Saverio M. Maviglia

Patient Safety Concerns Arising from Test Results That Return after Hospital Discharge

Context Poor communication between inpatient and outpatient providers precedes many preventable adverse events that occur shortly after discharge. Contribution Forty-one percent of 2644 patients on the hospitalist services of 2 academic hospitals had pending laboratory or radiology results at discharge. Physician-reviewers deemed approximately 9% of these results potentially actionable. Physician surveys done 14 days after results were first available showed that physicians were unaware of many results and thought that about 13% of them required urgent action. Cautions Findings may not apply to nonacademic or nonhospitalist settings. Implications We need good integrated systems to assure follow-up of tests that are pending at discharge. The Editors Good communication between inpatient and outpatient physicians at the transition from hospital to home is critical to patient safety. However, the amount and complexity of information that must be relayed at hospital discharge are often overwhelming. Unfortunately, when communication breaks down, patients are at risk: More than half of all preventable adverse events occurring soon after hospital discharge have been related to poor communication among providers (1). Recently, the challenges to high-quality transitions of care have been increasingly recognized (2), and several factors may be contributing to communication failures at discharge. Although the introduction of hospitalist programs across the United States has produced positive results (3-5), the discontinuity of care inherent in the hospitalist model increases the likelihood of communication failures and makes thorough communication at discharge essential (6). Discontinuity is also an issue in teaching hospitals, where physicians-in-training may be responsible for some or all of the communication at discharge and, under new work-hour restrictions, may frequently change services or work in shifts. Whatever the cause, discontinuity of care at the inpatient-to-outpatient transition has been shown to be associated with medical errors (7). Among these errors is a failure to follow up on the results of laboratory tests and radiologic studies that return after discharge. Although timely follow-up on test results has received attention from the Agency for Healthcare Research and Quality (8) and failure to follow up on results has been recognized by a large malpractice insurer (9) as accounting for one quarter of diagnosis-related malpractice cases, few studies have addressed follow-up on test results pending at hospital discharge. Moore and colleagues (7) studied test follow-up errors, which were defined as having a test result noted as pending at discharge in the inpatient medical record but not acknowledged in the outpatient chart. Using retrospective chart review, they found this type of error in the records of 8% of all discharged patients and 41% of all patients discharged with pending test results, but their study design did not allow them to determine 1) whether clinicians were aware of the results and did not document them or 2) the clinical consequences of these errors. To our knowledge, no other studies have prospectively examined the prevalence and characteristics of test results that return after discharge or physician awareness of them. We hypothesized that test results pending at discharge are frequently overlooked in the handoff from the inpatient physician to the outpatient physician and that some of these results might have important clinical consequences for patients. Accordingly, we sought to prospectively determine the prevalence and characteristics of these potentially actionable results, to determine how often physicians are unaware of these results, and to evaluate the satisfaction of inpatient physicians with current systems for following up on results returning after discharge. Methods We carried out our study on the general medicine hospitalist services at 2 academic tertiary care centers in Boston, Massachusetts (hospitals A and B). The human research committee for both hospitals reviewed and approved the study design. The hospitals belong to the same integrated care-delivery network and share a common electronic clinical data repository that includes test results, discharge orders and summaries, ambulatory notes, and medication and problem lists. These data are accessible at all inpatient and outpatient sites through the same electronic medical record. In addition, all physicians use the same e-mail system. Hospital A has 3 hospitalist inpatient teams that each consist of 1 hospitalist attending physician, 1 internal medicine resident, and 2 interns. At hospital A, the hospitalist attending physician is usually responsible for all communication to outpatient physicians at discharge, as well as for follow-up on all pending test results that return after discharge. Hospital B has 2 types of hospitalist services. One is nonhousestaff