A prospective experimental randomized controlled trial to investigate the effect of virtual reality (VR) technology on improving clinical pregnancy rates and lowering anxiety for IVF patients undergoing frozen embryo transfer (FET) Authors: Jak Ozsarfati, Samantha Yee, Michal Dviri, Anjila Roumia, Lilach Marom Haham, Andie Blankenstein, Ayesha Noman, Avishai Gasner, Shruti Pathak, Shivani Dhanani, Lianet Lopez, Carly Keshen, Karen Glass, Prati Sharma, Ari Baratz, Andree Gauthier-Fisher, Clifford Librach Published: 2026 ABSTRACT AbstractPurposeThe objective was to study the effect of exposure to virtual reality (VR) on clinical pregnancy rate (CPR) and anxiety levels in patients undergoing frozen embryo transfer (FET), one of the most stressful stages of the in vitro fertilization (IVF) treatment journey.MethodsA single-site prospective experimental randomized controlled trial was conducted between 05/2019 and 05/2023. Three hundred fifty FET patients using their own oocytes, or donated oocytes/embryos (oocyte provider’s aged 21–45) were enrolled and randomized in a 1:1 ratio to the VR and control groups. Clinical pregnancy rates (CPR) were assessed at 2, 6, and 12 weeks. Stress physiological parameters (heart rate, saliva samples, and systolic/diastolic blood pressure) and anxiety (STAI-State questionnaire) were assessed at baseline (T1), pre-FET (T2) and post-FET (T3). Patients were blinded to their group assignment at T1 and T2 measurements.ResultsPregnancy rates at 2 weeks were higher in the VR group than in the control group (52.6% vs. 44.6%, p = 0.134). The CPR increased significantly at 6 weeks (45.7% vs. 34.9%, p = 0.038) and 12 weeks (42.9% vs. 28.6%, p = 0.005) in the VR compared to the control group. In the multivariate logistic regression analyses, the total number of previous ovum pick-up, quality of the transferred embryo, and receiving VR exposure were the three significant predictors of increased CPR at 2, 6, and 12 weeks. Differences in the physiological stress parameters of all participants at three timepoints were found (p < 0.05) but did not differ between the two groups (p > 0.05). STAI-State anxiety scores among the VR group dropped much more substantially from T2 to T3 than in the control group (mean difference of 5.2 vs. 2.7, p = 0.026).ConclusionThe findings of this large-scale RCT suggest that one-time VR exposure before FET, along with other clinical factors such as the quality of transferred embryos, significantly increased CPR and decreased participants' stress and anxiety levels associated with the FET procedure.Trial registrationClinical Trial Registry: Clinical Trials.gov.Trial Registration Number: NTC04394962 https://clinicaltrials.gov/study/NCT04394962Registration Date: 2020–05-06.Date of Initial Enrollment: 2019–05-31.Supplementary InformationThe online version contains supplementary material available at 10.1007/s10815-026-03881-7.Keywords: Virtual reality, Randomized controlled trial, Embryo transfer, IVF, Stress reduction Introduction Approximately 26% of nulliparous women aged 15–49 living in the US have difficulties conceiving or carrying a pregnancy to term. Of these, approximately 12% have used assisted reproductive treatments (ART) to have one or more children [1]. Although in vitro fertilization (IVF) is an effective treatment for patients experiencing fertility challenges, advances in ovarian stimulation regimens and laboratory techniques do not guarantee a successful live birth, as evidenced by the latest US data showing that the overall live-birth deliveries of IVF cycles increased to 36.9% in 2020, from 31.5% in 2011 [1]. Furthermore, 18.8% of the 203,164 IVF cycles started in 2020 did not progress to embryo transfer (ET), and only 45.4% of the 165,041 ET cycles of that year resulted in live-birth deliveries [1]. Therefore, improving clinical pregnancy rates (CPR) is a priority for IVF centers worldwide. Many patients perceive ET as a crucial event during their IVF cycle [2]. The FET process itself can cause multifaceted stressors to patients, since only 10.6–48.5% of ET (depending on age) result in live-births [1]. One study found that the heightened anxiety experienced by IVF patients during ET was comparable to the preoperative mental state of women undergoing surgery [3]. It has been hypothesized that the intense anxiety evoked before, during and after ET may lead to various physiological responses and changes in reproductive functions and hormone-releasing patterns [4]. Experimental investigations are ongoing to better understand the exact mechanism of stress and reproductive function [5, 6]. Basic science studies indicate that higher stress levels result in poorer implantation rates due to reduced endometrial receptivity [5], reduced adrenergic receptor signaling in decidual cells [6], changes in uterine peristalsis [2, 7] and contraction frequency [2, 8]. Furthermore, numerous psychosocial studies have found that the physical and psychological burdens of the IVF process are the primary reasons for treatment postponement, discontinuation, and premature dropouts [9–11]. There is a great need for evidence-based stress reduction adjuvant tools during ET to improve patient experience and optimize IVF outcomes [2, 12, 13]. The use of complementary and alternative medicine to improve CPR has gained popularity among infertile populations despite systematic reviews and meta-analyses identifying controversial clinical evidence [13–15]. A small number of clinical trials have examined other alternative stress-relieving treatments and adjuvant interventions during IVF-ET to improve clinical outcomes. For example, a hypnosis case–control study of 194 IVF-ET patients [2], a retrospective andullation therapy (deep relaxation massage) study of 267 IVF-ET patients [16], and an RCT study of 219 IVF-ET patients using clown therapy [12] found significantly higher CPR in the exposure groups than in control groups. Researchers attributed the improved treatment outcomes to the relaxing effects of these therapies on reducing patients’ anxiety, calming uterine contractions, and enhancing endometrial receptivity for ET [2, 12, 16]. On the other hand, a harp therapy RCT of 181 IVF-ET patients [17], a music therapy RCT of 186 IVF-ET patients [18], and a music exposure RCT of 42 IVF-ET patients [19] found no significant differences in CPR between the experimental and control groups. Thus, the clinical effectiveness of these adjuvant interventions during IVF-ET on improving CPR remains inconclusive, with contradictory results which may stem from the heterogeneity of study designs, using self-reported [12] versus chart-collected clinical data, and issues in statistical analyses [20]. Furthermore, some adjuvant treatments require one-on-one intervention by certified professionals [2, 12, 17], which could be expensive and impractical to implement in confined medical settings such as IVF operating rooms. More convenient, cost-effective, easy-to-use adjuvant interventions during IVF-ET should be explored. The therapeutic use of virtual reality (VR) technology has gained medical attention over the past two decades, with growing evidence supporting its potential in the mental and physical health sectors [21–23]. VR is an advanced technological system that allows users to be immersed into a three-dimensional virtual environment using digitally generated multimodal stimuli that substitute real-world sensory perception [24]. Its