Curbside Consultation Practices and Attitudes Among Primary Care Physicians and Medical Subspecialists

Context.— Informal (curbside) consultations are an integral part of medical culture and may be of great value to patients and primary care physicians. However, little is known about physicians' behavior or attitudes toward curbside consultation.

Objective.— To describe and compare curbside consultation practices and attitudes among primary care physicians and medical subspecialists.

Design.— Survey mailed in June 1997.

Participants.— Of 286 primary care physicians and 252 subspecialists practicing in Rhode Island, 213 primary care physicians and 200 subspecialists responded (response rate, 76.8%).

Main Outcome Measures.— Self-reported practices of, reasons for, and attitudes about curbside consultation.

Results.— Of primary care physicians, 70.4% (150/213) and 87.5% (175/200) of subspecialists reported participating in at least 1 curbside consultation during the previous week. In the previous week, primary care physicians obtained 3.2 curbside consultations, whereas subspecialists received 3.6 requests for curbside consultations. Subspecialties most frequently involved in curbside consultations were cardiology, gastroenterology, and infectious diseases; subspecialties that were requested to provide curbside consultations more often than they were formally consulted were endocrinology, infectious diseases, and rheumatology. Curbside consultations were most often used to select appropriate diagnostic tests and treatment plans and to determine the need for formal consultation. Subspecialists perceived more often than primary care physicians that information communicated in curbside consultations was insufficient (80.2% vs 49.8%; P<.001) and that important clinical detail was not described (77.6% vs 43.5%; P <.001). More subspecialists than primary care physicians felt that curbside consultations were essential for maintaining good relationships with other physicians (77.2% vs 38.6%; P <.001).

Conclusions.— Curbside consultation serves important functions in the practice of medicine. Despite the widespread use of curbside consultation, disagreement exists between primary care physicians and subspecialists as to the role of curbside consultation and the quality of the information exchanged.

A CURBSIDE consultation is an informal process whereby a physician obtains information or advice from another physician to assist in the management of a particular patient.^1,2 The consultant is generally unfamiliar with the patient and has not reviewed the patient's chart or examined the patient.² Thus, in contrast with a formal consultation, the consultant's recommendations or comments are based almost exclusively on information provided by the physician seeking advice, rather than from the patient or medical record.

With increasing complexity in the diagnosis and management of medical problems, Americans value their access to subspecialists.³ One feature of managed care has been to limit patients' access to subspecialists in an effort to cut costs for third-party payers.^3,4 As a result, many primary care physicians have had to assume greater responsibility for the care of conditions that have previously been considered the realm of the subspecialist. Thus, it is possible that primary care physicians will use more curbside consultations to obtain clinical advice and information in an attempt to maintain quality care while limiting the number of formal referrals.^2,5

Despite the importance of this informal part of medical practice, there has been little research to characterize physician behavior related to curbside consultation. Previous studies investigating curbside consultation have been limited to reports in endocrinology, infectious disease, or gastroenterology.^1,2,6-8 In the current study, we describe curbside and formal consultation patterns among Rhode Island physicians and examine differences in attitudes about curbside consultation between primary care physicians and internal medicine subspecialists.

We used the Rhode Island Department of Health Board of Medical Licensure and Discipline database to identify potential physicians for our study. The database included 3114 physicians with active licenses as of June 1996. We excluded 2188 physicians who did not identify themselves as having a primary or secondary specialty of internal medicine (IM), family practice (FP), or general practice (GP). Of the 926 remaining physicians, we categorized 651 as primary care physicians and 275 as subspecialists. We defined primary care physicians as those who reported their primary specialty as IM, FP, or GP and listed either no secondary specialty or a secondary specialty as IM, FP, GP, or adolescent medicine. We defined subspecialists as physicians with a primary specialty of one of the following internal medicine subspecialties: allergy, allergy and immunology, cardiovascular diseases, cardiac electrophysiology, critical care medicine, diabetes, endocrinology, gastroenterology, geriatrics, hematology, infectious diseases, medical oncology, nephrology, pulmonary diseases, or rheumatology. Of the 651 primary care physicians, we randomly selected 50% to obtain a sample of 325. We included all 275 subspecialists in our sample. A total of 39 primary care physicians and 23 subspecialists were ineligible (no longer in practice, unreachable at current address, moved out of state, or deceased) and therefore were excluded.

We developed a self-administered survey with 4 sections (the survey is available from the authors). We defined curbside consultation as "an informal process whereby a physician obtains information or advice from another physician to assist in the management of a particular patient. The consultant neither reviews the patient's record nor examines the patient and does not document his/her recommendations." We defined formal consultation as "a process whereby a physician refers a patient to another physician. The consultant reviews the patient's record, examines the patient, and formally documents his/her recommendations."

