Consensus Statement on Visual Rehabilitation in Mild Traumatic Brain Injury

Abstract

Optometric visual rehabilitation therapy has been used for a variety of visual disorders. Descriptively named entities such as posttrauma visual syndrome, visual midline shift syndrome, and vertical heterophoria syndrome are frequently diagnosed by neuro-optometrists and/or behavioral optometrists in patients after stroke or head injury or in the setting of dizziness and/or headache. The scientific underpinnings of these diagnoses and treatments are weak, and published clinical studies comprise case reports and case series without comparison to control populations. Neuro-ophthalmologists are frequently questioned by patients about the utility of such treatment strategies. Many ophthalmologists and neurologists also are involved in the care of patients who carry these diagnoses and undergo these visual therapies. Involved physicians may benefit from guidance about the rationale, evidence, and level of evidence for the efficacy of these therapeutic approaches.

Traumatic brain injury (TBI) is a major cause of disability and death globally, and the associated cost continues to rise.^1,2 Mild TBI (mTBI) can be even more problematic than moderate or severe TBI from both an economic and diagnostic standpoint.¹ In moderate and severe TBI, the brain damage is obvious, but in mTBI, the damage is more subtle yet may still lead to both short- and long-term neurologic dysfunction. The terms concussion and mTBI often are used interchangeably, and repeated concussions, as might occur in sports or in the military, can lead to an aggregation of disability.³ An estimated 42 million people worldwide have a concussion every year.³ Although vehicular and industrial accidents are the most common causes of mTBI in adults, sports and bicycle accidents account for the majority of cases among 5- to 14-year-olds, usually from a fall.

Concussion is most frequently diagnosed and treated in acute care settings such as the emergency department or primary care provider office by physicians or advanced practitioners. Later, the diagnosis is perpetuated during follow-up with primary care physicians, neurologists, chiropractors, physical and occupational therapists, speech-language pathologists, and psychologists, leading to recommendations for tertiary levels of care with neurorehabilitative specialists (e.g., neuropsychologists, neurologists, neuro/behavioral optometrists, speech-language pathologists, and brain injury rehabilitation centers). For these later care providers, the initial observations of injury, the patient's response to injury, and symptoms immediately following the injury are essential information for establishing an appropriate diagnosis.

Despite the advances in diagnostic imaging and disease-specific biomarkers over the past several decades, the diagnosis of mTBI remains a clinical one.^4,5 Coupled with the often-confusing diagnostic taxonomy and heterogeneity in the case definitions of mTBI, this leads to costly and inaccurate under- and over-diagnoses. This imprecision may obscure the differentiation of functional neurologic disorders and/or even enable malingering in injuries involving liability, which in turn may lead to unwarranted studies and treatments that lead to excessive health care costs and potential harm.⁶

This consensus statement will focus on the commonly accepted elements that define the entity of mTBI⁴ from the World Health Organization Collaborating Task Force and the Centers for Disease Control and Prevention definition of mTBI:

mTBI is an acute brain injury resulting from mechanical energy to the head from external physical forces. Operational criteria for clinical identification include: (1) one or more of the following: confusion or disorientation, loss of consciousness for 30 minutes or less, post-traumatic amnesia for less than 24 hours, and/or other transient neurological abnormalities such as focal signs, seizure, and intracranial lesions not requiring surgery; and (2) Glasgow Coma Scale of 13-15 after 30 minutes post-injury or later upon presentation for healthcare. (3) These manifestations of mTBI must not be due to drugs, alcohol, medications, caused by other injuries or treatment for other injuries (e.g., psychological trauma, language barrier, or coexisting medical conditions) or caused by penetrating craniocerebral injury.⁴

