Christopher B. Forrest, MD, PhD

Conventionally, primary care physicians decided when to refer and to whom a patient should be referred.^1,2 Specialists’ assistance was sought for diagnostic or therapeutic dilemmas,^3,4 management of conditions that presented too infrequently to maintain clinical competence,⁵ and specialized procedures that fell outside a physician’s scope of practice.^3,4 In some cases, physicians referred because patients requested to see a specialist.^1,4

The reorganization of health care over the past few decades has dramatically altered the interactions between primary care physicians and specialists. The growth in multispecialty group practice arrangements⁶ has led to formal, organizationally defined linkages between practitioners. Managed health plans and medical groups⁷ encourage primary care physicians’ control over the referral process through such mechanisms as specialty referral authorizations, financial disincentives for making a referral, performance assessment of referral patterns, and referral guidelines. These changes have transformed a once-informal process into one rife with administrative restrictions on referral decision making.

The Ambulatory Sentinel Practice Network (ASPN) Referral Study was designed to describe and analyze primary care physicians’ referral decisions and their outcomes in the context of a changing health care system in the United States. The study occurred in the ASPN and other regional practice-based research networks. This report examines primary care physicians’ referral decisions. We conceptualized the referral decision as occurring in 2 phases: whether to refer followed by to whom to refer.²

Methods

Physician sample

Physicians were recruited from March 1997 to May 1998. Recruitment activities were directed to all physician members of ASPN, physicians affiliated with the Medical Group Management Association, local and regional networks (Minnesota Academy of Family Physicians Research Network, the Wisconsin Research Network, the Dartmouth Primary Cooperative Research Network (COOP), and the larger community of primary care physicians. The study was publicized via direct mailings to physicians, articles and notices in practice-based research network newsletters and journals, and presentations at conferences. Contact with physicians expressing interest was made by telephone.

Physicians were included in the study if they practiced in the United States and were not in residency or fellowship training. Of all physicians contacted, 342 expressed interest in the study and 182 completed some aspect of data collection. A total of 141 family physicians, 12 internists, and 1 pediatrician completed all phases of data collection. In this study, the 141 family physicians (41% members of ASPN) formed the physician sample. They delivered health care in 87 practices located in 31 states.

Procedures

Study protocols and materials, based on a similar practice-based research study conducted with pediatricians, were reviewed and approved by the Committee on Human Research of the Johns Hopkins School of Public Health and the Colorado Multi-Institutional Review Board.^4,8,9 We conducted a pilot test in 5 practices; this test led to further refinements of methods and questionnaires.

Data collection occurred from September 1997 to February 1999, with 94% of physicians collecting data in 1998 only. Before beginning data collection, physicians completed a questionnaire concerning their practices and personal characteristics. Each practice selected a coordinator who communicated with research staff, learned study protocols, trained office staff and physicians, and monitored data quality. Coordinators completed a questionnaire about the organizational and financial components of their practice. They kept a log of all visits made during 15 consecutive business days and occurring during regularly scheduled office hours. A business day was defined as a half or full work day, provided that the physician held routine office hours. Each patient’s date of birth (5% missing), sex (2% missing), and principal diagnosis (5% missing) were recorded.

The coordinator kept another log of all referrals made by physicians, nurses, and other office staff. Referrals made during telephone conversations with patients were included. A referral was defined as a recommendation that a patient have a face-to-face encounter with another practitioner. We excluded referrals made to laboratories, radiologic facilities, emergency departments, hospitals for inpatient admission, and “curbside consultations” (ie, when the referring physician obtains advice from a specialist but does not send the patient for a visit).

A medical record abstractor assigned ICD-9-CM codes to diagnoses provided by office staff. We matched ICD codes to an expanded set of diagnosis clusters (EDCs). EDCs group ICD codes into clinically homogeneous categories using the methods developed by Schneeweiss.¹⁰ (For more information on EDCs, see http://acg.jhsph.edu.)

When physicians made a referral, they completed a questionnaire (response rate 93.9%) with items concerning the referral decision. Reasons for referral were based on our previously developed taxonomy used in a pediatric referral study⁴ and focus groups of family physicians convened during an annual ASPN convocation.

At the study’s conclusion, physicians received a report that compared their referral practice patterns with those of the entire sample. To defray office expenses associated with data collection, each practice was given a $100 stipend in addition to $5 for each physician referral.

Generalizability analysis

We compared referral rates of the study sample with the National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office visits made to family physicians.^11,12 We pooled surveys from 1989 to 1994, inclusive, when the majority of the items in the survey instruments remained unchanged.¹³ (The 1995–1999 surveys did not contain information on whether the visit led to referral.)

We selected visits made by patients enrolled in non-HMO health plans (NAMCS) and health plans that had neither capitated primary care physician payment nor gatekeeping arrangements (study sample). This was done because of the known effect of managed care in general, and of gatekeeping specifically, on increasing referral rates^8,13 and the unequal distribution of managed care plans between the 2 samples. Unweighted visits yielded a sample size of 37,145; of these, 11,676 met the selection criteria.

The proportions of office visits referred were compared overall and by age, sex, and health condition. The 10 most frequently referred conditions in the study sample were used for the condition-specific referral rate assessments. Statistical significance was assessed by the chi-square statistic.

Results

Descriptive information on the 141 family physician sample is presented in Table 1. Physicians spent an average of 51.3 hours per week in their jobs. About 68% of their time was devoted to direct patient care. In most practices, a staff member coordinated administrative aspects of specialty referrals; 20% permitted patients to request a referral by leaving a voice mail message.

TABLE 1
FAMILY PHYSICIAN STUDY SAMPLE

Frequency of referral

The 141 family physicians had 34,519 office visits and made 2165 referrals during 1771 practice-days; 5.1% of office visits were referred. Physicians saw an average of 19.7 patients per day (range 7.0 to 48.4) and made 1.23 referrals per full practice-day (range 0 to 3.90). Referrals made during telephone conversations with patients accounted for 18.9% of all referrals made by physicians (range 0% to 100% per physician).

An additional 369 referrals (a rate of 0.21 referrals per practice-day) were made by staff. Overall, 68% of all referrals were made by physicians during visits with patients, 18% by physicians during telephone conversations with patients, 11% by staff with physician input, and 3% by staff without physician input. In 43.6% of referrals made during telephone conversations with patients, the telephone encounter was the first presentation to medical care for the health problem.

We compared percentages of office visits in which a referral was made in the study sample with percentages of such referrals by family physicians from the NAMCS surveys (1989–1994). The overall percentages did not significantly differ between the 2 groups (4.0% vs 3.7%, P > .05). Although physicians in the study sample were statistically less likely than NAMCS counterparts to refer children (1.6% vs 2.5%, P = .030), more likely to refer the elderly (4.8% vs 4.1%, P = .045), and more likely to refer females (4.1% vs 3.9%, P = .009), these differences were small. There were no differences between the groups in condition-specific referral rates. In sum, these results show that patients in the ASPN sample were equally likely to be referred as those in the NAMCS sample.

Reasons for referral

Table 2 shows the distribution of physicians’ reasons for making the referral. Physicians endorsed a mean of 1.8 different reasons for making the typical referral. Although patients requested to see a specialist for 13.6% of referrals, physicians recorded patient request as the only reason for referral just 1.1%.

We compared referrals made for uncommon conditions (lowest tertile of practice-prevalence) with common conditions (highest practice-prevalence tertile). The calculation of practice-prevalence was based on prior research: the numerator was visits made for the index condition, and the denominator was all visits in the sample.⁵ Uncommon conditions were more likely to be referred for medical management (38.5% vs 25.4%, P < .001), patient request (19.8% vs 12.3%, P = .005), and specialist request (4.9% vs 2.1%, P = .021). Common conditions were more likely to be sent to specialists because of failed current therapy (13.6% vs 3.8%, P < .001) and endoscopy (4.3% vs 0.5%, P = .013). There were no significant differences between the 2 groups in the chances of referral for advice on either diagnosis or treatment.

TABLE 2
REASONS FOR REFERRAL

Conditions referred

Table 3 presents condition-specific referral rates and the 2 most common types of specialists referred to for the top 25 referred health problems. (A complete listing of these data for all conditions reported by study physicians can be found in Table W1.) The 50 most commonly referred health problems accounted for 76% of all referrals made during office visits. Signs or symptoms accounted for 22.4% of all referrals. Condition-specific referral rates varied from a low of 1.9% for patients with otitis media to a high of 45.7% of visits referred for patients with cholelithiasis or cholecystitis. This range in referral rates translates into 24-fold variation in the chances of referral during an office visit based solely on the presenting problem.

TABLE 3
NUMBER OF OFFICE VISITS, REFERRAL RATES, AND SPECIALISTS REFERRED TO FOR TOP 15 REFERRED CONDITIONS*

Specialist selection

Referrals were made most often to surgical subspecialists (45.4%), followed by medical subspecialists (31.0%), nonphysician clinicians (12.1%), obstetriciangynecologists (ob/gyns) (4.6%), mental health professionals (4.2%), other physicians (2.0%), and generalists (0.8%). The 5 most common specialists to whom patients were referred were orthopedic surgeons (12.1%), general surgeons (9.1%), otolaryngologists (6.9%), gastroenterologists (6.6%), and dermatologists (6.0%). Among male patients, referral to urologists was the second most common type; among female patients, referral to ob/gyns was the third most common type.

Mental health referrals were made predominantly to psychologists (2.1% of all referrals), followed by psychiatrists (1.3%) and social workers (0.4%). The most common types of nonphysician clinicians referred to were physical therapists (4.5%), podiatrists (3.0%), nutritionists (1.5%), and audiologists (1.2%).

Referring physicians recommended a specific specialist to the patient for 86.2% of referrals. In descending rank order according to the mean importance rating (range 1 to 3), the reasons for selecting a particular specialist were personal knowledge of the specialist (2.6), quality of prior feedback (2.5), technical capacity (2.3), appointment availability (2.0), patient’s request (1.6), requirements of patient’s health plan (1.6), and proximity of the specialist to the patient’s home (1.6).

Table 4 shows the 3 most common health problems referred to 10 types of specialists. (An expanded version of this table that includes 29 specialists can be found in Table W2.) The majority of referrals for each type of specialist were for 1 to 3 health problems. Family physicians made 17.1% of all referrals to practitioners within their practices. Intrapractice referrals were significantly higher than the overall average for audiologists (40.0%, P = .031), nutritionists (45.2%, P = .004), and psychologists (46.3%, P < .001) and were lower for gastroenterologists (9.3%, P = .022) and rheumatologists (4.0%, P = .005).

TABLE 4
THREE MOST COMMON CONDITIONS REFERRED TO SELECTED SPECIALISTS*

Discussion

This study shows that family physicians manage 95% of office visits without specialty referral. About one third of referrals made from primary care practices occur during encounters other than office visits. Referrals made by staff or during telephone conversations may be part of an integrated sequence of contacts between patients and physicians. Nonetheless, assisting patients in selecting a specialist, transferring relevant patient information, and scheduling specialty appointments (referral coordination activities) are more difficult to perform when patients are not seen in the office,¹⁴ because time is limited and integrating care is poorly reimbursed, if at all. When such referral decisions are made appropriately, they provide an efficient mechanism for decreasing workload in a busy primary care practice. Inappropriately made, they can lead to increased expense, unnecessary time spent with specialists, and poorly coordinated care.

We found that the rates of referral were substantially different among the most commonly referred conditions. Prior work has shown that the frequency with which conditions present to primary care physicians explains about 75% of the variation in condition-specific referral rates.⁵ The mix and severity of comorbidities are important determinants of annual patient referral rates^15,16 and the chances of referral during a visit.⁵ Thus, the epidemiology of morbidity among a patient population is a critical factor that defines the boundaries between primary care physicians and specialists. The appreciation of these clinical determinants is crucial for any valid assessment of primary care physicians’ referral patterns.

Limitations

The study’s focus was on new referral decisions made by physicians to other practitioners. No information is provided about ongoing, long-term referrals in which the patient was already under the care of a specialist. The low rates of referral for conditions such as diabetes may be a consequence of this limitation. Patients with diabetes may already have been under the care of a specialist, thereby generating few new referrals. It is also important to note that even in health plans with gatekeeping arrangements, patients self-refer to specialty care¹³; this study did not include any information on self-referral. Patient self-referral appears to be most likely among sick patients, those with established relationships with a specialist, and patients who do not have a good relationship with a primary care physician.¹⁷

We did not obtain information on the number of telephone calls fielded by physicians each day. Without these data, we were unable to determine whether our methods had failed to capture some telephone referrals or to calculate telephone referral rates. In this study, family physicians made 18.9% of all referrals during telephone conversations, in contrast with pediatricians in another study⁴ who made 27.5% of all referrals by telephone. The difference in these proportions is not large and is probably explained by pediatricians’ greater use of the telephone for patient care.

It could be argued that the volunteer physicians in this study systematically differ from the typical family physician. The average number of visits per day among study physicians (19.7) is similar to a national estimate of 19.9 visits/day for family physicians in single specialty group practices.¹⁸ Furthermore, we found similar probabilities of referral overall and for the 10 most commonly referred conditions between study physicians and a national sample, suggesting that referral propensities between the 2 groups were similar.

