Journal of Health Organization and Management Creating a fractal-based quality management infrastructure Peter J. Pronovost Jill A. Marsteller Article information: To cite this document: Peter J. Pronovost Jill A. Marsteller , (2014),"Creating a fractal-based quality management infrastructure", Journal of Health Organization and Management, Vol. 28 Iss 4 pp. 576 - 586 Permanent link to this document: http://dx.doi.org/10.1108/JHOM-11-2013-0262 Downloaded on: 12 October 2016, At: 05:55 (PT) References: this document contains references to 21 other documents. To copy this document: [email protected] The fulltext of this document has been downloaded 477 times since 2014* Users who downloaded this article also downloaded: (2015),"Small but sophisticated: Entrepreneurial marketing and SME approaches to brand management", Journal of Research in Marketing and Entrepreneurship, Vol. 17 Iss 2 pp. 149-164 http://dx.doi.org/10.1108/ JRME-05-2014-0008 (2010),"Going green: women entrepreneurs and the environment", International Journal of Gender and Entrepreneurship, Vol. 2 Iss 3 pp. 245-259 http://dx.doi.org/10.1108/17566261011079233

Access to this document was granted through an Emerald subscription provided by emerald-srm:172729 [] For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information. About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services. Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download. The current issue and full text archive of this journal is available at www.emeraldinsight.com/1477-7266.htm

JHOM 28,4 Creating a fractal-based quality management infrastructure Peter J. Pronovost Johns Hopkins Medicine Armstrong Institute for and Quality, 576 , , USA, and Received 25 November 2013 Jill A. Marsteller Revised 18 March 2014 Health Policy and Management, Accepted 11 April 2014 Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA

Abstract Purpose – The purpose of this paper is to describe how a fractal-based quality management infrastructure could benefit quality improvement (QI) and patient safety efforts in health care. Design/methodology/approach – The premise for this infrastructure comes from the QI work with health care professionals and organizations. The authors used the fractal structure system in a health system initiative, a statewide collaborative, and several countrywide efforts to improve quality of care. It is responsive to coordination theory and this infrastructure is responsive to coordination theory and repeats specific characteristics at every level of an organization, with vertical and horizontal connections among these levels to establish system-wide interdependence. Findings – The fractal system infrastructure helped a health system achieve 96 percent compliance on national core measures, and helped intensive care units across the USA, Spain, and England to reduce central line-associated bloodstream infections. Practical implications – The fractal system approach organizes workers around common goals, links all hospital levels and, supports peer learning and accountability, grounds solutions in local wisdom, and effectively uses available resources. Social implications – The fractal structure helps health care organizations meet their social and ethical obligations as learning organizations to provide the highest possible quality of care and safety for patients using their services. Originality/value – The concept of deliberately creating an infrastructure to manage QI and patient safety work and support organizational learning is new to health care. This paper clearly describes how to create a fractal infrastructure that can scale up or down to a department, hospital, health system, state, or country. Keywords Management, Quality improvement, Health and safety, Safety, Health care, Safety measures Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) Paper type Viewpoint

