Can long-term historical data from electronic medical records improve surveillance for epidemics of acute respiratory infections? A systematic evaluation.

BACKGROUND: As the deployment of electronic medical records (EMR) expands, so is the availability of long-term datasets that could serve to enhance public health surveillance. We hypothesized that EMR-based surveillance systems that incorporate seasonality and other long-term trends would discover outbreaks of acute respiratory infections (ARI) sooner than systems that only consider the recent past. METHODS: We simulated surveillance systems aimed at discovering modeled influenza outbreaks injected into backgrounds of patients with ARI. Backgrounds of daily case counts were either synthesized or obtained by applying one of three previously validated ARI case-detection algorithms to authentic EMR entries. From the time of outbreak injection, detection statistics were applied daily on paired background+injection and background-only time series. The relationship between the detection delay (the time from injection to the first alarm uniquely found in the background+injection data) and the false-alarm rate (FAR) was determined by systematically varying the statistical alarm threshold. We compared this relationship for outbreak detection methods that utilized either 7 days (early aberrancy reporting system (EARS)) or 2-4 years of past data (seasonal autoregressive integrated moving average (SARIMA) time series modeling). RESULTS: In otherwise identical surveillance systems, SARIMA detected epidemics sooner than EARS at any FAR below 10%. The algorithms used to detect single ARI cases impacted both the feasibility and marginal benefits of SARIMA modeling. Under plausible real-world conditions, SARIMA could reduce detection delay by 5-16 days. It also was more sensitive at detecting the summer wave of the 2009 influenza pandemic. CONCLUSION: Time series modeling of long-term historical EMR data can reduce the time it takes to discover epidemics of ARI. Realistic surveillance simulations may prove invaluable to optimize system design and tuning.

Decision support during electronic prescription to stem antibiotic overuse for acute respiratory infections: a long-term, quasi-experimental study.

BACKGROUND: Interventions to support decision-making can reduce inappropriate antibiotic use for acute respiratory infections (ARI), but they may not be sustainable. The objective of the study is to evaluate the long-term effectiveness of a clinical decision-support system (CDSS) interposed at the time of electronic (e-) prescriptions for selected antibiotics. METHODS: This is a retrospective, observational intervention study, conducted within a large, statewide Veterans Affairs health system. Participants are outpatients with an initial visit for ARI. A CDSS was deployed upon e-prescription of selected antibiotics during the study period. From 01/2004 to 05/2006 (pre-withdrawal period), the CDSS targeted azithromycin and the fluoroquinolone gatifloxacin. From 05/2006 to 12/2011 (post-withdrawal period), the CDSS was retained for azithromycin but withdrawn for the fluoroquinolone. A manual record review was conducted to determine concordance of antibiotic prescription with ARI treatment guidelines. RESULTS: Of 1131 included ARI visits, 380 (33.6%) were guideline-concordant. For azithromycin, concordance did not change between the pre- and post-withdrawal periods, and adjusted odds of concordance was 8.8 for the full study period, compared to unrestricted antibiotics. For fluoroquinolones, guideline concordance decreased from 88.6% (39 of 44 visits) to 51.3% (59 of 115 visits), pre- vs. post-withdrawal periods (p < 0.005). The adjusted odds of concordance compared to "All Other Antibiotics" visits decreased from 24.4 (95% CI 9.0-66.3) pre-withdrawal to 5.5 (95% CI 3.5-8.8) post-withdrawal (p = .008). Concordance did not change between those same time periods for antibiotics that were never subjected to the intervention ("All Other Antibiotics"). CONCLUSIONS: A CDSS interposed at the time of e-prescription of selected antibiotics can shift their use toward ARI treatment guidelines, and this effect can be maintained over the long term as long as the CDSS remains in place. Removal of the CDSS after 3.5 years of implementation resulted in a rise in guideline-discordant antibiotic use.

Restrictive versus liberal red blood cell transfusion strategy after hip surgery: a decision model analysis of healthcare costs.

