Abstract
We use data from the 2012-2015 TRICARE Standard Survey to examine factors that affect civilian health care providers’ acceptance of patients covered by the U.S. Department of Defense’s TRICARE insurance program and Medicare. We find that 74% of physicians report that they accept new TRICARE patients compared with 83% accepting new Medicare patients; in contrast, only 36% of mental health providers report that they accept new Medicare and/or TRICARE patients. Among the most common reasons provided by both physicians and mental health providers for not accepting either insurance type are insufficient reimbursement or their specialty not being covered; lack of awareness of TRICARE is also frequently cited, particularly among mental health providers. These findings suggest that successful strategies to increase provider acceptance of TRICARE and Medicare may include improving reimbursement rates and specialty coverage and increasing provider awareness of TRICARE through outreach programs.
Introduction
Many Americans rely on public health insurance programs for their access to health care, but physician acceptance of these programs is sometimes an issue. In this article, we consider two such programs—TRICARE and Medicare—and their acceptance by civilian health care providers. In fiscal year 2016, TRICARE served 9.4 million military personnel and their family members through several coverage options, including a health maintenance organization (HMO) plan called TRICARE Prime and a preferred provider organization (PPO) option through the TRICARE Standard and Extra plans (Defense Health Agency, Decision Support Division, 2017). About 1.9 million adult beneficiaries, who were primarily family members of active duty personnel, and military retirees and their family members, could access care through TRICARE’s PPO option (Defense Health Agency, Decision Support Division, 2017). In contrast, Medicare provides care primarily to individuals aged 65 years and older through traditional fee-for-service Medicare or through Medicare Advantage HMO and PPO plans.
Studies have found that acceptance of new Medicare patients is high among physicians and is comparable to acceptance rates of new patients with private noncapitated insurance (Bishop et al., 2014; Cunningham et al., 2006; Medicare Payment Advisory Commission, 2012). Yet, despite reimbursement, coverage, and provider eligibility similarities between TRICARE and Medicare, concerns exist that some providers are less willing to accept new TRICARE patients compared with Medicare patients (Government Accountability Office, 2013). Recent work has also shown geographic variation in acceptance and reported access under TRICARE (Ben-Shalom et al., 2019). In this article, we use data from a congressionally mandated survey of civilian physicians to improve understanding of factors that affect access under TRICARE and the implications for access under other public insurance programs by comparing providers’ acceptance of new patients under TRICARE to acceptance under Medicare.
TRICARE Health Plans
Two primary health plan options are available to TRICARE beneficiaries: TRICARE Prime, which is a point-of-service HMO that covers treatment provided by military or civilian members of TRICARE’s provider network, and a PPO option, which offers fee-for-service coverage with enhanced cost sharing for services from network providers. TRICARE’s regional health care contractors are obligated to maintain the TRICARE network only within Prime Service Areas, which tend to be close to military facilities where there is a concentration of eligible beneficiaries. Active duty personnel and their family members are automatically eligible for free enrollment in Prime, and military retirees younger than age 65 may enroll by paying an annual enrollment fee. Family members of active duty personnel (including family members of activated reservists) and military retirees and their dependents are eligible for the PPO option. Before 2018, they could use the PPO, known as TRICARE Standard (for nonnetwork coverage) and Extra (for network coverage), without any requirement to enroll. On January 1, 2018, TRICARE Select replaced TRICARE Standard and Extra, offering similar benefits but requiring enrollment.
TRICARE’s PPO option covers care from all qualified, TRICARE-approved providers, including those not part of TRICARE’s civilian network. TRICARE nonnetwork providers receive reimbursement according to a fee schedule similar to Medicare’s, which often is lower than reimbursement from private payers (Lopez et al., 2020). Providers in TRICARE’s civilian network contract to provide services at a negotiated rate that is usually lower than reimbursement received by nonnetwork providers and has lower cost-sharing for beneficiaries, making network providers more affordable than nonnetwork providers. PPO beneficiaries must meet an annual deductible each fiscal year before TRICARE begins to pay. After the deductible, users pay a cost share based on the types of care and provider they see (network vs. nonnetwork). The deductible and cost share amounts vary by military status.