Today’s clinicians are asked to attack more and more patient safety issues without receiving additional resources for these tasks. They have improved hospital safety and quality by clocking in extra time and adding tasks, meetings, checks and re-checks, cross-monitoring, and other efforts that are above and beyond their regular duties. Clinicians deeply care about patients, and sincerely want to improve quality and safety. Still, they are crying “uncle” as new demands press them to enhance hospital safety and quality. Many feel they simply cannot add another initiative or program, at least not reliably or sustainably, without more support from their employer (Michtalik et al., 2013). Further, although many in health care realize the need to engage patients in decision Journal of Health Organization and making, there are few mechanisms available to meaningfully tap into patient preferences Management Vol. 28 No. 4, 2014 and ideas. pp. 576-586 r Emerald Group Publishing Limited 1477-7266 DOI 10.1108/JHOM-11-2013-0262 The authors thank Christine G. Holzmueller, BLA for editing the manuscript. Most health care organizations have only pieces of the quality infrastructure needed Fractal-based to ensure high quality, safe, patient-centered care and, in the end, patients suffer. quality Frontline workers feel personally accountable and expend significant effort recovering from mistakes, but rarely reduce the risk that the mistake will recur. They rely on management short-term solutions to correct errors, called single-loop learning (Argyris and Scho¨n, infrastructure 1974, 1978) or first-order problem solving (Tucker and Edmondson, 2003), rather than looking for the most effective systems-based solutions, known as double-loop learning 577 (Argyris and Scho¨n, 1974, 1978) or second-order problem solving (Tucker and Edmondson, 2003). Thus, we rarely learn from these mistakes at the system level, failing to reduce the risk that future patients will suffer the same harm. Clinicians and their organizations fail to learn from mistakes for multiple reasons, including an emphasis on individual vigilance within our own sphere of control, tight staffing and parsimony of resources (interest in efficiency), and removal of managers and other non-direct labor support from daily work activities, weakening any infrastructure to support quality improvement (QI) (Tucker and Edmondson, 2003). Also, in fee-for-service medicine, there are few financial incentives for health care organizations to invest in a quality management infrastructure. With poor performance on quality measures eating into provider revenue, health care leaders have a greater incentive, perhaps even a need, to create a quality management infrastructure. To improve safety, such an infrastructure should establish interdependence while allowing independence. Health care organizations need a structure that creates a chain of accountability to improve safety, defining accountable people at each level of the organization while encouraging innovation in how to improve. While safety and quality leaders have called for systems approaches to improve care, few organizations have the infrastructure to support such an approach. Hospitals and other health care organizations must provide an infrastructure that supports clinician-led efforts. They must also harness employees’ enthusiasm for patient safety, QI, and patient-centeredness to meet broader hospital-level (or even state- and national-level) initiatives by giving them the resources (equipment, evidence, and tools) and technical and motivational support needed to improve the system. The multi-level nature of the system requirements demands better coordination of available resources than typically done to date. Where can we look for solutions to the problem of

Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) coordination in multi-level systems? One place is biomimicry, which is the science of studying nature and applying its designs and processes to solve human problems. Nature has provided a model for addressing the quality management infrastructure problem, especially for complex and decentralized organizations – a fractal. Fractals, such as ferns, have self-similar patterns, wherein the whole object has the same shape as one or more of its parts, and all of the parts are connected to support and shape the larger structure in which it resides (Mandelbrot, 2012). For instance, the larger fern frond is composed of identically shaped fern leaves. Health care could accelerate improvements in safety with a quality infrastructure with such repeating patterns. We propose using a fractal organizational structure to inspire, implement, evaluate, and disseminate QI. The fractal approach is a repeating structure for building and supporting QI expertise, goal alignment, and communication at all levels of an organization (Figure 1). Its elements include: (1) a centralized improvement core at the root of the fractal structure that provides a resource group of experts in multiple improvement disciplines, including JHOM Health System Core of Expert Quality Leaders 28,4

Hospital Hospital Hospital – QI – QI Learning and sharing – QI leaders Learning and sharing 578 leaders leaders

Depts – Depts – Depts – Quality and Quality Quality Alignment of Safety Learning and sharing and Safety Learning and sharing and Safety goals, Officer Officer Officer resources, accountability, Patient and learning Involvement throughout Units – Units – levels of Units – QI QI organization QI leaders leaders Learning and sharing Learning and sharing leaders and CUSP and CUSP and CUSP team team team

Staff – Staff – Staff – Learning and sharing QI work Learning and sharing QI work QI work

