Specialist Participation in Healthcare Delivery Transformation: Influence of Patient Self-Referral

, , , ,
The American Journal of Managed Care, January 2014, Volume 20, Issue 1

Analyses of national trends indicate that a considerable proportion of new specialist visits among both Medicare and private insurance beneficiaries are self-referred.


Improving coordination of care and containing healthcare costs are prominent goals of healthcare reform. Specialist involvement in healthcare delivery transformation efforts like Accountable Care Organizations (ACOs) is necessary to achieve these goals. However, patients’ self-referrals to specialists may undermine care coordination and incur unnecessary costs if patients frequently receive care from specialists not engaged in such healthcare delivery transformation efforts. Additionally, frequent self-referrals may also diminish the incentive for specialist participation in reform endeavors like ACOs to get access to a referral base.


To examine recent national trends in self-referred new visits to specialists.

Study Design:

A descriptive cross-sectional study of new ambulatory visits to specialists from 2000 to 2009 using data from the National Ambulatory Medical Care Survey.


We calculated nationally representative estimates of the proportion of new specialist visits through self-referrals among Medicare and private insurance beneficiaries. We also estimated the nationally representative absolute number of self-referred new specialist visits among both groups of beneficiaries.


Among Medicare and private insurance beneficiaries, self-referred visits declined from 32.2% (95% confidence interval [CI], 24.0%-40.4%) to 19.6% (95% CI, 13.9%-23.3%) and from 32.4% (95% CI, 27.9%-36.8%) to 24.1% (95% CI,18.8%-29.4%), respectively. Hence, at least 1 in 5 and 1 in 4 new visits to specialists among Medicare and private insurance beneficiaries, respectively, are self-referred.


The current considerable rate of self-referred new specialist visits among both Medicare and private insurance beneficiaries may have adverse implications for organizations attempting to transform healthcare delivery with improved care coordination.

Am J Manag Care. 2014;20(1):e22-e26

  • The trend in the absolute number of self-referred new visits to specialists among Medicare and private insurance beneficiaries was relatively flat between 2000 and 2009.

  • The rate of self-referred new specialist visits in both beneficiary populations isconsiderable, with 1 in 5 and 1 in 4 new specialists visits among Medicare and private insurance beneficiaries, respectively, being self-referred.

  • The rate of self-referred new specialist visits among private-insurance and Medicare beneficiaries may adversely impact healthcare reform efforts to improve quality of care with better coordination.

The goals of healthcare reform include transforming the delivery of healthcare in order to improve coordination of care and contain healthcare-related costs.1 In pursuit of this goal, the Affordable Care Act proffered concepts such as Accountable Care Organizations (ACOs) and patient-centered medical homes (PCMHs) as vehicles to drive transformations in healthcare delivery.2 However, some of the concepts most strongly supported in the healthcare reform law are principally focused on primary care.2-4 For example, ACO beneficiaries are to be assigned based on where they obtain a plurality of primary care services.3 Hence, primary care providers are the only provider type that must be part of an ACO3; the roles of specialty providers are less clear.

Despite the fact that specialists have unclear roles in primary care—focused healthcare delivery transformation vehicles such as ACOs, an incentive for them to be involved is the potential for a referral base from organizations in which they participate.3,5-9 However, high rates of patient self-referral to specialists (hereafter referred to as self-referral) may undermine this referral base incentive and ultimately undermine efforts to improve care coordination and contain costs if specialists remain uninvolved with these organizations.5,9 Two previous studies that specifically examined self-referral patterns found a 70% self-referral rate among Medicare beneficiaries; among private insurance beneficiaries in indemnity plans, self-referral rates exceeded 40% within 9 of the 14 specialties studied.10,11

Hence, our goal in this study was to use the most recently available, nationally representative data to examine the current magnitude of, and trends in, self-referrals at the national level related to several specialties. We undertook this study to provide insight into whether self-referral could impede to achieving the goals of improved care coordination and healthcare-related cost containment.


We used data from the most recent 10 consecutive years (2000-2009) of the National Ambulatory Medical Care Survey (NAMCS), anationally representative survey of nonfederal, employed, officebased physicians engaged in direct patient care. This survey is frequently utilized to estimate and examine trends in national healthcare utilization.12-15 We pooled data from multiple years in order to increase the sample size within the specialties studied and also to examine the overall trend in self-referral rates over the last decade of available data.