Why family physicians refer

No value judgments can be made about the appropriateness of physicians’ reasons for referral. Physicians most commonly referred because they were uncertain about diagnosis or treatment and sought advice from another practitioner. For about 1 in 5 referrals, physicians recorded only a sign or symptom as the diagnosis, suggesting a reasonably high level of diagnostic uncertainty. Physicians’ tolerance of uncertainty varies markedly,¹⁹ making it difficult to judge questions about appropriateness of referrals that are made to reduce this uncertainty.

Another important reason for referral was that physicians deemed the management of the health problem to be outside their scope of practice. Physicians were more likely to refer a patient with a common problem after trying out a course of treatment than was the case for uncommon problems that were more likely to be referred for medical management.

Patients may raise the topic of possible referral. When physicians agree that referral is indicated, they almost always find other reasons for making the referral. Alternatively, physicians might make a decision to refer and justify it in part as being a result of patient request. Discussions on whether a referral is needed are common in primary care. Among referrals made in an Israeli family practice network, patients raised the topic of possible referral in 27% of cases.²⁰ In a study of 856 internal medicine visits, 45% of patients indicated some desire to discuss the need for referral with their physician; however, physicians recognized these desires only about half the time.²¹

Selecting a specialist

Our results show that primary care physicians prefer to send their patients to specialists with whom they have developed a relationship. Physicians in this study maintained a high level of involvement in specialist selection, providing patients with the name of a specific practitioner for 86.2% of referrals. The most important factor in selecting a specialist in our study was the same as that found nearly 20 years ago by Ludke¹: personal knowledge of the specialist. Physicians’ dissatisfaction with the specialty referral process in managed care settings^22,23 could be a result of their reduced choice of specialists with whom they have forged personal relationships.

Slightly more than 1 in 6 referrals were made to specialists in the referring physician’s practice, consistent with movement of primary care physicians into multispecialty groups. Whether intrapractice referral holds any advantage over referrals outside the practice, such as better coordination and appointment adherence, awaits future study.

Our results show that physicians must not only select a specific practitioner but also choose among different types of practitioners. Some patients were sent to nonphysician clinicians and physicians (eg, podiatrists and orthopedic surgeons for acquired foot deformity), whereas others were sent to medical or surgical subspecialists (eg, nephrologists and urologists for urinary tract symptoms). These patterns are likely to reflect the need for multidisciplinary specialty care for some conditions. For instance, patients with diabetes may see an ophthalmologist for retinopathy evaluations and an endocrinologist for medical management consultation. For some conditions, there appears to be considerable uncertainty regarding the boundaries between specialists.²⁴ Should a patient with a skin mass be sent to a general surgeon, a dermatologist, or a plastic surgeon? When should a patient with allergic rhinitis be sent to an allergist and when to an otolaryngologist? These referral patterns may reflect local care practices and specialist availability. They may also be a consequence of a surplus of specialists in this country and competition for patients.

In a survey of family physicians that was performed in the late 1980s, respondents reported that they were more likely to refer to internal medicine subspecialists than internists for adults, but preferred general pediatricians over pediatric subspecialists.²⁵ Our findings suggest that the trend for adult patients remains, but there has been a shift away from general pediatricians toward subspecialists for pediatric referrals. These new patterns may be a consequence of greater availability of pediatric subspecialists, greater exposure of family physicians to pediatric consultants, and a larger share of family physicians who have completed residency training.

Implications for physician training

Fifty conditions accounted for 76% of all specialty referrals made during office visits in this study. Interactions with most types of specialists are generally limited to a few conditions; 3 health problems accounted for more than half of referrals to most specialties. Educators should ensure that these commonly referred conditions are emphasized in curricula that provide family physicians with the skills necessary to expand their scope of practice, when appropriate; determine when a patient should be referred; and identify the type of practitioner to whom the patient should be sent.

Physicians in training should be taught the skills required to recognize the boundaries of their clinical uncertainty and scopes of practice. A challenge for educators is to assist trainees in determining when to tolerate clinical uncertainty while employing a watchful waiting approach and when to initiate a more aggressive evaluation, including when to obtain specialty referral. Modes of implementing these approaches are likely to differ across conditions. Thus, it makes sense in physician training to place the greatest emphasis on conditions for which family physicians commonly refer.

Under certain circumstances, patient request for a specialty consultation may be a sufficient and legitimate reason for referral. For example, as we found in this study, patients with uncommon health problems may seek reassurance from specialists skilled in the management of their specific condition. Managing access to specialists, particularly when the physician is acting as an administrative gatekeeper to referrals, can be challenging.²¹ When doctors and patients disagree on the need for referral, patients may become dissatisfied with their health care²⁶ and decide to self-refer to specialty care.¹⁷ In consideration of the increasing complexity of medical care, developing skills that help physicians discuss and negotiate access to specialized services with both patients and specialists has never been more timely.

Acknowledgments

This study was funded by grant no. R01 HS09377 from the Agency for Healthcare Research and Quality. James Werner and Laurie Vorel provided technical assistance with data collection and project implementation. Many physicians collected data for this study. Their time and devotion were invaluable to the success of this study. These physicians are listed by the states in which they practice. Arizona: Scott Ekdahl, DO; Arkansas: John Scott, MD; California: Andrew Ness, MD; Colorado: Howard Corren, MD; Nell Davis, MD; Timothy Dudley, MD; Audrey Farley, MD; Tillman Farley, MD; Charles Kay, MD; Joan FAMILY PHYSICIANS’ REFERRAL DECISIONS MacEachen, MD; George Maxted, MD; John Miller, MD; Kathy Miller, MD; Steven Milligan, MD; Frank Reed, MD; Louise Schottsteadt, MD; Lynne Spicer, MD; Laura Stein, MD; Lynn Strange, MD; Dan Sullivan, MD; Georgia: Linda Casteel, MD; Randy Cronic, MD; Bruno Denis, MD; Keith Ellis, MD; Kelly Erola, MD; Craig Fabel, MD; Russell Leubbert, DO; Richard Liotta, DO; Mark Majoch, MD; David Najjar, MD; James Snow, DO; Roslyn Taylor, MD; Illinois: Steven Lidvall, MD; Anna Meenan, MD; Eduardo Scholcoff, MD; Loyd Wollstadt, MD; Indiana: Paul Daluga, MD; Steven Phillipson, MD; Iowa: Ken Miller, MD; Janet Ryan, MD; Kansas: Wendell Ellis, DO; John R. Eplee, MD; Robert Moser, MD; Daniel Sontheimer, MD; Louisiana: Linda Stewart, MD; Michigan: Linda French, MD; John Hickner, MD; Minnesota: Ravi Balasubraman, MD; Dave Bucher, MD; William Davis, MD; Richard Gebhart, MD; Katie Guthrie, MD; Anthony Jaspers, MD; Timothy Komoto, MD; Glenn McCarty, DO; Stephen Mitrione, MD; Thomas Retzinger, MD; Paul Spilseth, MD; Ashlesha Tamboli, MD; Montana: Curt Kurtz, MD; Nevada: Coleen Lyons, MD; New Hampshire: Richard Douglass, MD; Paul Friedrichs, MD; Peter Hope, MD; Jonathan Mishcon, MD; New Jersey: John Orzano, MD; Winifred Waldron, MD; New York: Carmella Abraham, MD; R. Eugene Bailey, MD; Lorne Becker, MD; John DeSimone, MD; Miguel Diaz, MD; Rebecca Elliott, MD; John Glennon, MD; James Greenwald, MD; Glenn Griffin, MD; Eileen Hoffman, MD; L. Thomas Wolff, MD; North Carolina: Ed Bujold, MD; Thomas Detesco, MD; Dave Rogers, MD; Phil Sherrod, MD; Oklahoma: Laura Miller, DO; Mike Pontious, MD; Oregon: Douglas Eliason, DO; L.J. Fagnan, MD; Jerry Flaming, DO; Tom Flaming, DO; Jeffrey Humphrey, DO; Michael Kelber, MD; John Sattenspiel, MD; Pennsylvania: John Farmer, DO; Penitha Williams, MD; South Dakota: Fred Thanel, MD; Tennessee: Dan Brewer, MD; Michael Hartsell, MD; R. Louis Murphy, MD; John Parham, MD; Texas: Michael Averitt, DO; Sharon Barber, MD; Kim Patrick Bolton, MD; Robert Cortes, MD; Paul Gerdes, MD; Robert Henry, DO; Michael Kirkpatrick, MD; John Manning, MD; Shane Maxwell, MD; Luis Moreno, MD; Larry G. Padget, MD; Peter Sullivan, MD; Utah: Scott Endsley, MD; David Flinders, MD; Jim Giovino, MD; Eric Hogenson, MD; Dwayne Roberts, MD; Virginia: Duane Lawrence, MD; James Ledwith, MD; June Tunstall, MD; George Wortley, MD; Washington: John Anderson, MD; Elizabeth Wise, MD; West Virginia: Dan Doyle, MD; J. Michael Herr, DO; Wisconsin: Richard Anstett, MD, PhD; Walter Boisvert, MD; Lea Cornell, MD; Anne Eglash, MD; Rod Erickson, MD; Tom Frisby, MD; Terry Hankey, MD; Kevin Jessen, MD; Dan Landdeck, MD; Dave Lonsdorf, MD; Michael Pace, MD; Michael Saunders, MD; Catherine Soderqueist, MD; Jon Temte, MD; Vince Winklerprins, MD; Brian Woody, MD.

Conventionally, primary care physicians decided when to refer and to whom a patient should be referred.^1,2 Specialists’ assistance was sought for diagnostic or therapeutic dilemmas,^3,4 management of conditions that presented too infrequently to maintain clinical competence,⁵ and specialized procedures that fell outside a physician’s scope of practice.^3,4 In some cases, physicians referred because patients requested to see a specialist.^1,4

The reorganization of health care over the past few decades has dramatically altered the interactions between primary care physicians and specialists. The growth in multispecialty group practice arrangements⁶ has led to formal, organizationally defined linkages between practitioners. Managed health plans and medical groups⁷ encourage primary care physicians’ control over the referral process through such mechanisms as specialty referral authorizations, financial disincentives for making a referral, performance assessment of referral patterns, and referral guidelines. These changes have transformed a once-informal process into one rife with administrative restrictions on referral decision making.

The Ambulatory Sentinel Practice Network (ASPN) Referral Study was designed to describe and analyze primary care physicians’ referral decisions and their outcomes in the context of a changing health care system in the United States. The study occurred in the ASPN and other regional practice-based research networks. This report examines primary care physicians’ referral decisions. We conceptualized the referral decision as occurring in 2 phases: whether to refer followed by to whom to refer.²

Methods

Physician sample

Physicians were recruited from March 1997 to May 1998. Recruitment activities were directed to all physician members of ASPN, physicians affiliated with the Medical Group Management Association, local and regional networks (Minnesota Academy of Family Physicians Research Network, the Wisconsin Research Network, the Dartmouth Primary Cooperative Research Network (COOP), and the larger community of primary care physicians. The study was publicized via direct mailings to physicians, articles and notices in practice-based research network newsletters and journals, and presentations at conferences. Contact with physicians expressing interest was made by telephone.

Physicians were included in the study if they practiced in the United States and were not in residency or fellowship training. Of all physicians contacted, 342 expressed interest in the study and 182 completed some aspect of data collection. A total of 141 family physicians, 12 internists, and 1 pediatrician completed all phases of data collection. In this study, the 141 family physicians (41% members of ASPN) formed the physician sample. They delivered health care in 87 practices located in 31 states.

Procedures

Study protocols and materials, based on a similar practice-based research study conducted with pediatricians, were reviewed and approved by the Committee on Human Research of the Johns Hopkins School of Public Health and the Colorado Multi-Institutional Review Board.^4,8,9 We conducted a pilot test in 5 practices; this test led to further refinements of methods and questionnaires.

Data collection occurred from September 1997 to February 1999, with 94% of physicians collecting data in 1998 only. Before beginning data collection, physicians completed a questionnaire concerning their practices and personal characteristics. Each practice selected a coordinator who communicated with research staff, learned study protocols, trained office staff and physicians, and monitored data quality. Coordinators completed a questionnaire about the organizational and financial components of their practice. They kept a log of all visits made during 15 consecutive business days and occurring during regularly scheduled office hours. A business day was defined as a half or full work day, provided that the physician held routine office hours. Each patient’s date of birth (5% missing), sex (2% missing), and principal diagnosis (5% missing) were recorded.

The coordinator kept another log of all referrals made by physicians, nurses, and other office staff. Referrals made during telephone conversations with patients were included. A referral was defined as a recommendation that a patient have a face-to-face encounter with another practitioner. We excluded referrals made to laboratories, radiologic facilities, emergency departments, hospitals for inpatient admission, and “curbside consultations” (ie, when the referring physician obtains advice from a specialist but does not send the patient for a visit).

A medical record abstractor assigned ICD-9-CM codes to diagnoses provided by office staff. We matched ICD codes to an expanded set of diagnosis clusters (EDCs). EDCs group ICD codes into clinically homogeneous categories using the methods developed by Schneeweiss.¹⁰ (For more information on EDCs, see http://acg.jhsph.edu.)

