Journal of Pediatric Rehabilitation Medicine: An Interdisciplinary Approach 6 (2013) 193–202 193DOI 10.3233/PRM-130260IOS Press

“Wii-Hab” in critically ill children: A pilottrial

Farah Abdulsatara, Rachel G. Walkera, Brian W. Timmonsa,b and Karen Choongb,c,d,∗aChild Health and Exercise Medicine Program, McMaster University, Hamilton, ON, CanadabDepartment of Pediatrics, McMaster University, Hamilton, ON, CanadacDepartment of Critical Care, McMaster University, Hamilton, ON, CanadadDepartment of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada

Accepted 5 November 2013

Abstract.PURPOSE: To evaluate the safety and feasibility of virtual reality (VR) exercise as a novel acute rehabilitation intervention in aPediatric Critical Care Unit (PCCU) setting.METHODS: Children aged 3–18 years with an anticipated PCCU stay > 48 hours, and baseline normal to moderate cognitiveand functional disability were eligible. Exclusion criteria included: anticipated death, physical inability, or a contraindication tomobilization. Nintendo WiiTM Boxing was prescribed for a minimum of 10 minutes twice a day for 2 days. Primary outcomeswere feasibility and safety.RESULTS: Of 21 eligible patients, 12 (57.1%) were enrolled and 8 completed the study. 41.7% (5/12) were males, and themedian age was 11 (3,16) years. Four of the 8 participants who received the intervention were mechanically ventilated duringWiiTM play. Participants used the WiiTM a median of 2 times (1,5) over the 2-day intervention period, for a median total durationof 54.5 (15, 224) minutes. There were no adverse events attributable to the intervention. Upper limb activity during WiiTM wassignificantly greater than the average daily activity (p = 0.049). Grip strength did not change significantly from baseline (p =0.20).CONCLUSION: While the results of this pilot trial suggest that VR exercise may be safely applied in a subset of critically illchildren, we observed several threats to its feasibility in this population.

Keywords: Pediatrics, critical care, acute rehabilitation, virtual reality gaming

1. Introduction

Patients with critical illness are often bed-riddenfor prolonged periods of time due to perceived com-fort, safety, and maintenance of cardiorespiratory sta-bility. As a result, immobility and deep sedation arecommon prescriptions in pediatric critical care units(PCCU) [1]. The care of these patients has primarilybeen focused on resuscitation and reversal of acute or-gan failure, with little attention paid to the prevention

∗Corresponding author: Karen Choong, McMaster University,1280 Main Street West. Hamilton, ON L8S 4K1, Canada. Tel.: +1905 5212100 Ext 75617; E-mail: choongk@mcmaster.ca.

of weakness and neuromuscular dysfunction, and en-hancement of recovery during critical illness [2]. In-tensive Care Unit-acquired weakness (ICU-AW) is acommon and important sequelae of prolonged bedrestand immobility in the critically ill population. Affect-ing up to 60% of critically ill adults [3,4], ICU-AWcontributes to significant morbidity, hospital length ofstay, and mortality, and adversely affects the long-termfunctional outcomes and quality-of-life in survivors ofcritical illness and their caregivers [5–7]. While theevidence in pediatrics is limited, ICU-AW has alsobeen linked to mechanical ventilator dependence, pro-longed hospitalization, and persistent neuromuscularweakness in this population [7]. Prolonged immobilityand ICU-AW have been implicated as risk factors for

1874-5393/13/$27.50 c© 2013 – IOS Press and the authors. All rights reserved

194 F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial

poor neurocognitive and functional recovery in thesechildren [8].

Accumulating evidence suggests that acute rehabil-itation implemented while still in the ICU in criticallyill adults is safe, feasible, cost effective, and improvesshort-term patient outcomes [9–11]. In contrast, thereis a paucity of this research in pediatrics. While im-mobilization has been implicated as an important con-tributing factor to morbidity and adverse functional,cognitive, and quality-of-life outcomes, the safety andefficacy of acute rehabilitation has yet to be systemat-ically evaluated in critically ill children. Furthermore,there are numerous actual and perceived challengesto applying rehabilitation interventions in the PCCUenvironment given the nature of underlying illnessesand diversity in functional and cognitive abilities [12].Hence, novel approaches to enhance physical activ-ity in this population need to be evaluated. Currentevidence suggests that interactive virtual reality (VR)exercise results in physiological and energy expendi-ture responses similar to traditional forms of physicalactivity [13]. Furthermore, this modality of promot-ing physical activity is popular amongst children andtheir peers, and can elicit better compliance than tradi-tional, structured exercise [14]. We therefore hypothe-sized that VR exercise may be a novel method of com-plementing rehabilitation in the PCCU setting. The pri-mary objective of this pilot trial was to evaluate the fea-sibility and safety of VR exercise as a means of acuterehabilitation in critically ill children.