Clinical Community

Notes: The health system quality leaders provide the support and expertise needed for quality improvement efforts, whether it is improvement science, biostatistics, organizational theory, teamwork and communication training, or performance improvement strategies, such Figure 1. as Lean Sigma or human factors engineering. The different levels of the organization are Illustrates a fractal vertically aligned by common goals, accountability, learning, and quality leaders. Across quality management each level, every department and unit (represented by the small circles) has a person(s) or infrastructure in health care team for the quality work, and there is cross-learning and sharing among peers. Clinical communities (represented by the light grey area) may exist across hospitals at any level Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) information technology, analytics, organizational science, human factors engineering, and Lean Sigma; (2) training and salary support for and nurse QI professionals at each level of an organizational system (ideally); (3) clearly defined and aligned goals with standard measurement, transparent reporting, and assigned accountability; (4) horizontal peer-learning structures cutting across entities at the same level; and (5) meaningful involvement of patient representatives in QI work within the fractal structure. At the end fractal branch where it touches the patient, Comprehensive Unit-based Safety Program (CUSP) teams serve as frontline implementers. CUSP is an interprofessional, local, team-based method to improve teamwork, communication, and patient safety (Pronovost et al., 2005; Timmel et al., 2010). This program is currently in wide use in intensive care units (ICUs) and inpatient wards in the USA. Fractal-based Clinical communities among, e.g. hospitalists, serve as a horizontal peer-learning quality structure; a network of CUSP teams across hospitals also exists. We have seen the benefits of this fractal model within a health system, a state, and management several countrywide efforts to improve quality. The Johns Hopkins Health System used infrastructure this system to achieve X96 percent compliance on core (process) measures, and received the National Leader Top Performer Award from The Joint Commission 579 (Pronovost et al., 2013a). We used the fractal approach in , in partnership with the Michigan Health & Hospital Association, to improve care in ICUs including reducing central line-associated bloodstream infections (CLABSIs) (Pronovost et al., 2006, 2010). The hospital association convened hospitals, hospitals recruited teams from ICUs, teams influenced patient care, and all levels remained connected for support and accountability (Pronovost et al., 2008). We also used the fractal structure in our national efforts to reduce CLABSI in Spain and across the USA (Pronovost et al., 2013b; Palomar et al., 2013). In the USA, the national structure included the American Hospital Association, various components of the Department of Health and Human Services, and the Armstrong Institute, who forged links with state hospital associations, and each hospital association made links to hospitals, hospitals used existing links to units, and units enhanced links to clinicians and patients. In each case, every level was aligned via common goals and measures; each higher level met regularly with the next level down (e.g. hospital association convened hospitals) to establish a mechanism for training, peer learning, and accountability, and to encourage local ownership and innovation when implementing the work. Given the limited evidence for QI on a large scale, we believe the success of these efforts supports the benefits of a fractal model for quality management, although further evaluation of the model is warranted.

Why the fractal works The fractal model offers a simple and efficient coordination structure to manage learning in complex health care quality systems. We anticipate that in most health care organizations there are poorly defined or no connections at the unit/clinic, department, hospital or group practice, and health system levels for safety and QI efforts. In many

Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) hospitals, formal channels for quality assurance and improvement are limited to one isolated administrative department, and common organizational goals for quality are ill defined or unstated. In such cases, coordination mechanisms for quality management and organizational learning across all levels of the organization are largely unplanned and ad hoc. To understand the labyrinth of informal connections would require a blueprint only known by the coordinating architects. In such complex systems, communicating information, sharing resources, aligning quality goals, and ensuring accountability to realize goals is often haphazard across levels, and individuals may not understand their roles in the larger system. Such an ad hoc system likely lacks both formal and informal coordination mechanisms, which are administrative tools for achieving productive work and work flow among units toward common quality goals. Martinez and Jarillo (Martinez and Jarillo, 1989) summarized the work of key coordination theorists, including March and Simon, Galbraith, Ouchi, and Thompson, in the first two columns of Table I. Major kinds of coordination mechanisms include formal and informal mechanisms, and the more complex a system is, the larger the number of coordination strategies relied upon (Martinez and Jarillo, 1989). Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) 580 28,4 JHOM ftefatlmodel fractal the of features and mechanisms coordination Common I. Table

Most common mechanisms of coordination Features of the fractal model for managing QI