BACKGROUND: Red blood cell transfusion related to select surgical procedures accounts for approximately 2.8 million transfusions in the United States yearly and occurs commonly after hip fracture surgeries. Randomized controlled trials have demonstrated lack of clinical benefit with higher versus lower transfusion thresholds in postoperative hip fracture repair patients with cardiac disease or risk factors for cardiac disease. The economic implications of a higher versus lower hemoglobin (Hb) threshold have not yet been investigated. STUDY DESIGN AND METHODS: A decision tree analysis was constructed to estimate differences in healthcare costs and charges between a Hb transfusion threshold strategy of 8 g/dL versus 10 g/dL from the perspective of both Centers for Medicare and Medicaid Services (CMS) as well as hospitals. Secondary outcome measures included differences in transfusion-related adverse events. RESULTS: Among the 133,697 Medicare beneficiaries undergoing hip fracture repair in 2012, we estimated that 45,457 patients would be anemic and at risk for transfusion. CMS would save an estimated $11.3 million to $24.3 million in payments, while hospitals would reduce charges by an estimated $52.7 million to $93.6 million if the restrictive transfusion strategy were to be implemented nationally. Additionally, rates of transfusion-associated circulatory overload, transfusion-related acute lung injury, acute transfusion reactions, length of stay, and mortality would be reduced. CONCLUSIONS: This model suggests that the uniform adoption of a restrictive transfusion strategy among patients with cardiac disease and risk factors for cardiac disease undergoing hip fracture repair would result in significant reductions in clinically important outcomes with significant cost savings.

Epidemic surveillance using an electronic medical record: an empiric approach to performance improvement.

BACKGROUNDS: Electronic medical records (EMR) form a rich repository of information that could benefit public health. We asked how structured and free-text narrative EMR data should be combined to improve epidemic surveillance for acute respiratory infections (ARI). METHODS: Eight previously characterized ARI case detection algorithms (CDA) were applied to historical EMR entries to create authentic time series of daily ARI case counts (background). An epidemic model simulated influenza cases (injection). From the time of the injection, cluster-detection statistics were applied daily on paired background+injection (combined) and background-only time series. This cycle was then repeated with the injection shifted to each week of the evaluation year. We computed: a) the time from injection to the first statistical alarm uniquely found in the combined dataset (Detection Delay); b) how often alarms originated in the background-only dataset (false-alarm rate, or FAR); and c) the number of cases found within these false alarms (Caseload). For each CDA, we plotted the Detection Delay as a function of FAR or Caseload, over a broad range of alarm thresholds. RESULTS: CDAs that combined text analyses seeking ARI symptoms in clinical notes with provider-assigned diagnostic codes in order to maximize the precision rather than the sensitivity of case-detection lowered Detection Delay at any given FAR or Caseload. CONCLUSION: An empiric approach can guide the integration of EMR data into case-detection methods that improve both the timeliness and efficiency of epidemic detection.

Using the electronic medical record to identify community-acquired pneumonia: toward a replicable automated strategy.

BACKGROUND: Timely information about disease severity can be central to the detection and management of outbreaks of acute respiratory infections (ARI), including influenza. We asked if two resources: 1) free text, and 2) structured data from an electronic medical record (EMR) could complement each other to identify patients with pneumonia, an ARI severity landmark. METHODS: A manual EMR review of 2747 outpatient ARI visits with associated chest imaging identified x-ray reports that could support the diagnosis of pneumonia (kappa score  = 0.88 (95% CI 0.82∶0.93)), along with attendant cases with Possible Pneumonia (adds either cough, sputum, fever/chills/night sweats, dyspnea or pleuritic chest pain) or with Pneumonia-in-Plan (adds pneumonia stated as a likely diagnosis by the provider). The x-ray reports served as a reference to develop a text classifier using machine-learning software that did not require custom coding. To identify pneumonia cases, the classifier was combined with EMR-based structured data and with text analyses aimed at ARI symptoms in clinical notes. RESULTS: 370 reference cases with Possible Pneumonia and 250 with Pneumonia-in-Plan were identified. The x-ray report text classifier increased the positive predictive value of otherwise identical EMR-based case-detection algorithms by 20-70%, while retaining sensitivities of 58-75%. These performance gains were independent of the case definitions and of whether patients were admitted to the hospital or sent home. Text analyses seeking ARI symptoms in clinical notes did not add further value. CONCLUSION: Specialized software development is not required for automated text analyses to help identify pneumonia patients. These results begin to map an efficient, replicable strategy through which EMR data can be used to stratify ARI severity.

A sustainable strategy to prevent misuse of antibiotics for acute respiratory infections.

