Stress Responses of Examiners during Ophthalmic Examination Practices in Healthy Young Students

Featured Application

Stress reactions were measured in students enrolled in Goldmann perimetry practice. Although there was no significant increase in the levels of stress-related substances, three types of stress reactions were observed. First, there was an increased sympathetic tone at the beginning of practice. Second, students with higher self-confidence showed higher cortisol concentrations immediately before practice. Third, students with longer test durations showed higher levels of ACTH secretion during practice.

Abstract

The stress reaction of trainees is an issue in the practices of medical-related examinations that involve real-time decision making based on the examiner–subject interactions. The Goldmann perimetry (GP) test is one of these examinations. To evaluate the students’ stress reactions in the practice of the GP test, the stress-related substances and heart rate variability were measured in forty students enrolled in the practice. While there was no significant increase in stress-related substances during the practice, significantly increased sympathetic activities were observed at the beginning of the tests. Moreover, the plasma cortisol measured before the tests showed a significant positive correlation to the students’ self-confidence scores, indicating the students, especially those with higher self-confidence scores, were anxious for upcoming tests with unfamiliar subjects. Once the tests began, they felt relieved in the procedures they had learned repeatedly. On the other hand, while the average plasma ACTH decreased significantly during the test, the ACTH secretion correlated positively to the test duration, indicating that the skillful participants had less stress during the test. In the medical-related practices, pre-training on how to deal with unfamiliar subjects may be helpful for reducing the stress of the trainees, in addition to the procedure itself.

1. Introduction

The stress reaction of trainees is an acute issue in the training of medical-related disciplines [1,2,3,4,5,6,7] and in various fields involving real-time interaction and decision making for human safety, such as driving a vehicle [8,9,10] or operating a nuclear power plant [11,12]. In certain ophthalmological examinations, the examiner is required to make real-time decisions based on the interactions between the patient and the examiner. For example, the accuracy of the Goldmann visual field kinetic perimetry (GP) test depends on the behavior of both the patient and the perimetrist [13,14,15,16,17]. In this test, the patient must maintain fixation at the center of the dome and respond appropriately when the visual target projected onto the dome becomes visible to him/her, while the perimetrist manually moves the target in centripetal directions on the dome (Figure 1). The perimetrist plots the target’s position on the report paper at the moment the patient responds. This procedure is repeated until the entire direction is covered, changing the starting point.

Because different target sizes and luminance give different ISO sensitivity curves (isopters), the perimetrist must choose the most appropriate one from many types of visual targets, which are classified by area (0–V (V is the largest)) and intensity (1–4 (4 is the brightest)); intensity between two numbers is subdivided into a–e (e is the brightest) to complete the test. The required number of isopters (namely, target types) depends on the patient’s visual field deficits. Therefore, perimetrists should decide on these during the test. Additionally, an adequate lens must be inserted before the subject at an appropriate timing. At the same time, the perimetrist should ensure that patients fully understand the test procedure during the initial explanation, encourage patients’ fixation throughout the test, and promptly monitor patients’ responses. These real-time face-to-face interactions can be difficult, especially for inexperienced examiners. Therefore, we hypothesized that learning an interactive ophthalmological examination such as the GP test may elicit stress reactions in student learners. As the first step to address this issue, we measured the autonomic and stress reactions in orthoptics students during GP test practices.

2. Materials and Methods

2.1. Participants

Healthy fourth- or third-year undergraduate students (n = 40; female = 33, male = 7) from an orthoptics-related department of a university were recruited as participant perimetrists. The average age was 21.43 ± 0.36 years (mean ± standard deviation) at the time of the experiments. Before the experiments, they had learned the basic procedures of the GP test for 26 or 29 months by conducting them with fellow students. In this study, however, the participant subjects of the GP tests were recruited from as many different departments’ students as possible to minimize the possibility of examining the subjects they had known well. If the participant subjects of the GP tests were recruited from the department’s undergraduate students the same as the participant perimetrist, students in school years other than the participant perimetrists were selected. Because we focused on the responses of the participant perimetrists during GP tests in this study, we hereafter call the participant perimetrists “participants”, unless otherwise mentioned.

2.2. Protocol

The GP test experiments were conducted on the 12th, 19th, and 21st of March 2020, 26 December 2020, and 1st to 3rd of September 2021 in the GP practice room of the Department of Orthoptics and Visual Sciences, Niigata University of Health and Welfare. Four to seven participants conducted the tests at the same time. Blood biochemistry and salivary amylase tests were performed before and after the GP tests. The electrocardiograms (ECGs) were recorded during the tests. The participants recorded the sounds throughout the experiments using their mobile devices to reconstruct the time course of each test and we instructed them to speak aloud the isopters they were about to measure. The tests started with the invitation of the subjects into the GP practice room. Only a single eye was tested for each subject. All GP tests took 17–45 min (average: 30.6 min) and were conducted at least 2 h after the last meal. The resting ECGs were recorded approximately one week after the GP test experiments. Then, comments on the test were gathered from the participants, especially around the time when the LF/HF ratios were high.