and is staffed only by hospitalist and nonhospitalist attending physicians; the nonhospitalist attending physicians care for their own patients on this service, but for the purposes of the study, we categorized them as inpatient physicians. The other hospitalist service at hospital B is a teaching service of 4 teams, each with 1 hospitalist attending physician, 1 junior resident, and 3 interns. On these teams at hospital B, the junior resident is responsible for communication at discharge and follow-up on all pending test results. During the study, 16 hospitalists were responsible for patient discharges at hospital A, 15 hospitalist and 93 nonhospitalist attending physicians were responsible for discharges on the nonhousestaff service at hospital B, and 54 junior residents were responsible for discharges on the teaching service at hospital B. Patient Selection and Identification of Results Returning after Discharge Using the hospital computer systems, we prospectively identified 2644 consecutive patients discharged from February to June 2004. Shortly after each patients discharge, a research assistant entered into a database the patients identifying information, discharge diagnosis, and times and dates of hospital admission and discharge. He or she then tracked each patients pending test results by entering the patient on a watch list using a feature in a results-management system called Results Manager. Results Manager is a computer application that is fully integrated into the electronic medical record and is able to cull pending and final test results from the clinical data repository and to prioritize them on the basis of type of result and degree of abnormality. It was originally developed to track test results in the outpatient setting, and it has been evaluated and tested extensively in that setting but has not been used for inpatients (10). Data Collection We tracked test results with Results Manager for 14 days after patient discharge. A research assistant screened all laboratory and radiologic test results returning after discharge and excluded the results of tests done after discharge. Normal, near-normal, and stable results were excluded by using a predefined algorithm (Figure 1). If a result was abnormal, it was sent to 1 of 4 physician-reviewers who, using the electronic medical record, reviewed the discharge diagnosis; any related test results; and the discharge order, note, or summary (when available) to determine whether the result was potentially actionable. Any result mentioned in the discharge summary was excluded (these were most often final radiologic test results that did not differ from the preliminary results available to the inpatient team). Figure 1. Identifying results for physician review At both hospitals, the discharge order (including discharge diagnoses, medications, and follow-up appointments) was entered into the electronic medical record on the day of discharge and therefore was always available at the time of physician review. Of the 671 results that we reviewed, 525 (78%) were for patients who also had a dictated or typed discharge summary available at the time of review. When discharge summaries are completed after hospital discharge, inpatient physicians have access to the electronic medical record, including any test results that were not available on the day of discharge. The physician-reviewers are board-certified internists; 2 are hospitalists, and 2 are primary care physicians. If a physician-reviewer was involved in the care of a patient who had a result that required review, that result was sent to one of the other 3 reviewers. After reviewing the discharge order, the discharge summary, and related test results, the physician-reviewer used clinical judgment to determine whether the result required clinical action on the basis of the available information. A result was considered potentially actionable if it could change the management of the patient by requiring a new treatment or diagnostic test (or repeated testing), modification or discontinuation of a treatment or diagnostic testing, scheduling of an earlier follow-up appointment, or referral of the patient to another physician or specialist. The reviewer rated the result as definitely actionable, probably actionable, probably not actionable, or definitely not actionable. The reviewer also rated the urgency of the required action according to how soon it should occur: within 1 hour, 8 hours, 24 hours, 72 hours, 1 week, or 1 month. Surveys If the physician-reviewer defined a result as definitely actionable or probably actionable, either the inpatient physician or the primary care physician was surveyed by e-mail to determine whether he or she was aware of the result. At hospital A, the attending hospitalist was the inpatient physician surveyed; on the teaching service at hospital B, the junior resident was surveyed. On the nonhousestaff service at hospital B, the hospitalist or nonhospitalist attending physician was surveyed as the inpatient physician. The

