Phase 2: Assessment of effectiveness
Study design and setting
A parallel, multicenter, superiority RCT will be carry out in hospitals and primary and specialty care centers of Tenerife (Spain). Patients will be randomly allocated to intervention (web-based PtDA) or control group (general information about OA).
Eligibility criteria
Adult patients with a diagnosis of hip OA will be included in the trial. Exclusion criteria will be: previous hip arthroplasty, patients with terminal illnesses, non-fluent Spanish language, be participating in other trials, or mental or sensory alterations that impede or hinder the response to the measurement instruments or do not have the capacity to make decisions.
Randomisation and allocation concealment
Computer-based randomization will be performed centrally by an independent researcher. Randomization will be stratified by patient inclusion in waiting list for arthroplasty (yes/no). Sealed envelopes with each patient’ allocation will be prepared, which will be opened only after the patient sign the informed consent for participation in the study. The nature of the intervention makes impossible to blind patients and researchers to group allocation.
Recruitment and procedure
Eligible patients will be identified through the review of participating centres’ clinical charts. A member of the research team will contact them by telephone to explain the study and invite them to participate. Those who agree will complete the baseline assessment in the same phone call, and an appointment will be scheduled with a researcher at their referral primary healthcare centre or hospital. There, patients will be asked to read the project information sheet, which will explain in detail what their participation will consist of, and to sign the informed consent. Then, patients allocated to the intervention group will review the PtDA accompanied by a researcher, who will give support in navigation if necessary. Then, they will complete the questionnaires assessing the outcome measures, in the same web interface. Patients in the control group will follow the same procedure, but instead of the PtDA they will review general information about OA of the hip, knee, ankle and foot, information included in a brochure delivered in primary care centers of the Canary Islands Health Service, which will be also presented in a web format.
Outcome measures
The primary outcome of this study will be decisional conflict regarding the treatment for OA, assessed with the Decisional Conflict Scale (DCS) [18], immediately after reviewing the PtDA. This is a 16-item self-reported scale widely used in PtDA research. It has 5 subscales: feeling informed, values clarity, feeling supported, uncertainty and effectiveness. Scores are transformed to a 0–100 scale, with higher scores indicating more conflict. In previous studies in other health conditions, we have obtained good internal consistency values (0.88–0.90) [19–21]. The secondary outcomes will include:
- Knowledge of the disease and treatments: assessed with a 7-item questionnaire previously used in studies of PtDA in OA, specifically adapted for hip OA [22]. Items have a “true/false/don’t know” format. The percentage of correct responses represents the total score.
- Values about characteristics of treatments: patients will be asked to rate in a 0–10 scale the importance they attribute to different characteristics of treatment: mode of administration, improvement in pain and function, risk of mild and serious adverse effects, time until symptoms’ improvement, duration of benefits, and in the case of arthroplasty, period of rehabilitation and risk of needing surgery revision.
- Treatment preference (physiotherapy, medication, intra-articular injections, arthroplasty, not sure).
- Intention to undergo the preferred treatment: assessed with one item ranging 0 (not sure) to 5 (completely sure).
- Concordance between values and intention to undergo arthroplasty vs. conservative treatments: binary variable (yes/no) derived from the association between values and intention to undergo arthroplasty (see statistical analysis).
- Decision quality about arthroplasty: binary variable (yes/no) defined as a combination of adequate knowledge (≥60% of correct responses) and concordance.
- Satisfaction with the decision making process: assessed with the 12-item questionnaire developed by Barry et al. [23]. In a previous study with type–2 diabetes patients we obtained a Cronbach alpha value of 0.90 [20].
- Safety: patients will be asked whether participation in the study has produced any psychological undesired outcome such as anxiety, mood changes, uncertainty or any kind of concern.
At baseline, in the telephone call to recruit patients, assessment will include the primary outcome (DCS), as well as the following control variables: sociodemographic and clinical data; the Patient Activation Measure (PAM) [24], and general quality of life measured with the EuroQoL questionnaire (EQ–5D–5L) [25]. Knowledge will be assessed just before the application of the PtDA (or the OA general information in the control group). Immediately after the intervention, all the outcome measures will be assessed again.
Sample size calculation
Assuming type I and II errors of 0.05 and 0.20 respectively, a total of 124 patients are needed in order to detect a medium effect size (Cohen’s d = 0.5) in the primary dependent variable (DCS). No losses are expected given that the post-intervention assessment will take place immediately after the application of the PtDA.
Statistical analysis
Data will be analysed on an intention-to-treat basis, although given the nature of the intervention we do not expect relevant protocol deviations. The baseline equivalence of the randomized groups will be assessed by means of Student’s t-test and χ2 test, for continuous and categorical variables, respectively. To analyse the effect of the intervention on the primary outcome (DCS), mixed linear regression model with intervention as independent variable will be used, adjusting for DCS baseline scores and sociodemographic or clinical variables which show significant between-group differences at baseline. Centre will be included as a random intercept to control for the potential intraclass correlation. The same analysis will be carried out for the remaining continuous outcomes. Treatment preference will be analysed by means of χ2 test. Concordance between values and intention to undergo arthroplasty will be calculated by means of a method similar to the one proposed by Sepucha et al. [26]: first, a binary variable will be created, with a value of 1 for participants who prefer arthroplasty and scored 4 or 5 in intention to undergo it, and zero for the remaining patients. Second, a mixed logistic regression model will be performed with patients’ values as independent variables and intention to undergo arthroplasty (vs. other treatments) as dependent variable. Significant predictors will be retained in the model and the predicted score for each patient will be calculated. Patients with a predicted score equal or higher (lower) than 0.5 and intention to (not) undergo arthroplasty will be classified as concordant, and the remaining ones as not concordant. Finally, the effect of the DA in the number of concordant patients, as well as the number who make a quality decision, will be assessed with χ2 test.
Data monitoring, management, and confidentiality
This is a minimal risk study and we do not expect harms derived from participation. Therefore, an independent safety data monitoring committee is not needed. Nonetheless, researches who will apply the intervention and collect outcomes data will asked for any undesired outcome and register any behavioural or verbal sign which could indicate discomfort, annoyance or worry.
Study data will be stored on secure drives on a server’s institution, and managed according to the General Registry for the Protection of Personal Data (RGPD UE 679/2016) and Spanish Organic Law 3/2018, of December 5, on the Protection of Personal Data and Guarantee of Digital Rights. Only the study team will have access to the final dataset.
Dissemination policy
Results of this study will be published in a peer-review journal and presented in international conferences.