptions were also observed. Some participants answered “yes” to the multiple-choice question about receiving IR results but subsequently answered “no risk” or “no increased risk” in open-ended responses (Figure 1).

Some participants with IR for heart disease acknowledged their increased risk without providing details, while others interpreted their results in the context of already having manifestations of disease [e.g., responding with “high cholesterol” or “CAD” (coronary artery disease)]. Two individuals included the specific genes (KCNH2 and MYBPC3) in which they were found to have a pathogenic variant. Comparable to responses of those with IR for heart disease, some participants with IR for cancer named the specific cancers they were at higher risk for based on their results, including breast and colon cancer. Participants reported being identified as having a pathogenic variant in the following cancer-related genes: BARD1, BRCA1, CHEK2, MUTYH, PALB2, and PMS2. Those with IR for cancer indicated they learned about the actionability of their findings for healthcare management: “I have the BRCA2 gene, and I received consultation about my increased risk for varying cancers and…current recommendations for prevention…” (Female, 25 years). One participant who received an actionable result expressed frustration regarding the lack of insurance coverage for recommended management: “I have a CHEK2 mutation that puts me at higher risk of breast cancer and yearly breast MRIs are recommended but [employee’s health insurance company] disagrees and won’t pay for breast MRIs. VERY UPSETTING!!” (Female, 51 years). Others mentioned pursuing different preventative measures, such as colonoscopies and/or a hysterectomy due to a diagnosis of Lynch syndrome.

Both those with IR for cancer and heart disease indicated learning about risk factors for disease apart from pathogenic variants: “Family history, age at first period, race, and not having given birth all put me at an increased risk for cancer” (Female, 31 years, IR for cancer); “It was noted to screen with [a] family doctor or cardiolog[ist] due to [my mother’s] history” (Female, 54 years, IR for heart disease). One participant with IR for cancer indicated involvement in high-risk surveillance for colon cancer: “I was at an increased risk of breast and colon cancer. I already get colonoscopies every 5 years due to polyps in the past…” (Female, 62 years). Acknowledgment or awareness of a personal and/or family history of disease was also observed more generally in participants’ responses: “Cancer runs in my family” (Male, 29 years, IR for cancer); “I already have a heart condition” (Female, 54 years, IR for heart disease).

Responses from participants who reported informative PGx results ranged in comprehensiveness, forming a spectrum related to the information they learned about their prescription medication use from their wGT results (Figure 3). On one end of the spectrum, participants generally understood that the test performed was for medication response: “I may have a different reaction to certain medications” (Female, 27 years). Participants with informative PGx results tended to fall into this category. Some participants provided more complete responses, using PGx-specific terms such as “metabolism” or “dosage.” Infrequently, participants included specific genes and/or drug names: “[Ultra-rapid] metabolizer of CYP1A2 and poor metabolizer of CYP2C9” (Female, 39 years); “A mutation possibly affecting response to anticoagulants” (Female, 53 years).

Apart from the spectrum of PGx responses (Figure 3), other participants indicated their informative PGx results had personal relevance to their medication use, such as denoting medication changes: “It helped me decide which antidepressant to change to” (Female, 27 years). Some indicated test results may be relevant to future medication use or might be useful to revisit in the future: “If I should have a heart problem, what [medications] would be best and that my current anti-anxiety [medications] may not be the best fit for me” (Female, 53 years). Participants also expressed uncertainty or trouble remembering their PGx results: “Something about Coumadin, I think. I would have to go back and look” (Female, 55 years); “I have to share the information with a provider to learn more” (Female, 59 years). Responses also indicated that a discussion with a clinician could be helpful.

Discussion

We gained insights into employees’ understanding of, and psychosocial responses to, their wGT results by analyzing survey data from over 400 employees at a large healthcare system who underwent no-cost genetic testing and received their results. Most participants found their wGT results useful and easy to understand even though there was limited engagement in pretest education. With the metrics utilized, most employees did not experience psychosocial harm after receipt of wGT results, consistent with other studies examining DTC genetic testing [20–22] and population screening [6–9]. However, we identified a few predictors of negative psychosocial response to test results, including receiving IR results, self-identifying as non-Hispanic African American/Black, and younger age.