Past Studies of Physician Insurance Acceptance
Multiple studies have found that acceptance of new Medicare patients is high among physicians and has been relatively stable in recent years. Cunningham et al. (2006) found that approximately 73% of physician practices accepting any new privately insured patients were accepting all new Medicare patients and almost 97% were accepting at least some new Medicare patients in 2004-2005, with similar acceptance rates in 1996-1997. Bishop et al. (2014) found that 88% of physicians accepted new Medicare patients in 2005-2006 and 86% in 2009-2010. In both studies, trends in acceptance of privately insured patients are similar to those of Medicare patients, indicating that broader physician-related constraints (such as provider shortages or time constraints) may be a major factor in Medicare acceptance rates rather than Medicare-specific factors (such as reimbursement rates). Studies based primarily on a series of congressionally mandated surveys of providers and TRICARE beneficiaries known as the TRICARE Standard Surveys have tracked acceptance of new TRICARE patients since 2005 and found lower acceptance rates among physicians for TRICARE than for Medicare (Government Accountability Office, 2011, 2013).
Policy makers, government agencies, and health insurance beneficiaries have expressed considerable interest in the factors that affect acceptance rates of new patients with public or private health insurance (Government Accountability Office, 2013; Hing et al., 2015; Holgash & Heberlein, 2019). Several studies have found that large group practices are more likely than practices with one or two physicians to accept new patients and that acceptance of new patients is lower for physicians who have been practicing longer (Adams & Herring, 2008; Boukus et al., 2009; Cunningham & Nichols, 2005). In addition, psychiatrists are less likely than physicians in other specialties to accept new patients (Adams & Herring, 2008; Bishop et al., 2014; Boukus et al., 2009).
Studies of Medicaid have indicated that characteristics of a local market may affect acceptance of and access to care for new patients. Adams and Herring (2008) found that acceptance of new Medicaid patients is higher in locations that have higher Medicaid reimbursement rates, higher ratios of physicians per person, and more commercial Medicaid HMO penetration. Cunningham and Nichols (2005) found Medicaid acceptance was negatively related to the percentage of people in the market who were uninsured and positively related to the percentage age 65 and older. Daly and Mellor (2020) showed that acceptance of Medicaid patients by primary care physicians in Virginia varied across the state and was lower in urban communities with relatively large Hispanic populations.
New Contributions
While past literature has examined the factors that affect the acceptance rates of new patients generally as well as Medicare and Medicaid patients specifically, little is known about the factors that affect the acceptance rates of TRICARE patients. Our research seeks to fill that gap in the literature by addressing three research questions separately for physicians and mental health providers:
How does the acceptance rate for new TRICARE patients compare with the acceptance rate for new Medicare patients?
What reasons do providers give for not accepting TRICARE or Medicare and how do reasons differ by insurance type?
What provider and market characteristics are associated with accepting each insurance type?
We are able to explore these questions using data from the TRICARE Standard Surveys, a congressionally mandated survey of civilian physicians and mental health providers, which provides detailed information about the providers and their practices, including their acceptance of different insurance types.
Method
Data
TRICARE Standard Survey of Providers
TRICARE beneficiary groups expressed concern about access to care using the Standard/Extra plans. In response to these complaints, Congress mandated surveys of providers and beneficiaries in the 2008 and 2012 National Defense Authorization Act. Each year between 2008 and 2011, and between 2012 and 2015, two parallel surveys were conducted. One targeted civilian providers—physicians and nonphysician mental health providers; one targeted TRICARE beneficiaries who rely on these civilian providers. Collectively, these surveys are known as the TRICARE Standard Surveys.
