Prescription Drug Insurance Decision Making
Conditions
Keywords
Medicare Part D, decision support, prescription drugs, expert recommendation, health insurance choice
Brief summary
The objective of this study is to determine whether providing Medicare beneficiaries with a web-based patient-centered decision tool to help them choose among prescription medication coverage plans improves outcomes for patients including a greater likelihood of changing a plan, better coverage for prescribed drugs, less decisional conflict when choosing plans, and greater satisfaction with the choice process relative to current practice.
Detailed description
In this study, the investigators tested the effectiveness of two versions of a web-based tool (called CHOICE) to help people choose among Medicare Part D plans (Treatments A and B) relative to standard care (Control). Both treatment arms incorporated simplified design and automated importation of an individual's prescription drugs relative to standard care. The treatment arms varied based on whether they provided expert guidance on recommended plans. In the control arm, study participants were directed to the existing, publicly available Medicare.gov website and received instructions on how to download their drugs from the Palo Alto Medical Foundation (PAMF) patient-facing online personal health portal (myhealthonline). The study sample included PAMF patients who were enrolled in Part D plans (not Medicare Advantage) during the 2016 enrollment period. Prior to the 2017 open enrollment period (October 15 to December 7, 2016), we invited a subset of PAMF patients not covered by either MediCal or a Medicare Advantage plan, aged 66-85, residing in 4 counties served by PAMF, and with at least one active medication order to participate in a study examining the effectiveness of decision tools that provide personalized information on the financial implications of enrolling in different Part D plans. The primary study outcomes included 1) Plan switching, 2) Decisional conflict 3) Satisfaction with the choice process, and 4) Change in generosity of coverage of prescription drugs. The investigators measured the primary study outcomes using a combination of administrative data and a post open enrollment survey. The investigators also collected information on individual characteristics at the time of enrollment in the study and implemented a survey examining use of the intervention tool to assess patient experience at the time of use.
Interventions
BEHAVIORALExpert Recommendation
Decision support tool that provides personalized information on the financial implications of enrolling in different plans and expert recommendations of particular plans.
BEHAVIORALIndividual Analysis
Decision support tool that provides personalized information on the financial implications of enrolling in different plans.
BEHAVIORALUsual Care
Study subjects randomized to the control arm will receive information on how to download their prescription drug information from their electronic medical record and provided with a list of resources available in the community to help them choose a prescription drug plan.
Sponsors
Palo Alto Medical Foundation
Patient-Centered Outcomes Research Institute
Stanford University
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
66 Years to 85 Years
Healthy volunteers
No
Inclusion criteria
* The study population included Medicare beneficiaries who received care from the Palo Alto Medical Foundation, a large multi-specialty group in the San Francisco Bay Area ages 66-85, one per household enrolled in a Medicare Part D plan in 2016.
Exclusion criteria
* Enrolled in a Medicare Advantage Plan and/or MediCal
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Count of Participants Whose 2017 Plan Differed From Their 2016 Plan | within 50 days of the end of the open enrollment period. | Indicator of whether the self-reported plan of the participant differed before and after open enrollment and the participant reported that s/he changed plans during open enrollment. |
| Decisional Conflict | within 50 days of the end of the open enrollment period. | Low literacy decisional conflict scale (Linder et al., 2011), edited slightly for context of health insurance rather than treatment choice. The scale has 4 subscales (uncertainty, informed, values clarity and support) with 2 to 3 questions per subscale. Respondents can indicate yes, no, or unsure for each item. An answer of yes receives 0, unsure receives 2 and no receives 4 points. The sum of the responses to each question within a subscale is normalized to a scale of 25. The subscales are then summed to a total score ranging from 0 to 100 where 0 represents the lowest level of decisional conflict and 100 represents the highest level of decisional conflict. |
| Satisfaction With the Choice Process | within 50 days of the end of the open enrollment period | Response to the question of, How satisfied are you with the process of choosing a plan? with 4 potential responses: very satisfied, somewhat satisfied, somewhat dissatisfied and very dissatisfied. The count of participants who responded very satisfied is reported. |
| Change in Estimated Prescription Drug Spending | within 50 days of the end of the open enrollment period | Change in estimated prescription drug spending is the difference in estimated spending in US dollars, including both premiums and out-of-pocket spending on prescription drugs, between the participant's 2016 and 2017 plans based on their initial drug list. |
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Control Usual Care: Study subjects randomized to the control arm received access to information on resources available in the community to help them choose a prescription drug plan and on how to download their prescription drug information from their electronic medical record. | 313 |
