Background
Shared decision-making (SDM) is one of the main activities of patient-centred care [
1,
2]. It involves the exchange of information and negotiation between the clinician and the patient to agree on the best way to medically proceed for the individual patient [
3]. Often the decision-making process is complex - especially when preference sensitive aspects are involved. Various interventions have been developed to facilitate SDM.
Patient Decision Aids (PtDAs) are intended to support patients in making specific and deliberated choices among healthcare options [
4‐
6]. In contrast to more general health education materials (e.g. information leaflets), PtDAs specifically state the decision being considered and stress the relevance of a SDM process [
4‐
6]. Furthermore, PtDAs provide information on all available treatment options and help patients clarify what matters to them regarding these treatment options [
4‐
6]. A systematic review recently revealed that, for many different decisions and conditions, PtDAs can improve patients’ knowledge about options, risk perceptions, feelings of being informed and being certain about what matters to them [
7]. Furthermore, with the use of PtDAs, patients more often reach decisions that are consistent with their personal values [
7]. Finally, PtDAs can improve patient-doctor communication [
7].
The International Patient Decision Aids Standards (IPDAS) Collaboration states that the development of PtDAs should be systematic and include consultations of patients and health professionals [
4‐
6]. However, many studies of PtDA development projects do not report on having involved patients during their development [
8]. In response to this omission, the IPDAS’ evidence base has recently been updated to include a development process model that places more emphasis on patient involvement during PtDA development [
6,
8]. This process model provides a step-wise approach to careful and systematic development, evaluation and implementation of PtDAs. Although this new comprehensive model provides an overview of the entire development process, it does not provide guidance on
how to best involve patients and health professionals nor which research methods to use. The authors, therefore, urged PtDA developers to complement the IPDAS development process model with other guidelines, such as a user-centred design approach [
8]. In a user-centred design, specific research methods are used to consult with potential users relatively early within the developmental timeframe [
9,
10]. This approach allows developers to adopt and implement user-centred input, resulting in the product more adequately fulfilling users’ needs and, consequently, positively affecting user satisfaction [
9,
10].
While patients with rheumatic diseases often face long-term treatment decisions, only a few studies have been reported on PtDAs for initiating disease modifying anti-rheumatic drug (DMARD) therapy [
7,
11‐
13]. DMARDs are the core element of the management of inflammatory arthritis in order to control the disease process and to relieve or reverse symptoms [
14‐
16]. DMARDs form two major classes: synthetic chemical compounds (sDMARDs) and biological agents (bDMARDs). With regard to DMARDs, the decision-making process has become increasingly complex, as numerous therapeutic options are available. In addition, new treatment strategies are rapidly evolving, but without sufficient information on differential efficacy and safety [
14‐
16]. When weighing the options, elements to consider include treatment efficacy, approximate time-to-benefit, possible side effects, current and future risks, cost-effectiveness, route of administration and impact on daily life. Because this complex decision-making process concerns both clinical and preference-sensitive aspects, the choice of treatment needs to be based on a shared decision between the patient and rheumatologist.
The reported PtDAs for initiating DMARDs mainly focus on the decision whether to initiate one specific DMARD or a particular class of DMARDs [
7,
11,
13]. Yet, our previous study showed that patients would like to be informed about multiple specific DMARDs [
17]. In fact, previous research has shown that patients with a rheumatic disease are often less informed and less involved in decision-making than they would prefer [
17‐
24]. SDM barriers reported by patients include being unaware of having a choice, lack of medical knowledge and a power imbalance in the doctor-patient relationship [
17,
25].
In order to fulfil this need of patients with rheumatic disease, the aim of this study was to develop a tool that could compare multiple specific DMARDs. This paper describes the development of a web-based PtDA with use of the IPDAS development process model and user-centred design methods. The PtDA is intended for inflammatory arthritis patients who face the decision whether to initiate DMARDs.
