Prism adaptation treatment improves spatial neglect after severe traumatic brain injury: A case series

1 Introduction

Spatial neglect (SN) is characterized by failure or slowness to respond or initiate action towards the contralesional space, not due to a primary motor or sensory impairment (Rode et al., 2017; Heilman and Valenstein, 2012). This cognitive-motor disorder is caused by the disruption of cortical and subcortical networks responsible for visuospatial processing and attention (Corbetta and Shulman, 2011). SN can occur after injury to either cerebral hemisphere, but it is more common, severe, and persistent after right-brain injury (Stone et al., 1993). Most research has focused on the identification, functional impact, and treatment of SN after a stroke (Chen et al., 2015a). SN has been reported in approximately 30% of patients after a stroke and is associated with longer rehabilitation length of stay (LOS), increased risk of falls, decreased likelihood of discharge home, and poorer rehabilitation functional outcomes (Chen et al., 2015b; Wee and Hopman, 2008; Katz et al., 1999, Esposito et al., 2021).

However, SN is not unique to stroke survivors and has been described in other acquired brain injuries, including traumatic brain injury (TBI) (Brain, 1941; Chen et al., 2016; Chen et al., 2020; Mckenna et al., 2006; Riddoch, 1935; Jeong and Min, 2017). SN is estimated to occur in 31–45% of the TBI population undergoing rehabilitation and is associated with increased rehabilitation LOS and decreased functional improvement at discharge (Chen et al., 2016; Mckenna et al., 2006). Due to the high prevalence and negative influence on rehabilitation outcomes, it is essential to establish rehabilitation-specific interventions for SN after TBI.

While multiple strategies exist, there is currently no consensus on the evidence-based treatment of SN. Major practice guidelines for adult stroke rehabilitation recommend prism adaptation treatment (PAT) for SN management (Winstein et al., 2016). PAT is a visuomotor training process involving repeated arm movements to visual targets while wearing prism goggles that shift the visual field horizontally. Once the goggles are removed, the prism after-effect increases spatial exploration of the neglected space (Chen and Hreha, 2020; Rossetti et al., 1998). Repeated PAT sessions strengthen neural connectivity, with four to six sessions believed to be the minimum dose yielding a treatment effect for SN after a stroke (Goedert et al., 2015). Multiple randomized controlled trials in patients after stroke have shown SN improvement after undergoing PAT (Mizuno et al., 2011; Nys et al., 2008; Serino et al., 2009; Vaes et al., 2018). In addition to visuospatial abilities, PAT has also been shown to positively affect motor function and activities of daily living after a stroke (Champod et al., 2018; Goedert et al., 2014).

Interestingly, PAT has not been thoroughly evaluated as an intervention for SN after TBI. There is currently only one published case of post-TBI SN treated with PAT while in the inpatient rehabilitation setting (Jeong and Min, 2017). Another study randomized patients with SN following stroke and TBI into PAT and sham treatment groups (Vilimovsky et al., 2021). Two patients with TBI were included in the analysis (one assigned to PAT, one assigned to sham); however, data was analyzed in aggregate with the majority of data (11 PAT and 10 sham) from patients following stroke. One additional study retrospectively analyzed acquired brain injury patients with SN treated using PAT, including 17 patients post-TBI with potential stroke co-occurrence (Chen et al., 2022). Among patients with Catherine Bergego Scale (CBS) scores before and after PAT, grouped data analysis indicated that having TBI was predictive of greater total and motor functional independence measure (FIM) gains versus stroke. Further research is needed to establish rehabilitation practice standards for the evaluation and treatment of SN after TBI. In this case series, we aim to address this gap in the literature by detailing the cases of six patients after a TBI identified with SN and treated with PAT during acute inpatient rehabilitation, examining PAT dose, frequency, after-effect, and treatment response.

