Concept Mapping Versus Interviews
Conditions
Keywords
Concept Mapping, VOICe, Interviews, Patient-Centered Research
Brief summary
The goal of this study is to improve the methods with which researchers identify patient centered outcomes for use in research. Specifically, the investigators will test the application of concept mapping as compared to one-on-one interviews as a comprehensive and efficient method of identifying patient-important outcomes for use in research.
Detailed description
This is an observational study comparing two methods - concept mapping and one-on-one interviews. The investigators will test the methods themselves by comparing the comprehensiveness of the lists of patient-important outcomes obtained through each method (primary outcome) as well as the efficiency, as measured by resource intensiveness, of the methods (secondary outcome). The aims of this study will be tested with adult patients with moderately to poorly controlled diabetes mellitus who receive care at a large, urban, academic health system. In order to allow comparisons of the comprehensiveness of the methods in producing patient-important outcomes regardless of the healthcare setting in which outcomes are elicited, the investigators will recruit patients for the interview group from 3 different care settings: an acute care visit (in the emergency department), a post-acute care visit (within 1 week of a hospital discharge), and a routine primary care visit. Target sample size within each healthcare setting is 30 patients, which is the anticipated number needed for thematic saturation. The investigators will then recruit patient-participants from existing clinical and research databases for 3 separate concept mapping groups, each with a target of 20 patients. In Aim 2, the investigators will test the hypothesis that concept mapping in a research setting produces outcomes consistent with those elicited from patients in the real-world settings. In Aim 3, the investigators will assess the overall resource utilization, assessed primarily by time required from patients, the research team, and the advisory board, of qualitative interviews conducted to thematic saturation within a single healthcare setting compared to resources used for a single iteration of concept mapping. Findings from this work will improve patient-centered outcomes research methods by providing researchers with information about standardized scalable methods to identify patient-important outcomes for use in research studies, so that individual patients are able to select outcomes that are most useful to them. In addition to providing insight regarding methods for eliciting patient-centered outcomes, the study will also provide valuable information to stakeholders regarding how patients' priorities vary across the care continuum. The investigators will report how findings from concept mapping apply across healthcare settings, and the degree to which patient priorities (based on interviews) differ across healthcare settings. If concept mapping results (collected in the research setting) capture the priorities identified by patients across different healthcare settings, this finding would enhance researchers' ability to perform studies in different, or multiple, settings. If the findings do not generalize across settings, this will be a valuable finding demonstrating that patient priorities differ across interview settings and patient acuity and will validate researchers' needs to be specific about the setting in which they choose to perform research to ensure they are reaching their intended patient population. The work is being performed in close collaboration with the Patient and Key Stakeholder Advisory Board (PAKSAB) - equal partners with the research team who have been involved with the proposal from inception. PAKSAB members will be part of the research team throughout conducting the aims and assisting with all analysis and data interpretation.
Interventions
Patients will be engaged in open-ended, semi-structured qualitative interviews, which will be performed one-on-one either in person or over the phone (depending on the healthcare setting that they are recruited from). Qualitative interviews will be audio recorded, with the patient's permission, transcribed, de-identified and entered into NVivo software for coding and analysis.
The CM process consists of 3 steps that take place over 3 sessions: Step 1: Generation of Ideas- Participants brainstorm and generate responses to the focus statement. Once the group agrees that no new statements are being generated, the list of statements is reviewed within the group. Step 2: Structuring of Statements- Each participant is given a set of sort cards and asked to sort the statements into piles. Participants then rate each idea regarding importance. Research staff enters this information into the CM software. Concept Systems Global Software generates point maps using a technique that detects underlying similarities/differences between statements. The CM software then uses hierarchical cluster analysis to draw boundaries around the point map to create conceptual clusters. Step 3: Interpretation- The CM group revises the concept map. Participants review the cluster names suggested by the software and decide upon final naming of each cluster as a group.