therapeutic value has been explored in various clinical settings with promising results, including treatment for anxiety disorders [24, 25], phobias [26], social anxiety [27], posttraumatic stress disorder [28], eating disorders [29], pain management [22, 30, 31] and rehabilitation [32]. A recently published meta-analysis and systematic review of 11 VR studies involving 892 surgical adult patients identified the clinical benefit of VR distraction and exposure in reducing patients’ preoperative anxiety, heart rate and blood pressure [24]. The medical interest in using VR technology during minimally invasive gynecologic procedures has also been growing [33]. Recent clinical trial studies involving patients undergoing minor gynecological procedures [34, 35] found VR technology was effective in reducing acute pain and anxiety associated with procedures. Several other RCT studies found VR exposure did not result in significant pain relief during hysterosalpingography [36] and had no significant effects in reducing pain during oocyte retrieval procedures [37, 38]. In a large-scale RCT study investigating the effect of VR exposure with or without hypnosis among 342 patients before oocyte retrieval, no differences in anxiety levels and pregnancy rates were found between the study group (both VR and hypnosis exposure) and control group (VR exposure only) [39]. However, less is known about its clinical applicability among patients undergoing FET procedures, as no published clinical trial reports exist to date [33]. This single-site experimental RCT aimed to examine the effects of VR technology as a stress reduction adjuvant therapy during FET procedures. We hypothesized that VR exposure before FET would improve CPR (primary outcome) and reduce stress and anxiety levels in participants by calming their sympathetic autonomous nervous system (secondary outcome). Materials and methods Study protocols Figure 1 outlines the flowchart and timeline for this IRB-approved randomized controlled trial (VERITAS, #2396, 2019–03-01) registered on ClinicalTrials.gov (ID# NTC04394962) on 2020–05–06. All participants provided written informed consent for enrollment and had the opportunity to withdraw at any time. The inclusion criteria included: IVF patients aged 21–45 years undergoing FET using their own oocytes, or donated oocytes/embryos (where the oocyte provider age was between 21 and 45 years), for their own reproductive purposes. Exclusion criteria included any contraindication to VR technology such as epilepsy or a previous seizure, claustrophobia, current migraine, heart disease, visual and/or auditory deficits, use of a medical device such as a cardiac pacemaker or hearing aids, anxiety disorder or regular use of anxiety medication, and major uterine anomalies, Asherman’s syndrome, and/or uterine fibroids distorting the uterine cavity. Using a 2-sided significance level of p < 0.05 with 80% statistical power, a sample size of 350 patients (175 in each arm) would be required to detect changes in pregnancy rates from 30 to 45%. Study method flowchart and timeline Study recruitment and enrollment took place from May 2019 to May 2023, including pauses from April 2020 to August 2020 and December 2020 to February 2022, during the acute phases of the COVID-19 pandemic. In total, 869 IVF female patients were screened for eligibility. Of these, 685 were deemed eligible, invited to participate in the study, and presented with informed consent information, including details about random assignment, the timing of study measurements, and the nature of the VR intervention for those allocated to the study group. Of the 421 (61.4%) recruited, 71 were excluded for reasons including retrospective discovery of ineligibility (n = 21), withdrawal (n = 11) and FET cancellation (n = 39). A total of 350 (83.1% of recruited) were enrolled and completed the study (Fig. 1). The study coordinator used a computer-assisted block randomization method for allocation concealment to randomize participants into the VR and control groups in a 1:1 ratio and stratified by age groups (21–34 years and 35–45 years) prior to T2, on the day of FET. Heart rate and blood pressure were measured, saliva samples were collected using saliva collection devices (Salimetrics), and the State questionnaire of the State-Trait Anxiety Inventory (STAI) [40] was administered at baseline before starting FET protocols (T1) and on the FET day prior to transfer (T2). Participants were blinded to their group assignment at T1 and T2 measurements. Following T2 measurements, the study coordinator notified participants of the group they were randomized to and provided VR group participants with brief instructions on navigating the VR equipment.1 Only the nurses who provided direct patient care in the operating room were aware of the VR intervention; the embryologists who selected embryos for FET and physicians who performed FET were blinded to participants’ group assignment. VR participants then received 15–30 min of VR exposure to one of five possible VR environments of their choice (i.e., beach, underwater diving, natural vegetation, palm trees, and redwood forest) before the FET procedure. The VR exposure was passive; patients had full control of the VR headset using a hand-held device to navigate or stop at any point. Control group participants received routine care. Immediately after the FET procedure (T3), heart rate and blood pressure were taken, and the STAI-State questionnaire was administered on both group participants. The study coordinator asked participants about their VR experience (i.e., any unusual experience or adverse events, and feedback on the VR experience) and recorded their responses. Clinical data related to the following were collected from medical records and de-identified: participant characteristics, physical health, lifestyle, fertility history, embryo quality, FET cycle parameters, and pregnancy outcomes at 2, 6, and 12 weeks post-ET (see Table 1) in dichotomous (No/Yes), categorical or numerical format. The quality of the transferred blastocysts was assessed using a modified SART embryo scoring system (41): ‘Good’ grade for inner cell mass (ICM) grade A and trophectoderm (TE) trophectoderm (TE) grade A or B (AA or AB blastocysts); ‘Fair’ grade for ICM grade B and TE grade A, B or C (BB, BC, or BA blastocysts), and ‘Poor’ grade for any ICM grade C (CC or CB blastocysts). Table 1Patient and treatment characteristics of VR (n = 175) and Control (n = 175) groupsVariablesaAll (N = 350)VR Group (N = 175)Control Group (N = 175)p-valuebI. RecruitmentRandomization Control VR Total175 (50)175 (50)350 (100)0175 (100)175 (100)175 (100)0175 (100)N/ARecruitment yearc 2019 2020 2021 2022 2023 Total55 (15.7)28 (8.0)0186 (53.1)81 (23.1)350 (100)26 (14.9)14 (8.0)092 (52.6)43 (24.6)175 (100)29 (16.6)14 (8.0)094 (53.7)38 (21.7)175 (100)p = 0.920II. PATIENT CHARACTERISTICSPatient’s age at recruitment37.2 ± 4.4, 37 (21–48)37.2 ± 4.3, 37, 26–4537.1 ± 4.5, 37, 21–48p = 0.864Oocyte provider’s age at recruitment d35.3 ± 4.4, 35 (21–44)35.2 ± 4.3, 35, 23–4435.3 ± 4.6, 36, 21–44p = 0.923Ethnicity White Non-white Total212 (62.9)125 (37.1)337 (100)108 (63.9)61 (36.1)169 (100)104 (61.9)64 (38.1)168 (100)p = 0.704Relationship status Single Partnered Total28 (8.0)322 (92.0)350 (100)15 (8.6)160 (91.4)175 (100)13 (7.4)162 (92.6)175 (100)p = 0.694III. Physical health and lifestyle choicesBMI (kg/m2)25.7 ± 6.2, 24.3, 