Patient Education and Counseling 78 (2010) 177–183
Contents lists available at ScienceDirect
Patient Education and Counseling journal homepage: www.elsevier.com/locate/pateducou
Medical Decision Making
Developing a dyadic OPTION scale to measure perceptions of shared decision making Emma Melbourne, Kate Sinclair, Marie-Anne Durand, France Le´gare´, Glyn Elwyn * Department of Primary Care and Public Health, Cardiff University, Cardiff CF14 4XN, United Kingdom
A R T I C L E I N F O
A B S T R A C T
Article history: Received 8 January 2009 Received in revised form 2 July 2009 Accepted 7 July 2009
Objective: Our aim was to develop a measurement which enables research into the interdependent nature of clinical encounters. The prime objective was to develop an instrument capable of assessing the extent to which patients have been involved in (shared) decision making from two viewpoints—that of the patient and the clinician. Methods: To develop an initial ‘dyadic OPTION’ instrument, the twelve original third-person items were drafted in passive, first person plural forms. Using this version initially, three rounds of cognitive debriefing interviews were held. These were audio-recorded and analysed at the end of each round and the results used to revise the dyadic OPTION scale. Results: It was possible to modify the observer OPTION instrument into an instrument for completion by both clinicians and patients after a dyadic interaction. Cognitive debriefing revealed five areas of interpretative difficulty. Each item of the observer OPTION scale underwent modification in order to develop a dyadic version of the scale. Conclusions: The dyadic OPTION scale is acceptable and comprehensible by both clinicians and public respondents. Cognitive debriefing adapted and refined an existing scale and provided confidence that the core constructs of the scale (perceived involvement in decisions making) were understood. Practice implications: Further validation of the dyadic OPTION scale is required prior to its use in research settings. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Shared decision making Involvement Decision making Measurement Dyadic analysis
1. Introduction Although shared decision making in clinical encounters is clearly the result of an interaction between two individuals, attempts to measure this phenomenon have, so far, been limited to single viewpoints, most often from the viewpoint of an observer assessing a recording of the dialogue [1,2]. Whilst such assessments have provided valuable information to researchers in the field, they are inevitably limited in that they are restricted to what is audible or observable. They take no account, nor can they, assess the perceptions of those who are involved in the process of understanding the nature of decisions, negotiating their role in the decision process, and taking action to deliberate and decide. Yet there is evidence that patients differ in their evaluations of such encounters, their perception of involvement is different to the perception of clinicians [3]. We have little information as to
* Corresponding author at: Department of Primary Care and Public Health, School of Medicine, Cardiff University, Neuadd Meirionnydd, Heath Park, Cardiff CF14 4XN, United Kingdom. E-mail address:
[email protected] (G. Elwyn). 0738-3991/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pec.2009.07.009
whether these ‘internal’ perceptions correlate with those of external observers, or indeed as to how this interdependence in dyadic interaction affects the outcome of the encounter [4–6]. There is a need therefore to develop a measure which attempts to address this research gap and which could provide more information about the perceptions of those engaged in these complex, negotiated encounters. For consultations that require participation in decision making, and in particular where attempts are made to share decisions, the interactions will require both parties to address the issues of decisional equipoise, compare the features of options and achieve consensus about the best actions: in short, to achieve ‘shared decision making’ [7]. Although the term ‘dyadic analysis’ is relatively novel in healthcare settings, there is a growing interest in the concept of interdependence in health communication research [8]. Statistical methods used to analyse data gathered from health care encounters, including those focused on shared decision making, typically assume independent actors. However, in contrast, dyadic analysis methods assume an interaction effect between the individuals involved, and, rather than skate over the issue of mutual influence, accept that non-independence should be investigated, and in some occasions, may be the main outcome
178
E. Melbourne et al. / Patient Education and Counseling 78 (2010) 177–183
