Journal of American College Health

ISSN: 0744-8481 (Print) 1940-3208 (Online) Journal homepage: http://www.tandfonline.com/loi/vach20

Health insurance literacy and health service

utilization among college students

Tyler G. James, Meagan K. Sullivan, Leanne Dumeny, Katherine Lindsey,

JeeWon Cheong & Guy Nicolette

To cite this article: Tyler G. James, Meagan K. Sullivan, Leanne Dumeny, Katherine
Lindsey, JeeWon Cheong & Guy Nicolette (2018): Health insurance literacy and health
service utilization among college students, Journal of American College Health, DOI:
10.1080/07448481.2018.1538151

To link to this article: https://doi.org/10.1080/07448481.2018.1538151

View supplementary material

Published online: 10 Dec 2018.

Submit your article to this journal

Article views: 34

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=vach20
JOURNAL OF AMERICAN COLLEGE HEALTH
https://doi.org/10.1080/07448481.2018.1538151

MAJOR ARTICLE

Health insurance literacy and health service utilization among college students
Tyler G. James, MS, CHESa , Meagan K. Sullivan, BS, BAb, Leanne Dumeny, MSMc , Katherine Lindsey, MS,
CPC, CHDAd, JeeWon Cheong, PhDa, and Guy Nicolette, MD, CAQSMd,e
a
Department of Health Education and Behavior, University of Florida, Gainesville, Florida, USA; bDepartment of Social and Behavioral
Sciences, University of Florida, Gainesville, Florida, USA; cDepartment of Pharmacotherapy and Translational Research, University of Florida,
Gainesville, Florida, USA; dStudent Health Care Center, University of Florida, Gainesville, Florida, USA; eDepartment of Community Health and
Family Medicine, University of Florida, Gainesville, Florida, USA

ABSTRACT ARTICLE HISTORY

Objective: Health literacy and health insurance literacy affect healthcare utilization. The purpose Received 5 June 2018
of this study was to determine the relation between health insurance knowledge, self-efficacy, and Revised 1 October 2018
student healthcare utilization in the past year. Participants: A random sample of 1,450 respond- Accepted 14 October 2018
ents, over the age of 18, attending a public university in the southeastern United States com-
KEYWORDS
pleted a survey in March 2017. Methods: A model was constructed to test the effect of health College health; healthcare
insurance self-efficacy on the relation between knowledge and healthcare utilization in the past utilization; health insurance
year. Results: Health insurance knowledge (M ¼ 5.8, range 0–10) and self-efficacy (M ¼ 2.48, knowledge; health
range 1–4) were low. Self-efficacy was a significant moderator when explaining healthcare utiliza- insurance literacy; health
tion in the past year. Conclusions: College students have low knowledge and self-efficacy regard- insurance self-efficacy
ing health insurance. These findings can be used for developing policies and self-efficacy-based
health education programs that may increase student healthcare utilization.

Introduction related to the “young invincible” phenomenon (ie, adoles-

cent and young adult self-perception that “they are not at
Since the full force of the Patient Protection and Affordable
risk for poor health”).7 However, young adults do experience
Care Act (ACA) taking effect in January 2014, the United
high morbidity of chronic illness and mental health condi-
States has increased health insurance coverage for over 20 mil- tions, and high mortality of suicide and unintentional
lion adults, with large gains seen among young adults (ages injury.8 Thus, researchers need to identify patterns of util-
19–25). In fact, between October 2013 and early 2016, the ization to help increase college student healthcare engage-
uninsured rate among nonelderly adults declined 43%.1 ment and prevention efforts.
Improving health insurance coverage has been shown to Among socio-economic and contextual factors, health lit-
improve healthcare quality and outcomes with increased out- eracy has implications for health service utilization and pre-
patient utilization and preventive care, reduction in emergency ventative medicine. The Institute of Medicine defines health
department use among some patient populations—including literacy as “the degree to which individuals have the capacity
younger adults—and improved self-reported health.2–4 to obtain, process, and understand basic health information
When considering healthy lifestyles over the course of and services needed to make appropriate health decisions.”9
aging, college student populations present a unique oppor- Health literacy is a significant factor when considering a
tunity to deliver health education which may establish posi- person’s ability to navigate through complex health systems.
tive health behavior change for a lifetime. In a multi- Having been well studied in both college students and gen-
university study between January 2011 and May 2014, over eral populations, health literacy has been associated with
800,000 students used their university’s student health center socioeconomic indicators including race, disability, age, and
(SHC), compiling over 4 million patient encounters. level of education. In the general population, large segments
Preventive services had the highest number of visits; how- (23% to 37%) of the adult population have marginal or
ever, only 15.6% of all enrolled students were seen in this inadequate health literacy.10–13 College students are not pro-
category each year.5 This is despite a majority (96%) of col- tected from inadequate health literacy as some studies sug-
lege students reporting having health insurance coverage,6 gest that 1-in-15 college students are at risk for low health
and the rising prominence of SHCs as healthcare providers literacy, with nonwhite students presenting higher risk.14
tailored to the unique needs of college students. College stu- When searching for health information, college students
dent lack of engagement in preventive medicine could be report the Internet as a preferred platform.14 Yet, many