Study Population

The level of observation in NAMCS is the patient visit, with data entered by responding physicians/practices. We selected patient isits to all specialties by excluding visits to general/family practice, general internal medicine, and pediatric physicians. We then refined the study population by selecting only new visits in order to better assess self-referral as a proportion of the overall referral base for new patients to a practice. Hence, we excluded postoperative visits and visits from established patients.

Study Variables

As part of NAMCS, physicians report whether the patient whose visit was sampled for the survey was referred. If the answer to this item was no, we designated the visit as a self-referred visit. Thus, the referral variable was dichotomous: self-referred versus not self-referred. Approximately 11.5% of observations had missing variable values for the referral item. We used logistic regression multiple imputation to predict missing variable values for the referral item based on patientrelated variables including age, sex, and race.16 Other variables include the year the visit was sampled, geographic region, metropolitan statistical area status, major reason for the visit, primary payer for the visit, and the type of specialist seen. The multiple imputation prevented exclusion of observations without a variable value for the referral item and hence allowed preservation of the representativeness of the study sample.

The National Ambulatory Medical Care Survey includes an item with primary payer options for each encounter. The options are private insurance, Medicare, Medicaid/State Children’s Health Insurance Program, worker’s compensation, self-pay, and no charge/charity. In order to describe the trends in self-referral for Medicare and private insurance over the decade studied, we created indicator variables for visits for which Medicare or private insurance plans, respectively, were primary payers. Specialty of the physician responding to NAMCS was also documented in the survey, as self-reported by the physician.

Study Analyses

For each year of the study, we estimated separately for the Medicare and private-insurance populations the proportions of new visits that resulted from self-referral. Then, using data aggregated from all 10 years of the study, we estimated the proportion of new visits that resulted from self-referral for each of 10 major specialties sampled in the survey.

A recent study on trends in physician-generated referrals from 1999 to 2009 using NAMCS showed a doubling in the likelihood of physician referral during an ambulatory visit and a 34% increase in the overall number of ambulatory office visits nationally.17 In order to characterize the proportion of self-referred new visits among Medicare and private insurance beneficiaries in the context of the reported overall referral and ambulatory-visit trends, we applied sampling weights provided by NAMCS to our Medicare and private insurance self-referred visit proportion estimations in order to obtain nationally representative absolute figures for self-referred visits for each year of the study.



We examined 32,784 new visits surveyed between 2000 and 2009 within the 10 major specialties most commonly identified in NAMCS by respondents (in descending rank order, they are neurology, otolaryngology, dermatology, orthopedics, urology, general surgery, ophthalmology, cardiology, obstetrics/gynecology, and psychiatry). The presents the proportions of Medicare and private insurance self-referrals for new visits within each of the 10 specialties studied (2000-2009 pooled data). Self-referrals within specialties were generally comparable between Medicare and private insurance beneficiaries, with the exception of psychiatry and ophthalmology, in which self-referral rates among private insurance beneficiaries exceeded their Medicare counterparts by as much as 5 percentage points.

Figure 1A

Figure 1B

Medicare was the primary payer in 6288 (19%) of the surveyed visits, and private insurance was the primary payer in 19,331 (59%) of the surveyed visits. Among both Medicare and private-insurance beneficiaries, there was general decline in the proportion of self referrals between 2000 and 2009 (). Although there was a decrease in the proportion of self-referred new visits to specialists among Medicare and private insurance beneficiaries, the trend in the weighted absolute number of self-referred visits among Medicare and private insurance beneficiaries remained generally stable from 2000 to 2009: 3.1 million (95% confidence interval [CI], 2.5-3.6) to 3.9 million (95% CI, 3.2- 4.6) among Medicare beneficiaries and 13.3 million (95% CI, 11.9- 4.6) to 13.7 million (95% CI, 11.7-15.8) among private insurance beneficiaries ().