Automating Complex Guidelines for Chronic Disease: Lessons Learned

There is scant published experience with implementing complex, multistep computerized practice guidelines for the long-term management of chronic diseases. We have implemented a system for creating, maintaining, and navigating computer-based clinical algorithms integrated with our electronic medical record. This article describes our progress and reports on lessons learned that might guide future work in this field. We discuss issues and obstacles related to choosing and adapting a guideline for electronic implementation, representing and executing the guideline as a computerized algorithm, and integrating it into the clinical workflow of outpatient care. Although obstacles were encountered at each of these steps, the most difficult were related to workflow integration.

Using an Electronic Medical Record to Identify Opportunities to Improve Compliance with Cholesterol Guidelines

OBJECTIVE: To use an electronic medical record to measure rates of compliance with the National Cholesterol Education Program (NCEP) cholesterol guidelines for secondary prevention, to characterize the patterns of noncompliance, and to identify patient and physician-specific correlates of noncompliance. DESIGN: Cross-sectional descriptive analysis of data extracted from an electronic medical record. SETTING: Nineteen primary care clinics affiliated with a tertiary academic medical center. PATIENTS: All patients who visited their primary care physician in the preceding year who met criteria for secondary prevention of hypercholesterolemia. INTERVENTIONS: None. The main outcome was rate of compliance with NCEP cholesterol guidelines. MAIN RESULTS: Of 2,019 patients who qualified for secondary prevention, only 31% were in compliance with NCEP recommendations, although 44% were on lipid-lowering therapy. There was no low-density lipoprotein cholesterol (LDL-C) on record within the last three years for 771 (38%), and another 809 (40%) had a recent LDL-C that was above the recommended target of 100 mg/dL. Of the latter group, 374 (46%) were not on a statin, including 188 patients with an LDL-C>130 mg/dL. Compliance among secondary prevention patients with cerebrovascular or peripheral vascular disease, but not coronary disease, was even lower: 19% versus 36%, P<.0001. Most of the additional noncompliant patients never had an LDL-C checked. Patient-specific factors associated with compliance included having seen a cardiologist (45% vs 21%); having had a recent admission for myocardial infarction, unstable angina, or angina (41% vs 26%); being male (37% vs 24%); and being white (34% vs 26%). Patients over 79 and under 50 years old also were less likely to be compliant (22% vs 34% for 50–79 year olds). There were no significant differences in compliance rates based on physician-specific factors, such as level of training, gender, or panel size. CONCLUSION: We found poor compliance with nationally published and well-accepted guidelines on diagnosing and treating hypercholesterolemia in secondary prevention patients. Compliance was unrelated to physician or physician-specific characteristics, but it was especially low for women, African Americans, patients without a cardiologist, and patients with cerebrovascular and peripheral vascular disease.

Governance for clinical decision support: case studies and recommended practices from leading institutions

OBJECTIVE Clinical decision support (CDS) is a powerful tool for improving healthcare quality and ensuring patient safety; however, effective implementation of CDS requires effective clinical and technical governance structures. The authors sought to determine the range and variety of these governance structures and identify a set of recommended practices through observational study. DESIGN Three site visits were conducted at institutions across the USA to learn about CDS capabilities and processes from clinical, technical, and organizational perspectives. Based on the results of these visits, written questionnaires were sent to the three institutions visited and two additional sites. Together, these five organizations encompass a variety of academic and community hospitals as well as small and large ambulatory practices. These organizations use both commercially available and internally developed clinical information systems. MEASUREMENTS Characteristics of clinical information systems and CDS systems used at each site as well as governance structures and content management approaches were identified through extensive field interviews and follow-up surveys. RESULTS Six recommended practices were identified in the area of governance, and four were identified in the area of content management. Key similarities and differences between the organizations studied were also highlighted. CONCLUSION Each of the five sites studied contributed to the recommended practices presented in this paper for CDS governance. Since these strategies appear to be useful at a diverse range of institutions, they should be considered by any future implementers of decision support.