The economic burden of mTBI extends beyond morbidity, disability, and mortality and includes varied facets of treatment, labor costs, and effects on the fabric of the family and society. The financial cost may be staggering^1,7 yet poorly documented. For example, visual rehabilitation given to injured workers in Washington State over an 8-year period for a diagnosis of convergence insufficiency (CI), a common symptom in mTBI, costs the system $250,000 annually. In this cohort, independent examination by 1 author (C.S.) found that less than half of those treated actually had CI, and in that same period, only 18% of a consecutive cohort of 782 patients diagnosed with mTBI actually met the World Health Organization criteria for the disorder. Although VR is only a small component of the package of mTBI treatment paradigms, this example illustrates that a significant amount of the resources directed at treating CI may be misdirected. The purpose of this work is to review the current status of evidence regarding diagnosis and management of visual disorders in mTBI and to propose further work that will improve our care of patients with mTBI. We searched the English language literature using PubMed and used search terms that combined optometric vision therapy and TBI to identify a broad range of existing publications and reports.

Visual Diagnoses in mTBI

mTBI can be associated with disruption of both visual input to the brain (afferent visual system) and brain control of ocular motor systems (efferent system). A review of the potential disruptions in these systems in mTBI starts with both the clinical history and a careful assessment of the neurologic and ophthalmologic systems.

Efferent Visual Diagnoses in mTBI

Disorders of ocular alignment and coordination of binocular movement may be a defining feature of acute mTBI, as evidenced by research and subsequent development of tests that seek to permit rapid and reliable diagnosis of mTBI immediately after head trauma.⁸ These abnormalities usually remit within weeks,⁸ but subjects with mTBI may have ongoing symptoms including diplopia and dizziness for which the exact cause may be unclear. If assessment of ocular motor dysfunction is used in both the definition of mTBI and to suggest recovery from it, this presents a potential confounder in determining the value of an oculomotor rehabilitation treatment.

Diplopia

Diplopia is described frequently by subjects after mTBI, and trauma is among the most common causes of diplopia, potentially affecting cranial nerves, extraocular muscles, vergence, and supranuclear pathways associated with ocular alignment (Table). In mTBI, the incidence of direct injury to these structures is very low. Nonetheless, injury to any of the ocular motor nerves (cranial nerves [CNs] III, IV, and VI) may occur,⁹ although it has been suggested that CN IV is more susceptible to relatively minor trauma than CNs III and VI. Although CN injuries are rare in mTBI, diplopia has been reported by over 40% of US military veterans who had an mTBI from blast and nonblast causes; thus, although a supranuclear disturbance of ocular motor function is suspected, the precise location of the dysfunction is not known.¹⁰ Many subjects with mTBI with diplopia may have normal eye movements and small angles of strabismus, including phorias, that are better appreciated with Maddox rod testing than with prism cover evaluation.¹¹ Ocular motor alignment is important for optimal performance of many daily activities, such as self-care, reading, spatial awareness, limb coordination, postural control, and navigation. Correlation of diplopia with afferent TBI symptoms such as photophobia, blurred vision, and difficulty reading has not been demonstrated, and afferent symptoms and diplopia symptoms do not always resolve in conjunction.

Table.

Efferent Visual Complications of Traumatic Brain Injury

Convergence Insufficiency

As noted, CI is said to be present in about a third of patients with mTBI,¹² and some reading difficulties (an afferent symptom) in mTBI may reflect CI. Among the major unanswered questions about this vergence problem are the duration of symptoms and identifying predictors of recovery. Studies of civilian and military subjects with TBI demonstrate that the most will have symptom improvement or resolution within 6 months, but a subset will have persistent symptoms that continue to cause functional deficits. Identifying this subpopulation remains difficult, and no tools have been developed that reliably predict recovery of vergence. Finding parameters connected to poor recovery may allow for early identification of patients who need treatment while allowing those who are likely to improve spontaneously to avoid the time and expense of treatments, should they prove to be effective.

