Defining a Patient Population With Cirrhosis: An Automated Algorithm With Natural Language Processing.

OBJECTIVES: The objective of this study was to use natural language processing (NLP) as a supplement to International Classification of Diseases, Ninth Revision (ICD-9) and laboratory values in an automated algorithm to better define and risk-stratify patients with cirrhosis. BACKGROUND: Identification of patients with cirrhosis by manual data collection is time-intensive and laborious, whereas using ICD-9 codes can be inaccurate. NLP, a novel computerized approach to analyzing electronic free text, has been used to automatically identify patient cohorts with gastrointestinal pathologies such as inflammatory bowel disease. This methodology has not yet been used in cirrhosis. STUDY DESIGN: This retrospective cohort study was conducted at the University of California, Los Angeles Health, an academic medical center. A total of 5343 University of California, Los Angeles primary care patients with ICD-9 codes for chronic liver disease were identified during March 2013 to January 2015. An algorithm incorporating NLP of radiology reports, ICD-9 codes, and laboratory data determined whether these patients had cirrhosis. Of the 5343 patients, 168 patient charts were manually reviewed at random as a gold standard comparison. Positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity of the algorithm and each of its steps were calculated. RESULTS: The algorithm's PPV, NPV, sensitivity, and specificity were 91.78%, 96.84%, 95.71%, and 93.88%, respectively. The NLP portion was the most important component of the algorithm with PPV, NPV, sensitivity, and specificity of 98.44%, 93.27%, 90.00%, and 98.98%, respectively. CONCLUSIONS: NLP is a powerful tool that can be combined with administrative and laboratory data to identify patients with cirrhosis within a population.

Temporal trends in rates of patient harm resulting from medical care.

BACKGROUND: In the 10 years since publication of the Institute of Medicine's report To Err Is Human, extensive efforts have been undertaken to improve patient safety. The success of these efforts remains unclear. METHODS: We conducted a retrospective study of a stratified random sample of 10 hospitals in North Carolina. A total of 100 admissions per quarter from January 2002 through December 2007 were reviewed in random order by teams of nurse reviewers both within the hospitals (internal reviewers) and outside the hospitals (external reviewers) with the use of the Institute for Healthcare Improvement's Global Trigger Tool for Measuring Adverse Events. Suspected harms that were identified on initial review were evaluated by two independent physician reviewers. We evaluated changes in the rates of harm, using a random-effects Poisson regression model with adjustment for hospital-level clustering, demographic characteristics of patients, hospital service, and high-risk conditions. RESULTS: Among 2341 admissions, internal reviewers identified 588 harms (25.1 harms per 100 admissions; 95% confidence interval [CI], 23.1 to 27.2) [corrected]. Multivariate analyses of harms identified by internal reviewers showed no significant changes in the overall rate of harms per 1000 patient-days (reduction factor, 0.99 per year; 95% CI, 0.94 to 1.04; P=0.61) or the rate of preventable harms. There was a reduction in preventable harms identified by external reviewers that did not reach statistical significance (reduction factor, 0.92; 95% CI, 0.85 to 1.00; P=0.06), with no significant change in the overall rate of harms (reduction factor, 0.98; 95% CI, 0.93 to 1.04; P=0.47). CONCLUSIONS: In a study of 10 North Carolina hospitals, we found that harms remain common, with little evidence of widespread improvement. Further efforts are needed to translate effective safety interventions into routine practice and to monitor health care safety over time. (Funded by the Rx Foundation.).

Robust climate policies under uncertainty: a comparison of robust decision making and info-gap methods.

This study compares two widely used approaches for robustness analysis of decision problems: the info-gap method originally developed by Ben-Haim and the robust decision making (RDM) approach originally developed by Lempert, Popper, and Bankes. The study uses each approach to evaluate alternative paths for climate-altering greenhouse gas emissions given the potential for nonlinear threshold responses in the climate system, significant uncertainty about such a threshold response and a variety of other key parameters, as well as the ability to learn about any threshold responses over time. Info-gap and RDM share many similarities. Both represent uncertainty as sets of multiple plausible futures, and both seek to identify robust strategies whose performance is insensitive to uncertainties. Yet they also exhibit important differences, as they arrange their analyses in different orders, treat losses and gains in different ways, and take different approaches to imprecise probabilistic information. The study finds that the two approaches reach similar but not identical policy recommendations and that their differing attributes raise important questions about their appropriate roles in decision support applications. The comparison not only improves understanding of these specific methods, it also suggests some broader insights into robustness approaches and a framework for comparing them.

