Methods
Design
A two-arm prospective RCT was conducted. Study approval was received from The Ottawa Hospital Research Ethics Board (# 2006724-01H) and the trial was registered (ClinicalTrials.gov, NCT00743951).
Setting
Patients with osteoarthritis of the knee were recruited from an orthopaedic intake clinic, within a Canadian tertiary hospital. The sports medicine physician assessed surgical eligibility using the 7-item Western Wait List Hip Knee Priority Tool (HKPT) mapped onto the three criteria for total knee arthroplasty according to the clinical practice guidelines (moderate to severe pain, moderate to severe functional limitations, and abnormal radiographic findings). Although, the priority tool was originally developed and validated as a transparent and fair approach for prioritizing patients on waitlists, it is used in this clinical setting as a standardized assessment tool applied to all patients. From April 2006 to March 2007, 47% of patients at this clinic were assessed to have milder osteoarthritis and were directed back to their referring physician with suggestions for conservative management. The others were deemed eligible for surgical consideration for knee osteoarthritis.
Participants
Eligible knee osteoarthritis patients were those with access to a television with a VCR or DVD player. Those with inflammatory arthritis, previous TJA, uncorrected hearing or visual impairment, or unable to read, or understand English, were excluded.
Interventions
Experimental Group. The PtDA, developed by the Informed Medical Decisions Foundation and distributed through Health Dialog, is entitled Treatment Choices for Knee Osteoarthritis. It consists of a 50-minute video and accompanying booklet that provides information on various treatment options for knee osteoarthritis, including lifestyle changes, non-drug treatments, pain medication, injections, complementary therapies, and surgery. A description of the options, probabilities of benefits and harms for each option, and video-clips of patient experiences allows patients to clarify their values associated with outcomes of options. According to the International Patient Decision Aid Standards, this PtDA meets most criteria for content (12 of 15), development process (8 of 9), and effectiveness (1 of 2). For more details on the IPDAS score card and the PtDA go to: http://decisionaid.ohri.ca/AZsumm.php?ID=1191. Patients received a questionnaire, formatted as user-friendly booklet, assessing their knowledge, values, preferred treatment choice, decisional conflict, and comments or questions. These results were combined with the patients' clinical assessment findings to create a one-page preference report for the surgeon (see Figure 1).
(Enlarge Image)
Figure 1.
Summary report for surgeons [26].
Control Group. Patients in the usual education group received a standard information booklet prepared by the participating hospital for all patients undergoing joint replacement surgery. Information included preparation for surgery, recovery after surgery, and discharge plans. There was no information on benefits and harms of surgery or alterative options that could be used for decision making. Surgeons for patients in the control group received a half-page summary of patients' clinical assessment findings only.
Procedures
Eligible patients met with a research assistant who collected consent to participate in the study and baseline data. Baseline data included the Western Ontario McMaster Universities Osteoarthritis Index (WOMAC) measuring the patient's perceptions of knee pain, stiffness and function, the HKPT completed by the assessing physician, and demographic information. This data was used to populate the surgeon's clinical summary report (upper half of Figure 1).
After baseline data collection, patients were allocated to the PtDA intervention or usual education. The allocation schedule was computer-generated centrally by a statistician using a permuted block design with randomly varying block lengths of 4, 6, or 8. Allocations were concealed in numbered opaque sealed envelopes until after signed consent was obtained. Once allocated, patients were instructed to review their respective information (i.e., PtDA plus usual education booklet or usual education booklet only) at home and complete the accompanying questionnaire. Patients were not informed of the intervention characteristics. Although the research assistant was not blinded to group allocation, study outcomes for effectiveness were objective and obtained from clinic data (e.g. date of surgery or waitlist status).
Within two weeks from recruitment, the research assistant telephoned participants to obtain their answers to the questionnaire. This information was added to the surgeon's clinical summary report to create the surgeon preference report (Figure 1). Participants were contacted by telephone one year after recruitment to determine whether they had seen the orthopaedic surgeon, and if so, whether they had chosen surgery or alternative non-surgical options. Dates for surgeon consultation and surgery were collected from the hospital health information system.
