Fostering shared decision-making about prostate cancer screening among African American men patients and their primary care providers: a randomized behavioral clinical trial

Patient measures and outcomes

Patient outcomes are assessed using different measures and across different study intervals. The SPIRIT schedule of assessment and interventions (Table 2) includes three points of contact: visit #1 includes measures completed at enrollment and the randomization to the intervention or control arms. Visit #2 includes measures completed immediately after the medical encounter, and visit 3 includes measures completed 3 months post medical encounter. Patients randomized to the intervention (decisional aid) complete the training and respective assessment before the clinical encounter.

Table 2 Quantitative patient measuresProvider measures and outcomes

For providers, the SPIRIT schedule of assessment and interventions (Table 3) includes three points of contact: in visit #1, providers receive summary information about the study; give implicit consent to participate; provide demographic information; and report on their practices regarding PSA-based screening for prostate cancer. Providers joining the study receive an educational brochure with information about prostate cancer risk assessments, PSA screening guidelines, health disparities in prostate cancer and recommendation for African American men, and a summary of the SDM process and tips to apply SDM during the medical encounter. Visit #2 includes measures completed immediately after each medical encounter, and visit 3 includes the study evaluation survey completed if the provider leaves the clinical site or at study closeout. Before each medical appointment (visit #2) with a participant patient, providers receive an email with the confirmation of the appointment and a reminder of study procedures.

Table 3 Quantitative provider measuresOperationalization of variablesPrimary outcome

The assessment of the primary outcome, Level of Engagement in SDM process regarding PSA-based screening, is conducted through the rating of the audio-recorded medical encounters using the OPTION scale [10]. Two raters independently rate each conversation according to the level of competence observed in 12 communication behaviors stated in the OPTION scale. Sub-scores range from 0, representing “the behavior is not observed,” to 4, indicating “the behavior is exhibited to a very high standard.” The 12 sub-scores are added to produce an overall score from 0 to 48. The average of rating scores of the two raters are computed to get the final score. A third, independent rater, in case of need, may be used to average with the previous two scores to improve wide discrepancies in individual ratings. Additional to this independent assessment of the PCP-patient conversation using the OPTION scale, patients and PCPs report their own perceptions of the encounter using the SDM questionnaires [11, 12]. The mean scores of the patient and provider own ratings are compared to the OPTION scores and analyzed to identify differences in individual perceptions of the application of SDM process during the medical encounter versus actual performance (Table 4).

Table 4 Operationalization of primary outcome variablesSecondary outcomes

The study has two secondary outcomes: Patient quality of engagement in decision making (PQED) and preference-congruent decision making (PCDM).

PQED is assessed through an average difference in pre-posttest scores of four scales: Prostate Cancer and Screening Knowledge, Decisional Confidence, Decisional Self-efficacy, and Satisfaction with Decision (Table 5). Higher percentage scores mean higher quality of SDM process. Questionnaires are completed by patients in two different points of care: at enrollment (baseline) and at 3-month follow-up.

Table 5 Operationalization of secondary outcome variables

PCDM is measured as a comparison between the patient’s intention-to-screen versus the actual PSA-screenings reported in the medical records. Measures are assessed from the medical encounter (visit #2) to follow-up (visit #3). The outcome is accessed through 3 indicators: Intention-to-screen, Actual PSA-tests received, and Congruence; this is the match between intention-to-screen and actual PSA-screening behaviors (Table 5).

Exploratory outcomes

Exploratory outcomes include the comparative measures of acceptability of the study procedures (enrollment process, intervention, and medical encounters) and overall satisfaction with the study (Table 6). The measures are completed by patients during the follow-up (Visit #3) and by PCPs during study closeout

Table 6 Operationalization of tertiary outcome variablesData collection and management

To evaluate the efficacy of the decision aid, we collect quantitative data from the surveys administered to patients and providers as well as from patients’ electronic health record (Epic, eClinical, etc.). The study protocol is integrated, applicable to each independent site and approved by the respective collaborators. Step-by-step description of procedures (who does what and how) including documents and resources needed to complete each one of the study activities are described in the Manual of Procedures (MOP). All research team members receive the respective training regarding the procedures to be sure that the protocol is administrated as designed. Following the MOP, the Data Collection System (DCS) was designed to integrate the electronic collection of structured information from patients and PCPs joining the study (patient consent/HIPAA documentation, demographic data, and baseline assessments); the automatic randomization process; the respective surveys collecting data needed to measure the study outcomes; and the electronic delivery of the educational interventions (prostate cancer screening decision aid); as well as the electronic distribution of incentives (gift cards). The DCS includes quantitative and qualitative information that is collected not only when PCPs and patients complete the study activities but also during the support given by research personnel during the recruitment process, medical encounters, and follow-up surveys. The DCS was developed using Qualtrics XM platform, versions 2020-2022. Qualtrics is a powerful cloud-based customer experience management program that allows creating and distributing surveys online. It has the capability to automatically code, validate, and process data, as well as automatically run basic statistics and produce reports [22].

