Collaborative care plans combined with provider education resulted in significant reductions in referrals to specialists without an apparent increase in the rate of emergency department visits or hospital admissions.
Several strategies have been proposed to improve referrals and communication between primary care providers (PCPs) and specialists. In this article, we describe the effectiveness of collaborative care plans (CCPs) in reducing utilization of specialist resources in a capitated health plan based in a safety net hospital. To operationalize individual care plans, a single clinic called the Total Care Clinic (TCC) was launched. Midlevel providers were assigned to subspecialties and trained in specific algorithms of care that they were responsible for. Midlevel providers in the TCC were invited to attend in-house education opportunities. These interventions resulted in an overall 33.6-percentage-point reduction in the referral rate over 7 years of observation. The largest decrease in referrals was observed in gastroenterology, which resulted mostly from colon cancer screening with fecal immunochemical tests in place of colonoscopies. No increase in emergency department (ED) visits or hospital admissions accompanied the decreased referrals to specialists. Combining CCPs with provider education and placing select specialists in proximity of the PCPs resulted in significant referral reductions to specialists without increases in ED visits or hospital admissions.
Am J Manag Care. 2020;26(4):177-180. https://doi.org/10.37765/ajmc.2020.42835
The aging population and the increased prevalence of obesity are contributing to increasing numbers of patients with chronic diseases that require care by multiple specialties.1,2 In one survey, approximately 40% of Americans aged between 50 and 69 years saw more than 1 physician per year.3 It is expected that the number of referrals to specialists will increase manpower needs, significantly affecting healthcare costs. Optimizing the referral process to avoid overuse or underuse of consulting services may help address this. Poor communication and lack of appropriate indication for referral may result in patient harm due to unnecessary diagnostic or therapeutic interventions.4 The optimal rate of referrals by primary care providers (PCPs) is unknown5; however, significant variability of referral rates among providers suggests that patients are inconsistently using specialist resources.6
Unless the severity of a problem necessitates a more urgent encounter, a patient referral to a specialist should occur within 2 weeks. After the patient is seen, consultation results should be promptly made available to the referring PCP. This process of “closing the referral loop” in large healthcare systems with complex provider networks is poor.7 Proposed strategies for improving referrals and communication between PCPs and specialists include involvement of specialists in active educational programs for PCPs and use of structured referral sheets.8 In addition, alternatives to referrals, such as an in-house second opinion, have been tried.8 Passive dissemination of referral guidelines alone is unlikely to lead to improvements in referral practice. Relocation of specialists into primary care sites would make consults more accessible but may not reduce outpatient visits because the demand for consults may increase with increased accessibility.9 The current study’s aim was to determine the effectiveness of a cluster of interventions in reducing utilization of specialist resources in a capitated health plan based in a safety net hospital. Interventions included implementation of disease-specific algorithms for referrals, training and educating of PCPs to optimize the utilization of referral guidelines, and monitoring of the appropriateness of referrals by a designated subspecialty provider. Additionally, cardiology and gastroenterology, 2 specialties with high rates of utilization, made specialized providers available for consulting at the primary care site. These interventions, all aimed to facilitate communication among multiple providers about a patient’s diagnosis and treatment, are collectively referred to as collaborative care plans (CCPs).10
The study was conducted at the University of Florida Health Center in Jacksonville, Florida, a tertiary 695-bed safety net hospital. Administrative data of patients enrolled in a capitated plan funded by the City of Jacksonville (City Contract plan) were retrospectively reviewed. The time period of the study was 7 years (July 1, 2011, through June 30, 2018). Approval from the institutional review board was obtained. People who do not have health insurance and do not qualify for a federal health plan such as Medicare or Medicaid are eligible to enroll in the City Contract plan. Eligibility to enroll in this plan is reevaluated every 6 months except for those whose sole source of income is Social Security benefits. The latter group is eligible for a period of 1 year and constitutes a very small percentage of City Contract—eligible persons. As eligibility may change every 6 months, the population enrolled in the City Contract plan is in near-constant flux. Given the rapid turnover of enrollees, it is thus difficult to define the demographics of the population. The variation in the number of members enrolled in the City Contract plan over the study period is shown in the Figure.
The CCPs for each specialty were developed jointly by PCPs and specialists. Cardiology and gastroenterology each had an in-house specialty provider (a physician and a midlevel provider, respectively) available on-site. These in-house experts provided consultation directly to patients and were a resource for the other providers. The cardiologist spent 1 day a week in the clinic. The midlevel provider with gastroenterology had a full-time assignment at the site. The components of the CCPs are listed in Table 1, and specialty specific guidelines for referrals in the CCPs are summarized in the eAppendix (available at ajmc.com).