16.4–52.426.2 ± 6.4, 24.7, 16.9–49.425.3 ± 5.9, 24.2, 16.4–52.4p = 0.200Chronic disease No Yes Total158 (45.1)192 (54.9)350 (100)82 (46.9)93 (53.1)175 (100)76 (43.4)99 (56.6)175 (100)p = 0.519On regular medication No Yes Total220 (62.9)130 (37.1)350 (100)115 (65.7)60 (34.3)175 (100)105 (60.0)70 (40.0)175 (100)p = 0.269Exercise regularly No Yes Total51 (17.3)244 (82.7)295 (100)23 (15.2)128 (84.8)151 (100)28 (19.4)116 (80.6)144 (100)p = 0.339Smoking No Yes Total330 (94.8)18 (5.2)348 (100)163 (93.7)11 (6.3)174 (100)167 (96.0)7 (4.0)174 (100)p = 0.333Alcohol use No Yes Total176 (50.7)171 (49.3)347 (100)90 (52.0)83 (48.0)173 (100)86 (49.4)88 (50.6)174 (100)p = 0.628Recreational drug use No Yes Total327 (94.0)21 (6.0)348 (100)166 (95.4)8 (4.6)174 (100)161 (92.5)13 (7.5)174 (100)p = 0.260IV. Fertility historyGravidity No Yes Total133 (38.0)217 (62.0)350 (100)66 (37.7)109 (62.3)175 (100)67 (38.3)108 (61.7)175 (100)p = 0.912Parity No Yes Total277 (64.9)123 (35.1)350 (100)114 (65.1)61 (34.9)175 (100)113 (64.6)62 (35.4)175 (100)p = 0.911Infertility type Primary Secondary Other e Total130 (37.1)216 (61.7)4 (1.1)350 (100)65 (37.1)107 (61.1)3 (1.7)175 (100)65 (37.1)109 (62.3)1 (0.6)175 (100)p = 0.601Infertility etiology Female factor only Male factor only Female and male factors Unexplained Other e Total203 (58.0)18 (5.1)117 (33.4)8 (2.3)4 (1.1)350 (100)104 (59.4)10 (5.7)54 (30.9)4 (2.3)3 (1.7)175 (100)99 (56.6)8 (4.6)63 (36.0)4 (2.3)1 (0.6)175 (100)p = 0.729Infertility duration by month35.7 ± 26.2, 29, 0–16838.1 ± 25.8, 31.5, 0–10533.4 ± 26.4, 25, 0–168p = 0.105Recurrent implantation failure No Yes Total308 (88.0)42 (12.0)350 (100)157 (89.7)18 (10.3)175 (100)151 (86.3)24 (13.7)175 (100)p = 0.324Recurrent pregnancy loss No Yes Total294 (84.0)56 (16.0)350 (100)145 (82.9)30 (17.1)175 (100)149 (85.1)26 (14.9)175 (100)p = 0.560Total number of all types of pregnancy loss0.8 ± 1.3, 0, 0–100.9 ± 1.4, 0, 0–100.8 ± 1.1, 0, 0–6p = 0.470V. CYCLE CHARACTERISTICSUsed oocyte or embryo donationf No Yes Total318 (90.9)32 (9.1)350 (100)155 (88.6)20 (11.4)175 (100)163 (93.1)12 (6.9)175 (100)p = 0.138Oocyte provider’s total antral follicle count (AFC) d19.1 ± 13.0, 17, 0–10019.5 ± 13.9, 17, 0–10018.8 ± 12.1, 17, 0–50p = 0.588Oocyte provider’s AMH (pmol/L) d20.4 ± 19.2, 14.9, 0–127.821.0 ± 21.3, 15, 0–127.819.8 ± 16.8, 14.8, 0.1–92.2p = 0.585Total number of previous ovum pick-ups (OPU) g2.0 ± 1.4, 1, 1–142.0 ± 1.4, 1, 1–81.9 ± 1.5, 1, 1–14p = 0.336Had embryo transfer before or since last pregnancy No Yes Total194 (55.4)156 (44.6)350 (100)95 (54.3)80 (45.7)175 (100)99 (56.6)76 (43.4)175 (100)p = 0.667VI. EMBRYO CHARACTERISTICSTotal number of frozen embryos from the same cycle4.5 ± 3.4, 4, 1–214.6 ± 3.4, 4, 1–174.5 ± 3.4, 4, 1–21p = 0.813PGT-A No Yes Total92 (26.3)258 (73.7)350 (100)43 (24.6)132 (75.4)175 (100)49 (28.0)126 (72.0)175 (100)p = 0.466Number of embryos transferred 1 2 Total338 (96.6)12 (3.4)350 (100)172 (98.3)3 (1.7)175 (100)166 (94.9)9 (5.1)175 (100)p = 0.078Ploidy status of the transferred embryoh Euploid Mosaic Untested Total245 (70.0)18 (5.1)87 (24.9)350 (100)126 (72.0)9 (5.1)40 (22.9)175 (100)119 (68.0)9 (5.1)47 (26.9)175 (100)p = 0.683Quality of the transferred embryo of the frozen cyclei,j Poor Fair Good Total51 (14.6)198 (56.6)101 (28.9)350 (100)25 (14.3)107 (61.1)43 (24.6)175 (100)26 (14.9)91 (52.0)58 (33.1)175 (100)p = 0.170VII. FROZEN EMBRYO TRANSFER CYCLELuteal support Progesterone and estrogen Progesterone only Estrogen only No Luteal support272 (77.7)73 (20.9)2 (0.6)3 (0.9)130 (74.3%)41 (23.4%)1 (0.6%)3 (1.7%)142 (81.1%)32 (18.3%)1 (0.6%)0p = 0.200Endometrial thickness on the embryo transferred day9.9 ± 1.9, 9.6, 6.5–1810.0 ± 2.0, 9.5, 6.5–18.09.8 ± 1.8, 10, 7–17.0p = 0.339Catheter used Cook Guardia & TDT Tomcat Wallace Total124 (35.4)14 (4.0)138 (39.4)74 (21.1)350 (100)56 (32.0)8 (4.6)74 (42.3)37 (21.1)175 (100)68 (38.9)6 (3.4)64 (36.6)37 (21.1)175 (100)p = 0.538aData are presented as mean ± SD, median, range, or n (%). Missing data were not included in total countbp-values of the VR and control groups were assessed by t-tests, chi-square or Fisher’s exact test (where appropriate)cPatients were recruited from 05/2019–05/2023. Recruitment was paused several times during the acute phase of the COVID-19 outbreakdDonor’s age, AMH and AFC were used for oocyte/embryo donation cases. Otherwise, the participant’s age, AMH and AFC were usedeFour patients underwent IVF for reasons other than infertility, such as carriers of genetic diseases and fertility preservationfOne participant in the VR group used oocyte donation and two participants in the Control group used embryo donationgTotal count only includes patient’s own OPU; egg donation OPU was excludedhOne patient in the Control group received a double-embryo transfer of one euploid embryo and one mosaic embryo. This case was collapsed into the ‘euploid’ categoryiOf the twelve patients who received double-embryo transfer, the best embryo grade of either the first or the second embryo was used for categorizationjOnly six patients used day-3 embryos for their FET cycles Patient and treatment characteristics of VR (n = 175) and Control (n = 175) groups Control VR Total 175 (50) 175 (50) 350 (100) 0 175 (100) 175 (100) 175 (100) 0 175 (100) 2019 2020 2021 2022 2023 Total 55 (15.7) 28 (8.0) 0 186 (53.1) 81 (23.1) 350 (100) 26 (14.9) 14 (8.0) 0 92 (52.6) 43 (24.6) 175 (100) 29 (16.6) 14 (8.0) 0 94 (53.7) 38 (21.7) 175 (100) White Non-white Total 212 (62.9) 125 (37.1) 337 (100) 108 (63.9) 61 (36.1) 169 (100) 104 (61.9) 64 (38.1) 168 (100) Single Partnered Total 28 (8.0) 322 (92.0) 350 (100) 15 (8.6) 160 (91.4) 175 (100) 13 (7.4) 162 (92.6) 175 (100) No Yes Total 158 (45.1) 192 (54.9) 350 (100) 82 (46.9) 93 (53.1) 175 (100) 76 (43.4) 99 (56.6) 175 (100) No Yes Total 220 (62.9) 130 (37.1) 350 (100) 115 (65.7) 60 (34.3) 175 (100) 105 (60.0) 70 (40.0) 175 (100) No Yes Total 51 (17.3) 244 (82.7) 295 (100) 23 (15.2) 128 (84.8) 151 (100) 28 (19.4) 116 (80.6) 144 (100) No Yes Total 330 (94.8) 18 (5.2) 348 (100) 163 (93.7) 11 (6.3) 174 (100) 167 (96.0) 7 (4.0) 174 (100) No Yes Total 176 (50.7) 171 (49.3) 347 (100) 90 (52.0) 83 (48.0) 173 (100) 86 (49.4) 88 (50.6) 174 (100) No Yes Total 327 (94.0) 21 (6.0) 348 (100) 166 (95.4) 8 (4.6) 174 (100) 161 (92.5) 13 (7.5) 174 (100) No Yes Total 133 (38.0) 217 (62.0) 350 (100) 66 (37.7) 109 (62.3) 175 (100) 67 (38.3) 108 (61.7) 175 (100) No Yes Total 277 (64.9) 123 (35.1) 350 (100) 114 (65.1) 61 (34.9) 175 (100) 113 (64.6) 62 (35.4) 175 (100) Primary Secondary Other e Total 130 (37.1) 216 (61.7) 4 (1.1) 350 (100) 65 (37.1) 107 (61.1) 3 (1.7) 175 (100) 65 (37.1) 109 (62.3) 1 (0.6) 175 (100) Female factor only Male factor only Female and male factors Unexplained Other e Total 203 (58.0) 18 (5.1) 117 (33.4) 8 (2.3) 4 (1.1) 350 (100) 104 (59.4) 10 (5.7) 54 (30.9) 4 (2.3) 3 (1.7) 175 (100) 99 (56.6) 8 (4.6) 63 (36.0) 4 (2.3) 1 (0.6) 175 (100) No Yes Total 308 (88.0) 42 (12.0) 350 (100) 157 (89.7) 18 (10.3) 175 (100) 151 (86.3) 24 (13.7) 175 (100) No Yes Total 294 (84.0) 56 (16.0) 350 (100) 145 (82.9) 30 (17.1) 175 (100) 149 (85.1) 26 (14.9) 175 (100) No Yes Total 318 (90.9) 32 (9.1) 350 (100) 155 (88.6) 20 (11.4) 175 (100) 163 (93.1) 12 (6.9) 175 (100) No Yes Total 194 (55.4) 156 (44.6) 350 (100) 95 (54.3) 80 (45.7) 175 (100) 99 (56.6) 76 (43.4) 175 (100) No Yes Total 92 (26.3) 258 (73.7) 350 (100) 43 (24.6) 132 (75.4) 175 (100) 49 (28.0) 126 (72.0) 175 (100) 1 2 Total 338 (96.6) 12 (3.4) 350 (100) 172 (98.3) 3 (1.7) 175 (100) 166 (94.9) 9 (5.1) 175 (100) Euploid Mosaic Untested Total 245 (70.0) 18 (5.1) 87 (24.9) 350 (100) 126 (72.0) 9 (5.1) 40 (22.9) 175 (100) 119 (68.0) 9 (5.1) 47 (26.9) 175 (100) Poor Fair Good Total 51 (14.6) 198 (56.6) 101 (28.9) 350 (100) 25 (14.3) 107 (61.1) 43 (24.6) 175 (100) 26 (14.9) 91 (52.0) 58 (33.1) 175 (100) Progesterone and estrogen Progesterone only Estrogen only No Luteal support 272 (77.7) 73 (20.9) 2 (0.6) 3 (0.9) 130 (74.3%) 41 (23.4%) 1 (0.6%) 3 (1.7%) 142 (81.1%) 32 (18.3%) 1 (0.6%) 0 Cook Guardia & TDT Tomcat Wallace Total 124 (35.4) 14 (4.0) 138 (39.4) 74 (21.1) 350 (100) 56 (32.0) 8 (4.6) 74 (42.3) 37 (21.1) 175 (100) 68 (38.9) 6 (3.4) 64 (36.6) 37 (21.1) 175 (100) aData are presented as mean ± SD, median, range, or n (%). Missing data were not included in total count bp-values of the VR and control groups were assessed by t-tests, chi-square or Fisher’s exact test (where appropriate) cPatients were recruited from 05/2019–05/2023. Recruitment was paused several times during the acute phase of the COVID-19 outbreak dDonor’s age, AMH and AFC were used for oocyte/embryo donation cases. Otherwise, the participant’s age, AMH and AFC were used eFour patients underwent IVF for reasons other than infertility, such as carriers of genetic diseases and fertility preservation fOne participant in the VR group used oocyte donation and two participants in the Control group used embryo donation gTotal count only includes patient’s own OPU; egg donation OPU was excluded hOne patient in the Control group received a double-embryo transfer of one euploid embryo and one mosaic embryo. This case was collapsed into the ‘euploid’ category iOf the twelve patients who received double-embryo transfer, the best embryo grade of either the first or the second embryo was used for categorization jOnly six patients used day-3 embryos for their FET cycles Outcome measures Pregnancy rates were assessed based on biochemical beta hCG at 2 weeks (pregnancy rate), and ultrasonographic visualization of gestational sac(s) and fetal heartbeat at 6 (clinical pregnancy rate) and 12 weeks (ongoing pregnancy rate). Stress physiological parameters were assessed at baseline (T1), pre-FET (T2) and post-FET (T3) by measuring blood pressure and heart rate and administering the STAI-State questionnaire, a 20-item 4-point Likert scale measuring the state of anxiety (‘low’: 20–37, ‘moderate’: 38–44, and ‘high’: 45–80) (36). Cronbach’s alpha was 0.930 at T1, 0.937 at T2, and 0.926 at T3. Salivary alpha-amylase Samples were stored at − 80 °C for the duration of the study. Salivary alpha-amylase levels were analyzed using enzyme-linked immunosorbent assay (ELISA) in a subset of participants, following manufacturer’s instructions (Salimetrics). Alpha-amylase activity was analyzed in 90 sets of samples at T1, T2, and T3 (44 in the VR and 46 in the control groups) at three time points. Each sample was normalized to the control in the assay. Statistical analyses Data analysis was conducted using SPSS v.29 (IBM Corp., Armonk, NY, USA) using descriptive statistics, reliability tests, χ2 or Fisher’s exact (where applicable), independent samples t-tests, between-subject and within-subject repeated measures ANOVA, paired-samples t-tests for post-hoc pairwise comparisons, and bivariate logistic regressions and hierarchical logistic regressions to identify factors predicting the odds of pregnancy at 2 weeks, clinical pregnancy at 6 weeks (CP6wks) and ongoing clinical pregnancy at 12 weeks (CP12wks). Model assumptions (i.e., Hosmer and Lemeshow’s goodness-of-fit tests, and Levene’s, Box’s and Mauchly’s tests, p > 0.05) were checked prior to statistical interpretation. All statistical tests were two-sided, and p < 0.05 was considered statistically significant unless otherwise indicated. Results Participants The mean ages of study participants in the VR (N = 175) and control groups (N = 175) were similar (37.2 ± 4.3 vs 37.1 ± 4.5, p = 0.86); so was the mean oocyte provider age (35.2 ± 4.3 vs 35.3 ± 4.6, p = 0.92). Female factor was involved in 91.4% of the cases; 38% had no prior pregnancy history, and 64.9% were nulliparous. Primary infertility, recurrent implantation failure, and pregnancy loss comprised 37.1%, 12%, and 16% of the cases, respectively. Most participants used preimplantation genetic testing for aneuploidy (PGT-A) (73.7%), underwent autologous embryo transfer (90.9%), and underwent a single embryo transfer (96.6%). The quality of the transferred blastocysts was graded as ‘poor’ (14.6%), ‘fair’ (56.6%) or ‘good’ (28.9%) using a modified SART embryo scoring system [41]. Seventy percent of the FET cycles involved transferring a euploid embryo while the remainder transferred an untested (24.9%) or mosaic (5.1%) embryo. Participants in the VR and control groups had comparable clinical, treatment and cycle characteristics with no significant differences between them (p > 0.5) (Table 1). The mean duration of VR exposure was 26.1 ± 5.1 (range: 15–30 min). Primary outcome The overall pregnancy rates for all participants at 2, 6 and 12 weeks were 48.6%, 40.3%, and 35.7%, respectively (Fig. 2). Pregnancy rates at 2 weeks were higher in the VR group than in the control group (52.6% vs. 44.6%, p = 0.134). The VR group had significantly higher rates of CP6wks (45.7% vs. 34.9%, p = 0.038) and CP12wks (42.9% vs. 28.6%, p = 0.005) compared to the control group. Pregnancy rates of VR (n = 175) and Control (n = 175) groups at 2 weeks, 6 weeks and 12 weeks. Note: Data are presented as n (%). *p < 0.05; **p < 0.01 Univariate logistic regression analyses of three independent dichotomous pregnancy outcomes (Yes/No) at two, 6 and 12 weeks were first conducted using all 34 variables in Table 1 for hierarchical logistic regression analyses, which are shown in Table 2. Significant variables at a cutoff of p < 0.10 (see Supplementary Table 1) were entered into three hierarchical logistic regression models as predictors by block entry: patient characteristics (Block 1), cycle- and embryo-related characteristics (Block 2), and randomization (Block 3). For the pregnancy outcome at 2 weeks, seven significant variables at p < 0.10 in univariate analyses and one clinically meaningful variable (randomization, p = 0.135) were entered into the model: oocyte provider’s age at recruitment, infertility etiology (Block 1), oocyte provider’s antral follicle count, total number of previous ovum pick-up (OPU), total number of frozen embryos the transferred embryo originated from, ploidy status of the transferred embryo, quality of the transferred embryo (Block 2), and randomization (Block 3). In the final model, the three variables that were significant in predicting pregnancy at 2 weeks were: (a) having a lower number of previous OPU (OR 0.81, p = 0.035); (b) transferring a ‘good’ (OR 2.44, p = 0.028) or ‘fair’ grade embryo (OR 2.5, p = 0.012) as opposed to transferring a ‘poor’ grade embryo; and c) receiving VR exposure (OR 1.59, p = 0.047). The final model predicted 17.4% of the variance in pregnancy outcome at 2 weeks, with 1.4% contributed by VR exposure alone. The overall correct classification of the model was 65.7% (model sensitivity 61.0% and specificity 70.6%). Table 2Hierarchical logistic regression model of pregnancy at 2 weeks, clinical pregnancy at 6 weeks, and ongoing clinical pregnancy at 12 weeksVariablesaSignificant Variables at p < 0.10 cutoff in Univariate