Sponsors
Study design
Eligibility
Inclusion criteria
* Adult patient (age 18 and older) * English speaking * Provide informed consent * Diagnosis of moderately to poorly controlled diabetes mellitus (DM) defined as follows (for interview groups): * Acute care setting: patient in the Jefferson Emergency Department (ED) for a diabetes-related problem determined to require medical treatment * Post-acute care setting: patient was discharged from the Jefferson Family Medicine Associates (JFMA) hospital service within the past 7 days after admission for a diabetes-related problem * Primary care setting: patient currently attending a routine scheduled care visit (not urgent need) at the JFMA practice and has at least 2 measurements of HgbA1C \> 7.5 in the prior one year * Diagnosis of moderately to poorly controlled diabetes mellitus (DM) defined as follows (for concept mapping groups): * Acute care setting: patient within the Jefferson ED patient contact database (see below) who has had a visit to the Jefferson ED over the past 6 months (as determined on chart review) for a diabetes-related problem * Post-acute care setting: patient was discharged from the JFMA hospital service within the past 6 months after admission for a diabetes-related problem * Primary care setting: patient has had routine scheduled office visit within the past 6 months to the JFMA practice and has at least 2 measurements of HgbA1C \> 7.5 in the prior one year
Exclusion criteria
* Patient has had a significant permanent complication related to DM including: * End stage renal disease * History of amputation * Blindness related to diabetes complication * Patient undergoing medical clearance for a detox center or any involuntary court or magistrate order * Patient in police custody or currently incarcerated individual * Patient who has, in their clinician's best judgment, major communication barriers such as visual or hearing impairment or dementia that would compromise their ability to give written informed consent
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Comprehensiveness of Interviews as Compared to One Concept Mapping Group | Interviewed patients participated for 1 day; Concept mapping patients participated for 3 days | The investigators will use a qualitative content analysis approach to analyze interview transcripts from one healthcare setting, with one of the codes being goals. All ideas coded to goals that are in any way relevant to patients' diabetes care will be extracted to create a list of patient-important outcomes. The investigators will then determine the proportion of patient-important outcomes identified in the interviews from one healthcare setting that are present in the list of patient-important outcomes generated from the initial concept mapping group during the brainstorming session. The investigators will also identify the presence of unique outcomes found in each method. |
| Comprehensiveness of Interviews Compared to Three Concept Mapping Groups | Interviewed patients participated for 1 day; One group of concept mapping patients participated for 3 days | The investigators will use a qualitative content analysis approach to analyze interview transcripts, with one of the codes being goals. All ideas coded to goals that are in any way relevant to patients' diabetes care will be extracted to create a list of patient-important outcomes. The investigators will then determine the proportion of patient-important outcomes identified in interviews that are present in an aggregate list of patient-important outcomes generated from the brainstorming session of all three concept mapping groups. The investigators will also identify the presence of additional patient-important outcomes in three concept mapping groups that were not identified in interviews. |
| Comprehensiveness of Concept Mapping | 3 days for one concept mapping group | The investigators will measure the comprehensiveness of outcomes elicited in one concept mapping group compared to multiple groups. The investigators will assess concept mapping saturation, wherein we compare the patient-important outcomes that emerge from each CM group. The investigators will use the outcomes from our first group as the baseline data, and will determine the amount of new data added from including a second/third group. This assessment will allow us to draw a basic concept mapping saturation curve. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Comparison of Method Efficiency in Terms of Time | Interviewed patients participated for 1 day; Concept mapping patients participated for 3 days | The investigators will assess the efficiency of implementation of the two methods, with efficiency including time required by researchers (team of 3) and patient-participants to complete each method. The investigators will compare the efficiency of conducting one concept mapping iteration (24 participants) to the efficiency of performing one set of interviews done to theme saturation (30 interviews). This efficiency analysis is structured for what would need to be budgeted in a grant application, and will provide useful information for general planning needs for method implementation. |
| Comparison of Method Efficiency in Terms of Cost | Interviewed patients participated for 1 day; Concept mapping patients participated for 3 days | The investigators will assess the efficiency of implementation of the two methods, with efficiency assessed by cost to complete a single concept mapping iteration (24 participants) and cost to complete one set of interviews performed to saturation (30 interviews). This efficiency analysis is structured for what would need to be budgeted in a grant application, and will provide useful information for general planning needs for method implementation. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Interviews Interview participants will be recruited from 3 different care settings: an acute care visit (in the emergency department), a post-acute care visit (within 1 week of a hospital discharge), and a routine primary care visit. Target sample size within each healthcare setting is 30 patients, which is the anticipated number needed for thematic saturation. The total recruitment goal for this cohort is 90-120 participants.