2.3. Biochemistry and Electrophysiology

Plasma levels of ACTH, cortisol, dopamine, adrenaline, and noradrenaline were measured as markers of stress and sympathetic tone [18,19,20,21,22,23] using enzyme-linked immunosorbent assays (for ACTH, ENZ-KIT138-0001 ACTH ELISA, Enzo Life Sciences, Inc., Farmingdale, NY, USA; for cortisol, RE52061 Cortisol ELISA, Immuno-Biological Laboratories Co., Ltd., Gunma, Japan; for dopamine, adrenaline, and noradrenaline, BA-E-6300, BA-E-6200, BA-E-6100, ImmuSmol SAS, Bordeaux, France, respectively). Blood biochemical data were obtained from 39 participants because 1 female participant refused blood sampling. The salivary amylase was measured in all participants using a salivary amylase monitor (CM-3.1, Nipro, Osaka, Japan) [24,25].

The R-R intervals (RRIs) and the low-frequency component to high-frequency component ratios (LF/HF ratio) were calculated to evaluate sympathetic tone [26,27,28]. GP was suitable for these evaluations because the perimetrists stayed on the stool almost all the time in this test (Figure 1a) and, thus, the influences from bodily movements [29] were small. Electrocardiograms were recorded using custom-made battery-powered amplifiers (ECG-128AE, O’hara, Tokyo, Japan). The signals were band-pass filtered (the cutoff frequencies: 1.5 and 100 Hz) and sampled at 500 Hz with an analog-to-digital converter (USB-6211, National Instruments, Austin, TX, USA), then stored in a laptop computer. The RRIs were interpolated using cubic Bézier curves and resampled at 10 Hz. Premature beats were manually eliminated before interpolation. From the resampled RRI data, the low-frequency (LF, 0.04–0.15 Hz) and high-frequency (HF, 0.15–0.5 Hz) components were calculated using fast Fourier transformation (FFT) with a 60-sec Hanning window sliding by 5 s. Each test was divided into three periods: (a) the period before the I/4e isopter measurement (“Before I/4e”), (b) the 10-min period after (a) (“Next 10 min”), and (c) the last 10 min of the test (“Last 10 min”). If the whole time of the test was short, these 10-min periods could overlay each other. We employed “Before I/4e” as the first period because it was the time when the participant completed the first isopter (V/4e, see Figure 1c,d) of the subject and had an outlook of the further stages of the undergoing GP test. The detection of the QRS complexes and interpolation were performed using a laboratory-made application [30]. The FFT analyses were performed using GNU R (https://www.r-project.org (accessed on 4 April 2022)).

2.4. Questionnaires

Before and after the GP tests, the participants were asked 28 (14 before the GP tests, 14 after the GP tests) questions using the visual analog scale. The questions consisted of general ones (self-confidence in the GP test and self-evaluation of the GP test he/she had finished) and other GP test-specific ones. We collected the answers from all the participants. In this study, only the answers to two general questions were analyzed.

2.5. Statistical Analyses

Unless stated otherwise, statistical significance was set at p < 0.05. The statistical calculations were performed using GNU R. We employed a paired t-test to compare the concentrations of a substance sampled from the same participant before and after practice. One-way repeated measures analyses of variance (ANOVAs) and post hoc t-tests with adjustments for multiple comparisons were employed for participant-wise comparisons of heart rate variability (HRV) indices between the four time periods. Pearson’s product moment correlations were employed for raw data or their subtractions. Spearman’s rank correlation tests were used for the responses to questionnaires (the VAS results) and the ratios of two raw data. The repeated measures ANOVAs and the post hoc t-tests with adjustments for multiple comparisons were performed using ANOVA-kun (http://riseki.php.xdomain.jp/index.php?ANOVA君 (accessed on 12 May 2023)) on GNU R.

3. Results

In spite of our assumption that there might be increased secretion of stress-related substances, there were no significant increases in salivary amylase or plasma catecholamines during GP tests (Figure 2a–d). Moreover, the plasma ACTH level significantly decreased during the test (p = 0.0029, paired t-test, Figure 2e). Nevertheless, the increase in cortisol level was significantly related to an increase in ACTH level (p = 0.00017, ρ = 0.575, Spearman’s rank correlation coefficient, Figure 2g). Additionally, there was a significant correlation between the change in mean RRI during the test from the resting period and plasma dopamine increase (i.e., the more the HR increased, the less dopamine was secreted, p = 0.0324, r = 0.343, Pearson’s product moment correlation, Figure 3).

