Lau JS, Adams SH, Park MJ, Boscardin WJ, Irwin CE. Improvement in Preventive Care of Young Adults After the Affordable Care ActThe Affordable Care Act Is Helping. JAMA Pediatr. 2014;168(12):1101-1106. doi:10.1001/jamapediatrics.2014.1691
The 2010 Affordable Care Act (ACA) included expansion of insurance coverage for young adults and improved access to preventive care.
To examine the ACA’s initial effects on young adults’ receipt of preventive care.
Design, Setting, and Participants
Secondary data analysis using a pre-post design that compared health care use by young adults (aged 18 to 25 years) from 2009 and 2011 Medical Expenditure Panel Surveys. Data were collected through computer-assisted personal interviews of a nationally representative sample of the noninstitutionalized US population.
Main Outcomes and Measures
Differences by year in rates of receiving a routine examination in the past year, blood pressure screening, cholesterol screening, influenza vaccination, and annual dental visit. Three logistic regression models were developed to (1) compare pre-ACA (2009) and post-ACA (2011) rates of receiving preventive care and (2) determine if post-ACA increases in insurance coverage accounted for changes in preventive care rates. Model 1 was a bivariate model to determine differences in preventive care rates by year; model 2, a multivariable model adding insurance status (full-year private, full-year public, partial-year uninsured, and full-year uninsured) to determine whether insurance accounted for survey year differences; and model 3, a multivariable model adding covariates (usual source of care and sociodemographic variables) to determine whether they further accounted for differences by survey year or insurance status.
After ACA, young adults had significantly higher rates of receiving a routine examination (47.8% vs 44.1%; P < .05), blood pressure screening (68.3% vs 65.2%; P < .05), cholesterol screening (29.1% vs 24.3%; P < .001), and annual dental visit (60.9% vs 55.2%; P < .001) but not an influenza vaccination (22.1% vs 21.5%; P = .70). Full-year private insurance coverage increased (50.1% vs 43.4%; P < .001), and rates of lacking insurance decreased (partial-year uninsured, 18.4% vs 20.7%; P = .03; and full-year uninsured, 22.2% vs 27.1%; P < .001). Full-year public insurance rates remained stable (9.4% vs 8.8%; P = .53). Insurance status fully accounted for the pre- and post-ACA differences in routine examination and blood pressure screening and partially accounted for year differences for cholesterol screening and annual dental visits. Covariate adjustment did not affect year differences.
Conclusions and Relevance
The ACA provisions appear to increase insurance coverage and receipt of preventive services among young adults. Further studies are needed to replicate these findings as other ACA provisions are implemented.
The 2010 Affordable Care Act (ACA) provides an unprecedented expansion of health insurance coverage to young adults. As of 2011, a total of 3 million young adults had gained insurance coverage through the provision that extends dependent private insurance coverage up to 26 years of age, resulting in a decrease in the uninsured rate among young adults from 42% in 2010 to 36% in 2011.1- 3 The uninsured rate for this group is projected to continue to fall with implementation of 2 additional provisions: Medicaid expansion within the states and insurance exchanges, which are estimated to expand coverage by 7.2 and 4.9 million additional young adults, respectively, by 2014.4
Most morbidity and mortality during young adulthood are preventable, and many health problems that develop during young adulthood, including obesity, mental health disorders, and infectious diseases, can become chronic and affect future health during a life course.5- 10 Evidence-based guidelines for young adults have provided specific recommendations for identifying and proposing treatment to prevent the most deleterious effects of emerging health conditions during early adulthood.11 Yet, young adults have historically had the lowest rate of health insurance coverage of any age group,12 decreasing the likelihood that they will use preventive care.13,14 The ACA provisions to increase insurance coverage for young adults hold promise for improving access to preventive care. Because of the ACA, a wide range of preventive services is available, with no copayment, to young adults with private insurance and those who are newly eligible through Medicaid (in states adopting the ACA Medicaid expansion). These include services recommended by the US Preventive Services Task Force, vaccinations recommended by the Advisory Committee for Immunization Practices,15 and additional preventive services for women’s health and well-being developed under the auspices of the Health Resources and Services Administration.16
Earlier work13 assessing the receipt of preventive care services in a representative sample of young adults in California found that fewer than 1 in 5 received influenza vaccination, only 1 in 5 received emotional health screening or diet or exercise counseling, about 2 in 5 received sexually transmitted disease screening, and 1 in 2 received cholesterol screening. Those who lacked health insurance had significantly lower rates of receiving these services.