Interviews: Patients will be engaged in open-ended, semi-structured qualitative interviews, which will be performed one-on-one either in person or over the phone (depending on the healthcare setting that they are recruited from). Qualitative interviews will be audio recorded, with the patient's permission, transcribed, de-identified and entered into NVivo software for coding and analysis. | 89 |
| Concept Mapping Concept mapping participants will be recruited from existing clinical and research databases for 3 separate concept mapping groups, each with a target recruitment of 20 patients. The total recruitment goal for this cohort is 60 people.
Concept Mapping (CM): The CM process consists of 3 steps that take place over 3 sessions:
Step 1: Generation of Ideas- Participants brainstorm responses to a focus statement.
Step 2: Structuring of Statements- Each participant is given a set of sort cards and asked to sort the statements into piles. Participants then rate each idea regarding importance.
CM Software detects underlying similarities/differences between statements to generate point maps. The CM software then uses hierarchical cluster analysis to draw boundaries around the point map to create conceptual clusters.
Step 3: Interpretation- The CM group revises the concept map. Participants review cluster names suggested by the software and decide upon final naming of each cluster. | 52 |
| Total | 141 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Later determined to be ineligible | 2 | 0 |
| Overall Study | Participant interviewed twice | 1 | 0 |
| Overall Study | Recording malfunction | 2 | 0 |
| Overall Study | Withdrawal by Subject | 2 | 0 |
Baseline characteristics
| Characteristic | Interviews | Concept Mapping | Total |
|---|---|---|---|
| Age, Continuous | 54.6 years STANDARD_DEVIATION 13.8 | 55.6 years STANDARD_DEVIATION 14.7 | 54.9 years STANDARD_DEVIATION 14.1 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 8 Participants | 3 Participants | 11 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 80 Participants | 49 Participants | 129 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 1 Participants | 0 Participants | 1 Participants |
| Race/Ethnicity, Customized Race Black or African American | 60 Participants | 42 Participants | 102 Participants |
| Race/Ethnicity, Customized Race Caucasian/White | 24 Participants | 5 Participants | 29 Participants |
| Race/Ethnicity, Customized Race Other | 4 Participants | 4 Participants | 8 Participants |
| Race/Ethnicity, Customized Race Unknown or not reported | 1 Participants | 1 Participants | 2 Participants |
| Region of Enrollment United States | 89 participants | 52 participants | 141 participants |
| Sex: Female, Male Female | 49 Participants | 24 Participants | 73 Participants |
| Sex: Female, Male Male | 40 Participants | 26 Participants | 66 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 |
Outcome results
Comprehensiveness of Concept Mapping
The investigators will measure the comprehensiveness of outcomes elicited in one concept mapping group compared to multiple groups. The investigators will assess concept mapping saturation, wherein we compare the patient-important outcomes that emerge from each CM group. The investigators will use the outcomes from our first group as the baseline data, and will determine the amount of new data added from including a second/third group. This assessment will allow us to draw a basic concept mapping saturation curve.
Time frame: 3 days for one concept mapping group
Population: This analysis is of the number of patient-important outcomes per concept mapping group.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Interviews | Comprehensiveness of Concept Mapping | Two Groups | 36 Reported patient-important outcomes |
| Interviews | Comprehensiveness of Concept Mapping | One Group | 33 Reported patient-important outcomes |
| Interviews | Comprehensiveness of Concept Mapping | Three Groups | 38 Reported patient-important outcomes |
Comprehensiveness of Interviews as Compared to One Concept Mapping Group
The investigators will use a qualitative content analysis approach to analyze interview transcripts from one healthcare setting, with one of the codes being goals. All ideas coded to goals that are in any way relevant to patients' diabetes care will be extracted to create a list of patient-important outcomes. The investigators will then determine the proportion of patient-important outcomes identified in the interviews from one healthcare setting that are present in the list of patient-important outcomes generated from the initial concept mapping group during the brainstorming session. The investigators will also identify the presence of unique outcomes found in each method.