of interest [9]. To facilitate these investigations, new or adapted measurement instruments are needed in order to allow assessments from dual, or more, perspectives. It is noteworthy how few instruments exist that are capable of providing such data in health care settings [10]. However, instruments do exist which have been designed to measure shared decision making in clinical practice, most often from a third observer point of view [1,2,11]. Two of the most widely used have been the Braddock scale, representing six elements of informed decision making [12] and the observer OPTION scale, developed to measure the extent to which clinicians involve patients in decision making [13,14]. Both instruments require an independent, trained observer to rate achievement against defined competencies in audio or video-taped consultations. Although these tools have provided valuable data about decision making processes in clinical encounters, they are only able to represent external assessments: the perceptions of participants engaged in the interactions remain unavailable [4]. COMRADE [15] and the Perceived Involvement in Care (PICS) [16], are examples of instruments which aim to measure patients’ perception of involvement in decision making but neither were designed so that they could assess interdependence between participants in the encounter. We conclude therefore that a gap exists and that in order to apply the concept of dyadic analysis to clinical encounters, particularly to those where we wish to study perceived involvement and interdependence in shared decision making encounters, we either need to develop new tools or adapt existing tools. We took the view that an instrument used to tackle this task should meet the following criteria: to have a sound developmental pathway, clarity about its measurement construct, and evidence of item development and feasibility testing [17]. We also propose that tools capable of providing data for dyadic analysis need to have items that are identical for all respondents. In other words, item phrasing should be identical when administered to the clinician or to the patient. If we asked clinicians and patients to respond to items that were slightly different, we would run the risk of giving rise to different interpretations. We also decided to base a new scale on an existing tool, namely observer OPTION, a scale that has been rigorously developed to assess the degree to which clinicians involve patients in shared decision making, has psychometric data to support its uni-dimensional nature [13,14]. These principles guided the adaptation of observer OPTION [14], into a ‘dyadic’ version’ and the steps taken are reported in this article. The aim of this study was to develop a dyadic version of the OPTION tool which is acceptable and understood by both patients and physicians and provides a platform for analysing interdependence in the consultation, specifically with regard to the measurement of shared decision making. To achieve this aim, we conducted three cycles of cognitive debriefing interviews.
respond to written or verbal materials [19] by allowing an analysis of the points at which respondents may be experiencing difficulties in interpreting or mis-interpreting. Identifying and modifying such items increases the likelihood that respondents will respond appropriately in future versions [19]. Three cycles of cognitive debriefing interviews with two different groups of participants were planned. The first cycle of interviews considered the initial revision. Changes and improvements were made to the evolving versions of the dyadic OPTION instrument at the end of each cycle. The proposed sample size was based on previous cognitive debriefing studies [19]. Nine participants were planned for each cycle of interviews, and so for each cycle we set out to recruit six members of the general public and three clinicians. Members of the general public would be recruited from public spaces, such as cafe´s and other locations. A sample matrix was designed to achieve a balance between genders and stratified by preferred readership of a range of daily newspapers. Twenty-six million people of working age in the UK have levels of literacy and numeracy below those expected of school leavers [20], a figure which represents a significant proportion of potential patient respondents. The Gunning Fog Index indicates the number of years of formal education required to read text. Tabloids scored less than 14 and broadsheets greater than 14 on the scale and participants from each group were recruited. Clinicians were recruited by snowball sampling, using clinical contacts and existing academic networks. It was hypothesised that physicians were more likely to interpret items more consistently than their patients due to their extensive common education and as such a representative conceptual equivalence could be attained with fewer participant physicians than patients. Thus three general practitioners and six public respondents were recruited for each round of interviewing. Standard consent procedures were designed and implemented. 2.2. Data collection
2. Methods
Interviews were conducted by a single researcher (KS), who was familiar with the observer OPTION instrument, using a semistructured interview schedule, probing for understanding, item phrasing and completing field notes. All interviews were audiotaped and transcribed. Participants were asked to read the proposed ‘dyadic OPTION’ scale, where items had been rephrased to allow completion by either patients or clinicians. To contextualise responses, individuals were asked to think about their ‘most recent visit to a clinician’ where a decision had been required. Clinicians were asked to respond as if they had completed a consultation where they had made efforts to share an index decision during the encounter. Participants’ initial impressions and interpretations of items were sought immediately after each had been read. Anticipated and spontaneous probes were used in order to elicit additional information. At the close of the interview, opportunities were given to respond in general about the instrument and about its anticipated use in practice.