The grantee, Xavier University of Louisiana, is responsible for developing and maintaining the study password-protected online data collection system, including the measures stated in the IRB protocol, and coding schemas. As owner of the data, Xavier is responsible for the security of data access and privacy, and the implementation of the controls stated in the respective IRB protocol. Any changes in the measures require the approval by the study lead principal investigator (PI) and are appropriately documented in the “summary of changes and version control” table that is included at the beginning of each IRB document.

Study monitoring and safety plan

This is an educational intervention where participants read materials, watch videos and complete questionnaires and pre-post assessments. Considering that participation in this research presents no more than minimal risk and no life-threatening events, there are no adverse events anticipated in this study. The research team meets weekly to review/discuss any issue related to the study, including the inclusion and enrollment report. As members of the research team, the project coordinator and the medical residents and research assistants in each clinical site are responsible for the day-to-day support for the trial procedures including participant screening, recruitment, consent, and retention along the study timeline. The principal investigator and co-investigators in each clinical site are responsible for supervising the trial including data safety and monitoring and report, directly to the respective IRBs, any participant’s issue that may be related to participation in the research and that may pose a greater risk than previously recognized. The External Advisory Board (EAB) follow-ups the study and conducts two official reviews per year. The IRB revises the study reports and advance once per year. Both the EAB and IRB have authority to extend study activities and/or stop, suspend, or require modifications to the study, if needed. The study statistician is responsible for the production of interim analysis, including stopping points, and recommendations to terminate the trial, if needed. In addition, the Patient Advisory Board (PAB) provides revisions and feedback regarding study interventions, recruitment materials, and survey procedures. The PAB is formed by African American men, some of them being prostate cancer survivors, who represent the communities targeted in the study. The PAB members meet twice per year to review the advance of the study, follow-up with participants, and give recommendations on how to address specific patient and community needs regarding early detection of prostate cancer.

Qualtrics, the data collection system used in this study to capture and store data, is HIPAA-compliant, uses Transport Layer Security (TLS) encryption for all transmitted Internet data, has the functionality to ensure the confidentiality of survey results, and has data auditing to ensure any changes to data are recorded [22]. Access to study records is limited to IRB-approved members of the study team. The specialist in Qualtrics support is responsible for data protection, creation of user accounts including privileges to access study raw data, and maintenance of the Qualtrics system at the university level.

Statistical analysis

Preliminary analysis, prior to hypothesis testing, focuses on descriptive statistics of measures of frequency (count, percentage, frequency), central tendency (mean, median, mode), and dispersion (range, variance, standard deviation) to summarize the characteristics of the sample, check assumptions underlying analytic procedures, and evaluate whether randomization produced comparable patient groups on demographics (age, insurance, family PrCa history, previous PSA utilization, education, and health literacy level) and baseline scores. Exploratory/confirmatory factor analyses will be used to examine and report on the psychometric properties of all study measures.

Hypothesis testing for the Level of Engagement in SDM (primary outcome) and for Patient Quality of Engagement in Decision-Making (PQED) include two-tailed hypothesis tests with an α-level of 0.05. ANOVAs and chi-square tests will be used to assess group differences on scale scores using controlling procedures to account for multiple comparisons (Holm, Bonferroni, or other procedures). Although the intervention is tailored based on education and prostate cancer family history, it is possible these variables may nonetheless account for some variability in learning from the intervention. Family history of prostate cancer may be associated with increased familiarity with or a desire to have PSA screening. Controlling for the baseline status on outcome variables, where possible, reduces within-patient variance in effect estimates. We will conduct sensitivity analyses for the PQED subscales separately to identify predictors of missing data and characterize the robustness of the effects observed.

Analysis for preference-congruent decision making (PCDM) consist of binary logistic regressions to examine the impact of the intention-to-screen (dichotomous independent variable) on PSA utilization (dichotomous outcome) while controlling for covariates.

We will run separate analysis of intervention and control arms to examine possible associations between separate measures (Prostate Cancer and Screening Knowledge, Decisional Confidence, Decisional Self-efficacy, and Satisfaction with Decision) and the study primary (SDM) and secondary (PQED, and PCDM) outcomes.

In general patients’ age range, health insurance type, family prostate cancer history, previous PSA utilization, education level, and intention-to-screen at baseline (visit 1) as well as their PCPs’ gender, age, race, clinical site, and specialty (internal medicine, family medicine, etc.) will be explanatory variables under consideration in all analysis.

留言 (0)

沒有登入
gif