Use of virtual reality for symptom management in solid-tumor patients with implications for primary brain tumor research: a systematic review

Abstract

Background

Primary brain tumors (PBTs) remain incurable, with a typically poor prognosis and significant symptom burden for patients. Virtual reality (VR) can potentially alleviate some of the negative aspects of illness by allowing individuals to escape to environments where they can experience more positive thoughts and emotions. Given promising findings for VR use in other clinical populations, there is increasing interest to use VR for symptomatic improvement in oncology patients. The purpose of this review was to analyze the literature of VR-related interventions for symptom management in adult PBT and other solid-tumor patients, which will guide development of future VR interventions in these populations.

Methods

A systematic search of EMBASE, PubMed, Scopus, and Web of Science was performed using PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines with predefined eligibility criteria. Thirteen studies met the inclusion criteria and were selected for review.

Results

Findings showed promising evidence that VR can improve anxiety for solid-tumor patients, with mixed results reported for pain, distress, depression, and mood. There was significant heterogeneity in methodological approaches across the literature and the majority of studies were underpowered and lacked rigorous study designs. Qualitative findings demonstrated a high degree of participant satisfaction with VR use, with very few adverse side effects reported.

Conclusions

Findings from this review suggest that VR can be used as an innovative delivery system for targeted interventions to improve symptoms in PBTs and other solid-tumor patient populations, though additional well-designed clinical trials are needed to better establish its efficacy.

Although primary brain tumors (PBTs) are relatively rare compared to other cancers, an estimated 86 970 new cases of PBTs are expected to be diagnosed in the United States in 2019¹, with approximately 700 000 individuals currently living with these tumors.² Despite decades of laboratory and clinical research in this field, the majority of PBTs remain incurable with a typically poor prognosis,³ and patients often experience significant symptom burden from the time of diagnosis through treatment and survivorship. A recent study by Armstrong et al⁴ found that PBT patients were highly symptomatic, with more than half of the sample population reporting at least 10 concurrent symptoms and 40% reporting at least 3 symptoms at the moderate to severe level. In this study, the most common moderate to severe symptoms reported by patients were fatigue, drowsiness, difficulty remembering, disturbed sleep, and distress,⁴ which were similar across all tumor grades and are commonly reported by other solid-tumor patients as well.⁵

Not surprisingly, distress, anxiety, and other psychological disorders are more prevalent in the PBT population,^6,7 compared both to the general population and those with non-CNS tumors,^8,9 likely because of an often complicated and unpredictable clinical course, high symptom burden, and a poor prognosis for their disease.^4,10 Significant levels of distress has been associated with worse clinical outcomes for oncology patients in terms of their quality of life, adherence to treatment regimens, lower satisfaction with care, and poorer overall survival.¹¹ There is a need for innovative targeted interventions that are able to address these psychological symptoms in the PBT population in the clinical setting, which could improve both their psychological and physical health.

Virtual reality (VR) can be defined as a computer-generated simulation that allows users to explore and interact with a virtual environment in a way that makes them feel like they are actually present in that world.^12–14 As such, VR has the potential to alleviate some of the negative aspects of illness by allowing individuals to “escape” from their lives to pleasant environments where they can experience more positive thoughts and emotions.¹⁵ VR systems can be classified into 2 types: immersive and nonimmersive. Immersive VR can be characterized by full immersion in a virtual environment using a head-mounted display, which allows the user to lose awareness of time and the real world.¹⁶ In contrast, nonimmersive VR is often experienced using either a computer screen or some other nonwearable media platform that allows the user to explore the virtual environment while still remaining connected to the external world.

Past clinical research using VR as a therapeutic intervention has reported promising results both in adult and pediatric populations, including distraction during stressful medical procedures,¹⁷ improved pain control during burn dressing changes,^18,19 and decreased anxiety and depressive symptoms.^20,21 Given these encouraging findings in other patient populations, there is increasing interest to explore the use of VR interventions in oncology populations, particularly those with solid tumors who tend to experience a similar set of core symptoms throughout the disease trajectory. For PBT patients, we are particularly interested in the potential for VR to improve distress and anxiety because these can significantly affect patient quality of life, though there is potential to improve other symptoms as well based on past research in this field. Therefore, the purpose of this systematic review is to explore the existing literature for VR-related interventions for symptom management in adult PBT and other solid-tumor patients, which will guide future development of VR interventions in these populations. The research question guiding this systematic review was “For adult patients with solid tumors, what effect does VR have on their self-reported symptoms, such as anxiety and pain?”

Methods

A literature search of PubMed, EMBASE, Scopus, and Web of Science was performed to obtain relevant published literature on the use of VR for self-reported symptom management in solid-tumor patients, using the following keywords: cancer, chemotherapy, glioma, neoplasm, tumor, and virtual reality. The search parameters included peer-reviewed quantitative, qualitative, or mixed methods research that were published in the English language, with no publication year limitations given the scant research in this field. The database literature search retrieved 243 publications from PubMed, 1014 publications from EMBASE, 266 publications from Scopus, and 102 publications from Web of Science (see Supplementary 1). This search strategy yielded zero articles focusing on VR use in PBT patients, therefore we included studies in any solid-tumor patients for review. The main author (N.M.L.), along with a librarian, screened titles and abstracts of 1503 articles for relevance, and 3 authors (N.M.L., T.S.A., and A.L.K.) reviewed 27 eligible full-text articles and prepared the manuscript. Articles that evaluated the effects of VR on patient-centered outcomes (ie, pain, anxiety, distress) were included in the review, whereas those assessing effects on disease-related outcomes (ie, survival, toxicities) were excluded. Though solid tumor patients were the target population, there were several studies identified with mixed cancer populations (including solid tumors and hematologic tumors) that otherwise met inclusion criteria, which we ultimately decided to include in the sample. Reference lists of relevant studies were scanned for supplementary articles, which resulted in 4 additional studies included for review. Reviews, abstracts, case reports, and unpublished dissertations were excluded. Differences in reviewers’ extractions were resolved through group discussion. Figure 1 details the final literature selection process for the systematic review. Final inclusion and exclusion criteria are outlined in Table 1.