We asked physician respondents to estimate how many formal and curbside consultations they had requested (inpatient and outpatient) during the previous week and how many minutes they spent on average during these consultations. We asked respondents to indicate on a 4-point scale ranging from never to frequently how often they obtained curbside consultations from 5 different physical locations and how often they used curbside consultations for assistance in 8 different patient care situations. We also asked respondents to indicate their agreement on a 5-point Likert scale ranging from strongly disagree to strongly agree with 10 statements about curbside consultation, which addressed issues of general satisfaction, quality of information, professional relations, autonomy, time, financial incentives, and medical education. We included questions about physician practice characteristics and demographics.

Survey questions given to subspecialists sought to obtain information about their receiving requests to provide curbside consultations, whereas the questions given to primary care physicians sought information about their requesting curbside consultations. The survey sent to primary care physicians contained 2 additional questions: what characteristics of subspecialists do primary care physicians deem most important when selecting subspecialists to approach, and which 3 subspecialties do they most frequently consult formally and informally? The survey was pilot-tested for clarity and content by 7 physicians from the Division of General Internal Medicine and subspecialty departments of Rhode Island Hospital, Providence, and its contents were revised.

Each physician who was selected to participate was sent the survey with a self-addressed, postage-paid return envelope and a letter detailing the purpose of the study. The first mailing occurred in June 1997. Nonresponders received up to 3 follow-up reminder mailings 3, 5, and 10 weeks later. Each follow-up mailing included another copy of the survey. The study was approved by Rhode Island Hospital's institutional review board.

Survey responses were linked with data contained in the Rhode Island medical licensure file for descriptive statistics. Wilcoxon rank sum tests, χ² tests, and t tests were used as was appropriate. All analyses were performed using STATA software, Version 5.0 (STATA Corp, College Station, Tex).

Overall, 413 (76.8%) of 538 physicians responded to the survey. The response rate did not differ between primary care physicians (213/286, 74%) and subspecialists (200/252, 79%) and it did not differ across the 4 most common subspecialties (cardiology, gastroenterology, hematology/oncology, and pulmonary disease). There was no significant difference between responders and nonresponders with respect to sex, age, number of years in practice, or number of years licensed to practice medicine in Rhode Island.

Of the primary care physicians, 55% were internists, 38% were family physicians, and 7% were general practitioners (Table 1). Of the subspecialists, the largest groups were cardiology (27%), gastroenterology (16%), and hematology/oncology (16%) (Table 1). There was no significant difference between subspecialists and primary care physicians with respect to age, number of hours of reported direct patient care, or years licensed to practice medicine in Rhode Island. The majority of the respondents for both the primary care and subspecialist groups were men, although nearly twice as many primary care physicians were women compared with subspecialists (29% vs 12.5%; P<.001). Subspecialists were more likely than primary care physicians to be US medical school graduates and board certified.

The majority of both primary care physicians and subspecialists were in solo or single specialty practices. Less than 10% of reimbursement was derived from capitated contracts. Primary care physicians saw more patients than subspecialists in the outpatient setting (80 vs 51 visits per week; P<.001) but fewer inpatients (10 vs 18 visits per week; P<.001).

Nearly 30% of primary care physicians reported obtaining no curbside consultations during the previous week, whereas 12.5% of subspecialists reported providing no curbside consultations during the previous week. Physicians who did not obtain any curbside consultations were significantly older (51 years vs 45 years; P<.001), in practice longer (15.2 years vs 11.2 years; P<.001) and more likely to be in solo practice (42% vs 23%; P<.001). Primary care physicians who obtained curbside consultations obtained an average of 3.2 consultations per week. Subspecialists received 3.6 requests for curbside consultations per week. Both groups of physicians participated in formal consultations approximately 3 times more often than curbside consultations. Subspecialists estimated that curbside consultations require much less time to complete than formal consultations (8 minutes vs 51 minutes). When the curbside consultation practices were examined by subspecialty, specialists in infectious disease and endocrinology received more consultations in the previous week (6.8 and 4.2 per week, respectively) compared with other subspecialties: cardiology (3.4), gastroenterology (3.6), pulmonary disease (3.6), and hematology/oncology (2.8).

Primary care and subspecialist respondents both reported that curbside consultations occurred most frequently in person in the hospital, in person in the office setting, or by telephone (Table 2). Few physicians used e-mail sometimes or frequently as a route for curbside consultation. The location of consultations did not differ significantly across these subspecialties (data not shown). Primary care physicians felt that the quality of the consultant's formal consultations and the consultant's superior skills or knowledge base compared with that of other physicians in that specialty were the most important characteristics in selecting a curbside consultant (percentage responding fairly important or very important, 88.6% and 82.5%, respectively).

