Automating pending labs list into discharge summaries Authors: Khanh T Nguyen, Chukwunedum Aniemeka, Nia Feaster, Misha C Tran, Matthew T Cerasale Published: 2025 ABSTRACT AbstractIntroductionPending test results not communicated during the post-acute care transition may pose harm and lead to diagnostic errors in patients. This quality improvement initiative aimed to assess the effectiveness of pending labs communication by automating a list of pending labs into discharge summaries.MethodsA SmartLink was developed in the EpicCare electronic health record that populates a list of pending labs and pathology studies and was added to the standard discharge summary template for hospitalists and resident physicians. A 2-month pre-post review of randomly selected discharge encounters was conducted to assess the frequencies of discharge summaries reporting pending labs, discharges with pending labs, discharges with eventually abnormal pending labs and discharges with results reported in follow-up notes.ResultsOf the 291 pre-intervention encounters reviewed, 130 (44.7%) had one or more labs pending, compared with 115 (40.6%) of 283 post-intervention encounters. 35 (26.9%) of the pre-intervention discharge summaries mentioned all pending labs, compared with 104 (90.4%) post-intervention, p value <0.01. There was no difference in the frequency of discharge encounters with results reported in follow-up notes; however, a trend towards significance was observed for those with abnormal results.DiscussionImplementing an automated link that pulls pending studies into discharge summary templates drastically improved documentation of pending labs. However, given the complexity of post-acute care transitions, more work is needed to ensure receipt and follow-up of this crucial communication. Introduction The post-acute care transition is associated with harm from incomplete or ineffective communication.1 In one report, 41% of patients had at least one pending study at discharge, 43% of which were eventually abnormal and 9.4% were potentially actionable.2 The discharge summary (DCS) is a ubiquitous tool used by inpatient providers to communicate important clinical information, and its standardisation has improved providers’ satisfaction with quality of information while reducing the perception of medical errors.3 However, only between 12% and 46% of pending tests are reported on DCS.4 A few interventions, ranging from education, audit and feedback, and use of electronic health record (EHR) tools, have had modest improvements in the documentation of pending tests on discharge.5 We aimed to assess the effectiveness of pending labs documentation using a novel approach of automating both a pending labs SmartLink and a default discharge summary template. Methods In this quality improvement initiative, we set a specific-measurable-attainable-relevant-time bound (SMART) aim to improve pending labs documentation on discharge to 90% or more of general medicine discharges by 3 months of implementation of an automated pending labs list into discharge summary templates. A process map (online supplemental file 1) of how pending labs are relayed to patients was developed through focus group interviews with key stakeholders involved in the inpatient to outpatient care transition process, including representatives for inpatient physicians, social worker, case manager, primary care physician and quality improvement director. We employed two clinical vignettes with scenarios involving pending results on discharge, obtained from the Association of American Medical Colleges’ Diagnostic Safety Toolkit,6 to guide discussions during the focus group. Opportunities for interventions were identified from the process map, sorted with an impact matrix, and narrowed to a SmartLink that pulls a list of pending labs into discharge summaries (figure 1). SmartLinks are shortcuts that automatically pulls in patient data from EpicCare EHR into a note. Studies were considered pending if the study’s status was not finalised. The pending labs SmartLink was added to an updated DCS template and populated as the default template for internal medicine residents and hospitalists, effective 18 January 2024. Example of pending labs SmartLink in a portion of the discharge summary note, 2025 Epic Systems Corporation. Discharge encounters for general medicine patients were selected for review, by random number generation, in a 2-month pre-period (1 August to 31 September 2023) and post-period (19 January and 14 March 2024), excluding death during the encounter. Chart reviews were conducted by 9 residents and 2 hospitalists, with a secondary review by KN. The primary outcome was the frequency of DCS listing all pending labs. Secondary measures were discharges with pending labs, discharges with eventually abnormal pending labs and discharges with results reported on follow-up notes. Call notes, clinic notes and other outpatient notes after the discharge date in the EHR ‘Notes’ section were reviewed, including from primary care and specialist providers, to assess for mention of results. Reviewers did not assess EHR notes that were not immediately available within the ‘Notes’ section. Categorical variables and frequencies were analysed using the χ² test. This study was formally determined to be quality improvement, not human subjects research, and was therefore not overseen by the Institutional Review Board, per institutional policy. Results We reviewed 291 of 1561 pre-intervention and 283 of 1410 post-intervention discharges. 