Virtual Reality: Augmenting the Acute Pain Experience in Children Samantha Diaz-Hennessey, Eileen R. O'Shea, and Kyle King

Sickle cell disease affects approximately 100,000 Americans, and according to global newborn estimates, approximately 300,000 infants are born with sickle cell disease each year (Piel, Steinberg, & Rees, 2017). Vaso-occlusive crisis is a complication of the disease that causes severe pain. Children require prescriptions for pain medication, often opioids, during a vaso-occlusive crisis experienced at home, but they seek medical care in the emergency department (ED) when the pain is not improving. Fifty percent to 60% of visits to the ED among children with sickle cell disease are due to pain (Meier & Miller, 2012). The following review of the literature used the electronic databases CINAHL Plus, PubMed, and Cochrane Library. Key search terms included virtual reality, sickle cell crisis, pain, children, and pediatrics.

Sickle Cell Pain Control Methods

An expert panel report released by the National Heart, Lung, and Blood Institute (NHLBI) (2014) included recommendations for the management of pain in pediatric patients who present with vaso-occlusive crisis. The report emphasized the importance of 1) assessing pain rapidly using self-reported pain scales and observation, 2) assessing previously used analgesics and their effectiveness, and 3) administering analgesic therapy within 30 minutes of triage in the ED. The panel advised administering opioids according to individualized pain protocols and reassessing pain every 15 to 30 minutes. Although opioids and non-steroidal anti-inflammatory drugs (NSAIDs) are typically used to manage vaso-occlusive crisis pain in the ED (Puri, Nottage, Hankins, & Anghelescu, 2018), recommendations also include the use of non-pharma-cologic modalities.

Despite the national guidelines, children suffering from pain caused by vaso-occlusive crisis are still not achieving adequate pain relief (Vijenthira et al., 2012). Vijenthira and colleagues (2012) completed a chart review of 50 pediatric patients who were hospitalized for sickle cell pain. Intravenous morphine, acetaminophen, and ibuprofen were the most frequently administered medications, which followed the hospital protocol. However, non-pharmacologic interventions were not well documented and likely not used. Only 20% of reports mentioned using a hot pack and 37% of reports included distraction by watching television. Pain scores remained at a range from 4/10 to 10/10 using the Numerical Rating Scale (NRS) and/or FLACC scale throughout the hospital stay. Vijenthira and colleagues (2012) concluded that pain among this population should be addressed more aggressively and with more multimodal pain management strategies as patients continued to report pain even with medication management.

When pain is not adequately controlled during an ED visit, patients with sickle cell disease are admitted for inpatient pain management, and on average, this occurs with about 42% of these patients (Frei-Jones, Baxter, Rogers, & Buchanan, 2008). These frequent pain crises and hospital admissions affect pediatric patients, their families, and their quality of life (Fisak, Belkin, von Lehe, & Bansal, 2012).

Two barriers to adequately managing sickle cell pain/vaso-occlusive crisis in pediatric patients include the frequent necessity of several doses of narcotics and the discomfort level of ordering these medications among ED providers (Bergman & Diamond, 2013). Porter, Feinglass, Artz, Hafner, and Tanabe (2012) concluded that 53% of ED physicians believed approximately 20% of their patients with sickle cell disease were addicted to narcotics, and 30% of ED nurses were hesitant to give these patients high doses of narcotics. Historically, ED nurses have tried adding non-pharmacologic therapies and distraction techniques to help reduce pain without having to administer more medication (Dobson & Byrne, 2014).

VR for Pain Control

Healthcare providers are typically interested in the use of non-pharma-cologic interventions to augment pain in children because it is seen as an essential part of pain management (Wente, 2013). One innovative type of non-pharmacologic tool is virtual reality (VR). VR is a computer hardware and software device that creates simulated environments to look and feel real (Yamato, Pompeu, Pompeu, & Hassett, 2016). VR is easily administered and can be used as a distraction for children experiencing pain. Although empirical studies are sparse, evidence is building to support the use of VR to augment patients' pain. Schmitt and colleagues (2011) conducted a randomized, controlled study using a crossover design to examine the effects of VR as an adjunctive treatment for pediatric burn patients undergoing range-of-motion exercises as part of physical therapy. The sample consisted of 54 participants aged 6 to 19 years. Participants were medicated with standard pre-procedure analgesia, and everyone completed physical therapy with an equal amount of time spent with and without using VR in the same session. Schmitt and colleagues (2011) assessed the cognitive, affective, and sensory components of pain after each treatment condition and found a 44%, 32%, and 27% reduction respectively when VR was used compared to when VR was not used. There was a significant change in the experienced pain level when virtual reality was used. Kipping, Rodger, Miller, and Kimble (2012) assessed the effectiveness of the use of VR for adolescent burn patients during dressing changes. Nurses and caregivers observed pain and patients' self-reported pain. There was a statistically significant reduction in pain as observed by the nurses, and the length of time for the procedure was decreased in patients who used VR compared to those who received standard distraction.