CONTACT Tyler G. James tjames95@ufl.edu Department of Health Education and Behavior, University of Florida, PO Box 118210, Gainesville, FL 32611-
8210, USA.
Supplemental data for this article can be accessed here.
ß 2018 Taylor & Francis Group, LLC
2 T. JAMES ET AL.

college students do not have adequate eHealth literacy (ie, studies did not explore potential relations of these variables
skills or knowledge needed to access accurate health infor- to healthcare utilization indicators; yet, almost 25% of col-
mation online).15 When confronted to barriers accessing lege students indicate that their lack of understanding insur-
health information, students may turn to parents who also ance delayed or stopped their healthcare seeking.28 Given
have limited health literacy: a study of more than 6,100 the financial barriers of healthcare service utilization among
parents showed that 28.7% had below-basic or basic health young adult populations, health insurance literacy is critical
literacy and that 68.4% were unable to enter names and when making decisions to engage in healthcare as it has dir-
birth dates correctly on a health insurance form.16 It is pos- ect implications for effectively using services by estimating
sible that, like the general population, the inadequate health costs to keep patient financial responsibility low, thus mini-
and eHealth literacy impacts the uptake of positive health mizing a strong barrier to utilization.29 The purpose of this
behaviors among college populations.17 study was to assess and examine the relations among health
Young adult patients, recognizing the financial barriers insurance knowledge and self-efficacy of using and choosing
associated with accessing healthcare and having difficulty health insurance, and health service utilization outcomes
estimating the potential out-of-pocket cost for services, typ- among a larger sample of college students. Specifically, we
ically forgo health services.18,19 Given that the majority of hypothesize that (H1) health insurance self-efficacy will
this population has health insurance, it becomes necessary moderate the relation between health insurance knowledge
to address health insurance literacy when exploring factors and healthcare engagement with a provider in the past
that influence healthcare utilization.20 Health insurance liter- 12 months, and (H2) that self-efficacy will also moderate the
acy, a facet of health literacy, is the degree that a person has relation between knowledge and healthcare engagement at
the knowledge, ability, and confidence to select and use the SHC in the past 12 months.
health insurance plans.21 Thus, health insurance literacy can
be decomposed into multiple constructs: knowledge, self-
efficacy, and actual ability to select and use health- Methods
care coverage.22 Overview, setting, and participants
Health insurance literacy has not been as well-studied as
health literacy. However, recent research has shown dispar- We administered an anonymous web-based survey to a sam-
ities in health insurance literacy among the general public. ple of college students attending a large, public land-, sea-,
In a national survey conducted by the Kaiser Family and space-grant research university in the southeastern
Foundation (KFF), younger Americans (age 18–29), people United States with an enrolled student population over
with lower levels of education, and people who were unin- 52,000. The university has an accredited SHC in a central-
sured were less familiar with basic health insurance terms ized campus location, with outreach psychiatry services at
and calculations.23 Of those ages 18–29, 43% were consid- the campus counseling center. In addition, the university
ered low scorers (ie, getting 0–4 of the 10 questions correct) has a hard waiver health insurance mandate for students
and had the lowest mean score of all the age groups; these who are enrolled at least half-time in a degree seeking pro-
findings are further corroborated by other studies indicating gram; this mandate is to ensure students have adequate
low health insurance knowledge among highly educated coverage using the 10 essential health benefits outlined as
young adults, who had difficulty in defining “deductible”, part of the ACA. Students must submit proof of coverage
“coinsurance”, and “PPO” among other health insurance prior to each academic year; this coverage is verified by the
terms.24 Further, Bartholomae and colleagues found that, in University’s Health Compliance Program. Students who do
a community based sample of participants in a health insur- not have health insurance or do not have adequate coverage
ance literacy intervention in seven U.S. states, lower health are automatically enrolled in the school sponsored insur-
insurance literacy was associated with younger age, male ance plan.
gender, and lower income.25 Additional research has found This study was declared exempt by the university’s
that young consumers, including college students, are over- Institutional Review Board. The Office of Institutional
whelmed when exploring health insurance plans and that Research and Planning provided a simple random sample
they are less confident in their ability to choose.22,26,27 (without replacement) of 10,000 undergraduate and graduate
We postulate that the disparities in health insurance liter- students over the age of 18. Students without local address
acy (both knowledge and self-efficacy) seen in the general and who do not attend classes on-campus (ie, distance
population are also present in college students. Research learning students) were excluded from the sampling frame
among a small, nonprobability sample of college students in due to differing SHC access restrictions and exemption from
the U.S. indicates that students have low self-efficacy of the university health insurance mandate. Sampled partici-
choosing and using health insurance plans.27 Further, in an pants were invited via e-mail to take the survey in March
analysis of college student health insurance knowledge, col- 2017. Prospective participants were sent three recruitment e-
lege students scored a median of 75% correct answers, and mails over the course of two weeks. Recruitment e-mails
struggled with knowledge regarding health reimbursement provided a brief explanation of the survey topic, incentive,
accounts, cost-sharing, and plan navigation.28 To date, stud- and informed consent. After completing the survey, the first,
ies on college students have focused on measuring only middle, and last 10 participants who completed an unlinked
knowledge or only self-efficacy.27,28 In addition, these incentive form were provided $20 AmazonV R gift cards.
 JOURNAL OF AMERICAN COLLEGE HEALTH 3