To effectively operationalize individual care plans, a single clinic called the Total Care Clinic (TCC) was launched. The medical director appointed to this clinic worked collaboratively with the division chiefs in each specialty to coordinate care. Midlevel providers were assigned to subspecialties and trained in the specific algorithms of care they were responsible for. Faculty of each division were available to address the questions of the midlevel providers by phone or electronically. Midlevel providers in the TCC were invited to division meetings, journal clubs, grand rounds, and other in-house education opportunities to further develop their skills in the subspecialties. A designated consultant in each subspecialty monitored the appropriateness of referrals. If a referral did not meet the prespecified criteria, timely feedback was provided to the referring provider.
Statistical comparisons of the rate of referrals in each subspecialty were carried out using a z score.11 P <.01 was considered the level of statistical significance.
The number of enrollees who were referred to medical subspecialists is summarized in Table 2. The total number of enrollees in the City Contract plan varied because the individuals are enrolled and leave periodically depending on their economic status (Figure). The decline in the number of enrollees was attributed to the introduction of Affordable Care Act—related health plans, with many patients electing to enroll in alternative plans.
The referral rates for the various medical subspecialists over the 7-year observation period are summarized in Table 2. Year 1 is the year prior to implementation of the CCPs. Overall, the number of referrals decreased over time, and at the seventh year, there was an overall 33.6-percentage-point reduction of referrals to the medical subspecialties compared with year 1. The largest decrease in referrals (21.9-percentage-point change in the absolute rate of referrals) was observed in gastroenterology. Similarly, referral rates to endocrinology, cardiology, pulmonary, and nephrology were significantly reduced—by 5.5, 4.3, 2.3, and 1.0 percentage points, respectively (P <.005 for all)—whereas referrals to hematology/oncology, infectious diseases, and rheumatology were not altered significantly (Table 2). With the exception of changes in referrals to cardiology, the decline in referrals for other specialties occurred gradually. The referrals to cardiology, hematology/oncology, and rheumatology declined in the first 3 to 5 years of observation only to increase gradually over the subsequent 2 to 4 years.
During years 3, 4, 5, 6, and 7 of subspecialty data collection, there were 5025 (68.6%), 4702 (64.4%), 4488 (68.3%), 4424 (69.7%), and 3767 (64.8%) emergency department (ED) visits, respectively. There were 978 (13.4%), 975 (13.3%), 934 (14.2%), 802 (12.6%), and 734 (12.6%) hospital admissions, respectively. The data in parentheses are the percentages of enrollees in that year. Corresponding data for ED visits and hospital admissions during years 1 and 2 were not available.
Results of this study show that CCPs incorporating referral guidelines, provider education, and selective in-house placement of specialists in services with high demand resulted in an overall 33.6-percentage-point reduction in referral rate. The largest decrease in referrals was observed in gastroenterology, which was mostly the result of preferentially offering fecal immunochemical testing (FIT) rather than colonoscopies for colon cancer screening. FIT was introduced as part of the CCPs and facilitated by educating the providers about the validity of this test in colon cancer screening through group discussions and sharing of relevant supporting literature.
It is noteworthy that there were insignificant changes in the referral rates to hematology/oncology, infectious diseases, and rheumatology, as the referral guidelines for these services shared with the PCPs were the least restrictive to allow easy access for patients with ailments in the purview of these specialties.
The effectiveness of the interventions was most robust in the first 5 years of implementation of the program. Referral rates to cardiology, hematology/oncology, and rheumatology crept up in the last 2 years of observation. The transient nature of the changes in cardiology referrals was attributed to the discontinuation of the cardiology consultant after the third year of launching these collaborative plans.
It is also noteworthy that the available data for ED visits and hospital admissions during the observation period indicate that the reduction in referrals to subspecialists did not result in increased ED visits or hospital admissions.
Although the appropriate rate of referrals to various specialties is not known, there appears to be significant opportunity for healthcare cost savings. With increasing availability of specialist advice to patients and PCPs by telemedicine, email, or phone, more opportunities now exist to reduce outpatient referrals and reduce costs. It is possible that with this reduction, some patients are not receiving a referral when appropriate. This study design did not allow for assessment of the reduction in unnecessary care.
Because of their emphasis on improving quality and value of patient care, CCPs will have a pivotal role in accountable care organizations. It is noteworthy that not all CCPs are efficacious in reducing costs or improving the quality of referrals.12 The unique approach of the present study was to incorporate several potentially useful interventions simultaneously, augmenting the benefits driven from a single intervention.
Limitations of this study include lack of demographic data, due to high turnover of the enrollees, and lack of information about what health plans were chosen by former City Contract plan enrollees.