In the adult literature, Hoffman, Chambers, Meyer, Arceneaux, and Russell (2011) completed a study to evaluate how VR affects pain levels in patients with burns undergoing wound debridement in a hydrotherapy tank. Hoffman and colleagues (2011) found that patients who had initially reported the most severe pain had a 41% reduction in pain level while using VR. Although participants in this study were not children, results support that even the most severe pain can be reduced using VR.

The gate control theory helps explain how VR works to reduce pain. According to this theory, humans have limited capacity for attention. When focused on a noxious stimulus, it is perceived as painful. However, when focused on a pleasant stimulus, the noxious stimulus is perceived as less intense and severe (Li, Montano, Chen, & Gold, 2011). The purpose of this study was to determine the effectiveness of VR in reducing pain and the amount of pain medication needed in pediatric patients with sickle cell pain/vaso-occlusive crisis. Using the gate control theory, this study aimed to answer the following two research questions:

1. Do pain scores for pediatric sickle cell patients who are experiencing pain crisis in the ED decrease with the use of VR?

2. Does a VR intervention for children with sickle cell crisis in the ED decrease the use of pain medications?

Methods

Design

A quasi-experimental research design was used as the research method to evaluate the effectiveness of VR as a non-pharmacologic therapy for children experiencing vaso-occlusive crisis in the ED.

Sample

A total of 15 pediatric patients between the ages of 8 and 18 years with sickle cell crisis volunteered to participate in the study, which took place in a pediatric ED at an urban teaching hospital in the Northeast. Systematic sampling was used in which every other patient was assigned to the control group (those who received standard care alone) or the intervention group (those who received standard care along with VR). Standard care at the acute care hospital consisted of IV narcotics every 30 minutes as needed for up to three doses. Patients were included in the study if they presented to the ED with vaso-occlusive crisis, were between the ages of 8 and 18 years, were English-speaking, and had a parent or legal guardian with them. Patients were excluded from the study if they had used VR before, if they had presented with other conditions that could be contributing to their pain, or if they had a history of seizures or motion sickness because these were potential side effects of using VR. A G*Power analysis was conducted to determine a desired sample size of 128, with 64 participants in each the control group and the intervention group, using a standard power of 0.8, along with an effect size of 0.5, and an alpha set to 0.05.

Instruments

For all participants, pain levels were assessed using both a behavioral scale (FLACC scale) and a self-report scale (NRS).

The FLACC Behavioral Pain Assessment Scale includes an assessment of face, legs, activity, cry, and consolability as five different components of pain observed in a patient (Merkel, Voepel-Lewis, Shayevitz, & Malviya, 1997). Each of the five categories is scored on a scale of 0 to 2 based on the severity, with a total possible range of 0 to 10. A score of 0 indicates the patient has no particular facial expression or is smiling; has a normal position or is relaxed; is lying quietly and moves easily; has no cry; and is content and relaxed. A score of 10 indicates the patient has frequent to constant quivering chin or clenched jaw; is kicking his/her legs or has them drawn up; is arched, rigid, or jerking; is crying steadily, screaming, or sobbing with frequent complaints; and is difficult to console or comfort. This pain assessment tool is typically recommended for children aged 2 months to 7 years (Srouji, Ratnapalan, & Schneeweiss, 2010); however, multiple researchers have evaluated and confirmed validity and reliability of this scale in various pediatric age groups when patients are unable to self-report pain (Blount & Loiselle, 2009; Nilsson, Finnstrom, & Kokinsky, 2008; Voepel-Lewis, Zanotti, Dammeyer, & Merkel, 2010). Additionally, researchers Voepel-Lewis and colleagues (2010) reported criterion and construct validity of the FLACC scale, as well as interrater reliability, and found the Cronbach alpha to be 0.882.