Fig. 1.

PRISMA Flow Diagram for Systematic Review of the Effects Virtual Reality Interventions Have on Patient-Reported Symptoms in Solid-Tumor Patients PRISMA indicates Preferred Reporting Items for Systematic Reviews and Meta-Analyses; VR, virtual reality.

Table 1.

Inclusion and Exclusion Criteria for Systematic Review

Inclusion criteria	Exclusion criteria
1. Peer-reviewed published quantitative, qualitative, and mixed methods studies	1. Reviews, case reports, abstracts, and dissertations
2. Human studies	2. Non-English language
3. English language	3. Nonsolid-tumor cancers (if not part of a mixed population)
4. Adult and pediatric patients	4. Disease-related outcomes
5. All publication years included
6. Solid-tumor cancers (with some mixed populations)
7. Patient-centered outcomes

Results

In total, we identified 13 manuscripts from 1999 to 2018 that were eligible for inclusion, with only 1 study including a very small number of PBT patients in its sample population.²² The majority of the included studies had experimental designs (N = 12), with 3 randomized controlled trials (RCTs),^23–25 3 cross-over designs,^26–28 3 pretest-posttest designs,^22,29,30 and 3 feasibility studies.^15,31,32 In addition, there was 1 secondary data analysis included that pooled results from 3 previous experimental trials.³³ Sample sizes were typically small (10-30 patients), ranging from 10 to 137 patients across the literature, with the largest sample in a secondary data analysis of combined clinical trials data.³³ Of the 721 pooled participants across the literature, the majority were breast cancer patients^{23,26–28,30,33} (41%), followed by lung cancer^27,29,31,33 (11%) and leukemia^22,25 (11%) patients. Given that several of the included studies had mixed tumor types in their samples, there was a wide variety of cancers represented in the systematic review (see Table 2 for additional details on pooled sample population demographics and tumor types).

Table 2.

Characteristics of Pooled Sample Population (N = 721)

Participant characteristics
	Count	%
Age, y
Adults (≥ 18)	480	67%
Children	231	32%
Unspecified	10	1%
Sex
Male	230	32%
Female	491	68%
Race/ethnicity
Caucasian	354	49%
Asian	124	17%
African American	40	6%
Hispanic	14	2%
Other	25	3%
Unspecified	164	23%
Tumor type
Breast	299	41%
Leukemia	82	11%
Lung	77	11%
Colon	37	5%
Lymphoma	36	5%
Germ cell	23	3%
Osteosarcoma	14	2%
Unspecified solid tumor	9	1%
Bladder	6	1%
Brain	5	1%
Ovarian	4	1%
Histiocytosis	3	< 1%
Stomach	2	< 1%
Rectal	2	< 1%
Unspecifieda	122	17%

^aIndicates lack of identification of tumor type and/or a mixed clinical population.

Virtual Reality Interventions

Owing to technological advances in VR technology over the last 19 years, the types of VR interventions used in the included literature varied widely. A total of 9 studies included interventions using immersive VR headsets,^{15,23,26,28,32} which used a head-mounted display that portrayed engaging virtual environments, typically with an audio component. Two of the more recent immersive VR studies used a smartphone-based VR headset system to deliver the virtual experiences,^15,32 which is still a common approach in modern VR technology. The remaining 4 studies conducted interventions using nonimmersive VR technology,³¹ which displayed the virtual content on a television, computer screen, or both. Nature-related virtual environments were incorporated in the majority of VR interventions in these studies, including scenarios related to ocean exploration, walking through forests and cities, and aerial fly-over views of scenic places. The VR interventions were used for a variety of patient scenarios across the literature, including during outpatient chemotherapy,^26–28,33 inpatient hospitalizations,^{15,22,23,29,31} and stressful medical procedures,^24,25 with 2 studies not reporting VR intervention details.^30,32

Instruments Used

Although there was significant heterogeneity in the measurement approaches used for the symptoms assessed in this review, the most commonly used instruments were various iterations of the visual analog scale (VAS) and the State Anxiety Inventory (SAI), both of which have established psychometrics in oncology populations.^34,35 The VAS is commonly used in clinical research, and is often represented with a numerical rating scale that represents the intensity of the symptom the patient is experiencing.³⁶ Of the studies included in this review, 4 adult studies and 2 pediatric studies employed a version of the VAS to measure the effects of VR on a variety of symptoms, including pain and physical discomfort,^29–31 mood,^29,31 anxiety,²⁴ and fear.²⁵ The SAI is a shortened version of the State-Trait Anxiety Inventory, the most commonly used measurement tool for anxiety in applied psychology research, with demonstrated reliability and validity across numerous clinical populations (including oncology).^37–39 A version of the SAI was used in 6 studies included in the systematic review (5 adult studies and 1 pediatric study) assessing the effects of some sort of VR intervention on the patients’ self-reported anxiety.