Dizziness and Imbalance

Other areas of reported dysfunction are saccadic dysmetria and vestibular disorders (Table). Patients may exhibit fatigable saccadic accuracy, and increased saccadic latency has been proposed as a characteristic sign. It is not clear whether these findings reflect ocular motor dysfunction or broader issues with fatigue and attention. It is also not clear (and perhaps unlikely) that they are directly responsible for any visual symptoms, as neurologists frequently observe similar, asymptomatic saccadic abnormalities in patients with nontraumatic neurologic diseases. Nystagmus is rare in mTBI, but dizziness is common and may have a variety of causes.¹³ A subset of patients who experience dizziness will report symptoms being triggered by complex visual environments and/or rapid motion within the field of view, and this condition may represent persistent postural-perceptual dizziness¹⁴

Potential Relationship to Blast Injury

Whether the likelihood of visual and ocular motor dysfunction varies with the type of injury remains to be determined. One potentially important subgroup is that of blast injuries, often seen in the military. Nearly 80% of US military personnel mTBIs incurred during post-9/11 deployments were related to blast, and blast exposure results in more diffuse brain injury compared with non–blast-related mTBI.¹⁵ Blast thus may lead to a greater multifocal effect on the brain structures associated with ocular motor control (e.g., frontal and prefrontal cortices and cerebellum). Although there have been editorial opinions in academic journals calling for treatment of these problems, a lack of mechanistic understanding hampers our ability to prescribe directed and proven effective therapies.

Afferent Visual Diagnoses in mTBI

Although damage to the retinostriate pathway or optic nerves is rare in mTBI, a myriad of visual symptoms including blur, light sensitivity, difficulty reading, motion sensitivity, and eye fatigue have been reported to be associated with mTBI, and a variety of theories have been proposed to explain their genesis.

Postconcussion Syndrome

In the absence of injury to the eyes or visual pathways, afferent symptoms may arise as part of a postconcussive syndrome, a condition in which symptoms persist beyond the expected recovery period. As noted before, diagnosis of mTBI after the acute event⁵ remains dependent on historical data from the subject and witnesses at the scene. Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, (DSM-IV) and International Classification of Diseases, 10th Revision, (ICD-10) criteria for diagnosing postconcussive syndrome require the diagnosis of mTBI as a first step, along with the persistence of 3 or more symptoms¹⁶ that can be grouped into cognitive, somatic, affective, and sleep-arousal domains.¹³ Visual symptoms may be included in the somatic category, although they are not specifically listed as part of the DSM-IV or ICD-10 criteria.¹⁶ Nevertheless, blur and diplopia are included in other instruments for assessing the postconcussive state, such as the Rivermead Postconcussion Symptoms Questionnaire.¹⁷

The diagnosis of postconcussive syndrome is agnostic regarding the etiology or mechanism behind the symptoms. These can include cerebral effects on cognition; vestibular or ocular motor dysfunction; modulating effects of depression, anxiety, and other psychological factors; substance abuse; and functional disorders.⁵ However, there are other diagnostic terms in use that attribute a range of visual and nonvisual symptoms to impaired visual processing in mTBI without known mechanistic explanations.

Posttrauma Vision Syndrome

This term refers to a set of symptoms and signs attributed to an imbalance of focal and ambient visual systems.¹⁸ The explanation is speculative, and there is little proof that the wide variety of visual and nonvisual symptoms of Posttrauma vision syndrome all originate in visual or ocular motor dysfunction. Indeed, given the potential for diffuse injury like mTBI to affect multiple systems, including vestibular, cognitive, psychological, and musculoskeletal function, it seems improbable that visual dysfunction alone is causing the many complaints of patients with mTBI.

Midline Shift Syndrome

This diagnosis refers to a deviation of the subjective midline due to a mismatch between visual and proprioceptive information about the body's center of mass.¹⁹ Such deviations resemble findings with egocentric hemineglect from right parietal lesions or past pointing with unilateral vestibular lesions,²⁰ but it is not clear that similar anatomic problems underlie midline shift in mTBI and its reported presence has only a subjective basis. No anatomic explanation exists as to why midline shift should occur with mTBI.

