Patients’ views on the use of an Option Grid for knee osteoarthritis in physiotherapy clinical encounters: An interview study

Abstract Background Patient decision support tools have been developed as a means of providing accurate and accessible information in order for patients to make informed decisions about their care. Option Grids™ are a type of decision support tool specifically designed to be used during clinical encounters. Objective To explore patients’ views of the Option Grid encounter tool used in clinical consultations with physiotherapists, in comparison with usual care, within a patient population who are likely to be disadvantaged by age and low health literacy. Methods Semi‐structured interviews with 72 patients (36 who had been given an Option Grid in their consultation and 36 who had not). Thematic analysis explored patients’ understanding of treatment options, perceptions of involvement, and readability and utility of the Option Grid. Results Interviews suggested that the Option Grid facilitated more detailed discussion about the risks and benefits of a wider range of treatment options for osteoarthritis of the knee. Participants indicated that the Option Grid was clear and aided their understanding of a structured progression of the options as their condition advanced, although it was not clear whether the Option Grid facilitated greater engagement in shared decision making. Discussion and conclusion The Option Grid for osteoarthritis of the knee was well received by patient participants who reported that it helped them to understand their options, and made the notion of choice explicit. Use of Option Grids should be considered within routine consultations.


| INTRODUCTION
Shared decision making (SDM) is a process in which patients are supported by clinicians within the consultation to consider treatment choices that are aligned with their informed preferences. 1 Clinicians are required to elicit and respect patient preferences. There is a strong policy directive driving SDM in the United Kingdom, 2,3 which is mirrored in many other countries worldwide. 4,5 Patient decision support tools have been developed as a means of providing accurate and accessible information in order for patients to make informed decisions about their care. Evidence suggests that these tools given ahead of the consultation have a positive effect on increased patient knowledge, 6 but there is very limited evidence to show that they can improve SDM within the consultation. 7 However, there are studies showing that tools designed to be used within the clinical encounter can have a positive effect on the process of SDM. 8 Option Grids™ are a type of decision support tool specifically designed to be used during clinical encounters. They are designed to orientate patients to the existence of choice by displaying treatment options side by side, organized by the concerns or questions that frequently occur to patients, thereby increasing the potential for SDM. 9 The need for SDM is particularly relevant for patients with osteoarthritis of the knee, a common cause of physical disability in older adults that is associated with significantly reduced quality of life. 10 The prevalence of osteoarthritis of the knee is predicted to rise in line with an ageing and increasingly obese population. 11 Despite an increasing number of knee replacement operations being conducted in the UK, 12 there is evidence of large unwarranted clinical variation which may be influenced by participants' (lack of) awareness of the treatment options including knowledge and understanding of the risks and benefits of knee replacement surgery and significant differences being expressed by patients on whether surgery is the only reliable solution compared to others who wish to avoid surgery at all costs. 13 Osteoarthritis of the knee can be managed by lifestyle measures such as improving activity levels to build strength in the joint and weight loss. Other management options include medication such as painkillers and anti-inflammatories, steroid joint injections and knee replacement surgery.
Much of the research on SDM to date has been generated from trials where patient participants have had high educational and socio-economic backgrounds. Concern has been expressed that decision support tools might exacerbate health inequalities as most are developed in English and may not be accessible to people with poor levels of health literacy or who do not have English as a main language. 14 A recent systematic review has demonstrated that overall SDM interventions may benefit disadvantaged groups more than non-disadvantaged groups, although interventions which were specifically tailored to disadvantaged groups were found most effective at achieving benefits. 15 We had previously conducted a trial which demonstrated SDM increased when clinicians used the knee osteoarthritis Option Grid in consultations. 16 The aim of this substudy was to explore patients' views of the Option Grid encounter tool within a patient population who are likely to be disadvantaged by age and low health literacy, specifically eliciting patients' views about how treatment options had been presented, and perceptions of their involvement in decision making.

| METHODS
The current study reports data from interviews with patients recruited to a trial of the efficacy of an Option Grid decision support tool for osteoarthritis of the knee. The trial used a multiple interrupted time series (stepped wedge) design where patients were allocated sequentially to consult with an extended scope specialist physiotherapist who had consented to the trial. Patients are referred to the service from primary care and may then access a range of multidisciplinary specialist care for rheumatology, orthopaedics and musculoskeletal pain.
Participating clinicians (physiotherapists) received brief (30 minutes) training on how to use the Option Grid within a routine consultation after having consulted with six patients using usual care procedures (the control group).
The Option Grid for osteoarthritis of the knee was developed by a team of clinicians and academics with an interest in SDM, musculoskeletal conditions and pain management. User testing informed its refinement. 17 The Option Grid presents three recommended treatment options in a tabular format: pain killers, intra-articular corticosteroid injections, and knee replacement surgery for symptoms refractory to non-surgical treatment. 18 Patients' frequently asked questions are listed on the side of the grid, allowing readers to compare answers across the three treatment options. A copy of the grid is presented in Table 1.
Process and outcome measures were compared for patients participating in the intervention time period with those in the control time period. 19 Ethical approval was obtained from the local research ethics committee (Ref: 11/WA/0356).