The NRS is used to assess pain by asking a patient to rate his/her pain on a scale of 0 to 10, with 0 representing no pain, and 10 representing the worst pain possible (Myrvik et al., 2013). Until recently, there were limited data supporting the use of the NRS in children (Castarlenas, Jensen, von Baeyer, & Miro, 2017). Newer studies have evaluated the psychometric properties of this pain scale, confirming its appropriate use for children in clinical practice. Miro, Castarlenas, and Huguet (2009) found a strong association between the NRS and Faces Pain Scale -Revised (Hicks, von Baeyer, Spafford, van Korlaar, & Goodenough, 2001, which demonstrated convergent validity. The authors concluded that the NRS could be used to assess pain in children greater than age 8 years. Fernando, Rifaya, Asantha, Chandarathna, and Wijeratna (2017) compared the NRS with the Faces Pain Scale and the Verbal Pain Rating scale (von Baeyer, 2006) in children aged 4 to 12 years and found internal consistency (Cronbach alpha = 0.902).

Ethical Considerations

The Institutional Review Board approved this study prior to data collection. Informed consent was obtained from parents/legal guardians of participants, and assent was obtained from all participants. Minimal risks were involved with the study, and participants were made aware of risks prior to participating. Potential risks included nausea/motion sickness related to the VR and eyestrain. Benefits of reducing pain without using narcotic pain medication was also explained to the patients and their parents/legal guardians. Participants were made aware that participation was voluntary and that they could discontinue their participation in the study at any point without penalty. Participants and parents/legal guardians could ask questions prior to the start of the study and during the study, and all questions were answered. No identifying information was collected in the study to protect patients' right to privacy.

Intervention

The Google Daydream VR was used for patients who were in the intervention group for 15 minutes after they received their IV pain medication (standard care). The VR consisted of a headset that wears like a pair of goggles and a screen from a Google Pixel cell phone inside the headset that illuminated in front of the patient's eyes. While wearing the headset, the patient could move his/her head around to see a full 360-degree view of the image, which immersed the patient into a virtual world. Participants could choose from five different applications that were downloaded onto the Google Pixel cell phone. The choices of applications included 1) Wonderglade, an application with several interactive mini games, such as basketball and golf; 2) Ocean Rift, where the participant can explore the ocean and its inhabitants; 3) Karts Sprint, a car racing game; 4) Ace Fishing, a 3D fishing adventure; and 5) The Turning Forest, where a narrator reads a story while the participant encounters a magical forest. The headset blocked the viewer's vision of his/her real surroundings. The phone was set to Airplane Mode so Wi-Fi and cellular data were not used at the time the video was displayed.

Data Collection

The principal investigator (PI) (SDH) collected data at one acute care hospital ED setting, over 14 weeks. All patients used the NRS to self-report their pain levels at two points in time: once prior to the IV opioid administration (standard care) and then again 15 minutes post-medication administration. At the same two points in time, the PI also used the FLACC scale to record observations of the patient's pain levels. For all patients, the PI also conducted one more observational assessment, which occurred 5 minutes after the IV opioid was administered.

Participants in the intervention group were given their first dose of IV pain medication and then used the VR for 15 minutes. This amount of time was chosen because virtual reality has been successful in other studies (Dascal et al., 2017; Gershon, Zimand, Pickering, Rothbaum, & Hodges, 2004; Hoffman et al., 2011, Schmitt et al., 2011) when it was used for medical procedures, which typically lasted no more than 15 minutes. Using the FLACC scale, the PI observed the patient's movement, activity, and facial expressions while the patient was wearing the headset to determine the level of pain 5 minutes into the experience; this is the time at which IV morphine reaches its peak effect (Larijani, Goldberg, Warshal, & Gratz, 2005). The FLACC scale was used at this point to enable the PI to assess the patients' pain rather than interrupting the patients while using the VR to assess pain on a reported scale. Interrupting the experience could remind patients of their pain and potentially alter the true effect of the VR. After 15 minutes of using VR, the PI assessed the patients' level of pain by using the NRS and the FLACC scale again. Both scales were used to assess pain before and after the intervention, so a comparison could be made between the reported pain by the patients and the observed pain level from the PI.

Patients in the control group received standard care for sickle cell pain and their level of pain was assessed 5 minutes after they received their first dose of IV pain medication also using the FLACC scale. The PI assessed pain 15 minutes after the first narcotic medication was administered using the NRS and the FLACC scale again. The difference in pain level from the time of initial medication administration and 15 minutes later was compared. The FLACC scale was used to assess pain 5 minutes after the narcotic was given, so it could be compared to the pain level observed in patients in the intervention group with the same scale.

