Physician–Physician Communication: What’s the Hang-up?

What we have here is a failure to communicate.

Prison captain to prisoner Luke Jackson

Cool Hand Luke, 1967

Inadequate communication as patients transition across venues of care carries substantial risks. At the time of discharge, changes in medications may be missed, abnormal or pending test results ignored, and evolving aspects of the patient’s condition may not receive appropriate attention^1–3. The essential nature of timely communication at discharge has been recognized by the Joint Commission (TJC), which mandates that discharge summaries be completed within 30 days of discharge and specifies the core elements that need to be included⁴. Though the discharge summary is a key component of the transfer of information, inadequate communication at admission carries additional hazards. At the time of admission, inaccurate medication reconciliation may occur, key elements of the history missed, prior studies unnecessarily repeated, and important aspects of the home situation overlooked^5–7.

Given the importance, it could be assumed that communication with the primary care physician (PCP) is a routine staple of hospitalization and performed in a systematic manner along with other essential items, such as assessment of allergies and documentation of prior laboratory results. On the contrary, the inconsistent nature of physician–physician communication is well-established. A systematic review of the literature found that direct communication between inpatient physicians and PCPs occurred during only 3–20% of hospitalization⁸. Deficiencies noted have included poor quality of referral letters from general practitioners to inpatient physicians in Norway⁹; poor timeliness and quality of discharge summaries in Canada, England, and Australia^3,10–11; and discontinuity in the care plan after discharge at an academic hospital in the U.S.^2,12. These studies suggest that inadequate communication between inpatient and outpatient physicians is the norm rather than the exception.

Two studies in this issue of JGIM offer further insights into the nature of doctor-doctor communication in the inpatient arena. Bell and colleagues surveyed 1,772 PCPs for 1,078 hospitalized patients at six academic medical centers¹³. Of the 77% of PCPs who were aware that their patient was admitted, only 23% received direct communication from an inpatient physician at any point during hospitalization. Roy and colleagues examined communication patterns for readmitted patients at two academic medical centers to determine if the admitting team had contacted the prior inpatient team¹⁴. The results are consistent with other assessments of communication: only 43.7% of admitting teams had communicated with the prior inpatient teams.

Though the consequences of inadequate communication have been repeatedly demonstrated, the gains from enhanced communication have not been as clearly proven. The lack of definitive data may be due to inability to control for important confounders, inadequate power, inability to assess the quality of the communication, or intrinsic difficulties with communication preventing meaningful improvement in outcomes. In addition, most studies have examined the transfer of information at the end of hospitalization, and little examination has been done on communication at the time of admission. Benefit was shown in a randomized trial which found that an intervention to facilitate the transition from hospital to home, including a comprehensive discharge form completed by a discharge planning nurse and electronically transmitted to a nurse at the PCP’s office, markedly decreased the number of incomplete workups, though ED visits and readmissions were unchanged¹⁵. A large retrospective chart review found a trend towards decreased readmission (RR 0.74) for patients for whom a discharge summary was available at the time of their follow-up visit with their PCP¹⁶. Coleman and colleagues randomized 750 adults to usual care or to an intervention including a “transitions coach” and a patient-centered record designed to facilitate transfer of information across sites¹⁷. This comprehensive approach to enhancing communication achieved significant reductions in rehospitalization rates at 30 days (8.3 vs. 11.9 days) and was net cost-saving. Additional support comes from a study of the impact of discharge summaries in London, which noted that 24% of patients had management affected by delayed or poor discharge summaries as determined by their PCP³. In the study by Bell and colleagues in this issue, there was no significant association between the PCP having communicated with the inpatient team with the composite endpoint of death, readmission, or emergency department visits, though a nonsignificant 5% decrease in the composite outcome was noted¹³. Though this result could have been due to chance, the study was underpowered to find a small benefit. Given the likely scale of any potential gain, future studies will need to have sufficient power to detect modest improvements in outcomes. Also, though the authors adjusted for comorbidities, this study and others may be confounded by communication being more likely to occur in patients who are more complex and thus prone to complications¹⁴.

A concern is that the rapid growth of the hospitalist model may further exacerbate discontinuity of care. The use of hospitalists has expanded rapidly as fewer PCPs care for their hospitalized patients due to the need to achieve maximum productivity in the office, increasingly complex inpatient management, and increased pressure to minimize length of stay. The concern is supported by an administrative database review of 938,833 patients in Canada which found that the relative risks for readmission and death at 30 days were significantly decreased (by 3% and 5%, respectively) when patients were seen after discharge by the same physician who provided their care in the hospital¹⁸. Given the data demonstrating the association of discontinuity with impaired outcomes and that the hospitalist model is predicated on the belief that having physicians who limit their practice to inpatient management enhances patient care, the burden is on hospitalists to ensure that care is not compromised during the transitions.

Accomplishing a cultural change in the approach to inpatient-outpatient communication will be challenging. Roy and colleagues identified three major barriers to communication: being too busy (i.e., time); a perceived lack of benefit; and not knowing whom to contact, which was presumably due to lack of a readily-identifiable PCP¹⁴. The latter item is largely out of the control of inpatient physicians, and may only be addressed by a comprehensive assessment of healthcare in the U.S. The first two barriers, however, indicate that many inpatient physicians feel that communication with the PCP is of limited value and not a priority relative to other responsibilities. The finding that some clinicians feel that communication is unimportant suggests that more education and emphasis needs to be placed on the consequences of inadequate information transfer.

A necessary first step in transforming the current passive approach to transitions will be a call from national leaders and local champions to urge that communication beyond discharge summaries be considered an essential element of patient care. The Society of Hospital Medicine has been a leader in improving transitions of care, though the focus has been on enhancing the discharge process with less attention to communication at admission or during hospitalization^19,20. Local initiatives could include requiring chart documentation of communication, regular surveys of satisfaction of PCPs with hospitalist communication, and audits and feedback on the timeliness and quality of discharge summaries. A more systemic approach with potentially great impact would entail promoting widespread availability of health information technology (HIT) to allow providers to access information across venues and seamlessly integrate communication into the workflow of patient care. For example, electronic medical record (EMRs) allow providers at specified sites of care to access all current and prior progress notes, and some EMRs allow messages and results to be forwarded to other providers within the system.

Given the competing priorities and other responsibilities of inpatient clinicians, assignment of a clinician whose sole responsibility is to enhance communication, such as a discharge planner or, more comprehensively, a “transitions navigator,” could help facilitate transfer of information. The study by Coleman and colleagues suggests this role has the potential to improve outcomes, and may be cost-saving¹⁷. Trials confirming the effect at other hospitals and populations or demonstration projects yielding similar results in real-world settings would lend support to this model. Lastly, given the nihilism that some may harbor regarding the impact of communication, large rigorous trials demonstrating a clear association between communication and important outcomes, such as length of stay, readmission, and patient satisfaction, would likely also encourage discussion among providers.

High-quality communication between inpatient and outpatient physicians is essential to ensure patient safety during transitions. Prior studies have focused primarily on the discharge summary and have consistently shown the need for improvement. However, hospitalists and other inpatient physicians need to consider the discharge summary as a single method of interacting with their outpatient colleagues, and recognize that important information can be transmitted at admission and throughout hospitalization. Technological advances have made a myriad number of options available, including e-mail, fax, text messaging, and EMRs. Whether communication occurs via high-tech methods or simply by picking up the phone, these discussions need to become engrained in the culture and behavior of inpatient physicians.