It is no secret that diagnosis is part of clinical work for most speech-language pathologists. Diagnosis is valued most for its contribution to planning treatment, which is usually at the core of our clinical practices. Important as this contribution is, diagnosis can play another critical role—that of helping to localize and diagnose neurologic disease.
Despite astounding advances in neuroimaging and other diagnostic tests, many neurologic diseases remain difficult to diagnose. Neurodiagnostic tests may be normal or of uncertain meaning, and the medical history can be unrevealing or misleading. Sometimes, a motor speech disorder (MSD)—dysarthria or apraxia of speech—is the predominant sign of such illnesses.
The fact that MSDs (and other neurologic communication disorders) often can be linked to disease location, and sometimes to specific diseases, even when localization and diagnosis have not yet otherwise been established, is often under-recognized. More than 20 years ago, Aronson (1987) argued that this fact was "one of the best-kept secrets of our time," a secret ignored or underemphasized by both medicine and speech-language pathology. Now, in the next century, we should ask if this fact is still a secret. If it isn't, are we using it effectively to enhance clinical care?
Cases in Point
The following cases—which are slightly modified but unembellished—illustrate the contributions MSD diagnosis can make to medical localization and diagnosis. Accounts like these are encountered most frequently in large medical centers, but they undoubtedly present in a wide variety of settings. (Many of the issues addressed below, including some of the case examples and referral quotes, have been adapted from more detailed discussions in other publications.
A 74-year-old man had a 10-month history of progressive speech difficulty. Previous workups for stroke and myasthenia gravis were negative. Neurologic examination revealed only dysarthria. The neurologist requested speech consultation, stating "we will then decide what additional studies are appropriate."
Speech evaluation revealed bilateral face and tongue weakness, but no fasciculations; strained-hoarse-groaning voice quality; monopitch and monoloudness; hypernasality and weak pressure consonants; and slow rate. Vowel prolongation was strained-harsh, with vocal flutter. Speech alternating motion rates (e.g., rapid repetition of /p^p^p^/) were slow but regular.
The clinician reported that the patient had a "moderate spastic dysarthria that is strongly suggestive of bilateral upper motor neuron dysfunction affecting the bulbar speech muscles." The possibility of lower motor neuron involvement was also raised. Based on the speech diagnosis, the neurologist ordered electromyography (EMG) and nerve conduction studies that led to a diagnosis of amyotrophic lateral sclerosis (ALS). The patient was subsequently seen for management through an interdisciplinary ALS team.
A 63-year-old man had a one-month history of speech difficulty that the referring internist attributed to stroke. Neurologic examination found only dysarthria and concluded that the likely cause was stroke because of multiple vascular risk factors (e.g., hypertension, high cholesterol). Additional tests were ordered, including speech consultation.
During speech evaluation, the patient's speech was initially normal, but then became hypernasal with lingual consonant imprecision. Following 15–30 seconds of rest, it returned to normal. This pattern of deterioration, with improvement following rest, was replicated several times.
The clinician reported that the patient had a "flaccid dysarthria with rapid onset of speech difficulty associated with continuous speaking, followed by significant recovery after a short period of rest, strongly suspicious for myasthenia gravis." The neurologist then modified the workup. Subsequent EMG and acetylcholine receptor antibody testing were diagnostic of myasthenia gravis. The patient was placed on Mestinon and, later, prednisone. His dysarthria subsequently resolved, except during prolonged physical activity.
A 69-year-old woman had a two-year history of progressive speech difficulty. Neurologic examination was normal, with the exception of speech and slow eye movements. Subsequent EEG, MRI, laboratory tests, and neuropsychological assessment were unremarkable.
Speech-language assessment revealed normal language, with the exception of infrequent self-corrected function word and semantic errors. Speech was characterized by reduced rate, occasional segmentation of syllables, distorted substitutions, and articulatory groping. The patient had a nonverbal oral apraxia.