When physicians made a referral, they completed a questionnaire (response rate 93.9%) with items concerning the referral decision. Reasons for referral were based on our previously developed taxonomy used in a pediatric referral study⁴ and focus groups of family physicians convened during an annual ASPN convocation.

At the study’s conclusion, physicians received a report that compared their referral practice patterns with those of the entire sample. To defray office expenses associated with data collection, each practice was given a $100 stipend in addition to $5 for each physician referral.

Generalizability analysis

We compared referral rates of the study sample with the National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office visits made to family physicians.^11,12 We pooled surveys from 1989 to 1994, inclusive, when the majority of the items in the survey instruments remained unchanged.¹³ (The 1995–1999 surveys did not contain information on whether the visit led to referral.)

We selected visits made by patients enrolled in non-HMO health plans (NAMCS) and health plans that had neither capitated primary care physician payment nor gatekeeping arrangements (study sample). This was done because of the known effect of managed care in general, and of gatekeeping specifically, on increasing referral rates^8,13 and the unequal distribution of managed care plans between the 2 samples. Unweighted visits yielded a sample size of 37,145; of these, 11,676 met the selection criteria.

The proportions of office visits referred were compared overall and by age, sex, and health condition. The 10 most frequently referred conditions in the study sample were used for the condition-specific referral rate assessments. Statistical significance was assessed by the chi-square statistic.

Results

Descriptive information on the 141 family physician sample is presented in Table 1. Physicians spent an average of 51.3 hours per week in their jobs. About 68% of their time was devoted to direct patient care. In most practices, a staff member coordinated administrative aspects of specialty referrals; 20% permitted patients to request a referral by leaving a voice mail message.

TABLE 1
FAMILY PHYSICIAN STUDY SAMPLE

Frequency of referral

The 141 family physicians had 34,519 office visits and made 2165 referrals during 1771 practice-days; 5.1% of office visits were referred. Physicians saw an average of 19.7 patients per day (range 7.0 to 48.4) and made 1.23 referrals per full practice-day (range 0 to 3.90). Referrals made during telephone conversations with patients accounted for 18.9% of all referrals made by physicians (range 0% to 100% per physician).

An additional 369 referrals (a rate of 0.21 referrals per practice-day) were made by staff. Overall, 68% of all referrals were made by physicians during visits with patients, 18% by physicians during telephone conversations with patients, 11% by staff with physician input, and 3% by staff without physician input. In 43.6% of referrals made during telephone conversations with patients, the telephone encounter was the first presentation to medical care for the health problem.

We compared percentages of office visits in which a referral was made in the study sample with percentages of such referrals by family physicians from the NAMCS surveys (1989–1994). The overall percentages did not significantly differ between the 2 groups (4.0% vs 3.7%, P > .05). Although physicians in the study sample were statistically less likely than NAMCS counterparts to refer children (1.6% vs 2.5%, P = .030), more likely to refer the elderly (4.8% vs 4.1%, P = .045), and more likely to refer females (4.1% vs 3.9%, P = .009), these differences were small. There were no differences between the groups in condition-specific referral rates. In sum, these results show that patients in the ASPN sample were equally likely to be referred as those in the NAMCS sample.

Reasons for referral

Table 2 shows the distribution of physicians’ reasons for making the referral. Physicians endorsed a mean of 1.8 different reasons for making the typical referral. Although patients requested to see a specialist for 13.6% of referrals, physicians recorded patient request as the only reason for referral just 1.1%.

We compared referrals made for uncommon conditions (lowest tertile of practice-prevalence) with common conditions (highest practice-prevalence tertile). The calculation of practice-prevalence was based on prior research: the numerator was visits made for the index condition, and the denominator was all visits in the sample.⁵ Uncommon conditions were more likely to be referred for medical management (38.5% vs 25.4%, P < .001), patient request (19.8% vs 12.3%, P = .005), and specialist request (4.9% vs 2.1%, P = .021). Common conditions were more likely to be sent to specialists because of failed current therapy (13.6% vs 3.8%, P < .001) and endoscopy (4.3% vs 0.5%, P = .013). There were no significant differences between the 2 groups in the chances of referral for advice on either diagnosis or treatment.

TABLE 2
REASONS FOR REFERRAL

Conditions referred

Table 3 presents condition-specific referral rates and the 2 most common types of specialists referred to for the top 25 referred health problems. (A complete listing of these data for all conditions reported by study physicians can be found in Table W1.) The 50 most commonly referred health problems accounted for 76% of all referrals made during office visits. Signs or symptoms accounted for 22.4% of all referrals. Condition-specific referral rates varied from a low of 1.9% for patients with otitis media to a high of 45.7% of visits referred for patients with cholelithiasis or cholecystitis. This range in referral rates translates into 24-fold variation in the chances of referral during an office visit based solely on the presenting problem.

TABLE 3
NUMBER OF OFFICE VISITS, REFERRAL RATES, AND SPECIALISTS REFERRED TO FOR TOP 15 REFERRED CONDITIONS*

Specialist selection

Referrals were made most often to surgical subspecialists (45.4%), followed by medical subspecialists (31.0%), nonphysician clinicians (12.1%), obstetriciangynecologists (ob/gyns) (4.6%), mental health professionals (4.2%), other physicians (2.0%), and generalists (0.8%). The 5 most common specialists to whom patients were referred were orthopedic surgeons (12.1%), general surgeons (9.1%), otolaryngologists (6.9%), gastroenterologists (6.6%), and dermatologists (6.0%). Among male patients, referral to urologists was the second most common type; among female patients, referral to ob/gyns was the third most common type.

Mental health referrals were made predominantly to psychologists (2.1% of all referrals), followed by psychiatrists (1.3%) and social workers (0.4%). The most common types of nonphysician clinicians referred to were physical therapists (4.5%), podiatrists (3.0%), nutritionists (1.5%), and audiologists (1.2%).

Referring physicians recommended a specific specialist to the patient for 86.2% of referrals. In descending rank order according to the mean importance rating (range 1 to 3), the reasons for selecting a particular specialist were personal knowledge of the specialist (2.6), quality of prior feedback (2.5), technical capacity (2.3), appointment availability (2.0), patient’s request (1.6), requirements of patient’s health plan (1.6), and proximity of the specialist to the patient’s home (1.6).

Table 4 shows the 3 most common health problems referred to 10 types of specialists. (An expanded version of this table that includes 29 specialists can be found in Table W2.) The majority of referrals for each type of specialist were for 1 to 3 health problems. Family physicians made 17.1% of all referrals to practitioners within their practices. Intrapractice referrals were significantly higher than the overall average for audiologists (40.0%, P = .031), nutritionists (45.2%, P = .004), and psychologists (46.3%, P < .001) and were lower for gastroenterologists (9.3%, P = .022) and rheumatologists (4.0%, P = .005).

TABLE 4
THREE MOST COMMON CONDITIONS REFERRED TO SELECTED SPECIALISTS*

Discussion

This study shows that family physicians manage 95% of office visits without specialty referral. About one third of referrals made from primary care practices occur during encounters other than office visits. Referrals made by staff or during telephone conversations may be part of an integrated sequence of contacts between patients and physicians. Nonetheless, assisting patients in selecting a specialist, transferring relevant patient information, and scheduling specialty appointments (referral coordination activities) are more difficult to perform when patients are not seen in the office,¹⁴ because time is limited and integrating care is poorly reimbursed, if at all. When such referral decisions are made appropriately, they provide an efficient mechanism for decreasing workload in a busy primary care practice. Inappropriately made, they can lead to increased expense, unnecessary time spent with specialists, and poorly coordinated care.

We found that the rates of referral were substantially different among the most commonly referred conditions. Prior work has shown that the frequency with which conditions present to primary care physicians explains about 75% of the variation in condition-specific referral rates.⁵ The mix and severity of comorbidities are important determinants of annual patient referral rates^15,16 and the chances of referral during a visit.⁵ Thus, the epidemiology of morbidity among a patient population is a critical factor that defines the boundaries between primary care physicians and specialists. The appreciation of these clinical determinants is crucial for any valid assessment of primary care physicians’ referral patterns.

Limitations

The study’s focus was on new referral decisions made by physicians to other practitioners. No information is provided about ongoing, long-term referrals in which the patient was already under the care of a specialist. The low rates of referral for conditions such as diabetes may be a consequence of this limitation. Patients with diabetes may already have been under the care of a specialist, thereby generating few new referrals. It is also important to note that even in health plans with gatekeeping arrangements, patients self-refer to specialty care¹³; this study did not include any information on self-referral. Patient self-referral appears to be most likely among sick patients, those with established relationships with a specialist, and patients who do not have a good relationship with a primary care physician.¹⁷

We did not obtain information on the number of telephone calls fielded by physicians each day. Without these data, we were unable to determine whether our methods had failed to capture some telephone referrals or to calculate telephone referral rates. In this study, family physicians made 18.9% of all referrals during telephone conversations, in contrast with pediatricians in another study⁴ who made 27.5% of all referrals by telephone. The difference in these proportions is not large and is probably explained by pediatricians’ greater use of the telephone for patient care.

It could be argued that the volunteer physicians in this study systematically differ from the typical family physician. The average number of visits per day among study physicians (19.7) is similar to a national estimate of 19.9 visits/day for family physicians in single specialty group practices.¹⁸ Furthermore, we found similar probabilities of referral overall and for the 10 most commonly referred conditions between study physicians and a national sample, suggesting that referral propensities between the 2 groups were similar.

Why family physicians refer

No value judgments can be made about the appropriateness of physicians’ reasons for referral. Physicians most commonly referred because they were uncertain about diagnosis or treatment and sought advice from another practitioner. For about 1 in 5 referrals, physicians recorded only a sign or symptom as the diagnosis, suggesting a reasonably high level of diagnostic uncertainty. Physicians’ tolerance of uncertainty varies markedly,¹⁹ making it difficult to judge questions about appropriateness of referrals that are made to reduce this uncertainty.

Another important reason for referral was that physicians deemed the management of the health problem to be outside their scope of practice. Physicians were more likely to refer a patient with a common problem after trying out a course of treatment than was the case for uncommon problems that were more likely to be referred for medical management.

Patients may raise the topic of possible referral. When physicians agree that referral is indicated, they almost always find other reasons for making the referral. Alternatively, physicians might make a decision to refer and justify it in part as being a result of patient request. Discussions on whether a referral is needed are common in primary care. Among referrals made in an Israeli family practice network, patients raised the topic of possible referral in 27% of cases.²⁰ In a study of 856 internal medicine visits, 45% of patients indicated some desire to discuss the need for referral with their physician; however, physicians recognized these desires only about half the time.²¹

Selecting a specialist

Our results show that primary care physicians prefer to send their patients to specialists with whom they have developed a relationship. Physicians in this study maintained a high level of involvement in specialist selection, providing patients with the name of a specific practitioner for 86.2% of referrals. The most important factor in selecting a specialist in our study was the same as that found nearly 20 years ago by Ludke¹: personal knowledge of the specialist. Physicians’ dissatisfaction with the specialty referral process in managed care settings^22,23 could be a result of their reduced choice of specialists with whom they have forged personal relationships.

Slightly more than 1 in 6 referrals were made to specialists in the referring physician’s practice, consistent with movement of primary care physicians into multispecialty groups. Whether intrapractice referral holds any advantage over referrals outside the practice, such as better coordination and appointment adherence, awaits future study.

Our results show that physicians must not only select a specific practitioner but also choose among different types of practitioners. Some patients were sent to nonphysician clinicians and physicians (eg, podiatrists and orthopedic surgeons for acquired foot deformity), whereas others were sent to medical or surgical subspecialists (eg, nephrologists and urologists for urinary tract symptoms). These patterns are likely to reflect the need for multidisciplinary specialty care for some conditions. For instance, patients with diabetes may see an ophthalmologist for retinopathy evaluations and an endocrinologist for medical management consultation. For some conditions, there appears to be considerable uncertainty regarding the boundaries between specialists.²⁴ Should a patient with a skin mass be sent to a general surgeon, a dermatologist, or a plastic surgeon? When should a patient with allergic rhinitis be sent to an allergist and when to an otolaryngologist? These referral patterns may reflect local care practices and specialist availability. They may also be a consequence of a surplus of specialists in this country and competition for patients.

In a survey of family physicians that was performed in the late 1980s, respondents reported that they were more likely to refer to internal medicine subspecialists than internists for adults, but preferred general pediatricians over pediatric subspecialists.²⁵ Our findings suggest that the trend for adult patients remains, but there has been a shift away from general pediatricians toward subspecialists for pediatric referrals. These new patterns may be a consequence of greater availability of pediatric subspecialists, greater exposure of family physicians to pediatric consultants, and a larger share of family physicians who have completed residency training.

Implications for physician training

Fifty conditions accounted for 76% of all specialty referrals made during office visits in this study. Interactions with most types of specialists are generally limited to a few conditions; 3 health problems accounted for more than half of referrals to most specialties. Educators should ensure that these commonly referred conditions are emphasized in curricula that provide family physicians with the skills necessary to expand their scope of practice, when appropriate; determine when a patient should be referred; and identify the type of practitioner to whom the patient should be sent.

