Simon C. Mathews

Physician autonomy and informed decision making: finding the balance for patient safety and quality.

1. Appelbaum PC. Microbiology of antibiotic resistance in Staphylococcus aureus. Clin Infect Dis. 2007;45(suppl 3):S165-S170. 2. Pitout JD, Laupland KB. Extended-spectrum beta-lactamase-producing Enterobacteriaceae: an emerging public-health concern. Lancet Infect Dis. 2008;8 (3):159-166. 3. Queenan AM, Bush K. Carbapenemases: the versatile beta-lactamases. Clin Microbiol Rev. 2007;20(3):440-458. 4. Schwaber MJ, Klarfeld-Lidji S, Navon-Venezia S, Schwartz D, Leavitt A, Carmeli Y. Predictors of carbapenem-resistant Klebsiella pneumoniae acquisition among hospitalized adults and effect of acquisition on mortality. Antimicrob Agents Chemother. 2008;52(3):1028-1033. 5. Bratu S, Landman D, Haag R, et al. Rapid spread of carbapenem-resistant Klebsiella pneumoniae in New York City: a new threat to our antibiotic armamentarium. Arch Intern Med. 2005;165(12):1430-1435. 6. Vatopoulos A. High rates of metallo-beta-lactamase-producing Klebsiella pneumoniae in Greece: a review of the current evidence. Euro Surveill. 2008;13 (4):8023. 7. European Antimicrobial Resistance Surveillance System Web site. http://www .rivm.nl/earss/. Accessed November 14, 2008. 8. Rodriguez-Bano J, Alcala JC, Cisneros JM, et al. Community infections caused by extended-spectrum beta-lactamase-producing Escherichia coli. Arch Intern Med. 2008;168(17):1897-1902. 9. Schwaber MJ; Israel CRE Working Group. Infection control at the national level: containment of an outbreak of carbapenem-resistant Klebsiella pneumoniae (CRKP) in Israeli hospitals. Presented at: 48th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy/Infectious Diseases Society of America 46th Annual Meeting; Washington, DC; October 27, 2008. Presentation K-3509. 10. Klevens RM, Morrison MA, Nadle J, et al; Active Bacterial Core surveillance (ABCs) MRSA Investigators. Invasive methicillin-resistant Staphylococcus aureus infections in the United States. JAMA. 2007;298(15):1763-1771. 11. Bratu S, Brooks S, Burney S, et al. Detection and spread of Escherichia coli possessing the plasmid-borne carbapenemase KPC-2 in Brooklyn, New York. Clin Infect Dis. 2007;44(7):972-975. 12. Urban C, Bradford PA, Tuckman M, et al. Carbapenem-resistant Escherichia coli harboring Klebsiella pneumoniae carbapenemase beta-lactamases associated with long-term care facilities. Clin Infect Dis. 2008;46(11):e127-e130. 13. Paterson DL, Bonomo RA. Extended-spectrum beta-lactamases: a clinical update. Clin Microbiol Rev. 2005;18(4):657-686. 14. Schwaber MJ, Carmeli Y. Mortality and delay in effective therapy associated with extended-spectrum beta-lactamase production in Enterobacteriaceae bacteraemia: a systematic review and meta-analysis. J Antimicrob Chemother. 2007; 60(5):913-920. 15. Daikos GL, Karabinis A, Paramythiotou E, et al. VIM-1-producing Klebsiella pneumoniae bloodstream infections: analysis of 28 cases. Int J Antimicrob Agents. 2007;29(4):471-473.

Mining high-dimensional administrative claims data to predict early hospital readmissions

BACKGROUND Current readmission models use administrative data supplemented with clinical information. However, the majority of these result in poor predictive performance (area under the curve (AUC)<0.70). OBJECTIVE To develop an administrative claim-based algorithm to predict 30-day readmission using standardized billing codes and basic admission characteristics available before discharge. MATERIALS AND METHODS The algorithm works by exploiting high-dimensional information in administrative claims data and automatically selecting empirical risk factors. We applied the algorithm to index admissions in two types of hospitalized patient: (1) medical patients and (2) patients with chronic pancreatitis (CP). We trained the models on 26,091 medical admissions and 3218 CP admissions from The Johns Hopkins Hospital (a tertiary research medical center) and tested them on 16,194 medical admissions and 706 CP admissions from Johns Hopkins Bayview Medical Center (a hospital that serves a more general patient population), and vice versa. Performance metrics included AUC, sensitivity, specificity, positive predictive values, negative predictive values, and F-measure. RESULTS From a pool of up to 5665 International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnoses, 599 ICD-9-CM procedures, and 1815 Current Procedural Terminology codes observed, the algorithm learned a model consisting of 18 attributes from the medical patient cohort and five attributes from the CP cohort. Within-site and across-site validations had an AUC≥0.75 for the medical patient cohort and an AUC≥0.65 for the CP cohort. CONCLUSIONS We have created an algorithm that is widely applicable to various patient cohorts and portable across institutions. The algorithm performed similarly to state-of-the-art readmission models that require clinical data.