BACKGROUNDS: Over 50% of antibiotics prescriptions are for outpatients with acute respiratory infections (ARI). Many of them are not needed and thus contribute both avoidable adverse events and pressures toward the development of bacterial resistance. Could a clinical decision support system (CDSS), interposed at the time of electronic prescription, adjust antibiotics utilization toward consensus treatment guidelines for ARI? METHODS: This is a retrospective comparison of pre- (2002) and post-intervention (2003-2006) periods at two comprehensive health care systems (intervention and control). The intervention was a CDSS that targeted fluoroquinolone and azithromycin; other antibiotics remained unrestricted. 7000 outpatients visits flagged by an ARI case-finding algorithm were reviewed for congruence with the guidelines (antibiotic prescribed-when-warranted or not-prescribed-when-unwarranted). RESULTS: 3831 patients satisfied the case definitions for one or more ARI: pneumonia (537), bronchitis (2931), sinusitis (717) and non-specific ARI (145). All patients with pneumonia received antibiotics. The relative risk (RR) of congruent prescribing was 2.57 (95% CI = (1.865 to 3.540) in favor of the intervention site for the antibiotics targeted by the CDSS; congruence did not change for other antibiotics (adjusted RR = 1.18 (95% CI = (0.691 to 2.011)). The proportion of unwarranted prescriptions of the targeted antibiotics decreased from 22% to 3%, pre vs. post-intervention (p<0.0001). CONCLUSIONS: A CDSS interposed at the time of e-prescription nearly extinguished unwarranted use targeted antibiotics for ARI for four years. This intervention highlights a path toward sustainable antibiotics stewardship for outpatients with ARI.

Combining free text and structured electronic medical record entries to detect acute respiratory infections.

BACKGROUND: The electronic medical record (EMR) contains a rich source of information that could be harnessed for epidemic surveillance. We asked if structured EMR data could be coupled with computerized processing of free-text clinical entries to enhance detection of acute respiratory infections (ARI). METHODOLOGY: A manual review of EMR records related to 15,377 outpatient visits uncovered 280 reference cases of ARI. We used logistic regression with backward elimination to determine which among candidate structured EMR parameters (diagnostic codes, vital signs and orders for tests, imaging and medications) contributed to the detection of those reference cases. We also developed a computerized free-text search to identify clinical notes documenting at least two non-negated ARI symptoms. We then used heuristics to build case-detection algorithms that best combined the retained structured EMR parameters with the results of the text analysis. PRINCIPAL FINDINGS: An adjusted grouping of diagnostic codes identified reference ARI patients with a sensitivity of 79%, a specificity of 96% and a positive predictive value (PPV) of 32%. Of the 21 additional structured clinical parameters considered, two contributed significantly to ARI detection: new prescriptions for cough remedies and elevations in body temperature to at least 38°C. Together with the diagnostic codes, these parameters increased detection sensitivity to 87%, but specificity and PPV declined to 95% and 25%, respectively. Adding text analysis increased sensitivity to 99%, but PPV dropped further to 14%. Algorithms that required satisfying both a query of structured EMR parameters as well as text analysis disclosed PPVs of 52-68% and retained sensitivities of 69-73%. CONCLUSION: Structured EMR parameters and free-text analyses can be combined into algorithms that can detect ARI cases with new levels of sensitivity or precision. These results highlight potential paths by which repurposed EMR information could facilitate the discovery of epidemics before they cause mass casualties.

Racial/ethnic and age disparities in chemotherapy selection for colorectal cancer.

OBJECTIVES: To test the hypotheses that African American patients and older patients with stage IV colorectal cancer were less likely to receive newer chemotherapy agents. STUDY DESIGN: Retrospective cohort design. METHODS: Among 5068 Surveillance, Epidemiology, and End Results-Medicare patients diagnosed as having stage IV colorectal cancer between 2000 and 2002, a total of 2466 received chemotherapy and were included in the analysis. Irinotecan hydrochloride was the first of the "newer" chemotherapy agents and was marketed in 2000 as a first-line add-on agent. Descriptive statistics were generated, and a multivariable logistic regression was run to estimate the odds of receiving irinotecan among African American patients and older patients and within 2 months of chemotherapy initiation. RESULTS: African American patients had lower odds of initiating treatment with a newer chemotherapy than white patients (adjusted odds ratio, 0.641; 95% confidence interval, 0.453-0.907). An age disparity was also found, with all older age groups being significantly less likely to initiate treatment with a newer chemotherapy than the youngest age group: the adjusted odds of receiving newer chemotherapy agents (relative to patients aged 66-70 years) were lower and significant among patients aged 71 to 75, 76 to 80, and older than 80 years (odds ratios, 0.708, 0.527, and 0.213, respectively). CONCLUSIONS: Disparities in chemotherapy selection exist among patients receiving chemotherapy for stage IV colorectal cancer. On initiating chemotherapy, African American patients and older patients were less likely to receive a newer agent.