| Participants and setting
Inclusion criteria for patient participants were adults with a diagnosis of osteoarthritis of the knee made by their GP or extended scope physiotherapist who were able to give informed consent. Potentially eligible patient participants were identified from referrals to a single musculoskeletal service that provides integrated services to the local population in Oldham, Greater Manchester, UK. A member of the clinical team attempted to telephone all eligible patients and invited them to participate in the study. Patients who expressed interest were sent an invitation letter, information sheet and consent form with their appointment letter. Informed written consent was given when they attended their appointment. Participants were recruited until the required sample size of 72 was reached (six physiotherapists consulting six patients each pre-intervention and six patients each postintervention). 19 In total, 78 participants were recruited, but six later were found to be ineligible or withdrew from the study. All 72 patient participants were interviewed. Clinicians were also interviewed and data from this are reported elsewhere, 20 as are the main findings of the trial, 16 in addition to a discourse analysis of how the Option Grids were used when an interpreter was present in the consultation. 21

| Data collection
For the purposes of the trial, a range of patient participant sociodemographic variables were recorded including age, postcode, ethnicity, educational attainment and main language spoken. Patients were also asked to complete the REALM-R reading exercise, which is a brief eight-item measure of health literacy. 22 Scores of 6 and under are considered to indicate poor health literacy. A research nurse (KK) was present in the consultation and observed the use of the Option Grid in the consultation and made structured field notes recording how the Option Grid was used during the consultation (number of questions asked, who held the Option Grid etc.). The consultation was also audiorecorded, but audiorecorded data and field notes are not presented in this paper. Short semi-structured interviews with all patient participants were conducted immediately following the consultation in a private room by the same research nurse and were audiorecorded and transcribed verbatim. The aim of the interviews was to explore patients' views about the consultation, including whether they felt involved in the decision-making process, their views about how risks and benefits of treatment options had been presented to them, reasons for their prior and current treatment preferences, and T A B L E 1 Option grid for Osteoarthritis of the knee. This grid is designed to help you and your clinician decide how best to manage your knee pain and activity level. The first steps are to become as fit and close to your ideal weight as possible and to consider having physiotherapy. Surgery is normally recommended only after other treatments have been tried

Frequently asked questions Pain killers Joint injections (steroids) Knee replacement surgery
Will this reduce the pain I have in my knee?
It depends on which tablets are taken. Tablets like ibuprofen are effective for 50 in 100 people. Over the counter tablets, like paracetamol, including those that have codeine, are also effective.
Some people get good symptomatic relief after an injection, which may include pain relief and/or reduced swelling.
About 90 in 100 people who have this operation say it leads to relief of most or all of their pain, over time. 10 in 100 people say it does not lead to significant pain relief.
Will this treatment help improve which activities I can manage to do?
It may. As you get pain relief, you should be able to be more active and this in turn can also help to reduce pain. It helps to take painkillers before doing physical activity.
Yes, usually for up to a month or so. Plan to be more active as a result of the pain relief. Advice from a physical therapist may help.
Yes, the majority of patients experience improvement in their activity level. However, not everybody sees differences in their ability to walk or climb stairs.
Are there any risks to this treatment?
As with all medications, pain killers have some side-effects. For example, codeine often leads to constipation and prolonged use of antiinflammatory tablets like ibuprofen increases your risk of stomach bleeding.
There is a small risk of frequent injections causing cartilage damage, especially in weightbearing joints. Allergic reactions and infections due to joint injections are uncommon. You might feel slight pain at the injection site for a few days.
Wound infection needing treatment occurs in 5 in every 100 people. Blood clots in the leg occur in 2 in 100 people.
The risks from surgery increases if you have other conditions, such as heart or lung disease, are a smoker or are overweight.
How long will it take me to feel better after the treatment?
You may start experiencing pain relief within a few days of when you start taking the medication.
Most people who experience relief feel better within the first week or so after the injection.
Pain relief is gradual. You will stay in the hospital for around 3-5 days. Most people walk unaided after 3 months. Full recovery usually takes between 6 and 12 months.
Will I need to have more treatment or surgery?
If things don't get better, talk to your clinician about other treatment options.
Pain relief lasts for up to a month or so. You can only have up to 4 injections per year.
Most knee replacements can last 15 years, many last longer.
What are the outcomes for people with arthritis who have this treatment?
Many people cope well by using medication, being active and losing weight. Reducing your pain may help you achieve the benefits of exercise.
Some people have good relief by having injections when swelling and pain cause problems.
Surgery is usually considered after other options have been tried. About 80 in every 100 people are satisfied after having a knee replacement. About 20 in every 100 are not satisfied. their views about the use of the Option Grid within the consultation (if it was used). The interview schedule is outlined in Table 2 and was modified depending on whether an Option Grid was used and to allow the research nurse to incorporate her observations of the consultation. Data were collected over a period of 18 months.