Outcomes
The main feasibility outcomes were recruitment and questionnaire completion rates. Feasibility targets were: a) >70% study enrollment of eligible patients; b) >90% of patients completing questionnaires at home prior to surgeon consultation; and c) <10% missing data. Preliminary effectiveness outcomes were wait times, decision quality, preparation for decision making, decisional conflict and patient feedback on the PtDA.
Preliminary effectiveness outcomes were assessed to inform a future larger scale RCT. Wait times were calculated based on the number of days from screening to definitive choices (i.e., date of surgery or date of decision to decline the surgery either explicitly stated or based on date appointment was cancelled without rebooking). Decision quality was defined as the extent to which patients' decisions were informed and values congruent with their choice. Decision quality was deemed sufficient if a patient scored ≥66% on the knowledge test and if their predicted probability of surgery based on values corresponded with their actual choice. A score of 66% was chosen because the mean score for patients who had completed the knowledge test after viewing the PtDA was 68% and it is consistent with knowledge scores in trials of PtDAs.
Outcome Measurement Instruments
Hip-Knee Osteoarthritis Decision Quality Instrument. Patients' knowledge was assessed using four multiple choice questions (i.e., osteoarthritis progress over time, need for revision joint replacement, proportion of patients with reduced pain, and length of time for recovery) from the Hip-Knee Osteoarthritis Decision Quality Instrument. Knowledge scores were previously shown to be reproducible and to discriminate between those exposed to PtDAs and controls. Patients' values were measured by asking patients to rate the personal importance of the benefits and harms of outcomes for 6 items (e.g. relief of pain) on a 10-point rating scale with 1 indicating low importance and 10 indicating high importance. In a previous study, those who valued pain relief and return to normal activities were more likely to choose surgery while those who valued surgery avoidance were less likely to choose surgery. The validity of these items have previously been demonstrated: patients whose treatment choice was concordant with their values felt more confident and had less regret with their decision.
Decisional Conflict Scale. The SURE tool, a 4-item version of the Decisional Conflict Scale, was used to assess patients' perception of feeling sure, informed, supported, and clear about what mattered most. In patients considering treatment options, this tool was previously shown to have adequate internal consistency with Kuder-Richardson 20 coefficient of 0.7 and significant correlation between the Decisional Conflict Scale and SURE scores.
Preparation for Decision Making. Four of the 10 items on the Preparation for Decision Making Scale were used to determine patients' perceptions of the decision making process. Four items were chosen because of their relevance to surgical decisions and they discriminated between patients prepared for decision making with PtDAs and those who were not. The items include recognition that a decision needs to be made (discrimination value 2.12), knowledge that the best choice depends on what matters most to the patient (3.39), level of decision making involvement desired by the patient (2.61), and patient preparedness for discussion with the surgeon (3.08). This instrument reports good internal consistency (>0.91) and excellent item discrimination (range 2.12 – 3.80).
Data Management & Statistical Analysis
All data were entered twice into Microsoft Excel, verified for accuracy, and analyzed using SAS v. 9.1. Feasibility outcomes were summarized using descriptive statistics. Dichotomous effectiveness outcomes were describing using proportions with 95% confidence intervals and differences between arms were tested for statistical significance using chi-squared tests. Differences in the between groups median wait list times were described using Kaplan-Meier survival curves with 95% confidence intervals and were assessed for statistical significance using the log-rank test. Patients were censored at the end of the study, at the time of death, or loss to follow-up.
The mean scores on the knowledge test (calculated as the percentage of correct answers) were compared between the two groups using the two-sample t-test. The match between the patients' choice and their values for benefits/risks was calculated as a dichotomous measure. The predicted probability of surgery was calculated for each patient using a logistic regression equation derived from three items assessing the patient's values. The equation was [1 + exp (–S)] where S = -0.3384 + 0.3869 × Value Q6 – 0.6111 × Value Q7 + 0.1933 × Value Q8 (Q6 relieve pain; Q7 avoid surgery; Q8 return to usual activities). The predicted probabilities were rounded to 0 or 1 (0 = no surgery and 1 = surgery) to determine the predicted choice based on the patient's values. Mean scores on the Decisional Conflict Scale and Preparation for Decision Making scale were compared using the two-sample t-test. All tests were carried out at the two-sided 5% level of significance. Feedback from patients about the PtDA was analyzed descriptively.