Analysis of False Positive Errors of an Acute Respiratory Infection Text Classifier due to Contextual Features.

Text classifiers have been used for biosurveillance tasks to identify patients with diseases or conditions of interest. When compared to a clinical reference standard of 280 cases of Acute Respiratory Infection (ARI), a text classifier consisting of simple rules and NegEx plus string matching for specific concepts of interest produced 569 (4%) false positive (FP) cases. Using instance level manual annotation we estimate the prevalence of contextual attributes and error types leading to FP cases. Errors were due to (1) Deletion errors from abbreviations, spelling mistakes and missing synonyms (57%); (2) Insertion errors from templated document structures such as check boxes, and lists of signs and symptoms (36%) and; (3) Substitution errors from irrelevant concepts and alternate meanings for the same word (6%). We demonstrate that specific concept attributes contribute to false positive cases. These results will inform modifications and adaptations to improve text classifier performance.

Outcomes of irinotecan-based chemotherapy regimens in elderly Medicare patients with metastatic colorectal cancer.

BACKGROUND: Several population-based studies have confirmed the benefits of adjuvant chemotherapy with 5-fluorouracil/leucovorin for treatment of colorectal cancer. Few population-based studies have evaluated other chemotherapies that are now available for colorectal cancer management. OBJECTIVE: This study primarily sought to evaluate the survival benefit of first-line irinotecan use in a group of Medicare patients with stage IV (metastatic) colorectal cancer. METHODS: Data on chemotherapy users with a diagnosis of colorectal cancer reported between 1998 and 2002 were obtained from the Surveillance Epidemiology and End Results (SEER)-Medicare database. Irinotecan, marketed in 1997, was one of the newer chemotherapy agents in the available data. Chemotherapy episodes, defined as periods of continuous chemotherapy treatment with no gaps >90 days between successive claims, were identified. The first chemotherapy episode after diagnosis was used to identify lines of treatment: patients may have initiated irinotecan therapy within 2 months (first-line), used irinotecan later in the first episode (second-line), or not used irinotecan at all. Descriptive statistics were generated and a multivariable Cox proportional hazards model was used to determine the survival benefit of irinotecan. Secondary analyses explored the survival benefit in specific patient subgroups. The impact of irinotecan use on health care utilization also was assessed. RESULTS: Of 3327 chemotherapy users (mean/median age, 75 years), 842 (25.3%) initiated chemotherapy using irinotecan. No overall survival benefit for irinotecan was observed in the primary analysis comparing irinotecan initiators with all other chemotherapy users (including those who used irinotecan subsequently). Covariates that were negatively associated with survival included older age, presence of >1 comorbidity, a high tumor grade, lymph node involvement, and a primary tumor site in the colon. Surgery was positively associated with a lower hazard of death. In subgroup analyses that excluded subsequent irinotecan users, a survival benefit for irinotecan was observed but diminished over time. Irinotecan users had higher rates of hospitalizations possibly due to chemotherapy-related adverse effects. This retrospective claims study had limitations such as a lack of information on patient performance status, dosing, and the types of regimens used; hence, certain assumptions had to be made and selection bias may have been involved. CONCLUSIONS: The definitive survival advantage of irinotecan observed in clinical trials was not reproducible in this population of elderly Medicare patients. The results emphasize the need for expansion of trials to include a more diverse patient group as well as continued evaluation of more recent chemotherapies in real-world settings.

Optimizing A syndromic surveillance text classifier for influenza-like illness: Does document source matter?

Syndromic surveillance systems that incorporate electronic free-text data have primarily focused on extracting concepts of interest from chief complaint text, emergency department visit notes, and nurse triage notes. Due to availability and access, there has been limited work in the area of surveilling the full text of all electronic note documents compared with more specific document sources. This study provides an evaluation of the performance of a text classifier for detection of influenza-like illness (ILI) by document sources that are commonly used for biosurveillance by comparing them to routine visit notes, and a full electronic note corpus approach. Evaluating the performance of an automated text classifier for syndromic surveillance by source document will inform decisions regarding electronic textual data sources for potential use by automated biosurveillance systems. Even when a full electronic medical record is available, commonly available surveillance source documents provide acceptable statistical performance for automated ILI surveillance.