2. Methods

2.1. Study design and participants

This prospective clinical trial was approved by theHamilton Health Sciences/Faculty of Health SciencesResearch Ethics Board and conducted in the PCCUat McMaster Children’s Hospital. Consecutive patientsaged 3 to 18 years admitted to the PCCU with an an-ticipated length of stay of greater than 48 hours wereeligible. Exclusion criteria were as follows: anticipateddeath or withholding life-sustaining therapy, physicalinability to mobilize (e.g. secondary to upper limbsurgery), cardio-respiratory instability, language bar-rier, and inability to comprehend instructions or per-form the intervention (i.e. deep sedation, and/or severecognitive or functional disability, as defined by poorpediatric overall performance (POPC) and pediatriccerebral performance (PCPC) category scores of �

4) [15]. Informed consent and assent where appropriatewere obtained prior to patient enrolment. The primaryoutcomes of interest were feasibility and safety. Fea-sibility was defined as the ability to enroll participantsover an 8-month study period, implement the interven-tion and study procedures, and adhere to the researchprotocol. Safety was evaluated by the adverse eventrate (defined below). Secondary outcomes included up-per limb activity during the intervention compared tothe rest of the day, muscle strength during the study pe-riod, and caregiver and participant satisfaction with theintervention. Hand-grip strength is a recommended ob-jective tool for screening and monitoring ICU-AW inadults, but has not been validated in critically ill chil-dren [16]. We therefore assessed the feasibility of con-ducting this measurement in the population.

2.2. Study intervention

Nintendo WiiTM Boxing (hereafter referred to asWiiTM) was the specific VR intervention chosen forthis trial. WiiTM Boxing encourages antigravity-move-ment of the upper limbs, requires minimal manual dex-terity, and can be used in a recumbent position, mak-ing it suitable for our target population. One of the in-vestigators (FA) set up the WiiTM for each participantat each intervention period, and instructed the care-givers on its use. The schema for the study period andmeasurements are outlined in Fig. 1. Following enroll-ment, each participant was prescribed WiiTM play fora minimum of 10 minutes, twice a day for a maxi-mum of two days. However, participants were allowedto play longer and/or more frequently if they desired.For voluntary WiiTM -play times outside of the pre-scribed intervention periods, the participant had the op-tion of playing other game choices from the WiiTM

Sport pack. Caregivers were asked to record the fre-quency and duration of each WiiTM play period in anactivity diary.

2.3. Measurements

In order to compare upper-limb activity during theshort intervention period to non-intervention periodsduring the rest of the day, we measured upper-limbmovement using accelerometers, a validated methodfor measuring physical activity in children [17]. We at-tached Actigraph GT3X accelerometers to both wristsof each participant to measure upper-limb movementover the entire study period. Activity counts wererecorded in 3-s sampling intervals (i.e., epochs; time

F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial 195

Fig. 1. Study period and measurements.

period over which activity counts recorded by the ac-celerometer are summed) and data were downloadedand processed at the end of the study period ac-cording to standard procedures previously describedby our group [18]. Rather than transforming activitycounts into energy expenditure, we reported raw countsas an indication of upper-limb movement. Hand-gripstrength was measured by the investigator (FA) at base-line and daily during the study period, using the Grip-A DynamometerTM or the Martin Vigorimeter as ap-propriate for patient size [19].

Baseline demographic data and patient outcomeswere collected on standardized case report forms bythe investigators (FA, RGW). Severity of illness wasmeasured using validated pediatric scoring tools (thePediatric risk of mortality (PRISM III)) [20]. The Pe-diatric logistic organ dysfunction (PELOD) [21], thePediatric Cerebral Performance Category (PCPC), andthe Pediatric Overall Performance Category (POPC)scores on admission were used to quantify cogni-tive and functional ability of each patient at base-line [22]. Physiologic measurements (heart rate, bloodpressure, respiratory rate, and oxygen saturation) wererecorded at baseline, during, and after the interven-tion by the bedside nurse. However, data recorded dur-ing the intervention was only available for 3 patients.Adverse events were defined as the occurrence of anyof the following during the intervention: accidentaltube dislodgment, musculoskeletal injury, pain or dis-comfort (requiring more than the usual patient’s se-dation/analgesia), arrhythmia, persistent tachycardia,hypo-or hypertension, or tachypnea for age, increasedwork of breathing, and fall in oxygen saturation to< 85%. The relationship between adverse events and

the intervention was ascertained by the bedside nurseand/or physician of the participant. At the end of thestudy period, the parents and/or caregivers were askedto indicate their perception of the intervention on aself-administered questionnaire, which included a 7-point Likert Scale and a section for optional commen-tary.