Ambulatory hypercholesterolemia management in patients with atherosclerosis: Gender and race differences in processes and outcomes

OBJECTIVE: To determine whether outpatient cholesterol management varies by gender or race among patients with atherosclerosis, and assess factors related to subsequent cholesterol control.DESIGN: Retrospective cohort study.SETTING: Primary care clinics affiliated with an academic medical center.PARTICIPANTS: Two hundred forty-three patients with coronary heart disease, cerebrovascular disease, or peripheral vascular disease and low-density lipoprotein cholesterol (LDL-C)> 130 mg/dl.MEASUREMENTS AND MAIN RESULTS: The primary process of care assessed for 1,082 office visits was cholesterol management (medication intensification or LDL-C monitoring). Cholesterol management occurred at 31.2% of women’s and 38.5% of men’s visits (P=.01), and 37.3% of black and 31.7% of white patients’ visits (P=.09). Independent predictors of cholesterol management included female gender (adjusted risk ratio [ARR], 0.77; 95% confidence interval [CI], 0.60 to 0.97), seeing a primary care clinician other than the patient’s primary care physician (ARR, 0.23; 95% CI, 0.11 to 0.45), and having a new clinical problem addressed (ARR, 0.60; 95% CI, 0.48 to 0.74). After 1 year, LDLC <130 mg/dl occurred less often for women than men (41% vs 61%; P=.003), black than white patients (39% vs 58%; P=.01), and patients with only Medicare insurance than with commercial insurance (37% vs 58%; P=.008). Adjustment for clinical characteristics and management attenuated the relationship between achieving an LDL-C <130 mg/dl and gender.CONCLUSIONS: In this high-risk population with uncontrolled cholesterol, cholesterol management was less intensive for women than men but similar for black and white patients. Less intense cholesterol management accounted for some of the disparity in cholesterol control between women and men but not between black and white patients.

Automated coded ambulatory problem lists: evaluation of a vocabulary and a data entry tool

BACKGROUND Problem lists are fundamental to electronic medical records (EMRs). However, obtaining an appropriate problem list dictionary is difficult, and getting users to code their problems at the time of data entry can be challenging. OBJECTIVE To develop a problem list dictionary and search algorithm for an EMR system and evaluate its use. METHODS We developed a problem list dictionary and lookup tool and implemented it in several EMR systems. A sample of 10,000 problem entries was reviewed from each system to assess overall coding rates. We also performed a manual review of a subset of entries to determine the appropriateness of coded entries, and to assess the reasons other entries were left uncoded. RESULTS The overall coding rate varied significantly between different EMR implementations (63-79%). Coded entries were virtually always appropriate (99%). The most frequent reasons for uncoded entries were due to user interface failures (44-45%), insufficient dictionary coverage (20-32%), and non-problem entries (10-12%). CONCLUSION The problem list dictionary and search algorithm has achieved a good coding rate, but the rate is dependent on the specific user interface implementation. Problem coding is essential for providing clinical decision support, and improving usability should result in better coding rates.

A multi-layered framework for disseminating knowledge for computer-based decision support.

BACKGROUND There are several challenges in encoding guideline knowledge in a form that is portable to different clinical sites, including the heterogeneity of clinical decision support (CDS) tools, of patient data representations, and of workflows. METHODS We have developed a multi-layered knowledge representation framework for structuring guideline recommendations for implementation in a variety of CDS contexts. In this framework, guideline recommendations are increasingly structured through four layers, successively transforming a narrative text recommendation into input for a CDS system. We have used this framework to implement rules for a CDS service based on three guidelines. We also conducted a preliminary evaluation, where we asked CDS experts at four institutions to rate the implementability of six recommendations from the three guidelines. CONCLUSION The experience in using the framework and the preliminary evaluation indicate that this approach has promise in creating structured knowledge, to implement in CDS systems, that is usable across organizations.

