Faculty Bibliography

BACKGROUND: Most examples of successful medication reconciliation (MR) programs have reported on paper-based systems, the most common of which is a standardized MR form that often serves as a medication order form. An interdisciplinary process was undertaken by Bellevue Hospital, New York City, to develop a full, online MR program. PHASE 1. MOVING BEYOND PAPER: In 2005 Bellevue piloted a paper-based MR process. However, this effort was unsuccessful, so an online MR application that would be more accessible and easier to audit was initiated. The longitudinal outpatient medication list--the definitive, electronic medication list for patients in our system--formed the basis of the MR project. The list included every prescription written in the electronic health record (EHR). Historical medication could also be entered into the list, representing a useful function in the outpatient setting for patients who transfer their care to Bellevue and are already on chronic medications. In a two-month pilot in Summer 2006, compliance was achieved for only 20% of patients. PHASE 2. AUDITING AND MANDATORY FUNCTIONALITY: In April 2007, MR was made a mandatory part of the admission process; a blocking function in the EHR prevented medication orders if the admission MR had not been completed. Compliance rates subsequently increased to 90% throughout the hospital. To 'close the loop' in the reconciliation process, in November 2007, a discharge reconciliation was made a mandatory part of the discharge process, resulting in 95% compliance. LESSONS LEARNED: Successful implementation of admission and discharge MR suggested several lessons, including (1) mandatory functionality leads to adaptation and integration of MR into housestaff work flows and (2) an electronic MR is preferable to a paper-based process in organizations with an EHR and computerized physician order entry

BACKGROUND: As medical care moves towards an outpatient focus, monitoring systems for ambulatory patients are increasingly important. Because adverse outcomes due to medications are an important problem in outpatients, the authors developed an automated monitoring system for detecting adverse drug reactions (ADRs) in ambulatory patients. METHODS: The authors obtained a set of approximately 110,000 ambulatory care notes from the medicine clinic at Bellevue Hospital Centre for 2003-4, and manually analysed a representative sample of 1250 notes to obtain a gold standard. To detect ADRs in the text of electronic ambulatory notes, the authors used a 'trigger phrases' methodology, based on a simple grammar populated with a limited set of keywords. RESULTS: Under current functionality, this system detected 38 of 54 cases in the authors' gold standard set, of which 17 were true positives, for a sensitivity of 31%, a specificity of 98%, and a positive predictive value of 45%. Their proxy measure correlated with 70% of the ADRs in the gold standard. These values are comparable or superior to other systems described in the literature. CONCLUSIONS: These results show that an automated system can detect ADRs with moderate sensitivity and high specificity, and has the potential to serve as the basis for a larger scale reporting system

BACKGROUND: Experience with a quality improvement (QI) program undertaken to increase the use of beta-adrenergic blockade in at-risk patients at both a major academic medical center and a community hospital suggests barriers to implementation. METHODS: A retrospective and prospective cohort study was performed to establish the incidence and effectiveness of beta-blockade use pre- and postimplementation of a standardized screening tool and a major education program as part of a QI project. Data gathering involved a baseline phase pre-intervention; 6 weeks postintervention; and 3-6 months postintervention. RESULTS: During phase I (baseline) 56% of eligible received beta-blockers, but targeted measures (a pre-induction heart rate < 70 or a systolic blood pressure [BP] < 110 mmHg) were achieved in only 11% of patients. Phase II saw a significant overall increase in beta-blocker administration (79%) and efficacy (50%). However, during phase III (3-6 months postimplementation), the rate of beta-blocker administration fell to 61% overall, while overall efficacy remained stable at 52%. Significant differences between the academic and community hospitals were observed throughout the study. CONCLUSION: Implementation of a quality program for beta-blockade is significantly affected by the presence or absence of ongoing physician and staff education beyond the study period

To measure user expectations and knowledge of the issues involved in the transition from a paper-based system to an EHR, we created an on-line questionnaire, and invited physicians from several specialties at Beth Israel Medical Center (BIMC) to respond to it. As expected, most participants had positive expectations for the EHR and its features, but respondents were more skeptical of electronic clinical decision support systems (CDSS) than we had expected. The responses of this 'EHR-naive' group of physicians underscore the importance of managing expectations with the implementation of the EHR and of the delicate balance involved in preserving physician independence when proposing a clinical decision support system

Medical diagnostic decision support systems are some of the most visible applications of medical informatics. Many of these systems, however, were created in the pre-WWW era, when access to clinical knowledge was much more limited than it is currently. Using the Google(tm) Java API1, we created a simple program to extract potential diagnoses from a general web search. Our system performed at a level comparable to other medical expert systems