There were a number of other instruments used in the literature, given the wide variety of symptoms assessed across adult and pediatric patients. Some examples of the more frequently used measures include the Hospital Anxiety and Depression Scale,^29,30 the Piper Fatigue Scale,^26–28,33 and the Symptom Distress Scale,^26–28 all of which have established psychometric validation and their use has been well-established in oncology populations.^40–42 However, apart from studies conducted by the same research group (ie, the Schneider group) who were consistent in their chosen symptoms and instruments, many of the remaining instruments in this review were used in only 1 or 2 studies. Of note, the pediatric literature used instruments that have been validated and tested in younger populations, though this did contribute to measurement heterogeneity between adult and pediatric studies evaluating similar symptoms. For further details on instruments used, see Tables 3 and 4.

Table 3.

Measurement Instruments for Symptoms

Symptom assessed	Measurement instrument
Pain	VAS, CAS, CHEOPS
Anxiety	VAS, SAI, STAI, HADS, CSAS-C
Distress	SDS, ASDS-2, OSBD
Depression	HADS, CES-DC
Mood	VAS, FS
Fatigue	PFS, institution-developed scalea
Fear	GFS

Abbreviations: ASDS-2: Adapted Symptom Distress Scale; CAS: color analog scale; CES-DC: Center for Epidemiological Studies Depression Scale for Children; CHEOPS: Children’s Hospital of Eastern Ontario Scale; CSAS-C: Chinese version of State Anxiety Inventory for Children; FS: Fordyce Scale; GFS: Glasses Fear Scale; HADS: Hospital Anxiety and Depression Scale; OSBD: Observational Scale of Behavioral Distress; PFS: Piper Fatigue Scale; SAI: State Anxiety Inventory; SDS: Symptom Distress Scale; STAI: State-Trait Anxiety Inventory; VAS: visual analog scale.

^aRefers to an unnamed fatigue scale developed by the study institution.

Table 4.