Eliminating waste in US health care.

The need is urgent to bring US health care costs into a sustainable range for both public and private payers. Commonly, programs to contain costs use cuts, such as reductions in payment levels, benefit structures, and eligibility. A less harmful strategy would reduce waste, not value-added care. The opportunity is immense. In just 6 categories of waste--overtreatment, failures of care coordination, failures in execution of care processes, administrative complexity, pricing failures, and fraud and abuse--the sum of the lowest available estimates exceeds 20% of total health care expenditures. The actual total may be far greater. The savings potentially achievable from systematic, comprehensive, and cooperative pursuit of even a fractional reduction in waste are far higher than from more direct and blunter cuts in care and coverage. The potential economic dislocations, however, are severe and require mitigation through careful transition strategies.

Miles to go: an introduction to the 5 Million Lives Campaign.

BACKGROUND: The Institute for Healthcare Improvement (IHI)'s 5 Million Lives Campaign targets a reduction of five million instances of harm from December 2006 through December 2008. The campaign continues the six interventions of the 100,000 Lives Campaign and adds six more. DEFINITION OF MEDICAL HARM AND SETTING THE GOAL: The campaign's aim is to support the reduction of medical harm, so defined: "Unintended physical injury resulting from or contributed to by medical care (including the absence of indicated medical treatment), that requires additional monitoring, treatment, or hospitalization, or that results in death." The goal of a reduction of five million incidents of harm in two years is based on an estimate that 40 to 50 incidents occur per 100 admissions, for a total of 15 million incidents of medical harm each year in the United States. THE 5 MILLION LIVES CAMPAIGN'S "PLATFORM": This campaign's six new interventions address the prevention of pressure ulcers, reduction of methicillin-resistant Staphylococcus aureus (MRSA) infection, prevention of harm from high-alert medications, reduction of surgical complications, delivery of reliable and evidence-based care for congestive heart failure, and getting hospitals' boards of directors on board. CONCLUSION: Together with complementary partner initiatives, the 5 Million Lives Campaign is intended to act as a major driver of national improvement.

Development and Validation of an Automated Method to Identify Patients Undergoing Radical Cystectomy for Bladder Cancer Using Natural Language Processing.

INTRODUCTION: Measurement for quality improvement relies on accurate case identification and characterization. With electronic health records now widely deployed, natural language processing, the use of software to transform text into structured data, may enrich quality measurement. Accordingly we evaluated the application of natural language processing to radical cystectomy procedures for patients with bladder cancer. METHODS: From a sample of 497 procedures performed from March 2013 to October 2014 we identified radical cystectomy for primary bladder cancer using the approaches of 1) a natural language processing enhanced algorithm, 2) an administrative claims based algorithm and 3) manual chart review. We also characterized treatment with robotic surgery and continent urinary diversion. Using chart review as the reference standard we calculated the observed agreement (kappa statistic), sensitivity, specificity, positive predictive value and negative predictive value for natural language processing and administrative claims. RESULTS: We confirmed 84 radical cystectomies were performed for bladder cancer, with 50.0% robotic and 38.6% continent diversions. The natural language processing enhanced and claims based algorithms demonstrated 99.8% (κ=0.993, 95% CI 0.979-1.000) and 98.6% (κ=0.951, 95% CI 0.915-0.987) agreement with manual review, respectively. Both approaches accurately characterized robotic vs open surgery, with natural language processing enhanced algorithms showing 98.8% (κ=0.976, 95% CI 0.930-1.000) and claims based 90.5% (κ=0.810, 95% CI 0.686-0.933) agreement. For urinary diversion natural language processing enhanced algorithms correctly specified 96.4% of cases (κ=0.924, 95% CI 0.839-1.000) compared with 83.3% (κ=0.655, 95% CI 0.491-0.819). CONCLUSIONS: Natural language processing enhanced and claims based algorithms accurately identified radical cystectomy cases at our institution. However, natural language processing appears to better classify specific aspects of cystectomy surgery, highlighting a potential advantage of this emerging methodology.