In sum, for both groups, the initial assessment included the NRS reported by the patient and the FLACC pain score observed by the PI. Then the PI assessed pain using the FLACC scale 5 minutes after medication administration. Lastly, the assessment at the 15-minute mark, post-medication administration, included both the NRS score reported by the patient and the FLACC pain score observed by the PI.

The PI documented the pain level reported and observed at the start of the study, the pain level observed at the 5-minute time interval, and the pain level reported and observed 15 minutes after the initial pain medication with the respective times in an Excel spreadsheet. The changes in scores between the different assessment time intervals were calculated. The PI also recorded how many doses of pain medication the patient requested and the total length of stay in the ED. All data collected corresponded to a coded number to represent each participant without using private identification information. The PI obtained data for all patients, including the observed and reported pain levels, to keep the data collection consistent and accurate.

Statistical Analysis

Results collected in this research project were analyzed using an independent samples t test to compare the pain levels at all time intervals for participants in the intervention and control groups. Additionally, a linear mixed model in R was used to test effects of the intervention in relation to time. Confidence interval was 95%, and all p-levels were considered significant at the level of 0.05.

Results

Fifteen children with vaso-occlusive crisis participated in the study. Their ages ranged from 8 to 17 years. The median age of participants was 13 years. Sixty percent of participants were female, and 40% were male. All participants were African American (see Table 1). The average observed pain scores using the FLACC scale were significantly lower for patients using VR for 5 minutes (M=1.57, SD=1.51, n=7) than the average pain scores documented for participants in the control group (M=4.25, SD=2.12, n=8), t(13)=2.78, p=0.01 (see Figure 1). The average observed pain level using the FLACC scale was lower at 15 minutes in those who used VR (M=1.29, SD=1.70); however, scores did not differ significantly from those who did not use VR (M=3.25, SD=2.12), t(13)=1.96, p=0.07.

Self-reported pain scores using the NRS did not significantly differ on average whether the patients used VR for 15 minutes (M=5.71, SD=2.75) or received standard care alone (M=5.25, SD=2.25), t(13)=-0.36, p=0.73. The average pain level reported by patients was actually higher in those who used VR. Those who used VR had a greater decrease in observed pain from the initial assessment compared to the score observed 5 minutes later (M=3.29, SD=1.38). However, there was no significant change in the pain observed when compared to those in the control group (M=1.75, SD=2.05), t(13)=-1.67, p=0.11. Further, there was a greater decrease in pain observed from the time of the initial assessment to the one 15 minutes later for those who experienced VR (M=3.57, SD=0.53) when compared to those who received standard care only (M=2.75, SD=2.05), t(13)=-1.09, p=0.30, but this difference was not statistically significant.

The linear mixed model found the effect of the intervention reduced FLACC score by -1.65; although, this was not statistically significant (p=0.11). However, for every minute increase from the initial measurement, FLACC scores were reduced by -0.18, which is statistically significant (p<0.01). This model was also used to assess the effect of the intervention on reported pain scores, and the effect of VR did not reduce reported pain scores on average. In fact, there was an increase in reported pain scores by 0.13, though this was not statistically significant (p=0.90). Similar to the effect on FLACC scores, as every minute passed during the 15 minutes of assessment, reported pain scores decreased by -0.11; this was statistically significant (p<0.01).

Furthermore, the amount of pain medication doses needed did not differ for those in the intervention group (M=2.13, SD=0.99) compared to those in the control group (M=2.57, SD=1.13), t(13)=0.81, p=0.47 (see Table 2). Interestingly, the average length of stay was shorter for those who used VR (M=257.88 minutes, SD=130.74) than for those who received standard care alone (M=304.00 minutes, SD=146.74); however, this was not statistically significant, t(13)=0.64, p=0.53.

Discussion

This study set out to evaluate the effectiveness of virtual reality in children experiencing vaso-occlusive crisis from sickle cell disease. Although the statistical analysis does not lend to particularly successful results, it does suggest that VR affects patients' behaviors observed by the researcher. The FLACC scale scores (behavioral scale) were improved with patients who used VR for 15 minutes. This is consistent with the study conducted by Hua, Qiu, Yao, Zhang, and Chen (2015); nurses observed significantly lower FLACC scores in children who used VR while undergoing painful dressing changes. In contrast, patients in our study who used VR did not, on average, self-report a decrease in pain. This differs from the study done by Hua and colleagues (2015), who found a significant reduction in pain reported by patients after the dressing change was finished. However, patients experiencing pain due to vaso-occlusive crisis have pain that lasts much longer than the time it would take to do a dressing change, possibly lasting several days to more than 2 weeks (Fosdal, 2015). One might conclude the observed pain would be lower after a dressing change because the procedure is over, but pain for children in this study still existed after VR was removed. The gate control theory (Li et al., 2011) supports this conclusion in which as soon as VR was removed, the distraction was eliminated, and participants were reminded of their pain. Similarly, results from this pilot study may provide further evidence supporting the gate control theory.