| Data analysis
A thematic analysis of the data set was undertaken as described by  Table 3 presents characteristics of the 72 patients who were interviewed as part of the trial. Patient participants from each arm of the study were similar in terms of age and REALM-R score, although there were differences in gender ratio between groups. The mean REALM-R score for each group was slightly lower than the adult population mean of 6.8 (SD 2.1) found in a US validation study of the measure 22 ; 29% of the participants had educational attainment at A' level or above, compared to 39.5% for England and Wales as a whole. 25 We report the findings from the qualitative interviews grouped into three main sections which reflect grouped themes from the thematic analysis: (i) increased awareness of the treatment options, (ii) Option Grid is acceptable to patients and (iii) positive perception of involvement for most, but not all, patients. Data extracts are used to illustrate our data alongside participant identification number and socio-demographic characteristics (whether an Option Grid was used, gender, REALM score and age).

| Increased awareness of the treatment options
Participants who received the Option Grid appeared to demonstrate a greater awareness that there were a number of options available, although not all of the options were suitable for every patient: Observation data also indicated that if an Option Grid was not used in a consultation, it was more likely that the clinician would focus discussion on the risks and benefits of the most likely treatment option.
Consequently, patients were not made fully aware of the risks and benefits of the full range of treatment options. However, patients were not always critical of this strategy, rather they often recognized that this had occurred and attributed it to tailored information-giving appropriate to their particular symptoms. When asked about whether she had been told the risks and benefits of knee replacement surgery, participant 20 indicated that she thought this was only relevant to the consent process rather than the decision-making process: No she didn't tell me. I suppose that would come… I should imagine like when you have a pre-op they will tell you all the pros and cons then… they usually give you the percentage of the success rates as well.
(P20, without OG, female, REALM 8, age 67) From the patient interviews, it is difficult to ascertain clearly whether the Option Grid actually changed people's treatment preferences despite the interviewer asking whether the patient felt that the Option Grid had influenced their discussion and treatment decision.
Often patients were not able to give a clear answer to the question of what their prior treatment preference was, instead stating that they wanted to get their knee "fixed" or receive some "wonder drug" (patient 74). Patients also understandably found it difficult to distinguish between the information given by the clinician and the information that they received due to the Option Grid being used in the consultation (any added benefit of the Option Grid

| Option Grid is acceptable to patients
Participants who had received an Option Grid expressed positive perceptions in relation to its content, reporting that it was clear and helped them to understand their options.
Although there were a few participants who suggested that the grid had provided them with enough, possibly more than enough, information. However, the limitations for the three non-English speaking patients in the study were clear as the same participant indicated that she would have no future use for the Option Grid, as she could not read it. In contrast, most participants whose first language was English felt the Option Grid would be a useful "take home tool" that they could refer to in the future and show family members or friends who had not attended the consultation with them.
Data from the interviews also suggested that patients who had been given an Option Grid were provided with more of a sense of a structured sequence of their treatment options as their disease course progressed, although some patients who had not received the Option Grid were also able to identify this staged approach: For one patient, the perceived delay in receiving the Option Grid in the consultation rather than prior to the consultation was expressed as frustration as she believed had she been better informed before the consultation then the consultation might have concluded with a treatment decision rather than a recommendation to reflect more on options.