Provisions in the ACA to increase insurance coverage among young adults have the potential to improve their health now and during their life course through increased access to health care and preventive care in particular. Using Medical Expenditure Panel Survey (MEPS) data from the years 2009 and 2011, the purpose of this study was 2-fold: (1) to determine whether rates of preventive care visits and services for young adults increased following the initial implementation of the ACA in 2010; and (2) to determine if increases in insurance coverage after ACA implementation accounted for any changes in preventive care rates. These findings can provide an early indication of changes in health care use in this age group.
We examined MEPS data from the 2009 and 2011 public-use files; the 2009 data represent pre-ACA status and the 2011 data represent the initial stage of ACA implementation. The MEPS is a nationally representative household survey of the US population that collects data on socioeconomic characteristics, health, health insurance, use of medical care services, and health expenditures. The household member who is most knowledgeable about the family’s health care use serves as the respondent for each family member. The MEPS respondents are enrolled for 2 years of data collection, with a new panel beginning each year.17 This study protocol was approved by the Committee on Human Research at University of California, San Francisco, under the exempt status.
Our analyses included the MEPS 18- to 25-year old young adult subsamples of the 2009 (subsample, n = 3768; full sample, N = 34 920) and 2011 (subsample, n = 3717; full sample, N = 33 622) data sets. The upper limit of 25 years of age was selected to be consistent with the age group identified in the ACA for extended health insurance coverage with dependent insurance coverage that became effective in 2010.
Study variables were preventive care visits and preventive services outcomes. Participants responded to questions regarding approximately how long it had been since they attended a routine examination or received a blood pressure or cholesterol screening from a physician or other health professional. They were also asked about 2 additional items: how long it had been since they had an influenza vaccination and, on average, how often they receive an annual dental visit (referred to as past-year dental check-up). These 5 items were recoded into having had a visit or service vs not in the past year.
The survey year was dummy-coded as 2009 or 2011. A 4-item insurance variable was developed for 2009 and 2011 based on the presence and type of insurance reported for each month of the year: (1) full-year private insurance; (2) full-year public insurance; (3) partial-year uninsured; and (4) full-year uninsured. If a respondent reported having had private insurance for each month of the year, they were coded as having full-year private insurance. Those who reported having public insurance for all 12 months were coded as having full-year public insurance. Those who reported having coverage for some but not all months were coded as partial-year uninsured and those who reported having no coverage for the 12 months were coded as full-year uninsured. A small number reported both private and public coverage for the 12-month period (Table 1), but this group is not included in the analyses of outcomes.
Access to care and demographic variables used in the analyses as covariates included having a usual source of care, sex, language spoken at home, race or ethnicity, income level, educational level, student status, and employment status. Participants reported whether they had a usual place where they receive medical care. They reported their sex as male or female. Language spoken at home was recoded into English or non-English. The race or ethnicity variables were recoded into the following: non-Hispanic white (referred to as white); non-Hispanic black (referred to as black); non-Hispanic Asian (referred to as Asian); Hispanic; American Indian or Alaskan Native, Native Hawaiian or Pacific Islander, and multiracial. Household income, presented as a percentage of the federal poverty level ($22 050 for a 4-person family in 2009 and $22 350 in 2011), was recoded into a 4-item federal poverty level index (<100%; 100 to <200%; 200 to <400%; and ≥400%). Educational level was recoded into less than 12 years vs 12 years or more. Student status was recoded to student (full- or part-time) vs nonstudent. For employment status, participants reported whether they had a job that provided their main source of income, which was recoded to employed vs unemployed.
All analyses were conducted using Stata, version 12 (StataCorp). We applied weights available in the MEPS data sets and accounted for the complex survey design, which allowed us to generate frequency estimates that are nationally representative of the noninstitutionalized civilian population and to develop estimates of the study variables. We conducted χ2 analyses to determine whether the independent variable (insurance status) and the covariates (access and demographic characteristics) varied significantly between 2009 and 2011 (Table 1).