Primary care physicians who were employed by health maintenance organizations (HMOs, group or staff model) were more likely to obtain curbside consultations than those not employed by HMOs (5.6 vs 2.9 consultations per week; P=.02). Primary care physicians who estimated that at least 20% of their patients were enrolled in capitated contracts had similar rates of curbside consultation as physicians who estimated that fewer than 20% of their patients were enrolled in capitated contracts (2.6 vs 2.2 curbside consultations per week; P=.38).

The most common reasons cited by primary care physicians for obtaining curbside consultation were to help select an appropriate diagnostic test, to determine need for formal consultation, to select an appropriate treatment plan, and to interpret laboratory or radiology data. Subspecialists cited helping select an appropriate diagnostic test, diagnosing a specific medical problem or condition, selecting an appropriate treatment plan, and determining the need for a formal consultation as the most common reasons they received curbside consultations (Table 2). However, specialists in pulmonary diseases and infectious diseases were much more likely to report that they provide curbside consultations sometimes or frequently to select an appropriate treatment plan (87% and 100%, respectively) compared with the other specialties (range, 52%-79%). Hematology/oncology and pulmonary disease specialists were more likely to report receiving requests for curbside consultations to help assess prognosis (59% and 52%, respectively) compared with other specialties (range, 32%-46%).

Primary care physicians most frequently requested curbside consultations from cardiology, gastroenterology, and infectious diseases. Formal consultations were most frequently obtained from cardiology, gastroenterology, and neurology. Seven subspecialties had significant differences between rates of formal and curbside consultation. Endocrinology, infectious diseases, and rheumatology received significantly more curbside consultations compared with formal consultations. Cardiology, neurology, ophthalmology, and surgery received requests for significantly fewer curbside consultations compared with formal consultations (Table 3).

Approximately half (49%) of the primary care physicians reported that their patients' care was sometimes or frequently "taken over" by the consultants following a formal consultation. Most of the subspecialists (61%) also reported that this happened sometimes or frequently (Table 4).

Slightly more than half of all subspecialists and primary care physicians reported that they enjoyed obtaining or receiving curbside consultations (Table 4). However, subspecialists perceived more often than primary care physicians that the information communicated in curbside consultations was insufficient (80.2% vs 49.8%; P <.001) and that important clinical information was missed (77.6% vs 43.5%; P<.001). Subspecialists also felt that after curbside consultations, primary care physicians would feel less obligated to follow the consultants' recommendations (46.9% vs 26.9%; P<.001).

Subspecialists were more likely than primary care physicians to report that curbside consultations were essential for maintaining good relations with other physicians (77.2% vs 38.6%; P <.001). Nearly half of both groups agreed that curbside consultations saved money for the patient and third-party payer and that these consultations were an important way for physicians to stay current with medical knowledge. Less than one quarter of both groups felt that curbside consultations should be used more often to reduce the number of inappropriate formal consultations. These results did not differ when board certified and non–board certified physicians were compared.

When attitudes about curbside consultation were compared across subspecialties, infectious disease specialists appeared to differ from the other specialties (Table 4). Infectious disease specialists were less likely than other specialists to report that they enjoy curbside consultations (15% vs 42%) and more likely to agree that insufficient information is exchanged (92% vs 80%) and that primary care physicians are less obligated to follow recommendations made during a curbside consultation (69% vs 47%). They were also more likely to agree that consultants are less enthusiastic with curbside consultations because they do not get reimbursed (54% vs 40%) and were less likely to consider curbside consultations essential for maintaining good professional relations (54% vs 77%).

Our data suggest that the majority of primary care physicians and subspecialists in internal medicine participate in curbside consultations. However, these physicians have differing viewpoints about the purpose and quality of curbside consultations. Most subspecialists are concerned about the adequacy of this method of providing input into the care of patients, whereas half of primary care physicians agree that the quality of information obtained by curbside consultation is adequate.

Primary care physicians request curbside consultations from subspecialists for a variety of reasons. Frequently, they ask for assistance with diagnostic test selection and interpretation of laboratory or radiology data, and they use curbside consultations to help determine the need for formal consultation. To answer such clinical questions, accurate and complete data need to be exchanged. However, subspecialists often feel that during curbside consultations, insufficient clinical information is provided compared with a formal consultation. In addition, subspecialists often feel that important findings may have been missed by their primary care colleagues. These results may suggest that subspecialists do not trust either primary care physicians' ability to communicate information or their history-taking and examining ability. Our findings corroborate those of Myers,¹ who found that incorrect information about the patient's history, physical examination, or laboratory data was given during curbside consultation.