Adverse drug events caused by serious medication administration errors

Objective To determine how often serious or life-threatening medication administration errors with the potential to cause harm (potential adverse drug events) result in actual harm (adverse drug events (ADEs)) in the hospital setting. Design Retrospective chart review of clinical events following observed medication administration errors. Background Medication errors are common at the medication administration stage for inpatients. While many errors can cause harm, it is unclear exactly how often. Methods In a previous study where 14 041 medication administrations were directly observed, 1271 medication administration errors were discovered, of which 133 had the potential to cause serious or life-threatening harm and were considered serious or life-threatening potential adverse drug events. As a follow-up, clinical reviewers conducted detailed chart review of serious or life-threatening potential ADEs to determine if they caused an ADE. Reviewers assessed severity of the ADE and attribution to the error. Results Ten (7.5% (95% CI 6.98 to 8.01)) actual ADEs resulted from the 133 serious and life-threatening potential ADEs, of which 6 resulted in significant, three in serious, and one life threatening injury. Therefore 4 (3% (95% CI 2.12 to 3.6)) of serious or life threatening potential ADEs led to serious or life threatening ADEs. Half of the ADEs were caused by dosage or monitoring errors for anti-hypertensives. Conclusions Unintercepted potential ADEs at the medication administration stage can cause serious patient harm. At hospitals where 6 million doses are administered per year, about 4000 preventable ADEs would be attributable to medication administration errors annually.

The morningside initiative: collaborative development of a knowledge repository to accelerate adoption of clinical decision support.

The Morningside Initiative is a public-private activity that has evolved from an August, 2007, meeting at the Morningside Inn, in Frederick, MD, sponsored by the Telemedicine and Advanced Technology Research Center (TATRC) of the US Army Medical Research Materiel Command. Participants were subject matter experts in clinical decision support (CDS) and included representatives from the Department of Defense, Veterans Health Administration, Kaiser Permanente, Partners Healthcare System, Henry Ford Health System, Arizona State University, and the American Medical Informatics Association (AMIA). The Morningside Initiative was convened in response to the AMIA Roadmap for National Action on Clinical Decision Support and on the basis of other considerations and experiences of the participants. Its formation was the unanimous recommendation of participants at the 2007 meeting which called for creating a shared repository of executable knowledge for diverse health care organizations and practices, as well as health care system vendors. The rationale is based on the recognition that sharing of clinical knowledge needed for CDS across organizations is currently virtually non-existent, and that, given the considerable investment needed for creating, maintaining and updating authoritative knowledge, which only larger organizations have been able to undertake, this is an impediment to widespread adoption and use of CDS. The Morningside Initiative intends to develop and refine (1) an organizational framework, (2) a technical approach, and (3) CDS content acquisition and management processes for sharing CDS knowledge content, tools, and experience that will scale with growing numbers of participants and can be expanded in scope of content and capabilities. Intermountain Healthcare joined the initial set of participants shortly after its formation. The efforts of the Morningside Initiative are intended to serve as the basis for a series of next steps in a national agenda for CDS. It is based on the belief that sharing of knowledge can be highly effective as is the case in other competitive domains such as genomics. Participants in the Morningside Initiative believe that a coordinated effort between the private and public sectors is needed to accomplish this goal and that a small number of highly visible and respected health care organizations in the public and private sector can lead by example. Ultimately, a future collaborative knowledge sharing organization must have a sustainable long-term business model for financial support.

Structured representation for core elements of common clinical decision support interventions to facilitate knowledge sharing.

At present, there are no widely accepted, standard approaches for representing computer-based clinical decision support (CDS) intervention types and their structural components. This study aimed to identify key requirements for the representation of five widely utilized CDS intervention types: alerts and reminders, order sets, infobuttons, documentation templates/forms, and relevant data presentation. An XML schema was proposed for representing these interventions and their core structural elements (e.g., general metadata, applicable clinical scenarios, CDS inputs, CDS outputs, and CDS logic) in a shareable manner. The schema was validated by building CDS artifacts for 22 different interventions, targeted toward guidelines and clinical conditions called for in the 2011 Meaningful Use criteria. Custom style sheets were developed to render the XML files in human-readable form. The CDS knowledge artifacts were shared via a public web portal. Our experience also identifies gaps in existing standards and informs future development of standards for CDS knowledge representation and sharing.