Bivariate and multivariable logistic models were conducted to examine differences in the 5 outcomes between the pre-ACA (2009) and post-ACA (2011) periods (Tables 2, 3, and 4). Unadjusted differences in outcomes by survey year were determined through bivariate logistic regression analyses (model 1). In model 2, we added insurance status to the analysis to determine the extent to which insurance accounted for differences by survey year and its role in outcomes (ie, receipt of a routine examination and the 4 other preventive service outcomes). In model 3, we added covariate variables to determine whether they further accounted for differences by survey year or insurance level. Initially in the model development, we used the Wald test to determine whether the level of association between insurance categories and receipt of care or service outcomes varied significantly between the years. Because these interactions were not statistically significant, they were not retained in the models.
In the χ2 analyses of the demographic variables, only educational level varied significantly across the 2 years (Table 1). Rates of having at least a high school education were higher in 2011 compared with 2009 (81.2% vs 76.7%; P < .001). Having a usual source of care did not vary significantly by year. Insurance status varied significantly by year (P < .001). Tests of paired comparisons indicated that full-year private insurance increased from 43.4% to 50.1% (P < .001), partial-year uninsured status decreased from 20.7% to 18.4% (P < .03), and full-year uninsured status decreased from 27.1% to 22.2% (P < .001).
Rates of receiving a routine examination in the past year increased from 44.1% in 2009 to 47.8% in 2011 (P < .05) (Tables 2 and 3). When insurance was entered into the model (model 2), this difference was no longer significant, indicating that insurance coverage accounted for the significant difference between years for having had a routine examination (Table 3). Those with full-year public insurance were more likely to have attended a routine examination (P < .05) compared with those who had full-year private insurance, and those who were either partial- or full-year uninsured were less likely to attend a routine examination (P < .001 for both). Further adjusting for the covariates had no effect on year differences, and insurance status differences were somewhat attenuated.
A similar pattern was found for past-year blood pressure screening. Rates increased significantly from 65.2% to 68.3% in 2011 (P < .05). When insurance was entered into the model, the year was no longer significant. Those who were either partial- or full-year uninsured were less likely to have had their blood pressure screened (P < .001 for both) compared with those who had full-year private insurance. Controlling for covariates, the differences in insurance, while somewhat attenuated, remained significant.
Rates of having a cholesterol screening increased from 24.3% in 2009 to 29.1% in 2011 (P < .001). The degree of difference was somewhat attenuated but remained significant when insurance was entered into the model. Full-year uninsured young adults were significantly less likely to have their cholesterol screened (P < .001) compared with those who had full-year private insurance; these differences did not change when covariates were included in the models.
Rates of receiving the influenza vaccine did not vary significantly between 2009 (21.5%) and 2011 (22.1%) (P = .70). Those with full-year public insurance were more likely to have received the vaccine (P < .05) compared with those who had full-year private insurance, and full-year uninsured individuals were less likely to have received the vaccine (P < .001). When covariates were entered, the differences were greater between the individuals with full-year public and private insurance and did not change for the full-year uninsured.
Reports of receiving an annual dental visit increased from 55.2% to 60.9% (P < .001). When insurance was entered into the model, the degree of difference between years was attenuated but remained significant (P < .05). Compared with having full-year private insurance, all other insurance groups were less likely to report usually receiving an annual dental visit (P < .001). Controlling for covariates resulted in a slight increase in the degree of differences between 2009 and 2011, but the insurance differences either remained the same or were slightly attenuated.
This study demonstrates modest increases in rates of insurance coverage and receipt of preventive services in the past year among young adults in the year following enactment of the ACA. Consistent with recent studies,1- 4,18 there was a significant decrease in the percentage of young adults who lacked insurance for the entire year. To our knowledge, the increases in young adults’ receipt of routine examinations and preventive services after ACA found in this study have not been reported previously. This increase in insurance rates is primarily owing to the expansion of private dependent coverage to 26 years of age. Full implementation of ACA provisions, including those related to state-based marketplaces and Medicaid expansions (for states choosing this option), is likely to further expand coverage and has the potential to increase young adults’ use of preventive services.
These findings indicate that the expansion of insurance increased young adults’ receipt of preventive services. The finding that insurance accounted for the increase in young adults’ receipt of a routine examination in the past year suggests that young adults will take initiative to seek a routine examination when financial barriers are removed. The importance of insurance is further demonstrated by the finding that insurance accounted for the increase in receiving a blood pressure screening and accounted for part of the increases in receiving a cholesterol screening and annual dental visit. The inclusion of socioeconomic variables in the analysis had little effect on the significance of the relationship between insurance and the outcome measures of a routine examination, blood pressure screening, cholesterol screening, and annual dental visit. Those who were full- or partial-year uninsured fared worse on most or all outcomes compared with those who had full-year private insurance, replicating the results of previous studies on this age group.14,19 Taken together, these findings provide strong support for the critical role that insurance plays in moving young adults into the health care system for a general routine examination and receiving preventive services.