DISCUSSIONDeclining Trends in Proportion of Self-Referrals to Specialists

In this nationally representative study, we observed a marked decline in the proportion of new, non-postoperative ambulatory visits to specialists from self-referral among both Medicare and private insurance beneficiaries between 2000 and 2009—totaling a relative reduction of one-third or more over the decade for each group of beneficiaries (Figure 1A). However, the decline in the proportion of self-referred visits translated to a relatively even trend in the absolute number of self-referred specialist visits between 2000 and 2009, meaning that the absolute number of specialist visits resulting from self-referral did not increase or decrease substantially in the decade between 2000 and 2009 (Figure 1B).

This seemingly contradictory finding is perhaps better understood in the context of the recent study by Barnett and colleagues,17 who used NAMCS to demonstrate that the rate of physician referral nearly doubled and the overall number of ambulatory visits increased by more than one-third between 1999 and 2009. In essence, it appears that self-referred visits, although stable in magnitude, constitute a generally declining proportion of both increasing overall referrals and ambulatory visits in the US healthcare system.

Possible Implications of Self-Referral Trends for Specialists and Healthcare Delivery Reform

In addition to the increase in expense that results from specialist visits, physicians and national experts have also identified patient self-referral to specialists as a noteworthy impediment to care coordination efforts.18,19 Therefore, it is conceivable that a considerable volume of self-referred specialist visits in the national healthcare system will have implications for the ability of healthcare delivery transformation mechanisms such as ACOs to involve and fully engage specialists—and will similarly influence specialists’ motivation to be affiliated with the fast-growing number of reform-focused healthcare delivery vehicles. Furthermore, a considerable volume of self-referrals may also have implications for care coordination efforts since these visits bypass primary care providers, who are potentially more able to direct care management without the influence of uncoordinated self-referrals.

Based on the most recent national data available, we found that 1 in 5 new specialist visits among Medicare beneficiaries and 1 in 4 new specialist visits among private insurance beneficiaries are self-referred (Figure 1A). These results are important for healthcare elivery transformation endeavors like ACOs, which are intent on providing better, more cost-effective care through improved coordination.

It is also noteworthy that the proportion of self-referrals depend on the specialty. Fewer than 1 in 10 new visits to neurologists are self-referred among Medicare and private insurance beneficiaries, while approximately 1 in 5 new visits to procedural specialists such as orthopedists and otolaryngologists are self-referred, which has cost implications for the healthcare system (Table 1). Additionally, the differential self-referral rates provide informative data on where efforts may need to be focused in order to achieve the goals of improved care coordination and cost containment.

Ultimately, as providers, payers, and other stakeholders come together to form organizations to transform healthcare delivery, measuring the impact of self-referrals and developing strategies to mitigate their influence will be crucial to success in achieving the goals of healthcare reform. Initial steps can include tracking self-referrals and physician referrals in detail, and establishing clear and reproducible metrics that assess when referrals contribute to or adversely affect efforts to improve care coordination. These measures can help in development of organizationwide practices that optimize the contribution of referrals to care coordination.

Study Limitations

This analysis must be understood in light of certain limitations. Although this study quantified the magnitude of, and recent trends in, self-referred specialist visits, it was not possible to directly measure the impact on care coordination or healthcare cost with NAMCS because it contains crosssectional data with visits as the units of observation; hence, patients cannot be tracked longitudinally. Another limitation of using NAMCS is that our measures of self-referrals in the public and private healthcare sectors were aggregated at the national level and did not reflect the variations in individual healthcare markets with different profiles of benefits plans, which may influence self-referral rates. However, visits to specialists aggregately constitute approximately half of all outpatient visits in the United States, and referrals are the gateway to these visits.16 Therefore, there is value in understanding broad national trends in referral patterns as a portal to the use of specialist services. Author Affiliations: From the University of Michigan (OA, GS, JB, KCC, MMD), Ann Arborm MI.

Funding Source: Robert Wood Johnson Foundation Clinical Scholars Program and the Veterans Affairs Administration Midcareer Investigator Award in Patient-Oriented Research from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS)

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (OA, GS, JB); acquisition of data (OA, GS, JB); analysis and interpretation of data (OA, JB, MMD, GS); drafting of manuscript (OA, MMD); critical revision of manuscript for important intellectual content (OA, MMD, JD, KCC, GS); statistical analysis (OA, JB); supervision (MMD, KCC).