2.1. Design and participants
2.3. Data analysis
Using the observer OPTION tool as a starting point, an initial modified twelve item version was created (by GE), adapted so that the items could be read and completed by both patient and clinician at the end of a consultation. Potential ambiguities and difficulties in comprehension were anticipated and an interview schedule was prepared with suitable probes. Cognitive debriefing has become an accepted method in questionnaire development [18]. The goal of cognitive debriefing is to facilitate a more uniform interpretation of items by specified target audiences, a process which increases measurement validity. The method helps instrument developers understand how target audiences process and
At the end of each interview cycle, the data were analysed and categorised. All items were read and discussed between GE and KS after each round of interviewing. A summary was generated of the most salient issues prior to revision of the instrument ahead of the next cycle of interviews. This data summary enabled the interviewer (KS) and the scale originator (GE) to assess how each item had been interpreted. After discussion, items where comprehension difficulties or mis-interpretations were encountered were revised for further assessment. At the point when a cycle of interviews failed to offer new information, the data collection process would be stopped.
E. Melbourne et al. / Patient Education and Counseling 78 (2010) 177–183
3. Results 3.1. Participants characteristics Between November 2007 and April 2008, 27 cognitive debriefing interviews were conducted, during each round of interviewing. Six members of the public and three clinicians with educational or research affiliations participated in each cycle of interviews. Table 1 shows the breakdown of participant characteristics. Half of the 18 public respondents were male; seven of the nine clinicians were male. In each round, three of the public respondents regularly read The Sun, The Daily Mail or the Mirror, newspapers scoring less than 14 on the Gunning Fog Index and three regularly read broadsheets such as The Times, The Independent or the Guardian, scoring greater than 14 on the Gunning Fog Index. 3.2. Data from cognitive interviews Analysis of the interviews revealed four areas of difficulty. These were 1. Semantic difficulty. Significant variation in interpretation of the items and elements of the vocabulary was encountered. Particular attention was given to interpretations of items which were inconsistent with the scale’s intended area of interest, i.e. degrees of involvement in decision making. 2. Grammatical difficulty. Some sentence structures were consistently mis-interpreted. The decision to use the passive voice in order to maximise item applicability to both patients and clinicians rendered some items ‘unnatural’ and their construction required further attention. 3. Construct mismatch. Particular attention was given to items where it was clear that participant interpretation differed from that of the developers’ attempt to assess the core construct of involvement in decision making processes. Suggestions for improvement were particularly helpful and enabled new item formulations which better achieved the aims of brevity, ease of understanding and construct accuracy. 4. Personal preferences. The pursuit of clarity, brevity and grammatical consistency were used as justification and guidance for how to prioritise suggested modifications, aiming to achieve minimal respondent burden whilst maximising comprehension.
3.3. Synthesis of findings 3.3.1. First interview cycle The first nine interviews resulted primarily in vocabulary modification. Anticipated probes were used to assess the precursor
Table 1 Participant characteristics. Participant characteristic
First cycle
Second cycle
Third cycle
GP Male Female
2 1
2 1
3 0
Public respondents Male Female
3 3
3 3
3 3
Newspaper readership Gunning Fog < 14 Gunning Fog > 14 Mean age Age range
3 3 49.42 19–72
3 3 42.83 18–74
3 3 35 20–58
179
dyadic OPTION scale. Verbs and nouns were both modified. For example, the word ‘checked’ was considered to be condescending and was omitted. The word ‘management’ was considered by both patients and physicians to be applicable to medical treatment and as such was felt to be more appropriate than ‘deal with’ when considering treatment issues. Probing of apparent construct mismatch resulted in three significant changes, such as the substitution of ‘considered’ for ‘respected’ in item 10. When considering item 9, one elderly public respondent indicated that whilst tacit opportunities to ask questions during consultations often existed, he was unlikely to ask any and thus agree with the item unless specifically prompted. The issue was subsequently probed and a number of respondents indicated that this was the case and so ‘or prompts’ was added to the item as a result of this interview cycle. Several items which had double-barrelled queries, items 2, 11 and 12, were modified. Items which caused grammatical difficulties for respondents, items, 3, 8 and 11 were completely overhauled using a conditional reactive probe of ‘can you think of a better way to say it’ to respondents who understood the item. The need for an introductory statement to provide respondents with a context for completing the instrument was signalled by many respondents and a short statement was added. 3.3.2. Second interview cycle A second version of dyadic OPTION was produced, and a second cycle of interviews led to further comments. Many of the modifications made after the first cycle were accepted, including the use of the term ‘respected’. One participant responded ‘‘I like that term because you can respect a patient’s decision without having to agree with it.’’ In general, the changes that had been made led to greater acceptance and ease of comprehension. Item 8 had been given particular attention in the first cycle because of its complexity, and on specific probing it was much better understood after modification. The dual clause version of item 11 that entered this interview cycle was better understood than the initial version however, a number of possible paths were still identified and it was deemed necessary to further overhaul this item. Finally, the addition of ‘possibility of changing’ to item 12 was found to cause the item to be interpreted as an option not a necessity, but ‘changing’ still provoked dissent and so underwent modification during the review. The focus of interviewing during this iteration shifted slightly as the most significant semantic and grammatical difficulties were remedied. Leads arising from anticipated probing were extended upon and general issues that might affect the employment of dyadic OPTION began to surface. One course of emergent probing administered early in the iteration suggested that there was no point in talking about the advantages and disadvantages of different treatment options without considering the possible outcomes, this point was subsequently investigated by spontaneous probing throughout the remainder of the iteration. It was concluded that a construct mismatch existed because this was implicit to physicians but had up to that point been obscure to developers and ‘possible outcomes’ was subsequently included. 3.3.3. Third interview cycle During this cycle, changes made in the second cycle were analysed in more detail and participants were asked to complete the third version of the dyadic OPTION scale. Suggested modifications to item phrasing were minimal during this cycle: it was clear that grammatical or semantic difficulties had been addressed and eliminated during the first two cycles. Two items however, required attention. Item 4 used the item phrasing: ‘Different options (including doing nothing) were discussed’. This phrasing was considered ambiguous. The following change was made: ‘Different options, including the possibility of doing nothing, were discussed’. Item 11 had been repeatedly modified during previous
180
Table 2 Item development. Observer OPTION June 2006
First version of ‘dyadic OPTION’
Reason to modify
Second version
Reason to modify
Third version
Reason to modify Final version
Item 1 The clinician draws attention to an identified problem as one that requires a decision making process
A problem for which a decision was needed was identified
Syntactical difficulty
A health problem was identified where it was made clear that a decision was needed
None needed
A health problem was identified where it was made clear that a decision was needed
None needed
Item 2 The clinician states that there is more than one way to deal with the identified problem
It was agreed that more than one way to deal with the problem existed
Construct mismatchdouble barreled Syntactical difficulties
More than one way to deal with the health problem was identified
More than one way to manage the health problem was described
None needed
Item 3 The clinician assesses the patient’s preferred approach to receiving information to assist decision making (e.g. discussion, reading printed material, assessing graphical data, using videotapes or other media) Item 4 A health problem for which a decision was needed was identified
Methods of presenting information to best inform the decision were discussed
Syntactical difficulties and Different types of construct mismatch information (e.g. leaflets, websites) to help make the decision were given.
Syntactical difficulties-capacity to answer depends upon the problem Syntactical difficulties
Different sources of information (e.g. leaflets, websites, contact with other health professionals) to help make the decision were offered
Syntactical difficulties
Different sources of information (e.g. leaflets, websites, contact with other people) to help make the decision were offered
Different options (including doing nothing) were explored
Syntactical difficulties
Different options (including doing nothing) were discussed
None needed
Different options (including doing nothing) were discussed
Construct mismatch
Item 5 The clinician explains the pros and cons of options to the patient (taking no action is an option)
The pros and cons of the Syntactical difficulties, The advantages and different options were explored inappropriate vocabulary disadvantages of different options were discussed
The advantages and disadvantages of possible outcomes of options were discussed
None needed
Item 6 The clinician explores the patient’s expectations (or ideas) about how the problem(s) are to be managed Item 7 The clinician explores the patient’s concerns (fears) about how the problem(s) are to be managed Item 8 The clinician checks that the patient has understood the information Item 9 The clinician offers the patient explicit opportunities to ask questions during the decision making process Item 10 The clinician elicits the patient’s preferred level of involvement in decision making Item 11 The clinician indicates the need for a decision making (or deferring) stage
Ideas or expectations about how to deal with the problem were addressed
Construct mismatch— incomplete construct. Syntactical difficulties None needed
Different options (including the possibility of doing nothing) were discussed The advantages and disadvantages of possible outcomes of options were discussed
Ideas or expectations about managing the health problem were discussed Concerns or worries about how to manage the health problem were discussed It was made sure that the information had been understood There were opportunities to ask questions
None needed
None needed
Item 12 The clinician indicates the need to review the decision (or deferment)
The preference to take part in the decision (or not) was respected Either a decision was made, or there was an agreement to postpone making the decision The possibility of coming back to the decision in the future was discussed
Worries about how to manage Syntactical difficulties, Concerns or worries about the problem were explored inappropriate vocabulary how to manage the health problem were discussed The understanding of information given to help make a decision was checked Opportunities to ask questions existed
Grammatical difficultiesoverhaul
The wish to take part in the decision was considered
Syntactical difficulties and construct mismatch
It was clear that there was a need to make/postpone a decision
Construct mismatch— double barrelled
It was clear that opportunities to check/change a decision existed
Construct mismatch— double barrelled
Syntactical difficulties and construct mismatch