The era of applied genomic medicine is quickly approaching accompanied by the increasing availability of detailed genetic information. Understanding the genetic etiology behind complex, multi-gene diseases remains an important challenge. In order to uncover the putative genetic etiology of complex diseases, we designed a method that explores the relationships between two major terminological and ontological resources: the Unified Medical Language System (UMLS) and the Gene Ontology (GO). The UMLS has a mainly clinical emphasis; Gene Ontology has become the standard for biological annotations of genes and gene products. Using statistical and semantic relationships within and between the two resources, we are able to infer relationships between disease concepts in the UMLS and gene products annotated using GO and its associated databases. We validated our inferences by comparing them to the known gene-disease relationships, as defined in the Online Mendelian Inheritance in Man's morbidmap (OMIM). The proof-of-concept methods presented here are unique in that they bypass the ambiguity of the direct extraction of gene or disease term from MEDLINE. Additionally, our methods provide direct links to clinically significant diseases through established terminologies or ontologies. The preliminary results presented here indicate the potential utility of exploiting the existing, manually curated relationships in biomedical resources as a tool for the discovery of potentially valuable new gene-disease relationships

As cross-disciplinary research escalates, researchers are facing the challenge of linking disparate biomedical databases that have been developed without common indexes. Manually indexing these large-scale databases is laborious and often impractical. Solutions involving mediating terminologies have been proposed, but coordination of terms from the databases of interest to these mediating terminologies is also laborious, and regular synchronization between indexes is an additional problem. In this study we describe a novel method of linking heterogeneous databases using terminology networks constructed with automated mapping methods. Linkage was established between two disparate biomedical databases (SNOMED-CT and HDG), using two relevant intermediating databases (UMLS and OMIM). One gold standard of 514 distinct matches is used as proof-of-principle. In conclusion, as hypothesized, 1) Manually curated pathways provide high precision, but offer low recall, 2) the automated terminology pathways can significantly increase recall at acceptable precision. Taken together, our conclusion may suggest the combined manual and automated terminology networks could offer recall and precision in an incremental manner

Understanding clinical phenotypes through their corresponding genotypes is one of the principal goals of genetic research. Though achieving this goal is relatively simple with single gene syndromes, more complex diseases often consist of varied clinical phenotypes that may be the result of interactions among multiple genetic loci. Microarray technology has brought the phenotype -genotype relationship to the molecular level, using differently behaving cancers, for example, as the basis for comparing patterns of gene expression. With this feasibility study, we attempted to use similar methods of analysis at the clinical level, in order to evaluate our hypothesis that the clustering of clinical phenotypes would provide information that would be useful in elucidating their underlying genotypes. Because of its breadth of content and detailed descriptions, we used OMIM as our source material for phenotypic and genetic information. After processing the source material, we then performed self-organizing map and hierarchical clustering analysis on representative diseases by phenotypic category. Through pre-determined queries over this analysis, we made two findings of potential clinical significance, one concerning diabetes and another concerning progressive neurologic diseases. Our methods provide a formal approach to analyzing phenotypes among diverse diseases, and may help indicate fruitful areas for further research into their underlying genetic causes

We developed the Guideline Execution by Semantic Decomposition of Representation (GESDOR) model to share guidelines encoded in different formats at the execution level. For this purpose, we extracted a set of generalized guideline execution tasks from the existing guideline representation models. We then created the mappings between specific guideline representation models and the set of the common guideline execution tasks. Finally, we developed a generic task-scheduling model to harmonize the existing approaches to guideline task scheduling. The evaluation has shown that the GESDOR model can be used for the effective execution of guidelines encoded in different formats, and thus realizes guideline sharing at the execution level

Integration of disparate biomedical terminologies is becoming increasingly important as links between biological science and clinical medicine grow. Mapping concepts in the Gene Ontology (GO) to the UMLS may help further this integration and allow for more efficient information exchange among researchers. Using a gold standard of GO term--UMLS concept mappings provided by the NCI, we examined the performance of various published and combined mapping techniques, in order to maximize precision and recall. We found that for the previously published techniques precision varied between (0.61-0.95), and recall varied from (0.65-0.90), whereas for the hybrid techniques, precision varied between (0.66-0.97), and recall from (0.59-0.93). Our study reveals the benefits of using mapping techniques that incorporate domain knowledge, and provides a basis for future approaches to mapping between distinct biomedical vocabularies