Measures Disease Control’s (CDC) Behavioral Risk Factor Surveillance

Survey were modified to include SHCs and used to deter-
We measured demographics, health insurance knowledge
mine healthcare utilization in the past year. Specifically, par-
and self-efficacy, and healthcare service utilization. A list of
ticipants were asked (1) how many times they had visited
the items used and links to their sources are available in the
the SHC, and (2) how many times they were a patient at a
Supplemental File.
doctor’s office.

Demographics
Data analysis
Demographic items measured participant age, gender, race,
ethnicity, sexual identity, primary source of health insurance Data were cleaned and analyzed using the Statistical Package
coverage, and college student profile (eg, affiliation with a for the Social Sciences version 22.0 (SPSS, Inc., Chicago, IL,
social Greek organization, living location, and first-gener- USA). Demographic sub-groups (ie, gender, citizenship sta-
ation status). tus, student status, minority racial categories, sexual minor-
ity categories, living location, and insurance plan type) were
combined into dichotomous variables for analysis. Binary
Health insurance knowledge
variables were created for healthcare service utilization in
Health insurance knowledge was measured using the KFF the past 12 months to determine “any” versus “no” engage-
health insurance terminology quiz (see https://www.kff.org/ ment. Self-efficacy scores were calculated by averaging the
health-reform/poll-finding/assessing-americans-familiarity-with- response scale values across 21 items. In the current study,
health-insurance-terms-and-concepts/). This 10-item multiple- participants without missing data on the study variables
choice quiz measures health insurance literacy by testing (n ¼ 1,450) were included in the analysis. We constructed
respondent’s vocabulary and cost-calculation knowledge. Items
moderation models in logistic regression framework to
include questions such as “Which of the following is the best
determine the effect of the interaction of mean centered
definition of the term ‘health insurance premium’?” and “True
knowledge (main predictor) and self-efficacy (moderator)
or false: If your health insurance plan refuses to pay for a ser-
variables on healthcare engagement in the past 12 months at
vice that you think is covered and your doctor says you need,
the SHC or at any medical office (outcome); two separate
you can appeal the denial and possibly get the insurance com-
logistic regression models were estimated, one for each out-
pany to pay the claim.” “Don’t know” responses were coded
come variable. Key covariates (selected a priori) included
as incorrect, and the number of “correct” items was calculated
gender, age, being a first-generation college student, ever
for each respondent. We calculated a Kuder-Richardson
having a conversation with a parent about health insurance
Formula 20 (KR-20) score to measure internal consistency of
coverage, sexual identity, year in school, primary source of
the 10 dichotomous items.30 The KFF knowledge scale had
health insurance, being diagnosed with a chronic illness, liv-
acceptable internal consistency (q[KR-20] ¼ 0.73), indicating
ing location, ethnicity, race, citizenship status, social Greek
that the scores were reliable.
affiliation, and a self-defined preference for going to the
SHC for routine services. Prior to being included in the
Health insurance self-efficacy model, variables were assessed for multicollinearity; all vari-
Health insurance self-efficacy was measured using the ance inflation factors (VIFs) were within acceptable ranges.
Health Insurance Literacy Measure (HILM).22 The 21-item
survey conceptualizes health insurance literacy as choosing
Results
and using health insurance, both of which are divided into
confidence and likelihood of behavior subscales. For Sample characteristics
example, “How confident are you that you know how to
A total of 2,474 students participated in the survey (24.7%
estimate what you have to pay for your health care needs in
response rate), with 1,450 students answering all the ques-
the next year, not including emergencies?” Respondents
tions involved in the current analysis. Characteristics of the
indicated their likelihood or confidence of doing the specific
participants included in the current study are reported in
task through a 4-point Likert-type scale, with a “don’t
know” option; higher scores indicate higher self-efficacy. Table 1. Participants’ age range was 18–57 years old, with
The HILM has undergone previous psychometric evaluation an average age of 22.34 years old (SD ¼ 4.33 years). A
using a national probability sample of the adult population majority of the respondents were female (62.8%), white
in the United States.22 In the current study, the internal (65.3%), and undergraduate students (65.2%; see Table 1).
consistency score of the overall scale (Cronbach’s a ¼ 0.96) The majority of respondents (61.9%) were on their parents’/
indicated scores were reliable. guardians’ healthcare plan and had previously had a conver-
sation with their parents about their health insurance cover-
age (81.1%). In addition, most students had seen a
Healthcare service utilization healthcare provider at the SHC or other medical office in
Participants were asked to state their typical place to receive the past 12 months (55.8% and 85.6%, respectively).
healthcare services when they are sick and when they are Total health insurance knowledge ranged from 0 to 10,
receiving routine physicals. Items from the Center for with a mean of 5.8 (SD ¼ 2.4). Three of the ten health
4 T. JAMES ET AL.

