Does a Free Office Visit Affect Primary Care–Seeking Behavior? A Study of New Exchange Health Plan Enrollees in Mississippi

In recent years, several health plans in the Health Insurance Marketplace exchange were offering free doctors’ visits to their members, with the hope of identifying illnesses before they become more difficult or more expensive to treat (eg, antibiotics prescribed in a physician’s office could prevent a future hospitalization for pneumonia).¹ Although the rationale for free visits is sound, this is the first study to report empirical evidence that assesses the impact of a free visit offer on primary care–seeking behavior.

The positive association between utilization of primary care and overall health is well-established.^2-6 Not surprisingly, adults with a usual source of care are more likely to receive preventive services, such as the influenza vaccine.^2,7 Emergency department use is also lower for people who regularly visit their doctor.^2,8 Furthermore, those who experience the interpersonal continuity of care provided in primary care consistently have greater satisfaction with their healthcare.⁹

From a population perspective, populations in geographies with more primary care providers (PCPs) have better health, lower mortality, and lower healthcare costs.³ Yet, limited access to care continues to prevent communities from obtaining the benefits of primary care. In 2013, only 33% of uninsured adults reported a preventive care visit within the year compared with 74% of adults with employer coverage and 67% of adults with Medicaid.¹⁰ Given the known benefits of primary care, efforts to improve utilization have the potential to improve health outcomes, particularly in regions with low primary care use, significant health needs, or large uninsured populations who have limited access to care.

The Affordable Care Act (ACA) created the exchange, whereby individuals can directly purchase health insurance from a choice of health plans that have met federally mandated requirements and offer comparable benefit coverage. Although the patients’ health status and previous encounters with the healthcare system were not known at the time of enrollment in the inaugural year of the exchange, enrollees were expected to be uninsured or underinsured. Insuring the uninsured provided an opportunity to increase primary care use. As such, Humana Inc, an insurer with coverage in 40 of the 82 counties in Mississippi and the only health insurance option in 36 of those counties, developed a program that extended a free PCP office visit offer to new exchange enrollees. The opportunity to help establish patient–physician relationships was especially high in Mississippi, and only 74% of the state’s residents reported having a usual source of healthcare in 2012.¹¹

This study evaluated this free office visit offer by comparing primary care–seeking behaviors in the exchange population in Mississippi with similar populations in neighboring states who were not extended a free visit offer during the first year (2014) of the exchange health plans.

Methods
Study Design and Objectives

We conducted a nonrandomized concurrent control study using historical medical claims data from January 1, 2014, to December 31, 2014. The study objectives were to assess the odds of a PCP visit among enrollees who were offered a free nonpreventive visit compared with enrollees who did not receive the offer, and to measure the frequency of contacts with the healthcare system in sites of care that may be used in lieu of PCP office visits (ie, emergency department, urgent care centers, and retail-based clinics).

During the first months of coverage, the health plan mailed “Get your first doctor’s office visit for free” letters (Flesch-Kincaid readability test grade level 6.7) directly to all new enrollees in Mississippi, notifying them that there would be no payment required for their first office visit before June 30, 2014 which was eventually extended to August 31, 2014, to align with the extension in exchange enrollment (see Appendix at the end of this article). A similar letter was sent to PCPs to prepare them to answer any patient questions regarding the free visit offer.

Both letters specified that the free office visit applied to nonpreventive visits, because preventive services are mandated to be provided without patient out-of-pocket responsibility under the ACA.¹² For example, well-child visits are always covered without patient cost-sharing; however, a visit prompted by symptoms of an ear infection generally involves a patient cost. Under the free visit offer, the patient cost for the ear infection visit would be waived.

Visits billed using a Current Procedural Terminology (CPT) code indicative of an office or other outpatient evaluation and management service (CPT codes 99201-99205 and 99211-99215) were eligible.¹³ Automated claims processing determined the patient cost-sharing amount. Claims audits identified the first visit for a patient that included cost-sharing (ie, nonpreventive), and the patient cost-sharing was removed.