2.4. Statistical analyses

Given our feasibility objectives, we did not spec-ify a sample size apriori; rather, we planned to evalu-ate the number of participants enrolled over an eight-month study period [23]. Descriptive summaries wereused to present baseline patient demographics. Cate-gorical data are reported as percentages and continu-ous data as a mean (standard deviation [SD]) or me-dian (minimum [min], maximum [max]) where appro-priate. The Student’s matched pair t-test was used tocompare cardiorespiratory parameters at baseline andpost WiiTM play, and upper limb activity counts duringWiiTM play and the remainder of the day. One-way re-peated measures ANOVA was used to compare hand-grip strength over the study period. Spearman’s ρ wasused to explore the correlation between WiiTM playtime, severity of illness, and caregiver perception of theintervention respectively. All statistical analyses wereconducted using SPSS Version 18 (www.SPSS.com)and statistical significance was set at p � 0.05. All pre-sented p-values are two-tailed.

3. Results

This trial was conducted from January to August2010. Of 140 patients screened, 100 met exclusion cri-

196 F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial

Fig. 2. Study enrolment flowchart.

teria, 21 were approached for consent, and 12 (57.1%)consented to participate and were enrolled in the trial(Fig. 2). Table 1 presents the baseline demographicsfor each participant. There were 5 (41.7%) males andthe median age was 11 (3, 16) years. Median PRISMIII score on admission was 9.5 (0, 21), and the medianPCPC and POPC scores were 1 (1, 2) and 1 (1, 3), re-spectively.

Four of the 12 enrolled participants did not receivethe intervention (Participants No. 9–12, Table 1): 3were discharged prior to receiving the intervention, and1 withdrew due to dislike for the accelerometer. Themeasurements are therefore presented for the 8 partic-ipants who actually received the intervention, unlessotherwise indicated. Four (50%) of the 8 participantswho received the intervention were mechanically ven-tilated during the intervention. The median time fromPCCU admission to start of the intervention was 9.5days (1, 56). Participants played WiiTM for a medianof 25 (8,120) minutes per day, and a total of 54.5 (15,

224) minutes over the 2 days. One participant playedboth WiiTM boxing and bowling games, while the re-mainder played only WiiTM boxing. Two (25%) of8 patients met or exceeded the exercise prescription.Reasons that the exercise prescription was not met forthe remaining 6 patients are outlined in Fig. 2. With re-spect to clinical outcomes, the median PCCU and hos-pital length of stay was 14 (16, 66) and 23.5 (6, 85)days respectively. There was a significant worseningin participant’s POPC scores from baseline (1 ± 1) toPCCU discharge (2 ± 1) (mean ± SD change of 1.08± 1.00, p = 0.02), whereas the PCPC score was un-changed from baseline to discharge.

3.1. Upper limb activity and muscle strength

Mean upper-limb activity was significantly greaterduring WiiTM play as compared to activity during theremainder of the day when not playing WiiTM (57.12

F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial 197

Table 1Participant demographics

Participant Age Sex PCCU admission PRISM Baseline Baseline Mechanically Time from PCCU Total number Total duration(years) (M*/F†) diagnosis III‡ PCPC§ POPC¶ ventilated admission to of WiiTM of WiiTM

during intervention play times play timeintervention (days) (min)

1 14 M Scoliosis surgery 7 2 1 No 14 1 182 11 F Transverse

myelitis21 1 1 Yes 56 5 157

3 4 F Septic shock 18 1 1 Yes 10 1 154 13 M Diabetic

ketoacidosis10 1 1 No 9 4 224

5 10 M Propionicacidemia withcardiomyopathy

12 2 3 No 1 2 150

6 15 F Septic shock 11 1 1 Yes 19 2 287 12 M Empyema 2 1 2 Yes 3 2 778 16 M Prolonged QT

syndrome andcardiac arrest

5 1 1 No 4 2 32

9 11 F Epiglottitis 0 1 1 NA NA NA NA10 10 F Diabetic

ketoacidosis13 1 1 NA NA NA NA

11 9 F Asthmaexacerbation

2 1 1 NA NA NA NA

12 3 F Neuroblastoma 9 1 1 NA NA NA NA∗M, Male; †F, female; ‡PRISM III, Pediatric Risk of Mortality III score, assessment range 0–74, with higher scores indicating a greater risk ofdeath [20]; §PCPC, Pediatric Cerebral Performance Category; ¶POPC, Pediatric Overall Performance Category; (assessment range for POPCand PCPC scores 1–7, with higher scores indicating greater disability, e.g. 1 = normal vs. 7 = cardiorespiratory death) [22]; NA, Not applicable,intervention not applied; PCCU, Pediatric Critical Care Unit.

Fig. 3. Comparison of upper limb activity during WiiTM play-timeand the rest of the day for participants who received the intervention(n = 8). Upper limb activity expressed as a fold difference in av-erage activity counts during WiiTM play-time compared to activitycounts during the rest of the day.