Table 1. Sample characteristics. Table 2. Item difficulty of KFF health insurance knowledge scale.
Frequency (n) Percentage Item % Correct
Gender Definition of “health insurance premium” 79.5%
Female 910 62.8% Health insurance premium must be paid every month 78.5%
Male 536 37.0% Definition of “annual health insurance deductible” 67.4%
Transgender/genderqueer 4 0.3% Calculate out-of-pocket costs for hospital stay with deductible 48.3%
Race/ethnicity and copay
Biracial/multiracial 143 9.9% Definition of “annual out-of-pocket limit” 68.6%
American Indian or Alaskan Native 24 1.7% Definition of “health insurance formulary” 18.6%
Asian 252 17.4% Definition of a health plan “provider network” 80.9%
Black or African American 83 5.7% Not all doctors who provide care at in-network hospital may be 39.0%
Native Hawaiian or other Pacific Islander 1 0.1% in-network
White 947 65.3% Calculate out-of-pocket costs when insurer pays a portion of 24.8%
Hispanic/Latino(a) 312 21.5% allowed charges for out-of-network lab tests
Year in school Ability to appeal health plan denial 70.6%
Undergraduate 945 65.2% Note. KFF: Kaiser Family Foundation.
Graduate 505 34.8%
Sexual orientation
Straight 1307 90.1% type for students receiving preventive care, while 30.8%
LGBþ 143 9.9%
Social Greek organization member
reported using the SHC (see Supplemental Table 1). For
Yes 1116 77.0% nonpreventive visits, two-fifths (41.4%) of students preferred
No 334 23.0% to go to the SHC, followed by a doctor’s office (25.7%), or
First generation college student
Yes 432 29.8%
urgent care center (14.8%). However, 13% of students
No 1018 70.2% reported not going to a doctor’s office for preventive or
Living location nonpreventive care. In total, 46.9% of respondents said they
University housing 363 25.0%
Nonuniversity housing 1087 75.0% would go to the SHC for routine preventative services or
Health insurance plan when they are sick.
University-sponsored plan 386 26.6%
Parents’ health insurance 898 61.9%
Another plan 115 9.9%
I don’t have/not sure I have 23 1.6%
Health insurance knowledge, self-efficacy, and
Chronic disease diagnosis healthcare service utilization
Yes 190 13.1%
No 1260 86.9% To examine the moderating role of insurance self-efficacy,
Patient at student health center, past 12m we first examined use of healthcare services in physician offi-
Yes 809 55.8%
No 641 44.2% ces or the SHC in the past 12 months as the outcome. The
Patient at any healthcare office, past 12m model fit statistic was statistically significant, v2 ðdf ¼
Yes 1241 85.6% 18; n ¼ 1450Þ ¼ 152:829; p < 0:001; indicating the set
No 209 14.4%
Ever had conversation with parent about health insurance coverage of variables in the model were able to distinguish between
Yes 1176 81.1% respondents who had and had not engaged in healthcare in
No 274 18.9% the past 12 months. The model correctly classified 86.2% of
Looked at health insurance on the healthcare exchange/marketplace
Yes or current insurance 239 16.5% cases. As shown in Table 3, the interaction between self-effi-
No 966 66.6% cacy and knowledge was statistically significant indicating that
I don’t know what the Health Insurance 245 16.9% the relation between insurance knowledge and healthcare ser-
Marketplace/Exchange is.
Looked at university-sponsored health insurance vice use was moderated by insurance self-efficacy. Knowledge
Yes or current insurance 536 37.0% was not significantly related to healthcare utilization; however,
No 862 59.4%
I did not know the university has 52 3.6%
insurance self-efficacy was, such that higher self-efficacy was
sponsored plans for students related to greater probability of healthcare utilization. The sig-
Citizenship status nificant interaction between knowledge and self-efficacy indi-
U.S. citizen 1239 85.4%
Noncitizen 211 14.6% cates the relations between insurance knowledge and the
Permanent resident 44 3.0% probability to utilize healthcare service became stronger as
International student 162 11.2% self-efficacy increased. Significant covariates associated with
Undocumented resident 5 0.3%
increased odds of healthcare utilization were: being a U.S. citi-
zen, being diagnosed with a chronic disease, having a prefer-
insurance items had difficulties lower than 40% and, thus, ence to use the SHC for routine services or when sick, female
are considered more difficult items (see Table 2). Over two- gender, and Hispanic ethnicity; being a first-generation college
fifths (43.9%) of respondents answered seven or more of the student, and being on the school health insurance plan or not
questions correctly, indicating those students had higher having insurance were associated with decreased odds
health insurance knowledge. The mean score of the self-effi- of engagement.
cacy measure was 2.48 (SD ¼ 0.7). Health insurance know- The second model examined the use of SHC as the out-
ledge and self-efficacy were significantly, positively come. The fit statistic for this model was also statistically
correlated (Pearson’s r¼ 0.43, p < 0.001). significant, v2 ðdf ¼ 18; n ¼ 1450Þ ¼ 522:285; p < 0:001,
Going to an outside (non-SHC affiliated) doctor’s office and correctly classified 76.2% of SHC users. As shown in
or HMO was the most frequently (47.0%) reported facility Table 3, the interaction of self-efficacy and knowledge was
 JOURNAL OF AMERICAN COLLEGE HEALTH 5