A two-stage sampling design of the TRICARE Standard Surveys was developed to meet the requirements of the 2008 National Defense Authorization Act, which specified sampling 20 Prime Service Areas and 20 non-Prime Service Areas regions each year. For the provider sample frame, physicians, including psychiatrists, within each sampled stratum were identified using the American Medical Association master data file. Nonphysician mental health providers (social workers, psychiatric nurses, psychologists, marriage and family therapists, pastoral counselors, and mental health counselors) were identified from six sources including state licensing records, members’ lists of professional associations, and the National Plan Provider and Enumeration System from the Centers for Medicare & Medicaid Services. Separate random samples of mental health providers and nonpsychiatrist physicians were drawn from each stratum sampled for each year. In each year, providers from select health care markets that were identified by stakeholders as having possible access problems were oversampled. The surveys were addressed to the providers’ office; in this study, we refer to the responses as coming from the provider although they may come from the office staff instead. The response rate was 39%, ranging from 35% in 2015 to 42% in 2013. Sampling weights were assigned based on the probability of each respondent’s selection into the sample and were further adjusted for nonresponse.
The main survey questions of interest for our analyses are whether the provider accepts new TRICARE and/or Medicare patients and if not, the reasons why. Providers who do not accept either insurance type were asked to provide up to seven reason why not. We categorized responses into one of 14 reasons. We also use the physician characteristics—age, gender, practice type, specialty type, and TRICARE region—which we collected through the sample frame and survey questions. Fewer characteristics are available for mental health providers; for most, we know only the TRICARE region in which they practice and whether the provider is a psychiatrist or another type of mental health provider. We calculated estimates using sampling weights to be representative of the U.S. population of physicians and mental health providers in the reporting units described above. The final sample size for the analysis is 30,131 physicians and 19,259 mental health providers.
Local Area Characteristics
We draw on the American Community Survey for demographic characteristics of the local area (defined as reporting units) in which providers practice, the Defense Enrollment and Eligibility Reporting System for information on the number of reservists and residents eligible for TRICARE in the providers’ local area, and the provider sample frame for the number of providers in a local area. The variables we create from the American Community Survey, Defense Enrollment and Eligibility Reporting System, and provider sample frame include average per capita income, number of providers per 1,000 residents, number of residents older than age 65 per provider (a proxy for demand for Medicare), number of reservists per provider, and number of residents eligible for TRICARE per provider (a proxy for demand for TRICARE).
Analytic Approach
We present summary statistics for the provider sample, weighting the means to be representative of the average U.S. civilian provider, and compare the characteristics of our sample against other nationally representative statistics. We perform this and subsequent analyses separately for physicians (including primary care providers and specialists) and mental health providers because there are large differences in acceptance and insurance practices of these two types of providers.
We then calculate the acceptance rates of new TRICARE and Medicare patients by categorizing them into five groups: (1) does not accept any new patients, (2) accepts both TRICARE and Medicare, (3) accepts Medicare but not TRICARE, (4) accepts TRICARE but not Medicare, and (5) accepts neither TRICARE nor Medicare. We classify providers who report they do not know if they accept TRICARE and/or Medicare as not accepting patients with that coverage. We note in the results section when there are noteworthy differences in the results if these providers are analyzed separately or excluded.
For providers who do not accept TRICARE and/or Medicare, we present the proportion of providers who offered a reason that falls into each category. Providers could select multiple reasons, so the percentages for these reasons sum to more than 100%. For this analysis only, we exclude those who say they do not know if they accept the insurance type because they were not asked this question.
Finally, we measure the association between provider characteristics and the acceptance category to which they belong. We present statistics summarizing the characteristics of each acceptance type group, excluding the 4% of physicians and 8% of mental health providers who accept no new patients. We compare the characteristics of providers who reject one or both of the insurance programs with the characteristics of providers who accept both new TRICARE and Medicare patients. We then use regression analysis to measure the association between each characteristic and the probability of being in each acceptance type group, after controlling for other characteristics and conditional on accepting new patients.