Nearly half of subspecialists felt that primary care physicians would be less likely to follow a consultant's curbside advice compared with that which results from a formal consultation. Despite these concerns, 87.5% of subspecialists reported providing curbside consultations. Most subspecialists felt that curbside consultations were essential for maintaining good professional relations, acknowledging that curbside consultation is part of medical culture and that primary care physicians depend on it for patient care.

Nearly three quarters of primary care physicians requested curbside consultations, even though half also expressed concerns about the quality of information exchanged. This finding is consistent with a survey of physicians who requested curbside consultation from an infectious disease specialist at one hospital.¹ In that study, a majority of physicians requesting curbside consultations felt that inaccurate information was exchanged during a curbside consultation. Primary care physicians may continue to request curbside consultations despite these concerns because of the possible advantages of time and money saved and the educational information obtained.^1,8

In our study, primary care physicians and medical subspecialists reported that curbside consultations constituted one quarter of all consultation activity. This finding is similar to reports by specialists in endocrinology² and infectious diseases⁶ who prospectively tracked formal and curbside consultations. For subspecialists, the economics of curbside consultation is double-edged. On one hand, curbside consultation can lead to formal consultation and help maintain a referral base. On the other hand, subspecialists are not reimbursed for the time spent curbsiding. For cognitive subspecialties such as infectious diseases and endocrinology, in which providing information is often the only marketable service (as opposed to performing a procedure), curbside consultations may have substantial economic impact.⁵ We found that cognitive subspecialists received more curbside consultations than formal consultations and suspect that such disciplines may be more amenable to the brief, focused questions that characterize curbside consultation.⁹ Primary care physicians may feel more comfortable managing problems related to these disciplines independently because they either received better training in these areas or perceive the range of problems to be of lower acuity. In contrast, subspecialties that require specialized or hands-on physical examination (eg, dermatology) or the use of invasive procedures (eg, cardiology) received greater numbers of formal consultations.

The phenomenon of the consultant taking over the care of a patient is quite common according to both primary care physicians and subspecialists. This was also a concern of consulting physicians interviewed by an infectious disease specialist who tracked all his curbside consultations during a 12-month period.¹ The factors associated with takeovers of patient care have not been studied, but in our experience, formal referrals often result in follow-up care by the consultant, even when the referring physician did not request such concurrent follow-up care.

Curbside consultation may reduce the overall number of formal referrals required.⁸ This would allow primary care physicians to continue to care for their patients independently, a potential advantage to physicians practicing in a managed care or capitated setting. In our study, physicians practicing in a group-model or staff-model HMO were twice as likely to obtain curbside consultations than physicians not practicing in an HMO setting. An endocrinology group practice also reported that they were approached more frequently by HMO physicians than by fee-for-service physicians.² However, HMO-physician curbside consultations were more likely to result in formal consultations compared with fee-for-service physician curbside consultations.²

Physicians may use curbside consultation to keep current with medical information in addition to using journal articles, textbooks, and continuing medical education course material.¹⁰ Primary care physicians are responsible for patients presenting with a broad range of symptoms and illnesses. Covell et al¹⁰ observed that physicians contacted a colleague to answer 53% of their patient care questions resulting from an office visit. Nearly half of our surveyed primary care physicians agreed that curbside consultations are an important way to stay current with medical knowledge (data not shown). Information that primary care physicians seek may not yet be available in textbooks, and conducting a literature search may take more time than making a telephone call to a consultant.

Our study has several limitations. First, reported rates of curbside consultations may not necessarily reflect actual practice. However, our estimates of the duration and frequency of curbside consultation are similar to or less than those reported in other studies of informal consultation based on direct observation.^1,2,6,8 For example, we found that infectious disease specialists received an average of 6.8 curbside consultations in the previous week, which is similar to the number of curbside consultations reported in 2 prospective studies by infectious disease specialists (5.6 and 6.5 per week).^3,4 Second, our study was based in Rhode Island and may differ from curbside practices found in other states. There may be geographic variation in curbside consultation due to subspecialist availability and variations in physicians' perceptions of their own responsibilities. For instance, in Rhode Island, managed care remains relatively uncommon. Third, we assessed only curbside consultations requested by primary care physicians of subspecialists. We did not assess curbside consultations provided by primary care physicians or those requested by subspecialists. Finally, we did not assess the quality or appropriateness of curbside consultations and we did not evaluate the effect of curbside consultations on patient care.

In conclusion, curbside consultations are common and appear to be an integral part of medical culture. However, there remains disagreement between primary care physicians and subspecialists in Rhode Island regarding the quality of the information exchanged. Given this concern and the frequency with which curbside consultations occur, the effect of this common practice on patient outcomes needs to be studied. Until such data become available, physicians should keep in mind the potential disadvantages and advantages of curbside consultation.