Physicians in training should be taught the skills required to recognize the boundaries of their clinical uncertainty and scopes of practice. A challenge for educators is to assist trainees in determining when to tolerate clinical uncertainty while employing a watchful waiting approach and when to initiate a more aggressive evaluation, including when to obtain specialty referral. Modes of implementing these approaches are likely to differ across conditions. Thus, it makes sense in physician training to place the greatest emphasis on conditions for which family physicians commonly refer.

Under certain circumstances, patient request for a specialty consultation may be a sufficient and legitimate reason for referral. For example, as we found in this study, patients with uncommon health problems may seek reassurance from specialists skilled in the management of their specific condition. Managing access to specialists, particularly when the physician is acting as an administrative gatekeeper to referrals, can be challenging.²¹ When doctors and patients disagree on the need for referral, patients may become dissatisfied with their health care²⁶ and decide to self-refer to specialty care.¹⁷ In consideration of the increasing complexity of medical care, developing skills that help physicians discuss and negotiate access to specialized services with both patients and specialists has never been more timely.

Acknowledgments

This study was funded by grant no. R01 HS09377 from the Agency for Healthcare Research and Quality. James Werner and Laurie Vorel provided technical assistance with data collection and project implementation. Many physicians collected data for this study. Their time and devotion were invaluable to the success of this study. These physicians are listed by the states in which they practice. Arizona: Scott Ekdahl, DO; Arkansas: John Scott, MD; California: Andrew Ness, MD; Colorado: Howard Corren, MD; Nell Davis, MD; Timothy Dudley, MD; Audrey Farley, MD; Tillman Farley, MD; Charles Kay, MD; Joan FAMILY PHYSICIANS’ REFERRAL DECISIONS MacEachen, MD; George Maxted, MD; John Miller, MD; Kathy Miller, MD; Steven Milligan, MD; Frank Reed, MD; Louise Schottsteadt, MD; Lynne Spicer, MD; Laura Stein, MD; Lynn Strange, MD; Dan Sullivan, MD; Georgia: Linda Casteel, MD; Randy Cronic, MD; Bruno Denis, MD; Keith Ellis, MD; Kelly Erola, MD; Craig Fabel, MD; Russell Leubbert, DO; Richard Liotta, DO; Mark Majoch, MD; David Najjar, MD; James Snow, DO; Roslyn Taylor, MD; Illinois: Steven Lidvall, MD; Anna Meenan, MD; Eduardo Scholcoff, MD; Loyd Wollstadt, MD; Indiana: Paul Daluga, MD; Steven Phillipson, MD; Iowa: Ken Miller, MD; Janet Ryan, MD; Kansas: Wendell Ellis, DO; John R. Eplee, MD; Robert Moser, MD; Daniel Sontheimer, MD; Louisiana: Linda Stewart, MD; Michigan: Linda French, MD; John Hickner, MD; Minnesota: Ravi Balasubraman, MD; Dave Bucher, MD; William Davis, MD; Richard Gebhart, MD; Katie Guthrie, MD; Anthony Jaspers, MD; Timothy Komoto, MD; Glenn McCarty, DO; Stephen Mitrione, MD; Thomas Retzinger, MD; Paul Spilseth, MD; Ashlesha Tamboli, MD; Montana: Curt Kurtz, MD; Nevada: Coleen Lyons, MD; New Hampshire: Richard Douglass, MD; Paul Friedrichs, MD; Peter Hope, MD; Jonathan Mishcon, MD; New Jersey: John Orzano, MD; Winifred Waldron, MD; New York: Carmella Abraham, MD; R. Eugene Bailey, MD; Lorne Becker, MD; John DeSimone, MD; Miguel Diaz, MD; Rebecca Elliott, MD; John Glennon, MD; James Greenwald, MD; Glenn Griffin, MD; Eileen Hoffman, MD; L. Thomas Wolff, MD; North Carolina: Ed Bujold, MD; Thomas Detesco, MD; Dave Rogers, MD; Phil Sherrod, MD; Oklahoma: Laura Miller, DO; Mike Pontious, MD; Oregon: Douglas Eliason, DO; L.J. Fagnan, MD; Jerry Flaming, DO; Tom Flaming, DO; Jeffrey Humphrey, DO; Michael Kelber, MD; John Sattenspiel, MD; Pennsylvania: John Farmer, DO; Penitha Williams, MD; South Dakota: Fred Thanel, MD; Tennessee: Dan Brewer, MD; Michael Hartsell, MD; R. Louis Murphy, MD; John Parham, MD; Texas: Michael Averitt, DO; Sharon Barber, MD; Kim Patrick Bolton, MD; Robert Cortes, MD; Paul Gerdes, MD; Robert Henry, DO; Michael Kirkpatrick, MD; John Manning, MD; Shane Maxwell, MD; Luis Moreno, MD; Larry G. Padget, MD; Peter Sullivan, MD; Utah: Scott Endsley, MD; David Flinders, MD; Jim Giovino, MD; Eric Hogenson, MD; Dwayne Roberts, MD; Virginia: Duane Lawrence, MD; James Ledwith, MD; June Tunstall, MD; George Wortley, MD; Washington: John Anderson, MD; Elizabeth Wise, MD; West Virginia: Dan Doyle, MD; J. Michael Herr, DO; Wisconsin: Richard Anstett, MD, PhD; Walter Boisvert, MD; Lea Cornell, MD; Anne Eglash, MD; Rod Erickson, MD; Tom Frisby, MD; Terry Hankey, MD; Kevin Jessen, MD; Dan Landdeck, MD; Dave Lonsdorf, MD; Michael Pace, MD; Michael Saunders, MD; Catherine Soderqueist, MD; Jon Temte, MD; Vince Winklerprins, MD; Brian Woody, MD.

Conventionally, primary care physicians decided when to refer and to whom a patient should be referred.^1,2 Specialists’ assistance was sought for diagnostic or therapeutic dilemmas,^3,4 management of conditions that presented too infrequently to maintain clinical competence,⁵ and specialized procedures that fell outside a physician’s scope of practice.^3,4 In some cases, physicians referred because patients requested to see a specialist.^1,4

The reorganization of health care over the past few decades has dramatically altered the interactions between primary care physicians and specialists. The growth in multispecialty group practice arrangements⁶ has led to formal, organizationally defined linkages between practitioners. Managed health plans and medical groups⁷ encourage primary care physicians’ control over the referral process through such mechanisms as specialty referral authorizations, financial disincentives for making a referral, performance assessment of referral patterns, and referral guidelines. These changes have transformed a once-informal process into one rife with administrative restrictions on referral decision making.

The Ambulatory Sentinel Practice Network (ASPN) Referral Study was designed to describe and analyze primary care physicians’ referral decisions and their outcomes in the context of a changing health care system in the United States. The study occurred in the ASPN and other regional practice-based research networks. This report examines primary care physicians’ referral decisions. We conceptualized the referral decision as occurring in 2 phases: whether to refer followed by to whom to refer.²

Methods

Physician sample

Physicians were recruited from March 1997 to May 1998. Recruitment activities were directed to all physician members of ASPN, physicians affiliated with the Medical Group Management Association, local and regional networks (Minnesota Academy of Family Physicians Research Network, the Wisconsin Research Network, the Dartmouth Primary Cooperative Research Network (COOP), and the larger community of primary care physicians. The study was publicized via direct mailings to physicians, articles and notices in practice-based research network newsletters and journals, and presentations at conferences. Contact with physicians expressing interest was made by telephone.

Physicians were included in the study if they practiced in the United States and were not in residency or fellowship training. Of all physicians contacted, 342 expressed interest in the study and 182 completed some aspect of data collection. A total of 141 family physicians, 12 internists, and 1 pediatrician completed all phases of data collection. In this study, the 141 family physicians (41% members of ASPN) formed the physician sample. They delivered health care in 87 practices located in 31 states.

Procedures

Study protocols and materials, based on a similar practice-based research study conducted with pediatricians, were reviewed and approved by the Committee on Human Research of the Johns Hopkins School of Public Health and the Colorado Multi-Institutional Review Board.^4,8,9 We conducted a pilot test in 5 practices; this test led to further refinements of methods and questionnaires.

Data collection occurred from September 1997 to February 1999, with 94% of physicians collecting data in 1998 only. Before beginning data collection, physicians completed a questionnaire concerning their practices and personal characteristics. Each practice selected a coordinator who communicated with research staff, learned study protocols, trained office staff and physicians, and monitored data quality. Coordinators completed a questionnaire about the organizational and financial components of their practice. They kept a log of all visits made during 15 consecutive business days and occurring during regularly scheduled office hours. A business day was defined as a half or full work day, provided that the physician held routine office hours. Each patient’s date of birth (5% missing), sex (2% missing), and principal diagnosis (5% missing) were recorded.

The coordinator kept another log of all referrals made by physicians, nurses, and other office staff. Referrals made during telephone conversations with patients were included. A referral was defined as a recommendation that a patient have a face-to-face encounter with another practitioner. We excluded referrals made to laboratories, radiologic facilities, emergency departments, hospitals for inpatient admission, and “curbside consultations” (ie, when the referring physician obtains advice from a specialist but does not send the patient for a visit).

A medical record abstractor assigned ICD-9-CM codes to diagnoses provided by office staff. We matched ICD codes to an expanded set of diagnosis clusters (EDCs). EDCs group ICD codes into clinically homogeneous categories using the methods developed by Schneeweiss.¹⁰ (For more information on EDCs, see http://acg.jhsph.edu.)

When physicians made a referral, they completed a questionnaire (response rate 93.9%) with items concerning the referral decision. Reasons for referral were based on our previously developed taxonomy used in a pediatric referral study⁴ and focus groups of family physicians convened during an annual ASPN convocation.

At the study’s conclusion, physicians received a report that compared their referral practice patterns with those of the entire sample. To defray office expenses associated with data collection, each practice was given a $100 stipend in addition to $5 for each physician referral.

Generalizability analysis

We compared referral rates of the study sample with the National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office visits made to family physicians.^11,12 We pooled surveys from 1989 to 1994, inclusive, when the majority of the items in the survey instruments remained unchanged.¹³ (The 1995–1999 surveys did not contain information on whether the visit led to referral.)

We selected visits made by patients enrolled in non-HMO health plans (NAMCS) and health plans that had neither capitated primary care physician payment nor gatekeeping arrangements (study sample). This was done because of the known effect of managed care in general, and of gatekeeping specifically, on increasing referral rates^8,13 and the unequal distribution of managed care plans between the 2 samples. Unweighted visits yielded a sample size of 37,145; of these, 11,676 met the selection criteria.

The proportions of office visits referred were compared overall and by age, sex, and health condition. The 10 most frequently referred conditions in the study sample were used for the condition-specific referral rate assessments. Statistical significance was assessed by the chi-square statistic.

Results

Descriptive information on the 141 family physician sample is presented in Table 1. Physicians spent an average of 51.3 hours per week in their jobs. About 68% of their time was devoted to direct patient care. In most practices, a staff member coordinated administrative aspects of specialty referrals; 20% permitted patients to request a referral by leaving a voice mail message.

TABLE 1
FAMILY PHYSICIAN STUDY SAMPLE

Frequency of referral

The 141 family physicians had 34,519 office visits and made 2165 referrals during 1771 practice-days; 5.1% of office visits were referred. Physicians saw an average of 19.7 patients per day (range 7.0 to 48.4) and made 1.23 referrals per full practice-day (range 0 to 3.90). Referrals made during telephone conversations with patients accounted for 18.9% of all referrals made by physicians (range 0% to 100% per physician).

An additional 369 referrals (a rate of 0.21 referrals per practice-day) were made by staff. Overall, 68% of all referrals were made by physicians during visits with patients, 18% by physicians during telephone conversations with patients, 11% by staff with physician input, and 3% by staff without physician input. In 43.6% of referrals made during telephone conversations with patients, the telephone encounter was the first presentation to medical care for the health problem.

We compared percentages of office visits in which a referral was made in the study sample with percentages of such referrals by family physicians from the NAMCS surveys (1989–1994). The overall percentages did not significantly differ between the 2 groups (4.0% vs 3.7%, P > .05). Although physicians in the study sample were statistically less likely than NAMCS counterparts to refer children (1.6% vs 2.5%, P = .030), more likely to refer the elderly (4.8% vs 4.1%, P = .045), and more likely to refer females (4.1% vs 3.9%, P = .009), these differences were small. There were no differences between the groups in condition-specific referral rates. In sum, these results show that patients in the ASPN sample were equally likely to be referred as those in the NAMCS sample.

Reasons for referral

Table 2 shows the distribution of physicians’ reasons for making the referral. Physicians endorsed a mean of 1.8 different reasons for making the typical referral. Although patients requested to see a specialist for 13.6% of referrals, physicians recorded patient request as the only reason for referral just 1.1%.

We compared referrals made for uncommon conditions (lowest tertile of practice-prevalence) with common conditions (highest practice-prevalence tertile). The calculation of practice-prevalence was based on prior research: the numerator was visits made for the index condition, and the denominator was all visits in the sample.⁵ Uncommon conditions were more likely to be referred for medical management (38.5% vs 25.4%, P < .001), patient request (19.8% vs 12.3%, P = .005), and specialist request (4.9% vs 2.1%, P = .021). Common conditions were more likely to be sent to specialists because of failed current therapy (13.6% vs 3.8%, P < .001) and endoscopy (4.3% vs 0.5%, P = .013). There were no significant differences between the 2 groups in the chances of referral for advice on either diagnosis or treatment.