Specifically, we estimate a multinomial probit model where the dependent variable is a categorical variable that equals 1 if the provider accepts both TRICARE and Medicare (i.e., the omitted category), 2 if the provider accepts only Medicare, 3 if the provider accepts only TRICARE, and 4 if the provider accepts neither insurance type. The independent variables are provider characteristics (indicators for the age category, gender, practice type, and specialty type) and characteristics of local area (TRICARE region, per capita income, number of providers per 1,000 residents, number of residents older than 65 years per provider, number of reservists per provider, and number of TRICARE-eligible residents per provider). We estimated a separate model for physicians and mental health providers, and the robust standard errors are clustered by stratum. The only provider characteristic we include in the regressions for mental health providers is whether the mental health provider was a psychiatrist because other characteristics were not available. We convert the resulting coefficients into marginal effects that estimate the marginal impact on the probability of being in each acceptance group if the provider has the relevant characteristic of interest, holding all other characteristics constant.
Results
Summary Statistics
The top panel of Table 1 presents summary statistics of the characteristics of our analysis sample. Slightly over half of physicians are in a group practice, 29% are in solo practice, 30% are primary care providers (defined as those who report their specialty as family practice/general medicine or general internist), 29% are internal medicine specialists, and 13% are surgeons. These characteristics of our sample are similar to those of a national sample of physicians in 2008 (Boukus et al., 2009). The age and gender distribution of our sample is similar to that of all active physicians and physicians in training in the 2013 Physician Specialty Data Book (Center for Workforce Studies, Association of American Medical Colleges, 2015). As aforementioned, limited characteristics of mental health providers are available. In terms of local area characteristics, almost half of physicians and mental health providers are in the North TRICARE region, and the average income per capita for the strata in which providers are located is approximately $30,000. There are about 1.4 physicians and 1.5 mental health providers per 1,000 residents, about 3.4 TRICARE eligible beneficiaries per physician, and 3.6 TRICARE eligible beneficiaries per mental health provider.
Summary Statistics.
Note. Information on age, gender, and practice type were not collected for most mental health providers. Providers who said they do not know if they accept TRICARE and/or Medicare are counted as not accepting these insurance types. During the time of this study, TRICARE was managed in three separate regions in the United States: North (which covers states in New England, mid-Atlantic, and as far west as Michigan and Illinois), South (which stretches from Texas through Florida), and West (the remaining states including Alaska and Hawaii). Primary care provider includes those who report their specialty as family practice/general medicine or general internist. Obstetricians/Gynecologists and pediatricians (including those who also report being a primary care practitioner) are reported in a separate category.
Source. TRICARE Standard Survey of Civilian Providers, 2012-2015.
Acceptance Rates of New TRICARE and Medicare Patients
The bottom panel of Table 1 shows how the acceptance rates for TRICARE and Medicare differ. Two thirds of physicians report that they accept both TRICARE and Medicare. Only 5% of physicians accept new patients but neither TRICARE nor Medicare, and 4% accept no new patients. We find that 17% of physicians report that they accept Medicare and not TRICARE and 8% accept TRICARE but not Medicare, showing that physicians accept Medicare more frequently than TRICARE by approximately 9 percentage points. Overall, 74% of physicians report that they accept new TRICARE patients compared with 83% who report accepting new Medicare patients. A significant part of this gap is due to providers who do not know if they accept TRICARE, which we classify as nonacceptance. For example, 59% of physicians who accept Medicare but not TRICARE and 28% who do not accept either program are classified as such because they do not know if they accept TRICARE (not shown in table). In contrast, less than 1% of physicians who accept TRICARE but not Medicare and 11% who do not accept either program are classified in that way because they do not know if they accept Medicare (not shown in table). This disparity indicates that physicians are less familiar with the TRICARE program than Medicare.