The need for systems integration in health care.

ROGRESS IN PATIENT SAFETY AND QUALITY HAS BEEN slow, despite increasing recognition of risk across the health care system. Efforts to reduce harm to patients or to improve quality of care often focus on a single, local intervention or a collection of local interventions, usually seeking to improve a single care process. Although valuable, this approach is incremental, resulting inmodest,thoughneeded,improvements.Mostqualityimprovement efforts miss a larger opportunity to improve and redesign the fabric of health care. It appears that a systemsintegration approach that incorporates the fundamental building blocks of health care, from equipment and technologytoclinicalinsightandworkflowprocesses,isneeded

Decreasing central-line-associated bloodstream infections in Connecticut intensive care units.

Abstract: Central‐line–associated bloodstream infections (CLABSIs) are a significant cause of preventable harm. A collaborative project involving a multifaceted intervention was used in the Michigan Keystone Project and associated with significant reductions in these infections. This intervention included the Comprehensive Unit‐based Safety Program, a multifaceted approach to CLABSI prevention, and the monitoring and reporting of infections. The purpose of this study was to determine whether the multifaceted intervention from the Michigan Keystone program could be implemented in Connecticut and to evaluate the impact on CLABSI rates in intensive care units (ICUs). The primary outcome was the NHSN‐defined rate of CLABSI. Seventeen ICUs, representing 14 hospitals and 104,695 catheter days were analyzed. The study period included up to four quarters (12 months) of baseline data and seven quarters (21 months) of postintervention data. The overall mean (median) CLABSI rate decreased from 1.8 (1.8) infections per 1,000 catheter days at baseline to 1.1 (0) at seven quarters postimplementation of the intervention. This study demonstrated that the multifaceted intervention used in the Keystone program could be successfully implemented in another state and was associated with a reduction in CLABSI rates in Connecticut. Moreover, even though the statewide baseline CLABSI rate in Connecticut was low, rates were reduced even further and well below national benchmarks.

Achieving Secondary Prevention Low-Density Lipoprotein Particle Concentration Goals Using Lipoprotein Cholesterol-Based Data

Background Epidemiologic studies suggest that LDL particle concentration (LDL-P) may remain elevated at guideline recommended LDL cholesterol goals, representing a source of residual risk. We examined the following seven separate lipid parameters in achieving the LDL-P goal of <1000 nmol/L goal for very high risk secondary prevention: total cholesterol to HDL cholesterol ratio, TC/HDL, <3; a composite of ATP-III very high risk targets, LDL-C<70 mg/dL, non-HDL-C<100 mg/dL and TG<150 mg/dL; a composite of standard secondary risk targets, LDL-C<100, non-HDL-C<130, TG<150; LDL phenotype; HDL-C≥40; TG<150; and TG/HDL-C<3. Methods We measured ApoB, ApoAI, ultracentrifugation lipoprotein cholesterol and NMR lipoprotein particle concentration in 148 unselected primary and secondary prevention patients. Results TC/HDL-C<3 effectively discriminated subjects by LDL-P goal (F = 84.1, p<10−6). The ATP-III very high risk composite target (LDL-C<70, nonHDL-C<100, TG<150) was also effective (F = 42.8, p<10−5). However, the standard secondary prevention composite (LDL-C<100, non-HDL-C<130, TG<150) was also effective but yielded higher LDL-P than the very high risk composite (F = 42.0, p<10−5) with upper 95% confidence interval of LDL-P less than 1000 nmol/L. TG<150 and TG/HDL-C<3 cutpoints both significantly discriminated subjects but the LDL-P upper 95% confidence intervals fell above goal of 1000 nmol/L (F = 15.8, p = 0.0001 and F = 9.7, p = 0.002 respectively). LDL density phenotype neared significance (F = 2.85, p = 0.094) and the HDL-C cutpoint of 40 mg/dL did not discriminate (F = 0.53, p = 0.47) alone or add discriminatory power to ATP-III targets. Conclusions A simple composite of ATP-III very high risk lipoprotein cholesterol based treatment targets or TC/HDL-C ratio <3 most effectively identified subjects meeting the secondary prevention target level of LDL-P<1000 nmol/L, providing a potential alternative to advanced lipid testing in many clinical circumstances.