Structural and 1. Departmentalization or grouping of organizational units, shaping Ensures some level of representation in each department and at formal mechanisms the formal structure each level 2. Centralization or decentralization of decision making through the Some expertise remains centralized, while authority to apply hierarchy of formal authority evidence is decentralized 3. Formalization and standardization: written policies, rules, job New job descriptions for QI – trained staff in every department; descriptions, and standard procedures, through instruments such standardization of practices where evidence exists as manuals, charts, etc. 4. Planning: strategic planning, budgeting, functional planning, QI goal alignment across levels scheduling, etc. 5. Output and behavior control: financial performance, technical Clear goals and measures and accountability for meeting QI reports, sales and marketing data, etc. and direct supervision goals Other mechanisms, 6. Lateral or cross-departmental relations: direct managerial contact, Placement of boundary spanners/integrators (staff ) in all units/ more informal and temporary or permanent teams, task forces, committees, departments; structure is “parallel,” or outside the normal subtle integrators, and integrative departments hierarchy 7. Informal communication: personal contacts among managers, Horizontal and vertical communication networks (groupings of management trips, meetings, conferences, transfer of both like unit-types and disparate types) managers, etc. 8. Socialization: building an organizational culture of known and Training in the science of safety (e.g. systems awareness and shared strategic objectives and values by training, transfer of thinking; non-punitive response to error) and in quality managers, career path management, measurement and reward improvement skills systems, etc.

Sources: Modification of Table I (Martinez and Jarillo, 1989). Permission to republish was granted by Palgrave Macmillan on September 4, 2013 A fractal system is a planned coordination mechanism with quality management and Fractal-based organizational learning represented at every level of the organization (Figure 1). quality Further, there are connections both vertically across organizational levels to support accountability and horizontally among units at the same level to support peer learning. management At every level, a fractal quality management system has a structure of skilled infrastructure professionals, resources, goals, measures, and accountability for QI and learning. Vertical connections tie the clinical area or unit to the next higher unit in the hospital, 581 and to the higher level goals, resources, and support for accountability. Through vertical connections, unit-level QI staff can access needed expertise that exists at higher organizational levels to supplement their own skills and carry out local safety and QI efforts. Horizontal connections bring peers together, such as frontline providers on a given unit or hospitalists across multiple hospitals in the same health system. These connections support the sharing of real world experiences, the spread of best practices, organizational learning, and system-wide innovation. A fractal quality management structure could be the formal reporting structure or a parallel structure for existing formal reporting structures and, in keeping with the complexity of the system (Martinez and Jarillo, 1989), supplement formal coordination mechanisms with informal ones, including lateral relations, informal communication networks, and socialization. Applied to health care, the fractal model pools the strengths within an organization to offer balance and greater success. For example, self-organizing teams are often innovative and highly motivated, but may lack accountability, autonomy, and resources. Top-down management has strong accountability and considerable resources, but sometimes stifles innovation by failing to rouse intrinsic motivation to change among clinicians. All levels of the organization are linked through this fractal system, allowing both the local independence and system-wide interdependence needed to succeed. Below, we describe how a fractal health care quality management system would work.

Creating fractals for QI Fundamentally, the fractal model requires an organizational investment in clinicians, managers, and other quality and safety-related professionals, as well as involvement of the patient perspective. One constant factor in the patient safety movement is

Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) a clinician’s inclination to adopt new behaviors when initiatives tap their personal drive to change, that is, when intrinsic rather than extrinsic motivators are employed. An intervention to improve safety succeeds and is sustained when inspired clinicians own it and, through social norms, hold one another accountable for adopting it (Seidman, 2007; Pink, 2009). For example, caregivers will not consistently use an infection prevention checklist or a communication tool unless they believe it is valid, modify it to fit their context, and settle on a locally feasible implementation approach (Pronovost et al., 2006; Marsteller et al., 2012). Interventions that are done to, rather than with, professionals are highly resisted, only partially or superficially implemented, and are generally ineffective because they do not fit the local context in which care is delivered (Dixon-Woods et al., 2011). Many well-intentioned change efforts have failed by ignoring this crucial factor. Thus, simply designating a hospital patient safety or quality officer and issuing safety and quality directives is insufficient. At the unit level, the fractal model taps frontline staff in departments and units to take the lead in applying patient safety and QI efforts to their local context. Designated staff leaders need time away from clinical duties to devote to this work, special training in QI and patient safety methods, JHOM technical support in areas such as measurement, and small budgets to support QI 28,4 activities at that level. Patient perspectives should be incorporated at each level as well. The work of each level is somewhat different. In addition, the time and resources required to conduct the work increases moving from lower to higher organizational levels, although the number of staff members in these roles concurrently diminishes moving up the levels. For instance, unit-level nurse and physician quality leaders could 582 devote 20 percent of their time to improving quality. A unit could have multiple quality and safety leaders. Duties at this level include engaging other staff in QI activities, educating peers on evidence-based practices, monitoring implementation, engaging a patient representative, and feeding back process and outcome results both horizontally and vertically. At the department level, one designated quality leader would devote 50 percent or more of their time to quality and safety across the units in the department. Duties at this level include implementing organization-wide quality and safety goals, helping frontline staff prioritize issues to work on, ensuring needed resources are available, providing assurances of management’s commitment, helping structure, execute and evaluate quality and safety initiatives, and holding units accountable for delivering on system-wide, department, and unit-level goals. Patient representatives might be engaged at this level and above on quality committees. At the hospital level, full-time safety and quality staff are required. These staff will help set hospital-level goals, develop strategies and tactics to realize those goals, provide technical support to all quality and safety projects across the hospital, track performance, review performance with QI staff at other levels, ensure compliance with regulation, provide guidance on priorities for quality and safety improvement, and ensure departments are accountable for system-wide and departmental quality and safety goals. The fractal system creates efficiency by connecting local units to core resources from across the organization that they can access when needed.

Building capacity for QI and organizational learning As a system transitions to a fractal model, patient safety and quality leaders at every level must develop or expand their skills. At the unit level, nurse and physician quality leaders should learn fundamental patient safety concepts and tools, such as how to

Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) improve safety culture and teamwork, identify safety hazards, provide patient-centered care, implement evidence-based practices, manage projects, lead change, critique data quality, and implement interventions to reduce safety risks. This may also apply for small departments. For larger departments, the designated quality leaders should receive the same fundamental patient safety concepts and training, plus advanced training in implementation, such as (Lean Sigma “Black Belt”), and in teamwork, leadership, and evaluation methodology, largely through a master’s degree or a doctorate in public health. Hospital-level leaders would undergo the same training as department leaders, plus training in targeted areas, such as waste and efficiency, teamwork training, organizational behavior, and human-technology interface, to ensure health care systems are designed for safety. In addition, the hospital or health system quality infrastructure may include staff with advanced degrees in statistics, sociology, psychology, informatics, and human factors and industrial or systems engineering. Structurally, the fractal model links staff, units, departments, and hospitals horizontally and vertically to support organizational learning. Peers at each level should deliberate together on solutions to safety hazards and work with other levels of the organization to share learning and receive support. Staff may use CUSP to identify hazards and mitigate them (Pronovost et al., 2005; Timmel et al., 2010). CUSP teams can Fractal-based be linked horizontally across hospitals and vertically across levels to permit joint quality problem solving and peer learning. Clinical communities, which comprise professionals or safety teams with common clinical foci from various areas in a hospital or health management system, can establish this horizontal link. These are long-term learning networks that infrastructure work on both common and local harms and receive support from a system-level quality resource, such as a hospital safety office or an institute. 583 Establishing goals and measures of QI Each level of the organization needs goals and measures to judge goal attainment. While system-wide goals cascade down through an organization, each level should also set goals and measures that are both intrinsically valued by staff at that level and aligned with the overall organizational mission and objectives. The lower levels (clinical areas or units) often have the largest number of goals because they have to meet system-wide, hospital, department, and unit goals. Paradoxically, the lowest organizational levels typically have the fewest resources to improve. The fractal model hopes to address this paradox by linking local units to resources at other levels of the organization.