2 Methods

A retrospective chart review was performed on patients diagnosed with severe TBI admitted to a single inpatient rehabilitation facility between November 2018 and June 2022. This retrospective review was determined to meet criteria for an exempt study by the St. Peter’s Health Partners’ Institutional Review Board on 03 March 2022 based on Health and Human Services’ policy on protection of human subjects in 45 CFR 46.101(b) (approval number 22-0222-3). A waiver of informed consent was granted. Patients included were those identified with SN through standardized CBS assessment (score > 0) who received at least one session of PAT before rehabilitation discharge. Patients with a visual field cut, prior stroke history, or TBI and stroke co-occurrence were excluded. SN evaluation and PAT intervention were integrated into the rehabilitation standard of care rather than receiving additional therapy sessions. Patient demographics, TBI severity, neuroimaging findings, and rehabilitation course data were obtained. TBI severity was determined by post-traumatic amnesia (PTA) duration, with severe TBI defined as PTA lasting greater than seven days post-injury.

Occupational therapists (OTs) screened for SN on day four of the rehabilitation admission using a standardized CBS via the Kessler Foundation Neglect Assessment Process (KF-NAP^®) (Chen et al., 2015a; Chen and Hreha, 2015; Azouvi, 1996). The KF-NAP^® is a standardized method to administer the CBS, a 10-item observational tool for evaluating the functional impact of SN, including limb awareness, personal belongings, dressing, grooming, gaze orientation, navigation, collisions, eating, and cleaning up after a meal (Chen and Hreha, 2015). Each item receives a score of 0 (no neglect), 1 (mild neglect), 2 (moderate neglect), or 3 (severe neglect). The total score is calculated by the total sum score divided by the number of scored items multiplied by ten. The total score ranges from 0 to 30, indicating SN severity as mild (CBS score 1–10), moderate (CBS score 11–20), or severe (CBS score 21–30) (Chen and Hreha, 2015; Azouvi, 1996). If patients were unable to participate in the initial CBS evaluation, OTs continued to assess the feasibility by monitoring wakefulness, agitation, command following, and participation in activities of daily living (ADLs). The CBS was performed once appropriate, based on clinical judgment.

When patients were identified with SN, defined as a CBS score≥1, trained OTs administered PAT sessions using the Kessler Foundation Prism Adaptation Treatment (KF-PAT^®) protocol and portable kit (Chen and Hreha, 2020). Each session required approximately 15 minutes to complete, consisting of 60 arm reaching movements while wearing goggles fitted with 20-diopter wedged prism lenses that shift the visual field 11.4° horizontally. One patient with delayed processing required additional time to complete sessions but did not wear goggles for greater than 20 minutes as recommended by the KF-PAT^® protocol. Those with left SN received PAT with prism lenses thicker on the left for a leftward or negative value after-effect, and patients with right SN received PAT with prism lenses thicker on the right for a rightward or positive value after-effect (Chen and Hreha, 2020). The recommended PAT dose for SN after a stroke is ten sessions over two weeks (Chen and Hreha, 2020). OTs documented the number of PAT sessions, barriers encountered, and duration over which PAT was given.

Before and after each PAT session, OTs administered standardized proprioceptive pointing (PP) and visuoproprioceptive pointing (VPP) tasks to determine the after-effect immediately following prism removal. During PAT, repetitive arm movements result in visuomotor system adaptation and increased exploration of the neglected side. After the prism goggles are removed, the visual information is no longer shifted. However, there is a persistent exploration of the neglected side due to visuomotor adaptation, referred to as the after-effect (Chen and Hreha, 2020; Fortis et al., 2010; Frassinetti et al., 2002). The after-effect for left SN increases leftward spatial exploration for pointing tasks, while the after-effect for right SN increases rightward spatial exploration for pointing tasks. PP and VPP task values were recorded before and after each PAT session to assess for prism after-effect (Chen and Hreha, 2020). Statistical significance testing was performed using IBM SPSS version 26 (IBM Corporation, 2019) and G*Power (version 3.1.9.7).