Review of Virtual Reality Articles

Authors, y	Population and sample	Primary objective	Study design and instruments	Key findings and limitations
Adult
Bani Mohammad and Ahmad, 201823	80 participants, ages 30 to 70 y, mean 52 y ■ Female patients ■ Breast cancer Setting: ■ Individuals in inpatient oncology units who have chronic pain ■ Jordan	To assess effectiveness of immersive VR distraction  technology in reducing pain and anxiety among female patients with breast cancer	Design: ■ Randomized controlled trial ■ Power analysis VR intervention: ■ One 15-min session of immersive VR using headset Instruments: ■ VAS, pain ■ SAI, anxiety ■ MMSE, cognitive function Other measures: ■ Time since diagnosis, morphine dose, treatment type, marital status, educational level, cancer stage	Findings: ■ Significant difference in postintervention anxiety and pain scores between groups, no preintervention differences ■ Significant difference in mean anxiety and pain scores between preintervention and postintervention in both groups ■ 1 session of immersive VR plus morphine had significant reduction in pain and anxiety compared with morphine alone Limitations: ■ Single institution ■ Headsets relatively expensive
Baños et al, 201331	19 participants, ages 29 to 85 y, mean 61 y ■ Primarily males ■ Breast, lung, stomach, rectal, bladder cancer Setting: ■ Individuals in inpatient oncology units ■ Spain	To explore feasibility of a psychological intervention that uses VR technology to induce positive emotional states in oncology patients with advanced disease	Design: ■ Feasibility study VR intervention: ■ Four 30-min sessions for 1 wk using nonimmersive VR on TV and computer Instruments used: ■ VAS, mood, physical discomfort, satisfaction ■ Satisfaction with Intervention Scale ■ Open-ended qualitative questionnaire Other measures: ■ Age range, marital status, educational level, causes of hospitalization and emotional state	Findings: ■ No significant reductions in physical discomfort following VR sessions ■ Participants reported increase in positive emotions and decrease in negative emotions ■ No major difficulties with use of devices ■ Main perceived benefits were distraction, entertainment, and relaxation ■ 4 patients reported fatigue, 1 patient reported slight increase in preexisting dizziness Limitations: ■ Single institution ■ Small sample size, no power analysis ■ Half of participants did not complete all 4 VR sessions
Espinoza, et al, 201229	33 participants, ages 41 to 85 y, mean 62 y ■ Primarily males ■ Metastatic lung, breast and bladder cancer Setting: ■ Individuals in inpatient oncology units ■ Spain	To assess ability of VR to induce positive emotions in adult oncology inpatients	Design: ■ Pretest-posttest design VR intervention: ■ Four 30-min sessions for 1 wk using nonimmersive VR on TV and computer Instruments used: ■ HADS, anxiety and depression ■ Fordyce Scale, happiness ■ VAS, mood and physical discomfort Other measures: ■ KPS, age range, educational level, and causes of hospitalization	Findings: ■ Significant improvement in anxiety and depression levels after session 1 ■ Only depression scale and total HADS significant ■ Significant improvements in physical discomfort after sessions 2 and 3 ■ Participants reported significantly increased happiness and decreased sadness after sessions 2 and 4 Limitations: ■ Single institution ■ Small sample size, no power analysis ■ One-third of participants did not complete all VR sessions ■Insufficient data provided to support findings and conclusions (no data tables and minimal statistics reported)
Li et al, 201632	10 participants, ages of participants not specified ■ Primarily females ■ Sample tumor types not specified Setting: ■ Location of intervention not specified ■ United States	To assess usability testing of VR intervention and determine participant susceptibility to adverse effects	Design: ■ Feasibility study VR intervention: ■ One 30-min session using immersive smartphone-based Google cardboard headset Instruments used: ■ Semistructured qualitative interview ■ MSSQ-Short, susceptibility to motion sickness	Findings: ■ No reports of significant motion sickness while using VR headset, even among patients showing anxiety or predisposition to motion sickness ■ No significant correlation between predisposition to motion sickness (MSSQ) and reports of motion sickness during the intervention Limitations: ■ Single institution ■ Small sample size, no power analysis ■ Age of participants, tumor type, and location of VR intervention not specified
Mosadeghi et al, 201615	30 participants, mean age 50 y ■ Primarily male and Caucasian ■ Mixed population with other diagnoses than cancer, tumor types not specified Setting: ■ Individuals in inpatient units ■ United States	To evaluate acceptability of VR technology in a diverse cohort of hospitalized patients	Design: ■ Feasibility study VR intervention: ■ Four 3- to 5-min sessions using immersive smartphone-based Samsung VR headset Instruments used: ■ Semistructured qualitative interview (think-aloud exercises) Other measures: ■ Age range, race/ethnicity, reason for hospitalization	Findings: ■ Those who agreed to use VR were significantly younger than those who refused ■ No differences in sex, race, ethnicity, or reason for hospital admission ■ 2 patients did not complete study because of VR-related nausea or discomfort from VR headset (too heavy) ■ 1 patient reported minor, transient dizziness that self-resolved ■ 86% reported positive experience and indicated VR could improve pain and anxiety through distraction, and provide “escape” from confines and boredom of hospital Limitations: ■ Single institution ■ Small sample size, no power analysis ■ Unclear what tumor types were included in study
Oyama et al, 199930	22 participants Ages 33 to 75y ■ Female ■ Breast and ovarian cancer Setting: ■ Location of intervention not specified ■ Japan	To improve quality of life of bedridden cancer patients with a support system incorporating VR technology	Design: ■ Pilot study, pretest-posttest design VR intervention: ■ One 6- to 7-min session using nonimmersive VR on TV with a scent system Instruments used: ■ HADS, anxiety and depression ■ VAS (physical discomfort) ■ Unspecified fatigue scale (developed at their institution) ■ Qualitative questionnaires Other measures: ■ Age range	Findings: ■ Most participants showed increase in positive emotions and decrease in negative emotions following VR use ■ 5 participants reported fatigue using system Limitations: ■ Single institution ■ Small sample size, no power analysis ■ Insufficient data provided to support findings and conclusions (no data tables and minimal statistics reported)
Schneider et al, 200326	16 participants Ages 50 to 77 y mean 58 y ■ Females and primarily Caucasian ■ Breast cancer Setting: ■ Individuals receiving outpatient chemotherapy ■ United States	To examine effects of VR distraction intervention on chemotherapy-related distress in older women with breast cancer	Design: ■ Pilot study, cross-over design VR intervention: ■ One session (10 min-2 h) using immersive Sony VR headset Instruments used: ■ MMSE, cognitive function screening ■ PFS, fatigue ■ SAI, anxiety ■ SDS, distress ■ Open-ended qualitative questionnaire Other measures: ■Age range, diagnosis, ethnic identification, and participation rate	Findings: ■ Significant improvement in anxiety immediately following chemotherapy when using VR ■ Mean SDS and PFS scores lower after VR use, though not significant ■ No significant changes in distress, fatigue or anxiety 2 d later (though tend toward lower scores with VR condition) Limitations: ■ Single institution ■ Very small sample, no power analysis ■ Variability in VR duration across participants ■ Marginal significance for anxiety improvement