Our findings of perceived test utility and ease of understanding among employees are consistent with studies of clinical and DTC models [4,5,32,33]. Perceived test utility and ease of understanding were significantly associated with experiencing positive feelings after wGT. Employees expanded upon their perceived test utility in open-ended responses, indicating their results applied to their current circumstances and helped them plan for the future. Ease of understanding was also significantly associated with less uncertainty about wGT. While most employees may have found their wGT results useful and easy to understand, some open-ended responses seemed to suggest confusion, with a sub-analysis from the same dataset showing evidence of incorrect risk perceptions and self-directed medication changes among a subset of participants [27].

While we observed overall low levels of measured psychosocial harm among our entire sample, receiving IR results, specifically, was associated with higher negative emotional responses to testing, which is consistent with the broader genetic testing literature [6–9]. Although personal and family history of disease alone were not predictors of such responses, it should be noted that most participants with IR for cancer and/or heart disease reported having a family history of disease [27]. In the open-ended responses from those with IR results, some employees with IR for heart disease interpreted their results in the context of already having manifestations of disease, potentially conflating genetic risk with symptoms. Those with IR for cancer tended to provide specific details about their results (e.g., gene names, medical recommendations) and discussed what their results meant for their future healthcare management. Participants who received informative PGx results had far more difficulty recalling their results than those who received IR results (Figure 2)—reflected in open-ended responses. We observed a range in the comprehensiveness of participants’ conclusions drawn from testing. Few provided details about their precise findings, and some cited trouble remembering results and the need for a provider to explain their results, mirroring studies assessing patients’ understanding of PGx results [34]. Possible explanations may include uncertainty and low comfort with PGx information among the general population [7,34] and clinicians [35,36], highlighting the challenges in integrating PGx results into clinical care [37]. Compared to genetic testing for hereditary heart disease and PGx, cancer genetic testing has a longer history, greater media presence [38], and availability of nationally standardized guidelines (e.g., National Comprehensive Cancer Network^® [39]).

Participants reported underutilization of pretest educational materials in this study. Pretest counseling with a GC is not routine for wGT [27], and instead, alternative service delivery models, including webinars, videos, and written materials, are commonly offered to educate employees [27]. Roughly half of the participants reported reading the laboratory’s educational brochure, and fewer than half reported participating in other pretest activities. However, participants who reported speaking with a PCP or other clinician before wGT (13%) were significantly more likely to experience negative emotions about wGT. This finding was unexpected given that studies have suggested individuals seek support from both genetic and non-genetic clinicians in decision-making before genetic testing [40,41], and, therefore, should be viewed as preliminary, warranting further study.

While examining predictors of feelings after wGT, we found that self-identifying as non-Hispanic African American/Black was significantly associated with experiencing more negative emotions, uncertainty, and fewer positive feelings about such testing. Our sample only included 24 participants in this race/ethnicity category, with this subsample not differing from the larger sample with regard to other demographic and health characteristics. The broader literature on responses to genetic testing suggests African American/Black individuals may be more likely to have concerns about genetic and racial discrimination [42,43] and the privacy of genetic information [44–46]. Hypothetical studies examining employee perceptions have also suggested higher levels of employer mistrust among Black employees considering undergoing wGT [25,47]. While these are possible explanations, the exploration of these systemic issues was beyond the scope of this study.

Strengths & Limitations

In the age of genomic medicine, wGT is a relatively new and understudied avenue for genetic testing, with available studies having primarily explored hypothetical scenarios [47,48]. Our study provides data regarding psychosocial responses to wGT from employees at a large US healthcare system. However, our findings must be considered in light of limitations. Although wGT was offered to all employees, our findings potentially lack generalizability given limited diversity in sex, race and ethnicity, education, and health status, and all participants were recruited from a single employer. Additionally, since our research team was not connected to the testing laboratory (i.e., we did not directly offer this wGT program or conduct the testing), we could not access all pretest education materials offered to employees before undergoing wGT, nor could we verify employees’ self-reported wGT results. Given that wGT results were self-reported, we could not confirm results, demographics, and factors influencing open-ended responses. While the FACToR questionnaire is a validated instrument to assess psychosocial responses, specifically in the week immediately following wGT, most participants completed the survey after this period. Furthermore, the clinical significance of heightened FACToR subscale scores has yet to be determined [30]. Participants subjectively rated their perceptions and understanding of their results, but no formal objective measures of comprehension were used in our study. In addition, we did not examine other factors potentially contributing to the identified associations, especially between self-reporting as non-Hispanic African American/Black and negative psychosocial responses. Finally, we only captured employees’ short-term reactions to their wGT results, limiting our understanding of the long-term impacts of such testing.

