Three Suggestions for Translating Apraxia of Speech Treatment Research in Clinical Practice

Published treatment research typically represents the effects of therapy applied under ideal conditions, which may vary considerably from the conditions under which therapy typically occurs in clinical practice. For example, your patient may only be able to attend therapy one time per week, but published findings from Sound Production Treatment (SPT) reflect treatment administered at least three times a week. As a clinician, you may wonder if SPT should even be considered for your patient. The following steps address issues surrounding the use of treatment research in your own practice, with an emphasis on the treatment of acquired apraxia of speech (AOS).

1.   Treatment Intensity: Know What is Typical and Adjust Accordingly

Researchers often plan their studies to maximize the effects of the tested treatment. In doing so, they may administer treatment in a manner that is unfeasible in clinical practice (e.g., length of sessions, number of sessions). Even if the researcher has considered clinical applicability, practicing clinicians will likely have patients who cannot meet the schedule of treatment reported in the research literature. If the patient cannot receive as much, or as frequent treatment as has been reported, the clinician should not eliminate a particular therapy as a treatment option.

In AOS research, intensity of treatment has received very limited study. We know little about the effects of factors such as the number of therapy sessions or appointment spacing on treatment outcomes. For example, with SPT and other AOS therapies, we do not know the effects of administering treatment with different intensities.

It would be valuable to know the number and length of treatment sessions typically needed to achieve the improvements reported by researchers. You could then inform your patient about expected outcomes under “ideal” conditions, explaining that reduced intensity of treatment (fewer sessions per week or less than a whole session dedicated to that particular treatment) may increase the time required to change behavior or may result in lack of change.

Most of the AOS treatment literature is based on single-case experimental designs. These provide information about individual research participants. The clinician can then compare their patient’s characteristics to those of the research participants. The clinician can compare his/her patient’s characteristics to those of reported research participants to obtain an idea of expected response patterns such as how many sessions were needed to see gains and how much change was achieved after a given period of time.

2.   Measurement of Change: Treatment Data ≠ Probe Data

Outcome measures from single-case experimental design studies are usually derived from probes repeatedly administered prior to, during, and following treatment phases.^1,2 Probes are brief assessments of the behavior under study and are conducted during a time when treatment is not taking place (often just prior to the daily treatment session).

Repeated probing is a necessary experimental design component. However, recurring probing is not typically practical for clinicians due to the time involvement. Instead, clinicians usually track progress during the treatment session itself. For example, with SPT, clinicians may tally how often the target item was produced correctly on the first step of the treatment hierarchy. Unfortunately, performance during treatment does not always accurately reflect change as measured by probes. Specifically, patients often perform better during the treatment session than during probes when there is no feedback or instruction being provided.

In order to ensure that the patient made gains that will be maintained to the degree reflected in the treatment literature, clinicians should assess performance during probes and treatment. This does not mean that clinicians need to probe repeatedly. When it appears that a patient is performing at a consistently high level of accuracy during treatment, the clinician should then conduct a “non-treatment” probe – ideally completed at the start of the next session to measure short-term maintenance. The probe may confirm the patient’s high levels of performance or may suggest that more treatment is necessary.

3.   Generalization: Plan Ahead and Measure – Do Not Assume

With SPT, substantial generalization to untrained manifestations of treated items is well documented. For example, if the cluster “st” is trained in 8-10 treatment items such as “star”, “step”, “stain”, “steal”, “stone”, “stock”, “stew”, “stove”, then generalization to untrained items such as “stir” and  “stick” is likely to occur. The caveat with this type of generalization (called response generalization) is that treatment items should represent a variety of vowel and phonetic contexts.

Generalized production of treated items to different elicitation contexts has not been thoroughly studied with SPT. This type of generalization is called stimulus generalization. Positive stimulus generalization effects are found with some SPT participants, but there is not currently enough data to predict which patients will achieve generalization. Little is known about stimulus generalization effects of other AOS treatments.

Prior to treatment, clinicians should select a few items for measuring response generalization. These items should be similar to treatment items (as in the example above). Clinicians should also think about measuring production of treatment items in different elicitation contexts prior to treatment. For example, if therapy is going to be focused on single word production, then measuring production of treatment words in phrases may be used to measure stimulus generalization.

The take-home generalization lessons are:

• Do not assume that generalization occurs automatically
• Create reference points prior to treatment for comparison to post-treatment measures

Translating Research into Practice

There is a lack of information about the effects of AOS treatments, such as SPT, when implemented in clinical settings. Our knowledge about the outcomes of AOS treatments is primarily derived from research treatment administered under ideal conditions, which may not correspond to clinical practice. Clinicians should be aware of the potential mismatch between clinical and research conditions and can take a few simple steps to help in translating research to practice.