Discussion
We have described in detail the development of a PtDA for patients with inflammatory arthritis that helps them to choose between DMARDs. This PtDA was developed using the IPDAS development process model [
8] and user-centred design methods [
9,
10]. Based upon the needs assessments of both patients and health professionals, we constructed a web-based PtDA consisting of the following parts: 1) general information about SDM, inflammatory arthritis and DMARDs; 2) an application to compare specific DMARDs attributes; 3) exercises to gain insight into the patient’s preferences, worries and questions; and 4) a printed summary of the patient’s notes, preferences, worries and questions to be discussed with the rheumatologist at the next consultation. The results of the alpha tests revealed that the developed PtDA largely satisfied the needs of both patients and health professionals and thus has the potential of being a valuable tool for patients who need to choose between DMARDs.
The overall process of development was satisfactory. The IPDAS development process model is relatively new and has yet to be substantially tested. Nevertheless, this process model proved to be systematic and helpful to our iterative development of the PtDA as well as compatible with user-centred design methods. In addition, the user-centred design methods proved to be helpful in gaining valuable insights into different stakeholders’ needs with regard to the PtDAs content and design and how it should be integrated into daily practice.
Firstly, rapid prototyping (i.e. the use of paper prototypes) proved to be of additional value to the needs assessment interviews. Patients (but also clinicians) often have difficulty conceptualizing what a PtDA is and how it should look and function, which might limit them in expressing their needs. With the use of rapid prototyping, it was easier for users to express their wishes and needs and to give critical input. For this reason, we recommend using rapid prototyping in the development process of future PtDAs.
Secondly, according to the IPDAS development process model, health professionals’ perceptions of patients’ needs for information and decision support should be assessed. We recommend conducting this step after having elucidated the patients’ needs. In our study we intentionally conducted the study first among patients and presented the results of this study during the session with the health professionals. By doing so, the results of the needs assessments among patients were largely confirmed. But perhaps more essentially, this procedure proved to be effective in creating support among more sceptical health professionals for the development of the PtDA. Health professionals who initially questioned the added value of a PtDA had less misgivings and were more willing to use it.
Thirdly, we recommend not only asking health professionals about their perception of patients’ needs, but also asking them about their own needs and thoughts on implementing a PtDA into their practice. Their practical and expert knowledge on the decision-making process can be of great value for the integration of a PtDA into the patient pathway and daily workflow of health professionals, and consequently enhance the adoption and implementation of the PtDA. The adoption and implementation of PtDAs using a referral model (i.e. health professionals inviting eligible patients to use the PtDA) is often challenged by indifference on the part of health professionals [
43]. This indifference may stem from a lack of confidence in the content of the PtDAs and concerns about disruption of established workflows [
43]. Our PtDA is still being successfully used after conclusion of the project (after beta testing), and newly developed DMARDs have since been added by the health professionals. This indicates that the iterative and extensive involvement of health professionals and the acknowledgement of their needs for the PtDA were important in creating ownership.
Finally, as the scope of the internet grows, PtDAs will be more and more computerized and web-based. These formats may offer many opportunities, not only for rapid adjustments of the PtDAs, but also for studying usage and usage behaviour in detail. For instance, the amount of log-ins, page-views, and time spent on the PtDA could be logged, but also patterns of usage (e.g. How do users navigate? Which elements and combination of elements are often used? When do users drop-out?) and users’ input (e.g. selected preferences, worries and questions) [
44,
45]. This information could be used to gain more insight into users’ (evolving) needs and improve the PtDAs usage, usability and impact. Therefore, we recommend adding researchers to the stakeholders of web-based PtDAs and advising researchers to include logging anonymous user data as a requirement for the PtDA.
