In this issue, Pearson et al.1 report that gastroenterologists in a staff model HMO spent substantial time on “curbside” consultations. Curbside consultations are important in physician-to-physician communication but challenging to study because of their informal nature; more work of this type is needed to increase our understanding of the area. The time required for curbside consults should be included when physician workloads are assessed. While curbside consults do require time, they require less time of the patient and the consulted physician than the formal consultation that might otherwise have occurred. Therefore, physicians desiring to reduce their workload, or to keep it from growing, may view curbside consultations as preferable to formal consultations. Physicians in health care systems that penalize them for requesting formal consultations may also consider curbside consultations a desirable alternative, even though physicians who request curbside consults often assume greater responsibility for, and spend more time on, the clinical problem than they would by requesting a formal consult.
Besides influencing physician workload, curbside consults may have other effects, both desirable and undesirable, that should be considered. One potential advantage of curbside consults is that the requesting physician receives information sooner and has the opportunity for direct interaction with the consultant. Such direct communication may not occur with a formal consult, particularly in large systems of care. Direct, immediate communication may be of greater educational value for the requesting physician, and it may lead to more thorough or efficient work-ups before formal consultations. In addition, curbside consults are convenient for patients and save some of the costs associated with a formal consultation. By limiting the role of the consultant, curbside consults preserve the comprehensiveness of care given by the requesting physician.
There are, of course, potential disadvantages to curbside consultations. At least two questions arise relating to quality of care. First, can the consultant formulate recommendations, without directly evaluating a patient, that are as useful as those that would result from a formal consultation? Second, does implementation of recommendations by the requesting physician with less involvement from the consultant result in a process of care and outcomes that are equal to those following a formal consultation? While it is tempting to assume that consultants and referring physicians always know when a curbside consult can substitute for a formal consultation, the reality is that we don’t actually know what is lost, or gained, in the curbside process.
Curbside consultation is just one of the many possible variations of consultation. Other variations include multidisciplinary teams, telemedicine, and computer-based communications (the most notable of which is electronic mail). While these and other variations have the potential to enhance physician-to-physician communication, it is important that we carefully assess their impact on the process and outcomes of care.
Research on formal consultation in the United States has been fairly limited, especially compared to the much greater attention this topic receives from researchers in the United Kingdom.2 This discrepancy is surprising, given the growing scrutiny of the generalist-specialist interface by patients, physicians, government, and the press, who are increasingly concerned about the possible constraints on specialty care by managed care. Fortunately, the Agency for Health Care Policy and Research has identified the generalist-specialist interface as a priority research area and has funded work, currently underway, addressing the deficiency of knowledge in this area.
Although more accessible to study than informal consultation, the formal referral and consultation process remains a challenging research topic from several perspectives. Variations in referral rates have been studied in relation to patient, provider, and health-system characteristics.3–7 Unfortunately, different studies have used different methods to determine referral rates, making it difficult to compare results across studies.8 Measuring reliable referral rates at the individual provider level is particularly complex, especially when referrals are infrequent, as they are in a healthy population. The small-numbers problem is worse when studying referral rates to a specific specialty. Moreover, there are important problems when adjusting provider-level referral rates for case mix.9 Finally, as the study by Pearson and colleagues reminds us, referral rates fail to capture all sources of physician expertise that actually affect patient care.
Even after addressing these methodological issues, there remains an additional, perhaps more fundamental challenge to understanding referral and consultation. While referral rates can help identify important trends and potential over- or underuse of certain types of care, the real meaning of referral rates ultimately lies in understanding the impact of physician specialty on the process and outcomes of care. For example, one of the best studies examining the impact of physician specialty found few differences in outcomes for hypertension and diabetes among patients of generalists and specialists.10 Studies identifying the circumstances that benefit from specialty consultation will allow us to move beyond simply observing variations in referral rates to interpreting specialty-specific, condition-specific referral rates as either too high or too low for specific patient populations, thus allowing us to develop evidence-based referral guidelines for specialty consultations.—Steven J. Borowsky, MD, MPH,Minneapolis VAMC, HSR&D Field Program, Center for Chronic Disease Outcomes Research, Minneapolis, MN.