A Simple Goal Elicitation Tool Improves Shared Decision Making In Outpatient Orthopaedic Surgery: A Randomized Controlled Trial

Abstract

Introduction

Shared decision making involves educating the patient, eliciting their goals, and collaborating on a decision for treatment. Goal elicitation is challenging for physicians as previous research has shown that patients do not bring up their goals on their own. Failure to properly elicit patient goals leads to increased patient misconceptions and decisional conflict. We performed a randomized controlled trial to test the efficacy of a simple goal elicitation tool in improving patient involvement in decision making.

Methods

We conducted a randomized, single blind study of new patients presenting to a single, outpatient surgical center. Prior to their consultation, the intervention group received a demographics questionnaire and a goal elicitation worksheet. The control group received a demographics questionnaire only. After the consultation, both groups were asked to complete the Perceived Involvement in Care Scale (PICS) survey. We compared the mean PICS scores for the intervention and control groups using a nonparametric Mann Whitney Wilcoxon test. Secondary analysis included a a qualitative content analysis of the patient goals.

Results

Our final cohort consisted of 96 patients (46 intervention, 50 control). Both groups were similar in terms of demographic composition. The intervention group had a significantly higher mean PICS score compared to the control group (9.04 (SD 2.15) vs. 7.54 (SD 2.27), p < 0.01). 39% of patient goals were focused on receiving a diagnosis or treatment, while 21% of patients wanted to receive education regarding their illness or their treatment options.

Discussion

A single-step goal elicitation tool was effective in improving patient perceived involvement in their care. This tool can be efficiently implemented in both academic and non-academic settings.

INTRODUCTION

Healthcare delivery is transitioning towards patient-centered decisions based on their values and preferences through a collaborative decision making process with the physician (1). One such model is shared decision making, which involves educating the patient about their options, eliciting their goals and preferences for treatment, and making a healthcare decision that reflects their values and preferences (2,3). Studies have shown that when patients experience shared decision making and feel more involved in their care, they have increased trust in their physicians, increased confidence in their decision, and improved self-efficacy, satisfaction, and perceived quality of care (4–9).

Implementation of shared decision making can take several forms, including the use of decision aids (e.g. brochures or handouts for education), preference elicitation tools, or goal elicitation tools (10–13). The third of these options, goal elicitation tools, have the potential to inform goal-directed treatment decisions and outcomes assessment. Past research shows that goals for treatment and recovery are often unexpressed by the patient or unsolicited by the physician (14,15). However, because decision aids and prompts increase patient participation and question asking (16–18), there is potential for a tool that triggers a discussion on goals. While physicians may feel they can approximate their patients goals informally, prior studies demonstrate that physicians and health systems tend to misjudge what patients believe is important for their treatment and recovery (19–21). Unfortunately, existing goal elicitation tools can be time-consuming and resource draining, particularly for physicians outside of academic centers (22). Additionally, these tools are often condition or specialty specific, making general implementation difficult (23,24). A simple, time-efficient, generalizable method of eliciting patient goals may address these barriers and engage patients and improve shared decision making.

Therefore, we studied a novel, single-step goal elicitation tool as a method of improving shared decision making in the clinical setting. We conducted a randomized, controlled study to test the null hypothesis that goal elicitation does not impact patient perceived involvement in care. We also completed a qualitative content analysis of the goals listed by patients in the intervention group.

METHODS

Patients

We completed an institutional review board-approved, prospective, single-blinded, randomized, controlled trial of patients seeking care at a multispecialty orthopaedic surgery clinic from July 2018 to November 2018. Seven orthopaedic surgeons in subspecialties including hand/upper extremity, spine, total joint arthroplasty, sports, foot and ankle, and trauma, who are part of an ongoing collaboration, participated in the study. The surgeons were informed that they would be taking part in a shared decision making study. They were blinded to the randomization of their patients as well as the contents and design of the goal elicitation questionnaire (including their patients’ specific goals) and outcome measure. New patients were approached for inclusion in this study. Patients were considered eligible for enrollment if they were presenting to the clinic for the first time, were 18 years or older, English literate, and able to provide informed consent. The trial was registered at ClinicalTrials.gov (Trial number: NCT-0364S135).

Randomization and Intervention

Patients were asked to participate in a study on shared decision making. Patients who were consented to participate in this study were randomized to either the control group or the intervention group. Randomization was carried out prior to the start of the study using block randomization in permuted blocks in groups of 2 and 4. The randomization list was kept in an Excel spreadsheet that could only be accesses by the investigator enrolling patients. We were unable to blind the randomization from the investigators that distributed the questionnaires to the patients.