AnalysisHierarchical Logistic Regression ModelBLOCK 1Odds Ratio (95% Confidence Interval), p-valueBLOCK 2Odds Ratio (95% Confidence Interval), p-valueBLOCK 3 (Final Model)Odds Ratio (95% Confidence Interval), p-valuePregnancy at 2 weeksbNagelkerke R2 (ΔR2)N/A7.4%16% (8.6%)17.4% (1.4%) eBLOCK 1: Patient characteristicsOocyte provider’s age at recruitmentp = 0.065OR 0.95 (CI: 0.91–1.0), p = 0.058OR 1.0 (CI: 0.94–1.1), p = 0.933OR 1.0 (CI: 1.0–0.94), p = 0.964Infertility etiologyNoneFemale factor onlyMale factor onlyBoth female and male factorsUnexplainedP = 0.019RefOR 0.92 (CI: 0.13–6.6), p = 0.930OR 0.12 (CI: 0.03–5.57), p = 0.091OR 1.41 (CI: 0.28–27.25), p = 0.738OR 0.97 (CI:0.21–42.62), p = 0.981RefOR 1.07 (CI: 0.14–8.42), p = 0.946OR 0.12 (CI: 0.01–1.46), p = 0.096OR 1.82 (CI: 0.23–14.53), p = 0.572OR 0.90 (CI:0.07–11.06), p = 0.934RefOR 1.23 (CI: 0.15–10.25), p = 0.847OR 0.13 (CI: 0.01–1.66), p = 0.116OR 2.18 (CI: 0.26–18.47), p = 0.477OR 1.04 (CI:0.08–13.22), p = 0.977BLOCK 2: Cycle and embryo related characteristicsOocyte provider’s total antral follicle count (AFC)0.045–OR 1.0 (CI:0.99–1.02), p = 0.706OR 1.0 (CI:0.98–1.02), p = 0.779Total number of previous ovum pick-ups (OPU)0.010–OR 0.83 (CI:0.68–1.0), p = 0.049*OR 0.81 (CI:0.67–0.99), p = 0.035*Total number of frozen embryos the transferred embryo(s) originated from0.010–OR 1.08 (CI:1.0–1.17), p = 0.065OR 1.08 (CI:1.08–1.0), p = 0.064Ploidy status of the transferred embryoUntestedMosaicEuploid0.065–RefOR 1.07 (CI:0.62–1.85), p = 0.797OR 0.53 (CI:1.5–1.9), p = 0.325RefOR 1.06 (CI:0.61–1.83), p = 0.838OR 0.53 (CI:1.5–1.9), p = 0.329Quality of the transferred embryo of the frozen cyclePoor gradeFair gradeGood grade0.006–RefOR 2.5 (CI:1.22–5.1), p = 0.012*OR 2.3 (CI:1.05–5.06), p = 0.038*RefOR 2.5 (CI:1.22–5.13), p = 0.012*OR 2.44 (CI:1.1–5.4), p = 0.028*BLOCK 3: InterventionRandomizationControl GroupVR Group0.110c––RefOR 1.59 (CI:1.01–2.53), p = 0.047*Clinical pregnancy at 6 weekscNagelkerke R2 (ΔR2)N/A4.3%10.6% (6.3%)13.2% (2.6%) fBLOCK 1: Patient characteristicsInfertility etiologyNoneFemale factor onlyMale factor onlyBoth female and male factorsUnexplainedp = 0.068RefOR 1.87 (CI: 0.19–18.32), p = 0.590OR 0.38 (CI: 0.03–5.57), p = 0.476OR 2.75 (CI: 0.28–27.25), p = 0.386OR 3.0 (CI:0.21–42.62), p = 0.417RefOR 2.24 (CI: 0.22–22.69), p = 0.496OR 0.38 (CI: 0.03–5.85), p = 0.489OR 3.51 (CI: 0.34–36.07), p = 0.291OR 2.83 (CI:0.19–41.97), p = 0.450RefOR 2.83 (CI: 0.27–29.87), p = 0.387OR 0.45 (CI: 0.03–7.12), p = 0.570OR 4.65 (CI: 0.43–49.99), p = 0.204OR 3.60 (CI:0.24–55.17), p = 0.358BLOCK 2: Cycle and embryo related characteristicsTotal number of previous ovum pick-ups (OPU)p = 0.016–OR 0.81 (CI: 0.67–0.97), p = 0.022*OR 0.79 (CI: 0.65–0.95), p = 0.013*Total number of frozen embryos the transferred embryo(s) originated fromp = 0.053–OR 1.05 (CI: 0.98–1.13), p = 0.132OR 1.05 (CI: 0.98–1.13), p = 0.147Quality of the transferred embryo of the frozen cyclePoor gradeFair gradeGood gradep = 0.030–RefOR 2.13 (CI: 0.95–4.76), p = 0.065OR 2.12 (CI: 1.01–4.42), p = 0.046*RefOR 2.29 (CI: 1.01–5.18), p = 0.046*OR 2.13 (CI: 1.01–4.48), p = 0.046*BLOCK 3: InterventionRandomizationControl GroupVR Groupp = 0.039––RefOR 1.87 (CI: 1.18–2.96), p = 0.008*Ongoing clinical pregnancy at 12 weeksdNagelkerke R2 (ΔR2)N/A1.6%7.8% (6.4%)11.3% (3.5%) gBLOCK 1: Patient characteristicsPatient’s age at recruitmentp = 0.096OR 0.98 (CI: 0.92–1.05), p = 0.568OR 1.0 (CI: 0.94–1.07), p = 0.993OR 1.0 (CI: 0.94–1.07), p = 0.997Oocyte provider’s age at recruitmentp = 0.060OR 0.96 (CI: 0.91–1.03), p = 0.228OR 0.98 (CI: 0.92–1.05), p = 0.604OR 0.98 (CI: 0.92–1.05), p = 0.586BLOCK 2: Cycle and embryo related characteristicsTotal number of previous ovum pick-ups (OPU)p = 0.014––OR 0.82 (CI: 0.67–1.0), p = 0.047*OR 0.80 (CI: 0.65–0.98), p = 0.029*Total number of frozen embryos the transferred embryo(s) originated fromp = 0.017––OR 1.05 (CI: 0.97–1.12), p = 0.218OR 1.04 (CI: 0.97–1.12), p = 0.254Quality of the transferred embryo of the frozen cyclePoor gradeFair gradeGood gradep = 0.007–RefOR 2.80 (CI: 1.23–6.41), p = 0.015*OR 2.95 (CI: 1.22–7.16), p = 0.017*RefOR 2.78 (CI: 1.21–6.42), p = 0.016*OR 3.18 (CI: 1.29–7.82), p = 0.012*BLOCK 3: InterventionRandomizationControl GroupVR Groupp = 0.006–––RefOR 2.07 (CI: 1.30–3.30), p = 0.002***p < 0.05; **p < 0.01a ‘ref’ refers to the reference categorybSeven variables were found to be significant at the p < 0.10 cutoff from univariate logistic analyses of pregnancy at 2 weeks (see Supplementary Table 1). The ‘randomization’ variable is selected because of its clinical meaning although its p-value (0.110) is slightly more than the cutoff of p < 0.10. These eight variables were used for the hierarchical logistic regression analysiscSix variables were found to be significant at a p < 0.10 cutoff from univariate logistic analyses of clinical pregnancy at 6 weeks (see Supplementary Table 1). They were used for the hierarchical logistic regression analysisdSix variables were found to be significant at p < 0.10 cutoff from univariate logistic analyses of ongoing clinical pregnancy at 12 weeks (see Supplementary Table 1). They were used for the hierarchical logistic regression analysiseFinal model of pregnancy at 2 weeks: sensitivity 61.0% (108/177), specificity 70.6% (120/170), and overall correct classification 65.7%fFinal model of clinical pregnancy at 6 weeks: sensitivity 44.0% (62/141), specificity 80.7% (167/207), and overall correct classification 65.8%gFinal model of ongoing clinical pregnancy at 12 weeks: sensitivity 26.4% (33/125), specificity 88.3% (197/223), and overall correct classification 66.1% Hierarchical logistic regression model of pregnancy at 2 weeks, clinical pregnancy at 6 weeks, and ongoing clinical pregnancy at 12 weeks Infertility etiology None Female factor only Male factor only Both female and male factors Unexplained Ref OR 0.92 (CI: 0.13–6.6), p = 0.930 OR 0.12 (CI: 0.03–5.57), p = 0.091 OR 1.41 (CI: 0.28–27.25), p = 0.738 OR 0.97 (CI:0.21–42.62), p = 0.981 Ref OR 1.07 (CI: 0.14–8.42), p = 0.946 OR 0.12 (CI: 0.01–1.46), p = 0.096 OR 1.82 (CI: 0.23–14.53), p = 0.572 OR 0.90 (CI:0.07–11.06), p = 0.934 Ref OR 1.23 (CI: 0.15–10.25), p = 0.847 OR 0.13 (CI: 0.01–1.66), p = 0.116 OR 2.18 (CI: 0.26–18.47), p = 0.477 OR 1.04 (CI:0.08–13.22), p = 0.977 Ploidy status of the transferred embryo Untested Mosaic Euploid Ref OR 1.07 (CI:0.62–1.85), p = 0.797 OR 0.53 (CI:1.5–1.9), p = 0.325 Ref OR 1.06 (CI:0.61–1.83), p = 0.838 OR 0.53 (CI:1.5–1.9), p = 0.329 Quality of the transferred embryo of the frozen cycle Poor grade Fair grade Good grade Ref OR 2.5 (CI:1.22–5.1), p = 0.012* OR 2.3 (CI:1.05–5.06), p = 0.038* Ref OR 2.5 (CI:1.22–5.13), p = 0.012* OR 2.44 (CI:1.1–5.4), p = 0.028* Randomization Control Group VR Group Ref OR 1.59 (CI:1.01–2.53), p = 0.047* Infertility etiology None Female factor only Male factor only Both female and male factors Unexplained Ref OR 1.87 (CI: 0.19–18.32), p = 0.590 OR 0.38 (CI: 0.03–5.57), p = 0.476 OR 2.75 (CI: 0.28–27.25), p = 0.386 OR 3.0 (CI:0.21–42.62), p = 0.417 Ref OR 2.24 (CI: 0.22–22.69), p = 0.496 OR 0.38 (CI: 0.03–5.85), p = 0.489 OR 3.51 (CI: 0.34–36.07), p = 0.291 OR 2.83 (CI:0.19–41.97), p = 0.450 Ref OR 2.83 (CI: 0.27–29.87), p = 0.387 OR 0.45 (CI: 0.03–7.12), p = 0.570 OR 4.65 (CI: 0.43–49.99), p = 0.204 OR 3.60 (CI:0.24–55.17), p = 0.358 Quality of the transferred embryo of the frozen cycle Poor grade Fair grade Good grade Ref OR 2.13 (CI: 0.95–4.76), p = 0.065 OR 2.12 (CI: 1.01–4.42), p = 0.046* Ref OR 2.29 (CI: 1.01–5.18), p = 0.046* OR 2.13 (CI: 1.01–4.48), p = 0.046* Randomization Control Group VR Group Ref OR 1.87 (CI: 1.18–2.96), p = 0.008* Quality of the transferred embryo of the frozen cycle Poor grade Fair grade Good grade Ref OR 2.80 (CI: 1.23–6.41), p = 0.015* OR 2.95 (CI: 1.22–7.16), p = 0.017* Ref OR 2.78 (CI: 1.21–6.42), p = 0.016* OR 3.18 (CI: 1.29–7.82), p = 0.012* Randomization Control Group VR Group Ref OR 2.07 (CI: 1.30–3.30), p = 0.002** *p < 0.05; **p < 0.01 a ‘ref’ refers to the reference category bSeven variables were found to be significant at the p < 0.10 cutoff from univariate logistic analyses of pregnancy at 2 weeks (see Supplementary Table 1). The ‘randomization’ variable is selected because of its clinical meaning although its p-value (0.110) is slightly more than the cutoff of p < 0.10. These eight variables were used for the hierarchical logistic regression analysis cSix variables were found to be significant at a p < 0.10 cutoff from univariate logistic analyses of clinical pregnancy at 6 weeks (see Supplementary Table 1). They were used for the hierarchical logistic regression analysis dSix variables were found to be significant at p < 0.10 cutoff from univariate logistic analyses of ongoing clinical pregnancy at 12 weeks (see Supplementary Table 1). They were used for the hierarchical logistic regression analysis eFinal model of pregnancy at 2 weeks: sensitivity 61.0% (108/177), specificity 70.6% (120/170), and overall correct classification 65.7% fFinal model of clinical pregnancy at 6 weeks: sensitivity 44.0% (62/141), specificity 80.7% (167/207), and overall correct classification 65.8% gFinal model of ongoing clinical pregnancy at 12 weeks: sensitivity 26.4% (33/125), specificity 88.3% (197/223), and overall correct classification 66.1% For the CP6wks, the five variables that were significant at p < 0.10 in univariate analyses were entered into the model: infertility etiology (Block 1), the total number of previous OPU, the total number of frozen embryos the transferred embryo originated from, the quality of the transferred embryo (Block 2), and randomization (Block 3). In the final model, three variables were found to be significant in predicting CP6wks: (a) having a lower number of previous OPU (OR 0.79, p = 0.013); (b) transferring a ‘good’ (OR 2.13, p = 0.046) or ‘fair’ grade embryo (OR 2.29, p = 0.046) as opposed to transferring a ‘poor’ grade embryo; and (c) receiving VR exposure (OR 1.87, p = 0.008). The final model predicted 13.2% of the variance in clinical pregnancy at 6 weeks, with 2.6% contributed by VR exposure alone. The overall correct classification of the model was 65.8% (model sensitivity 44.0% and specificity 80.7%). For the ongoing CP12wks, the six variables that were significant at p < 0.10 in univariate analyses were entered into the model: patient’s age as well as oocyte provider’s age at recruitment (Block 1), total number of previous OPU, total number of frozen embryos the transferred embryo originated from, quality of transferred embryo (Block 2), and randomization (Block 3). In the final model, three variables were found to be significant in predicting CP12wks: (a) having a lower number of previous OPU (OR 0.80, p = 0.029); (b) transferring a ‘good’ (OR 3.18, p = 0.012) or ‘fair’ grade embryo (OR 2.78, p = 0.016) as opposed to transferring a ‘poor’ grade embryo; and (c) receiving VR exposure (OR 2.07, p = 0.002). The final model predicted 11.3% of the variance in CP12wks, with 3.5% contributed by VR exposure alone. The overall correct classification of the model was 66.1% (model sensitivity 26.4% and specificity 88.3%). Secondary outcomes Figure 3a–c shows the adjusted means of systolic and diastolic blood pressure, and heart rate after controlling for BMI. The adjusted means of the salivary alpha-amylase levels are shown in Fig. 4. The highest mean heart rates were recorded at T2 before FET (84.1 ± 0.8) of all participants compared to T1 (74.4 ± 0.6) and T3 (82.5 ± 0.8). Significant adjusted mean differences in systolic blood pressure (F(1.7, 591.4) = 3.5, p = 0.037), heart rate (F(1.5, 509.4) = 4.7, p = 0.018) and salivary alpha-amylase (F(2,176) = 20.3, p < 0.001) were found between T1 vs T2 (p < 0.001), T2 vs T3 (p < 0.001), and T1 vs T3 (p < 0.001) in post-hoc pairwise comparisons, although the diastolic blood pressure at different timepoints did not differ (p = 0.335). On the other hand, no interaction effects of time and randomization of these four physiological parameters were found, and the VR and control groups were observed to have similar patterns of systolic (p = 0.693) and diastolic (p = 0.835) blood pressure, heart rate (p = 0.629), and salivary alpha-amylase levels (p = 0.920). Heart rate, systolic blood pressure, diastolic blood pressure, and State-Trait Anxiety Inventory-State (STAI-State) Scale of VR (n = 175) and Control (n = 175) groups at T1 (baseline), T2 (pre-FET), and T3 (post-FET)a. aData are presented as adjusted mean ± standard error (95% confidence interval) for heart rate and blood pressure controlled for BMI. Greenhouse–Geisser’s Test with an adjusted degree of freedom was used when Mauchly’s Test of Sphericity assumption was violated. Post-hoc pairwise comparisons were conducted only when p < 0.05. For STAI-State scores, data were presented as mean ± standard deviation. bAdjusted means of systolic blood pressure controlling for BMI of all patients: 112.5 ± 0.7 (CI: 111.2–113.9) at T1, 121.6 ± 0.7 (CI: 120.2–123.1) at T2, and 124.1 ± 0.7 (CI: 122.7–125.5) at T3. Repeated measures of time: F(1.7, 591.4) = 3.5, p = 0.037*, ηP2 = 0.010; repeated measures of time and randomization: F(1.7, 591.4) = 0.3, p = 0.693, ηP2 = 0.001. Post-hoc pairwise comparisons of time: T1 vs. T2, p < 0.001***; T2 vs. T3, p < 0.001***; T1 vs. T3, p < 0.001***. cAdjusted means of diastolic blood pressure controlling for BMI of all patients: 80.5 ± 0.5 (CI: 79.5–81.6) at T1, 76.5 ± 0.5 (CI: 75.5–77.5) at T2, and 77.4 ± 0.5 (CI: 76.3–78.4) at T3. Repeated measures of time: F(1.7, 607.7) = 1.1, p = 0.335, ηP2 = 0.003; repeated measures of time and randomization: F(1.8, 607.7) = 0.2, p = 0.835, ηP2 = 0.000. dAdjusted means of heart rate controlling for BMI of all patients: 74.4 ± 0.6 (CI: 73.2–75.6) at T1, 84.1 ± 0.8 (CI: 82.6–85.6) at T2, and 82.5 ± 0.8 (CI: 80.8–84.1) at T3. Repeated measures of time: F(1.5, 509.4) = 4.7, p = 0.018*, ηP2 = 0.013. Repeated measures of time and randomization: F(1.5, 509.4) = 0.4, p = 0.629, ηP2 = 0.001. Post-hoc pairwise comparisons of time: T1 vs. T2, p < 0.001***; T2 vs. T3, p < 0.001***, T1 vs. T3, p < 0.001***. eSTAI-State questionnaire is a 20-item 4-point Likert scale with a total score ranging from 20 to 80. ‘Low’ anxiety—between 20 and 37, ‘moderate’ anxiety between 38 and 44, and ‘high’ anxiety—between 45 and 80. Cronbach’s alpha coefficients were 0.930 at T1, 0.937 at T2, and 0.926 at T3. fMeans of STAI-State questionnaire scores of all patients: 41.9 ± 0.6 at T1, 40.1 ± 0.6 at T2, and 36.1 ± 0.6 at T3. Repeated measures of time: F(1.8, 633.4) = 56.3, p < 0.001***, ηP2 = 0.139. Post-hoc pairwise comparisons of time: T1 vs. T2, p = 0.002**; T2 vs. T3, p < 0.001***; T1 vs. T3, p < 0.001***. Repeated measures of time and randomization: F(1.8, 633.4) = 3.8, p = 0.026*, ηP.2 = 0.011. Post-hoc pairwise comparisons of time and randomization (VR Group): T1 vs. T2, p = 