Next, we evaluated the LF/HF ratios and RRIs during the tests. As exemplified in Figure 4a, the LF/HF ratio exceeded the mean value in the resting time (dotted black line) for most of the test; occasionally, it exceeded the resting value + 6σ (dashed black line). Excesses were mostly observed at the beginning of the GP test and then gradually decreased as the test progressed, with several periodic surges. These surges corresponded to small accidents or decision making the participants encountered in the GP test. For example, the participant felt uneasy when she completed the explanation of the test while the other participants were continuing explanations (arrowhead 1), when she missed expected responses from the subject (arrowhead 2), when she decided to insert the corrective lens before the subject (arrowhead 3), and, finally, when she misguided the initial position of the blind spot investigation (arrowhead 4). Similarly, the RRIs (solid blue line), another index of sympathetic tone, were short at the beginning of the GP test and gradually became longer as the test progressed. In the pooled data (Figure 4b,c), both indices were the highest before the I/4e measurement. To obtain pooled data, we compared the LF/HF ratios and RRIs within the three periods of the GP tests (see Materials and Methods). One-way repeated measures ANOVAs revealed significant main effects of the period on both LF/HF ratio and RRI (SS = 812850.14, df = 3, MS = 270950.05, F = 89.87, p < 0.0001; SS = 85.7032, df = 3, MS = 28.57, F = 24.89, p < 0.0001, respectively; the p-values of Mendoza’s multisample sphericity tests were <0.0001 for both). In both HRV indices, significant differences were found in all the combinations between four periods (three GP periods and resting) except for “Next 10 min” and “Last 10 min” (post hoc t-tests with Shaffer’s modified sequentially rejective Bonferroni procedure; the alpha level was set to 0.05).

As shown in Figure 5, the ACTH increase ratio during the GP tests correlated significantly with the difference in the LF/HF ratios between “Before I/4e” and “Next 10 min” (p = 0.0226, ρ = −0.366, Spearman’s rank correlation coefficient). In this regard, there was a significant correlation between the self-confidence scores of the participants and cortisol levels before the GP tests (p = 0.00521, ρ = 0.489, Spearman’s rank correlation coefficient, Figure 6), suggesting that participants with higher mental preparation might secrete more cortisol prior to the tests. On the other hand, the increase in ACTH correlated significantly with the duration of the tests (p = 0.0151, r = 0.386, Pearson’s product moment correlation, respectively, Figure 7), suggesting that the longer the test duration, the more participants were stressed.

There were many comments regarding anxieties about unfamiliar subjects who they were going to test or were testing (

Table 1. The comments of the perimetrist participants (those from the multiple participants only).

Periods	Comments
Before I/4e	I was nervous at first. (5) I was nervous about explaining the test to the patient. (5) I was not confident whether the eye patch was fixed appropriately to the patient’s eye. (2) I had never tested an unknown patient. (1) I wondered whether a corrective lens would be required or not. (1) I had never experienced such a situation where multiple tests were conducted in the same room. (1) I received no response, although the target should have been seen. (1) I attempted to start the test without an eye patch. (1)
Next 10 min	I found that I had misunderstood the target position for blind spot detection.
Middle	I wondered whether the corrective lens was required or not. (2) I had to change the corrective lenses a few times because of unexpected results. (2) Were intermediate isopters needed to be plotted? (2) I could not detect the isopter (*1) as expected. (1) I could not detect the blind spot. (1) I was nervous when asking the patient “do you want a rest?” (1)
Last 10 min	Was the test completed now? (3) I worried about the contours of the detected isopters. (3) I forgot to speak out voice for recording. (3) Was the target position appropriate for the detection of the blind spot? (3) I was uneasy to continue the test because the other perimetrist participants were about to complete their tests. (2)
General	Sometimes, I was puzzled by unexpected responses from the patient. (1)

The numbers in parentheses indicate the number of similar answers, excluding the displayed ones. All the comments were originally in Japanese. “Middle” period indicates the period between the end of “Next 10 min” and the beginning of “Last 10 min” (Figure 4a). If these two periods are overlapping, the answers within them are classified as “Middle”. “General” indicates the comments without specific references to time. *1: I/1e.

), especially in the first period of the tests (“Before I/4e”). Comments related to the patients decreased after this period; those related to the GP test itself (the contours of detected isopters, timing of corrective lens insertion, unexpected responses from the patients, etc.) increased (“Next 10 min” and later).

