ارزیابی مقبولیت مخاطب برای اقدامات اولویت دار در لکنت زبان
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|33550||2012||12 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Communication Disorders, Volume 45, Issue 5, September–October 2012, Pages 378–389
Purpose To assess the feasibility of using one or more of four standard economic preference measures to assess health-related quality of life in stuttering, by assessing respondents’ views of the acceptability of those measures. Method and results A graphic positioning scale approach was used with 80 adults to assess four variables previously defined as reflecting the construct of respondent acceptability (difficulty of decision making, clarity of text, reasonableness for decision making, and comfort in decision making) for four types of preference measurement approaches (rating scale, standard gamble, time trade-off, and willingness to pay). A multivariate repeated measures analysis of variance (p < .001) and follow-up univariate repeated measures analyses of variance (all p < .01) were all significant, indicating that respondents perceived differences among the preference measurement methods on all four acceptability variables. Conclusion The rating scale was perceived as the easiest, clearest, most reasonable, and most comfortable tool, but it is not a measure of utility (an economic term for desirability or worth). If utility is the objective, such as for cost-utility analyses in stuttering, then the present results suggest the use of standard gamble (rather than time trade-off). These results also support the use of willingness to pay assessments for cost-benefit analyses in stuttering. These findings supplement results previously obtained for other chronic conditions. Learning outcomes: The reader will be able to: (1) describe how four standard economic preference measures [rating scale (RS), time trade-off (TTO), standard gamble (SG), and willingness to pay (WTP)] can be used in economic analyses; (2) describe how RS, TTO, SG and WTP can be measured; and (3) describe how respondents perceive the use of RS, TTO, SG and WTP in measuring changes in stuttering.
Stuttering is often described in terms of two groups or types of variables. The first is speech parameters, such as speech rate or observable stuttering events. The second is nonspeech parameters, such as emotions, social abilities, and educational or vocational performance. There are differences of opinion as to the relative importance of the two types of variables, and as to the presence or direction of any causal relationship between them (compare, for example, the assessment and management recommendations provided by Costello and Ingham (1984), Hill (1999), Manning (1999), and Onslow, Packman, and Harrison (2003)). Nevertheless, most authors agree that both types are relevant to some greater or lesser degree, and most assessment protocols address attitudinal or emotional variables in addition to speech variables (e.g., Guitar, 2005 and Shipley and McAfee, 2008). Treatment recommendations are then built on the combination of speech parameters and nonspeech parameters, with different elements weighted in ways that reflect the client's and the clinician's (or researcher's) views as to their relative importance or causal relationship. One problem facing clients, clinicians, and researchers in stuttering, however, is that none of the instruments that have been developed to measure nonspeech variables [e.g., Andrews and Cutler's (1974) S-24, or Brutten and colleagues’ Communication Attitude Test (Brutten and Dunham, 1989, De Nil and Brutten, 1990 and De Nil and Brutten, 1991)] meets more than approximately half of a set of very basic psychometric standards for use either with individual clients or with groups of research participants (see Franic & Bothe, 2008, for detailed descriptions and analyses). Evidence that they result in reliable and valid measures of a well-defined construct is not available for most of these instruments, nor is evidence that they can be administered or interpreted as intended by their developers. In fact, in some cases, clear evidence to the contrary is available [see Franic and Bothe's (2008)Appendix A and Appendix B; see also Ingham, 2012 and Ulliana and Ingham, 1984]. This lack of focused, dependable, and defensible measurement instruments related to the nonspeech parameters in stuttering has important implications for theory, assessment, and management for this disorder. In response to this problem, some researchers have recently begun to explore how to measure the constructs of quality of life and health-related quality of life in stuttering ( Bothe et al., 2010, Bramlett et al., 2006, Craig et al., 2009, Franic and Bothe, 2008, Franic et al., 2012, Koedoot et al., 2011 and Yaruss and Quesal, 2006). Of the two, health-related quality of life (HRQL) is more directly relevant to stuttering; it is a summary outcome variable designed to measure clients’ overall perceptions about their lives with respect to domains often described as physical, social, role, and emotional functioning ( McHorney & Tarlov, 1995). The similarities between these domains and the traditional speech and nonspeech categories used to measure stuttering provide clear links between HRQL methods and questions of longstanding interest and demonstrated clinical relevance in stuttering. In addition, HRQL measures have the advantage of having been extensively studied, well validated, and widely used and recommended in the health outcomes literature and in health-related policy development. HRQL measures are known to reliably and validly capture the combined results of multiple variables in a way that can be responsive to health-care interventions and that can serve as a common metric for describing and comparing the overall impact of clients’ abilities, disabilities, or limitations in functioning. In short, HRQL measurement methods have several advantages over existing nonspeech measures used in stuttering, and they could help to address several measurement and management problems facing clients, clinicians, and researchers in stuttering. 