I think definitely it would have helped an awful lot where as now I have to take it home and read it and I might have been able to make a decision and have discussed it and then I can
have that treatment-so that definitely had I seen it before.
(P55, with OG, female, REALM 8, age 87) Participants were also able to reflect that although they had been given the Option Grid in the consultation, they, or members of their family, would be able to reread it and reflect on it after the consultation.
You know, when I get home I'd most probably read it completely properly again you know but as far as I'm concerned it did give you a lot of information.

| Positive perception about involvement for most, but not all, patients
When asked whether the Option Grid promoted their involvement in decision making, most participants answered positively. However, most participants in both groups appeared to feel that they had made their own treatment decision, with guidance from the clinician, making it difficult to determine the added benefit of the Option Grid from patients' reports of their involvement: [Clinician] helped me make the decision but it was my de-  Furthermore, there is some evidence that patients with low literacy struggled to engage with the Option Grid decision aid.

| Strengths and weaknesses
Interviews were conducted immediately following the consultation and so patients would be expected to have reasonable recall of their experiences of their encounter. The data set of 72 interviews is comparatively large for a qualitative study and patients recruited were similar in respect to age and gender compared to all patients deemed eligible for the study and in respect to age and gender were typical of the population that is normally affected by osteoarthritis of the knee.
Although the interviews were conducted by the research nurse, participants were encouraged to be candid in sharing their views about the consultation.
Lessons learned from a review of SDM implementation programmes suggest that local context is a vital factor in implementation. 26 The study site has a history of partnership working to promote self-care, providing skills and training in SDM and has the support of clinical leaders, which are factors associated with other local change programmes that have often fared better at implementing new models of patient-centred care. Patients in the "usual care" group were therefore likely to be consulting with clinicians who were practising SDM to some degree. The addition of the Option Grid in the "intervention arm" of the study, as a means of structuring the discussion of treatment options, was therefore an adaptation to the way in which the physio-  identified that the decision aid created a space for discussion of treatment options. Our data also indicate that patients find encounter tools a useful and flexible scaffold around which to deliberate their options during the consultation. Data from this paper, which demonstrates patient support for the Option Grid, complement our main findings that use of the Option Grid increases the amount of SDM in the consultation. 16 It also complements our data derived from interviews with clinicians (reported elsewhere 20 ) which demonstrates that clinicians, although initially sceptical, became more confident that Option Grids were acceptable to patients, allowed them to discuss options with a more neutral approach and encouraged engagement from patients.

| Comparison to other literature
Together, these substudies indicate support for the Option Grid from both parties within the clinician-patient dyad.
Previous survey research has found that patients reported more discussion of pros than cons in relation to 10 common treatment decisions. 31 Encounter tools, such as Option Grids, which prompt clinicians and patients to consider both risks and benefits of treatment options, are likely to lead to better informed patients and less decision regret.
Whilst the Option Grid displays all the treatment options side by side, the verbalization and navigation of the grid was directed by the clinician and the options were intentionally limited on occasion. Previous qualitative work with paediatric rheumatology clinicians found that discussion of treatment options were not only based on clinician preferences but also limited by the clinician's assessment of the clinical situation, 32 and this also appears to be the case in our study.
Evidence suggests that socially disadvantaged patients are more vulnerable to power imbalances within the clinician patient relationship. 33 Even patients who are well educated and well informed often struggle to express their knowledge during consultations and participate meaningfully in treatment decisions. 33 Such power imbalances typically result in patients adopting the role of a passive or compliant patient, a tendency for patients to underplay their knowledge and experiences, and so defer the decision to the clinician whom they see "owns" the knowledge. Although our main trial data indicated that overall in our population of older adults with lower than average health literacy the use of the Option Grid increased levels of SDM, increased patient knowledge and readiness to decide, 16 in our data we observed some evidence that patients with lower health literacy scores were less positive about being involved in their treatment decision and adopted a fairly passive approach to treatment decision regardless of whether or not they had received an Option Grid. However, there is evidence that many patients feel that they are also not able to participate in decision making rather than express that they do not wish to participate. 27 Using a short decision tool such as an Option Grid is one way of transferring the knowledge ownership and making it explicit to the patient that there is a choice to be made. Clinicians should be mindful that encounter tools such as Option Grids may go some way to empowering and including patients who are disadvantaged by language, low educational attainment and age, but some patients who are most disadvantaged may require additional support in the consultation. This may include spending additional time reading the Option Grid aloud to patients, or encouraging interpreters to do so, or possibly through providing summarized information about treatment options through other visual methods such as simplified images.
We found that many patients wanted other individuals to be involved in treatment decisions. These individuals included primary care physicians, but more commonly patients stated they wanted to discuss their options and preferences with family members. By involving others, patients were drawing on a form of "distributed health literacy" 34 in which they work collaboratively utilizing the skills of others to help understand health information. This notion also provides support that the Option Grids can be seen not just as a standalone tool for facilitating a conversation within the consultation, but also a tool to be used with the wider family and patient network including primary care physicians.

| Conclusions
The findings of this qualitative study, amongst an older patient population who have lower than average health literacy, support the development of Option Grids as a tool to facilitate SDM for patients with osteoarthritis of the knee. The implementation of Option Grids should be considered within routine consultations as patients are supportive of their use.