Table 3. Moderation model of healthcare service utilization in any setting and at the SHC.
Any Healthcare Engagement, SHC engagement,
Predictors Past 12 months OR (95% CI) Past 12 months OR (95% CI)
Knowledge  self-efficacy 1.91 (1.35–2.72) 1.06 (0.98–1.13)
Knowledge (centered) 1.01 (0.93–1.09) 0.98 (0.91–1.04)
Self-efficacy (centered) 1.65 (1.29–2.12) 1.25 (1.02–1.53)
Female gender 1.73 (1.25–2.39) 1.40 (1.07–1.84)
Prefer SHC 1.91 (1.35–2.72) 15.96 (11.80–21.60)
U.S. citizen 2.64 (1.64–4.24) 2.14 (1.34–3.42)
Live in university housing 1.25 (0.83–1.86) 1.43 (1.04–1.95)
Social Greek member 1.52 (0.98–2.37) 1.20 (0.88–1.63)
First-generation college student 0.71 (0.51–1.00) 0.84 (0.63–1.12)
Undergraduate 1.05 (0.67–1.67) 1.45 (0.97–2.15)
Age 0.98 (0.93–1.02) 0.98 (0.94–1.03)
Hispanic ethnicity 1.64 (1.08–2.49) 1.21 (0.88–1.65)
Nonwhite race 0.70 (0.49–1.01) 0.80 (0.59–1.09)
Conversation with parent regarding health insurance 0.70 (0.46–1.07) 1.14 (0.81–1.62)
Sexual minority (LGBþ) 1.56 (0.86–2.81) 1.04 (0.67–1.62)
Diagnosed with chronic disease 2.08 (1.13–3.86) 1.28 (0.88–1.87)
Primary insurance: Don’t have or don’t know (ref: parents’ plan) 0.50 (0.31–0.82) 0.82 (0.52–1.29)
Primary insurance: School health insurance (ref: parents’ plan) 0.60 (0.38–0.95) 1.76 (1.19–2.62)
Constant 3.94 0.14
Note. p< 0.05.
SHC: Student Health Center.