Vertical Heterophoria Syndrome

This diagnosis proposes that mTBI can cause a small vertical misalignment of the eyes from utricular or brainstem damage.²¹ The otoliths are vulnerable to head trauma,²² and proponents of this syndrome assert that it is responsible for symptoms ranging from blur, diplopia, and dizziness to anxiety, headache, light sensitivity, and feeling overwhelmed in crowds.²¹ However, such broad symptoms are not typical in subjects with nontraumatic otolithic dysfunction.²³ Furthermore, there is no evidence that these mTBI symptoms correlate with signs of otolith damage with vestibular testing. Of concern, the diagnosis of vertical heterophoria syndrome (VHS) has been based simply on whether subjects report symptomatic improvement with the use of a vertical prism, rather than on objective measures of ocular alignment. In addition, other studies have failed to find increased prevalence of vertical dysconjugacy in subjects with mTBI compared with healthy subjects.²⁴

Adjunctive Diagnostic Tests for Afferent and Efferent Disorders

Standard assessments of convergence and accommodation have been described elsewhere.²⁵ Measures of exotropia at near vs far and the near point of convergence are easily incorporated into a clinical examination, and recording quantitative assessments of both accommodation and convergence can aid in determining improvement over time. Autorefractors can be useful in diagnosing altered accommodative dynamics, and appropriate refractive correction should be provided to patients before CI is diagnosed.²⁶ The Convergence Insufficiency Symptom Scale appears to be a reliable and efficient way of following symptoms in adults.²⁷

The Developmental Eye Movement Test times patients while they read numbers in vertical columns and then in horizontal arrays, with the ratio between the 2 taken as an indirect index of horizontal saccadic efficiency. This test has been recommended as a measure of saccadic function in mTBI and to assess its response to vision therapy.²⁸ However, the results of this test do not actually correlate with any direct measure of saccades, although they do correlate with measures of reading and visual processing speed.²⁹ Of concern, the test's poor between-session reproducibility in children has raised concerns about its utility as a diagnostic or monitoring tool.³⁰

The King-Devick test is a very similar test, also originally designed to be part of a reading evaluation, but now promoted as a tool for acute concussion assessment. Patients read numbers arranged in rows on 3 screens with successively decreasing spacing between the lines, and the reading time and error rate are recorded. It appears to be an indirect measure of saccadic speed and accuracy. Although there is debate about its accuracy for detecting concussion, there are few data on its use in diagnosing or monitoring postconcussive syndrome.⁸

Neither the Developmental Eye Movement Test nor the King-Devick Test should be interpreted as indices of specific visual or ocular motor functions. Just as with real reading, many functions contribute to performance on these tests, including saccades, fixation stability, convergence, accommodation, attention, language ability, and cognitive processing speed.

Treatment

Ocular Motor Training

Ocular motor training is a popular approach that involves a series of in-office visits for eye movement exercises intended to treat disorders of vergence, versions, and/or accommodation. The multicenter Convergence Insufficiency Treatment Trial, a randomized placebo-controlled study of children and young adults with primary CI, provided evidence of the efficacy and superiority of office-based treatments over home exercises and placebo treatment.³¹ Proof that accommodative or convergence therapy works in the mTBI variant is lacking, however, as recognized by a recent pilot study.³² There is 1 retrospective review³³ and 1 prospective uncontrolled study of 19 patients³⁴ that claim benefit. Improvement in vergence and accommodation was reported in a small series of 12 subjects with mTBI using a crossover design with placebo vs training on 3 exercises.²⁶ However, methodologic issues with this study led a Cochrane review to classify its evidence having very low certainty.³⁵ Noted issues included the small sample size, drop out of 3 of the 12 subjects, flawed statistical analyses that relied on t tests that were not corrected for multiple comparisons, incongruent results from the ANOVAs that were performed, and the use of a crossover design for a situation that might not permit a washout period.³⁵