Schneider et al, 200428	20 participants, ages 27 to 55 y, mean 43 y ■ Female and primarily Caucasian ■ Breast cancer Setting: ■ Individuals receiving outpatient chemotherapy ■ United States	To explore use of VR as a distraction intervention to relieve symptom distress in women receiving chemotherapy for breast cancer	Design: ■ Cross-over design VR intervention: ■ One session (varied duration, 45–90 min) using immersive Sony VR headset Instruments used: ■ SDS, distress ■ SAI, anxiety ■ PFS, fatigue ■ Open-ended qualitative questionnaire Other measures: ■ Age range, diagnosis, and ethnic identification, cycles of chemotherapy	Findings: ■ Significant decreases in distress and fatigue occurred immediately following chemotherapy treatments with VR (nonsignificant for anxiety) ■ No significant decreases in distress, anxiety, or fatigue 2 d later (though all trended toward improvement after using VR) ■ Participants had significantly altered perception of time when using VR ■ VR distraction intervention well received by participants and easy to implement in clinical setting Limitations: ■ Single institution ■ Small sample size, no power analysis ■ Variability in VR duration across participants
Schneider and Hood, 200727	123 participants, ages 32 to 78 y, mean 54 y ■ Primarily female and Caucasian ■Breast, colon, and lung cancer Setting: ■ Individuals receiving outpatient chemotherapy ■ United States	To explore VR as a distraction intervention to relieve distress in adults receiving chemotherapy treatments for breast, colon, and lung cancer	Design: ■ Cross-over design VR intervention: ■ One session (varied duration, average 58 min) using immersive i-Glasses VR headset Instruments used: ■ PFS, fatigue ■ SAI, anxiety ■ ASDS-2, distress ■ PQ, distraction ■ Qualitative evaluation of VR intervention Other measures: ■ Age range, diagnosis, and race	Findings: ■ Participants had significantly altered perception of time when using VR ■ No main effect for anxiety, though significant cross-over effect noted ■ No significant differences in distress immediately or 2 d after chemotherapy treatments ■Improvement trend noted for all symptoms immediately after chemotherapy for those using VR ■ Participants believed head-mounted VR device was easy to use, they reported no cybersickness, and 82% would use VR again Limitations: ■ Single institution ■ 23 patients did not complete study ■ Variability in VR duration across participants
Schneider, 201133	137 participants, ages 27 to 78 y, mean 52 y ■ Primarily female and Caucasian ■ Breast, lung, colon cancer Setting: ■ Individuals receiving outpatient chemotherapy ■ United States	Explore influence of age, sex, state anxiety, fatigue, and cancer diagnosis on difference between participant-perceived time elapsed and actual time elapsed during chemotherapy treatment while immersed in VR environment	Design: ■ Retrospective, secondary analysis of pooled data from 3 trials VR intervention: ■ One session (varied duration, average 63 min) using immersive VR headset Instruments used: ■ STAI, anxiety ■ PFS, fatigue ■ Elapsed time, perceived and actual of chemotherapy duration Other measures: ■ Age, sex, and race/ethnicity	Findings: ■ Actual time elapsed during chemotherapy with VR intervention averaged 63 min for pooled sample ■ Most participants underestimated duration of VR intervention during chemotherapy sessions (varied by cancer diagnosis) ◦ 23 min (breast) ◦ 12 min (colon) ◦ < 4 min (lung) Limitations: ■ Low baseline anxiety and fatigue in participants (floor effect) ■ Retrospective design
Pediatric
Gershon et al, 200424	59 participants, ages 7 to 19 y, mean 13 y ■ Even sex distribution ■ Primarily Caucasian ■ Leukemia, lymphoma, and solid-mass tumors Setting: ■ Individuals in outpatient oncology unit who require port access ■ United States	To pilot and test feasibility of immersive VR technology to reduce anxiety and pain associated with an invasive medical procedure in children with cancer	Design: ■ Randomized controlled trial VR intervention: ■ One 5- to 10-min session using immersive (headset) or nonimmersive (computer) VR Gorilla program during port procedure Instruments: ■ VAS, pain and anxiety ■ CHEOPS, pain/distress ■ HR, physiological arousal Other measure: ■ Age	Findings: ■ VR groups had significantly lower HR during procedure ■ Medium effect size (η2 = 0.09) ■ VR groups had significantly lower RN-rated pain during procedure ■ Control group had significantly higher distress during procedure ■ Significant weak correlation (r = 0.3) found between HR during procedure and total CHEOPS score ■ Younger participants had higher HR during procedure compared to older participants Limitations: ■ Single institution ■ Small sample size, no power analysis ■ No true control group (they practiced with VR before procedure)
Li et al, 201122	122 children, ages 8 to 16 y ■ Mostly male and Chinese ■ Leukemia, lymphoma, osteosarcoma, germ cell tumors, and brain tumors Setting: ■ Individuals admitted to inpatient pediatric oncology unit on active treatment ■ Hong Kong	To examine effectiveness of therapeutic play, using VR computer games, in minimizing anxiety and reducing depressive symptoms in Hong Kong Chinese children hospitalized with cancer	Design: ■ Quasi-experimental pretest-posttest design ■ Power analysis VR intervention: ■ Five 30-min sessions in 1 wk using nonimmersive PlayMotion VR system (computers and projectors) conducted in small groups (maximum 4 children) Instruments: ■ CSAS-C, anxiety ■ CES-DC, depressive symptoms	Findings: ■ High baseline state anxiety scores for both groups ■ VR group had fewer depressive symptoms on day 7 ■ Moderate effect size (η2 = 0.06) ■ No significant difference in anxiety scores between groups on day 7 (though both decreased slightly) Limitations: ■ Single institution ■ Unbalanced study arms ■ Underpowered study (attrition and recruitment challenges) ■ Could be capturing natural progression of being hospitalized rather than intervention effects
Windich-Biermeier et al, 200725	50 children, ages 5 to 18 y, mean 11 y ■ Mostly male and Caucasian ■ Leukemia, lymphoma, solid tumors, and histiocytosis Setting: ■ Individuals in outpatient oncology unit receiving active treatment ■ United States	To evaluate effect of self-selected distractors on pain, fear, and distress in children and adolescents with cancer during venous port access or venipuncture procedures	Design: ■ Randomized controlled trial VR intervention: ■ One session (varied duration) with a self-selected distractor (including an immersive VR headset) to be used during procedure Instruments: ■ CAS, pain ■ Glasses Fear Scale, fear ■ OSBD, distress ■ IV Poke Questionnaire	Findings: ■ No significant differences in mean pain scores by group ■ Significant improvement in RN-reported fear scores from during procedure to afterward between groups ■ Control group had more fear during procedure ■ Differences in fear scores from before, during, and after procedure were not significant for child- or parent-rated scores ■ RN-rated distress scores significantly improved between 2 groups from during procedure to after procedure Limitations: ■ Single institution ■ Small sample size, no power analysis ■ Only 18% of sample chose VR distractor

