Study design

A cross-sectional survey study of newly diagnosed HGSOC patients was conducted at two academic centers in Toronto, Ontario. Eligible participants were recruited from the University Health Network (UHN) from December 2019–August 2021 and from Sunnybrook Health Sciences Centre from December 2020 to August 2021.

Study population & recruitment

Individuals with a new diagnosis of HGSOC were eligible to participate if they had: 1) pathology confirmed diagnosis of HGSOC; and 2) reflex BRCA1/2 tumour genetic testing results reported in the past 3 months. Individuals were excluded if they had already received results of germline genetic testing for hereditary ovarian cancer risk or were unable to complete study surveys due to language barriers or cognitive capacity.

Eligible participants with an upcoming appointment were flagged for their gynecologic oncologist (one of the nine participating surgeons), who disclosed BRCA1/2 tumour genetic testing results. Agreeable patients were approached by the study team to obtain informed consent. Consenting participants indicated their preferred method of completing the study survey (online, telephone or mail). Surveys were issued 1 week after disclosure of tumour genetic testing results; a reminder call or email was sent 1 and 2 weeks after the survey was provided.

MeasuresClinical & sociodemographic data

Relevant clinical data (age at diagnosis, cancer stage, primary treatment, personal history of breast cancer, family history of breast/ovarian cancer, and tumour genetic testing results) were abstracted from the medical record. Survey items were used to obtain additional socio-demographic data (race/ethnicity, partner status, household income, level of education obtained) as well as participant recall of their tumour genetic testing results.

Life Orientation Test – Revised (LOT-R) scale

Dispositional optimism was evaluating using the LOT-R, which is a 10-item scale consisting of five-point Likert scale questions to measure dispositional optimism [24]. The six scored items include: 3 items for optimism and 3 items for pessimism (reversed scored), with the remaining 4 items considered filler questions. Total scores range from 0 to 24. There was no missing data among scored items in our dataset. Cronbach’s alpha was used to calculate the internal consistency of the scales in our cohort. The LOT-R scale demonstrated acceptable internal consistency (α = 0.79).

Impact of Event Scale (IES)

The IES was used to measure cancer-related distress. The scale consists of 15 four-point response items to measure subjective distress following a specific event or stressor [25]. In the current study, the event was defined as “having a diagnosis of ovarian cancer”. The scale is comprised of intrusion (7 items; score 0–35) and avoidance (8 items: score 0–40) subscales; total scores range from 0 to 75, with thresholds for subclinical (< 9), mild [9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25], moderate [26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43] and severe (44+) impact of an event [26]. One participant had missing data for 5/15 items and was excluded from analyses involving IES. No other missing item-level data was detected. Cronbach’s alpha was calculated with acceptable internal consistency for intrusion (α = 0.83), avoidance (α = 0.77), and total scores (α = 0.84).

Multidimensional Impact of Cancer Risk Assessment (MICRA) questionnaire

The MICRA questionnaire was developed to measure the psychological impact of genetic testing result disclosure [27]. It is comprised of 21 four-point response items, 19 of which are scored (total possible score:0–95). The MICRA questionnaire includes subscales for genetic testing-related distress (6 items: score 0–30), uncertainty (9 items: score 0–45), and positive experiences (4 items, reverse scored: 0–20). Positive experience items are reverse scored, meaning higher scores indicate a less positive experience. The MICRA questionnaire does not have established thresholds; however, higher scores indicate higher levels of psychological impact of genetic testing. Permission to use the MICRA questionnaire was obtained from the Functional Assessment of Chronic Illness Therapy (FACIT) Organization (FACIT.org). Per FACIT instructions, mean imputation was used to account for item-level missing data. Due to an initial error in survey development, one item of the MICRA questionnaire “Being uncertain about what my tumour genetic test result means about my cancer risk” (included in total and uncertainty scores), was missing from the survey of 20 participants. Analyses involving the MICRA questionnaire were run with and without this item and its inclusion did not significantly impact study findings. The total scale had an acceptable internal consistency (α = 0.86); internal consistencies were also calculated for distress (α = 0.85), uncertainty (α = 0.76) and positive experience (α =0.59) subscales. Due to the low internal consistency in our cohort, scores for the positive experience subscale are not reported.

Knowledge of hereditary breast/ovarian cancer

Participant knowledge of hereditary breast/ovarian cancer was assessed using a modified BRCA knowledge scale. Developed by the National Center for Human Genome Research, the BRCA knowledge scale consists of 11 true/false items measuring four aspects of hereditary cancer: the prevalence of BRCA1/2 mutations, the patterns of inheritance, cancer risks associated with BRCA1/2 mutations, and risk management options for women with BRCA1/2 mutations [28]. Four true/false questions, developed in consultation with clinical experts, were added to measure knowledge of tumour versus germline genetic testing and potential treatment implications of tumour genetic testing results (Additional file 2). Knowledge scores were presented as the total percentage of correct responses; incomplete items were scored with a value of zero.

Satisfaction with reflex BRCA1/2 tumour testing

Participant reported satisfaction was assessed using a five-point Likert scale question.

Data analysis

Descriptive statistics were used to provide a summary of study variables. Variables were reported using means and standard deviations or frequencies and percentages, for continuous and categorical variables, respectively. To address the primary objective of describing the psychological impact of tumour genetic testing, MICRA mean scores, along with standard deviations, were calculated for the total score and two subscales (distress, uncertainty).

Secondary objectives were addressed as follows: Univariable and multivariable analyses were undertaken to determine factors associated with all MICRA scores (total and subscale scores). MICRA scores were compared across levels of categorical variables using an independent t-test. Linear regression was used to measure the association of continuous variables with MICRA scores. Ordinary Least Square regression was used to undertake multivariable analysis, where included variables (dispositional optimism (total LOT-R score), age at diagnosis, having children (yes/no), positive tumour genetic testing results (yes/no), and cancer related distress (total IES score) were selected a priori based on literature review and consultation with clinicians. Primary treatment (surgery/neoadjuvant chemotherapy) was added to the multivariable models post-hoc given its highly significant association with multiple MICRA outcomes. Hierarchial regression was completed where cancer-related distress was included in a second regression block. A minimum of 60 participants were required for analysis as 10 participants per predictor variable is considered appropriate for regression models with more than six predictor variables [29]. Regression coefficients, 95% confidence intervals, the adjusted coefficient of determination, and p values were reported. Participant knowledge and satisfaction were reported using descriptive statistics.

Statistical analyses were completed using IBM SPSS Statistics for Windows, version 28 (Armonk NY, USA: IBM Corp) and statistical significance was reported using a two-tailed α = 0.05. Given the exploratory nature of this study, statistical corrections for multiple testing were not completed.