Inclusion Criteria

All Mississippi residents who used the exchange to enroll in a health insurance plan provided by Humana, between January 1, 2014, and August 31, 2014, were eligible for the free office visit and were included in the analysis. Eligible individuals insured by Humana before January 1, 2014, were excluded. No information was available on whether new enrollees had other health insurance before January 1, 2014.

To form a concurrent control group with a rural-­urban distribution comparable with Mississippi, the popu­lation demographics of exchange enrollees in Humana plans in other states were evaluated. Such comparisons are common in observational studies that evaluate the impact of state-specific policies.^14-18 Tennessee provided the closest comparator for rurality; however, the sample size was less than 25% of that of Mississippi. Thus, we selected Georgia as a second comparator state with a larger sample size. Georgia and Tennessee are from the same geographic region and have similar demographics, state health rankings, and unmet health needs as Mississippi.^11,19 Chronic disease rates have been consistently high among these states; however, cost presents a significant barrier to care for a considerable proportion of adults in Mississippi (22%), Tennessee (19%), and Georgia (20%).^11,20

Outcome Variable

The primary outcome of this study was the utilization of nonpreventive PCP visits. PCP visits were identified using medical claims that contain dates of service, sites of service, provider type, International Classification of Diseases, Ninth Revision, Clinical Modification, and CPT codes for all inpatient and outpatient encounters. The percentage of enrollees who had at least 1 nonpreventive PCP visit during the free visit period (January 1, 2014-August 31, 2014) and during the 2014 calendar year (January 1, 2014-December 31, 2014) was computed.

The number of nonpreventive PCP visits per 1000 persons was also calculated for the calendar year and was weighted by length of follow-up, because people could have been enrolled in the health plan from 1 month to 12 months of duration in the study. The rates weighted for follow-up time are expressed in number of visits per 1000 person-years. The level of medical decision-making complexity for the first nonpreventive PCP office visit was reported based on the CPT code reported on the medical claim. The mean time to the first PCP visit (preventive or nonpreventive) was also reported during the calendar year.

Visits to other sites of care included emergency department, urgent care, and retail-based clinics, and were reported as visits per 1000 person-years. Emergency department visits were stratified by emergent versus nonemergent visits based on diagnosis and billing codes. Retail-based clinics are defined as walk-in medical facilities typically located in pharmacies, grocery stores, and retailers that provide basic medical care almost exclusively by nurse practitioners or physician assistants.^21,22

Covariates

Our study attempted to control for underlying differences between the Mississippi and the concurrent control groups using regression modeling and covariates available in the data. Age, sex, the plan “metal type,” the rural-urban designation, and the month of first enrollment (January-June) were all considered potential covariates. The plan metal types ranged from platinum to bronze, as defined by the ACA (ie, platinum, gold, silver, bronze, catastrophic).¹² Platinum plans require the highest monthly premiums and pay for 90% of medical care costs, whereas bronze plans require lower monthly premiums and pay for 60% of medical care costs.¹²

Catastrophic health plans pay for a lower percentage of medical care costs, but they require the lowest monthly premium. The rural-urban commuting area (RUCA) codes were used to develop the rural-urban designation. The RUCA classification uses whole numbers (1-10) to delineate metropolitan, micropolitan, small town, and rural commuting areas based on the size and direction of the primary (largest) commuting flows.²³ For the purpose of this study, these 10 codes were collapsed into the 4 rural-urban categories of urban, suburban, large town rural, and small town rural.

Statistical Analyses

The demographic characteristics were reported using descriptive statistics. To understand the baseline differences between the state groups that are not attributable to the free office visit letter, categorical variables (eg, sex, urban-rural designation) were compared using chi-square tests of significance and continuous variables (eg, mean age) were compared using analysis of variance.

The utilization of nonpreventive PCP visits was analyzed in 2 ways. First, event/trials logistic regression models, which involve binary data and account for multiple events over time, were used to assess the odds of a nonpreventive PCP visit in Mississippi compared with Georgia and Tennessee combined and by state. To account for the varying length of follow-up, the number of nonpreventive PCP visits was transformed (ie, nonpreventive PCP visits/number of days enrolled). Multiple linear regression models were used to assess the differences in the mean number of days to the first nonpreventive PCP visit between the groups. To meet the appropriate statistical assumptions, a square root transformation was performed; multiple testing was controlled for using Tukey’s comparisons. All models were adjusted by covariates.

