Multifaceted Program to Improve Interpersonal Skills of Physicians

Rationale: Interpersonal skills, encompassing communication and empathy, are key components of effective medical consultations and represent physician core competences that are most desired by patients. High-quality communication relates with enhanced patient satisfaction, greater adherence to treatment, better health outcomes, and decreased risk of malpractice claims. Many organizations have therefore implemented structured training programs to improve physician communication skills in North America and Europe. Yet limited evidence exists on the effectiveness of these programs in improving physician interpersonal skillsMost studies used unvalidated instruments that may not accurately reflect interpersonal skills, were underpowered to detect clinically meaningful differences in interpersonal skills due to relatively limited sample size, lacked a control group, or used a study design that was prone to evaluation bias. In addition, few studies examined whether these training programs dedicated to improve physician communication skills altered patient satisfaction therapeutic alliance or modified physician self-efficacy and professional achievement.

The "Four Habits Model" is a training program that has been developed within the US Health Maintenance Organization Kaiser Permanente and implemented for teaching effective communication to thousands of clinicians in this organization over the two last decades. The Four Habits Model refers to basic medical interview tasks that are organized within four dimensions, namely, Invest in the beginning, Elicit the patient's perspective, Demonstrate empathy, and Invest in the end. Since its development, the Four Habits Model has been successfully implemented outside the Kaiser Permanente system, in various settings in North America and Western Europe.

Our research team recently carried out the cross-cultural adaptation of the Four Habits Coding Scheme (4-HCS) scale into French. The 4-HCS is a standardized instrument designed to assess physician communication skills from an external rater's perspective, relying on the conceptual framework of the Four Habits Model.

The investigators hypothesize that the implementation of a multifaceted training program based on the Four Habits Model improved physician interpersonal skills without altering consultation duration, increased patient satisfaction and therapeutic alliance, and modified physician self-efficacy and professional achievement.

Primary objective: To determine whether the implementation of a multifaceted training program relying on the conceptual framework of the Four Habits Model improved the communication and interpersonal skills for hospital physicians during consultations compared with control physicians receiving no intervention.

Secondary objective:

To compare study arms with regard to:

• mean duration of medical consultation
• For patients:
• patient satisfaction with the consultation
• therapeutic alliance
• For physicians:
• 4-HCS subscale scores
• Professional achievement

Methodology: Prospective, randomized (with a 1:1 allocation ratio), controlled, open-label, two parallel arms, superiority interventional trial in one study site. The unit of randomization is the physician. Randomization will be performed by minimization, taking into account the status (attending versus non-attending) and specialty (medical versus surgical) of the physician.

Timeline:

• Pre-intervention trial period: The participating physicians will be invited to videotape medical consultations with at least four consecutive patients (consenting to the research) over a 3-month period. The video-recorded consultations will be independently evaluated by two raters using the 4-HCS scale
• Randomization: Physicians will be randomly assigned to study arms
• Intervention: Physicians in the intervention arm will receive the training program consisting of two half-day group workshops with a review of the skills needed to build effective patient relationships and a discussion of difficult consultations. Physicians in the control group will not receive any specific intervention at this stage.
• Post-intervention trial period: The participating physicians in the two study arms will be invited to videotape medical consultations with at least four consecutive patients (consenting to the research) over a 3-month period Personalized feedback will then be given on the acquisition of skills of the physicians in the intervention arm.

Number of subjects:

224 consecutive patients included (112 per arm) nested within 28 physicians (14 per arm). A sample of 56 patients included by 14 physicians in each arm would confer a power greater than 80% to show an average difference of 7.5 points in the 4-HCS score (alpha risk of 0.05 in bilateral situations). Each arm of the trial will include 56 pre-intervention and 56 post-intervention patients, for a total of 224 patients.