0.012*; T2 vs. T3, p < 0.001***; T1 vs. T3, p < 0.001***. Post-hoc pairwise comparisons of time and randomization (Control Group): T1 vs. T2, p = 0.05; T2 vs. T3, p < 0.001***; T1 vs. T3, p < 0.001***. *p < 0.05; **p < 0.01; ***p < 0.001 Salivary alpha-amylase levels in VR (n = 44) and Control (n = 46) groups at T1 (baseline), T2 (pre-FET), and T3 (post-FET). aMeans ± standard error (95% confidence interval) of all patients: 4.0 ± 3.2 (CI: 3.4–4.7) at T1, 3.6 ± 0.3 (CI: 3.0–4.1) at T2, and 5.0 ± 0.3 (CI: 4.4–5.7) at T3. Repeated measures of time: F(2, 176) = 20.3, ηP2 = 0.187, p < 0.001**; repeated measures of time and randomization: F(2, 176) = 0.1, ηP.2 = 0.001, p = 0.920. Post-hoc pairwise comparisons of time: T1 vs. T2 (p = 0.027*); T2 vs. T3 (p < 0.001**); T1 vs. T3 (p < 0.001**). *p < 0.05; **p < 0.001 With regard to the subjective appraisal of anxiety assessed by the STAI-State Scale scores (Fig. 3d), ‘moderate’ anxiety is reported at T1 (41.9 ± 0.6) and T2 (40.1 ± 0.6) while ‘low’ anxiety is reported at T3 (36.1 ± 0.6). Significant mean differences in the main effect of time (F(1.8, 633.4) = 56.3, p < 0.001) were found between T1 and T2 (p = 0.002), T2 and T3 (p < 0.001) and T1 and T3 (p < 0.001) when all participants were analyzed. Furthermore, significant interaction effects of time and randomization were found (F(1.8, 633.4) = 3.8, p = 0.026), in which the anxiety scores of VR participants dropped significantly from T2 to T3 compared to the control group (mean difference of 5.2 vs. 2.7). Patient experience of VR intervention Four participants reported experiencing mild nausea and dizziness during VR exposure, but all completed the intervention. No unusual events were reported by the remainder of the participants. Overall, 141 participants (81%) had a good experience with the VR intervention and would recommend it before FET. Discussion Principal findings Findings from this RCT show that one-time VR exposure of 15–30 min before the FET procedure contributed significantly to improved pregnancy outcomes at 2, 6 and 12 weeks, and helped to reduce participants’ anxiety levels from pre-FET to post-FET when compared to the control group. The three factors that were found to be significant in predicting the odds of positive pregnancy outcomes at 2, 6 and 12 weeks in all three hierarchical logistic regression models were: (a) having a lower number of previous OPU, (b) transferring a ‘good’ or ‘fair’ grade embryo as opposed to a ‘poor’ grade embryo, and (c) receiving VR exposure before FET. Moreover, only 2% reported experiencing physical discomfort during VR exposure, and most participants had a positive VR experience and would recommend its use. Our findings add empirical evidence to a growing body of research investigating the clinical applicability of VR technology on patient experiences and its effects on improving clinical outcomes. Results in the context of what is known The number of patients using assisted reproductive treatments in Canada has grown by 64% between 2013 and 2022 [42]. To date, limited empirical data exist on the effects of various non-pharmacologic, auditory-visual-sensory adjuvant therapies during IVF-ETs. Levitas et al. (2006) used a case–control experimental design involving 194 IVF-ET patients to examine the effects of hypnosis, in which the experimental group received ET after being hypnotized to a trance state [2]. Friedler et al. (2011) used a quasi-randomized experiential design involving 219 patients with supplementary control group participants to study the effects of medical clowning following ET [12]. Okhowat et al. (2015) used a retrospective observational study to examine the effects of deep massage therapy before FET on 56 patients with a control group of 211 patients [16]. These three studies found the intervention groups had significantly higher CPR compared to their control groups, in which the exposures to hypnosis before ET [2] and medical clowning following ET [12] increased the pregnancy odds by 2.67 or 7.58 times, respectively. However, direct comparison of our CPR with existing studies that found either a significant increase [2, 12, 16] or no difference [17–19] in CPR would be misleading due to varied inclusion criteria and different IVF practice environments when the studies were conducted. Nonetheless, out of the studies listed, our RCT is much more methodologically rigorous, and our sample size of 350 participants is by far the largest among studies on this topic [2, 12, 16–19]. Our study tracked the pregnancy outcomes at three timepoints of 2, 6 and 12 weeks, while most of the existing studies [2, 12, 17] measured the ongoing CPR by the presence of a gestational sac on ultrasound at one timepoint. In this study, the VR group had a higher pregnancy rate at 2 weeks (52.6% vs. 44.6%), significantly higher CPR at 6 weeks (45.7% vs. 34.9%) and significantly higher ongoing CPR at 12 weeks (42.9% vs. 28.6%) compared to the control group. Receiving VR intervention was also found to be one of the three significant contributors of increasing the odds of pregnancy at 2, 6 and 12 weeks in multivariate regression analyses. This finding is novel as it suggests that the effectiveness of VR intervention, along with other important clinical factors such as the quality of the transferred embryo, was reducing the incidence of pregnancy losses following successful conception, although the exact mechanisms of how it works are not well understood. Moreover, the clinical and FET cycle characteristics of participants in the VR and control groups did not differ (see Table 1). The percentages of participants using PGT (75% vs. 72%), transferring euploid (72% vs. 68%) or a ‘fair’/‘good’ grade-quality blastocyst (86% vs. 85%) were also similar. Therefore, the higher pregnancy rate at 2 weeks, and significantly higher CPR at 6 and 12 weeks among VR participants could not be explained by the sample characteristics of the study and control groups. In 2022, 61.6% of the 18,119 autologous ET cycles occurring in Canada did not result in ongoing clinical pregnancies [42]. Therefore, experiencing anxiety is very likely when patients fear a possible unsuccessful ET outcome [43]. In this study, one of the stress physiological parameters, heart rate, was highest before FET for participants in both groups. The state of anxiety of all participants was ‘high’ at both baseline and pre-FET, confirming that ET is an anxiety-provoking procedure for many IVF patients [2, 43]. However, the subjective anxiety scores assessed by the STAI-State questionnaire of VR participants dropped more substantially and significantly from pre-FET to post-FET than those receiving routine care, thereby confirming the therapeutic effect of VR technology as an effective anxiety reduction tool from the participants’ perspective. In a clinical trial examining the use of harp therapy on anxiety reduction in IVF-ET patients using the same standardized anxiety scale as our study, they found the intervention group had a more significant drop of STAI-State scores from pre to post ET than those in the control group [17]. A recent RCT study with 30 patients undergoing planned oocyte cryopreservation also confirmed the therapeutic effect of VR exposure on preoperative anxiety reduction measured by STAI scores [44]. Similarly, our findings show that VR intervention was effective in reducing patients’ pre-FET anxiety by putting their minds in a more relaxed state