It was made sure that the information had been understood There were opportunities or prompts to ask questions
None needed
None needed
Construct mismatch— unnecessarily complex The desire/preference to take Syntactical part in the decision (or not) difficulties was respected The decision or an agreement Construct mismatch—double to postpone the decision barrelled was made
The possibility of changing the decision in the future was discussed
Syntactical difficulties
None needed
None needed
None needed
None needed
Syntactical difficulties
Ideas or expectations about managing the health problem were discussed Concerns or worries about managing the health problem were discussed It was made sure that the information had been understood There were opportunities to ask questions
The preference to take part in the decision (or not) was respected Either a decision was made, or there was an agreement to postpone making the decision The possibility of coming back to the decision was discussed
E. Melbourne et al. / Patient Education and Counseling 78 (2010) 177–183
Syntactical difficulties, Ideas or expectations about inappropriate vocabulary managing the health problem were discussed
A health problem was identified where it was made clear that a decision was needed More than one way to manage the health problem was described
E. Melbourne et al. / Patient Education and Counseling 78 (2010) 177–183
181
Table 3 The dyadic OPTION instrument (December 2008). Most encounters about health problems lead to decisions of one sort or another. These questions ask about the idea of being or feeling involved in decisions, for example, having an opinion or deciding whether to take medication, and if so which one, or what to do next. Answer the questions from your point of view by putting a tick in one box for each question. Please answer every question. What was the main problem you talked about? Please describe it in a few words. . . 1. A health problem was identified, where it was made clear that a decision was needed. 2. More than one way to manage the health problem was described. 3. Different sources of information (e.g. leaflets, websites, contact with other people) to help make the decision were offered. 4. Different options (including the possibility of doing nothing) were discussed. 5. The advantages, disadvantages and possible outcomes of options were discussed. 6. Ideas or expectations about managing the health problem were discussed. 7. Concerns or worries about managing the health problem were discussed. 8. It was made sure that information had been understood. 9. There were opportunities to ask questions. 10. The preference to take part in the decision (or not) was respected. 11. During the consultation, a decision was made; or there was an agreement to postpone making the decision. 12. The possibility of coming back to the decision was discussed.
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
Strongly agree
Agree
Disagree
Strongly disagree
iterations due to grammatical difficulties and in this interview cycle it seemed that respondents were finding this item very simple to answer. Anticipated probing revealed that respondents were applying this item to any health problem they had previously consulted with as opposed to specifically those discussed in the reference consultation. As a result it was decided to explicitly ground the item in time by adding ‘during the consultation. . .’ The introductory comments underwent further modification. Several respondents noted that not all consultations contained discussions about specific ‘decisions’ and the introductory statement was modified accordingly. It was also necessary to introduce the concept of an ‘index’ problem, i.e. to indicate that although most consultations contain multiple decision making time points, it was sufficient for the purposes of the assessment to consider one decision making issue, where involvement had been maximal. Table 2 shows how items were modified during the three interview cycles. After the third interview cycle, few new insights were gained and ‘category saturation’ had been attained [21]. No further interviews were thus conducted. Table 3 shows the final version of the scale. 4. Discussion and conclusions 4.1. Discussion This study demonstrates that, although difficulties occurred, it was possible to modify the observer OPTION instrument to be an instrument capable of being completed by both clinicians and patients after a dyadic interaction. The cognitive debriefing
interviews revealed five areas of interpretative difficulty, namely: construct clarity, syntactical and grammatical problems, contextual positioning of the instrument and issues of personal preference. Interestingly, despite the apparent methodological variation between cognitive debriefing research groups, problems encountered are broadly similar. Five kinds of problems were identified in the consumer assessment of health providers and services survey in 2005 [18]. These included items with ambiguous terms that caused them to be consistently understood or (misunderstood), corresponding to the syntactical difficulties identified in this study; and questions that were not measuring the constructs that they were supposed to measure—corresponding to our construct mismatch. An advantage of the cognitive debriefing process, previously acknowledged is the identification of vague and unclear questions due to syntactical difficulties and construct mismatch [22,23]. The majority of data informing modifications due to semantic and grammatical difficulties was elicited during the first two rounds of cognitive debriefing, largely by the proactive administration of anticipated and spontaneous probes. Studies in the literature are vague about their preparation and use of probes. It is our opinion that the probes administered and resulting data should be viewed as one unit for analysis. Indeed a consistent feature of the literature by leading cognitive debriefing practitioners is the need for conscientious analysis of the probing techniques to allow further methodological development. It was to address this request and thus enable analysis of the question response patterns that we presented the probes we used and the data elicited as one unit, and urge teams taking on future cognitive debriefing to do the same.