nonsignificant indicating that self-efficacy did not moderate Importantly, our results indicate that health insurance lit-
the relation between knowledge and SHC utilization. eracy in this population was low. Similar to the KFF study,
Significant covariates associated with higher odds of engage- knowledge of health insurance terms and practices such as
ment were: having higher insurance self-efficacy, having a formularies, identifying in-network doctors, and calculating
preference to use the SHC for routine services or when sick, out-of-pocket costs had the highest item difficulties.23 In our
being a U.S. citizen, being on the school health insurance sample, the definition of “health insurance formulary” had
plan, living in university housing, and female gender. the highest item difficulty, with 18.6% of students correctly
identifying the definition; this finding was similar for the
national sample of respondents collected by KFF in that
Comment only 33% of the sample correctly identified the definition.23
Interestingly, Noble and colleagues28 found that 74% of col-
Previous studies among college student populations indicate
lege students correctly identified the vocabulary term of
that, despite reporting high rates of health insurance cover-
“formulary.” This difference may be attributed to sample
age, college students have low self-efficacy and knowledge differences or differences in item wording. Further, as
regarding health insurance.27,28 Low health insurance liter- observed in prior studies,22,25,27 the mean self-efficacy score
acy may influence healthcare utilization among college pop- was low, indicating lower confidence in selecting and using
ulations. Accordingly, this study explored the relation a health insurance plan. Results from our primary analyses
between health insurance knowledge and self-efficacy, and indicate that health insurance knowledge and self-efficacy
healthcare utilization in college students. are important factors when explaining college student
A majority (98%) of respondents in our sample reported healthcare utilization. As hypothesized, health insurance
having health insurance coverage; this is not surprising self-efficacy did moderate the relation between knowledge
given that, at the time of this study, health insurance man- and any healthcare utilization in the past 12 months.
dates were in effect at both the national and university level. Student likelihood to use healthcare service was higher for
In addition, a majority of respondents had previously had a the students with higher insurance self-efficacy than those
conversation with their parents regarding their coverage with lower self-efficacy. When we examined the SHC utiliza-
(81%); we did not assess the content of these conversations tion, self-efficacy did not moderate the relation between
but, depending on parental health insurance literacy and the knowledge the SHC utilization, although higher self-efficacy
depth of these conversations, they may have influenced stu- was associated with greater SHC utilization.
dent literacy and health utilization outcomes. In terms of
healthcare service utilization, however, respondents showed
differential patterns. Specifically, in the past year, 56% had Implications for practice
seen a provider at the SHC while 86% had seen a provider These results inform college student healthcare utilization
off-campus. It is unknown why 30% of the students who and highlight the growing importance of health insurance
had engaged in healthcare did not do so at the SHC. literacy in health service access. Having health insurance can
However, some students seem to understand the concept of decrease financial barriers to healthcare. Colleges and uni-
a preferred provider as almost half (47%) of respondents versities throughout the U.S. have instituted health insurance
said they would go to the SHC for routine preventive serv- mandates, complemented by the Affordable Care Act’s pre-
ices or when they are sick. vious individual mandate, aid in removing financial barriers.
6 T. JAMES ET AL.