to prepare for the procedure. Plausible explanations of the effectiveness of VR are its capacity to divert and distract patients from their current reality, reducing their attention focusing on the physical and mental aspects associated with the FET, and ultimately enhancing mindfulness and relaxation [45]. Research implications This is the first large-scale RCT to test the effectiveness of applying VR technology in ART. The next stage of this study is to conduct sub-analyses of participants with different clinical diagnoses and prognoses to compare the effectiveness of VR technology among different patient groups. Strengths and limitations Studies examining the effects of non-pharmacological adjuvant tools during FET are minimal. To our knowledge, there are no published clinical trial reports on the innovative use of VR technology as an adjuvant tool on improving FET outcomes. Not only is our study the first reported clinical trial on using VR technology in FET patients, but our sample size is also well powered, with 350 IVF-FET patients randomized prospectively into two arms on a 1:1 ratio, and participants were blinded to their group assignment at T1 and T2 measurements. One of the strengths of this study was that all the independent clinical variables used for analysis were collected from clinical records, thus minimizing subjectivity and self-report errors while ensuring data integrity and accuracy. Outcome variables were tracked at three time points; variability within subjects was eliminated when participants’ baseline measures were used as their own control. Our study had several limitations. First, the VR group received a one-time VR intervention before their FET procedure, which may restrict the effectiveness of VR therapy. Second, the primary outcomes were assessed by tracking the CPR at 2, 6 and 12 weeks following the specific FET procedure under investigation. We did not evaluate the cumulative pregnancy rate per IVF cycle or per patient. Live birth outcomes were not included, which would be a more important outcome measure than CPR. Third, participants’ stress levels were assessed using multiple objective (physiological parameters) and subjective (standardized STAI-State questionnaire) measures across three time points. Although participants were blinded to their group assignment at T1 and T2 measurements, unconscious bias in stress perception at T3 is possible once group allocation becomes known, which may introduce some threat to the internal validity of subjective stress appraisal. Finally, the RCT began in 2019, before the COVID-19 outbreak, and continued during the pandemic with several pauses as per public health safety restrictions at healthcare institutions. The pandemic presented IVF facilities with an unprecedented challenge when delivering fertility care. Patients undergoing fertility treatment during that period had been shown to experience adverse mental health sequelae [46, 47], although isolating the mental health impacts of COVID-19 was impossible with our methodology. Nonetheless, our randomized treatment vs. control groups were equally exposed to the pandemic; therefore, this is unlikely to affect our results. Conclusions Our findings demonstrate that VR technology is an effective adjuvant tool that is easily implemented in an IVF clinic, with the potential for improving FET outcomes and lowering patients’ anxiety during the FET procedure. Larger multi-site RCTs are needed to replicate this study to validate the clinical applicability of VR technology in FET procedures, and to better understand the physiological mechanisms of VR in reducing pregnancy losses following successful conception (i.e., how it works, why it works, and the required clinical conditions to make it work). Future studies should also examine if multiple VR exposures could further optimize the cumulative CPR and improve IVF patients’ overall quality of life and mental health well-being [48]. VR technology is built for repeated usage with costs increasingly affordable. It is an easy, non-invasive, and convenient option to offer patients as they prepare for their FET procedure. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 19 KB) Acknowledgements The authors wish to acknowledge Dr. Jordana Mashiach’s inputs to the study design, Katy Mousavi’s assistance with patient recruitment, Sahar Jahangiri for IRB coordination and study monitoring, Rocco Vari in the sourcing and set up of the VR equipment, and the nursing team at CReATe for their assistance with study protocol activities. Author contributions J.O.: recruitment, data collection, data interpretation and manuscript preparation; S.Y.: Data analysis, interpretation and manuscript preparation; M.D.: study design, recruitment, data collection, analysis and interpretation; A.R.: recruitment and data collection; L.M.H.: recruitment and data collection; A.B.: A.N. and A.G., recruitment and data collection, S.P. & S.D.: assistance with recruitment and data collection, ELISA; L.L.: ELISA, C.K.: data interpretation and manuscript preparation; K.G., P.S., A.B.: study design and study protocol support; A.G.F.: study supervision, data analysis and interpretation, manuscript review; C.L.: study design and supervision, data interpretation, and manuscript review. All authors approved the manuscript for submission. Funding The study was funded by CReATe Fertility Centre, through reinvestment of clinical earnings. Data availability De-identified study data or statistical analysis data will be shared, for approved purposes, upon request. Declarations Ethical approval VERITAS, #2396, 2019–03-01. Consent to participate All research participants provided written informed consent. Consent for publication All authors reviewed and approved this version of the manuscript and consented to its publication. Competing interests The authors declare no competing interests. Footnotes VR vendor Oculus Rift VR system includes the Rift Headset, two Oculus Sensors, and a pair of Oculus Touch controllers to provide an immersive 360-degree view of a three-dimensional VR world with built-in audio to enhance the experience. Attestation statement • The subjects in this trial were eligible to be concomitantly involved in another randomized trial that is not expected to impact outcomes. • Data regarding any of the subjects in the study has not been previously published. • The results of this clinical trial have not been previously published. • Data will be made available to the editors of the journal for review or query upon request. The original online version of this article has been revised. The author, Anjila Roumia, affiliated with CReATe Fertility Centre, Toronto, Canada, was inadvertently omitted from the author list. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Change history 8/3/2026 The original online version of this article has been revised. The author, Anjila Roumia, affiliated with CReATe Fertility Centre, Toronto, Canada, was inadvertently omitted from the author list. Change history 8/5/2026 A Correction to this paper has been published: 10.1007/s10815-026-03989-w Contributor Information Andree Gauthier-Fisher, Email: andree@createivf.com. Clifford Librach, Email: drlibrach@createivf.com. References Associated Data Supplementary Materials Supplementary file1 (DOCX 19 KB) Data Availability Statement De-identified study data or statistical analysis data will be shared, for approved purposes, upon request.