Acceptance rates of mental health providers are markedly different than physicians’, with 40% of mental health providers not accepting either program (including those who do not know if they accept these programs) and only 20% reporting that they accept both TRICARE and Medicare. Indeed, 18% of mental health providers are not accepting any new patients, 16% report that they accept Medicare but not TRICARE, and 16% report that they accept TRICARE but not Medicare, showing that overall acceptance rates by mental health providers are low (less than 40%) for both Medicare and TRICARE but similar for the two programs. However, similar to physicians, mental health providers are much less likely to know if they accept TRICARE than Medicare (not shown in table).
Reasons for not Accepting New TRICARE and/or Medicare Patients
Table 2 shows that issues with reimbursement and their specialty not being covered are reasons both physicians and mental health providers offer frequently to explain why they do not accept TRICARE or Medicare. Problems with reimbursement tend to be more prevalent among those who accept neither insurance type than among those who accept just one or the other. Among physicians who do not accept either insurance type, 17% say they do not accept Medicare and 22% TRICARE because of reimbursement, compared with only 8% who do not accept only Medicare because of reimbursement and 16% who do not accept only TRICARE because of reimbursement. We see a similar pattern for mental health providers, although the percentages are slightly lower. Among physicians who do not accept either insurance type, 32% do not accept Medicare and 18% do not accept TRICARE because their specialty is not covered. Among physicians who accept TRICARE but not Medicare, 58% say it is because their specialty is not covered; this high rate is due to OB/GYNs and pediatricians, who are infrequently seen by Medicare beneficiaries. If we exclude OB/GYNs and pediatricians from the sample (not shown in table), only 17% of physicians who accept TRICARE but not Medicare say it is because their specialty is not covered. Among mental health providers who do not accept either program type, 19% do not accept Medicare and 22% do not accept TRICARE for this reason. Among mental health providers who accept TRICARE but not Medicare, 28% say it is because their specialty is not covered.
Reasons Providers Give for not Accepting New TRICARE and/or Medicare Patients.
Note. Sample excludes providers who said that they do not know if they accept TRICARE and/or Medicare.
Source. TRICARE Standard Survey of Civilian Providers, 2012-2015.
A large number of both physicians and mental health providers report that they do not know why they do not accept TRICARE and/or Medicare or gave miscellaneous reasons we could not categorize. Inability to explain the reason(s) might be because survey responses often came from administrative staff or it might be because some larger practices set policies with little input from responding providers. Responses in these categories are particularly prevalent for physicians who do not accept TRICARE but do accept Medicare. However, 31% of these physicians said they do not know why or gave a miscellaneous answer compared with 15% of providers who accept TRICARE but not Medicare.
The frequency of other reasons differs depending on the insurance acceptance type or provider type. For example, among physicians who accept neither TRICARE nor Medicare, 14% say it is due to past problems with TRICARE and 17% say it is due to past problems with Medicare, compared with 5% who accept only Medicare and 3% who accept only TRICARE. This reason is less common among mental health providers, but the pattern is similar: 10% of mental health providers who do not accept either TRICARE or Medicare do so because of past problems, compared with 4% of those who accept only one of the two insurance types. In contrast, mental health providers are more likely than physicians to mention not being aware of TRICARE: 26% of mental health providers who accept Medicare but not TRICARE and 23% who accept neither program report they were not asked to accept it or do not know about TRICARE (compared with 11% and 7% of physicians, respectively). Lack of awareness is also prevalent among mental health providers in the full sample: 22% of mental health providers report being unaware of TRICARE compared with 7% of physicians (not shown in table).