1.1. Selecting among HRQL measurement methods Despite these advantages, several complexities arise as attempts are made to measure HRQL or related variables (known as preferences or economic utility, and referring generally to the value placed by a person on a good, service, state, or condition) for stuttering. Among the first is that HRQL can be measured in many ways, including the rating scale (RS), standard gamble (SG), time trade-off (TTO), and willingness to pay (WTP) methods (all are described in Appendix A for interested readers). All four have their advocates, and all four are strongly recommended for use in certain circumstances, but there are multiple competing and even contradictory reasons to select any one or another of these methods. Their underlying theoretical assumptions differ, for example, as do the specific purposes of the analyses conducted. Cost and time for administration also vary across the methods ( Gold et al., 1996 and Torrance, 1987). The RS method is generally accepted to be cognitively less challenging, faster, easier for respondents to complete, and less demanding for interviewers to administer, but it results in scores that cannot be used in certain ways (i.e., it does not produce utilities; see Gold et al., 1996 and Torrance et al., 2001). Similarly, the TTO method was developed in part because the SG method was viewed as too difficult for some respondents to manage, but whether TTO procedures are actually easier remains open to some debate ( Gold et al., 1996). Patrick, Starks, and Cain (1994) reported contradictory findings: Their respondents found the SG task easier to understand, less frustrating, and more accurately reflecting their preferences than TTO. Interestingly, the developer of the TTO method also favors the use of SG over TTO in utility elicitation ( Torrance et al., 2001). In addition to these many and conflicting general considerations, it is also important to consider both the actual and the perceived goodness of fit between the methods and the specific health conditions being assessed, or what has been referred to as the feasibility of any particular method for use in any particular application. Some investigators have addressed this question through indirect feasibility analyses, such as assessing response rates, identifying the numbers of respondents whose answers seem to reflect a misunderstanding of the questions, or identifying other features of obtained data that suggest that respondents were confused by the tasks ( Ryan, Watson, & Amaya-Amaya, 2003). In the case of stuttering, Bramlett et al. (2006) began this process by comparing the RS, SG, and TTO methods with a group of 76 respondents. Their results showed that all three methods could be completed by all respondents, and that all three methods appeared to result in meaningful data that met some expected patterns. Franic et al. (2012) continued the work by assessing WTP and a composite measure known as quality-adjusted life years. In the case of the WTP data, two of 80 original respondents provided extreme responses that were formally determined to be statistical outliers, but otherwise the results of this study again suggested that WTP questions could be feasible and useful in stuttering. Taken together, Bramlett et al.’s (2006) and Franic et al.’s (2012) studies suggest, in short, that all four HRQL measurement methods might be feasible for use in stuttering. What remains to be determined, and the focus of the present study, is whether respondents perceive differences among these four methods (RS, SG, TTO, and WTP questions) when the methods are directly compared. This is the second way of assessing the feasibility of alternative measures: asking respondents to evaluate the different methods, on the assumption that, even if they understand and are willing to answer all questions, they might find one approach more or less reasonable, or more or less conducive to decision making (Gold et al., 1996), for any number of reasons. One possibility that must be considered in extending to a new health area, for example, is that respondents’ views of the questions themselves might vary depending on whether they are being asked to consider such disparate conditions as chemotherapy-induced nausea and vomiting, chronic joint pain, a progressive or potentially fatal disease, or stuttering (Franic, Pathak, & Gafni, 2003). Flowers, Garber, Bergen, and Lenert (1997) provided a model for this type of research with their study of four characteristics of WTP questions in the context of possible necessary treatment for Gaucher's disease. The characteristics assessed by Flowers et al. were four domains that defined the construct of the acceptability of a preference measurement method: the difficulty of the questions, the clarity of the text, the reasonableness of the questions, and respondents’ overall comfort in using the approach for decision making. Flowers et al.’s results confirmed that most respondents found their WTP questions to be very clear, and found the WTP approach to be reasonable or very reasonable for decision making. However, 38.5% of respondents found WTP to be somewhat uncomfortable or very uncomfortable to use as a tool for decision-making in the contexts used by Flowers et al., and 45.2% found most or all decisions difficult to make in the WTP paradigm, data that might suggest the need to either change the specific format of the questions or use another approach. Ryan et al. (2003), similarly, discussed the need to consider the “acceptability to respondents” (p. 719) of different types of questions. The purpose of the present study, therefore, was to add to the growing literature about assessing quality of life in stuttering by evaluating respondent attitudes toward four standard economic preference methods: rating scales (RS), time trade-off (TTO), standard gamble (SG), and willingness to pay (WTP). All four of these methods had previously been demonstrated to be feasible for use in stuttering, but respondents’ views of the methods themselves, or of the methods’ acceptability for decision-making in stuttering, had never been assessed. The objective of this study, in other words, was to assess the value or appropriateness of the different tools that can be used to measure quality of life in stuttering. This information is a necessary part of any attempts to adopt standard HRQL measurement methods for use in stuttering, something that could serve as an important part of continuing efforts to measure and manage the speech and nonspeech elements of this disorder most appropriately.