182
E. Melbourne et al. / Patient Education and Counseling 78 (2010) 177–183
By using a series of cognitive debriefing interviews, the existing observer OPTION scale was adapted into a tool that we consider suitable and now available for research into the interdependent nature of clinical encounters. Data from a dyadic version of the OPTION scale is capable of collecting data about involvement in decision making from the dual perspectives of clinicians and patients and will enable the collection of data suited for dyadic analysis [9]. This is a new area of research in healthcare and one which requires specifically designed instruments. We have used the term dyadic OPTION scale to differentiate this scale from the observer (third party assessment) version. A strength of this study is the use of a well described method to develop a new tool [17,19]. We adopted an iterative design and assessment cycle and this allowed the modification and refinement of instrument items in response to respondents’ comments. By stratifying our sample by preferences for daily newspaper readership we also ensured the sample diversity. Grammatical and semantic problems were addressed in the first two rounds of interviewing whilst the introductory statement was amended after the final round of interviewing. Three cycles of interviews attained data saturation, with minimal modifications being made to the individual items after the third round of interviewing. A further strength was the use of a well established measurement instrument [14,24] as the basis for the design of a new measure. The main limitation of the study is that a single interviewer conducted the analysis of the data prior to discussion with the original scale developer. Although this ensured continuity and consistency during interviewing and analysis, added rigour could have been achieved by dual analysis by an independent second researcher. 4.2. Conclusion There is evidence that health services researchers are beginning to use a dyadic and relationship-centred approach to clinical decision making [10,25–27]. In other words, health services researchers are moving from studying groups of patients and health professionals separately to studying both simultaneously. For example, five of the six most recently developed instruments had corresponding patient versions [28–35]. Moreover, the authors of two of these instruments, only obtained evidence of validity and reliability data for the combined use of the physician’s and patient’s questionnaires [28,30,36–40]. This observation suggests that, increasingly, interdependence is being recognised as an inherent factor within the patient–physician relationship which ultimately impacts upon the outcome of the consultation. Furthermore, the patient–physician relationship is an important component of physicians’ satisfaction with their job [27]. Physicians’ judgements about their experience with individual patients both reflect and shape what takes place during office visits and beyond [41]. This balance supports what has previously been described on the basis of personal communication needs which can potentially be fulfilled in the clinical encounter [41]. Therefore, future research in the field of clinical decision making should foster the use of patient and clinician versions of a similar instrument. In line with the growing interest for shared decision making, this may allow for a more comprehensive assessment of the complexity of the clinical decision making process and thus of its dynamic and reciprocal nature [32]. Saba has highlighted the need to evaluate the ‘subjective experience of partnership’ when considering the degree of collaboration achieved within a consultation [3]. Previous attempts to assess shared decision making have led to the development of instruments such as the Patients’ Perceived Involvement in Care Scale (PICS) [16] and observer OPTION [13,14]. PICS is one of the few questionnaires which attempts to assess the patient’s role in the
