The association between patients' preferred treatment after the use of a patient decision aid and their choice of eventual treatment

Abstract Objective To investigate the association between patients' preferred treatment and eventual treatment. Second, to compare patients with surgical treatment to watchful waiting in order to identify predictive factors for surgery. Methods A single‐centre retrospective study was performed between December 2015 and August 2018. Patients (≥18 years) who used a patient decision aid (PDA) for gallstones or inguinal hernia were included. After their first surgical consultation, patients received access to an online PDA. The patients' preferred treatment after the PDA was compared with their choice of eventual treatment. Multivariable regression analyses were performed for predictive factors for surgery. Results In total, 567 patients with gallstones and 585 patients with an inguinal hernia were included. Of the patients with gallstones, 121 (21%) preferred watchful waiting, 367 (65%) preferred surgery, and 79 (14%) were not sure. The patients' preferred treatment was performed in 85.9%. Frequent pain attacks (OR 2.1, 95% CI 1.1‐3.9, P = .020) and preference for surgery (OR 4.4, 95% CI 1.9‐10.1, P = .001) independently predicted surgery. Of the patients with an inguinal hernia, 77 (13.2%) preferred watchful waiting, 452 (78.8%) preferred surgery, and 56 (9.6%) were not sure. The patients' preferred treatment was performed in 86.0%. The preference for surgery (OR 5.2, 95% CI 2.5‐10.6, P < .001) independently predicted surgery and worry about complications predicted avoidance of surgery (OR 0.5, 95% CI 0.2‐1.0, P = .037). Conclusion This study, reflecting current clinical care, shows that patients' preferred treatment after using a PDA matches their eventual treatment choice in 86% of patients with gallstones or an inguinal hernia. In these patients, symptoms and patients' preference for surgery independently predicts eventual choice of surgery.


| INTRODUC TI ON
An increasing number of patients want to be involved when a decision needs to be made between treatment options, commonly called shared decision making (SDM). 1 As time is a limiting factor for SDM in the consulting room, SDM is frequently facilitated by (online) patient decision aids (PDAs). PDAs present comparative information about advantages and disadvantages of available options and evaluate personal values and preferences of the patient. [2][3][4] PDAs are proven to be effective in improving knowledge, reducing decisional conflict and, moreover, changing patients' preferred treatment. 5 Our group recently showed that PDAs reduce operation rates in patients with gallstones or an inguinal hernia. 6 This change may be caused by the impact of the PDA on patients' preferences and cause a decision shift. 7 Patients' preference on treatment is influenced by several factors. First, the preference is influenced by the patients' symptoms and their concern about the course of the condition. Second, the physician's advice is important in development of patients' preferred treatment. [8][9][10] Third, treatment options, risks associated with each option and expectations of results can influence patients' preference. 7,11,12 A PDA additionally informs a patient on treatment options and associated risks, but also explores patients' values to determine what is most important for the individual patient. Current PDAs consist of a personal value clarification exercise, which helps patients to identify their own values and find the treatment option most consistent with their preference. However, it is unknown which personal values are important in the preference for surgery. 13 Knowledge concerning factors influencing patients' preferences and treatment decisions are important to improve SDM. 14 Gallbladder surgery and hernia repair are both, to a degree, considered to be examples of preference-sensitive care. 15 International guidelines support watchful waiting in selected patients if symptoms are mild and the condition does not show signs of potential complications. [16][17][18] Cholecystitis or biliary pancreatitis are potential complications of gallstones, while in case of an inguinal hernia, a complicated course may result in a bowel incarceration. Nevertheless, a non-operative treatment is important to consider, especially since it is known that cholecystectomy is a sub-optimal solution to treat pain in patients with gallstones and abdominal symptoms and the long-term outcomes of hernia repair are variable. [19][20][21] These variations in outcomes make it essential that patients are invited to participate in treatment decisions based on information about potential complications, surgical morbidity or the outcome after conservative management.
Current research mostly focuses on the patients' preferences, but information about the association between preferences and their choice of eventual treatment is not yet reported. 7 Therefore, the aim of this study was to investigate the association between the preferred treatment option after PDA usage and the eventual choice of treatment. Secondly, we identified clinical factors and patient values that predicted a decision to have surgery.