± 46.60 vs. 9.36 ± 4.12 counts, p = 0.049, n = 8).Figure 3 displays the ratio of average upper-limb activ-ity counts during WiiTM play to the remainder of theday. There was a median 5-fold higher upper-limb ac-tivity during WiiTM play-time as compared to the ac-tivity during the remainder of the day. There was no

significant difference in grip strength on days 2, 3, or 4,represented as a percentage of baseline values (F (3,18)= 1.71, p = 0.20, n = 7). Average grip strength overdays 2, 3, and 4 was 51.3 ± 102.1% of baseline. Themean ± SD for the change in grip strength was 40.4 ±75.9% on day 2, 2.70 ± 116.4% on day 3, and 60.8 ±114.4% on day 4.

3.2. Physiologic and safety outcomes

The mean ± SD baseline vital sign measurementswere as follows: heart rate 103 ± 18 beats per min, res-piratory rate 30± 13 breaths per min, mean blood pres-sure 68± 32, and SPO2 96± 1%. There were no statis-tical differences in any of these physiological measure-ments post WiiTM play as compared to baseline (Ta-ble 2). There were no accidental tube dislodgements,reported changes in pain or sedation requirements, orother adverse events during the study period.

3.3. Caregiver feedback

The response rate for the questionnaire was 75%(6/8). The mean (SD) scores out of a total of 7 for thecaregiver perception of the intervention were 5.7 (±

198 F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial

Table 2Vital sign measurements of the 8 participants at baseline and following WiiTM play

Baseline Post-intervention Change (mean ± SD) p value∗

Heart rate (beats/min)Median (min, max) 108 (64,123) 103 (69, 138) 2.57 ± 14.47 0.65Mean ± SD 103 ± 18 103 ± 22

Respiratory rate (breaths/min)Median (min, max) 30 (9, 50) 28 (13, 38) −0.14 ± 10.96 0.97Mean ± SD 30 ± 13 29 ± 9

Mean blood pressureMedian (min, max) 74 (63, 99) 75 (65, 96) −0.33 ± 2.52 0.84Mean ± SD 68 ± 32 78 ± 13

SPO†2 (%)

Median (min,max) 96 (94,98) 98 (94,100) 1.29 ± 1.50 0.06Mean ± SD 96 ± 1 98 ± 2

∗p value for difference between Baseline and Post-intervention, calculated via Student’s matched pair t-test; †SPO2, oxygen saturation measuredby pulse oxymetry.

Table 3Comments from the parent/caregiver satisfaction questionnaire

Participant Comment1 “I wish that he wasn’t on so much medication for pain which inhibited his drive to want to play. Other than that I think this is

a great idea. I think it helped him with his own feeding by arm strengthening.”2 “Thank you for bringing it she really enjoyed it.”3 “I think the main reason she didn’t want to do the Wii was that the game didn’t appeal to her (she was more motivated by

Dora). She was also quite resistant to any activity, not just this one.”4 “This activity really helped with soreness and stiffness from inactivity. Physically, it loosened up stiff and sore muscles and

mentally took his mind off of what was happening and even brought on some smiles. It helped all around, not just physically.”5 Participant: “it was hard at first but then it got easier. The game was kind of boring.”6 “Answer was neutral [for satisfaction] because he was having a ‘bad day’ and did not play as much as he would have.

Ordinarily, he would ‘LOVE’ Wii and would have played a lot. Also, he had visitors and they provided emotional supportrather than Wii.”

Fig. 4. Caregiver impression of the WiiTM intervention for partici-pants who received the intervention (n = 8). Data are presented asmean (SD) score for each item in the questionnaire. Parent rating isbased on a Likert scale from 1 to 7, with 1 representing “strongly dis-satisfied” or “strongly disagree” and 7 representing “very satisfied”or “strongly agree.”

1.8) for enjoyment, 6.9 (± 0.4) for their impressionof whether the intervention was safe, and 5.3 (± 1.8)for potential benefit provided to their child (Fig. 4).

Fig. 5. Intubated, mechanically ventilated patient during WiiTM

play. (Colours are visible in the online version of the article; http://dx.doi.org/10.3233/PRM-140260)

We found a strong correlation (ρ = 0.86, p = 0.02)between the total WiiTM play-time and caregiver per-ception of the child’s enjoyment with the intervention.A weaker correlation (ρ = 0.73, p = 0.07) was ob-served between total WiiTM play-time and perception

F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial 199

of benefit. There was no correlation between sever-ity of illness (PRISM III score) and total WiiTM play-time (ρ = 0.07, p = 0.88). Table 3 presents some ofthe comments recorded in the questionnaire. Figure 5depicts an intubated, mechanically ventilated patientwhile playing WiiTM.