Abbreviations: ASDS-2: Adapted Symptom Distress Scale; CAS: color analog scale; CES-DC: Center for Epidemiological Studies Depression Scale for Children; CHEOPS: Children’s Hospital of Eastern Ontario Pain Scale; CSAS-C: Chinese version of State Anxiety Inventory for Children; HADS: Hospital Anxiety and Depression Scale; HR, heart rate; MMSE: Mini Mental State Exam; MSSQ-Short: Motion Sickness Susceptibility Questionnaire-Short Form; OSBD: Observational Scale of Behavioral Distress; PFS: Piper Fatigue Scale; PQ: Presence Questionnaire; RN, registered nurse; SAI: State Anxiety Inventory; SDS: Symptom Distress Scale; STAI: State-Trait Anxiety Inventory; VAS: visual analog scale; VR: virtual reality.

Main Symptom-Related Results

Anxiety

The most commonly assessed symptom in the literature was anxiety, which was measured in 8 of the 13 articles (6 adult studies and 2 pediatric studies) in 475 patients. Although the type and duration of the VR interventions varied, most studies assessing anxiety used a single VR session in their designs. Measurement of anxiety was fairly consistent across the literature, with the SAI or the Hospital Anxiety and Depression Scale instruments used for adults and the VAS or Chinese version of the SAI used for children, all of which have been validated in oncology populations.³⁹ There were 5 studies that reported reduced anxiety following VR use, all of which had experimental designs, including 1 RCT,²³ 2 cross-over designs,^26,27 1 single-arm pilot study,³⁰ and 1 pretest-posttest design.²⁹ The most promising anxiety evidence was from 3 well-designed clinical trials, conducted primarily in breast cancer patients, that reported significant reductions in anxiety immediately after the VR intervention^23,26,27 with a moderate effect size (0.44).²⁶ Of the 3 studies with nonsignificant findings (1 RCT,²⁴ 1 pretest-posttest design,²² and 1 cross-over design²⁸), 2 of them were in pediatric mixed cancer populations with differing tumor types in their samples (mainly hematologic cancers) compared to the adult studies. Additionally, the VR intervention approaches in the pediatric literature were atypical compared to adult approaches, including group VR sessions in a hospital playroom²² and exploration of a virtual gorilla environment,²⁴ which were likely less anxiolytic compared to other VR scenarios used. Despite the majority reporting improvement in anxiety, these studies were often underpowered with small sample sizes, some had significant attrition or difficulty recruiting during data collection, and in some cases the duration of the VR intervention varied widely among participants. Additionally, 2 studies reported significant anxiety improvement but did not provide adequate statistical information to support their claims.^29,30

Pain or physical discomfort

Another commonly assessed symptom in relation to use of VR was pain or physical discomfort, which was measured in 5 of the 13 articles (3 adult studies and 2 pediatric studies) in 241 patients. There was a wide variety of approaches regarding the VR interventions employed in these studies, but measurement of pain was uniform across the literature with the VAS instrument (or the color analog scale–pediatric version), which aids in comparison of findings. In addition to the VAS, Gershon et al²⁴ also incorporated changes in heart rate to assess physiologic arousal, which could be a surrogate for pain and was the only biomarker used in any of the included literature in this review. Of the 5 studies assessing pain, 3 studies (1 RCT in adult breast cancer,²³ 1 pretest-posttest design study in adult metastatic solid tumors, and 1 RCT in pediatric hematologic and solid tumors²⁹) reported significant improvements in this symptom following a VR intervention. The most promising pain findings in adults were in the RCT by Bani Mohammad and Ahmad,²³ which found a 7-point reduction in mean pain scores on the VAS following a single VR intervention, compared to a 2.5-point reduction in the control group. In the pediatric RCT by Gershon and colleagues,²⁴ they too found evidence for pain reduction following VR use, with significantly lower nurse-rated child pain and lower heart rate during the procedure in the intervention group, with a medium effect size (η ² = 0.09). Although Espinoza et al²⁹ also reported significant improvement in pain scores, these authors did not provide sufficient statistical data to fully support their conclusions. Global limitations of the studies assessing pain were similar to that of anxiety with regard to small samples and underpowered studies, significant participant attrition, and a lack of reported statistical information to support the authors’ claims. However, these pain studies also had greater heterogeneity of tumor types in their samples (with more hematologic and metastatic cancers represented), there was an instance of control group contamination in an RCT,²⁴ and in 1 study only 18% of the sample chose VR as their distraction intervention.²⁵ Collectively, these design flaws likely contributed to less-compelling or nonsignificant findings for pain reduction with VR use.

Distress

Another symptom commonly investigated in this literature related to VR use was distress, which was measured in 5 of the 13 included articles (3 adult studies and 2 pediatric studies) in 268 patients. In adults, distress was measured with the Symptom Distress Scale instrument in 3 studies conducted by Schneider et al,^26–28 whereas in children researchers used observational instruments,^24,25 including the Children’s Hospital of Eastern Ontario Pain Scale and the Observational Scale of Behavioral Distress. Of the 5 studies that evaluated this symptom, 3 showed improvement in distress following a VR intervention (2 pediatric RCTs^24,25 in mixed cancer populations and 1 cross-over design in adult breast cancer patients²⁸), all of which used VR for distraction during painful medical procedures or a chemotherapy infusion. In the 2004 study by Schneider and colleagues,²⁸ although they did find a significant improvement in distress immediately following the VR intervention (effect size = 0.3), this effect did not persist 2 days after the intervention. Importantly, in the RCT by Gershon et al,²⁴ they used a behavioral pain scale to assess child distress, which seems an inappropriate choice for measurement of this symptom. Two other studies performed by the Schneider group^26,27 had nonsignificant distress findings following immersive VR interventions, though they reported an improvement trend in both experiments. In addition to small sample sizes and lack of power, another limitation to studies evaluating distress that could affect findings is the specific content chosen for the VR interventions. For example, the Schneider group had more stimulating, intellectual content that allowed participants to solve mysteries and learn about history, whereas other researchers tended to incorporate content aimed at relaxation.