3 Results

The sample consisted of six male patients who sustained a severe TBI, after either a fall (n = 3), motorcycle accident (n = 2), or gunshot wound (n = 1) with a mean age of 50.5 years (range: 20–87). Varying types of intracranial pathology were present, including intraparenchymal hemorrhage, subdural hematoma, subarachnoid hemorrhage, and epidural hematoma. Four patients had right-sided injuries and two had left-sided injuries, occurring in either the frontal or temporal regions, with no patients having diffuse axonal injury. All six patients had evidence of mass effect on neuroimaging. Five patients underwent neurosurgical intervention, including craniectomy (n = 3) and craniotomy (n = 2). The average hospital LOS prior to rehabilitation admission for cases one through five were 24.8 days (range: 6–39). Approximately 8 months had elapsed between TBI onset and the rehabilitation admission where PAT was performed for case six, during which the patient underwent cranioplasty. This patient had previously undergone acute rehabilitation, where he was unable to participate in CBS assessment or PAT due to severe cognitive deficits identified through the patient’s inability to complete the Brief Interview for Mental Status and limited memory/recall ability on the Staff Assessment for Mental Status (Section C of the CMS-issued IRF-PAI, Version 3.0). Neuropsychology consultation confirmed attention, orientation, and cognitive linguistic deficits. The mean rehabilitation LOS where PAT was performed was 32.8 days (range: 20–60) (Table 1).

The initial CBS assessment was attempted on day four of the rehabilitation admission when PAT was performed. However, it was only able to be completed with three patients. The remaining patients underwent initial CBS assessment with an average delay of 12 days (range: 8–18). Multiple barriers to completing initial CBS were present in three cases, including the presence of a disorder of consciousness, poor arousal with limited command following, agitation due to a post-traumatic confusional state, and severe cognitive deficits. Patients were identified with mild (n = 2), moderate (n = 2), or severe (n = 2) SN, with a mean CBS score of 16.8 (range: 8.75–24.3). Four patients were found to have left SN, with right SN occurring in two patients (Table 2). For three of the six patients, handedness corresponded to the neglected side.

From admission to the first PAT session, the mean duration was 10.5 days (range: 5–19). During the rehabilitation course, patients received an average of 8.8 PAT sessions (range: 5–10). All patients completed each PAT session without protocol deviations. On average, PAT sessions occurred over 12.7 days (range: 5–18) with a frequency of 0.75 sessions/day (range: 0.56–1.0). A mean negative after-effect was seen for all four patients with left SN, and a mean positive after-effect was present for the two patients with right SN for both pointing tasks (Table 2). Cases 3 and 4 could not receive 10 PAT sessions during their inpatient rehabilitation stay due to LOS limitations; upon final CBS assessment it was determined that their spatial neglect had resolved and further PAT sessions were not necessary.

The average discharge CBS score was 5.2 (range: 0–12.2), with the mean CBS score decreasing by 11.6 points (range: 7.8–21.8). Cases one to five, who were within 100 days of their TBI, experienced a clinically meaningful improvement in SN severity after undergoing PAT, with severe SN improving to mild or moderate SN, moderate SN improving to mild SN, and mild SN resolving. Case six, who underwent PAT approximately 8 months post-injury, remained within the moderate SN classification but his CBS score improved by 7.8 points. Patients were discharged to either home (n = 3) or a subacute rehabilitation facility (n = 3) (Table 2).

To calculate effect size, the means and standard deviations of the initial and final CBS scores were used. Pearson’s r for the initial versus final CBS scores was determined to be 0.64 (SPSS v26, 2019 IBM Corp). Because of the large effect size (2.28) observed for the initial versus final CBS scores, significance testing could be performed. Using a two-tailed Wilcoxon Signed-Rank Test, a sample size of 6 with the observed effect size achieves a power of 0.99 (alpha error probability = 0.05, G*Power v3.1.9.7). Despite the small sample size, the change in initial versus final CBS scores was significant (P = 0.031).

This study presented six cases of SN after TBI, detailing the successful use of PAT integrated into rehabilitation standard of care for SN management. Four patients with left SN after right-brain injury and two patients with right SN after left-brain injury were identified by standardized CBS administration with mild (n = 2), moderate (n = 2), or severe (n = 2) SN. On average, these patients received 8 PAT sessions at a frequency of 0.75 sessions/day. All patients showed prism after-effect and improved SN symptoms after receiving PAT, with a statistically significant mean decrease of 11.6 points through standardized CBS. For patients within 100 days of TBI, this clinically translated into a decrease in SN severity classification, with severe SN improving to mild or moderate, moderate SN improving to mild, and mild SN resolving. This case series is the largest to date showing the successful treatment of SN after TBI with PAT in the rehabilitation setting, and the first to demonstrate evidence of prism after-effect and functional SN improvement measured by the CBS.