4. Discussion

Improved mortality amongst adults with critical ill-ness has increased the prevalence of the “post-intensivecare syndrome,” a phenomenon of acquired func-tional, cognitive, and mental health sequelae observedamongst survivors [24]. Long-term sequelae rangingfrom significant functional disability to neuropsycho-logical impairments can be seen in up to 69% of sur-vivors of critical illness, resulting in significant eco-nomic burden to patients and the health care sys-tem [25,26]. While there is a paucity of evidence inpediatrics, it is recognized that children can also suf-fer neurocognitive and functional morbidity follow-ing critical illness [27], which in turn can negativelyimpact school performance, their quality of life, andthat of their caregivers [22]. Prolonged immobility andICU-AW have been implicated as risk factors for poorfunctional recovery in these children [8]. Subsequently,there is growing interest in preventing these significantshort- and long-term critical-illness related morbidi-ties. Emerging research focused on acute ICU-basedrehabilitation in adults suggests that early mobiliza-tion can improve ICU outcomes and recovery follow-ing discharge [28]. In contrast, there is very little re-search amongst children [29].

In recent years, VR technologies such as WiiTM

have been employed in the physical rehabilitation ofpatients. In addition to being novel and enjoyable,VR exercise applies relevant concepts of rehabilitation(i.e. multiple repetitions, varying intensities, and task-oriented training of the extremities) [30]. VR exercisehas been demonstrated to be safe and potentially effec-tive in the rehabilitation of adult stroke patients [31]and in children with developmental delay and cere-bral palsy [32]. It is an attractive means of encouragingphysical activity amongst children as it is fun, popu-lar amongst their peers, can be used by those who arenot accustomed to playing video games, and can en-gage participants in vigorous exercise without perceiv-ing the activity to be difficult [33]. However, while theuse of gaming systems by health care providers is in-creasing, there is still limited research on their poten-

tial therapeutic benefits [34]. Furthermore, VR gamesare not without risks, as injuries associated with recre-ational (i.e. non-rehabilitation) WiiTM use have beenreported [35]. The use of VR games as an adjunctiverehabilitation technique in the critically ill populationhas not been adequately researched.

WiiTM use during physical therapy has been re-ported in a case series of 22 critically ill adults [36],however our prospective study is the first, to ourknowledge, to evaluate the feasibility of VR gaming asan acute rehabilitation technique in critically ill chil-dren. In our small sample of participants, the brief 10-minute per session prescription for VR exercise ap-peared to be safe, and resulted in a significant increasein upper-limb activity during the intervention, whencompared to non-intervention periods of the day. Thepatient’s participation with the VR intervention corre-lated with their enjoyment of the game.

We observed the following threats to feasibility inthis study. The enrollment rate was slower than antic-ipated, primarily due to the limited availability of re-search staff. As the focus of this pilot study was safety,patient stability and physician comfort with the inter-vention had to be fulfilled. These factors and the re-quirement of participants to understand the interven-tion led to a small selection of patients who ultimatelyfulfilled all eligibility criteria. Subsequently, the timefrom PCCU admission to intervention may not nec-essarily be considered “early”, although what definesearly mobilization in the critically ill population is cur-rently debated in the literature [28,37]. The major-ity (75%) of patients did not complete the full 2-dayintervention due to excessive sedation, transfer fromPCCU, or inability to comply with the intervention.This speaks to the narrow patient selection and timeli-ness of this type of intervention amongst critically illchildren; by the time these patients are perceived to bestable enough to participate, they may be close to beingdischarged from the PCCU. Nevertheless, it is impor-tant to emphasise that we were able to apply the inter-vention in 4 of the 8 participants while they were me-chanically ventilated, in whom methods to enhance ac-tive physical activity are limited. Furthermore, we werealso able to apply the intervention in patients who, dueto their critical illness, experienced a deterioration intheir baseline functional ability. The challenges that weexperienced in conducting this trial in a critically ill pe-diatric population are not unique and have previouslybeen identified [38], and can serve to inform future re-search design and efforts.

We assessed primarily one type of VR game in thispilot trial, which, based on the feedback from the ques-