A newly emerging trend to improve patient access to specialists is the development of the medical neighborhood.13 This is an extension of the medical home that has been shown to improve the patient experience and minimize healthcare disparities.14 The medical neighborhood facilitates coordination of care between PCPs and specialists. It is considered an advanced form of collaborative care in which on-site consultants and PCPs have the opportunity to interact on a daily basis. The American College of Physicians has developed frameworks to develop these relationships.15
No single model will fit every practice. In addition, the effect of changes in referral rates on patient outcomes is uncertain. These limitations notwithstanding, it is important that PCPs continue to refer patients who need referrals.
Our model of combining CCPs with provider education and placing select specialists in proximity of PCPs resulted in significant reductions in referrals to specialists in a safety net hospital with limited resources. This reduction in utilization of healthcare resources did not lead to increased ED visits or hospital admissions.
The authors thank Ms Kelly Britt, of University of Florida Jacksonville Physicians Inc, for collating the administrative data used in this study.Author Affiliations: Department of Medicine (ADM, RS, CP) and Department of Family and Community Medicine (ES), University of Florida College of Medicine, Jacksonville, FL.
Source of Funding: None.
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 (ADM, RS, CP, ES); acquisition of data (ADM, RS); analysis and interpretation of data (ADM, RS, CP); drafting of the manuscript (ADM, CP, ES); critical revision of the manuscript for important intellectual content (ADM, RS, CP, ES); statistical analysis (CP); and administrative, technical, or logistic support (RS).
Address Correspondence to: Arshag D. Mooradian, MD, Department of Medicine, University of Florida College of Medicine, 653-1 W 8th St, 4th Floor — LRC, Jacksonville, FL 32209. Email: firstname.lastname@example.org.REFERENCES
1. van Oostrom SH, Gijsen R, Stirbu I, et al. Time trends in prevalence of chronic diseases and multimorbidity not only due to aging: data from general practices and health surveys. PLoS One. 2016;11(8):e0160264. doi: 10.1371/journal.pone.0160264.
2. Chronic diseases in America. CDC website. cdc.gov/chronicdisease/resources/infographic/chronic-diseases.htm. Updated October 23, 2019. Accessed March 6, 2020.
3. Reichman M. Optimizing referrals & consults with a standardized process. Fam Pract Manag. 2007;14(10):38-42.
4. Roland M, Coulter A, eds. Hospital Referrals. Oxford, United Kingdom: Oxford University Press; 1992.
5. Gwynne M, Page C, Reid A, Donahue K, Newton W. What’s the right referral rate? specialty referral patterns and curricula across 13 collaborative primary care residencies. Fam Med. 2017;49(2):91-96.
6. Wilkin D. Patterns of referral: explaining variation. In: Roland M, Coulter A, eds. Hospital Referrals. Oxford, United Kingdom: Oxford University Press; 1992:76-91.
7. Patel MP, Schettini P, O’Leary CP, Bosworth HB, Anderson JB, Shah KP. Closing the referral loop: an analysis of primary care referrals to specialists in a large health system. J Gen Intern Med. 2018;33(5):715-721. doi: 10.1007/s11606-018-4392-z.
8. Akbari A, Mayhew A, Al-Alawi MA, et al. Interventions to improve outpatient referrals from primary care to secondary care. Cochrane Database Syst Rev. 2014;(4):CD005471. doi: 10.1002/14651858.CD005471.pub2.
9. Winpenny EM, Miani C, Pitchforth E, King S, Roland M. Improving the effectiveness and efficiency of outpatient services: a scoping review of interventions at the primary—secondary care interface. J Health Serv Res Policy. 2017;22(1):53-64. doi: 10.1177/1355819616648982.
10. Mauksch L, Safford BH. Engaging patients in collaborative care plans. Fam Pract Manag. 2013;20(3):35-39.
11. 2-sample z-test to compare sample proportion. Epitools website. epitools.ausvet.com.au/content.php?page=z-test-2. Accessed March 6, 2020.
12. Solberg LI, Crain AL, Maciosek MV, et al. A stepped-wedge evaluation of an initiative to spread the collaborative care model for depression in primary care. Ann Fam Med. 2015;13(5):412-420. doi: 10.1370/afm.1842.
13. Feuerstein JD, Sheppard V, Cheifetz AS, Ariyabuddhiphongs K. How to develop the medical neighborhood. J Med Syst. 2016;40(9):196. doi: 10.1007/s10916-016-0557-7.
14. Lee K, Palacio C, Alexandraki I, Stewart E, Mooradian AD. Increasing access to health care providers through medical home model may abolish racial disparity in diabetes care: evidence from a cross-sectional study. J Natl Med Assoc. 2011;103(3):250-256. doi: 10.1016/s0027-9684(15)30293-5.
15. Kirschner N, Greenlee MC. The patient-centered medical home neighbor: the interface of the patient-centered medical home with specialty/subspecialty practices. American College of Physicians website. acponline.org/system/files/documents/advocacy/current_policy_papers/assets/pcmh_neighbors.pdf. Published 2010. Accessed March 6, 2020.