I-CATCH: A Novel Bundle to Improve Postcall Morning Handoffs

Efforts to improve patient handoffs are critical for patient safety. Despite research demonstrating that patient handoff bundles such as I-PASS reduce medical errors and near misses,^1,2  efforts to replicate these interventions have been met with implementation challenges.³  A systematic review of educational interventions to improve patient handoffs concluded that “inadequate reporting of interventions, especially as they relate to educational theory, pedagogy, curricula, and resource requirements, continues to impede replication.”⁴ 

A previous study at our institution showed that nearly 40% of on-call issues on general internal medicine wards were not handed off to the daytime team the next morning.⁵  These omissions could lead to delays in recognizing and acting on clinically important information.⁶  To address this challenge, we supplemented our existing I-PASS training with a focus on morning handoff, and developed and implemented a morning handoff bundle based on the mnemonic I-CATCH (Identify patient; Characterize situation; Action–what was done overnight?; To do for the team in the morning; Confirm the Handoff). We sought to determine whether supplementing I-PASS with the I-CATCH bundle improved morning handoffs.

We compared morning handoff practices before (November 2015 to February 2016) and after (March 2016 to May 2016) I-CATCH bundle implementation across 4 general internal medicine wards at Sunnybrook Health Sciences Centre, a large tertiary-care academic hospital affiliated with the University of Toronto. At baseline, general internal medicine residents and faculty on the inpatient service already received I-PASS evening handoff training, consisting of two 1-hour interactive sessions per month. We supplemented this with a monthly 1-hour session of dedicated I-CATCH training for residents and faculty on the inpatient service. Faculty provided direct observation and feedback on resident handoff skills intermittently during the I-CATCH intervention phase and subsequent I-CATCH training sessions.

We developed the I-CATCH handoff bundle by drawing on elements of the I-PASS handoff bundle, such as dedicated space for handoffs,⁷  and included successful elements from a prior study of morning handoffs (ie, “running the list” by going through patients one-by-one).⁵  This bundle included: (1) monthly resident and staff physician training sessions; (2) structured communication using the I-CATCH mnemonic; and (3) dedicated time and space for handoffs. Two investigators (J.S.Z. and B.M.W.) developed and refined the I-CATCH bundle using rapid-cycle change methodology (see online supplemental material for descriptions of the I-CATCH bundle and plan-do-study-act cycles).⁸ 

The main outcome measure was the proportion of overnight on-call issues handed off by the on-call resident to the daytime team. Secondary outcomes evaluated the use of structured communication—defined by the proportion of handoffs that adhered to all 5 elements of I-CATCH.

We collected data using a similar methodology as in our prior study of morning handoffs.⁵  We randomly selected 1 of the 4 internal medicine teaching teams for data collection each day, excluding the first week of the 4-week rotation to allow residents to receive I-CATCH training. Prior to morning rounds, a research assistant identified overnight issues by reviewing records for all patients covered by the overnight resident (excluding patients newly admitted overnight because these patients were discussed in greater detail as part of case review), and then directly observed morning handoffs to determine the proportion of those on-call issues verbally handed off to the daytime team.

The research assistant also determined whether the resident used a structured communication approach and gathered data on other important handoff processes, such as systematically running the list, using a dedicated room, performing a direct handoff instead of using an intermediary, and number of distractions. We also collected demographic data for on-call resident physicians as well as measures reflecting on-call workload, including the number of new patients admitted overnight and the number of existing patients covered by the on-call resident.

The Sunnybrook Research Ethics Board approved this study.

We computed summary statistics and univariate analyses to compare handoff and resident characteristics before and after the I-CATCH intervention. We used a multivariable logistic regression analysis with a priori covariate selection to evaluate the association between the I-CATCH intervention and the completeness of morning verbal handoffs and to calculate the corresponding adjusted odds ratios (ORs) and 95% confidence intervals (CIs). We considered P < .05 statistically significant.

Statistical analyses used SAS University Edition (SAS Institute Inc, Cary, NC), and QI Macros (KnowWare International Inc, Denver, CO) was used to generate statistical process control charts.

We observed morning handoffs for 34 days before and 29 days after I-CATCH implementation, with 242 overnight issues identified pre-I-CATCH (22 residents observed) and 193 overnight issues post-I-CATCH (21 residents observed). There were more handoff distractions in the pre-I-CATCH phase (median of 2 [IQR 0–4] versus 0 [IQR 0–2], P < .014), but otherwise there were no differences in resident physician or morning handoff characteristics between the 2 phases (table 1).

Overall, the on-call resident handed off 58% of overnight events to the daytime team. There was no association between I-CATCH implementation and issues handed off in the morning (OR = 0.96; 95% CI 0.52–1.47; P = .85; table 2). However, running the list (OR = 1.74; 95% CI 1.1–2.77; P = .019), progress note documentation (OR = 3.80; 95% CI 2.19–6.60; P < .001), and direct handoff (OR = 4.84; 95% CI 1.43–16.42; P = .011) were associated with an increased likelihood of morning handoff of overnight events (table 2).

Completeness of morning handoff (ie, adherence to all 5 elements of the I-CATCH mnemonic) was greater after implementation of the I-CATCH bundle (56% [61 of 109] versus 30% [43 of 145], P < .001). The P chart (figure) showed special cause variation (ie, nonrandom variation) immediately before and after implementation of the I-CATCH bundle, suggesting that the improvement is likely due to the intervention contrasting with secular trends.

We improved the use of a structured communication approach among residents handing off on-call issues in the morning, but the overall proportion of on-call issues handed off did not increase. Systematically running the list, ensuring direct handoff, and writing a progress note about the issue correlated with higher rates of morning handoff.

Our intervention may not have achieved its intended impact due to inconsistent application of all handoff elements. Resident and faculty adherence to I-CATCH bundle elements was variable (table 1). Our use of rapid-cycle change methods afforded us the ability to further characterize implementation barriers. At times, teams felt that direct handoffs and systematically running the list did not fit naturally with team discussions in the morning, highlighting the difficulty we faced integrating morning handoff processes with existing workflow.

Based on resident feedback, we planned to modify our electronic health record to facilitate overnight documentation (eg, inclusion of a morning handoff feature with a summary list of overnight issues) to cue on-call residents and remind them to hand off these issues.^9,10  Our information technology (IT) department was unable to prioritize these changes. Also, our sample was limited and the time for assessing the I-CATCH intervention was short—29 days over a span of 3 months. It is possible that better adherence to handoff elements and improved handoff of overnight events may require more time and better familiarity with I-CATCH.

This study adds to the growing literature that suggests patient handoff interventions, like other well-known patient safety interventions, such as the surgical safety checklist¹¹  and the central line bundle,¹²  are difficult to replicate.^3,4  Morning handoff interventions also must attend to broader workflow changes to ensure direct handoff and increase adherence to key processes such as running the patient list and ensuring consistent documentation of overnight issues. These may be targets for future interventions to improve handoff practices.

Supplementing I-PASS training with the I-CATCH program improved resident use of a structured approach to communicate overnight issues at morning handoff, but did not increase the proportion of on-call issues that were handed off by residents.