TABLE 2
REASONS FOR REFERRAL

Conditions referred

Table 3 presents condition-specific referral rates and the 2 most common types of specialists referred to for the top 25 referred health problems. (A complete listing of these data for all conditions reported by study physicians can be found in Table W1.) The 50 most commonly referred health problems accounted for 76% of all referrals made during office visits. Signs or symptoms accounted for 22.4% of all referrals. Condition-specific referral rates varied from a low of 1.9% for patients with otitis media to a high of 45.7% of visits referred for patients with cholelithiasis or cholecystitis. This range in referral rates translates into 24-fold variation in the chances of referral during an office visit based solely on the presenting problem.

TABLE 3
NUMBER OF OFFICE VISITS, REFERRAL RATES, AND SPECIALISTS REFERRED TO FOR TOP 15 REFERRED CONDITIONS*

Specialist selection

Referrals were made most often to surgical subspecialists (45.4%), followed by medical subspecialists (31.0%), nonphysician clinicians (12.1%), obstetriciangynecologists (ob/gyns) (4.6%), mental health professionals (4.2%), other physicians (2.0%), and generalists (0.8%). The 5 most common specialists to whom patients were referred were orthopedic surgeons (12.1%), general surgeons (9.1%), otolaryngologists (6.9%), gastroenterologists (6.6%), and dermatologists (6.0%). Among male patients, referral to urologists was the second most common type; among female patients, referral to ob/gyns was the third most common type.

Mental health referrals were made predominantly to psychologists (2.1% of all referrals), followed by psychiatrists (1.3%) and social workers (0.4%). The most common types of nonphysician clinicians referred to were physical therapists (4.5%), podiatrists (3.0%), nutritionists (1.5%), and audiologists (1.2%).

Referring physicians recommended a specific specialist to the patient for 86.2% of referrals. In descending rank order according to the mean importance rating (range 1 to 3), the reasons for selecting a particular specialist were personal knowledge of the specialist (2.6), quality of prior feedback (2.5), technical capacity (2.3), appointment availability (2.0), patient’s request (1.6), requirements of patient’s health plan (1.6), and proximity of the specialist to the patient’s home (1.6).

Table 4 shows the 3 most common health problems referred to 10 types of specialists. (An expanded version of this table that includes 29 specialists can be found in Table W2.) The majority of referrals for each type of specialist were for 1 to 3 health problems. Family physicians made 17.1% of all referrals to practitioners within their practices. Intrapractice referrals were significantly higher than the overall average for audiologists (40.0%, P = .031), nutritionists (45.2%, P = .004), and psychologists (46.3%, P < .001) and were lower for gastroenterologists (9.3%, P = .022) and rheumatologists (4.0%, P = .005).

TABLE 4
THREE MOST COMMON CONDITIONS REFERRED TO SELECTED SPECIALISTS*

Discussion

This study shows that family physicians manage 95% of office visits without specialty referral. About one third of referrals made from primary care practices occur during encounters other than office visits. Referrals made by staff or during telephone conversations may be part of an integrated sequence of contacts between patients and physicians. Nonetheless, assisting patients in selecting a specialist, transferring relevant patient information, and scheduling specialty appointments (referral coordination activities) are more difficult to perform when patients are not seen in the office,¹⁴ because time is limited and integrating care is poorly reimbursed, if at all. When such referral decisions are made appropriately, they provide an efficient mechanism for decreasing workload in a busy primary care practice. Inappropriately made, they can lead to increased expense, unnecessary time spent with specialists, and poorly coordinated care.

We found that the rates of referral were substantially different among the most commonly referred conditions. Prior work has shown that the frequency with which conditions present to primary care physicians explains about 75% of the variation in condition-specific referral rates.⁵ The mix and severity of comorbidities are important determinants of annual patient referral rates^15,16 and the chances of referral during a visit.⁵ Thus, the epidemiology of morbidity among a patient population is a critical factor that defines the boundaries between primary care physicians and specialists. The appreciation of these clinical determinants is crucial for any valid assessment of primary care physicians’ referral patterns.

Limitations

The study’s focus was on new referral decisions made by physicians to other practitioners. No information is provided about ongoing, long-term referrals in which the patient was already under the care of a specialist. The low rates of referral for conditions such as diabetes may be a consequence of this limitation. Patients with diabetes may already have been under the care of a specialist, thereby generating few new referrals. It is also important to note that even in health plans with gatekeeping arrangements, patients self-refer to specialty care¹³; this study did not include any information on self-referral. Patient self-referral appears to be most likely among sick patients, those with established relationships with a specialist, and patients who do not have a good relationship with a primary care physician.¹⁷

We did not obtain information on the number of telephone calls fielded by physicians each day. Without these data, we were unable to determine whether our methods had failed to capture some telephone referrals or to calculate telephone referral rates. In this study, family physicians made 18.9% of all referrals during telephone conversations, in contrast with pediatricians in another study⁴ who made 27.5% of all referrals by telephone. The difference in these proportions is not large and is probably explained by pediatricians’ greater use of the telephone for patient care.

It could be argued that the volunteer physicians in this study systematically differ from the typical family physician. The average number of visits per day among study physicians (19.7) is similar to a national estimate of 19.9 visits/day for family physicians in single specialty group practices.¹⁸ Furthermore, we found similar probabilities of referral overall and for the 10 most commonly referred conditions between study physicians and a national sample, suggesting that referral propensities between the 2 groups were similar.

Why family physicians refer

No value judgments can be made about the appropriateness of physicians’ reasons for referral. Physicians most commonly referred because they were uncertain about diagnosis or treatment and sought advice from another practitioner. For about 1 in 5 referrals, physicians recorded only a sign or symptom as the diagnosis, suggesting a reasonably high level of diagnostic uncertainty. Physicians’ tolerance of uncertainty varies markedly,¹⁹ making it difficult to judge questions about appropriateness of referrals that are made to reduce this uncertainty.

Another important reason for referral was that physicians deemed the management of the health problem to be outside their scope of practice. Physicians were more likely to refer a patient with a common problem after trying out a course of treatment than was the case for uncommon problems that were more likely to be referred for medical management.

Patients may raise the topic of possible referral. When physicians agree that referral is indicated, they almost always find other reasons for making the referral. Alternatively, physicians might make a decision to refer and justify it in part as being a result of patient request. Discussions on whether a referral is needed are common in primary care. Among referrals made in an Israeli family practice network, patients raised the topic of possible referral in 27% of cases.²⁰ In a study of 856 internal medicine visits, 45% of patients indicated some desire to discuss the need for referral with their physician; however, physicians recognized these desires only about half the time.²¹

Selecting a specialist

Our results show that primary care physicians prefer to send their patients to specialists with whom they have developed a relationship. Physicians in this study maintained a high level of involvement in specialist selection, providing patients with the name of a specific practitioner for 86.2% of referrals. The most important factor in selecting a specialist in our study was the same as that found nearly 20 years ago by Ludke¹: personal knowledge of the specialist. Physicians’ dissatisfaction with the specialty referral process in managed care settings^22,23 could be a result of their reduced choice of specialists with whom they have forged personal relationships.

Slightly more than 1 in 6 referrals were made to specialists in the referring physician’s practice, consistent with movement of primary care physicians into multispecialty groups. Whether intrapractice referral holds any advantage over referrals outside the practice, such as better coordination and appointment adherence, awaits future study.

Our results show that physicians must not only select a specific practitioner but also choose among different types of practitioners. Some patients were sent to nonphysician clinicians and physicians (eg, podiatrists and orthopedic surgeons for acquired foot deformity), whereas others were sent to medical or surgical subspecialists (eg, nephrologists and urologists for urinary tract symptoms). These patterns are likely to reflect the need for multidisciplinary specialty care for some conditions. For instance, patients with diabetes may see an ophthalmologist for retinopathy evaluations and an endocrinologist for medical management consultation. For some conditions, there appears to be considerable uncertainty regarding the boundaries between specialists.²⁴ Should a patient with a skin mass be sent to a general surgeon, a dermatologist, or a plastic surgeon? When should a patient with allergic rhinitis be sent to an allergist and when to an otolaryngologist? These referral patterns may reflect local care practices and specialist availability. They may also be a consequence of a surplus of specialists in this country and competition for patients.

In a survey of family physicians that was performed in the late 1980s, respondents reported that they were more likely to refer to internal medicine subspecialists than internists for adults, but preferred general pediatricians over pediatric subspecialists.²⁵ Our findings suggest that the trend for adult patients remains, but there has been a shift away from general pediatricians toward subspecialists for pediatric referrals. These new patterns may be a consequence of greater availability of pediatric subspecialists, greater exposure of family physicians to pediatric consultants, and a larger share of family physicians who have completed residency training.

Implications for physician training

Fifty conditions accounted for 76% of all specialty referrals made during office visits in this study. Interactions with most types of specialists are generally limited to a few conditions; 3 health problems accounted for more than half of referrals to most specialties. Educators should ensure that these commonly referred conditions are emphasized in curricula that provide family physicians with the skills necessary to expand their scope of practice, when appropriate; determine when a patient should be referred; and identify the type of practitioner to whom the patient should be sent.

Physicians in training should be taught the skills required to recognize the boundaries of their clinical uncertainty and scopes of practice. A challenge for educators is to assist trainees in determining when to tolerate clinical uncertainty while employing a watchful waiting approach and when to initiate a more aggressive evaluation, including when to obtain specialty referral. Modes of implementing these approaches are likely to differ across conditions. Thus, it makes sense in physician training to place the greatest emphasis on conditions for which family physicians commonly refer.

Under certain circumstances, patient request for a specialty consultation may be a sufficient and legitimate reason for referral. For example, as we found in this study, patients with uncommon health problems may seek reassurance from specialists skilled in the management of their specific condition. Managing access to specialists, particularly when the physician is acting as an administrative gatekeeper to referrals, can be challenging.²¹ When doctors and patients disagree on the need for referral, patients may become dissatisfied with their health care²⁶ and decide to self-refer to specialty care.¹⁷ In consideration of the increasing complexity of medical care, developing skills that help physicians discuss and negotiate access to specialized services with both patients and specialists has never been more timely.

Acknowledgments

This study was funded by grant no. R01 HS09377 from the Agency for Healthcare Research and Quality. James Werner and Laurie Vorel provided technical assistance with data collection and project implementation. Many physicians collected data for this study. Their time and devotion were invaluable to the success of this study. These physicians are listed by the states in which they practice. Arizona: Scott Ekdahl, DO; Arkansas: John Scott, MD; California: Andrew Ness, MD; Colorado: Howard Corren, MD; Nell Davis, MD; Timothy Dudley, MD; Audrey Farley, MD; Tillman Farley, MD; Charles Kay, MD; Joan FAMILY PHYSICIANS’ REFERRAL DECISIONS MacEachen, MD; George Maxted, MD; John Miller, MD; Kathy Miller, MD; Steven Milligan, MD; Frank Reed, MD; Louise Schottsteadt, MD; Lynne Spicer, MD; Laura Stein, MD; Lynn Strange, MD; Dan Sullivan, MD; Georgia: Linda Casteel, MD; Randy Cronic, MD; Bruno Denis, MD; Keith Ellis, MD; Kelly Erola, MD; Craig Fabel, MD; Russell Leubbert, DO; Richard Liotta, DO; Mark Majoch, MD; David Najjar, MD; James Snow, DO; Roslyn Taylor, MD; Illinois: Steven Lidvall, MD; Anna Meenan, MD; Eduardo Scholcoff, MD; Loyd Wollstadt, MD; Indiana: Paul Daluga, MD; Steven Phillipson, MD; Iowa: Ken Miller, MD; Janet Ryan, MD; Kansas: Wendell Ellis, DO; John R. Eplee, MD; Robert Moser, MD; Daniel Sontheimer, MD; Louisiana: Linda Stewart, MD; Michigan: Linda French, MD; John Hickner, MD; Minnesota: Ravi Balasubraman, MD; Dave Bucher, MD; William Davis, MD; Richard Gebhart, MD; Katie Guthrie, MD; Anthony Jaspers, MD; Timothy Komoto, MD; Glenn McCarty, DO; Stephen Mitrione, MD; Thomas Retzinger, MD; Paul Spilseth, MD; Ashlesha Tamboli, MD; Montana: Curt Kurtz, MD; Nevada: Coleen Lyons, MD; New Hampshire: Richard Douglass, MD; Paul Friedrichs, MD; Peter Hope, MD; Jonathan Mishcon, MD; New Jersey: John Orzano, MD; Winifred Waldron, MD; New York: Carmella Abraham, MD; R. Eugene Bailey, MD; Lorne Becker, MD; John DeSimone, MD; Miguel Diaz, MD; Rebecca Elliott, MD; John Glennon, MD; James Greenwald, MD; Glenn Griffin, MD; Eileen Hoffman, MD; L. Thomas Wolff, MD; North Carolina: Ed Bujold, MD; Thomas Detesco, MD; Dave Rogers, MD; Phil Sherrod, MD; Oklahoma: Laura Miller, DO; Mike Pontious, MD; Oregon: Douglas Eliason, DO; L.J. Fagnan, MD; Jerry Flaming, DO; Tom Flaming, DO; Jeffrey Humphrey, DO; Michael Kelber, MD; John Sattenspiel, MD; Pennsylvania: John Farmer, DO; Penitha Williams, MD; South Dakota: Fred Thanel, MD; Tennessee: Dan Brewer, MD; Michael Hartsell, MD; R. Louis Murphy, MD; John Parham, MD; Texas: Michael Averitt, DO; Sharon Barber, MD; Kim Patrick Bolton, MD; Robert Cortes, MD; Paul Gerdes, MD; Robert Henry, DO; Michael Kirkpatrick, MD; John Manning, MD; Shane Maxwell, MD; Luis Moreno, MD; Larry G. Padget, MD; Peter Sullivan, MD; Utah: Scott Endsley, MD; David Flinders, MD; Jim Giovino, MD; Eric Hogenson, MD; Dwayne Roberts, MD; Virginia: Duane Lawrence, MD; James Ledwith, MD; June Tunstall, MD; George Wortley, MD; Washington: John Anderson, MD; Elizabeth Wise, MD; West Virginia: Dan Doyle, MD; J. Michael Herr, DO; Wisconsin: Richard Anstett, MD, PhD; Walter Boisvert, MD; Lea Cornell, MD; Anne Eglash, MD; Rod Erickson, MD; Tom Frisby, MD; Terry Hankey, MD; Kevin Jessen, MD; Dan Landdeck, MD; Dave Lonsdorf, MD; Michael Pace, MD; Michael Saunders, MD; Catherine Soderqueist, MD; Jon Temte, MD; Vince Winklerprins, MD; Brian Woody, MD.

