Introduction

Improving inpatient handoffs is a major focus of efforts to enhance continuity of care and patient safety.1–9 In contrast, ambulatory handoffs of patients by a graduating resident to a junior colleague have received relatively little attention in the medical literature, with the limited published data focusing primarily on patient satisfaction,10,11 safety concerns,12,13 balancing resident workloads,13,14 and improving care at the end of academic year transition.15 Ambulatory malpractice claims data demonstrate that breakdowns in handoffs contribute to about 20% of cases, suggesting that communication issues and consequences occur in both ambulatory and inpatient settings.16

In 2010, an estimated 14 800 residents with a continuity clinic finished residency,17 likely accounting for more than a million transfers of care in the United States. Given the magnitude of this annual shift, it is important to better understand the components and risks of resident outpatient handoffs. Outpatient handoffs are unique: they need to convey important medical and psychosocial issues, such as the patient's living situation and access to health care; they occur in an almost simultaneous fashion, and they involve transition to a less experienced provider.

A practical way to facilitate improved, written, inpatient handoffs has been to apply a standardized electronic handoff tool18–20 and to place it within an electronic medical record (EMR), as used by some inpatient services.21 As this aspect has not been addressed in the medical literature, we undertook a pilot study to identify the factors affecting the quality and performance of outpatient handoffs, hypothesizing that the number of completed handoffs and the quality of information conveyed in them could be increased by a simple standardized handoff intervention. We created a prototype EMR intervention encouraging graduating residents to perform a standardized outpatient handoff via an EMR note, and studied the number and quality of the handoffs performed.

Methods

Study participants were residents graduating from our internal medicine and combined internal medicine-pediatrics residency program. At the end of each academic year, graduating residents' patients are assigned to an incoming intern. Of 15 graduating residents, all but 1 (93%) volunteered to participate.

Residents were randomized via an Internet-based tool (http://www.randomize.org) into an intervention and a control group. Seven residents (50%) were randomized to the intervention and sent an e-mail instructing them to write a short handoff note using the clinic's EMR system (Centricity, General Electric, Fairfield, CT). They were encouraged to prioritize writing handoff notes for patients who were medically complex. The instructions were worded as follows: “I would ask you to write a short note on these patients identified as yours in the diabetes QI [quality improvement] project, or any other patients of yours that have multiple medical issues.” The e-mail included explicit instructions to open an empty text box and explained which key elements to include (table 1). Residents were asked not to discuss the contents of the instructions to prevent contamination of the control group. A second, identical reminder e-mail was sent 1 week later to the intervention group. The 7 residents in the control group received an e-mail stating that no further instructions were forthcoming. Subjects in both groups were given 23 days to complete handoffs.

TABLE 1 Handoff Instructions and Key Elements to Be Included^a

View Fullscreen

The chief resident and faculty sent an e-mail to residents in both groups to compose an e-mail handoff note as had been the practice in the past and conveyed the names of the interns to whom they would transfer their patients. Only the intervention group received detailed EMR instructions. None of the residents in the study received protected time to complete handoff notes.

Every resident's EMR entries were reviewed from 14 days before the randomization date until 4 weeks after the end of residency. All handoffs were identified and scored by 2 independent, unblinded reviewers (M.J.D. and J.M.C.) using a predetermined point system. As no validated evaluation tool exists to assess outpatient handoffs, the scoring system was adapted from a validated, electronic, inpatient, handoff model.14 Zero or one point was awarded for each of 10 categories (table 2), and a weighted κ was calculated to evaluate interrater reliability.

TABLE 2 List of Categories From Which Handoffs Were Scored^a

View Fullscreen

The Georgetown University Institutional Review Board approved the study, and all residents provided informed consent.

Results

Residents in the intervention group collectively completed 16 handoff notes (range, 0–6 per resident), and 5 of the 7 residents (71%) in the intervention group performed standardized EMR handoffs. Of those 5 residents, 3 (60%) performed 2 handoffs, 1 resident (20%) performed 4 handoffs, and 1 resident (20%) performed 6 handoffs. In addition, 3 patients of the 16 handoffs (19%) by residents in the intervention group were handed off via e-mails that were not available to the investigators. Residents in the control group completed zero EMR handoff notes.

Evaluation of the handoff notes produced a wide range of scores (range, 3–7 out of 10). Potential reasons for this include variation in patients' medical complexity, resident effort, or resident's judgment in recognizing the need for a handoff. The mean score was 5.56, and scores had excellent interrater correlation (weighted κ, 0.77).

The categories contained in each handoff are found in the figure. Examination of the notes showed that most handoffs provided a paragraph of background information on the patient and then focused on recent clinical issues.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

Our pilot study suggests that a simple e-mail intervention instructing residents to use a standardized handoff process resulted in a modest increase in the number of outpatient handoff notes written in the EMR.

The published literature on inpatient and outpatient handoffs suggests that handoff content is critical, including new and existing clinical or psychosocial issues, pending tests, and clinical severity, whereas handoffs should also convey advice regarding longitudinal care.16 Many of the handoffs in our study focused on short-term concerns and tasks needing completion. We speculate that this may stem from residents' experience with inpatient handoffs, which have a short time horizon. Given the difficulties of arranging face-to-face handoffs at the end of residency and patients that often follow up with providers other than the primary resident, we thought a standardized procedure within the EMR would be most beneficial.

Our study has several limitations. First, few handoffs were performed by the residents in the study, and several residents in the intervention group e-mailed patient lists with short summaries to incoming interns outside of the EMR. These were not available for evaluation. It also is possible that handoffs within the EMR and those outside of the electronic system were not identified by the investigators.

Including a standard template within the EMR may facilitate more consistent inclusion of important items in the handoffs, but further study of this aspect is needed.

We were unable to estimate the size of each resident's patient panel, which precluded an estimate of effect size. Although handing off all patients in a panel would be ideal, each resident likely had patients not handed off, who would have benefitted, suggesting a need to set expectations in terms of handoff quantity. Targeted, educational intervention on the importance of handoffs is necessary to change culture and to produce greater compliance with handoff expectations. Informal communication with the residents after the conclusion of the study revealed that lack of time was the primary barrier to performing handoffs. Providing protected time and setting expectations for the number of handoffs may be a successful way of increasing the number of handoffs performed.

Conclusions

The opportunity for the outgoing and incoming care providers to communicate is brief and improving resident communication at the end of residency has the potential to improve patient safety. Our study showed that a simple intervention increased the number of handoffs of ambulatory patients by graduating residents. A standardized tool promoting efficient and comprehensive handoffs, allowing protected time, and providing clear expectations for handoffs are steps we identified to improve this process. Future investigation should focus on validating a tool to measure handoff quality and the effect of handoffs on provider and patient satisfaction, patient safety, and clinical outcomes.