200 F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial

tionnaire, appears to have influenced the degree of par-ticipation. Wii Sport BoxingTM can elicit a moder-ate to vigorous aerobic response, and enable healthyyoung adults to meet the recommended daily physicalactivity levels [33]. The VR game used in this trial wasnot specifically designed for rehabilitation; Wii SportBoxingTM encourages primarily upper limb movementin bed-ridden patients. Furthermore, this game choiceclearly does not appeal to all children. These are im-portant factors to consider not only in the design of fu-ture studies of this nature, but also in the evaluation offactors that influence compliance with a rehabilitationintervention amongst participants and providers. Anx-iety and enjoyment have been identified as factors thatcan influence motivation [39]. As such, we observeda correlation in this trial between enjoyment and com-pliance with the intervention. Conversely, if the childis disinterested, or the benefits are not perceived, theyare less likely to adhere to the intervention [40]. Whilewe did not find a correlation between severity of ill-ness and duration of WiiTM play, we did observe a cor-relation between enjoyment with the game and dura-tion of play. Hence, the variability in total WiiTM playduration in this cohort may indeed have been influ-enced by enjoyment and perception of benefit. We didnot observe any overall change in grip-strength in thisstudy, a secondary clinical outcome objective. Whiledynamometry is a reliable, rapid, and simple method ofmanual muscle testing, which may overcome some ofthe limitations of the Medical Research Council scale,both of these are volitional tests that are affected bya patient’s wakefulness, attention, and motivation [41,42]. Our data suggest that while this method is sug-gested as a simple screening test for ICU-AW, its utilityand validity in PCCU patients requires further study.

Whether acute rehabilitation in critically ill childrencan prevent adverse sequelae of immobility and im-proved functional recovery in these patients is an arearipe for research. While rehabilitation should ideallybe operationalized by physiotherapists who can ap-ply therapies specific to the patients’ needs and mon-itor individual patient progress, this is not feasible inmany PCCUs given resource limitations [43]. Hence,complementary treatments to physiotherapy that othercaregivers can facilitate may be of benefit. There aremultiple unique challenges to operationalizing reha-bilitation strategies in critically ill children, some ofwhich were identified in this trial. First, it remains un-clear which patients are at highest risk of morbidity andhence may benefit most from such interventions. Sec-ond, the majority of children admitted to PCCU have

complex chronic conditions, and heterogenous base-line cognitive and function abilities [44]. Subsequently,some of the exercise interventions currently used incritically ill adults may not be applicable in this pop-ulation. Third, while adult ICU studies have identifiedthat exercise is safe, feasible, and improves clinicaloutcome, the timing, nature, and efficacy of acute re-habilitation in the PCCU is uncertain. Novel rehabili-tation techniques that are safe and effective, yet engag-ing and enjoyable – specifically in the pediatric popu-lation – may encourage participation, self-efficacy, andcaregiver and provider buy-in, which was the rationalefor considering the use of VR exercise to enhance re-habilitation in the PCCU setting. Furthermore, VR ex-ercise is inexpensive (current retail cost of $99.99 perunit), does not require specialized resources or person-nel to implement, and can be used in the inpatient andoutpatient settings. As the benefits of combining VRexercise and rehabilitation are increasingly recognized,this technology may find a role as a complement tophysical therapy in critically ill children. However, fur-ther research is needed to evaluate the impact and costeffectiveness of VR exercise as a rehabilitation strategyin this population.

5. Conclusion

Research on acute rehabilitation in critically ill chil-dren is in its infancy. While VR gaming has beenused effectively in the rehabilitation of specific pedi-atric populations, there is a limited selection of crit-ically ill children in whom this intervention may besafely applied. This pilot trial exposed many chal-lenges and limitations that need to be considered whendesigning future interventional studies on acute VR-based rehabilitation in the PCCU environment. Giventhe assumed morbidity and negative outcomes associ-ated with immobilization in critically ill children, andthe promise that rehabilitation during critical illnessis effective and improves recovery, the magnitude ofthis problem needs to be defined, its determinants eval-uated, and novel complementary approaches in thisunique population need to be considered in order to im-prove the outcomes and overall recovery in criticallyill children.

Acknowledgement

We thank the patients and their families for their par-ticipation in this study.

F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial 201

The listed authors adhere to the ICJME definitionof authorship. All of the listed authors contributed tothe study design, acquisition of the data, analysis andinterpretation of the data, statistical analysis, draftingand final approval of the manuscript.

This study was not funded.

Conflict of interest

The authors report no conflict of interest.

References

[1] I.A. Jenkins, S.D. Playfor, C. Bevan, G. Davies, A.R. Wolf.Current United Kingdom sedation practice in pediatric inten-sive care. Paediatr Anaesth. 2007 Jul; 17: 675-683.

[2] E. Vasilevskis, W. Ely, T. Speroff, B. Pun, L. Boehm, R.Dittus. Reducing iatrogenic risks: ICU-acquired delirium andweakness – crossing the quality chasm. Chest. 2010; 138:1224-1233.

[3] D. Needham, W. Wang, S. Desai. Intensive care unit expo-sures for long-term outcomes research: development and de-scription of exposures for 150 patients with acute lung injury.J Crit Care. 2007; 22: 275-284.

[4] P. Morris, A. Goad, C. Thompson, K. Tayloer, B. Harry, L.Passmore, et al. Early intensive care unit mobility therapyin the treatment of acute respiratory failure. Crit Care Med.2008; 36: 2238-2243.