Practice Implications & Future Directions

As advances are made in genetic testing and more patients seek access, one-on-one pretest genetic counseling is less feasible [16]. While we identified low levels of measured psychosocial harm in this study, measures should be taken to identify those requiring further psychosocial support and education for genetic testing, whether in clinical or non-clinical settings. Furthermore, additional support may be needed for individuals who do not have one-on-one genetic counseling to interpret and apply their results accurately. In our study, employees seemed to remember their cancer genetic test results and related details more than their heart disease and PGx results. This highlights opportunities for additional education and enhanced awareness related to genetic testing for heart disease and PGx.

In light of our findings demonstrating an association between self-identifying as non-Hispanic African American/Black and negative psychosocial responses, future studies should also explore strategies to support all employees undergoing wGT. Studies have suggested prioritizing culturally sensitive education, psychosocial support [42,49], tailored messaging, and racial representation in the genetics workforce [50] to address systemic issues in genetics care among racial and ethnic minorities. Addressing systemic issues is important to mitigate the cycle of health disparities experienced by minority populations.

To determine whether these results are generalizable, future research should include diverse populations, specifically, diversity of race and ethnicity, gender, occupation, and personal/family history of common diseases. Additionally, protocols could implement independent confirmation of wGT results and examination of test-related beliefs and accuracy of understanding [51]. Lastly, as our study only ascertained immediate, short-term psychosocial reactions to wGT results, future studies should utilize longitudinal approaches to understand the full impact of such testing on employees. Overall, our study highlights opportunities for developing pre and post-test educational programs that ensure genetic testing implications are fully understood by employees and other populations undergoing genetic testing.

Conclusions

Our study contributes to the limited literature on wGT by exploring the psychosocial responses of employees at a large US healthcare system who utilized this employment benefit. We found that responses to undergoing wGT aligned with other studies on psychosocial responses to DTC genetic testing and population genomic screening. These findings suggest that while wGT generally presents low levels of measured psychosocial harm, some employees may require additional support in understanding and adapting to their wGT results. All adverse psychosocial outcomes are important to address. This research strengthens our understanding of employee responses to wGT and the potential benefits and harms of genetic testing in unselected populations.

Supplementary Material

Supplemental Materials

See “Supplemental Materials.docx” for supplemental tables titled “Supplemental Table 1” and “Supplemental Table 2.”

ARTICLE HIGHLIGHTS.

Workplace Genetic Testing

• Workplace genetic testing (wGT) is a recently introduced employment benefit that occurs outside clinical settings. Little is known about employees’ understanding of, and responses to, wGT results.

Psychosocial Responses

• wGT was associated with overall low levels of measured psychosocial harm among employees. However, the findings suggested a greater likelihood of negative psychosocial responses among employees with increased risk for cancer/heart disease and self-identifying as non-Hispanic African American/Black, specifically.

Employee Understanding

• Participants provided a range of open-ended responses about their results, indicating difficulty in interpreting, recalling, and acting on results received outside of a clinical setting.

Implications & Conclusions

• This research contributes to the limited literature on wGT, including employees’ understanding of, and psychosocial responses to, such testing. These findings suggest that while wGT may not pose a substantial risk of psychosocial harm overall, certain groups may be more likely to experience such harm.

Data Sharing

The authors will make relevant data available for the purposes of verifying or contextualizing the conclusions we have drawn in this publication. Any disclosure will be constrained by the need to protect the privacy of respondents. Requests should be sent to the corresponding author (EC) with a description of the reason for the request and the qualifications of those requesting the data. The corresponding author (EC) will also accept requests for a copy of the survey instrument.