Connecting the fractals, establishing system-wide accountability There must be accountability to improve patient safety and quality. For this, leaders from each level of the fractal system meet regularly with entities at the next lower level to share learning, ensure they have sufficient resources and skills, periodically review process and outcome measures, problem solve, and act as quality stewards, promoting awareness of and responsiveness to the goals set. For example, health system quality leaders meet with hospital chief executive officers (CEOs) to review performance, hospital CEOs meet with department leaders, and department leaders meet with unit leaders. There is tension when creating a health system that both learns and is accountable. The higher management levels must support the lower levels, ensuring they have the right skills, resources, and information to succeed and create structures to support peer learning. Yet management must also ensure that units and departments deliver results.

Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) In the pursuit of results, the higher levels must always remember to facilitate success at the lower levels. Higher levels can clarify and hold staff accountable for performance while encouraging local innovation in how to improve performance. To help create a structure to support this balance, quality leaders can work closely with the unit, department, or hospital manager at their level as well as the quality leader at the next higher level. Managers should evaluate these employees with involvement from the quality management leadership. We have discussed the challenges associated with QI efforts that are primarily technical challenges, such as lack of valid measures and staff skilled in improvement work. These challenges are more concrete, easy to target and, in many cases, easier to solve than adaptive (psychological/social/cultural) challenges, which are dynamic and complex. Adaptive change is essentially the gateway to implementing and sustaining QI, particularly culture change among professionals. Millar (2013) describes the use of Improvement Leaders’ Guides to socialize QI tools and techniques in health care settings. Intrinsic motivation, something we have stressed in this paper, is starting to emerge as a pertinent factor for engaging and changing clinician behavior (Nantha, 2013). We caution those developing interventions to explore the context in which they JHOM will be implemented to understand what intrinsic motivators support or inhibit the 28,4 change. What motivates a physician may not motivate a nurse. Project leaders must get both the technical and adaptive work right: a fractal structure can help. Although some institution-wide projects may be identified at higher levels of the fractal quality management infrastructure, greatest buy-in usually occurs for projects identified at the lowest level. Unit-level work is supported by vertical connections to 584 higher organizational levels, including engaged experts and leaders who can guide the work and help staff overcome barriers to success. In addition, unit work and learning is enhanced by vertical linking of peer units. It is not feasible or efficient for each unit or department to develop or hire all the expertise needed to be successful, whether it is biostatistics, organizational theory, teamwork and communication training, or performance improvement strategies, such as Lean Sigma or human factors engineering. Our experience using a fractal model for QI and patient safety in several projects involving multi-level entities has been extremely positive. In addition, the use of this model as a coordinating structure is supported by important principles in the organizational behavior and management literatures. When the fractal model’s characteristics are present throughout a health care organization, workers are inspired by common goals, all hospital levels are linked and supporting learning and accountability, solutions are grounded in local wisdom, patient perspectives are considered and resources are present to effect change. What’s more, the fractal model does not require complicated diagrams or blue prints – those who are part of the structure intuitively grasp the relationships between different parts of the whole. Fractals are efficient and effective structures in nature that show significant promise for helping health care organizations improve patient safety and quality.