Other symptoms

In addition to anxiety, pain, and distress, other symptoms assessed in relation to a VR intervention included depression and mood, as well as fatigue. Four of the 13 studies (3 adult studies in primarily breast cancer and 1 pediatric study with a mixed tumor population) analyzed the effects of VR use on depression and mood in 196 patients. For adults, there was 1 feasibility study³¹ and 2 pretest-posttest designs,^29,30 whereas the single pediatric study also had a pretest-posttest design.²² Even though all 4 of these studies reported significant improvement in depressive symptoms and mood, only the study by Li et al²² reported reliable findings. In this pediatric study, they found that the VR group had fewer depressive symptoms 7 days after admission following several VR sessions during the week, with a moderate effect size reported (η ² = 0.06). In the 3 adult studies, they did not report adequate statistical data to substantiate their claims that VR could improve mood for patients.^29–31 Fatigue was a symptom evaluated solely by the Schneider group in 4 of the 13 studies, which included 3 cross-over designs^26–28 and 1 secondary data analysis³³ of 3 larger trials, all of which were conducted in adult solid-tumor populations (primarily breast cancer) and included a total of 296 patients. These 4 studies were more homogeneous because they were conducted by the same researchers, so as a result their patient populations and study designs were quite similar and their findings more comparable. Of these 4 studies, only the 2004 study²⁸ found a significant reduction in fatigue immediately following the VR intervention (effect size = 0.41), though similar to distress, this effect did not persist 2 days later. The remaining 3 studies had nonsignificant findings for VR’s effect on fatigue during chemotherapy sessions. A key limitation of these studies was that participants might not have had high fatigue at baseline, which was noted in one study,³³ therefore the VR intervention was unlikely to reduce that symptom by any significant amount. Additionally, the specific VR content the researchers chose in these studies might not have been an appropriate choice to improve affective symptoms.

See Table 4 for further details about the main symptom-related findings from the articles included in the review.

Qualitative Results

Eight of the 13 included studies incorporated qualitative assessments into their study designs to explore patient satisfaction with the VR intervention and their thoughts about its ability to improve their symptoms, including feedback from 290 participants. The types of qualitative assessments varied across studies, with the most common being open-ended questionnaires,^{15,26,28,30–32} whereas some implemented other types of satisfaction assessments related to the VR intervention^25,27 (see Table 4 for details). The vast majority of patients had a positive experience with the VR intervention, reporting that it was pleasant, easy to use, relaxing, improved their mood, and was able to distract them effectively from a variety of unpleasant experiences. In a select few cases, participants reported having trouble initially figuring out how to navigate the VR interface, though with practice and assistance from the researcher, they were able to learn how to use the headsets and remotes effectively to complete the VR interventions. There were very few side effects reported with use of the VR headsets, though a few individuals reported fatigue,^30,31 dizziness or motion sickness,^15,31,32 and nausea¹⁵ while participating in the VR interventions.

Discussion

This systematic review evaluated the existing literature assessing the effects of VR interventions on self-reported symptoms in patients with solid tumors, though other tumor types were included because of mixed sample populations in some studies. Although our literature search did not identify any articles focusing on VR use solely in PBT populations, by reviewing this past work we were able to gain a better understanding of the potential therapeutic effects of VR interventions in other solid-tumor populations that often experience symptoms similar to those of PBT patients,⁴ which will provide the foundation for development of future VR interventions in the PBT population.

The most commonly studied populations in this review were adult patients with breast and lung cancer, which is not surprising given these are the most prevalent cancers worldwide.⁴³ In contrast, the pediatric literature had mixed tumor populations, including solid tumors, a variety of hematologic cancers, as well as the few brain tumors represented in the review.^22,24,25 Although hematologic malignancies are quite different from solid-tumor malignancies with regards to presentation, treatment, and for some of the symptoms patients experience,⁴⁴ the symptoms that were assessed in relation to VR use in this review are relatively universal across all types of cancer. The strongest evidence for VR use in past literature for other clinical populations has been for improved pain control^18,19 and reduced anxiety and depressive symptoms,²¹ all of which are commonly experienced by cancer patients, regardless of the type of cancer they have. As such, we felt it appropriate to include studies with varying tumor types to broadly characterize the efficacy of VR interventions for symptomatic improvement in oncology patients.

There was significant heterogeneity of instruments used to measure the symptoms assessed related to VR use, which is likely due to the high number of symptoms included in the review and the wide age range of participants. For example, the VAS instrument was used to measure 4 different symptoms in this review (pain/physical discomfort, mood, anxiety, and fear), and several different forms of the instrument were used based on the population it was used in (adult vs pediatric). Despite the broad number of instruments used, the vast majority were validated in the setting and sample in which they were used, but the measurement heterogeneity did make cross-study comparisons challenging. Future clinical trials using VR interventions in oncology populations should strive to adopt a standardized measurement approach for their symptoms outcome measures, which will allow greater comparability of findings across different patient populations.