Provider Characteristics by Acceptance Type
Table 3 shows characteristics of providers overall and by insurance acceptance type. Those who are female and pediatricians make up a higher proportion of physicians who do not accept TRICARE and/or Medicare compared with physicians who accept both new TRICARE and Medicare patients, with pediatricians comprising over half of physicians who accept TRICARE but not Medicare. Physicians and mental health providers who do not accept TRICARE tend to practice in areas where there is a higher per capita income and a higher number of providers per resident than those who do accept TRICARE. In contrast, certain specialists make up a higher proportion of physicians who accept both TRICARE and Medicare compared with physicians who do not accept one or both insurance types, such as internal medicine specialists or ophthalmologists. Among mental health providers, those who accept both TRICARE and Medicare are more likely to be psychiatrists compared with those who do not accept either insurance type; the reason is that psychiatrists are more likely to be covered by TRICARE and Medicare than other mental health providers. For example, among psychiatrists who do not accept either TRICARE or Medicare, 8% said they did not accept TRICARE because it did not cover their specialty and 12% said they did not accept Medicare because it did not cover their specialty (not shown in table). In contrast, 17% of nonpsychiatrist mental health providers said they did not accept TRICARE and Medicare because it did not cover their specialty (not shown in table). Physicians and mental health providers who accept TRICARE practice in areas where there are more beneficiaries eligible for TRICARE than providers who do not accept TRICARE.
Provider Characteristics by Insurance Acceptance Type.
Note. During the time of this study, TRICARE was managed in three separate regions in the United States: North (which covers states in New England, mid-Atlantic, and as far west as Michigan and Illinois), South (which stretches from Texas through Florida), and West (the remaining states, including Alaska and Hawaii). Primary care provider includes those who report their specialty as family practice/general medicine or general internist. Obstetricians/Gynecologists and pediatricians (including those who also report being a primary care practitioner) are reported in a separate category. Sample excludes providers who said that they do not accept any new patients.
Source. TRICARE Standard Survey of Civilian Providers, 2012-2015.
Difference from the mean for the “accepts both new TRICARE and Medicare patients” group is statistically significant at the .05 level; ** at the .01 level.
Table 4 presents results from a multinomial regression that examines the relationship between provider and market characteristics and the decision of which insurance types to accept. We have converted the estimated coefficients into marginal effects to show the relationship between the presence of each characteristic and the change in the probability of accepting each insurance type. For example, internal medicine specialists are 11 percentage points more likely to accept both insurance types than are primary care providers, 4 percentage points less likely to accept Medicare but not TRICARE, 5 percentage points less likely to accept TRICARE but not Medicare, and 3 percentage points less likely to accept neither.
Regression Adjusted Relationships Between Provider Characteristics and Insurance Acceptance Type for Providers That Accept New Patients.
Note. During the time of this study, TRICARE was managed in three separate regions in the United States: North (which covers states in New England, mid-Atlantic, and as far west as Michigan and Illinois), South (which stretches from Texas through Florida), and West (the remaining states, including Alaska and Hawaii).
Difference from zero is statistically significant at the .05 level; ** at the .01 level.
Source. TRICARE Standard Survey of Civilian Providers, 2012-2015.
For the most part, the findings from the regression analysis reinforce the relationships we observe in the descriptive results; when we exclude those who do not know if they accept Medicare or TRICARE, there are small differences in statistical significance compared with when we include them, but the sign and magnitude of the estimates still hold. Physicians who are female and pediatricians are less likely to accept both new TRICARE and Medicare patients, and more likely to accept neither insurance type than physicians who are males and primary care providers. However, some specialists (such as internal medicine specialists and ophthalmologists) are more likely than primary care providers to accept both new TRICARE and Medicare patients and less likely to accept one or neither insurance type. Psychiatrists are less likely than other mental health providers to refuse both TRICARE and Medicare or to refuse Medicare but accept TRICARE because psychiatrists are more likely to be covered by these programs.
The more beneficiaries eligible for TRICARE per physician or mental health provider in the local area, the more likely providers are to accept TRICARE and Medicare. Conversely, physicians in areas with a higher number of residents over age 65 per physician are less likely to reject Medicare (with or without TRICARE) and more likely to accept both Medicare and TRICARE. An increase in the number of physicians per resident is associated with a greater likelihood of accepting Medicare but refusing TRICARE.