A Model for the Departmental Quality Management Infrastructure Within an Academic Health System.

As quality improvement and patient safety come to play a larger role in health care, academic medical centers and health systems are poised to take a leadership role in addressing these issues. Academic medical centers can leverage their large integrated footprint and have the ability to innovate in this field. However, a robust quality management infrastructure is needed to support these efforts. In this context, quality and safety are often described at the executive level and at the unit level. Yet, the role of individual departments, which are often the dominant functional unit within a hospital, in realizing health system quality and safety goals has not been addressed. Developing a departmental quality management infrastructure is challenging because departments are diverse in composition, size, resources, and needs. In this article, the authors describe the model of departmental quality management infrastructure that has been implemented at the Johns Hopkins Hospital. This model leverages the fractal approach, linking departments horizontally to support peer and organizational learning and connecting departments vertically to support accountability to the hospital, health system, and board of trustees. This model also provides both structure and flexibility to meet individual departmental needs, recognizing that independence and interdependence are needed for large academic medical centers. The authors describe the structure, function, and support system for this model as well as the practical and essential steps for its implementation. They also provide examples of its early success.

Management's Discussion and Analysis: A tool for advancing quality and safety.

The Institute of Medicines landmark 1999 report, To Err is Human, set the stage for progress in quality and safety in health care. Yet 15 years later, recent findings still suggest that potentially up to 400,000 preventable deaths occur yearly. Part of the challenge in addressing this gap in health care is that systems for accountability are underdeveloped and standardization is lacking in the way quality and safety data are presented to health system leaders and boards. Creating a robust performance management system in health care to provide standardization and accountability similar to that found in financial accounting has been suggested; however, this represents a transformational step that is still incipient and requires broad health care stakeholder involvement. In the interim, it is possible to adapt financial concepts to quality and safety at a local level, with individual hospitals and health systems leading the way for greater progress. At Johns Hopkins Medicine (JHM), the academic medical center that includes The Johns Hopkins Hospital and Health System, we made a deliberate effort to bring greater oversight and accountability to quality and safety by modeling after financial reporting. Specifically, just as our institutions Board of Trustees Finance committee has direct oversight for every dollar spent and received throughout JHM, a similar goal was set for the hospitals Board of Quality. However, the nature of quality and safety does not lend itself to convenient and straightforward analysis of balance sheets, income statements, and cash flow statements. There is no summary profit and loss that can be determined, and as a result, there is no standardized way of calculating a bottom line for quality performance. Health care does have various process and outcome measures of quality and safety, indeed hundreds of measures, but these are often disparate and are not easily aggregated. As a result, it is not straightforward to provide an assessment on the overall “state of quality” at a given institution. We have attempted to bridge this gap by recognizing that both qualitative and quantitative information are important in quality and safety. It is in this setting that at our institution we adapted the financial reporting concept of Managements Decision and Analysis (MD&A) into a tool for quality and safety. Combined with a cascading system for accountability, this tool has provided a standardized way of reporting and addressing quality throughout our health system. As stipulated by the Federal Accounting Standards Advisory Board, the Managements Decision and Analysis is a mandatory section in federal financial reporting that provides a broad overview to “address the reporting entitys performance measures, financial statements, systems and controls, compliance with laws and regulations, and actions taken or planned to address