The key findings from this review were promising improvements in anxiety, pain or physical discomfort, distress, and effective distraction with use of a VR intervention, which mirrors past VR research in other clinical populations.^18–21 The most convincing evidence for improvement of anxiety symptoms was in an RCT with adult breast cancer patients,^23,26,27 though interestingly these findings were not replicated in the pediatric studies that assessed anxiety.^22,24 These null findings in children could be due to a variety of factors, including differing scenarios in which VR was used compared to the adult literature (ie, some environments might be more anxiolytic than others), differing tumor types in children (ie, anxiety might be more prevalent in those with solid tumors vs hematologic tumors), and generally weaker study designs in the pediatric trials. The ability of VR to improve pain and physical discomfort, as well as distress, was mixed, though several studies had encouraging findings^23–25,29 that need to be replicated in future RCTs in more homogeneous samples. The ability of VR to effectively distract patients from unpleasant experiences was also established through this review, primarily through participant feedback during qualitative semistructured interviews, though the exact mechanism for how VR is able to distract patients and improve their symptoms remains unclear.

Findings from the qualitative assessments of the VR experience were largely positive from participants, with most reporting a significant improvement in their stress level, state of mind, and mood following the intervention. Given that the included literature spanned a total of 19 years (1999 through 2018), synthesizing the participant critiques of the VR technologies used across these studies was impractical. The one common complaint that was reported about the VR experiences was that the participants had trouble initially understanding how to use the VR headsets and the remote controls, and how to navigate the VR interfaces. Keeping this in mind, for future VR research it will be important for researchers to first orient patients to the technology prior to the intervention to avoid participant frustration and potentially negating any positive effects the intervention is intended to have. Importantly, there were very few adverse side effects reported that were associated with the VR interventions, and for those who did experience symptoms, they typically experienced that symptom at baseline.

Implications for Primary Brain Tumor Patients

This review provides preliminary support for the use of VR-based interventions for psychological symptoms, in particular distress and anxiety, both of which occur with high prevalence in patients with PBT, particularly related to uncertainty related to imaging results.^8,9 The term scanxiety describes the distress related to often-debilitating anxiety cancer patients can experience in the period surrounding their diagnostic imaging studies and leading up to their clinic appointments,⁴⁵ which can be particularly difficult for PBT patients who have frequent surveillance imaging and an often uncertain clinical trajectory. Given the promising findings from this review, VR seems to be an ideal delivery approach to provide PBT patients with an immersive relaxation intervention that could potentially alleviate some of the psychological symptoms they experience, as well as their ability to self-manage these symptoms moving forward. Future research evaluating use of VR for symptom management in a PBT population would be beneficial, given they tend to be understudied compared to other cancer types, and also because of the potential correlation between neurological symptoms and the more common cancer symptoms that patients experience. It is unclear how certain symptoms specific to this population might affect usability and effectiveness of VR systems, including visual field cuts, expressive and receptive aphasia, nausea, and scalp sensitivity at surgical sites, which should be assessed in future interventional research in this field. Additionally, exploring the logistics of implementing VR in a clinical setting is needed to determine the impact on clinical workflows, as well as how easily VR devices can be sterilized in between patients. Future clinical trials are needed using VR in PBT patients to establish the feasibility and efficacy of this kind of technology in a clinical setting.

Limitations

There are several limitations to this report that are important to acknowledge to understand their potential impact on findings. One major limitation is the breadth of tumor types represented in the review, including solid tumors and hematologic cancers, which was a result of mixed tumor populations in the pediatric literature. Importantly, for our purposes, there were only 5 participants with PBTs across all included studies, which accounts for less than 1% of the pooled sample. As a result, these findings might not be reflective of the impact (either positive or negative) that VR interventions might have on symptoms in our population of interest. Another limitation is that the majority of included studies had small sample sizes (with only 2 studies conducting a priori power analyses) and also had significant difficulties with recruitment and retention, which further decreased power in those studies. A broad limitation is how diverse the studies were with regards to their VR interventional approaches and measurement of symptoms, which limited our ability to compare their findings and assess the efficacy of VR for symptom management. One methodological limitation was that the duration of the VR interventions varied considerably across studies, which could skew findings if there is a significant dose-effect present. A statistical limitation was related to the lack of adequate data reported in several studies, including absent data tables, unclear significance thresholds, and no effect sizes reported, all of which made interpreting the validity of those findings difficult. Last, only 1 study included a physiologic biomarker in its design to correlate with changes in self-reported outcome measures,²⁴ which could have strengthened the evidence for VR’s effectiveness in improving symptoms.

Conclusions

Within the PBT population, recent studies have found that psychological needs are highly unmet among these patients,⁸ with few interventions available to address the adverse psychological symptoms they experience throughout their illness trajectory. The findings of this systematic review suggest that VR is an innovative approach for targeted interventions that could improve not only the psychological symptoms that PBT patients experience, particularly surrounding clinical evaluations, but could also positively affect their quality of life, tolerance of oncologic therapies, and potentially their survival. Well-established mindfulness and guided-breathing techniques that have demonstrated ability to improve distress, anxiety, and other psychological symptoms could be ideal VR-based interventions to help elicit relaxation states for PBT patients and provide them with skills to better self-manage these symptoms moving forward. Further well-designed clinical trials using comparable interventions and technologies are needed to better demonstrate the efficacy of VR for improving symptoms in this population.

Conflict of interest statement.

None declared.

Funding

This work was supported by The Intramural Research Program of the National Institutes of Health, within the National Cancer Institute.