Prior to seeing their physician, the intervention group received a demographics questionnaire as well as a goal elicitation tool. Demographics included were age, sex, race, education, income and relationship status. The goal elicitation tool asked the patient to list three goals for their consultation that day. The control group received the same demographics questionnaire but did not receive a goal elicitation tool. In both the intervention and control groups, their goal elicitation tool and/or demographic questionnaire were taken from the patient prior to the physician entering the room. The patients were given no instruction as to whether they should talk about their goals with their physician. This was done to minimize bias as well as isolate the effect of the goal elicitation tool. The physician was not given any information of the substance of the goals and was not told into which study group their patient was randomized to minimize the potential for contamination. Following their consultation, patients in both groups were asked to complete the Perceived Involvement in Care Scale (PICS), a tool commonly used to assess the degree of shared decision making that has occurred (25).

Outcome Measure

The primary outcome measure was the patient’s perceived involvement directly after the patient’s consultation via the PICS questionnaire. The PICS questionnaire is a validated and widely-implemented tool used to assess a patient’s perceived level of involvement in their decision making (25). The questionnaire consists of 13 “yes or no” questions that ask the patient about different aspects of their consultation that indicate the degree of involvement in decision-making. The 13 questions are separated into three subsets that convey different aspects of patient involvement: physician facilitation of patient involvement (Subset A, five questions); information exchange (Subset B, four questions); and patient participation in the decision (Subset C, four questions). All questions are “yes/no.” A “yes” equals 1 point; a “no” equals 0 points. The maximum score is 13 points: 5 for Subset A; 4 for Subset B; and 4 for Subset C. A higher score indicates higher perceived involvement.

Statistical Analysis

An a priori sample size estimate based on effect size (Cohen’s d) was performed using pilot data including 20 patients (10 intervention, 10 control). Cohen’s d was calculated using group means and SDs from pilot data. Based on a Cohen’s d of 0.97, we determined that 48 patients (24 control, 24 intervention) were needed to provide 90% power to detect a significant difference between the mean PICS scores of the control and intervention groups, using a two-tailed T-test (alpha = 0.05). The patients from pilot data collection period were not included in the final analysis. Patients with blank questionnaires were excluded and were not included in an intention-to-treat analysis because they had not received any intervention. Patients that provided less than 3 goals were included in an intention-to-treat analysis.

The intervention and control group were initially analyzed for differences among groups based on demographic characteristics. After assessing the distribution of the PICS scores in each cohort, we used nonparametric Fisher exact tests to compare the groups when using categorical variables (e.g. insurance type, education status), and Mann-Whitney-Wilcoxon tests when analyzing continuous variables (e.g. age). We used Mann-Whitney-Wilcoxon tests to test the null hypothesis that goal elicitation did not impact PICS score. All statistical analysis were performed with SAS (SAS University Edition, Copyright © 2019, SAS Institute Inc., Cary, NC, USA). We considered a change of 1 in PICS score clinically meaningful. The minimum clinically important difference for PICS has not been established. We based our estimate on the knowledge that improvement in patient involvement can lead to clinically relevant outcome improvement (9,25).

Qualitative content analysis

Patient goals from the intervention group were listed individually in an Excel spreadsheet. Two reviewers independently open-coded the individual goals by assigned a short phrase that described the goal. The two reviewers then discussed and resolved any discrepancies in coding. The open codes were then grouped into subcodes and the subcodes were grouped into codes by the two reviewers. The codes categorized patient goals as one the following: 1) Care experience, 2) Education and Information, 3) Pain and Symptoms, 4) Diagnosis and Treatment, and 5) Recovery. We calculated frequencies of subcode and code categories.

Funding

Financial support for this study was provided in part by a grant from the National Institutes of Health (Mentored Patient-Oriented Research Career Development Award [K23AR073307]).

RESULTS

Randomized trial results

One-hundred twenty-six (126) patients were assessed for eligibility (Figure 1). Ten patients were not enrolled due to failing to meet inclusion criteria (9 non-English speakers, 1 under 18 years). Eleven patients declined enrollment. The remaining 105 patients were randomized into control and intervention groups (54 control, 51 intervention). At time of analysis, an additional 9 patients were removed due to incomplete questionnaires (4 control with blank demographic questionnaires, 5 intervention with blank demographic questionnaires and/or no listed goals). Patients were not omitted from analysis if they listed less than 3 goals. Thirty-two (33) patients wrote 3 goals, 9 wrote 2 goals, and 4 wrote 1 goal. No patients were lost to follow-up as all data was collected at a single office visit.

Figure 1.

Flowchart of study enrollment and data allocation and analysis

Our final cohort included 96 patients (50 control, 46 intervention). The cohorts were similar in terms of demographic characteristics (Table 1). Patients of both cohorts presented with a wide spectrum of orthopaedic conditions (Table 2). The intervention group had a significantly higher mean PICS score compared to the control group (9.04 (SD 2.15, CI 95% [8.42–9.55]) vs. 7.54 (SD 2.27, CI 95% [6.91–8.17]), Mann-Whitney-Wilcoxon results: p < 0.01, z = −3.37, effect size r = 0.34).