| Expert Recommendation Study subjects randomized to the Expert Recommendation arm received access to a decision support tool that provided personalized, expert recommendations of particular plans as well as personalized information on the financial implications of enrolling in different plans, including plan premiums and spending on prescription drugs, and the Medicare star ratings (a measure of average customer satisfaction). | 316 |
| Individual Analysis Study subjects randomized to the Individual Analysis arm received access to a decision support tool that provided personalized information on the financial implications of enrolling in different plans, including plan premiums and spending on prescription drugs, and the Medicare star ratings (a measure of average customer satisfaction). | 299 |
| Total | 928 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 | FG002 |
|---|---|---|---|---|
| Overall Study | Survey Non-response | 71 | 94 | 92 |
Baseline characteristics
| Characteristic | Control | Expert Recommendation | Individual Analysis | Total |
|---|---|---|---|---|
| Age, Continuous | 72.68 years STANDARD_DEVIATION 4.64 | 72.03 years STANDARD_DEVIATION 4.55 | 72.11 years STANDARD_DEVIATION 4.5 | 72.68 years STANDARD_DEVIATION 4.64 |
| Estimated prescription drug spending | 1496.80 US dollars STANDARD_DEVIATION 1689.97 | 1701.38 US dollars STANDARD_DEVIATION 2068.86 | 1580.56 US dollars STANDARD_DEVIATION 1493.58 | 1593.44 US dollars STANDARD_DEVIATION 1771.73 |
| Race/Ethnicity, Customized Asian | 35 Participants | 17 Participants | 18 Participants | 70 Participants |
| Race/Ethnicity, Customized Black | 0 Participants | 1 Participants | 1 Participants | 2 Participants |
| Race/Ethnicity, Customized Hispanic | 7 Participants | 3 Participants | 6 Participants | 16 Participants |
| Race/Ethnicity, Customized Other or missing | 28 Participants | 33 Participants | 34 Participants | 95 Participants |
| Race/Ethnicity, Customized White | 243 Participants | 262 Participants | 240 Participants | 745 Participants |
| Sex: Female, Male Female | 166 Participants | 150 Participants | 150 Participants | 466 Participants |
| Sex: Female, Male Male | 147 Participants | 166 Participants | 149 Participants | 462 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk | EG002 affected / at risk |
|---|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 | 0 / 0 |
Outcome results
Primary
Change in Estimated Prescription Drug Spending
Change in estimated prescription drug spending is the difference in estimated spending in US dollars, including both premiums and out-of-pocket spending on prescription drugs, between the participant's 2016 and 2017 plans based on their initial drug list.
Time frame: within 50 days of the end of the open enrollment period
Population: Study participants who responded to questions about their 2017 plan choice in the final survey administered after open enrollment had ended.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Control | Change in Estimated Prescription Drug Spending | -178.90 US dollars | Standard Deviation 1029.93 |
| Expert Recommendation | Change in Estimated Prescription Drug Spending | -249.59 US dollars | Standard Deviation 899.81 |
| Individual Analysis | Change in Estimated Prescription Drug Spending | -197.29 US dollars | Standard Deviation 599.38 |
Primary
Count of Participants Whose 2017 Plan Differed From Their 2016 Plan
Indicator of whether the self-reported plan of the participant differed before and after open enrollment and the participant reported that s/he changed plans during open enrollment.
Time frame: within 50 days of the end of the open enrollment period.
Population: Set of participants who responded to questions about plan enrollment in the final survey.
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Control | Count of Participants Whose 2017 Plan Differed From Their 2016 Plan | 85 Participants |
| Expert Recommendation | Count of Participants Whose 2017 Plan Differed From Their 2016 Plan | 110 Participants |
| Individual Analysis | Count of Participants Whose 2017 Plan Differed From Their 2016 Plan | 86 Participants |
Primary
Decisional Conflict
Low literacy decisional conflict scale (Linder et al., 2011), edited slightly for context of health insurance rather than treatment choice. The scale has 4 subscales (uncertainty, informed, values clarity and support) with 2 to 3 questions per subscale. Respondents can indicate yes, no, or unsure for each item. An answer of yes receives 0, unsure receives 2 and no receives 4 points. The sum of the responses to each question within a subscale is normalized to a scale of 25. The subscales are then summed to a total score ranging from 0 to 100 where 0 represents the lowest level of decisional conflict and 100 represents the highest level of decisional conflict.
Time frame: within 50 days of the end of the open enrollment period.
Population: Study participants who responded to decisional conflict questions on the final survey.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Control | Decisional Conflict | 21.06 units on a scale | Standard Deviation 22.56 |
| Expert Recommendation | Decisional Conflict | 20.86 units on a scale | Standard Deviation 21.99 |
| Individual Analysis | Decisional Conflict | 19.56 units on a scale | Standard Deviation 22.14 |
Primary
Satisfaction With the Choice Process
Response to the question of, How satisfied are you with the process of choosing a plan? with 4 potential responses: very satisfied, somewhat satisfied, somewhat dissatisfied and very dissatisfied. The count of participants who responded very satisfied is reported.
Time frame: within 50 days of the end of the open enrollment period
Population: Study participants who responded to question about satisfaction with the choice process on the final survey
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Control | Satisfaction With the Choice Process | 122 Participants |
| Expert Recommendation | Satisfaction With the Choice Process | 148 Participants |
| Individual Analysis | Satisfaction With the Choice Process | 134 Participants |