Previous studies have used varying assessments to diagnose SN after TBI, including the unilateral neglect subscale of the occupational therapy adult perceptual screening test, the visual and spatial perception subsets of the Loewenstein occupational assessment (LOTCA), the star cancellation test component of the behavioral inattention test, line bisection, letter cancellation, and the CBS via the KF-NAP^® (Chen et al. 2016; Mckenna et al., 2006; Jeong and Min, 2017). The CBS uniquely assesses the impact of SN on functional performance in the personal, peri-personal, and extra-personal spaces through direct observation of spontaneous behaviors during ten everyday activities (Chen and Hreha, 2015; Chen et al., 2012). However, there is no established standard for SN evaluation after TBI.

Completing the initial CBS assessment was challenging in this case series due to multiple patient-specific factors. Identified barriers that prevented the initial CBS evaluation included disorders of consciousness, poor arousal with limited command following, agitation due to a post-traumatic confusional state, and severe cognitive deficits. These patients were monitored for clinical improvement, including improved wakefulness, following simple motor commands consistently, decreased agitation, and increased participation in activities of daily living through occupational therapist assessment of eating, grooming, oral hygiene, toileting, bathing, dressing, and toilet/shower transfer during therapy sessions. Once patients met these clinical criteria, the CBS was completed with an average delay of 12 days (Table 2). In addition, case six was unable to undergo CBS assessment during his inpatient rehabilitation course due to severe cognitive deficits. Reassessment on his second admission, 8 months after his TBI, showed moderate SN that improved with PAT. This highlights the importance of re-assessing patients for SN who may not have initially been candidates for CBS evaluation. Additional research is needed to establish the ideal assessment measure for SN evaluation after TBI.

There is minimal literature evaluating PAT for SN management after TBI, with no current case studies performing a functional SN assessment before and after PAT in patients with a TBI. One previous report details a case of right SN after TBI successfully treated with 10 PAT sessions over two weeks (Jeong and Min, 2017). The patient sustained a traumatic subarachnoid hemorrhage resulting in tetraplegia and moderate cognitive impairments. Improvement in SN was evaluated by the LOTCA visual and spatial perception subsets, line bisection, and letter cancellation, with improvement in upper extremity motor function, ambulation, and independence with activities of daily living. However, a functional SN assessment, such as the CBS, was not administered before and after PAT. Also, the presence of prism after-effect after each PAT session was not described.

Another study included two patients with TBI (one receiving PAT and one receiving sham) along with patients after stroke in a randomized pilot trial evaluating SN treatment (Vilimovsky et al., 2021). One additional study retrospectively analyzed functional recovery of SN after receiving PAT in patients across acquired brain injury diagnoses, including 17 patients with TBI (Chen et al., 2022). Both of these studies utilized the CBS for functional assessment; however, because of the aggregate nature of the studies, there is no insight into the SN improvement specific to TBI in the individual receiving PAT. Further research studies are needed to establish PAT efficacy for the management of SN after a TBI.

This study is the first to assess the impact of PAT on functional SN outcome measures for right and left SN after a TBI in individual patients, showing a statistically significant mean CBS score improvement of 11.6 after receiving an average of 8.8 PAT sessions (Table 2). Multiple recent retrospective-matched studies have shown the functional benefit of PAT for SN treatment while in the inpatient rehabilitation setting. One study found that patients with SN receiving PAT had greater functional and cognitive improvements at rehabilitation discharge, assessed through FIM score change (Chen et al., 2021). Another study showed that patients after stroke with severe SN receiving an average of 8 PAT sessions had significant median CBS and FIM score improvements by 10 and 25 points, respectively (Gillen et al., 2022). This CBS score change is similar to the findings of the current case series, supporting the use of PAT for improved rehabilitation outcomes in patients after TBI with SN. Additional studies are needed to clarify the functional impact of PAT for SN treatment after TBI.