The utility and efficacy of training saccades and pursuit in mTBI are more uncertain. Claims of benefit are complicated by the fact that with one exception for pursuit,³⁶ all studies have combined training of versions, vergence, and accommodation in their treatment program, making it difficult to isolate the effects of each intervention. Although some data suggest benefit in symptoms and saccadic measures with training, studies in more severe TBI showed that ocular motor deficits also improved with time in untreated subjects.³⁷ Similarly, a recent prospective randomized pilot study of 14 hospitalized patients with TBI with oculomotor deficits showed improvement with either structured eye exercises or standard of care.³⁸ We thus conclude that there are some grounds but insufficient proof for using convergence or accommodative treatment in mTBI, and the rationale or evidence for training version eye movements such as pursuit and saccades is lacking.

Binasal Occlusion

Anecdotal case reports have claimed that placing opaque tape or filters on the nasal half of a patient's spectacle lenses improves visually provoked dizziness, balance problems, sensitivity to busy environments, and movement-induced eye pain in mTBI.³⁹ The rationale for this intervention is unclear. Some authors claim that the line of tape provides a stable frame of reference in the visual field; others suggest that visual motion sensitivity reflects inadequate filtering of peripheral motion or abnormal visual processing in the magnocellular pathway or cortical motion regions and that binasal occlusion works by reducing peripheral motion input.⁴⁰ If this were the explanation, though, monocular occlusion should also work, but there are no data on that.

Three group studies support this treatment. One did not report on clinical symptoms or signs; the other 2 studies had 10 and 15 subjects with TBI with visual motion sensitivity.⁴¹ Although their outcomes focused on visually evoked potential changes, they also reported that their subjects had reduced nausea and disorientation, less illusory motion, better fixation, walking, and grasping. Neither study had a placebo intervention, a control group, or an objective measure to validate the symptomatic report.

Prisms

There is 1 report on yoked prisms (yoked means that the same prism in the same direction is applied to both eyes) for midline shift syndrome in 36 patients with a variety of disorders, including 10 who had experienced a concussion.¹⁹ Gait was analyzed for sway on a pressure-sensitive mat, with and without yoked prisms. Prisms induced a shift in balance, but there were no data to show that this would reduce falls in daily life. The study was not masked, and there was no comparison intervention. The data for the patients with mTBI were not analyzed separately.

The use of a monocular vertical prism as treatment for VHS is not supported. Existing studies have circular logic: the diagnosis of the syndrome was based on subjective report of benefit from the prism during an assessment, and these subjects were then evaluated later with a questionnaire about benefit when wearing the prism.²¹ None reported a placebo intervention or a control group.

Light Filters

Light sensitivity is also a common posttraumatic symptom, and it seems logical that wearing filtered lenses would improve this symptom. However, there is no firm evidence on whether specific types of filters are better and whether symptoms and function improve with filters. One study of 7 subjects with mTBI with photophobia found that some filters improved contrast sensitivity and reading times, with uncontrolled subjective reports of reduced light intolerance and headaches at home after the study.⁴² A study of 12 subjects with mTBI with photophobia compared a few minutes of use of red, blue, neutral density, or individualized filters, and a no lens condition, in random order. There was little effect of lenses on reading, and effects on light intolerance were not mentioned. Paradoxically, a retrospective review of 62 cases of mTBI found that improved light intolerance after a year was less likely in patients who wore tinted lenses.⁴² Although patients who improved may have stopped wearing such lenses, the authors expressed concern that lens use might inhibit long-term adaptation that reduces light intolerance. There is evidence that FL-41 tinted lenses can improve photophobia in patients with migraine headache, but prospective data on the effect in patients with TBI-associated photosensitivity do not exist.⁴³ A prospective study would be needed to address this concern, with controls to avoid a confounding effect from the anticipated benefit from concurrent headache treatment.