Randomization:

To ensure allocation concealment, participating physicians will be randomized at the end of the pre-intervention study period. A third party will generate allocation sequence, with a 1:1 ratio, using minimization stratified by physician specialty (i.e., medical versus surgical) and attending status (i.e., attending physician versus non-attending physician). The investigators will use the software RStudio with package Minirand.

Blinding:

In this open label trial, the participating physicians will not be blinded to allocation. Only the physicians who will rate video-recorded consultations and the statistician will be blinded to study arm.

Statistical analysis of data:

A statistical analysis plan (SAP) will be developed prior to database lock, reviewed by the principal investigator and an independent statistician, and approved by the steering committee. Any post-hoc or unplanned analyses not specified in the SAP will be clearly identified as such in the final statistical report (FSR) and manuscripts for publication. No formal interim analysis is planned.

Analysis populations A statistical analysis plan (SAP) will be developed prior to database lock, reviewed by the principal investigator and an independent statistician, and approved by the steering committee. Any post-hoc or unplanned analyses not specified in the SAP will be clearly identified as such in the final statistical report (FSR) and manuscripts for publication. No formal interim analysis is planned.

Analysis populations The intention-to-treat (ITT) population will consist of all observations for participating physicians who have been randomized. Patients and physicians will be analyzed in the study arm assigned by randomization. The per-protocol (PP) population will consist of all observations for randomized physicians without any major deviation from the protocol (non-compliance with the multifaceted training program) and evaluable. The numbers of patients and physicians in ITT and PP populations will be presented by study arm throughout a CONSORT-style flow-chart extension for cluster randomized trials.

Baseline characteristics Baseline and demographic characteristics will be summarized for both ITT and PP populations. Baseline patient and physician characteristics will be compared between the two study arms.

Analysis of the primary outcome The primary outcome analysis (i.e., 4-HCS overall score) will be conducted within the ITT population and, for sensitivity reason, repeated within the PP population. For this purpose, the investigators will use a difference-in-differences approach, with a two-sided alpha level of 0.05. To account for patient clustering within participating physicians, the investigators will analyze 4-HCS overall score using random-intercept linear regression model for continuous dependent variable.

Analysis of the secondary outcomes. The analysis of secondary outcomes will be exploratory in nature. Inferential comparisons for participating physicians between study arms will be performed using the Student t or Wilcoxon rank-sum test for unpaired data for continuous outcome variables. To account for patient clustering within participating physicians, the investigators will analyze secondary outcome measures using random-intercept linear regression model for continuous dependent variable. All tests of secondary outcome analyses will be performed on both ITT and PP populations at a two-sided alpha level of 0.05.

Subgroup analysis. No subgroup analysis is planned for the primary and secondary outcome measures. Missing data. The completeness of study data will be reported for baseline characteristics and outcome variables. The investigators will perform multivariate imputation using chained equations (MICE) for replacing missing primary and secondary outcome values

Study feasibility:

The investigators identified the American multidimensional Four Habits Coding Scheme (4-HCS) scale, which produces a composite score of the quality of interpersonal skills that can be used in initial or continuing education. No instrument for measuring interpersonal skills was available in French. Therefore, the investigators proceeded with the cross-cultural adaptation of this scale. The cross-cultural adaptation of this scale was used to evaluate simulated filmed consultations of approximately 200 students from the Grenoble Faculty of Medicine during the period 2017-2018. This experiment allowed for the verification of the feasibility of the evaluation of interpersonal skills via this scale and made it possible to constitute a sufficient sample of videos to validate the psychometric properties of the scale.

Thus, this tool can be used in initial or continuing education for the evaluation of interpersonal skills and allows for the measurement of the progression of these skills within the framework of an intervention aimed at improving them.

Interest of the study:

The intervention should increase the physician's skills in establishing a caregiver-patient relationship and in managing difficult consultations, particularly those with high emotional load, without significantly altering the duration of the consultation. The proposed trial should document the effect of this intervention on specific patient (satisfaction, therapeutic alliance) and clinician (professional fulfillment) outcomes.