consultation. It considers the patient’s involvement in the consultation as a whole as opposed to OPTION which focuses more on the decision making aspects of the encounter. This leads to difficulties when trying to analyse the data from PICS specifically to measure involvement in shared decision making. In addition to this, many of the PICS items reflect the semantic, grammatical and construct mismatch problems which we encountered during this study. Acknowledging interdependence within a medical consultation is central to understanding and evaluating the encounter and realising the true relevance of such an interaction. We cannot therefore address this issue without attempting to assess the consultation from the dual perspectives of the actors involved. This problem is at the centre of current debates about the assessment of shared decision making [4]. Dyadic OPTION is the first tool to allow us access to this dual perspective whilst using identical items and construct and we hope it use will provide new insight the complex interdependent nature of clinical encounters, and in particular to the intricacies of attempts of achieve shared decision making. 4.3. Implications for practice This study has demonstrated that the dyadic OPTION instrument is well understood by clinical and patient respondents and has the potential to gather valid data about perceived patient involvement in shared decision making. We hope that by using cognitive debriefing to develop dyadic OPTION, we have been able to conceptualise difficult, inconsistently practiced constructs as accurate, accessible, comprehensible items and that these items allow truthful representative responses with minimal processing burden to respondents. However, cognitive debriefing alone does not guarantee that this has been achieved. We recommend therefore that further validation be carried out on target clinical populations to ascertain the scale’s true reliability and validity in practice before it is used for the collection of dyadic interdependence data. Conflict of interest The authors report no conflicts of interest. References [1] Elwyn G, Edwards A, Mowle S, Wensing M, Wilkinson C, Kinnersley P, Grol R. Measuring the involvement of patients in shared decision making: a systematic review of instruments. Patient Educ Couns 2001;43:5–22. [2] Dy SM. Instruments for evaluating shared medical decision making: a structured literature review. Med Care Res Rev 2007;64:623–49. [3] Saba G, Wong S, Schillinger D, Fernandez A, Somkin C, Wilson C, Grumbach K. Shared decision making and the experience of partnership in primary care. Ann Fam Med 2006;4:54–62. [4] Edwards A, Elwyn G. Inside the black box of shared decision making: distinguishing between the process of involvement and who makes the decision. Health Expect 2006;9:307–20. [5] Stevenson FA. General practitioners’ views on shared decision making: a qualitative analysis. Patient Educ Couns 2003;50:291–3. [6] Kiesler DJ, Auerbach SM. Optimal matches of patient preferences for information, decision-making and interpersonal behavior: evidence, models and interventions. Patient Educ Couns 2006;61:319–41. [7] Elwyn G, Edwards A, Kinnersley P, Grol R. Shared decision making and the concept of equipoise: the competences of involving patients in healthcare choices. Br J Gen Pract 2000;50:892–9. [8] Le´gare´ F, Elwyn G, Fishbein M, Fre´mont P, Frosch D, Gagnon MP, Kenny DA, Labreque M, Stacey D, St-Jacques S, van der Weijden. Translating shared decision-making into health care clinical practices: proof of concepts. Implementation Sci IS 2008;3:2. [9] Kenny Dak DA, Cook WL. Dyadic data analysis. New York: The Guilford Press; 2006. [10] Le´gare´ F, Moher D, Elwyn G, LeBlanc A, Gravel K. Instruments to assess the perception of physicians in the decision-making process of specific clinical encounters: a systematic review. BMC Med Inform Decis Making 2007;7:30.
E. Melbourne et al. / Patient Education and Counseling 78 (2010) 177–183 [11] Mullen PD, Allen JD, Glanz K, Fernandez ME, Bowen DJ, Pruitt SL, Glenn BA, Pignone M. Measures used in studies of informed decision making about cancer screening: a systematic review. Ann Behav Med 2006;32:188– 201. [12] Braddock CH, Fihn SD, Levinson W, Jonsen AR, Pearlman RA. How doctors and patients discuss routine clinical decisions informed decision making in the outpatient setting. J Gen Intern Med 1997;12:339–45. [13] Elwyn G, Edwards A, Wensing M, Hood K, Robling M, Atwell C, Grol R. Shared decision making: developing the OPTION scale for measuring patient involvement. Qual Safe Health Care 2003;12:93–9. [14] Elwyn G, Hutchings H, Edwards A, Rapport F, Wensing M, Cheung W, Grol R. The OPTION scale: measuring the extent that clinicians involve patients in decision-making tasks. Health Expect 2005;8:34–42. [15] Edwards A, Elwyn G, Hood K, Robling M, Atwell C, Holmes-Rovner M, Kinnersley P, Houston H, Russell I. The development of COMRADE—a patientbased