Primary care physicians make specialty referrals to obtain advice for clinically uncertain diagnostic evaluations or treatment plans, to obtain a specialized service that falls outside their scope of practice, because of patient or third-party requests, or because of a combination of these reasons.¹ The clinical reasons for these referral decisions include characteristics of the presenting health problem, the burden and severity of comorbidities, and patient preferences for various treatments and outcomes.

Previous research has shown that certain features ascribed to morbidities influence the likelihood of specialty referral. The type of diagnosis is the most obvious determinant. In one study,² adults with malignancies were 5 times more likely to be referred than those with respiratory illnesses. Similarly, in the Netherlands, Van Suijlekom-Smit and colleagues³ found more than an 8-fold variation among childhood diagnosis groups in the likelihood of referral. For patients with similar diagnoses, research has found that severe variants are more likely to be referred.^4-7 Specialty referral is also influenced by the array and complexity of comorbid conditions.⁸

The conceptual foundations of primary care provide further insight into how clinical factors may influence referral to specialty care. A defining feature of primary care is the provision of a comprehensive set of services that meets the majority of a population’s health needs.^9,10 Primary care physicians develop greater experience and expertise for health problems with which they are familiar than those that occur less often. It follows that they would seek specialist assistance for uncommon health problems. However, empirical evidence for this effect is currently lacking.

Our goal was to test the hypothesis that the frequency with which a condition is seen by primary care physicians (practice prevalence) influences the likelihood of referral from primary to specialty care. We use the term practice prevalence to mean the frequency of presentation to primary care physicians and to distinguish it from the frequency of occurrence in the community. Also, we examine the impact of other clinical factors on primary care physicians’ referral decisions, including patient age, sex, comorbidities, and the medical versus surgical nature of the target condition’s management.

Methods

Data Source and Study Sample

We used the 1989 to 1994 National Ambulatory Medical Care Surveys (NAMCS) to examine referrals made to specialist physicians during visits with primary care physicians. NAMCS is a nationally representative survey of office-based physician visits in the United States. Each year, a multistage probability sample of nonfederally funded US physicians who are engaged in patient care activities (excluding radiologists, anesthesiologists, and pathologists) is selected from the master files of the American Medical Association and the American Osteopathic Association. For 1 week each selected physician completes a questionnaire for a 20% to 100% systematic sample of patient visits. Details of the survey methodology and the survey instrument are presented elsewhere.¹¹ The distribution of patient age and sex remained consistent over the 6 years of data collection we used.¹² The 1995 to 1998 surveys were not used, because information on referral was not collected. Using the 1994 and 1998 NAMCS, Forrest and Whelan¹³ found that primary care practice patterns did not substantively differ over time. The pooled data set contained 219,830 visits, of which 78,107 (35.5%) were with generalists (self-reported specialty designation was family/general practice, general pediatrics, or general internal medicine).

Clinical Factors

To examine the referral characteristics of different types of conditions, International Classification of Diseases-ninth revision-Clinical Modification (ICD-9-CM) codes were grouped into clinically similar categories that we called expanded diagnosis clusters (EDCs). The EDCs were based on the diagnosis clusters originally developed by Schneeweiss and coworkers.¹⁴ EDCs were assigned to the vast majority of primary care visits in the data set, matching 93.6% of the unique diagnosis codes. EDCs were grouped into clinical domains based on the nature of the problems and the specialty most responsible for the care. We also classified each EDC according to whether the primary treatment approach was medical or surgical.

A practice-prevalence ratio was calculated for each condition. The numerator was the number of visits for the EDC under consideration, and the denominator was the total number of visits in the data set. These ratios were further divided into tertiles (ie, high-, medium-, and low-frequency EDCs.)

To account for the number and severity of comorbid conditions with which patients presented, we assigned a comorbidity index to each visit. The index was based on the aggregated diagnostic groups (ADGs), the building blocks of The Johns Hopkins Adjusted Clinical Groups Case-Mix system.¹⁵ The 32 ADGs represent morbidity groups that contain ICD-9-CM diagnosis codes that are similar with respect to likelihood of persistence, expected need for health care resources, and other clinical criteria. In the NAMCS, up to 3 ICD-9-CM codes were assigned per visit; thus, each visit was assigned 1 to 3 ADGs. A comorbidity index score was obtained by summing ADG-specific resource intensity weights.* Larger weights suggest more complex health problems and greater expected resource intensity. In previous research,¹³ the comorbidity index increased with patient age and distinguished type of primary care delivery site by morbidity burden. The comorbidity index was divided into tertiles representing high, medium, and low levels of comorbidity.

Data Analysis

The unadjusted association of each clinical factor with the probability of referral was evaluated in bivariate analyses using the chi-square statistic. We further analyzed the relationship between a condition’s practice prevalence and its referral rate using scatter plots and Pearson correlation coefficients. Because both the practice prevalence and referral rate variables were right skewed, we used a logarithmic transformation to normalize both. For the scatter plots we excluded conditions with unstable referral rate estimates because of small sample sizes. Specifically, a condition was included if the ratio of the difference between the 95% confidence interval upper and lower limits of the referral rates divided by the condition-specific referral rate was less than 1. For example, multiple sclerosis was excluded because it presented at a rate of just 6 per 10,000 visits, had a referral rate of 7.1%, and the difference between the 95% confidence interval limits divided by the referral rate was 2.7.

We conducted a multiple logistic regression analysis with the patient visit as the unit of analysis for examining the independent effect of practice prevalence of the presenting problem on the chances of referral. Controlling variables in this analysis included age, sex, comorbidity index tertile, and the medical versus surgical nature of the principal diagnosis. To determine if the clinical complexity of a visit as assessed by the comorbidity index modified the effect of practice prevalence, we also added interaction terms for these variables to the final model. The goodness of fit of the final model was tested using the Hosmer-Lemeshow statistic, which suggested adequate model fit.¹⁶ The regression analyses used the generalized estimating equation¹⁷ to control for the intraphysician correlation of clustered visits.

Results

The distribution of clinical characteristics is shown in Table 1. The population was somewhat more represented by women (57.5%) than men, and by children/adolescents (33.3%) and seniors (39.4%) than middle-aged persons. The average comorbidity index value was 0.63, but there was considerable variability and the distribution was right skewed (median=0.31; interquartile range=0.05-0.83). The mean EDC practice-prevalence estimate for patient visits was 51 per 1000 primary care visits, but this distribution was also right skewed (median=25; interquartile range=10-106).

In the bivariate analyses, the probability of referral was significantly related to age, sex, practice prevalence of the condition, patient comorbidity, and whether the condition was a surgical problem Table 2. For the 65 condition categories with stable referral rate estimates Figure 1, the correlation between the logarithms of the practice-prevalence ratios and the referral rates was -0.87 (P <.001). Also, this relationship remained after stratifying for the 35 medical (r=-0.82; P <.001) and 24 surgical conditions (r=-0.88; P=.001).

Table 3 shows the multiple logistic regression results modeling the probability of referral during an office visit as a function of the clinical factors. There were no significant differences in the chances of referring patients aged 11 to 64 years once measures of morbidity, such as patient comorbidity and practice prevalence of the principal diagnosis, were controlled. Men were 19% more likely to be referred than women, and medical conditions were 39% less likely to be referred than surgical ones.

Only one of the practice prevalence/comorbidity interaction terms was significantly different from 0: commonly occurring conditions presenting among patients with high levels of comorbidity. This finding implies that the comorbidity has a stronger influence on the chances of referral for patients presenting with common problems than those presenting with less common problems.

Table 4 shows the estimated probabilities of referral based on differences in practice prevalence and comorbidity. These probabilities were obtained from the b coefficients in Table 3. The reference group for the probability estimates is women aged 18 to 44 years with health problems categorized as medical conditions. The chances of referral varied as much as 8-fold based on only the practice prevalence of the principal diagnosis and level of comorbidity.

Discussion

Our results support the hypothesis that the frequency with which patients’ health problems present to primary care physicians (practice prevalence) has a strong inverse relationship with the chances of referral to specialty care. Primary care physicians were more likely to send patients with uncommon problems to specialists and retain those with the most common conditions. This finding highlights the responsible judgment primary care physicians employ in recognizing the boundaries of their scope of practice. Practice prevalence is a defining feature of the primary care–specialty care interface.

Referring patients with uncommon problems to specialists is a rational way to organize medical care. Outcomes are related to the volume of patients managed with a specific condition.¹⁸ Specialists need to care for an adequate number of patients with uncommon problems to maintain clinical competence. Patient self-referral, however, which dilutes the prevalence of health problems presenting to specialists, may result in potentially invasive and expensive diagnostic approaches to patients more appropriately evaluated by primary care physicians.¹⁹

In addition to a condition’s practice prevalence, the number and severity of comorbidities managed during the visit influenced primary care physicians’ decisions to make specialty referrals. Also, we found an interaction effect between high practice prevalence and high levels of comorbidity. In other words, patients with uncommon conditions were commonly referred, regardless of the complexity of other conditions. The chances of referral markedly increased for patients with common conditions when they also presented with co-existing medically complex health problems. Thus, the rare presentations for which specialist assistance is sought may be a result of either the practice prevalence of the presenting problem or the overall complexity of a patient.

Men were more commonly referred than were women, after accounting for differences in the nature of their problems. A possible explanation for this finding is that because women make more office visits over a year than men,²⁰ their probability of referral during any given visit will be lower given roughly equal chances of referral between the 2 groups during the course of a year.

Further Research

We demonstrated that the potential need for surgical interventions was an important predictor of referral. Even after other clinical factors were controlled, medical conditions were 39% less likely to be referred than surgical ones. This is not surprising given that primary care physicians generally perform only minor office-based surgical procedures. But which surgical procedures should be in the scope of practice of primary care physicians? This question deserves further research and could be addressed in part by an analysis that is similar to the one presented here. Common outpatient procedures are candidates for inclusion as primary care services. Secondary considerations include the requirements and expense of necessary equipment, technical personnel, and training. Research that builds epidemiologic profiles of office-based procedures would be helpful in determining how responsibilities should be divided between generalists and specialists for these technical services.

Limitations

Several limitations in our study’s data source warrant consideration. First, the data set of visits provided information on primary care physicians’ referral decisions and did not elucidate whether patients actually received specialty care. Second, the sample was restricted to visits made to generalist physicians, excluding both obstetrician-gynecologists and medical subspecialists who may act as primary care physicians. Third, the NAMCS data set did not include hospital-based physicians, who are known to have higher referral rates than their office-based counterparts.¹³ Fourth, the unit of analysis was the visit rather than the patient. Patients with certain chronic conditions may have higher referral rates than suggested by our data if the measure used is the percentage of persons obtaining specialty care over a year. The advantage of focusing on the visit is that physician referral decisions can be examined rather than specialist use. Fifth, some conditions had lower than expected referral rates (eg, appendicitis had a referral rate of 46%), because the denominator for the referral rates was all visits made to generalists for the condition, which included both new presentations and follow-up visits. Finally, because of data limitations we did not assess the extent to which condition prevalence within an individual physician’s own practice affects his or her referral behavior.