Address correspondence to: Oluseyi Aliu MD, MS, 2800 Plymouth Road Building 10, Ann Arbor, MI 48109. E-mail: Oluseyi@umich.edu.1. Centers for Medicare & Medicaid Services. ACA Savings Report 2012: The Affordable Care Act: Lowering Medicare Costs by Improving Care. http://www.cms.gov/apps/files/aca-savings-report-2012.pdf. Published April 23, 2012. Accessed June 1, 2013.

2. HealthCare.gov. Read the Law: The Affordable Care Act, Section by Section. http://www.healthcare.gov/law/full/. Published March 23, 2010. Accessed June 1, 2013.

3. Wise CG, Alexander JA, Green LA, Cohen GR, Koster CR. Journey toward a patient-centered medical home: readiness for change in primary care practices. Milbank Q. 2011;89(3):399-424.

4. Centers for Medicare & Medicaid Services, HHS. Medicare program; Medicare Shared Savings Program: Accountable Care Organizations: final rule. Fed Regist. 2011;76(212):67802-67990.

5. Higgins A, Stewart K, Dawson K, Bocchino C. Early lessons from accountable care models in the private sector: partnerships between health plans and providers. Health Aff (Millwood). 2011;30(9):1718-1727.

6. Goodney PP, Fisher ES, Cambria RP. Roles for specialty societies and vascular surgeons in accountable care organizations. J Vasc Surg. 2012;55(3):875-882.

7. Shortell SM, Casalino LP. Accountable Care Systems for Comprehensive Healthcare Reform. Fresh-Thinking website: http://www.fresh-thinking.org/docs/workshop_070301/ShortellCasalinoDeliverySystemModelsRevise9.pdf. Published March 2007. Accessed September 28, 2012.

8. Society of Thoracic Surgeons. ACO—a new model in health care savings becomes law. http://www.sts.org/news/aco%E2%80%93-new-model-health-care-savings-becomes-law. Published 2012. Accessed September 28, 2012.

9. Mechanic RE, Santos P, Landon BE, Chernew ME. Medical group responses to global payment: early lessons from the “Alternative Quality Contract” in Massachusetts. Health Aff (Millwood). 2011; 30(9):1734-1742.

10. Shea D, Stuart B, Vasey J, Nag S. Medicare physician referral patterns. Health Serv Res. 1999;34(1, pt 2):331-348.

11. Forrest CB, Reid RJ. Passing the baton: HMOs’ influence on referrals to specialty care. Health Aff (Millwood). 1997;16(6):157-162.

12. Kalsekar I, Record S, Nesnidal K, Hancock B. National estimates of enrollment in disease management programs in the United States: an analysis of the National Ambulatory Medical Care Survey data. Popul Health Manag. 2010;13(4):183-188.

13. Ananthakrishnan AN, McGinley EL, Saeian K, Binion DG. Trends in ambulatory and emergency room visits for inflammatory bowel diseases in the United States: 1994-2005. Am J Gastroenterol. 2010; 105(2):363-370.

14. Cheung R, Mannalithara A, Singh G. Utilization and antiviral therapy in patients with chronic hepatitis C: analysis of ambulatory care visits in the US. Dig Dis Sci. 2010;55(6):1744-1751.

15. Stojanovski SD, Rasu RS, Balkrishnan R, Nahata MC. Trends in medication prescribing for pediatric sleep difficulties in US outpatient settings. Sleep. 2007;30(8):1013-1017.

16. Raghunathan TE. What do we do with missing data? some options for analysis of incomplete data. Annu Rev Public Health. 2004;25: 99-117.

17. Barnett ML, Song Z, Landon BE. Trends in physician referrals in the United States, 1999-2009. Arch Intern Med. 2012;172(2):163-170.

18. Greenfield S, Nelson EC, Zubkoff M, et al. Variations in resource utilization among medical specialties and systems of care: results from the medical outcomes study. JAMA.


19. O’Malley AS, Tynan A, Cohen GR, Kemper N, Davis MM. Coordination of care by primary care practices: strategies, lessons and implications. Res Brief. 2009;(12):1-16.