[5] W. Schweickert, J. Hall. ICU-Acquired Weakness. Chest.2007; 131: 1541-1549.

[6] R. Stevens, D. Dowdy, R. Michaels, P. Mendez-Tellez, P.Pronovost, D. Needham. Neuromuscular dysfunction ac-quired in critical illness: A systematic review. Intensive CareMed. 2007; 33: 1876-1891.

[7] B.L. Banwell, R.J. Mildner, A.C. Hassall, L.E. Becker, J. Va-jsar, S.D. Shemie. Muscle weakness in critically ill children.Neurology. 2003 Dec 23; 61: 1779-1782.

[8] S. Williams, I. Horrocks, R. Ouvrier, J. Gillis, M. Ryan. Crit-ical illness polyneuropathy and myopathy in pediatric inten-sive care: A review. Pediatr Crit Care Med. 2007; 8: 18-22.

[9] P. Bailey, G. Thomsen, V. Spuhler, R. Blair, J. Jewkes, L.Bezdijan, et al. Early activity is feasible and safe in respiratoryfailure patients. Crit Care Med. 2007; 35: 139-145.

[10] C. Burtin, B. Clerckx, C. Robbeets, P. Ferdinande, D. Langer,T. Troosters, et al. Early exercise in critically ill patients en-hances short-term functional recovery. Crit Care Med. 2009;37: 2499-2505.

[11] W. Schweickert, A. Pohlman, C. Nigos, A. Pawlik, C. Es-brook, L. Spears, et al. Early physical and occupational ther-apy in mechaniclly ventilated, critically ill patients: a ran-domised controlled trial. Lancet. 2009; 373: 1874-1882.

[12] K. Choong, K.K. Koo, H. Clark, R. Chu, L. Thabane, K.E.Burns, et al. Early Mobilization in Critically Ill Children: ASurvey of Canadian Practice. Crit Care Med. 2013 Mar 15.

[13] A. Barnett, E. Cerin, T. Baranowski. Active Video Games forYouth: A Systematic Review. Journal of Physical Activity andHealth. 2011; 6: 724-737.

[14] D.E. Warburton, S.S. Bredin, L.T. Horita, D. Zbogar, J.M.Scott, B.T. Esch, et al. The health benefits of interactive videogame exercise. Appl Physiol Nutr Metab. 2007 Aug; 32: 655-663.

[15] D. Fiser. Assessing the outcome of pediatric intensive care. JPediatr. 1992; 121: 68-74.

[16] C.E. Baldwin, J.D. Paratz, A.D. Bersten. Muscle strength as-sessment in critically ill patients with handheld dynamome-try: An investigation of reliability, minimal detectable change,and time to peak force generation. J Crit Care. 2013 Feb; 28:77-86.

[17] S. Trost, P. Loprinzi, R. Moore, K. Pfeiffer. Comparison ofaccelerometer cut points for predicting activity intensity inyouth. Med Sci Sports Exerc. 2011; 43: 1360-1368.

[18] J. Obeid, T. Nguyen, L. Gabel, B. Timmons. Physical activityin Ontario preschoolers: Prevalence and measurement issues.Appl Physiol Nutr Metab. 2011; 36: 291-297.

[19] J. Desrosiers, R. Hébert, G. Bravo, l.E. Duti. Comparison ofthe Jamar dynamometer and the Martin vigorimeter for gripstrength measurements in a healthy elderly population. ScandJ Rehabil Med. 1995; 27: 137-143.

[20] M. Pollack, K. Patel, U. Ruttimann. PRISM III: An updatedPediatric Risk of Mortality score. Crit Care Med. 1996; 24:743-752.

[21] S. Leteurtre, A. Martinot, A. Duhamel, F. Proulx, B. Grand-bastien, J. Cotting, et al. Validation of the paediatric logis-tic organ dysfunction (PELOD) score: prospective, observa-tional, multicentre study. Lancet. 2003; 362: 192-197.

[22] D.H. Fiser, N. Long, P.K. Roberson, G. Hefley, K. Zolten,M. Brodie-Fowler. Relationship of pediatric overall perfor-mance category and pediatric cerebral performance categoryscores at pediatric intensive care unit discharge with outcomemeasures collected at hospital discharge and 1- and 6-monthfollow-up assessments. Crit Care Med. 2000 Jul; 28: 2616-2620.

[23] L. Thabane, J. Ma, R. Chu, J. Cheng, A. Ismaila, L.P. Rios, etal. A tutorial on pilot studies: the what, why and how. BMCMed Res Methodol. 2010; 10: 1.

[24] M.A. Harvey. The truth about consequences-post-intensivecare syndrome in intensive care unit survivors and their fami-lies. Crit Care Med. 2012 Aug; 40: 2506-2507.