Specialist visits can be initiated by primary care physician referral, patient self-referral, or specialist-to-specialist cross-referral. Although our database did not permit us to examine each of these pathways, other research suggests that primary care physician referral is the predominant route, particularly in health maintenance organizations.¹²

Conclusions

Our findings provide evidence that the boundaries between primary care physicians and specialists are defined in part by prevalence of health problems and the overall complexity of patients. Future research should focus on identifying modifiable characteristics of the physician-patient interaction, physicians, their practices, and the health system that influence referral decisions, after accounting for clinical factors. The appreciation of relevant clinical factors is critical to the fair application of administrative and financial constraints on physicians’ abilities to refer. Managed care plans that penalize physicians for high referral behavior, without adjusting for practice prevalence and comorbidity work, are contrary to the goal of providing quality patient care in the most appropriate settings. With more precise definitions of the clinical determinants of referral for populations, health systems can better gauge generalist and specialist workforce requirements.

Acknowledgments

This work was supported by the Agency for Healthcare Research and Quality grants #R01 and #HS09377. Barbara Starfield inspired this work and provided comments on the manuscript. We also thank Barbara Bartman, Norm Smith, MD, MPH, and Jonathan Weiner MD, MPH, for their review and comments on the manuscript. Mia Kang and Sarah von Schrader provided excellent technical assistance.

Related Resources

• Agency for Healthcare Research and Quality, Primary Care Subdirectory Page—includes research articles on primary care referral patterns and coordination of care among referring physicians and specialists. http://www.ahrq.gov/research/primarix.htm

Primary care physicians make specialty referrals to obtain advice for clinically uncertain diagnostic evaluations or treatment plans, to obtain a specialized service that falls outside their scope of practice, because of patient or third-party requests, or because of a combination of these reasons.¹ The clinical reasons for these referral decisions include characteristics of the presenting health problem, the burden and severity of comorbidities, and patient preferences for various treatments and outcomes.

Previous research has shown that certain features ascribed to morbidities influence the likelihood of specialty referral. The type of diagnosis is the most obvious determinant. In one study,² adults with malignancies were 5 times more likely to be referred than those with respiratory illnesses. Similarly, in the Netherlands, Van Suijlekom-Smit and colleagues³ found more than an 8-fold variation among childhood diagnosis groups in the likelihood of referral. For patients with similar diagnoses, research has found that severe variants are more likely to be referred.^4-7 Specialty referral is also influenced by the array and complexity of comorbid conditions.⁸

The conceptual foundations of primary care provide further insight into how clinical factors may influence referral to specialty care. A defining feature of primary care is the provision of a comprehensive set of services that meets the majority of a population’s health needs.^9,10 Primary care physicians develop greater experience and expertise for health problems with which they are familiar than those that occur less often. It follows that they would seek specialist assistance for uncommon health problems. However, empirical evidence for this effect is currently lacking.

Our goal was to test the hypothesis that the frequency with which a condition is seen by primary care physicians (practice prevalence) influences the likelihood of referral from primary to specialty care. We use the term practice prevalence to mean the frequency of presentation to primary care physicians and to distinguish it from the frequency of occurrence in the community. Also, we examine the impact of other clinical factors on primary care physicians’ referral decisions, including patient age, sex, comorbidities, and the medical versus surgical nature of the target condition’s management.

Methods

Data Source and Study Sample

We used the 1989 to 1994 National Ambulatory Medical Care Surveys (NAMCS) to examine referrals made to specialist physicians during visits with primary care physicians. NAMCS is a nationally representative survey of office-based physician visits in the United States. Each year, a multistage probability sample of nonfederally funded US physicians who are engaged in patient care activities (excluding radiologists, anesthesiologists, and pathologists) is selected from the master files of the American Medical Association and the American Osteopathic Association. For 1 week each selected physician completes a questionnaire for a 20% to 100% systematic sample of patient visits. Details of the survey methodology and the survey instrument are presented elsewhere.¹¹ The distribution of patient age and sex remained consistent over the 6 years of data collection we used.¹² The 1995 to 1998 surveys were not used, because information on referral was not collected. Using the 1994 and 1998 NAMCS, Forrest and Whelan¹³ found that primary care practice patterns did not substantively differ over time. The pooled data set contained 219,830 visits, of which 78,107 (35.5%) were with generalists (self-reported specialty designation was family/general practice, general pediatrics, or general internal medicine).

Clinical Factors

To examine the referral characteristics of different types of conditions, International Classification of Diseases-ninth revision-Clinical Modification (ICD-9-CM) codes were grouped into clinically similar categories that we called expanded diagnosis clusters (EDCs). The EDCs were based on the diagnosis clusters originally developed by Schneeweiss and coworkers.¹⁴ EDCs were assigned to the vast majority of primary care visits in the data set, matching 93.6% of the unique diagnosis codes. EDCs were grouped into clinical domains based on the nature of the problems and the specialty most responsible for the care. We also classified each EDC according to whether the primary treatment approach was medical or surgical.

A practice-prevalence ratio was calculated for each condition. The numerator was the number of visits for the EDC under consideration, and the denominator was the total number of visits in the data set. These ratios were further divided into tertiles (ie, high-, medium-, and low-frequency EDCs.)

To account for the number and severity of comorbid conditions with which patients presented, we assigned a comorbidity index to each visit. The index was based on the aggregated diagnostic groups (ADGs), the building blocks of The Johns Hopkins Adjusted Clinical Groups Case-Mix system.¹⁵ The 32 ADGs represent morbidity groups that contain ICD-9-CM diagnosis codes that are similar with respect to likelihood of persistence, expected need for health care resources, and other clinical criteria. In the NAMCS, up to 3 ICD-9-CM codes were assigned per visit; thus, each visit was assigned 1 to 3 ADGs. A comorbidity index score was obtained by summing ADG-specific resource intensity weights.* Larger weights suggest more complex health problems and greater expected resource intensity. In previous research,¹³ the comorbidity index increased with patient age and distinguished type of primary care delivery site by morbidity burden. The comorbidity index was divided into tertiles representing high, medium, and low levels of comorbidity.

Data Analysis

The unadjusted association of each clinical factor with the probability of referral was evaluated in bivariate analyses using the chi-square statistic. We further analyzed the relationship between a condition’s practice prevalence and its referral rate using scatter plots and Pearson correlation coefficients. Because both the practice prevalence and referral rate variables were right skewed, we used a logarithmic transformation to normalize both. For the scatter plots we excluded conditions with unstable referral rate estimates because of small sample sizes. Specifically, a condition was included if the ratio of the difference between the 95% confidence interval upper and lower limits of the referral rates divided by the condition-specific referral rate was less than 1. For example, multiple sclerosis was excluded because it presented at a rate of just 6 per 10,000 visits, had a referral rate of 7.1%, and the difference between the 95% confidence interval limits divided by the referral rate was 2.7.

We conducted a multiple logistic regression analysis with the patient visit as the unit of analysis for examining the independent effect of practice prevalence of the presenting problem on the chances of referral. Controlling variables in this analysis included age, sex, comorbidity index tertile, and the medical versus surgical nature of the principal diagnosis. To determine if the clinical complexity of a visit as assessed by the comorbidity index modified the effect of practice prevalence, we also added interaction terms for these variables to the final model. The goodness of fit of the final model was tested using the Hosmer-Lemeshow statistic, which suggested adequate model fit.¹⁶ The regression analyses used the generalized estimating equation¹⁷ to control for the intraphysician correlation of clustered visits.

Results

The distribution of clinical characteristics is shown in Table 1. The population was somewhat more represented by women (57.5%) than men, and by children/adolescents (33.3%) and seniors (39.4%) than middle-aged persons. The average comorbidity index value was 0.63, but there was considerable variability and the distribution was right skewed (median=0.31; interquartile range=0.05-0.83). The mean EDC practice-prevalence estimate for patient visits was 51 per 1000 primary care visits, but this distribution was also right skewed (median=25; interquartile range=10-106).

In the bivariate analyses, the probability of referral was significantly related to age, sex, practice prevalence of the condition, patient comorbidity, and whether the condition was a surgical problem Table 2. For the 65 condition categories with stable referral rate estimates Figure 1, the correlation between the logarithms of the practice-prevalence ratios and the referral rates was -0.87 (P <.001). Also, this relationship remained after stratifying for the 35 medical (r=-0.82; P <.001) and 24 surgical conditions (r=-0.88; P=.001).

Table 3 shows the multiple logistic regression results modeling the probability of referral during an office visit as a function of the clinical factors. There were no significant differences in the chances of referring patients aged 11 to 64 years once measures of morbidity, such as patient comorbidity and practice prevalence of the principal diagnosis, were controlled. Men were 19% more likely to be referred than women, and medical conditions were 39% less likely to be referred than surgical ones.

Only one of the practice prevalence/comorbidity interaction terms was significantly different from 0: commonly occurring conditions presenting among patients with high levels of comorbidity. This finding implies that the comorbidity has a stronger influence on the chances of referral for patients presenting with common problems than those presenting with less common problems.

Table 4 shows the estimated probabilities of referral based on differences in practice prevalence and comorbidity. These probabilities were obtained from the b coefficients in Table 3. The reference group for the probability estimates is women aged 18 to 44 years with health problems categorized as medical conditions. The chances of referral varied as much as 8-fold based on only the practice prevalence of the principal diagnosis and level of comorbidity.

Discussion

Our results support the hypothesis that the frequency with which patients’ health problems present to primary care physicians (practice prevalence) has a strong inverse relationship with the chances of referral to specialty care. Primary care physicians were more likely to send patients with uncommon problems to specialists and retain those with the most common conditions. This finding highlights the responsible judgment primary care physicians employ in recognizing the boundaries of their scope of practice. Practice prevalence is a defining feature of the primary care–specialty care interface.

Referring patients with uncommon problems to specialists is a rational way to organize medical care. Outcomes are related to the volume of patients managed with a specific condition.¹⁸ Specialists need to care for an adequate number of patients with uncommon problems to maintain clinical competence. Patient self-referral, however, which dilutes the prevalence of health problems presenting to specialists, may result in potentially invasive and expensive diagnostic approaches to patients more appropriately evaluated by primary care physicians.¹⁹

In addition to a condition’s practice prevalence, the number and severity of comorbidities managed during the visit influenced primary care physicians’ decisions to make specialty referrals. Also, we found an interaction effect between high practice prevalence and high levels of comorbidity. In other words, patients with uncommon conditions were commonly referred, regardless of the complexity of other conditions. The chances of referral markedly increased for patients with common conditions when they also presented with co-existing medically complex health problems. Thus, the rare presentations for which specialist assistance is sought may be a result of either the practice prevalence of the presenting problem or the overall complexity of a patient.

Men were more commonly referred than were women, after accounting for differences in the nature of their problems. A possible explanation for this finding is that because women make more office visits over a year than men,²⁰ their probability of referral during any given visit will be lower given roughly equal chances of referral between the 2 groups during the course of a year.

Further Research

We demonstrated that the potential need for surgical interventions was an important predictor of referral. Even after other clinical factors were controlled, medical conditions were 39% less likely to be referred than surgical ones. This is not surprising given that primary care physicians generally perform only minor office-based surgical procedures. But which surgical procedures should be in the scope of practice of primary care physicians? This question deserves further research and could be addressed in part by an analysis that is similar to the one presented here. Common outpatient procedures are candidates for inclusion as primary care services. Secondary considerations include the requirements and expense of necessary equipment, technical personnel, and training. Research that builds epidemiologic profiles of office-based procedures would be helpful in determining how responsibilities should be divided between generalists and specialists for these technical services.

Limitations

Several limitations in our study’s data source warrant consideration. First, the data set of visits provided information on primary care physicians’ referral decisions and did not elucidate whether patients actually received specialty care. Second, the sample was restricted to visits made to generalist physicians, excluding both obstetrician-gynecologists and medical subspecialists who may act as primary care physicians. Third, the NAMCS data set did not include hospital-based physicians, who are known to have higher referral rates than their office-based counterparts.¹³ Fourth, the unit of analysis was the visit rather than the patient. Patients with certain chronic conditions may have higher referral rates than suggested by our data if the measure used is the percentage of persons obtaining specialty care over a year. The advantage of focusing on the visit is that physician referral decisions can be examined rather than specialist use. Fifth, some conditions had lower than expected referral rates (eg, appendicitis had a referral rate of 46%), because the denominator for the referral rates was all visits made to generalists for the condition, which included both new presentations and follow-up visits. Finally, because of data limitations we did not assess the extent to which condition prevalence within an individual physician’s own practice affects his or her referral behavior.

Specialist visits can be initiated by primary care physician referral, patient self-referral, or specialist-to-specialist cross-referral. Although our database did not permit us to examine each of these pathways, other research suggests that primary care physician referral is the predominant route, particularly in health maintenance organizations.¹²

Conclusions

Our findings provide evidence that the boundaries between primary care physicians and specialists are defined in part by prevalence of health problems and the overall complexity of patients. Future research should focus on identifying modifiable characteristics of the physician-patient interaction, physicians, their practices, and the health system that influence referral decisions, after accounting for clinical factors. The appreciation of relevant clinical factors is critical to the fair application of administrative and financial constraints on physicians’ abilities to refer. Managed care plans that penalize physicians for high referral behavior, without adjusting for practice prevalence and comorbidity work, are contrary to the goal of providing quality patient care in the most appropriate settings. With more precise definitions of the clinical determinants of referral for populations, health systems can better gauge generalist and specialist workforce requirements.