Discussion
Our analyses show that almost two thirds of physicians report that they accept both TRICARE and Medicare, 5% report accepting new patients but not TRICARE nor Medicare, and 4% do not accept any new patients. Physicians more commonly report that they accept Medicare than TRICARE, with 17% reporting that they accept Medicare and either do not accept or know if they accept TRICARE; only 8% report that they accept TRICARE but do not accept or know if they accept Medicare. Mental health providers are less likely than physicians to report accepting either TRICARE or Medicare patients: 40% do not accept or know if they accept new patients from either program. Providers are less likely to know if they accept TRICARE than Medicare.
Provider acceptance varies according to the characteristics of the insurance program, of the providers themselves, and of the markets in which they provide their services. When asked why they accept no new patients from each program, the most common reasons physicians and mental health providers offered tend to be similar for both TRICARE and Medicare programs; for example, insufficient reimbursement and specialty not covered are among the most frequently listed reasons for not accepting either insurance type. These findings indicate that improving reimbursement or specialty coverage could increase acceptance. However, providers—especially mental health providers—who refuse TRICARE but not Medicare often cite lack of awareness of the program as a reason, which illustrates the significance of information dissemination for programs such as TRICARE that are smaller and more specialized than Medicare.
The greater size of Medicare’s beneficiary population compared with TRICARE may contribute to its higher level of reported acceptance in two ways. First, the larger number of beneficiaries in Medicare results in much higher potential revenue from accepting Medicare patients than TRICARE patients. Second, it means that physicians are much more likely to encounter patients asking them to accept Medicare than patients asking them to accept TRICARE. Our regression results are consistent with both explanations. We find that a higher number of residents older than age 65 per physician increases the probability of accepting Medicare and reduces the probability of physicians accepting TRICARE but not Medicare, which is consistent with the hypothesis that higher potential revenue is associated with a higher acceptance rate. The association between the ratio of physicians to residents and the higher probability of physicians accepting Medicare but not TRICARE is consistent with the second explanation because higher physician density means an individual physician is less likely to encounter a TRICARE user. Physicians who accept Medicare but not TRICARE are also the most likely to report they are unaware of TRICARE or have not been asked to accept it, which is consistent with the second explanation.
Acceptance rates are much lower for mental health providers than physicians: 40% of mental health providers do not accept or know if they accept either insurance type. Though a large proportion of those who refuse both insurance types report it is because their specialty is not covered, their assumptions about what TRICARE covers might be incorrect, and their refusal could be the result of lack of accurate information. Similarly, mental health providers are also more likely than physicians not to be aware of TRICARE. Because insurance coverage of mental health is often less generous than other health services, providers may be less reliant on insurance (public or private) and more likely to rely on patients to self-finance. Because the share of beneficiaries seeking a mental health provider is less than the share needing a physician (National Center for Health Statistics, 2015; Substance Abuse and Mental Health Services Administration, 2017), mental health providers are less likely than physicians to encounter a TRICARE beneficiary and might therefore have less information.
In summary, Medicare and TRICARE are similar in many ways but their differences indicate how health plan characteristics might influence acceptance and access. Our findings show that although most providers who accept one accept the other, the two programs compete on the margins and providers are less likely to be familiar with TRICARE than Medicare. Although TRICARE offers more generous coverage than Medicare in some dimensions, such as mental health, information about benefits might not be salient to providers if few TRICARE beneficiaries are in their markets. These findings demonstrate the importance of outreach and publicity not only for TRICARE but also for other smaller public programs that differ from and may compete with Medicare for provider services. Outreach could include webinars and training events conducted by TRICARE administrators and informational packets distributed to health care providers and office managers who are at-risk of not being aware of the TRICARE program. Administrators of smaller programs such as TRICARE must work to distinguish themselves from Medicare and overcome disadvantages in markets that favor Medicare. The cost of these efforts must be weighed against the number of beneficiaries whose access would be improved.
Footnotes
Acknowledgements
The authors would like to thank Swaati Bangalore for her excellent programming support. The statements contained herein are those of the authors and do not necessarily reflect the views or policies of the U.S. Department of Defense.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This analysis was funded by the U.S. Department of Defense, Contract No. HHSP23320095642WC/HHSP23337054T.