Effect and Efficiency of an Embedded Palliative Care Nurse Practitioner in an Oncology Clinic.

PURPOSE: To test a simultaneous care model for palliative care for patients with advanced cancer by embedding a palliative care nurse practitioner (NP) in an oncology clinic. METHODS: We evaluated the effect of the intervention in two oncologists' clinics beginning March 2014 by using implementation strategies, including use of a structured referral mechanism, routine symptom screening, integration of a psychology-based cancer supportive care center, implementation team meetings, team training, and a metrics dashboard for continuous quality improvement. After 1 year of implementation, we evaluated key process and outcome measures for supportive oncology and efficiency of the model by documenting tasks completed by the NP during a subset of patient visits and time-motion studies. RESULTS: Of approximately 10,000 patients with active cancer treated in the health system, 2,829 patients had advanced cancer and were treated by 42 oncologists. Documentation of advance care planning increased for patients of the two intervention oncologists compared with patients of the other oncologists. Hospice referral before death was not different at baseline, but was significantly higher for patients of intervention oncologists compared with patients of control oncologists (53% v 23%; P = .02) over the intervention period. Efficiency evaluation revealed that approximately half the time spent by the embedded NP potentially could have been completed by other staff (eg, a nurse, a social worker, or administrative staff). CONCLUSION: An embedded palliative care NP model using scalable implementation strategies can improve advance care planning and hospice use among patients with advanced cancer.

Building the Infrastructure for Value at UCLA: Engaging Clinicians and Developing Patient-Centric Measurement.

PROBLEM: Evolving payer and patient expectations have challenged academic health centers (AHCs) to improve the value of clinical care. Traditional quality approaches may be unable to meet this challenge. APPROACH: One AHC, UCLA Health, has implemented a systematic approach to delivery system redesign that emphasizes clinician engagement, a patient-centric scope, and condition-specific, clinician-guided measurement. A physician champion serves as quality officer (QO) for each clinical department/division. Each QO, with support from a central measurement team, has developed customized analytics that use clinical data to define targeted populations and measure care across the full treatment episode. OUTCOMES: From October 2012 through June 2015, the approach developed rapidly. Forty-three QOs are actively redesigning care delivery protocols within their specialties, and 95% of the departments/divisions have received a customized measure report for at least one patient population. As an example of how these analytics promote systematic redesign, the authors discuss how Department of Urology physicians have used these new measures, first, to better understand the relationship between clinical practice and outcomes for patients with benign prostatic hyperplasia and, then, to work toward reducing unwarranted variation. Physicians have received these efforts positively. Early outcome data are encouraging. NEXT STEPS: This infrastructure of engaged physicians and targeted measurement is being used to implement systematic care redesign that reliably achieves outcomes that are meaningful to patients and clinicians-incorporating both clinical and cost considerations. QOs are using an approach, for multiple newly launched projects, to identify, test, and implement value-oriented interventions tailored to specific patient populations.

An overview of measurement activities in the partnership for patients.

The Partnership for Patients, launched in April 2011, is a national quality improvement initiative from the Department of Health and Human Services that has set ambitious goals for U.S. providers to improve patient safety and care transitions. This paper outlines the initiative's measurement strategy, describing four measurement-related objectives: (1) to track national progress toward the program goals that U.S. hospitals reduce preventable adverse events by 40% and readmissions by 20%; (2) to support local quality improvement measurement in participating hospitals by providing the appropriate tools, training, and programmatic structure; (3) to obtain feedback on hospital and contractor progress, in close to real time, so the project can be effectively managed; and (4) to evaluate the program's impact on adverse event and readmission rates.

The Impact of Air Quality on Hospital Spending.