Acknowledgments

This work was supported by the Agency for Healthcare Research and Quality grants #R01 and #HS09377. Barbara Starfield inspired this work and provided comments on the manuscript. We also thank Barbara Bartman, Norm Smith, MD, MPH, and Jonathan Weiner MD, MPH, for their review and comments on the manuscript. Mia Kang and Sarah von Schrader provided excellent technical assistance.

Related Resources

• Agency for Healthcare Research and Quality, Primary Care Subdirectory Page—includes research articles on primary care referral patterns and coordination of care among referring physicians and specialists. http://www.ahrq.gov/research/primarix.htm

Primary care physicians make specialty referrals to obtain advice for clinically uncertain diagnostic evaluations or treatment plans, to obtain a specialized service that falls outside their scope of practice, because of patient or third-party requests, or because of a combination of these reasons.¹ The clinical reasons for these referral decisions include characteristics of the presenting health problem, the burden and severity of comorbidities, and patient preferences for various treatments and outcomes.

Previous research has shown that certain features ascribed to morbidities influence the likelihood of specialty referral. The type of diagnosis is the most obvious determinant. In one study,² adults with malignancies were 5 times more likely to be referred than those with respiratory illnesses. Similarly, in the Netherlands, Van Suijlekom-Smit and colleagues³ found more than an 8-fold variation among childhood diagnosis groups in the likelihood of referral. For patients with similar diagnoses, research has found that severe variants are more likely to be referred.^4-7 Specialty referral is also influenced by the array and complexity of comorbid conditions.⁸

The conceptual foundations of primary care provide further insight into how clinical factors may influence referral to specialty care. A defining feature of primary care is the provision of a comprehensive set of services that meets the majority of a population’s health needs.^9,10 Primary care physicians develop greater experience and expertise for health problems with which they are familiar than those that occur less often. It follows that they would seek specialist assistance for uncommon health problems. However, empirical evidence for this effect is currently lacking.

Our goal was to test the hypothesis that the frequency with which a condition is seen by primary care physicians (practice prevalence) influences the likelihood of referral from primary to specialty care. We use the term practice prevalence to mean the frequency of presentation to primary care physicians and to distinguish it from the frequency of occurrence in the community. Also, we examine the impact of other clinical factors on primary care physicians’ referral decisions, including patient age, sex, comorbidities, and the medical versus surgical nature of the target condition’s management.

Methods

Data Source and Study Sample

We used the 1989 to 1994 National Ambulatory Medical Care Surveys (NAMCS) to examine referrals made to specialist physicians during visits with primary care physicians. NAMCS is a nationally representative survey of office-based physician visits in the United States. Each year, a multistage probability sample of nonfederally funded US physicians who are engaged in patient care activities (excluding radiologists, anesthesiologists, and pathologists) is selected from the master files of the American Medical Association and the American Osteopathic Association. For 1 week each selected physician completes a questionnaire for a 20% to 100% systematic sample of patient visits. Details of the survey methodology and the survey instrument are presented elsewhere.¹¹ The distribution of patient age and sex remained consistent over the 6 years of data collection we used.¹² The 1995 to 1998 surveys were not used, because information on referral was not collected. Using the 1994 and 1998 NAMCS, Forrest and Whelan¹³ found that primary care practice patterns did not substantively differ over time. The pooled data set contained 219,830 visits, of which 78,107 (35.5%) were with generalists (self-reported specialty designation was family/general practice, general pediatrics, or general internal medicine).

Clinical Factors

To examine the referral characteristics of different types of conditions, International Classification of Diseases-ninth revision-Clinical Modification (ICD-9-CM) codes were grouped into clinically similar categories that we called expanded diagnosis clusters (EDCs). The EDCs were based on the diagnosis clusters originally developed by Schneeweiss and coworkers.¹⁴ EDCs were assigned to the vast majority of primary care visits in the data set, matching 93.6% of the unique diagnosis codes. EDCs were grouped into clinical domains based on the nature of the problems and the specialty most responsible for the care. We also classified each EDC according to whether the primary treatment approach was medical or surgical.

A practice-prevalence ratio was calculated for each condition. The numerator was the number of visits for the EDC under consideration, and the denominator was the total number of visits in the data set. These ratios were further divided into tertiles (ie, high-, medium-, and low-frequency EDCs.)

To account for the number and severity of comorbid conditions with which patients presented, we assigned a comorbidity index to each visit. The index was based on the aggregated diagnostic groups (ADGs), the building blocks of The Johns Hopkins Adjusted Clinical Groups Case-Mix system.¹⁵ The 32 ADGs represent morbidity groups that contain ICD-9-CM diagnosis codes that are similar with respect to likelihood of persistence, expected need for health care resources, and other clinical criteria. In the NAMCS, up to 3 ICD-9-CM codes were assigned per visit; thus, each visit was assigned 1 to 3 ADGs. A comorbidity index score was obtained by summing ADG-specific resource intensity weights.* Larger weights suggest more complex health problems and greater expected resource intensity. In previous research,¹³ the comorbidity index increased with patient age and distinguished type of primary care delivery site by morbidity burden. The comorbidity index was divided into tertiles representing high, medium, and low levels of comorbidity.

Data Analysis

The unadjusted association of each clinical factor with the probability of referral was evaluated in bivariate analyses using the chi-square statistic. We further analyzed the relationship between a condition’s practice prevalence and its referral rate using scatter plots and Pearson correlation coefficients. Because both the practice prevalence and referral rate variables were right skewed, we used a logarithmic transformation to normalize both. For the scatter plots we excluded conditions with unstable referral rate estimates because of small sample sizes. Specifically, a condition was included if the ratio of the difference between the 95% confidence interval upper and lower limits of the referral rates divided by the condition-specific referral rate was less than 1. For example, multiple sclerosis was excluded because it presented at a rate of just 6 per 10,000 visits, had a referral rate of 7.1%, and the difference between the 95% confidence interval limits divided by the referral rate was 2.7.

We conducted a multiple logistic regression analysis with the patient visit as the unit of analysis for examining the independent effect of practice prevalence of the presenting problem on the chances of referral. Controlling variables in this analysis included age, sex, comorbidity index tertile, and the medical versus surgical nature of the principal diagnosis. To determine if the clinical complexity of a visit as assessed by the comorbidity index modified the effect of practice prevalence, we also added interaction terms for these variables to the final model. The goodness of fit of the final model was tested using the Hosmer-Lemeshow statistic, which suggested adequate model fit.¹⁶ The regression analyses used the generalized estimating equation¹⁷ to control for the intraphysician correlation of clustered visits.

Results

The distribution of clinical characteristics is shown in Table 1. The population was somewhat more represented by women (57.5%) than men, and by children/adolescents (33.3%) and seniors (39.4%) than middle-aged persons. The average comorbidity index value was 0.63, but there was considerable variability and the distribution was right skewed (median=0.31; interquartile range=0.05-0.83). The mean EDC practice-prevalence estimate for patient visits was 51 per 1000 primary care visits, but this distribution was also right skewed (median=25; interquartile range=10-106).

In the bivariate analyses, the probability of referral was significantly related to age, sex, practice prevalence of the condition, patient comorbidity, and whether the condition was a surgical problem Table 2. For the 65 condition categories with stable referral rate estimates Figure 1, the correlation between the logarithms of the practice-prevalence ratios and the referral rates was -0.87 (P <.001). Also, this relationship remained after stratifying for the 35 medical (r=-0.82; P <.001) and 24 surgical conditions (r=-0.88; P=.001).

Table 3 shows the multiple logistic regression results modeling the probability of referral during an office visit as a function of the clinical factors. There were no significant differences in the chances of referring patients aged 11 to 64 years once measures of morbidity, such as patient comorbidity and practice prevalence of the principal diagnosis, were controlled. Men were 19% more likely to be referred than women, and medical conditions were 39% less likely to be referred than surgical ones.

Only one of the practice prevalence/comorbidity interaction terms was significantly different from 0: commonly occurring conditions presenting among patients with high levels of comorbidity. This finding implies that the comorbidity has a stronger influence on the chances of referral for patients presenting with common problems than those presenting with less common problems.

Table 4 shows the estimated probabilities of referral based on differences in practice prevalence and comorbidity. These probabilities were obtained from the b coefficients in Table 3. The reference group for the probability estimates is women aged 18 to 44 years with health problems categorized as medical conditions. The chances of referral varied as much as 8-fold based on only the practice prevalence of the principal diagnosis and level of comorbidity.

Discussion

Our results support the hypothesis that the frequency with which patients’ health problems present to primary care physicians (practice prevalence) has a strong inverse relationship with the chances of referral to specialty care. Primary care physicians were more likely to send patients with uncommon problems to specialists and retain those with the most common conditions. This finding highlights the responsible judgment primary care physicians employ in recognizing the boundaries of their scope of practice. Practice prevalence is a defining feature of the primary care–specialty care interface.

Referring patients with uncommon problems to specialists is a rational way to organize medical care. Outcomes are related to the volume of patients managed with a specific condition.¹⁸ Specialists need to care for an adequate number of patients with uncommon problems to maintain clinical competence. Patient self-referral, however, which dilutes the prevalence of health problems presenting to specialists, may result in potentially invasive and expensive diagnostic approaches to patients more appropriately evaluated by primary care physicians.¹⁹

In addition to a condition’s practice prevalence, the number and severity of comorbidities managed during the visit influenced primary care physicians’ decisions to make specialty referrals. Also, we found an interaction effect between high practice prevalence and high levels of comorbidity. In other words, patients with uncommon conditions were commonly referred, regardless of the complexity of other conditions. The chances of referral markedly increased for patients with common conditions when they also presented with co-existing medically complex health problems. Thus, the rare presentations for which specialist assistance is sought may be a result of either the practice prevalence of the presenting problem or the overall complexity of a patient.

Men were more commonly referred than were women, after accounting for differences in the nature of their problems. A possible explanation for this finding is that because women make more office visits over a year than men,²⁰ their probability of referral during any given visit will be lower given roughly equal chances of referral between the 2 groups during the course of a year.

Further Research

We demonstrated that the potential need for surgical interventions was an important predictor of referral. Even after other clinical factors were controlled, medical conditions were 39% less likely to be referred than surgical ones. This is not surprising given that primary care physicians generally perform only minor office-based surgical procedures. But which surgical procedures should be in the scope of practice of primary care physicians? This question deserves further research and could be addressed in part by an analysis that is similar to the one presented here. Common outpatient procedures are candidates for inclusion as primary care services. Secondary considerations include the requirements and expense of necessary equipment, technical personnel, and training. Research that builds epidemiologic profiles of office-based procedures would be helpful in determining how responsibilities should be divided between generalists and specialists for these technical services.

Limitations

Several limitations in our study’s data source warrant consideration. First, the data set of visits provided information on primary care physicians’ referral decisions and did not elucidate whether patients actually received specialty care. Second, the sample was restricted to visits made to generalist physicians, excluding both obstetrician-gynecologists and medical subspecialists who may act as primary care physicians. Third, the NAMCS data set did not include hospital-based physicians, who are known to have higher referral rates than their office-based counterparts.¹³ Fourth, the unit of analysis was the visit rather than the patient. Patients with certain chronic conditions may have higher referral rates than suggested by our data if the measure used is the percentage of persons obtaining specialty care over a year. The advantage of focusing on the visit is that physician referral decisions can be examined rather than specialist use. Fifth, some conditions had lower than expected referral rates (eg, appendicitis had a referral rate of 46%), because the denominator for the referral rates was all visits made to generalists for the condition, which included both new presentations and follow-up visits. Finally, because of data limitations we did not assess the extent to which condition prevalence within an individual physician’s own practice affects his or her referral behavior.

Specialist visits can be initiated by primary care physician referral, patient self-referral, or specialist-to-specialist cross-referral. Although our database did not permit us to examine each of these pathways, other research suggests that primary care physician referral is the predominant route, particularly in health maintenance organizations.¹²

Conclusions

Our findings provide evidence that the boundaries between primary care physicians and specialists are defined in part by prevalence of health problems and the overall complexity of patients. Future research should focus on identifying modifiable characteristics of the physician-patient interaction, physicians, their practices, and the health system that influence referral decisions, after accounting for clinical factors. The appreciation of relevant clinical factors is critical to the fair application of administrative and financial constraints on physicians’ abilities to refer. Managed care plans that penalize physicians for high referral behavior, without adjusting for practice prevalence and comorbidity work, are contrary to the goal of providing quality patient care in the most appropriate settings. With more precise definitions of the clinical determinants of referral for populations, health systems can better gauge generalist and specialist workforce requirements.

Acknowledgments

This work was supported by the Agency for Healthcare Research and Quality grants #R01 and #HS09377. Barbara Starfield inspired this work and provided comments on the manuscript. We also thank Barbara Bartman, Norm Smith, MD, MPH, and Jonathan Weiner MD, MPH, for their review and comments on the manuscript. Mia Kang and Sarah von Schrader provided excellent technical assistance.

Related Resources

• Agency for Healthcare Research and Quality, Primary Care Subdirectory Page—includes research articles on primary care referral patterns and coordination of care among referring physicians and specialists. http://www.ahrq.gov/research/primarix.htm