Commentary: Establishing Safety and Quality as Core Values A Hospital Road Map

Although still immature, the field of quality and safety in medicine has gained significant traction since the landmark publication of the Institute of Medicine’s To Err Is Human in 1999. The confluence of many system-level factors, including more robust research, increasing regulatory pressure, more vocal consumer advocacy, and health care reform, have created a health care environment that is very much aware of safety and quality. Not surprisingly, hospitals are promoting themselves as centers of excellence and recipients of safety and quality awards. Unfortunately, there is little transparency in the advertising that now permeates the health care marketplace. For safety and quality efforts to succeed, these concepts must go beyond superficial marketing and instead become core values within health care systems and be based on robust data demonstrating improved patient outcomes. We outline a basic road map for transforming existing hospitals and health systems into safety-driven and quality-driven institutions. This process involves the following steps: establishing a local safety culture, creating a chain of accountability within a physician management infrastructure, measuring and reporting valid outcomes, developing robust tools for quality improvement, and participating in a system of peer review while also expanding sources of internal input. As in aviation, where all employees in the air and on the ground are tasked with the safety of travelers, health care must establish a culture where all employees work to ensure the safety of patients. Creating a safety culture may seem to be an amorphous task, but it can be accomplished systematically. Furthermore, it is fundamental to the sustainability of future quality and safety initiatives. Although culture can be framed institutionally, it is ultimately local—living within each clinician, existing within patient care areas. One proven model that incorporates this belief is the Comprehensive Unit-based Safety Program. The first step of the program is to educate all staff on the science of safety. Basic tenets include the following: safety is a property of the health care system; safe design includes standardizing work, creating independent checks (checklists) for key processes, and learning from mistakes; and wise decisions are made when there is diverse and independent input. The next step is for staff to identify and then learn from defects in their work areas. Additionally, local units should adopt a senior executive to become part of their safety team. Last, staff are encouraged to use and develop tools to support teamwork. Quality and safety efforts are implemented at the unit level but have the coordination and support of senior leaders at the organization level. As a result, the Comprehensive Unit-based Safety Program provides a strategic plan to improve culture and has been associated with improved culture in more than 100 intensive care units in Michigan and in all patient care areas at large academic medical centers. However, these locally driven efforts must be supported by an overarching infrastructure. This is where a chain of accountability and a physician management infrastructure play a significant role. Physicians routinely accept responsibility for individual patient outcomes; however, this degree of ownership currently does not extend to the population level (eg, for patients cared for on a unit or in an entire hospital). Although some hospital areas such as intensive care units have a physician management framework in place, this model is not widespread. Physicians bring a unique perspective to quality and safety because of their close relationships with patients. They should be the central nodes within a quality and safety network. This network certainly should include the nursing, regulatory, and administrative officers who currently lead safety efforts, but as long as physicians believe that safety is “someone else’s job,” there will not be sufficient progress. As such, physicians need to help define quality measures that are valid and meaningful, and then work to improve performance; only then can senior leaders hold them accountable for improving performance. Yet accountability should not stop with the physician; it should extend all the way to the chief executive. However, for physicians and senior management to become engaged, 424332 AJMXXX10.1177/1062860611424332Mathe ws and PronovostAmerican Journal of Medical Quality XX(X)

Focus on Quality: An Opportunity to Execute Health Care Reform

Although the groundbreaking passage of the health care reform bill marks a significant milestone for the US health care system, the crisis in health care remains: 47 to 54 million Americans lack insurance, health care costs compromise our global competitiveness and the welfare of future generations, and the quality of care remains erratic and often low. The new law, the Patient Protection and Affordable Care Act, fundamentally addresses the important issue of access but does less to aggressively reduce cost and actively advance quality measures. We believe that addressing quality in health care is the next step in executing broader reform. Specifically, we propose the creation of a health care regulatory and enforcement body, equivalent in many respects to the Securities and Exchange Commission (SEC), to be an integral part of advancing national quality improvement in health care.

Creating a purpose-driven learning and improving health system: The Johns Hopkins Medicine quality and safety experience

Health care has often relied on independent silos of medical research to drive progress and innovation. However, this approach does not adequately address the complexities and opportunities within the modern health care environment. We posit that creating a learning and improving health system that is purpose‐driven will ultimately lead the next transformation in health care. We share the experience within Johns Hopkins Medicine that established a learning and improving health system in quality and safety. The system is built around a clear and compelling patient‐centered purpose and leverages a fractal framework that provides horizontal links for peer learning and vertical links for accountability. It dismantles traditional research and clinical silos and combines basic and applied research with health system operations. As a result, the system aligns the goals and strengths of a diverse set of stakeholders including clinicians, patients, researchers, and administrators toward a common goal.