In late 2017, the ACA’s mandate was repealed. Policy ana- health insurance literacy and healthcare utilization among
lysis indicates that the individual mandate provided “small college students. Future practice and research should focus
and inconsistent effects” in reducing the uninsured popula- on building health insurance self-efficacy and knowledge
tion in 2014 and 2015.31 Further, in states that failed to among college students using targeted messages across mul-
expand Medicaid, high uninsured rates continue among col- tiple communication channels.
lege aged students.32 Given that health insurance will no
longer be mandated at the national level, SHCs and health
promotion practitioners should advocate for a university- Conflict of interest disclosure
level insurance mandate for students to help reduce financial The authors have no conflicts of interest to report.
barriers to healthcare; however, this may also increase com-
petition for SHCs as students can see other providers cov-
ered by their insurance plans. ORCID
Increasing the insured population is not enough. Our Tyler G. James http://orcid.org/0000-0002-0694-4702
analyses indicate that health insurance literacy, specifically Leanne Dumeny http://orcid.org/0000-0003-3184-1235
self-efficacy, matters for healthcare service utilization among
a high insured population. These results have implications
for college health education and promotion practice; specif-
References
ically, health education specialists should design programs to
improve health insurance literacy to increase student health- 1. Uberoni N, Finegold K, Gee E. Health Insurance Coverage and
care engagement. In addition, our results provide a demo- the Affordable Care Act, 2010–2016. Washington D.C.:
Department of Health and Human Services, Office of the
graphic profile of college students who report
Assistant Secretary for Planning and Evaluation; 2016.
underutilization of SHCs and healthcare in general. More 2. Sommers BD, Blendon RJ, Orav EJ, Epstein AM. Changes in
efforts should be directed to increase service utilization rates utilization and health among low-income adults after Medicaid
in male students, students who are not U.S. citizens, and expansion or expanded private insurance. JAMA Intern Med.
those who do not live in university-owned housing. These 2016;176(10):1501–1509.
efforts of increasing health insurance coverage and health 3. Nikpay S, Freedman S, Levy H, Buchmueller T. Effect of the
Affordable Care Act Medicaid expansion on emergency depart-
insurance literacy among college populations may have posi- ment visits: evidence from state-level emergency department
tive effects on treatment utilization addressing prevalent databases. Ann Emerg Med. 2017;70(2):215–225.e6. doi:10.1016/
health disparities among college populations (eg, anxiety and j.annemergmed.2017.03.023
depression, substance misuse, sexual and reproductive health 4. Hernandez-Boussard T, Burns CS, Wang NE, Baker LC,
conditions) in addition to long-term positive effects on Goldstein BA. The Affordable Care Act reduces emergency
department use by young adults: evidence from three states.
healthcare service utilization and, in turn, health outcomes Health Aff Proj Hope. 2014;33(9):1648–1654. doi:10.1377/
over the students’ lifespan. hlthaff.2014.0103
5. Turner JC, Keller A. College health surveillance network: epi-
demiology and health care utilization of college students at US
Limitations and strengths 4-year universities. J Am Coll Health. 2015;63(8):530–538.
6. American College Health Association. American College Health
These findings should be considered in the context of the Association National College Health Assessment: Spring 2017
strengths and limitations of the current study. Our survey Reference Group Data Report. Hanover, MD; 2017.
relied on self-reported data related to healthcare utilization, 7. Bibbins-Domingo K, Pe~ na MB. Caring for the “young
which is subject to social desirability and recall bias. In add- invincibles.” J Gen Intern Med. 2010;25(7):642–643.
8. National Center for Health Statistics. Health, United States, 2016:
ition, the current study is cross-sectional, limiting the causal
With Chartbook on Long-Term Trends in Health. Hyattsville,
relations between the study variables. Strengths of our study MD: National Center for Health Statistics; 2017.
include a large sample of college students with diverse back- 9. Kindig DA, Panzer AM, Nielsen-Bohlman L. Health Literacy: A
grounds selected using a probabilistic sampling approach, Prescription to End Confusion. Washington, DC: National
and the inclusion of relevant socio-demographic and con- Academies Press; 2004.
10. Macabasco-O’Connell A, DeWalt DA, Broucksou KA, et al.
textual variables that allowed us to reduce omitted variable
Relationship between literacy, knowledge, self-care behaviors,
bias in our findings on the relations among the and heart failure-related quality of life among patients with heart
study variables. failure. J Gen Intern Med. 2011;26(9):979–986.
11. McKee MM, Paasche-Orlow MK, Winters PC, et al. Assessing health
literacy in Deaf American Sign Language users. J Health Commun.
Conclusion 2015;20(suppl 2):92–100. doi:10.1080/10810730.2015.1066468
12. Menendez ME, Mudgal CS, Jupiter JB, Ring D. Health literacy in
This study provides relevant information regarding con- hand surgery patients: a cross-sectional survey. J Hand Surg Am.
structs that influence healthcare utilization in college stu- 2015;40(4):798–804.
dents. Even when controlling for relevant socio- 13. Walker J, Pepa C, Gerard PS. Assessing the health literacy levels
of patients using selected hospital services. Clin Nurse Spec.
demographic characteristics, increased health insurance self-
2010;24(1):31–37.
efficacy was significantly associated with increased odds of 14. Mackert M, Champlin S, Mabry-Flynn A. Exploring college stu-
healthcare engagement. These insights can be used to justify dent health literacy: Do methods of measurement matter? J Stud
the development of health education programing to increase Aff Res Pract. 2017;54(3):1–11.
 JOURNAL OF AMERICAN COLLEGE HEALTH 7