[25] M. Herridge, C. Tansey, M. A, G. Tomlinson, N. Diaz-Granados, A. Cooper, et al. Functional Disability 5 years afterAcute Respiratory Distress Syndrome.

[26] M. van der Schaaf, A. Beelen, D.A. Dongelmans, M.B.Vroom, F. Nollet. Poor functional recovery after a critical ill-ness: A longitudinal study. J Rehabil Med. 2009 Nov; 41:1041-1048.

[27] R.O. Hopkins. Does critical illness and intensive care unittreatment contribute to neurocognitive and functional morbid-ity in pediatric patients? J Pediatr (Rio J). 2007 Nov-Dec; 83:488-490.

[28] J. Adler, D. Malone. Early Mobilization in the Intensive CareUnit: A Systematic Review. Cardiopulmonary Physical Ther-apy Journal. 2012; 23: 5-13.

[29] M. Munkwitz, R.O. Hopkins, R.R. Miller Iii, P.M. Luckett,E.L. Hirshberg. A perspective on early mobilization for adultpatients with respiratory failure: Lessons for the pediatricpopulation. J Pediatr Rehabil Med. 2010; 3: 215-227.

[30] A. Henderson, N. Korner-Bitensky, M. Levin. Virtual Realityin Stroke Rehabilitation: A systematic reivew of its effective-ness for upper limb motor recovery. Top Stroke Rehabil. 2007;14: 52-61.

[31] G. Saposnik, M. Levin, G. Outcome Research Canada Work-

202 F. Abdulsatar et al. / “Wii-Hab” in critically ill children: A pilot trial

ing. Virtual reality in stroke rehabilitation: A meta-analysisand implications for clinicians. Stroke. 2011 May; 42: 1380-1386.

[32] Y. Salem, S.J. Gropack, D. Coffin, E.M. Godwin. Effective-ness of a low-cost virtual reality system for children withdevelopmental delay: A preliminary randomised single-blindcontrolled trial. Physiotherapy. 2012 Sep; 98: 189-195.

[33] P. Bosch, J. Poloni, A. Thronton, J. Lynskey. Heart Rate Re-sponse to Nintendo Wii Boxing in Young Adults. Cardiopul-monary Physical Therapy Journal. 2012; 23: 13-19.

[34] M.J.D. Taylor, D. McCormick, T. Shawis, R. Impson, M. Grif-fin. Activity-promoting gaming systems in exercise and reha-bilitation. The Journal of Rehabilitation Research and Devel-opment. 2011; 48: 1171.

[35] T. Fysh, J.F. Thompson. A Wii problem. J R Soc Med. 2009Dec; 102:502.

[36] M.E. Kho, A. Damluji, J.M. Zanni, D.M. Needham. Feasibil-ity and observed safety of interactive video games for phys-ical rehabilitation in the intensive care unit: A case series. JCrit Care. 2012 Apr; 27: 219 e211-216.

[37] C.M. Lee, E. Fan. ICU-acquired weakness: What is prevent-ing its rehabilitation in critically ill patients? BMC Med. 2012Oct 3; 10: 115.

[38] D. Macrae. Conducting clinical trials in pediatrics. Crit CareMed. 2009; 37: S136-139.

[39] S. Yohannan, P. Tufaro, H. Hunter, L. Orleman, S. Palmatier,C. Sang, et al. The utilization of Nintendo WiiTM duringburn rehabilitation: A pilot study. J Burn Care Res. 2012; 33:36-45.

[40] K. Jack, S.M. McLean, J.K. Moffett, E. Gardiner. Barriersto treatment adherence in physiotherapy outpatient clinics: Asystematic review. Man Ther. 2010; 15: 220-228.

[41] C.E. Baldwin, J.D. Paratz, A.D. Bersten. Muscle strength as-sessment in critically ill patients with handheld dynamome-try: An investigation of reliability, minimal detectable change,and time to peak force generation. J Crit Care. 2012 Apr 18.

[42] E.A. Bittner, J.A. Martyn, E. George, W.R. Frontera, M. Eik-ermann. Measurement of muscle strength in the intensive careunit. Crit Care Med. 2009 Oct; 37: S321-330.

[43] R. Cremer, F. Leclerc, J. Lacroix, D. Ploin, G.R.C.D.i.P.S.Group. Children with chronic conditions in pediatric intensivecare units located in predominantly French-speaking regions:Prevalence and implications on rehabilitation care need andutilization. Crit Care Med. 2009 Apr; 37: 1456-1462.

[44] J.D. Edwards, A.J. Houtrow, E.E. Vasilevskis, R.S. Rehm,B.P. Markovitz, R.J. Graham, et al. Chronic conditions amongchildren admitted to U.S. pediatric intensive care units: Theirprevalence and impact on risk for mortality and prolongedlength of stay*. Crit Care Med. 2012 Jul; 40: 2196-2203.