Relevance. Clinical reasoning is a crucial component of a physician's professional activity, directly impacting the quality of care provided. Applying clinical reasoning in a real-life setting can often be challenging due to physicians' emotional states – anxiety, fear of making mistakes, and a sense of responsibility for their actions. Medical education requires research and application of training and technologies that promote the development of emotional management skills, including modern, high-tech virtual reality simulators designed to fully immerse physicians in the clinical environment. Such technologies can create a controlled stressful situation in which trainees can experience emotions similar to those encountered in a real-life patient interaction. Can a virtual simulator truly not only develop professional skills and clinical competencies, but also help physicians learn to manage their own anxiety?
The aim – comparison of state anxiety indicators of residents before and after working on a virtual reality simulator simulating the situation of a scheduled outpatient appointment.
Material and methods. A prospective, randomized, controlled study with blinded data analysis was conducted in January–May 2024. Research participants (RP) were volunteer second-year residents in the following specialties: 31.08.36 Cardiology; 31.08.49 Therapy; 31.08.53 Endocrinology; 31.08.35 Infectious Diseases; 31.08.42 Neurology; 31.08.56 Neurosurgery; 31.08.01 Obstetrics and Gynecology; 31.08.02 Anesthesiology-Resuscitation; 31.08.68 Urology; 31.08.18 Neonatology; 31.08.67 Surgery. The total number of RP (screening) was 118 people, 107 people were included in the study. Before the study, the RPs familiarized themselves with its terms and conditions, ensured their anonymity, and signed voluntary informed consent. After signing, randomization was carried out using the «envelope method» in order to create two groups that did not contact each other during the study: Group 1 – control (traditional training without the use of a VR simulator, n=57) and Group 2 – study (training using a VR simulator, n=50). Participants in Group 2 were asked to use VR glasses to complete a case in virtual reality, which involved an outpatient appointment with a patient. After completing the case, each RP was asked to conduct an outpatient appointment with a real person – a simulated patient. Before completing the virtual case, both groups' trait and state anxiety levels were measured; after completing the virtual case, a second measurement of SA levels was conducted for participants in Group 2. Trait and state anxiety levels were measured using the Spielberger State-Trait Anxiety Inventory (STAI) as adapted by Yu.L. Khanin.
Statistical analysis of the data was performed using Statistica v.13 and Excel. Data quality was checked using distribution histograms. The Shapiro–Wilk method was used to determine the type of data distribution. The main methods of statistical data analysis were: for comparative data analysis – the Wilcoxon test and/or the sign test; for comparing the parameters of independent samples (control and study groups) – the Mann–Whitney test. Results are presented as M±SD, where M is the arithmetic mean, SD is the standard deviation; Me is the median, and CI are confidence intervals, including the minimum and maximum values (min & max) or the and 10th, 95th percentiles for an informative presentation of the studied sample. Dichotomous and ordinal qualitative data are expressed as frequencies (n) and percentage distributions of features. Differences were considered statistically significant at p<0.05.
Results. The RP and control groups did not differ significantly in age, education level, gender, level of trait anxiety, or experience in examining patients. Comparison of the state anxiety (SA) parameter revealed a statistically significant decrease after Group 2 worked on the virtual simulator: before training – M1=39.06, CI (5–95%) 36–42; after working on the simulator – M2=36.3, CI (5–95%) 33–39; p=0.001. The SA indicator after working on the simulator decreased in 70% of participants (35 out of 50 measurements). Among the participants with a decreased parameter, 9 RP (25.7%) experienced a decrease in the SA indicator by 10 or more points, 12 RP out of 35 (34.3%) – by 5–10 points, and 14 RP (40%) – by 1–4 points. In 26% of RP in group No. 2, the SA indicator after working on the simulator increased, while in 4% of RP it did not change. In the case “Tumor of the head of the pancreas, mechanical jaundice”, which was completed by residents of the specialties “surgery”, “therapy”, “infectious diseases”, “neurology”, “neurosurgery”, “obstetrics and gynecology”, “urology” and “neonatology” (n=28), the SA indicator, after a single virtual outpatient appointment, statistically significantly decreased by 4.8% [M1=41.5, CI (5–95%) 37–45 versus M2=39.5, CI (5–95%) 35–43, p=0.045].
Cardiologists and anesthesiologists who completed the case “Arterial hypertension stage II” (n=18) also demonstrated an even greater statistically significant decrease in the SA parameter by 13.6% [M1=35.3, CI (5–5%) 29–40 vs M2=30.5, CI (5–95%) 26–34, p=0.001]. Endocrinologists (n=4) who dealt with the case “Thyrotoxicosis with diffuse goiter, newly diagnosed atrial fibrillation” demonstrated, on the contrary, an increase in SA by 5.6% [M1=38.8, CI (5–95%) 34–43 vs M2=41, CI (5–95%) 28–53], but statistical analysis showed that the differences are not significant (p=0.46). In 5 out of 10 cases, the frequency of answering “Absolutely true” to questions implying a positive assessment of one's state increased, while in one case it remained unchanged. Participants chose “Absolutely true” less frequently when answering questions such as “I am calm”, “I am not in any danger”, “I feel a sense of inner satisfaction” and “I am self-confident”. In 3 out of 10 cases, the frequency of choosing “No, that's not true” to questions implying a negative assessment of one's state increased, while in two cases it remained unchanged, and in five cases it decreased. Statistical analysis showed that the differences in the frequency of answer selection when comparing the overall frequency indices in both groups of questions are not significant (in the group of questions implying a positive assessment of one’s condition: Me1=12, Q1–Q3=11–13 vs Me2=11, Q1–Q3=9–16; p=0.85; in the group of questions implying a negative assessment of one’s condition: Me1=39, Q1–Q3=37–48 vs Me2=39, Q1–Q3=36–44; p=0.52).
Conclusion. After completing a single virtual reality case, the state anxiety of resident physicians statistically significantly decreased in 70% of participants, with the majority (60%) experiencing a decrease in SA of 10 points or more.
A comparison of SA by case nosology revealed a statistically significant decrease in this indicator during the surgical and cardiology cases by 4.8% and 13.6%, respectively.
A detailed analysis of the participants' emotional state in response to individual questions requires additional psychological research, likely with an increase in the frequency of virtual simulator sessions.