15. Stellefson M, Hanik B, Chaney B, Chaney D, Tennant B, 23. Norton M, Hamel L, Brodie M. Assessing Americans’ Familiarity
Chavarria EA. eHealth literacy among college students: A sys- with Health Insurance Terms and Concepts. Menlo Park, CA:
tematic review with implications for eHealth education. J Med Kaiser Family Foundation; 2014. https://www.kff.org/health-
Internet Res. 2011;13(4):e102. reform/poll-finding/assessing-americans-familiarity-with-health-
16. Yin HS, Johnson M, Mendelsohn AL, Abrams MA, Sanders LM, insurance-terms-and-concepts/.
Dreyer BP. The health literacy of parents in the United States: A 24. Wong CA, Asch DA, Vinoya CM, et al. Seeing health insurance
nationally representative study. Pediatrics. 2009;124(Suppl 3): and HealthCare.gov through the eyes of young adults. J Adolesc
S289–S298. Health. 2015;57(2):137–143. doi:10.1016/j.jadohealth.2015.04.017
17. Hsu W, Chiang C, Yang S. The effect of individual factors on 25. Bartholomae S, Russell MB, Braun B, McCoy T. Building health
health behaviors among college students: the mediating effects of insurance literacy: evidence from the Smart Choice Health
eHealth literacy. J Med Internet Res. 2014;16(12):e287. InsuranceTM program. J Fam Econ Issues. 2016;37(2):140–155.
18. Cohen RA, Martinez ME. Health Insurance Covarage: Early Release 26. Quincy L. Early Consumer Testing of Actuarial Value Concepts.
of Estimates from the National Health Interview Survey, 2011. Washington, DC: Consumer Union & Kleinman Communication
Hyattsville, MD: National Center for Health Statistics; 2012. https:// Group; 2011.
www.cdc.gov/nchs/data/nhis/earlyrelease/insur201206.pdf. 27. Yang L. Young adults’ attitudes and perceptions on health insur-
19. Lau JS, Adams SH, Irwin CE, Ozer EM. Receipt of preventive ance and their health insurance literacy levels. May 2016. https://
health services in young adults. J Adolesc Health. 2013;52(1): cornerstone.lib.mnsu.edu/cgi/viewcontent.cgi?referer¼ https://
42–49. doi:10.1016/j.jadohealth.2012.04.017 www.google.com/&httpsredir¼1&article¼1616&context¼etds.
20. Kim J, Braun B, Williams AD. Understanding health insurance 28. Nobles AL, Curtis BA, Ngo DA, Vardell E, Holstege CP. Health
literacy: a literature review. Fam Consum Sci Res J. 2013;42(1): insurance literacy: a mixed methods study of college students. J Am
3–13. Coll Health. 2018;0(ja):1–37. doi:10.1080/07448481.2018.1486844
21. Quincy L. Measuring health insurance literacy: a call to action. 29. Levitt L. Why health insurance literacy matters. J Am Med Assoc.
Consumers Union in partnership with University of Maryland, 2015;313(6):555–556.
College Park and American Institutes of Research. 2012. https:// 30. Cortina JM. What is coefficient alpha? An examination of theory
consumersunion.org/pub/Health_Insurance_Literacy_Roundtable_ and applications. J Appl Psychol. 1993;78(1):98.
rpt.pdf 31. Frean M, Gruber J, Sommers BD. Premium subsidies, the man-
22. Paez KA, Mallery CJ, Noel H, et al. Development of the Health date, and Medicaid expansion: coverage effects of the Affordable
Insurance Literacy Measure (HILM): conceptualizing and meas- Care Act. J Health Econ. 2017;53:72–86.
uring consumer ability to choose and use private health insur- 32. Lookout Mountain Group. Update for Uninsured College Student
ance. J Health Commun. 2014;19(Suppl 2):225–239. Population. Bozeman, MT; 2018.