ACQ Vol 13 no 2 2011

In addition, there are some common pitfalls that need to be avoided when measuring change over time. Of course, one can only truly evaluate whether a change has occurred if one measures performance on the same test, because of differences in test difficulty. Hence, at each assessment the same test (and the same items within a test) should be used to measure, for example, semantics, naming, repetition, reading. Moreover, should different items be completed on each occasion (e.g., a smaller number of items tested on the second assessment than the first), then the true extent of change can only be evaluated through comparing performance on those items that were assessed on both testing occasions. It is therefore important to choose, wherever possible, assessment tasks that are sensitive to a range of severities. Alternatively, if it is not possible to use the same test and the same items on each occasion, at least one testing session should use both a harder and an easier test, prior to no longer using the harder test. One final point to note is that some relatively easy tests (e.g., PALPA word-picture matching, subtest 47) may appear to give a picture of no change in performance when an individual is scoring close to ceiling (i.e., close to the maximum possible on the test). However, despite the lack of change in the score, a decline in the underlying skill may have occurred (e.g., semantic processing in PALPA 47). This is because on some assessments, even if language skills are impaired, it is still possible to score perfectly. It is important to also consider issues of reliability and sensitivity in relation to assessment beyond the impairment level. For example, it is only when sufficient detail has been documented regarding the nature of participation and communicative activities undertaken, that change can accurately be tracked in that participation. To return to our example of using the telephone: that an individual has difficulty in using the telephone may be too general a measure to be able to detect any change. However, if more detailed questioning had provided greater detail as to the extent and nature of the difficulty, then it may have been possible to determine whether there had indeed been a change over time in the degree of difficulty in telephone use. Incorporating client ratings into the extent of difficulty or depth of feelings can also add sensitivity. Tracking change as a result of treatment In the treatment of non-progressive aphasia, it is important that any assessment of treatment outcomes must show (a) whether the ability being treated is changing over the course of treatment, and (b) whether the change is actually due to the treatment as opposed to some other factor that happens to influence the outcome (Nickels, 2002; Wilson, 1987). While these are clearly also relevant to treatment of progressive aphasia, the situation is complicated because of the expectation of decline without treatment. Although improvement above the level seen at initial assessment is one possible outcome if therapy is effective, it is not the only one. One might also see an outcome of no change (i.e., the client maintains his or her current level of ability), or a slowing of deterioration that allows the individual to continue in desired activities for a longer period of time than would have been possible without the intervention (Rapp & Glucroft, 2009). To know whether a treatment is effective, the ability being treated must also be compared with a “control” condition (perhaps items of similar difficulty that are not being treated, or another ability that is not expected

allow for in-depth discussion of concerns, priorities, and goals, but also to gain an understanding of the nature of the functional impact of language impairments (e.g., not just that there is a problem with using the telephone, but just how this problem manifests itself). As Simmons-Mackie and Damico (2001) point out, clinicians routinely do obtain information from clients and families regarding activities, social relationships, and feelings. However, critically, this information is rarely documented systematically, rendering it a far less potent source of information. In sum, we advocate hypothesis-driven assessment and systematically documented in-depth interviews as the most efficient way of identifying the current status of language impairment and function and of enabling truly collaborative treatment planning for people with progressive aphasia. To investigate change over time Unfortunately, it is central to the nature of the disorder that people with progressive language impairments will show a decline in their language functions over time. An important role for the speech pathologist is therefore to track this decline in order that the rate and nature of change can be used to plan for future communication strategies and adjust the focus of treatment. For example, if one skill (e.g., writing) is declining at a more rapid rate than another (e.g., spoken output), a communication strategy that relies on a rapidly declining skill would seem to be ill advised, but an impairment-directed treatment aimed at maintaining this skill may be important (depending on the personal priorities of the individual). The speech pathologist might also provide education about the need for strategies that proactively address the potential consequences of further decline in that ability (e.g., how will the client handle future management of personal, health, legal, and financial matters which currently rely on writing?). In addition, as with any client group, it is vital to evaluate the nature of the change brought about through treatment. Choosing assessments for tracking change In order that assessments can accurately document change (both spontaneous change and change brought about through treatment), they must be reliable (i.e., if there is no change they will show the same level of performance on retest) and sensitive (i.e., if there is change this will be reflected in the scores) (Howard & Hatfield, 1987). Both of these criteria require a relatively large sample of behaviour (e.g., many items to name aloud, many observations of attempted conversational repair; multiple ratings of similar communicative behaviours – ordering a pizza, ordering a taxi, ordering a curry). The adequacy of the sample is another factor to consider when evaluating the suitability of a general language battery for use in assessment of progressive aphasia. While over the battery as a whole there are usually large numbers of stimuli, within each subtest the numbers of items are generally (and necessarily) small. Thus change in a specific skill as a result of a greater rate of decline, or as a result of improvement, if therapy has been targeted at that area, may neither be represented in the overall score nor be significant on its own because of the small number of stimuli. Hence, our recommendation would be, wherever possible, to use specialised assessments that are aimed at each level of processing of interest and that contain enough stimuli to enable sensitivity and reliability in the measurement of change as and when it occurs.

75

ACQ Volume 13, Number 2 2011

www.speechpathologyaustralia.org.au