Table 1:

Demographic Characteristics

		Control (n = 50)	Intervention (n = 46)	P Value
Age		51.76 (SD: 19.85)	54.11 (SD: 17.40)	0.236
Sex				0.567
	Male	23 (46.9%)	20 (46.5%)
	Female	26 (53.1%)	23 (53.5%)
Race				0.
	White	32 (64.0%)	28 (65.1%)
	African American	1 (2.0%)	2 (4.65%)
	Asian	11 (22.0%)	11 (25.6%)
	Hispanic	5 (10.0%)	1 (2.33%)
	Other	1 (2.0%)	1 (2.33%)
Income Bracket				0.204
	< 50K	11 (23.4%)	11 (26.8%)
	50 – 100K	12 (25.5%)	5 (12.2%)
	100–150K	4 (8.51%)	7 (17.1%)
	150 – 200K	1 (2.13%)	5 (12.2%)
	200 – 250K	3 (6.38%)	3 (7.32%)
	>250K	16 (34.0%)	10 (24.4%)
Employment Status				0.696
	Full Time, Part Time	26 (53.1%)	22 (47.8%)
	Retired	12 (24.5%)	15 (32.6%)
	No work outside home, disabled, unemployed	7 (14.3%)	4 (8.7%)
	Student	4 (8.16%)	5 (10.9%)
Relationship Status				0.2243
	Married, Domestic Partner	31 (67.4%)	34 (81.0%)
	Single	15 (32.6%)	8 (19.0%)
Insurance				0.927
	Medicaid	7 (14.0%)	6 (13.0%)
	Medicare	11 (22.0%)	13 (28.3%)
	Private	28 (56.0%)	24 (52.2%)
	Other	4 (8.0%)	3 (6.5%)
Education Status				0.216
	Some High School, High School, or Trade School	8 (16.0%)	13 (28.2%)
	Higher Education (Bachelors, Masters, Doctorate)	42 (84.0%)	33 (71.8%)

FOOTER: Used Mann Whitney Wilcoxon for continuous variables, Fisher exact for categorical variables

Table 2:

Orthopaedic conditions of patient cohort

		Intervention (n = 46)	Control (n = 50)	Total (n = 96)
Arthritis	Hip and Knee	8	10	18
	Other (Thumb, Ankle)	2	4	6
Ligament/Tendon/Soft Tissue	Hand	8	5	13
	Hip and Knee	6	5	11
	Foot	4	4	8
Fractures	Upper Extremity Fracture	4	6	10
	Lower Extremity Fracture	4	7	11
Other	Back Pain/Scoliosis/Stenosis	3	6	9
	Hip Impingement	4	2	6
	Bunion/Hammer toe	3	1	4

Qualitative results

A plurality of patient goals was related to diagnosis and treatment (39%). Of those goals, one-third (1/3^rd) each focused on either receiving a diagnosis or receiving treatment, while the remaining goals were related to developing a treatment plan or delaying or avoiding surgery. Additionally, approximately one-fifth (20%) of goals were related to receiving information or education about their condition or their options for treatment. A complete breakdown of the subcoding and coding of goals is included in Table 3.

Table 3:

Qualitative Content Analysis of Patient Goals

	N	Frequency (of total goals)
Care experience	6	5%
Positive Doctor/Patient relationship	3	2%
Positive overall experience	3	2%
Education/Information	26	21%
Information regarding condition	9	7%
Information regarding treatment options and risks	16	13%
Pain and symptoms	21	17%
Pain relief	14	12%
Symptom relief	4	3%
Pain avoidance	2	2%
Diagnosis/Treatment	47	39%
Receive diagnosis	16	13%
Receive treatment	16	13%
Plan for treatment	11	9%
Delay/avoid surgery	4	3%
Recovery	21	17%
Return to activities of daily life	5	4%
Plan for/time to recovery	7	6%
Heal/Improve current status	5	4%
TOTAL	121

DISCUSSION

Goal elicitation is a necessary step in the shared decision making process. However, implementation of goal elicitation remains understudied and without a specific structure or process in orthopaedic surgery (22). In this randomized, controlled trial, our single-step goal elicitation tool significantly improved patient perceived involvement in care, which is a critical aspect of shared decision making.

