|
" Optimizing Health Insurance Selection among Tufts University Employees "
Hijaz, Sarah Hatoum
Lischko, Amy
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1057502
|
| Doc. No
|
:
|
TL56619
|
| Main Entry
|
:
|
Hijaz, Sarah Hatoum
|
| Title & Author
|
:
|
Optimizing Health Insurance Selection among Tufts University Employees\ Hijaz, Sarah HatoumLischko, Amy
|
| College
|
:
|
Tufts University School of Medicine
|
| Date
|
:
|
2019
|
| Degree
|
:
|
D.P.H.
|
| student score
|
:
|
2019
|
| Note
|
:
|
255 p.
|
| Abstract
|
:
|
The growing trend towards expanding employee involvement in health care choices, including health insurance plan selection, is rooted in the idea that doing so will encourage employees to be more cognizant of the consequences of such choices and, as a result, make more high-quality decisions. Within the health insurance market, however, this is particularly challenging, especially when considering the complexity of the information required for employees to make appropriate health plan choice. As part of an effort by Tufts University to offer more of a diverse selection of health insurance plan options, Tufts adopted a new health insurance plan menu starting in calendar year 2015. This action provided an opportunity to learn more about how employees weigh various health plan attributes to make decisions, and to apply this understanding towards the development and evaluation of a decision-making tool (DMT) specific to this population. This project was conducted in three phases. Phase I involved the qualitative assessment of employee health plan decision-making in order to validate and further explore findings from prior research, thereby ensuring relevance within the Tufts employee population. Consistent with the literature, participants considered health plan cost, quality, and access in making health plan choices. The level of importance of each feature varied depending on participant characteristics, such as age and health. In-depth interviews also validated the appropriateness of a dual process cognitive framework in approaching health plan decision-making. As such, the study team identified health insurance literacy and decisional conflict as important DMT evaluation outcome measures. Finally, the study team observed status quo bias among interview participants, indicating the importance of including this as a secondary evaluation outcome measure. Interview findings also strongly suggest that Tufts employees would benefit from and desire additional decision-making support. Phase II involved the incorporation of literature review and Phase I findings towards the development of a DMT to support employees in making plan selections that meet their and their families’ needs. Building on this foundation, the DMT was designed to present a decision-making setting that minimized cognitive burden, alleviated anxiety and other negative emotional states, and disrupted status quo bias. In terms of DMT structure, we aimed to ensure that users were able to navigate the DMT as intended, enjoyed using the tool, and were responsive to our choice architecture goal of enhancing active comparative assessment of health plan options in relation to personal needs, preferences, and utilization patterns. We deliberately avoided using choice architecture in this setting as a mechanism to direct users towards or away from any particular health plan option. Phase III used a randomized controlled design to evaluate whether the DMT improved employee health insurance literacy, decisional conflict, and status quo bias. We randomized 126 participants to the standard provided information (“control”) arm and 140 participants to standard provided information plus DMT (“DMT”) arm. Of these, 117 participants on the control arm and 130 participants on the control arm completed study participation. Using an intention to treat analysis, the DMT provision alone had no significant effect on primary outcome measures, health insurance literacy (DMT arm score difference – control arm score difference = 0.05, p = 0.31) and decisional conflict (DMT arm score difference – control arm score difference = 1.83, p = 0.35), although there was effect modification by gender and numeracy. There were also no significant differences between arms for most of our status quo bias measures, including plan switching (logistic regression beta coefficient = 1.91, SE = 0.92, p = 0.18) and conducting any research (yes/no) into health plan options (Chi squared p-value = 0.44). However, reported time spent researching health plan options was higher in the DMT arm (Chi-squared p-value = 0.04). Separate treatment on the treated (TOT), use (used the DMT vs. did not use the DMT among those on the DMT arm), and dose response analysis findings suggest that actually using vs. simply being provided the DMT may influence outcome measures, including plan switching (TOT: OR = 2.87, SE = 1.44, p = 0.04; Use: OR = 9.62, SE = 10.19, p = 0.03). DMT users also reported more time researching health plan options (TOT: Chi-squared p-value < 0.01; Use: Chi-squared p-value < 0.01) and were more likely to report that they considered changing health plans during Open Enrollment (TOT: Chi-squared p-value < 0.01; Use: Chi-squared p-value < 0.01). Our findings supplement research identifying status quo bias within currently insured populations, and suggest that factors outside of ensuring access to health plan information need to be taken into account when considering how to address this bias. Of note in this study, DMT users seemed be more motivated to utilize health plan information when compared to members of the control arm and compared to DMT non-users in the DMT arm. This may in part explain why they were also more likely to experience improvements in several outcome measures upon accessing the DMT. We suggest that currently insured consumers may not equally benefit from the provision of health plan information. Research into the usefulness of targeted messaging towards those most in need of motivation to actively reassess their plan choice is warranted. This project emphasizes importance of understanding and supporting health plan consumer choice. More research is needed to understand how DMT features influence diverse consumers operating within different choice settings. It is increasingly clear that a single tool format is unlikely to be appropriate in all settings and for all consumers. Ideally, further research efforts will facilitate the development of strategies that support individuals in making the most appropriate decisions for their health and finances.
|
| Descriptor
|
:
|
Health care management
|
|
|
:
|
Organizational behavior
|
|
|
:
|
Public health
|
| Added Entry
|
:
|
Lischko, Amy
|
| Added Entry
|
:
|
Tufts University School of Medicine
|
| |