References Argyris, C. and Scho¨n, D.A. (1974), Theory in Practice: Increasing Professional Effectiveness, Jossey-Bass Publishers, San Francisco, CA. Argyris, C. and Scho¨n, D.A. (1978), Organizational Learning: A Theory of Action Perspective, Addison Wesley, Reading, MA. Dixon-Woods, M., Bosk, C.L., Aveling, E., Goeschel, C.A. and Pronovost, P.J. (2011), “Explaining Michigan: developing an ex post theory of a quality improvement program”, Milbank Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) Quarterly, Vol. 89 No. 2, pp. 167-205. Mandelbrot, B. (2012), The Fractalist: Memoir of a Scientific Maverick, Pantheon Books, New York, NY. Marsteller, J.A., Sexton, J.B., Hsu, Y.J., Hsiao, C.J., Holzmueller, C.G., Pronovost, P.J. and Thompson, D.A. (2012), “A multicenter, phased, cluster-randomized controlled trial to reduce central line-associated bloodstream infections in intensive care units”, Critical Care Medicine, Vol. 40 No. 11, pp. 2933-2939. Martinez, J.I. and Jarillo, J.C. (1989), “The evolution of research on coordination mechanisms in multinational corporations”, Journal of International Business Studies, Vol. 20 No. 3, pp. 489-514. Michtalik, H.J., Yeh, H.C., Pronovost, P.J. and Brotman, D.J. (2013), “Impact of attending physician workload on patient care: a survey of hospitalists”, Journal of the American Medical Association Internal Medicine, Vol. 173 No. 5, pp. 375-377. Millar, R. (2013), “Framing quality improvement tools and techniques in healthcare. The case of improvement leaders’ guides”, Journal of Health Organization and Management, Vol. 27 No. 2, pp. 209-224. Nantha, Y.S. (2013), “Intrinsic motivation: how can it play a pivotal role in changing Fractal-based clinician behavior?”, Journal of Health Organization and Management,Vol.27No.2, pp. 266-272. quality Palomar, M., Alvarez-Lerma, F., Riera, A., Diaz, M.T., Torres, F., Agra, Y., Larizgoitia, I., management Goeschel, C.A. and Pronovost, P.J. on behalf of the Bacteremia Zero Working Group infrastructure (2013), “Impact of a national multimodal intervention to prevent catheter-related bloodstream infection in the ICU: the Spanish experience”, Critical Care Medicine, Vol. 41 No. 10, pp. 2364-2372. 585 Pink, D. (2009), Drive: The Surprising Truth About What Motivates Us, Penguin Group (USA) Inc., New York, NY. Pronovost, P.J., Berenholtz, S.M., Goeschel, C., Thom, I., Watson, S.R., Holzmueller, C.G., Lyons, J.S., Lubomski, L.H., Thompson, D.A., Needham, D., Hyzy, R., Welsh, R., Roth, G., Bander, J., Morlock, L. and Sexton, J.B. (2008), “Improving patient safety in intensive care units in Michigan”, Journal of Critical Care, Vol. 23 No. 2, pp. 207-221. Pronovost, P.J., Demski, R., Callender, T., Winner, L., Miller, M.R., Austin, J.M. and Berenholtz, S.M. National Leadership Core Measures Work Groups (2013a), “Demonstrating high reliability on accountability measures at the ”, Joint Commission Journal on Quality and Patient Safety,Vol.39No.12,pp.531-544. Pronovost, P.J., Goeschel, C.A., Colantuoni, E., Watson, S., Lubomski, L.H., Berenholtz, S.M., Thompson, D.A., Sinopoli, D., Cosgrove, S., Sexton, J.B., Marsteller, J.A., Hyzy, R.C., Welsh, R., Posa, P., Schumacher, K. and Needham, D. (2010), “Sustaining reductions in catheter related bloodstream infections in Michigan intensive care units: observational study”, British Medical Journal,Vol.340,p.c309. Pronovost, P.J., Marsteller, J.A., Weeks, K., Watson, S.R., Berenholtz, S.M., Goeschel, C.A., Pham, J.C., Winters, B.D., Lubomski, L.H., Thompson, D.A., George, C., Wyskiel, R.M., Federowicz, M., Battles, J.B., Hines, S.C., Sawyer, M.D. and Combes, J.R. (2013b), “A national initiative to reduce central line-associated bloodstream infections: a model for reducing preventable harm”, Health Aff Blog, available at: http://healthaffairs.org/blog/2013/09/23/a-national- initiative-to-reduce-central-line-associated-bloodstream-infections-a-model-for-reducing- preventable-harm/ (accessed November 21, 2013). Pronovost, P., Needham, D., Berenholtz, S., Sinopoli, D., Chu, H., Cosgrove, S., Sexton, B., Hyzy, R., Welsh, R., Roth, G., Bander, J., Kepros, J. and Goeschel, C. (2006), “An intervention to decrease catheter-related bloodstream infections in the ICU”, New England Journal of Medicine, Vol. 355 No. 26, pp. 2725-2732. Downloaded by JHU Libraries At 05:55 12 October 2016 (PT) Pronovost, P., Weast, B., Rosenstein, B., Sexton, J.B., Holzmueller, C.G., Paine, L., Davis, R. and Rubin, H.R. (2005), “Implementing and validating a comprehensive unit-based safety program”, Journal of Patient Safety, Vol. 1 No. 1, pp. 33-40. Seidman, D. (2007), How: Why We Do Anything Means Everythingy.in Business (and in Life), John Wiley & Sons Inc., Hoboken, NJ. Timmel, J., Kent, P.S., Holzmueller, C.G., Paine, L.A., Schulick, R.D. and Pronovost, P.J. (2010), “Impact of the comprehensive unit-based safety program (CUSP) on safety culture in a surgical inpatient unit”, Joint Commission Journal on Quality and Patient Safety, Vol. 36 No. 6, pp. 252-260. Tucker, A.L. and Edmondson, A.C. (2003), “Why hospitals don’t learn from failures: organizational and psychological dynamics that inhibit system change”, California Management Review, Vol. 45 No. 2, pp. 1-18.