Although results from the linear mixed model did not find a significant difference in the effect of VR on pain scores, it did reveal that with time, both the observed and reported pain scores decreased regardless of whether participants used VR. This is most likely because of the effect of the opioids; everyone in the study received standard care, and opioids decrease pain as the drug is absorbed.

Lastly, though not statistically significant, the average length of stay was shorter when patients used VR. Those in the control group stayed, on average, one hour longer than those in the intervention group. This is an important finding because it may have implications to decreasing costs for these patients in the future.

Limitations

There are limitations to this study, including small sample size and possible researcher bias. According to the G*Power analysis, the sample size should have been 128 with 64 participants in each the control group and the intervention group. The sample size was much smaller than this because there was only one investigator collecting data at one acute care hospital setting, over 14 weeks. The PI was the only one to assess patients' FLACC scales to keep the study findings consistent; however, this method could have contributed to researcher bias. There is also a lack of capability to make generalizations based on results beyond the specific sample population because a convenience sample was used and due to the use of inferential statistics on a very small sample size.

Length of stay could have been increased for a number of other factors that do not relate to the patient's pain, including difficulty obtaining IV access and prolonged waiting time due to ED overcrowding on busy days. Additionally, this study is limited because an objective pain measure (FLACC scale) was used in conjunction with the NRS to assess a subjective pain experience. The Hawthorne effect should also be considered a limitation because participants may have reported decreased pain scores to please the investigator knowing they were part of a research study.

Conclusion

Virtual reality (VR) is an innovative healing modality that may be used for children experiencing severe pain. Pediatric patients experiencing pain from vaso-occlusive crisis are often undertreated for their pain; therefore, nurses can use VR to help decrease their pain without administering more narcotic pain medication. It is easy to use, and children enjoy the experience (Schmitt et al., 2011). Although this study was a pilot study, findings suggest that VR is a distraction technique that eases patients' pain; however, the technique may not have long-term effects when only used for 15 minutes. VR was not effective in decreasing the self-reported pain in patients, and it had no effect on the number of doses of narcotics the patients needed. However, this study helps confirm that the addition of non-pharmacologic treatments may be beneficial for children experiencing pain. Further examination needs to determine why there are differences between self-reported pain levels and observed pain levels. Continued study needs to be conducted using a larger sample size to further evaluate the effect of VR on children experiencing pain crisis due to sickle cell disease in the emergency department.

Using the article above, answer the following questions:

1.Identify the dependent variable.

Identify the independent variable.

2.What is the level of measurement of the dependent variable?

3.What is the total sample size (n)?

4.What measurement tools did the researchers use to collect data?

5.Did the researchers report validity and reliability statistics for the measurement tools? If so, what statistical test was reported? What were the results of the statistical test? What do these results indicate?

6.The researchers reported the following statistical results.

Observed pain scores using the FLACC scale for the experimental group at 5 minutes: mean=1.57, standard deviation=1.51, n=7

Observed pain scores using the FLACC scale for the control group at 5 minutes: mean=4.25, standard deviation=2.12, n=8

t(obtained) = 2.78

t(critical) at 13 degrees of freedom = 2.16

p = 0.01

What is your conclusion based on these results?

7.The researchers reported the following statistical results.

Observed pain scores using the FLACC scale for the experimental group at 15 minutes: mean=1.29, standard deviation=1.70, n=7

Observed pain scores using the FLACC scale for the control group at 15 minutes: mean=3.25, standard deviation=2.12, n=8

t(obtained) = 1.96

t(critical) at 13 degrees of freedom = 2.16

p = 0.07

What is your conclusion based on these results?

8.Why did the researchers perform a power analysis prior to conducting the study? What were the results of the power analysis?

9.What was one major limitation of this study?Which type error (Type I or Type II) is most likely to occur in this study?Define this error.How can the researchers decrease the risk of this type error?