Air pollution is harmful to human health, but little is known about the costs of pollution-related health care. If such care imposes a significant burden on insurance companies and employers, they would have substantial stakes in improving air quality. Reduced medical spending could also benefit public programs such as Medicare and Medicaid. This study estimated the amount of medical spending by private health insurers and public purchasers, such as Medicare, that is related to air pollution. Specifically, the authors determined how much failing to meet air quality standards cost various purchasers of hospital care in California over 2005-2007. The results indicate that substantial reductions in hospital spending can be achieved through reductions in air pollution.

Evaluation of organizational culture among different levels of healthcare staff participating in the Institute for Healthcare Improvement's 100,000 Lives Campaign.

BACKGROUND: Little is known about how hospital organizational and cultural factors associated with implementation of quality initiatives such as the Institute for Healthcare Improvement's (IHI) 100,000 Lives Campaign differ among levels of healthcare staff. DESIGN: Evaluation of a mixed qualitative and quantitative methodology ("trilogic evaluation model"). SETTING: Six hospitals that joined the campaign before June 2006. PARTICIPANTS: Three strata of staff (executive leadership, midlevel, and frontline) at each hospital. RESULTS. Surveys were completed in 2008 by 135 hospital personnel (midlevel, 43.7%; frontline, 38.5%; executive, 17.8%) who also participated in 20 focus groups. Overall, 93% of participants were aware of the IHI campaign in their hospital and perceived that 58% (standard deviation, 22.7%) of improvements in quality at their hospital were a direct result of the campaign. There were significant differences between staff levels on the organizational culture (OC) items, with executive-level staff having higher scores than midlevel and frontline staff. All 20 focus groups perceived that the campaign interventions were sustainable and that data feedback, buy-in, hardwiring (into daily activities), and leadership support were essential to sustainability. CONCLUSIONS: The trilogic model demonstrated that the 3 levels of staff had markedly different perceptions regarding the IHI campaign and OC. A framework in which frontline, midlevel, and leadership staff are simultaneously assessed may be a useful tool for future evaluations of OC and quality initiatives such as the IHI campaign.

Racial and ethnic disparities in hospital care resulting from air pollution in excess of federal standards.

This study investigates racial and ethnic disparities in hospital admission and emergency room visit rates resulting from exposure to ozone and fine particulate matter levels in excess of federal standards ("excess attributable risk"). We generate zip code-level ambient pollution exposures and hospital event rates using state datasets, and use pollution impact estimates in the epidemiological literature to calculate excess attributable risk for racial/ethnic groups in California over 2005-2007. We find that black residents experienced roughly 2.5 times the excess attributable risk of white residents. Hispanic residents were exposed to the highest levels of pollution, but experienced similar excess attributable risk to whites. Asian/Pacific Islander residents had substantially lower excess attributable risk compared to white. We estimate the distinct contributions of exposure and other factors to these results, and find that factors other than exposure can be critical determinants of pollution-related disparities.

Performance characteristics of a methodology to quantify adverse events over time in hospitalized patients.

OBJECTIVE: To assess the performance characteristics of the Institute for Healthcare Improvement Global Trigger Tool (GTT) to determine its reliability for tracking local and national adverse event rates. DATA SOURCES: Primary data from 2008 chart reviews. STUDY DESIGN: A retrospective study in a stratified random sample of 10 North Carolina hospitals. Hospital-based (internal) and contract research organization-hired (external) reviewers used the GTT to identify adverse events in the same 10 randomly selected medical records per hospital in each quarter from January 2002 through December 2007. DATA COLLECTION/EXTRACTION: Interrater and intrarater reliability was assessed using κ statistics on 10 percent and 5 percent, respectively, of selected medical records. Additionally, experienced GTT users reviewed 10 percent of records to calculate internal and external teams' sensitivity and specificity. PRINCIPAL FINDINGS: Eighty-eight to 98 percent of the targeted 2,400 medical records were reviewed. The reliability of the GTT to detect the presence, number, and severity of adverse events varied from κ=0.40 to 0.60. When compared with a team of experienced reviewers, the internal teams' sensitivity (49 percent) and specificity (94 percent) exceeded the external teams' (34 and 93 percent), as did their performance on all other metrics. CONCLUSIONS: The high specificity, moderate sensitivity, and favorable interrater and intrarater reliability of the GTT make it appropriate for tracking local and national adverse event rates. The strong performance of hospital-based reviewers supports their use in future studies.