This study was approved by the Schulman Institutional Review Board. All analyses were completed using SAS Enterprise Guide version 5.1 (SAS Institute, Inc; Cary, NC).

Results

A total of 39,121 enrollees met the study criteria in Mississippi; 225,877 in Georgia; and 8703 in Tennessee (Table 1). The age and sex distributions were similar across the 3 states. Rurality differed substantially by state. The plan metal types were similar in Mississippi and Georgia, whereas Tennessee had a larger proportion of catastrophic and gold plans. People in Mississippi tended to enroll later in the calendar year compared with people from the other states (ie, the month of enrollment).

In all, 12,576 (32.1%) Mississippians visited their PCP for a nonpreventive service during the free visit offer period (Figure 1). Nonpreventive PCP visits were significantly less frequent in Mississippi than in Tennessee, but were significantly more frequent than in Georgia during the free visit period and the full calendar year (P <.05; Figure 1). Reflecting the same directional trend, the rate of nonpreventive PCP visits per 1000 person-years was 1876.1 in Mississippi, 1642.7 in Georgia (P <.05 vs Mississippi), and 2598.1 in Tennessee (P <.05 vs Mississippi). The percentage of nonpreventive PCP visits that were coded as moderate or high complexity was highest in Georgia (63.2%), followed by Tennessee (53.0%) and Mississippi (44.3%).

After adjusting for confounders, the odds of a nonpreventive PCP visit were not significantly different between Mississippi and the Georgia-Tennessee group during the free visit period (odds ratio [OR], 0.99; 95% confidence interval [CI], 0.97-1.00; P = .11; Table 2). Similarly, there were no meaningful differences after extending to 1 year of follow-up (OR, 0.99; 95% CI, 0.98-1.00; P = .02). Individual comparisons of each state showed no difference between Mississippi and Georgia, but there were lower odds of nonpreventive PCP visits for Mississippi versus Tennessee.

The adjusted mean number of days from enrollment to the first PCP visit (preventive or nonpreventive) ranged from 56.3 (95% CI, 53.9-58.8) in Tennessee to 71.7 (95% CI, 69.9-73.5) in Georgia and 72.3 (95% CI, 70.3-74.3) in Mississippi. Compared with Mississippi, the time to the first PCP visit was significantly shorter in Tennessee (P <.0001), but did not differ significantly in Georgia (P = .673).

Relative to the Georgia-Tennessee group, patients in Mississippi had a significantly higher rate of emergent and nonemergent emergency department visits per 1000 person-years and lower rates of urgent care and retail clinic visits per 1000 person-years (P <.001 for all comparisons). In state-by-state comparisons, emergent and nonemergent emergency department visits were significantly higher in Mississippi than in Georgia, but there was less of a difference between Mississippi and Tennessee (Figure 2).

Adjusting for confounders, the odds of an emergent emergency department visit were 33% greater in Mississippi than in the Georgia-Tennessee group (OR, 1.33; 95% CI, 1.28-1.39), and the odds of a nonemergent emergency department visit were 40% greater in Mississippi (OR, 1.40; 95% CI, 1.35-1.44) during the calendar year. Conversely, patients in Mississippi were less likely to visit an urgent care center (OR, 0.10; 95% CI, 0.09-0.11) or retail clinic (OR, 0.13; 95% CI, 0.11-0.17) than patients in the Georgia-Tennessee group.

Discussion

Despite being eligible for a free nonpreventive visit, Mississippi enrollees were no more likely to visit a PCP than their Georgia-Tennessee counterparts. In fact, they were significantly less likely to visit a PCP than people in Tennessee after controlling for confounders. Given the complexity of barriers to obtaining healthcare services, this study suggests that removing only 1 barrier—in this case, cost—may not be sufficient to change behavior.