4. Discussion

We measured the stress responses in students conducting a practice of the GP test, one of the most important and frequently conducted ophthalmic tests. There was no significant release of the plasma stress-related hormones during the test (Figure 2), indicating that our student practices are appropriately conducted in terms of stress response as a whole. In our usual practices, stress reactions should be less than the current results because the trainees are dealing with more familiar subjects (i.e., their classmates). It is unlikely that the reduction in ACTH resulted from the negative feedback control of ACTH secretion caused by the increased plasma cortisol. First, there was no significant increase in cortisol levels. Second, the increase in cortisol levels was significantly related to the increase in ACTH. Additionally, there was a significant correlation between the change in mean RRI during the test from the resting period and plasma dopamine increase.

Participants with better self-confidence were more anxious at the beginning of the test and their stress-reactions disappeared as they gradually felt relieved as the test progressed. Moreover, our data on the ACTH secretion versus the test duration, which were about two times longer than those of experienced perimetrists, suggested that the skills and stress responses of participants during the GP tests correlated negatively. Therefore, it is unlikely that these responses were nonspecific to the experimental procedures. Instead, these data can be explained well if participants were anxious about upcoming GP tests before the first blood sampling; once the GP tests began, they felt rather relieved conducting the procedure that they had practiced many times. This interpretation was supported by comments from participants that consisted of many items regarding anxieties about unfamiliar subjects.

Based on their psychophysical nature, lots of ophthalmological examinations, other than the GP test, require interactions between the patient and the examiner and the examiner’s real-time decisions based on them. Therefore, good practice in these examinations with divergent subjects is essential. Belinda et al. measured the salivary cortisol and HR in physiotherapy students undergoing clinical education in laboratories with standardized patients or a hospital with real patients [31]. They discovered a temporally increased HR before the encounter with the patient under both conditions but no significant changes in salivary cortisol concentration, which supports our findings. Some studies recorded HRVs from students enrolled in medical-related degrees using ambulatory [32] or wearable [33] ECG devices and reported increased sympathetic activity during in-turn clinical practice in hospitals. However, in these studies, data were taken from three-hour periods, which consisted of various hospital procedures. In contrast, our study focused on the effect of a single ophthalmological test, which is usually completed in less than one hour. Therefore, our data may indicate the importance of repeated training in a particular test to relieve stress in students. This interpretation is supported by a study of stress response in surgical education, in which electrodermal activity during laparoscopic surgery decreased with postgraduate years [34].

Stress is a prevalent problem in specialists, such as information technology workers [35], especially those who work under extended work shifts [36], leading to possible physical illnesses [37]. Physical problems and stress-related vital signs, including HR and socio-emotional competencies, affect students’ performance in medical-related disciplines [38]. Peer training, rather than individual training, reduces the stress on trainees in nursing [39]. Unfortunately, this stress-relieving method cannot be applied to our case because a GP test must be conducted by a perimetrist alone in front of the patient. Although there are no objective methods to quantitatively evaluate each participant’s skill in the GP test, our data from the ACTH versus test durations, which were about two times longer than those of experienced perimetrists, suggested that the skill and stress of participants during the test correlated negatively. There are two possible explanations for this finding. If the lower skill causes stress reactions, more intensive training of the GP test might be helpful. Otherwise, when the stress reactions lower the skill, the test skill might be improved through stress management training [40]. Ubukata et al. newly proposed an objective method for evaluating the test results using a GP trainer [41]. Using such a device, we will be able to address this issue. According to recent studies, stress reactions depend on UCLA questionnaire scores of the subject [42] and gender [43]. Studies with a larger number of participants classified by appropriate questionnaires and gender may be required to further investigate the effects of GP tests on the examiner’s stress reactions. On this basis, we should teach students how to cope with unfamiliar subjects. In addition, more intense off-house training should be performed. Since the university is planning to elongate the internship periods, the effect of off-house training on the participants’ stress should be further investigated. Guidance on students’ lifestyles may be helpful for improving self-esteem and relief from stress reactions [44].

There are certain limitations in this current study. Most importantly, our data might underestimate the stress reactions because of an inevitable sampling bias (students who were unconfident in the GP test would not volunteer to the experiment). Second, the resting levels of stress-related substances in the plasma were not measured in this study. Because of these limitations, our data might underestimate the stress reactions. It should also be noted that this study was conducted during the COVID-19 pandemic period. Since we held most lecture classes online at that time, every practice was precious time for students to see their classmates face-to-face. This might reduce the stress reaction in the practice.