Time frame: Interviewed patients participated for 1 day; Concept mapping patients participated for 3 days
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Interviews | Comprehensiveness of Interviews as Compared to One Concept Mapping Group | Total number of outcomes | 26 Reported patient-important outcomes |
| Interviews | Comprehensiveness of Interviews as Compared to One Concept Mapping Group | Number of unique outcomes | 6 Reported patient-important outcomes |
| Concept Mapping | Comprehensiveness of Interviews as Compared to One Concept Mapping Group | Total number of outcomes | 33 Reported patient-important outcomes |
| Concept Mapping | Comprehensiveness of Interviews as Compared to One Concept Mapping Group | Number of unique outcomes | 13 Reported patient-important outcomes |
Comprehensiveness of Interviews Compared to Three Concept Mapping Groups
The investigators will use a qualitative content analysis approach to analyze interview transcripts, with one of the codes being goals. All ideas coded to goals that are in any way relevant to patients' diabetes care will be extracted to create a list of patient-important outcomes. The investigators will then determine the proportion of patient-important outcomes identified in interviews that are present in an aggregate list of patient-important outcomes generated from the brainstorming session of all three concept mapping groups. The investigators will also identify the presence of additional patient-important outcomes in three concept mapping groups that were not identified in interviews.
Time frame: Interviewed patients participated for 1 day; One group of concept mapping patients participated for 3 days
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Interviews | Comprehensiveness of Interviews Compared to Three Concept Mapping Groups | Total Number of Outcomes | 26 Reported patient-important outcomes |
| Interviews | Comprehensiveness of Interviews Compared to Three Concept Mapping Groups | Unique Number of Outcomes | 3 Reported patient-important outcomes |
| Concept Mapping | Comprehensiveness of Interviews Compared to Three Concept Mapping Groups | Total Number of Outcomes | 38 Reported patient-important outcomes |
| Concept Mapping | Comprehensiveness of Interviews Compared to Three Concept Mapping Groups | Unique Number of Outcomes | 15 Reported patient-important outcomes |
Comparison of Method Efficiency in Terms of Cost
The investigators will assess the efficiency of implementation of the two methods, with efficiency assessed by cost to complete a single concept mapping iteration (24 participants) and cost to complete one set of interviews performed to saturation (30 interviews). This efficiency analysis is structured for what would need to be budgeted in a grant application, and will provide useful information for general planning needs for method implementation.
Time frame: Interviewed patients participated for 1 day; Concept mapping patients participated for 3 days
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Interviews | Comparison of Method Efficiency in Terms of Cost | Patient Incentives | 750 Dollars |
| Interviews | Comparison of Method Efficiency in Terms of Cost | Transcription | 1,000 Dollars |
| Interviews | Comparison of Method Efficiency in Terms of Cost | Data Analysis Software | 1,380 Dollars |
| Concept Mapping | Comparison of Method Efficiency in Terms of Cost | Patient Incentives | 3,000 Dollars |
| Concept Mapping | Comparison of Method Efficiency in Terms of Cost | Transcription | 0 Dollars |
| Concept Mapping | Comparison of Method Efficiency in Terms of Cost | Data Analysis Software | 2,000 Dollars |
Comparison of Method Efficiency in Terms of Time
The investigators will assess the efficiency of implementation of the two methods, with efficiency including time required by researchers (team of 3) and patient-participants to complete each method. The investigators will compare the efficiency of conducting one concept mapping iteration (24 participants) to the efficiency of performing one set of interviews done to theme saturation (30 interviews). This efficiency analysis is structured for what would need to be budgeted in a grant application, and will provide useful information for general planning needs for method implementation.
Time frame: Interviewed patients participated for 1 day; Concept mapping patients participated for 3 days
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Interviews | Comparison of Method Efficiency in Terms of Time | Research team trainings | 20 Hours |
| Interviews | Comparison of Method Efficiency in Terms of Time | Research team travel | 60 Hours |
| Interviews | Comparison of Method Efficiency in Terms of Time | Patient participation | 1 Hours |
| Interviews | Comparison of Method Efficiency in Terms of Time | Research team conducting method | 38 Hours |
| Interviews | Comparison of Method Efficiency in Terms of Time | Research team participant recruitment | 66 Hours |
| Interviews | Comparison of Method Efficiency in Terms of Time | Research team data analysis | 133 Hours |
| Interviews | Comparison of Method Efficiency in Terms of Time | Patient travel | 0 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Research team data analysis | 4 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Patient travel | 2 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Patient participation | 6 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Research team trainings | 11 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Research team participant recruitment | 59 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Research team travel | 12 Hours |
| Concept Mapping | Comparison of Method Efficiency in Terms of Time | Research team conducting method | 18 Hours |