Although there are no known studies to date that assessed the impact of a free initial nonpreventive primary care visit, the relationship between patient cost-sharing and health service utilization has been investigated for more than 3 decades. The landmark RAND Health Insurance Experiment concluded that participants exposed to cost-sharing used fewer health services.²⁴ Similarly, observational assessments of increased copays for office visits have shown a reduction in such services.^25,26

However, these studies also revealed an unintended consequence of increased inpatient hospitalizations. Although the RAND study showed no differences in the quality of care received, it set the stage for reducing cost-sharing for high-value services.²⁴ In turn, several studies have reported that lowering or removing copays for maintenance medications is associated with improved adherence.^27-29

A 2008 study evaluated correlations between cost-sharing and mammography rates, concluding that even small increases in cost-sharing were associated with significantly lower mammography rates, especially in geographic areas with lower education and income.³⁰ The elimination of cost-sharing requirements for preventive services by the ACA underscores the relationship between costs and patient care. For example, a recently published study examined mammography use before and after Medicare eliminated cost-sharing and reported a modest 3.5% adjusted absolute increase in screenings.³¹

This body of evidence suggests that removing the copay for an initial nonpreventive PCP visit should result in increased primary care visits; yet, the lack of difference in primary care use between Mississippi and the concurrent control group of other state exchange populations is inconsistent with this hypothesis. Accordingly, other barriers beyond direct patient costs may explain these null findings.

One possible explanation for the lower-than-expected PCP visits is access barriers. Mississippi has one of the lowest rates of primary care physicians per capita in the country, with 59% of the counties in Mississippi designated as health professional shortage areas.³² Thus, obtaining an appointment with a physician may have been difficult. Although the primary care visit was free, child care, transportation, and time away from work were not, and therefore may have inhibited care-seeking by low-income families.

These findings also provide insight into primary care–seeking behaviors in all 3 states. Primary care utilization was highest in Tennessee and was lowest in Georgia; enrollees in Tennessee went to their doctors sooner after enrollment than their Mississippi and Georgia counterparts. After primary care offices, emergency departments were the most common site of care.

Mississippians were much more likely to visit the emergency department than retail or urgent care clinics. This suggests that these other sites of care may be either less common in Mississippi or at further distances from residents. Although Tennessee had the highest rates of primary care, this did not translate to the lowest rates of emergency department visits among the 3 states. Enrollees in Georgia made greater use of urgent care facilities and used the emergency department less frequently than did participants in the other 2 states.

Limitations

This study has limitations to its internal and external validity. Paper mailings were selected for this population based on local recommendations; however, it is not known what percentage of the Mississippi participants actually read the free offer letter. Therefore, the study design took a population-based approach to evaluate the impact of the free visit offer holistically.

The lack of randomization and the concurrent control group design mean that the study groups are not strictly comparable. Although the analysis controlled for confounders represented by the available data, control for other variables shown to affect the use of health services was not possible; additional factors include prevalent comorbidities, race, income, education level, health literacy level, employment, distance to medical facilities, and many other social determinants of health.

If factors that could have a negative impact on primary care–seeking behavior were more common in Mississippi than in the comparator states, it is possible that the study underestimated the actual difference. However, given the other contributors to healthcare-seeking behavior that have been identified in the literature, any true effect from the free visit offer may be relatively small.

Furthermore, the study design only allows the measurement of association, and does not permit causal inferences.

Finally, the generalizability of this study is limited by the nature of the population assessed—members of an exchange plan within a medically underserved state with a high prevalence of chronic diseases.

Conclusions

Despite being eligible for a free nonpreventive visit, adults enrolled in an exchange plan in Mississippi were no more likely to have a nonpreventive PCP office visit than their Georgia-Tennessee counterparts. The prevailing perspective in healthcare policy suggests that removing cost-sharing should increase the utilization of healthcare services, yet this viewpoint did not hold true in a population with substantial socioeconomic and healthcare barriers. Thus, the removal of direct patient cost alone may not be sufficient to overcome the full range of cultural and socioeconomic barriers to healthcare access. These findings should be used to inform innovative public health and policy approaches to improving primary care access.

Despite the study’s limitations, the lack of a demonstrated effect suggests that health plans and PCPs should work to identify and solve additional barriers to healthcare access beyond direct patient cost. Health plans that are offering similar programs should rigorously evaluate the impact of these programs to build the evidence base to include populations with fewer access and health issues. Furthermore, some of the offerings provide all PCP visits at no cost to the patient, which may have a stronger influence on primary care–seeking behaviors.¹

Appendix Excerpt from Letter

Get your first doctor’s office visit for free

We care about your health. So we will pay for your first office visit with a family doctor if it occurs by June 30, 2014 (later extended to August 31, 2016). Make sure your family doctor takes your Humana plan. Covered services will be paid at 100%.