| Study design and study population
A single-centre retrospective study was performed in a community hospital in the Netherlands. Patients were eligible for participation in the study if they were ≥18 years of age and used a PDA between December 2015 and August 2018. All patients with an emergency indication for surgery were excluded.
PDAs for patients with gallstones or an inguinal hernia were integrated into the standard workflow at the surgical outpatient clinic in December 2015. Physicians were trained in SDM to support the implementation of PDAs. This training covered the use of the PDA with a patient. Moreover, a workgroup was formed, consisting of at least a medical doctor, a project manager (from the hospital), a nurse of the outpatient clinic and a department manager of the outpatient clinic. Objectives for the use of the PDAs were set, as well as the care process for patients. Workshops were organized to inspire about SDM, to explain how the PDAs work and to train communication skills to empower patients to participate in the decision-making process, assess patients' preferences and health values.

| Outcome and data collection
To obtain information about the association between preference and eventual choice of treatment after PDAs, we compared the pa-

| Statistical analysis
Patients who had not completed the PDA, whose data about their eventual choice of treatment were not available from the hospital register, and patients who had completed multiple PDAs for the same condition were excluded from the analysis. The data of patients with gallstones and inguinal hernia were analysed and pre- To identify predictive factors for performing surgery, first univariate logistic regression analyses were performed with their eventual choice of treatment (watchful waiting/surgery) as dependent variable and patients' preferred treatment (watchful waiting, not sure and surgery), age, sex, disease burden and personal values, as independent variables. These were all included in the multivariable model. The outcomes of the univariable and multivariable analyses were presented as an odds ratio (OR) and 95% confidence interval (CI).
Associations with a P-value <.05 were considered statistically significant. All missing values were considered to be at random and were excluded from analyses. Analyses were performed using SPSS statistics version 25.0 (IBM).
Baseline characteristics are described in Table 1.  (Table 1). In total, 395 patients (69.7%) underwent a gallbladder removal. Figure 2A shows the association between the patients' preferred treatment at the end of the PDA and their eventual choice of treatment.

| Patients' preferred vs eventual choice of treatment
Of the patients with an inguinal hernia, 77 (13.2%) preferred watchful waiting, 56 of patients (9.6%) were not sure, and 452 (78.8%) preferred surgery. In total, 147 patients (25.1%) were treated conservatively and 438 patients (74.9%) had surgery. Figure 2B shows the association between the patients' preferred

| Certainty about preferred treatment
For both conditions, patients who preferred surgery were more certain about their preference compared to patients who preferred watchful waiting. In patients with gallstones, median scores

| Predictive factors for surgery
In Tables 2 and 3, patients with a surgery were compared with patients who underwent a watchful waiting strategy. Patients with gallstones who underwent surgery were younger compared to patients who waited (50.1 ± 15.1 vs 55.6 ± 14.9, P < .001). Surgical patients were characterized by reporting more pain attacks, major complaints and less concern about surgery compared to patients with a watchful waiting policy (P < .001 for each comparison). In multivariable analyses, only frequent pain attacks (OR 2.1, 95% CI 1.1-3.9, P = .020) and preference for surgery (OR 4.4, 95% CI 1.9-10.1, P = .001) were independently predictive factors for surgery.
In patients with an inguinal hernia, similar outcomes were found.
Surgically treated patients were younger compared to patients who waited (56.8 ± 14.4 vs 61.1 ± 14.4, P = .002). Surgical patients were characterized by more discomfort, less concern about surgery, less concern about risks of surgery and more concern about bowel incarceration, compared to patients with watchful waiting (P < .001 for all comparisons). In multivariable analyses, the preference for surgery (OR 5.2, 95% CI 2.5-10.6, P < .001) independently predicted surgery and worry about complications of surgery was negatively associated with surgery (OR 0.5, 95% CI 0.2-1.0, P = .037). The majority of patients with gallstones or inguinal hernia prefer surgery after using a PDA. It could well be that patients' preference for surgery contributes to the nationwide operation rates in the Netherlands above 70% for both conditions. 19 Despite the finding that watchful waiting is a good alternative in selected patients with symptomatic gallstones or an inguinal hernia, 16,23,24 it looks like most patients prefer a quick and definitive solution for their symptoms. [25][26][27][28] That is in line with our finding that the major-

CO N FLI C T O F I NTE R E S T S
All authors have completed the ICMJE uniform disclosure form at and declare the following: this research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Glyn Elwyn has edited and pub-

E TH I C A L A PPROVA L
The study was approved by the appropriate institution and/or national research ethics committee (registration number 2018-4587).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author after reasonable request.