Surprisingly, there was no change in the receipt of influenza immunization in the past year. Given that this immunization is available in many non–health care settings, we would have expected an increase in insurance coverage to make this immunization a particularly easy service to which to gain access to. Research suggests that the decision to obtain an influenza vaccine is multifactorial,20- 22 including personal beliefs associated with the vaccine, understanding of influenza information, and concern about getting influenza.20,22 Given that none of the study variables appear to have affected the flat rate from 2009 to 2011, it could be that personal beliefs regarding vaccine necessity or efficacy have a strong influence on this outcome among young adults.
Our finding of an increase in annual dental visits in the past year suggests that better access to dental care may be an additional benefit of the ACA. This increase may be explained by the employer-based dependent coverage that likely accounted for most of the increase in young adults’ insurance rates. Employer-based benefits may also include dental insurance, linking expansion of medical insurance to expansion of dental insurance. It remains to be seen whether expansion of Medicaid (in states choosing this option) and the availability of insurance through the state-based insurance marketplaces will be linked to a similar increase in dental care.
One finding not directly related to our study questions requires further examination. Those with full-year public insurance were more likely to have received a routine examination in the past year than those with full-year private insurance. This difference was no longer significant, however, in the full model that accounted for all covariates. A brief analysis of the multivariable model suggests that race and sex covariates accounted for this difference, a finding that may be explained by the large portion of minority women with full-year public insurance. As Medicaid expansion proceeds in some states, it will be important to monitor whether this pattern continues.
This study has limitations. First, these findings represent only 1 year of data. Findings from subsequent years may show different results. For example, it is possible that the Medicaid expansion and initiation of state-based exchanges, which are being implemented in 2014, may result in a more dramatic increase in routine examinations. In addition, measures of the content of preventive care are limited to the measures monitored by MEPS. Including recommended measures in national data collection would improve monitoring of receipt of critical services. The cholesterol screening rate only gives a glimpse of whether young adults have received the test in the recent past but not according to the timing recommended by current practice guidelines. The MEPS method of identifying the household respondent as the person with the most knowledge about the health care use of everyone in the household may pose a source of bias for all household members’ responses other than the respondent. This potential bias may be particularly true in the case of young adults’ survey responses. Parents and caregivers may not be fully aware of their young adults’ health care use in cases of confidential care or when the young adult is living away from home during college. Finally, the employment status measure, a covariate in this analysis, does not distinguish between full- and part-time employment because this factor is not directly assessed in the MEPS.
These findings provide us with an early perspective on the initial effects of the ACA’s insurance expansion on young adults’ use of preventive care. The ACA provisions that aim to increase insurance rates among young adults and increase use of preventive care appear to have accounted for some increase in receipt of preventive services. Continued monitoring of young adults’ insurance status and use of preventive care will be needed to determine whether these findings are replicated as the full ACA implementation proceeds.
Accepted for Publication: July 15, 2014.
Corresponding Author: Josephine S. Lau, MD, MPH, Division of Adolescent and Young Adult Medicine, University of California, San Francisco, 3333 California St, Ste 245, San Francisco, CA 94118 (email@example.com).
Published Online: October 27, 2014. doi:10.1001/jamapediatrics.2014.1691.
Author Contributions: Dr Lau had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Lau, Adams, Irwin.
Acquisition, analysis, or interpretation of data: Lau, Park, Boscardin.
Drafting of the manuscript: Lau, Adams, Park, Irwin.
Critical revision of the manuscript for important intellectual content: Adams, Park, Boscardin, Irwin.
Statistical analysis: Lau, Boscardin.
Obtained funding: Park, Irwin.
Administrative, technical, or material support: Lau, Park, Irwin.
Study supervision: Adams, Irwin.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grants U45MC 00002 and U45MC 00023 from the Maternal and Child Health Bureau, Health Resources and Services Administration, US Department of Health and Human Services, and University of California, San Francisco, Clinical and Translational Science Institute grant UL1 TR000004 from the National Center for Advancing Translational Sciences, National Institutes of Health.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.