The ideal timing and quantity of PAT sessions for SN management after TBI has not been established and is based on SN stroke research. The current recommended PAT dose for SN after a stroke is ten sessions over two weeks (Chen and Hreha, 2020). However, no evidence-based guidelines support this protocol, with randomized controlled studies showing variable SN improvement (Serino et al., 2009; Ten Brink et al., 2017; Turton et al., 2010). Other studies assessing four to seven PAT sessions have also shown mixed outcomes, with some studies reporting positive results in as few as one to two sessions (Nys et al., 2008; Vaes et al., 2018; Mancuso et al., 2012; Fortis et al., 2011; Saj et al., 2013). These studies were performed in varying stroke populations, within the subacute to chronic phase of recovery, further complicating these findings. Overall, the minimum PAT dose resulting in treatment effect for SN after a stroke is believed to be four to six sessions, persisting for three to four weeks after treatment (Goedert et al., 2015). All patients in this study received at least the minimum effective PAT dose and had decreased SN symptoms and severity. The five patients receiving PAT within 100 days of injury showed improvement in SN severity, with severe SN improving to mild or moderate, moderate SN improving to mild, and mild SN resolving. The one patient who underwent PAT eight months after TBI had a CBS improvement of 7.8 points but remained in the moderate SN category (Table 2). PAT dose and timing to induce a treatment effect in SN after TBI is unknown, warranting additional research.

Improved spatial exploration of the neglected hemispace after a PAT session is assessed by prism after-effect. Research performed on SN after a stroke suggests that patients who do not experience prism after-effect will not benefit from PAT for SN treatment (Fortis et al., 2010; Frassinetti et al., 2002). To date, no studies have evaluated prism after-effect in patients with SN specifically following TBI. All six patients in this case series showed prism after-effect with PP and VPP tasks. The four patients with left SN showed a more leftward mean after-effect for pointing tasks after PAT. The two patients with right SN showed a more rightward mean after-effect for pointing tasks after PAT (Table 2). Two patients (cases 2 and 4) exhibited greater VPP than PP, potentially due to visual deficits. Vision exams for these patients indicated intermittent blurred vision for case 2 and mild blurriness reported by case 4. This case series is the first to show that prism after-effect is experienced with PAT for SN after TBI. Additional studies are needed to determine if prism after-effect in pointing tasks is predictive of SN improvement after TBI, potentially identifying patients who would most benefit from PAT.

Injury characteristics may contribute to developing SN after TBI, including intracranial lesion location and mass effect on neuroimaging. After a stroke, SN manifests as decreased awareness of the contralesional space (Rode et al., 2017; Heilman and Valenstein, 2012; Corbetta and Shulman, 2011). However, pathophysiology significantly differs between stroke and TBI. While a stroke is classically unilateral and focal, TBI can be focal, multifocal, or diffuse. This may contribute to differences in SN presentation after stroke and TBI, with studies showing diffuse brain injury after TBI increases the risk of left SN (McKenna et al., 2006; Pavlovskaya et al., 2007). In addition, the presence of injury-related mass effect on neuroimaging has also been associated with left SN (Chen et al., 2020). The current study consisted of patients after a severe TBI with either focal or multifocal intracranial hemorrhage localized to one cerebral hemisphere, without diffuse axonal injury (Table 1). The lack of diffuse brain injury may explain why right SN occurred after left-sided brain injury and left SN was seen after right-sided brain injury. In addition, all patients had evidence of mass effect on neuroimaging, with five requiring surgical decompression (Table 1). Additional studies are needed to clarify the underlying mechanisms and risk factors for developing SN after TBI.

There are multiple limitations to this case series. First, patients who had SN after TBI but did not receive PAT during inpatient rehabilitation were not included in this study, potentially introducing bias. Second, due to the retrospective nature of this study, SN assessment measures were limited to those completed through the rehabilitation standard of care. Finally, natural recovery of SN after TBI or other rehabilitation interventions performed may have contributed to SN improvement.

5 Conclusion

There is minimal evidence-based guidance for identifying and managing SN after TBI in the rehabilitation setting. While PAT has been successfully used to treat SN following a stroke, it has not been thoroughly evaluated for SN after TBI. This study details six cases of SN after severe TBI who received five to ten PAT sessions as part of the rehabilitation standard of care, showing evidence of prism after-effect and functional SN improvement through CBS. Additional studies are needed to verify the efficacy of PAT for SN after TBI, clarify the PAT timing and dose needed to induce treatment effect, and assess PAT benefit for mild, moderate, and severe SN after TBI.