The clinician reported "a mild apraxia of speech (AOS) that is more pronounced than a subtle aphasia," and also stated that "progressive AOS, when it is a prominent manifestation of degenerative neurologic disease, and worse than any accompanying aphasia, is frequently associated with tauopathies such as progressive supranuclear palsy or corticobasal degeneration." (Tauopathies are a group of non-Alzheimer's degenerative diseases characterized by abnormal accumulation of tau, a protein, in neurons or glial cells.)
Over the next five years, she became nearly mute and reliant on augmentative and alternative communication. She comprehended language reasonably well and enjoyed hearing from people by phone. An autopsy revealed pathology consistent with progressive supranuclear palsy.
A 34-year-old man was referred for consideration of deep-brain stimulation surgery for generalized dystonia. The neurosurgeon felt surgery would be appropriate, pending completion of a comprehensive baseline workup, which included speech-language assessment.
During speech evaluation, the patient complained of consistently abnormal voice and episodic inability to speak. The oral mechanism was normal, except that the thyrohyoid space narrowed abnormally during phonation. Voice was strained-harsh-raspy, pitch and loudness variability were reduced, and speech rate was slow. His dysphonia worsened during reading; during counting it became a strained, aphonic whisper. The clinician felt the abnormal pattern was "not consistent with dystonia or any other recognizable dysarthria type." Therapy was undertaken and within 10 minutes his voice quality nearly normalized and remained so during the rest of the examination and subsequent discussion.
The clinician concluded, "the patient's speech disorder most likely reflects musculoskeletal tension and not any underlying dystonia." This conclusion was discussed with the referring neurologist, and led to questions about the diagnosis of dystonia in the limbs. Subsequent EMG and additional clinical examination led to a conclusion that the "generalized dystonia" was psychogenic. Surgery was canceled.
Important lessons from these cases include:
- MSDs can be the only, the initial, or the predominant sign of neurologic disease.
- MSD diagnoses have implications for localization, can narrow or expand diagnostic possibilities, and can provide strong or confirmatory evidence for neurologic diagnosis. In some medical practices there is an expectation from physicians that speech diagnosis can make these contributions.
- Recognizing speech abnormalities that are incompatible with neurologic disease or that run counter to neurologic diagnosis can modify medical diagnostic efforts and help avoid inappropriate medical/surgical treatments.
- Many people with MSDs and undiagnosed neurologic disease have an MSD long before they have a need for speech-language treatment. However, SLPs can actively contribute to medical diagnostic efforts by recognizing the meaning of a patient's speech problem as it relates to localization and possible cause.
How Did We Come by This Skill?
The underpinnings for diagnostic applications reflect many decades of clinical-descriptive and related research efforts. In the late 1960s, Darley, Aronson, and Brown provided a detailed, perceptually based description of the dysarthrias and identified several distinct types based on groupings of abnormal speech characteristics—each type reflecting localization within the neuromotor system and presumed pathophysiology (see Table 1 [PDF]). Many subsequent studies of MSDs (including AOS) as they occur in numerous neurologic diseases have further refined our clinical understanding of MSDs and their neuroanatomic, acoustic, and physiologic correlates. As a result, SLPs now have a better sense of the MSD types most likely to be present in a variety of common and uncommon neurologic conditions such as stroke, ALS, traumatic brain injury, Parkinson's disease, cerebral palsy, corticobasal degeneration, progressive supranuclear palsy, spasmodic dysphonia, Freidreich's ataxia, paraneoplastic cerebellar degeneration, and Wilson's disease. An MSD can be a prominent sign in several of these conditions. This research base has made us better diagnosticians, practitioners, and clinical researchers.
Clinical research continues to improve our diagnostic capabilities. For example, it recently has been documented that AOS can be the predominant communication problem in neurodegenerative disease (Duffy, 2006) and that it seems most often tied to diseases with prominent motor impairments such as progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), or ALS. Thus, when an isolated or predominant AOS of neurodegenerative origin is identified, the SLP is in a position to suggest that CBD, PSP, and ALS should be among the neurologist's diagnostic considerations.