Conclusions Regarding Treatment

All forms of vision therapy for mTBI must be regarded as having unproven benefit. For ocular motor training, there are reasonable grounds to believe that vergence or accommodative training may be of symptomatic benefit, but the rationale and evidence for training saccades and pursuit is poor. Similarly, there is no scientific rationale or evidence to support binasal occlusion, yoked prisms, or monocular vertical prisms. It is plausible that light filters may have some symptomatic benefit for light intolerance, but more evidence is required, particularly regarding the concern that filters may impede long-term recovery. Nonetheless, an open mind regarding potential benefit should be maintained until properly performed, scientifically rigorous studies are concluded and interpreted through the peer review process.

Outcome Measures

Interpretation of the results of therapeutic interventions as noted above has been hampered by subjectivity in the assessment of pre- and post-treatment patients, outcomes being determined based on improved performance on the tests used to define the apparent deficit and for which specific therapies aimed at test-taking tasks are designed, and other methodological factors. At a minimum, outcome measures should be designed such that they are administered by personnel masked to the mTBI diagnosis and therapeutic interventions that have been performed. In addition, outcomes must be measured by validated (if only internally) tools that also are designed to minimize learning effects that could indicate a false response to a rehabilitative strategy.

Most importantly, there are few randomized trials in which subjects with mTBI have been assigned to sham therapies for comparison.³⁸ The lack of a control group in published studies is particularly concerning, given the fact that symptoms and signs of mTBI will improve over time in most individuals.

Nevertheless, mTBI symptoms do persist in a minority of patients. It has been reported that at least one-third of patients with mTBI with visual concerns may not present for evaluation until 1 or 2 years postinjury.¹⁵ Some particular subsets of concern include those with multiple mTBIs, who may be at increased risk for persistent visual symptoms, and those with dual sensory impairments (auditory/vestibular and visual),¹⁴ who also may not recover as well. Treatments may be more important for these mTBI subsets, but systematic studies have yet to be conducted.

Recommendations

Prospective studies of rehabilitative strategies must be performed not only by experienced clinicians and researchers who are familiar with this problem but also by the broader neuro-ophthalmic and neurologic communities. Minimizing conflicts of interest in the design and interpretation of these treatment techniques must be a mainstay of any research strategy. We acknowledge that we may have overlooked some studies in our search strategy; however, even recent research publications appear to be case series of interventions without a control group for comparison.⁴⁴ We recommend the following steps:

1. Collaborative research efforts among neurologic TBI experts, optometrists, and ophthalmologists should be pursued.

2. Such research efforts should be funded by external granting agencies as much as possible. Prospective scientific review to ensure sound study design with minimization of bias and use of appropriate controls and outcome measures must be employed.

3. Results of these studies should be published in the peer-reviewed scientific literature.

4. While such studies are being conducted, a systematic review (through the Cochrane Eyes and Vision Group methodology or similar) may help to clarify the state of evidence regarding current therapeutic methods and knowledge gaps.

Visual symptoms may occur in patients with mTBI, and most patients note spontaneous symptom resolution. Several visual rehabilitative strategies have been proposed to alleviate persistent symptoms and to aid in functional recovery. Symptoms have been categorized into named syndromes for which a pathophysiologic basis is not always evident, and reports of treatment efficacy generally lack randomization to sham treatment or masking of subjects or examiners to the intervention being given. Unanswered questions include the following: (1) Which patients with mTBI are most likely to develop persistent visual symptoms, and are there structural or functional biomarkers to identify them prospectively? (2) Are visual rehabilitative treatments more effective than standard mTBI care methods? and (3) Is there an optimal time for prescribing visual rehabilitation, and are outcomes improved if early intervention occurs? Efforts are underway to design study protocols that can provide further objective data on both the pathogenesis of visual dysfunction after mTBI and the utility of visual rehabilitation relative to standard of care treatment for mTBI. These data will allow for more targeted and effective treatments for our patients with mTBI who may report symptoms that last for months or years after injury.

Appendix. Authors

Study Funding

The authors report no targeted funding.

Disclosure

The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.