Does Virtual Reality Increase Simulator Sickness During Exposure Therapy for Post-Traumatic Stress Disorder?

Background and Introduction

Exposure therapy through virtual reality (VR) has shown efficacy in the treatment of trauma and anxiety disorders. ¹ However, patient use of VR can result in temporary simulator-related side effects (SSEs) in symptom clusters such as oculomotor problems, disorientation, and/or nausea. ² Accordingly, the measurement of SSEs is an integral component of competent and ethical use of virtual reality exposure (VRE). ³

However, common SSEs overlap with symptoms of anxiety. ⁴ For example, SSEs include sweating, nausea, headache, vertigo, and confusion. These reactions overlap considerably with symptoms of anxiety, which may include sweating, dizziness or lightheadedness, fainting, and indigestion or discomfort in the abdomen. ⁵

Exposure therapy is a widely used evidence-based treatment for trauma and anxiety disorders ⁶ that involves the patient's intentional confrontation of feared stimuli. ⁷ As patients end maladaptive avoidance, state-based symptoms of anxiety are common. This is a problem for researchers conducting VRE. Providers need to know whether patient reactions represent negative side effects of VR use or the expected anxiety symptoms of patients who are ending maladaptive avoidance during exposure therapy, particularly early in the course of treatment. We are not aware of any research that compared the longitudinal patient-reported SSEs among patients receiving an exposure-based treatment who do and do not use VR.

Materials and Methods

This study represents a secondary analysis of data from a previously published trial ⁸ that compared the efficacy of prolonged exposure (PE) and VRE for soldiers with post-traumatic stress disorder (PTSD). Throughout the trial, we assessed SSEs in both treatment groups. A total of 108 soldiers were randomized to PE or use of the Virtual Iraq/Afghanistan system. ⁹ Fifteen patients did not participate in enough sessions to begin the exposure components of therapy or did not provide any data on the symptom measures. As a result, this study used data from 93 (PTSD; PE = 44; VRE = 49) participants. The Simulator Sickness Questionnaire (SSQ) ² is a validated 16-item self-report measure and one of the most widely used assessments of SSEs. To make administration at every session feasible, this study adapted the original measure by selecting only one symptom from each of the three factors of the original measure (occulomotor, disorientation, and nausea). Two of the items used in this study assessed the symptom of dizziness (eyes open/eyes closed), one item assessed for headache, and a fourth item inquired about sickness in the stomach. Patients placed an “X” in one of the four boxes for each item: None, Slight, Moderate, or Severe. For this study, we considered a patient endorsement of Moderate or Severe to reflect the presence of SSE. The brief SSQ was administered at the end of each of 10 treatment sessions. Data were analyzed using generalized estimating equations with a log link and a Poisson error distribution. The outcome variable for each treatment session was the count of symptoms identified as moderate/severe (score of 0–4 possible). Data for a total of 564 treatment sessions were recorded.

Results

Before the first use of VR for therapeutic trauma exposure (which occurred in session 3), there was no significant difference in the prevalence of SSEs among participants in the VRE group (23.58%) compared with those in the PE group (32.24%; prevalence risk ratio = 0.73, 95% confidence interval [CI] = 0.51–1.05). The frequency of moderate or severe SSE by treatment group and session are available in Table 1. The degree of linear relationship between baseline SSE frequency and baseline anxiety scores was 0.41 (95% CI = 0.22–0.60). There was no difference in the frequency of moderate/severe SSEs at the start of imaginal exposure (marginal mean PE = 0.48, marginal mean VRE = 0.46, rate ratio [RR] = 0.96, 95% CI = 0.60–1.53). Moderate/severe SSE frequency decreased by ∼10% for each subsequent session among the participants of the PE group (RR_within = 0.90, 95% CI = 0.84–0.97). Controlling for age, gender, baseline anxiety symptoms, and SSE symptom counts at the first two sessions of therapy, the rate of SSE reduction for participants in the VRE group (RR_within = 0.94, 95% CI = 0.89–1.00) was not statistically distinguishable from that of the PE group (RR_between = 1.04, 95% CI = 0.95–1.15). The minimum detectable RR estimate between study groups was 1.10, suggesting adequate power to detect a small difference if it occurred in the data.

Conclusions and Discussion

This study found that VR did not account for any differences in self-reported simulator sickness among soldiers receiving the same exposure therapy with or without VR. Consistent with previous studies of the relationship of anxiety to SSEs, these results may suggest caution in the interpretation of SSEs in the context of clinical VRE. Future studies might examine the extent to which the constructs of SSE and anxiety are unique, and new measures that assist clinicians in accurately distinguishing between anxiety and SSEs may be needed. Limitations include the unknown validity of the brief adaptation of the SSQ, the possibility of different constructs being assessed by treatment groups, and the assessment of SSE at the end of each session (∼30 min post-VR use).