These and similar observations may even have implications for predicting autopsy findings. For example, a recent study of 17 patients with neurodegenerative AOS and/or aphasia found that AOS of equal or greater severity than any accompanying aphasia predicted PSP or CBD pathology at autopsy in 90% of the patients, and biochemistry consistent with tauopathy in 100% of the patients (Josephs et al., 2006). PSP or CBD, and thus tauopathy, were present when AOS was not present, but only in 17%.
If replicated, these findings suggest that a prominent neurodegenerative AOS tends to predict not only clinical neurologic diagnosis, but also post-mortem diagnosis, something that could be crucial to early identification when effective medical treatments for tauopathies become available. This finding is a strong argument—beyond those that can be made from a speech-language treatment perspective—for explicitly distinguishing AOS from aphasia.
Lend Me Your Ears—They're Valuable!
MSDs are recognized primarily by how they sound, reflecting abnormalities at one or more of the respiratory-phonatory, resonatory, and articulatory levels of the speech system. For accurate diagnosis, clinicians infer whether what they hear reflects weakness, spasticity, lack of coordination, rigidity or reduced range of motion, involuntary movements, difficulty with "higher-level" motor planning or programming, or combinations of those deficits. The history, examination of oral and breathing mechanisms at rest and during nonspeech tasks, and acoustic and physiological measurements can all enrich the evaluation and provide supportive diagnostic clues. But the conclusions drawn from the auditory-perceptual speech evaluation are the diagnostic gold standard.
In this age of sophisticated instrumentation and a health care system that rewards the use of technology and under-recognizes thinking and judgment, relying on the ears for diagnosis is disconcerting to many. Like any assessment strategy, auditory-perceptual judgments have shortcomings—but they are essential to clinical diagnosis. Those uncomfortable with this method of diagnosis might be reassured by the fact that auditory-perceptual analysis is also an important diagnostic tool for numerous medical and non-medical professions. It stands to reason that well-trained ears should be of at least equal importance to SLPs.
Improving Clinical Practice
How can we do better? Improving the art and science of MSD diagnosis requires ongoing and improved research, enhanced clinical training, and better communication with our medical colleagues in a manner consistent with our self-perceived clinical abilities.
The following are examples of efforts that might be made in each of these areas:
- Research. Clinical treatment studies and acoustic, physiologic, and neuroimaging studies of MSDs are most meaningful when they take care to describe the perceptual features of the MSDs that are the subject of study. This inclusion too often is not the case. The unintended consequence of this omission is that the speech abnormalities under study are obscured to a degree that limits our ability to understand and generalize findings in clinically relevant ways.
- Clinical training. Training and experience, added to a working knowledge of relevant neuroanatomy and neurophysiology, underpin the ability to accurately and reliably distinguish among the MSDs. Developing this skill undoubtedly requires guided exposure to numerous exemplars of the deviant features of MSDs and how they combine to represent specific MSD types. Such perceptual training seems uncommon in many training programs. It needs to be provided.
- Communication and self-concept. If SLPs want to be recognized for their ability to contribute to medical localization and diagnosis, they must embrace this dictum: no evaluation report should end without a diagnosis and statements about its implications for localization and compatibility with neurologic diagnoses under consideration. All too often these statements are missing. Providing them as a matter of routine requires a modification of one's professional self-concept from "clinician" to "diagnostician and clinician." It requires replacing concern that such contributions will be viewed as unwanted incursions into medical terrain with optimism that they will be viewed as welcome contributions to patient care.
Our Role in Diagnosis
Diagnosis of MSDs can make important contributions to the localization and diagnosis of neurologic disease. This fact is not widely recognized by medicine, nor consistently made evident by SLPs. To "out" this secret we need to improve descriptions of abnormal speech characteristics in MSD research, improve clinical training, and consistently provide explicit statements of MSD diagnoses and their meaning in clinical reports and communication with medical colleagues.
These changes require a professional culture and self-image shift to value our clinical roles not only as therapists but also as diagnosticians who have contributions to make in medical diagnosis.