Note:

• Our offer does not cover any tests (like x-rays or blood work) the doctor may order unless they are preventive. Of course, your first preventive exam is covered at 100%. This is true any time.
• Visits to an emergency room and urgent care centers are not covered by this offer. Our offer does not cover specialist visits.
• When Humana gets claims from your doctors, we usually send you a form called an EOB (Explanation of Benefits). You may get 2 EOBs for this doctor visit. This first EOB will show how your benefits would have paid without our offer. The second one will show that Humana paid at 100%. Please ignore the first EOB.

We hope you use this chance to improve your health. Please share this letter with your doctor. Let your doctor know that Humana will pay at 100% for covered services.

Source of Funding
This study was funded by Humana Inc.

Author Disclosure Statement
Mr Cordier is a stockholder of Humana. Dr Beech, Dr Happe, Dr Trunk, Mr Haugh, Mr Kwong, Dr Gopal, and Dr Beveridge reported no conflicts of interest.

Dr Beech is Executive Director, Myrlie Evers-Williams Institute, Associate Vice Chancellor for Population Health, and Professor, Family Medicine and Pediatrics, University of Mississippi Medical Center, Jackson; Mr Cordier is Manager, Clinical Analytics, Humana, Louisville, KY; Dr Happe is Director of Research and Publications, Office of the Chief Medical Officer, Humana; Dr Trunk is Medical Director, Contracting Negotiations, Humana; Mr Haugh is Director, Clinical Analytics, Humana; Mr Kwong was Analytics Consultant, Clinical Analytics, Humana, at the time of the study; Dr Gopal is Vice President, Clinical Analytics, Humana; Dr Beveridge is Chief Medical Officer, Humana.

Preliminary findings from this study were presented as a poster at the 2015 AcademyHealth Annual Research Meeting.