Using cost-effectiveness analysis to sharpen formulary decision-making: the example of tiotropium at the Veterans Affairs health care system.

OBJECTIVE: To identify a cost-saving subset of criteria for the use of tiotropium at a Veterans Affairs Medical Center based on a cost-effectiveness analysis with ipratropium as the comparator. METHODS: Retrospective analysis of electronic medical records for the calendar year 2004 was conducted. The sample was drawn from a population at the Baltimore Veterans Affairs Medical Center that had a confirmed diagnosis of chronic obstructive pulmonary disease (COPD) and had filled prescriptions for ipratropium. The tiotropium sample was based on a modeled cohort of COPD patients who had received tiotropium. The analysis was conducted from the perspective of the Veterans Affairs Health Care System. The outcome was the incremental cost-effectiveness of tiotropium versus ipratropium. RESULTS: The incremental cost-effectiveness ratio (ICER) was $2360 per avoided exacerbation. Tiotropium cost-effectiveness increased with COPD severity and was cost-saving in patients with very severe disease (ICER = $-1818) and in patients with a previous COPD-related hospitalization (ICER = $-4472). The ICER was most sensitive to the relative effectiveness and price of tiotropium. Results identified the levels of treatment effectiveness and price beyond which tiotropium would become cost-saving relative to ipratropium. CONCLUSIONS: The results support the existing Veterans Affairs practice of offering tiotropium to patients with COPD-related hospitalizations. Periodic review of the effectiveness data to determine whether tiotropium would be cost-saving in patients with very severe COPD is suggested. Cost-effectiveness analyses that identify practical criteria-for-use should become an integral part of the formulary process.

Application of Natural Language Processing to VA Electronic Health Records to Identify Phenotypic Characteristics for Clinical and Research Purposes.

Informatics tools to extract and analyze clinical information on patients have lagged behind data-mining developments in bioinformatics. While the analyses of an individual's partial or complete genotype is nearly a reality, the phenotypic characteristics that accompany the genotype are not well known and largely inaccessible in free-text patient health records. As the adoption of electronic medical records increases, there exists an urgent need to extract pertinent phenotypic information and make that available to clinicians and researchers. This usually requires the data to be in a structured format that is both searchable and amenable to computation. Using inflammatory bowel disease as an example, this study demonstrates the utility of a natural language processing system (MedLEE) in mining clinical notes in the paperless VA Health Care System. This adaptation of MedLEE is useful for identifying patients with specific clinical conditions, those at risk for or those with symptoms suggestive of those conditions.

Adaptation of the NegEx algorithm to Veterans Affairs electronic text notes for detection of influenza-like illness (ILI).

We tested automated detection of influenza-like illness (ILI) from free-text clinician notes extracted from the VA electronic medical record using a simple negation algorithm coupled with string matching for relevant ILI symptoms mapped to UMLS concepts. Additionally, we documented negation problems encountered and adaptations made to the negation algorithm when applied to the full texts of a diverse array of VA note types and templates.

Vancomycin-resistant enterococci: a road map on how to prevent the emergence and transmission of antimicrobial resistance.

Nosocomial acquisition of microorganisms resistant to multiple antibiotics represents a threat to patient safety. Here we review the mechanisms that have allowed highly resistant strains belonging to the Enterococcus genus to proliferate within our health-care institutions. These mechanisms indicate that decreasing the prevalence of resistant organisms requires active surveillance, adherence to vigorous isolation, hand hygiene and environmental decontamination measures, and effective antibiotic stewardship. We suggest how to tailor such a complex, multidisciplinary program to the needs of a particular health-care setting so as to maximize cost-effectiveness.

Prescription of amiodarone through a computerized template that includes both decision support and executive functions improves the monitoring for toxicities.

The antiarrhythmic drug amiodarone causes severe side effects that warrant active monitoring. In a given patient, monitoring for amiodarone toxicities involves checking past results of five laboratory/imaging tests and reordering four of these tests every 6 to 12 month, a process that is both complex and time-consuming. We hypothesized that a software tool that automated the retrieval and ordering of these tests in a context-sensitive manner would improve the completeness of surveillance for drug toxicities.