Prior attempts to design goal elicitation tools have shown their effectiveness while also highlighting the difficulties of their implementation. A review of 11 goal elicitation and goal setting instruments found that while the majority of the instruments were effective in facilitating patient-centered goal elicitation, each instrument had unique drawbacks in ease of implementation (22). For example, the Goal Attainment Setting (GAS) tool was found to be successful in goal setting, goal negotiation and appraisal/feedback of goals for patients (26). However, the GAS tool was difficult to use with elderly or cognitively impaired patients and was time consuming to implement (27,28). Other studies have used conjoint analyses to assess goals of care, and while these tools are effective in improving patient knowledge and eliciting patient goals, they require some expertise in development (10,11,29). Our single-step goal elicitation tool is simple to implement, requires minimal resources and time, and can be incorporated into the care pathway of any orthopaedic condition to improve patient perceived involvement and facilitate the shared decision making process. Additionally, because we did not provide any instructions to the patients as to whether they should discuss their goals with their physician, nor provide physicians with these goals to promote their discussion, the patients experienced an increase in perceived involvement regardless of whether the goals were discussed or not. This may warrant further investigation into how the goal elicitation tool affects patient confidence and empowerment.

Improving patient perceived involvement in shared decision-making through a goal licitation process allows physicians, patients, and health systems to reap the benefits purveyed by increased shared decision-making. A Cochrane review on the effects of shared decision-making found that patients who experienced increased involvement in their decision had decreased decisional conflict about their options and were more satisfied with their treatment and their physician in their recovery period (30). In this same review, improved shared decision-making also led to decreased usage of discretionary screening and diagnostic tests as well as greater non operative treatment over major elective or invasive surgery (30). There is also evidence to suggest that improved shared decision-making through goal elicitation leads to increased adherence to medication and drug therapies (31,32). By implementing a goal elicitation tool, physicians can improve their patients’ clinical experience and may find its use leads to decreased resource use without negatively affecting quality of care – a goal of value-based health delivery models. Future studies should examine the effect of the goal elicitation tool on these factors directly. For example, does use of a goal elicitation tool lead to improved health?

In this study, we did not provide physicians with copies of their patients’ goals to assess the act of writing goals out (and not their review and discussion). There is the potential that active discussion of written goals can improve PICS scores even more. Physicians can refer to the qualitative results of this study or implement our goal elicitation tool in their practice to gain some insight into what patients are expecting from a clinical visit.

The results of this study should be considered in light of its limitations. The patient sample in the study is from a single, academic institution in a suburban setting. Further work to validate the results in other patient populations is needed. Additionally, we chose a change in PICS score of 1 (out of 13) as clinically significant difference based on the abundance of evidence that supports that more involved patients make more informed, patient-centered decisions with positive health outcomes. Although no minimally clinically important difference for the PICS has been established, the evidence supports that any improvement in involving patients is clinically relevant. Additionally, in order to maintain the blinding status of the physicians involved in this study, we did not provide instructions to the patients in the intervention group as whether they should discuss their written goals. In our study, we did not supply physicians with their patients’ goals. If physicians implement goal elicitation in the clinical setting, they will most likely review and address patient goals, rather than just elicit them. This may have limited the potential for greater perceived involvement; however, this also demonstrated that the process of goal elicitation alone was enough to improve perceived involvement. Some may contend that patients may have more than 3 goals for their consultation and that by limiting the number of goals, we are suppressing their opportunity to participate in their care. We acknowledge that patients may have additional goals; however, our results demonstrate that asking patients for up to 3 goals is sufficient in triggering a positive increase in participation while maintaining a realistic implementation process.

Additionally, the spectrum of injuries and conditions in our patient population may make the results of our study less applicable to a specific orthopaedic specialty. However, we believe our diversity of conditions improves our study’s applicability to orthopaedic surgery and musculoskeletal conditions in general. Additionally, some specialists may feel that certain injuries (such as displaced intraarticular fractures) do not have multiple options for treatment and therefore do not lend themselves to a goal-directed discussion and shared decision and that this may have influenced our results. However, understanding a patient’s goals and fostering collaborative discussion with the patient may help patients feel involved in their care even in situations without several clinical options. Furthermore, we compared the mean PICS scores between patients with trauma-related and non-traumatic injuries and found no significant difference (Table 4).

Table 4:

Orthopaedic condition type: mean PICS scores

Condition type	Mean PICS	Mann-Whitney-Wilcoxon test
Trauma-related (n = 25)	7.8 (SD 2.2), CI 95% [7.52–8.08]	p = 0.49
Non-traumatic (n = 71)	8.4 (SD 2.4), CI 95% [7.88–8.97]

We present a simple, single-step, goal elicitation tool that improves patient perceived involvement among patients with orthopaedic conditions. This tool can be easily implemented within the workflow of a busy clinic and should be accessible to physicians in all clinical settings. Further research can determine this goal elicitation tool’s effectiveness on other patient reported outcomes and patient health, such as trust in physician, confidence in decisions, and treatment adherence. Our qualitative content analysis of the goals in this study can serve as a starting point for further analysis of patient goals and understanding which aspects of care are most important to the patient.