outcome measure to evaluate the effectiveness of risk communication and treatment decision making in consultations. Patient Educ Couns 2002; 50:311–22. [16] Lerman CE, Brody DS, Caputo GC, Smith DG, Lazaro CG, Wolfson HG. Patients’ perceived involvement in care scale. J Gen Intern Med 1990;5:29–33. [17] Streiner DL, Norman GR. Health measurement scales. New York: Oxford University Press; 1989. [18] Levine RE, Fowler FJ, Brown JA. Role of cognitive testing in the development of the CAHPS1 hospital survey. Health Serv Res 2005;40:2037–56. [19] Willis GB. Cognitive interviewing: a tool for improving questionnaire design. Sage Publications Inc.; 2004. [20] Bourn J. Skills for life: improving adult literacy and numeracy. National Audit Office; 2004. [21] Strauss AL, Corbin J, Lynch M. Basics of qualitative research: grounded theory procedures and techniques. CA: Sage Newbury Park; 1991. [22] Rothgeb J, Willis G, Forsyth B. Questionnaire pretesting methods: do different techniques and different organizations produce similar results; 2001. p. 2001. [23] Presser S, Blair J. Survey pretesting: do different methods produce different results? Sociol Methodol 1994;73–104. [24] Weiss MC, Peters TJ. Measuring shared decision making in the consultation: a comparison of the OPTION and Informed Decision Making instruments. Patient Educ Couns 2008;70:79–86. [25] Kashy DA, Kenny DA. The analysis of data from dyads and groups. Handbook of research methods in social and personality psychology 2000;451–77. [26] Kenny DA, Cook W. Partner effects in relationship research: conceptual issues, analytic difficulties, and illustrations. Pers Relat 1999;6:433–48. [27] Suchman AL. A new theoretical foundation for relationship-centered care. J Gen Intern Med 2006;21:40–4.
183
[28] Sewitch MJ, Abrahamowicz M, Dobkin PL, Tamblyn R. Measuring differences between patients’ and physicians’ health perceptions: the patient–physician discordance scale. J Behav Med 2003;26:245–64. [29] Legare F, O’Connor AM, Graham ID, Wells GA, Tremblay S. Impact of the Ottawa decision support framework on the agreement and the difference between patients’ and physicians’ decisional conflict. Medical Decis Making 2006;26:373. [30] Harmsen JAM, Bernsen RMD, Meeuwesen L, Pinto D, Bruijnzeels MA. Assessment of mutual understanding of physician patient encounters: development and validation of a mutual understanding scale (MUS) in a multicultural general practice setting. Patient Educ Couns 2005;59:171–81. [31] Guimond P, Bunn H, O’Connor AM, Jacobsen MJ, Tait VK, Drake ER, Graham ID, Stacey D, Elmslie T. Validation of a tool to assess health practitioners’ decision support and communication skills. Patient Educ Couns 2003;50:235–45. [32] Dolan JG. A method for evaluating health care providers’ decision making: the provider decision process assessment instrument. Med Decis Making 1999; 19:38. [33] Cegala DJ, Coleman MT, Turner JW. The development and partial assessment of the medical communication competence scale. Health Commun 1998;10:261– 88. [34] Campbell C, Lockyer J, Laidlaw T, MacLeod H. Assessment of a matched-pair instrument to examine doctor–patient communication skills in practising doctors. Med Educ 2007;41:123–9. [35] Cegala DJ, McNeilis KS, McGee DS. A study of doctors’ and patients’ perceptions of information processing and communication competence during the medical interview. Health Commun 1995;7:179–203. [36] Dobkin PL, De Civita M, Abrahamowicz M, Bernatsky S, Schulz J, Sewitch M, Baron M. Patient–physician discordance in fibromyalgia. J Rheumatol 2003; 30:1326–34. [37] Dobkin PL, Sita A, Sewitch MJ. Predictors of adherence to treatment in women with fibromyalgia. Clin J Pain 2006;22:286. [38] Sewitch M, McGill U, Dept. of Occupational H. Effect of Discordant Physician– patient Perceptions on Patient Adherence in Inflammatory Bowel Disease: National Library of Canada = Bibliothque nationale du Canada; 2003. [39] Sewitch MJ, Dobkin PL, Bernatsky S, Baron M, Starr M, Cohen M, Fitzcharles MA. Medication non-adherence in women with fibromyalgia. Br Soc Rheumatol 2004;648–54. [40] Sewitch MJ, Abrahamowicz M, Barkun A, Bitton A, Wild GE, Cohen A, Dobkin PL. Patient nonadherence to medication in inflammatory bowel disease. Am J Gastroenterol 2003;98:1535–44. [41] Suchman AL, Roter D, Green M, Lipkin Jr M. Physician satisfaction with primary care office visits. Collaborative Study Group of the American Academy on Physician and Patient. Med Care 1993;31:1083–92.