130 (44.7%) pre-intervention discharges had one or more labs pending, compared with 115 (40.6%) post-intervention. For pre-intervention, 49.2% of encounters with pending labs had abnormal results, compared with 36.5% post-intervention. For the primary outcome, 35 (26.9%) of pre-intervention DCS mentioned all pending labs, compared with 104 (90.4%) post-intervention, p value <0.01. There was no difference in the frequency of discharge encounters with follow-up notes reporting results for all pending labs. For discharges with eventual abnormal labs, 20.3% had follow-up notes mentioning results pre-intervention, compared with 29.2% post-intervention, p=0.33 (table 1). Pathology results were the predominant abnormal results mentioned in follow-up notes and clinically relevant culture results had the largest relative improvement in follow-up note documentation (table 2). Table 1ResultsPre-InterventionN (%)Post-InterventionN (%)P valueDischarge encounters reviewed291283--Discharges with lab(s) pending at discharge130 (44.7)115 (40.6)0.38Discharges with eventual abnormal results64 (49.2)42 (36.5)--Discharge summaries listing all pending labs35* (26.9)104 (90.4) <0.01 Discharges with follow-up notes reporting results (all)17 (13.1)15 (13.0)0.99Discharges with follow-up notes reporting results (abnormal results only)13 (20.3)12 (29.2)0.33Bolded data highlights the primary outcome measure. *5 encounters with incomplete lists were not included. Results Bolded data highlights the primary outcome measure. 5 encounters with incomplete lists were not included. Table 2Pending labs with abnormal resultsLab typeMentioned Pre (n/N)Mentioned Post (n/N)% Mentioned Pre% Mentioned PostAbsolute difference (pp)Relative difference (%)Pathology8/179/1247.175.027.959.4Cultures, microbiology--Clinically relevant*1/111/59.120.010.9120.0--Not clinically relevant†0/191/120.08.38.3--Immunology, serology3/51/760.014.3–45.7–76.2Metabolic, vitamins1/120/68.30.0–8.3–100.0*Culture results were considered to be clinically relevant if they were different from the preliminary results on discharge or they could have changed management. †Finalised culture results that were the same as that from preliminary results or would not have change management were considered not clinically relevant. Pending labs with abnormal results Culture results were considered to be clinically relevant if they were different from the preliminary results on discharge or they could have changed management. Finalised culture results that were the same as that from preliminary results or would not have change management were considered not clinically relevant. Discussion Our findings build on work that has highlighted significant gaps in communication of pending studies at discharge and the potential of EHR tools to address these gaps. To our knowledge, no prior published work have paired a pending labs SmartLink with a default DCS template. Notably, our project had a greater increase in documentation with automation, compared with prior interventions that used manual prompts (12% vs 22%),7 DCS enhancement plus email reminders (13% to 46%)3 and manual input of an automated pending labs list (18% to 43%).8 This difference is likely attributed to integrating the SmartLink into the DCS template as a default feature, bypassing manual input requirements that are often limited by human factors such as oversight, recall and time constraints. However, while pending results documentation improved significantly, there was no change in documentation of post-discharge results. This may reflect the complexity of the post-discharge follow-up process. There was a trend towards significance in notes that mentioned abnormal results post-discharge, possibly indicating an effect, since abnormal results are more likely to be reported, but not detected in the small sample. Our project had several limitations. First, it was conducted at a single academic hospital, potentially limiting generalisability. The sample size was small. Only about 20% of discharges were reviewed, which could have impacted detection of less common events, such as follow-up note mentioning abnormal and clinically relevant results. Additionally, we only conducted one Plan-Do-Study-Act cycle, thus limiting assessment of the intervention overtime. Though, we suspect that given automation of the intervention, sustainability would not be impacted. The project’s strengths lie in its novel use of automation and focus on a practical, low-cost, scalable intervention with broad implementation across a large medical centre, involving both residents and hospitalists. We also assessed documentation of post-discharge results, a clinically relevant process measure not previously evaluated in similar studies involving DCS interventions.3 5 7 8 Conclusion Automating the documentation of pending labs into a DCS template can ensure documentation of pending studies on discharge. Future efforts should focus on integrating automated EHR tools with interventions that ensure direct communication with outpatient providers and patients, such as automated discharge summary delivery and review of pending labs at post-discharge visits. Supplementary material