References
1. Galewitz P. Obamacare insurers sweeten plans with free doctor visits. Kaiser Health News. January 4, 2016. http://khn.org/news/obamacare-insurers-sweeten-plans-with-free-doctor-visits/. Accessed January 11, 2016.
2. Friedberg MW, Hussey PS, Schneider EC. Primary care: a critical review of the evidence on quality and costs of health care. Health Aff (Millwood). 2010;29:766-772.
3. Starfield B, Shi L, Macinko J. Contribution of primary care to health systems and health. Milbank Q. 2005;83:457-502.
4. Beasley JW, Starfield B, van Weel C, et al. Global health and primary care research. J Am Board Fam Med. 2007;20:518-526.
5. Boukus ER, Cunningham PJ. Mixed signals: trends in Americans’ access to medical care, 2007-2010. Track Rep. 2011:1-6.
6. Starfield B. Insurance and the U.S. health care system. N Engl J Med. 2005;353:418-419.
7. Blewett LA, Johnson PJ, Lee B, Scal PB. When a usual source of care and usual provider matter: adult prevention and screening services. J Gen Intern Med. 2008;23:1354-1360.
8. Petersen LA, Burstin HR, O’Neil AC, et al. Nonurgent emergency department visits: the effect of having a regular doctor. Med Care. 1998;36:1249-1255.
9. Saultz JW, Albedaiwi W. Interpersonal continuity of care and patient satisfaction: a critical review. Ann Fam Med. 2004;2:445-451.
10. Garfield R, Licata R, Young K. The uninsured at the starting line: findings from the 2013 Kaiser Survey of Low-Income Americans and the ACA. Kaiser Family Foundation. February 6, 2014. http://kff.org/uninsured/report/the-uninsured-at-the-starting-line-findings-from-the-2013-kaiser-survey-of-low-income-americans-and-the-aca/. Accessed November 17, 2015.
11. Radley D, McCarthy D, Lippa J, et al. Aiming higher: results from a scorecard on state health system performance, 2014. The Commonwealth Fund. April 30, 2014. www.commonwealthfund.org/publications/fund-reports/2014/apr/2014-state-scorecard. Accessed November 17, 2015.
12. Mach AL, Fernandez B. Private health insurance market reforms in the Affordable Care Act. February 10, 2016. www.fas.org/sgp/crs/misc/R42069.pdf. Accessed November 17, 2015.
13. American Medical Association. Current Procedural Terminology: 2015 Standard Edition. Parsippany, NJ: American Medical Association; 2014.
14. Sommers BD, Maylone B, Nguyen KH, et al. The impact of state policies on ACA applications and enrollment among low-income adults in Arkansas, Kentucky, and Texas. Health Aff (Millwood). 2015;34:1010-1018.
15. Knudsen HK, Lofwall MR, Havens JR, Walsh SL. States’ implementation of the Affordable Care Act and the supply of physicians waivered to prescribe buprenorphine for opioid dependence. Drug Alcohol Depend. 2015;157:36-43.
16. Zur J, Mojtabai R. Medicaid expansion initiative in Massachusetts: enrollment among substance-abusing homeless adults. Am J Public Health. 2013;103:2007-2013.
17. Blum AB, Kleinman LC, Starfield B, Ross JS. Impact of state laws that extend eligibility for parents’ health insurance coverage to young adults. Pediatrics. 2012;129:426-432.
18. Sabik LM, Tarazi WW, Bradley CJ. State Medicaid expansion decisions and disparities in women’s cancer screening. Am J Prev Med. 2015;48:98-103.
19. United Health Foundation; American Public Health Association. America’s Health Rankings Annual Report: A Call to Action for Individuals and Their Communities. December 2015. www.americashealthrankings.org/reports/annual. Accessed December 17, 2015.
20. Mississippi State Department of Health. 2013 Behavioral Risk Factor Surveillance System annual prevalence report. December 18, 2014. www.msdh.state.ms.us/brfss/brfss2013ar.pdf. Accessed November 17, 2015.
21. Bohmer R. The rise of in-store clinics—threat or opportunity? N Engl J Med. 2007;356:765-768. Erratum in: N Engl J Med. 2007;356:2437.
22. Hansen-Turton T, Ryan S, Miller K, et al. Convenient care clinics: the future of accessible health care. Dis Manag. 2007;10:61-73.
23. United States Department of Agriculture Economic Research Service. Rural-urban commuting area codes. www.ers.usda.gov/data-products/rural-urban-commuting-area-codes.aspx. Accessed November 17, 2015.
24. RAND Health. The Health Insurance Experiment: a classic RAND study speaks to the current health care reform debate. 2006. www.rand.org/content/dam/rand/pubs/research_briefs/2006/RAND_RB9174.pdf. Accessed November 5, 2015.
25. Trivedi AN, Moloo H, Mor V. Increased ambulatory care copayments and hospitalizations among the elderly. N Engl J Med. 2010;362:320-328.
26. Chandra A, Gruber J, McKnight R. Patient cost-sharing and hospitalization offsets in the elderly. Am Econ Rev. 2010;100:193-213.
27. Eaddy MT, Cook CL, O’Day K, et al. How patient cost-sharing trends affect adherence and outcomes: a literature review. P T. 2012;37:45-55.
28. Nair KV, Miller K, Saseen J, et al. Prescription copay reduction program for diabetic employees: impact on medication compliance and healthcare costs and utilization. Am Health Drug Benefits. 2009;2(1):14-24.
29. Chernew ME, Shah MR, Wegh A, et al. Impact of decreasing copayments on medication adherence within a disease management environment. Health Aff (Millwood). 2008;27:103-112.
30. Trivedi AN, Rakowski W, Ayanian JZ. Effect of cost sharing on screening mammography in Medicare health plans. N Engl J Med. 2008;358:375-383.
31. Sabatino SA, Thompson TD, Guy GP Jr, et al. Mammography use among Medicare beneficiaries after elimination of cost sharing. Med Care. 2016;54:394-399.
32. US Department of Health & Human Services Health Resources & Services Administration. HPSA find. http://datawarehouse.hrsa.gov/tools/analyzers/hpsafind.aspx. Accessed September 10, 2015.