Further reading Stavert, R.R. and Lott, J.P. (2013), “The bystander effect in medical care”, The New England Journal of Medicine, Vol. 368 No. 1, pp. 8-9. JHOM About the authors 28,4 Dr Peter J. Pronovost, MD, PhD, is the Director of the Armstrong Institute for Patient Safety and Quality and the Senior Vice President for Patient Safety and Quality for Johns Hopkins Medicine, and a Professor of Anesthesiology and Critical Care Medicine, Surgery, and Health Policy and Management at the . He is a practicing Anesthesiologist and Critical Care Physician and a Health Services Researcher with expertise in patient safety, QI, and clinical 586 outcomes in the intensive care unit and surgery. Dr Peter J. Pronovost is the corresponding author and can be contacted at: [email protected] Dr Jill A. Marsteller, PhD, MPP, is an Associate Professor of Health Policy and Management at the Johns Hopkins Bloomberg School of Public Health and is jointly appointed in the Armstrong Institute for Patient Safety and Quality of the Johns Hopkins School of Medicine. She has been involved in health services research for 20 years and specializes in organizational behavior and theory as applied to QI in health care delivery. Downloaded by JHU Libraries At 05:55 12 October 2016 (PT)

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints This article has been cited by:

1. Steven J. Kravet, Jennifer Bailey, Renee Demski, Peter Pronovost. 2016. Establishing an Ambulatory Medicine Quality and Safety Oversight Structure. Academic Medicine 91:7, 962-966. [CrossRef] 2. Nadia Hensley, Tracey L. Stierer, Colleen G. Koch. 2016. Defining Quality Markers for Cardiac Anesthesia: What, Why, How, Where to, and Who’s on Board?. Journal of Cardiothoracic and Vascular Anesthesia . [CrossRef] 3. John Matthew Austin, Peter J. Pronovost. 2016. Improving performance on core processes of care. Current Opinion in Allergy and Clinical Immunology 16:3, 224-230. [CrossRef] 4. Lisa Ishii, Peter J. Pronovost, Renee Demski, Gill Wylie, Michael Zenilman. 2016. A Model for Integrating Ambulatory Surgery Centers Into an Academic Health System Using a Novel Ambulatory Surgery Coordinating Council. Academic Medicine 91:6, 803-806. [CrossRef] 5. Simon C. Mathews, Renee Demski, Peter J. Pronovost. 2016. Redefining Accountability in Quality and Safety at Academic Medical Centers. Quality Management in Health Care 25:4, 244-247. [CrossRef] 6. Peter J. Pronovost, Christine G. Holzmueller, Nancy E. Molello, Lori Paine, Laura Winner, Jill A. Marsteller, Sean M. Berenholtz, Hanan J. Aboumatar, Renee Demski, C. Michael Armstrong. 2015. The Armstrong Institute. Academic Medicine 90:10, 1331-1339. [CrossRef] Downloaded by JHU Libraries At 05:55 12 October 2016 (PT)