Compared to most previously reported PtDAs, the PtDA in this study encompasses many treatment options [
46]. Although patients have the right to be informed about all treatment options [
47] and one of the quality domains of the IPDAS is to provide all options to patients [
5,
6], we had a valuable discussion with the health professionals about whether to give patients access to all available medication options. Our previous studies [
17,
26] showed that patients not only worry about the side effects and potential risks of their current or proposed treatment, but also had significant worries about the risks of future treatments and about ‘running out of options,’ should the proposed medication fail to work. To decrease this uncertainty, patients expressed a need to have an overview of all available options, for the time being as well as for the future. However, patients will most certainly be overwhelmed by all the different options and their pros and cons. To guide patients through this plethora of options, we chose to provide them in writing (the referral card) a clear personal recommendation of their most appropriate medication options. To respect their needs and rights, we also provided an overview of all other medication options and gave patients access to this information as well. This format may also be suitable for PtDAs that address multiple treatment options for other conditions, such as asthma or diabetes.
Previously developed PtDAs for initiating DMARDs mainly differ in the amount of treatment options that are included and how it is integrated in the patient pathway and distributed to patients. Most of these PtDAs focus on the decision whether to initiate one specific DMARD or a particular class of DMARDs, are to be used outside the clinical encounter and the plan for distribution is often unclear [
7,
11,
13]. Only one other PtDA includes all DMARD options and a clear plan for integration in daily clinical care. It consists of a card deck to be used during the medical encounter and is developed for patients with limited health literacy or limited English language proficiency [
12].
One limitation of the currently developed PtDA is that it does not present outcome probabilities. This is due to the large number of treatment options included and the lack of evidence on differential efficacy and safety. Presenting outcome probabilities is a quality domain of the IPDAS [
5,
6]. Not presenting outcome probabilities limits the comparison of treatment options. However, the PtDA does present the negative and positive features in equal detail and a structured manner of the available treatment options for clinical and practical elements as well as the possible implications for daily life. This way of presenting outcomes was regarded as useful by the patients.
Furthermore, it should be noted that some wished-for attributes of the PtDA were not realized due to their technical complexity, limited time and finances, and privacy issues. A few (elderly) patients stated they did not want a computerized version of the PtDA because they feared that they lacked sufficient computer skills or did not have access to a computer/internet. Since it was only a few patients who stated this and because of limited time and finances, we did not develop a paper and pencil PtDA. However, we chose to acknowledge this need by having a computer available for patient use in the hospital and having a nurse guide the patient through the PtDA decision-making process.
The content of the PtDA is now tailored to the individual based on gender and desire to have children, but not on risk profile. With the insufficient evidence on differential efficacy and safety of DMARDs, this attribute remains a challenge for future research.
Due to time and financial limitations, it was also infeasible to develop a digital referral to the PtDA accompanied by a personal recommendation for appropriate medication options. Nevertheless, it is technically interesting to digitalize this process and may even improve the uptake. Sending the summary with the patients’ notes, preferences, worries and questions directly to the health professional was also not realized because it raised privacy issues. Perhaps in the future, the PtDA could incorporate an option that would allow patients to send their summary to their health professional. Such additional attributes might also help increase the users’ uptake.
Compared to the majority of PtDA developments, our PtDA substantially attempted to include all stakeholders. However, only small groups of participants were involved and all (patients and health professionals) were recruited from two hospitals. Although the number of participants in all our steps actually match the recommended numbers (see [
27,
48,
49]), our results may not be generalizable. When further developing or distributing this PtDA, attention should be paid to involve a larger and wider group of users.
Additionally, we have not compared and assessed our development of a web-based PtDA using the IPDAS process model and user-centred design methods with the development a web-based PtDA in a different way. To assess this, two web-based PtDAs need to be developed using different methods, but using the same ideas or content as the starting point. In our view, this seems unfeasible and undesirable. From our study, however, we can say that our methodology did allow us to clarify needs and we were able to adapt the PtDA to these needs.
Finally, in this paper we have not addressed evaluating the effectiveness or impact of the PtDA. To evaluate whether the PtDA is successful in improving patient participation and supporting SDM, we have conducted a post-test only study with a historical comparison group (beta testing). The results are published elsewhere [
https://arthritis-research.biomedcentral.com/articles/10.1186/s13075-016-1138-3]. To assess the use, appreciation and perceived impact from an expert perspective, we have recently conducted a focus group study with health professionals. These results are also being analysed at this moment.