A system is an interconnected set of elements organized to achieve a function or purpose. A system of care is an interconnected care delivery system, usually in a geographically contiguous region, organized to provide the opportunity to improve processes and outcomes of care. A critical element of any effective system of care is ongoing measurement and response to the quality of care. This chapter describes systems theory, experience with cardiovascular systems of care, quality improvement theory, experience with cardiovascular quality improvement programs, and lessons learned from these efforts about how further improvements and optimization of cardiovascular care can be achieved.
Systems theory is the interdisciplinary study of systems, with the goal of elucidating common principles that can be applied to diverse systems. The founding of systems theory is attributed to multiple individuals. The origin of systems theory has been attributed to the Industrial Revolution, when the relationships among structure, function, and output of manufacturing processes were evaluated with science, logic, and reductionism. Albert Einstein promulgated the concept that multiple perspectives exist, with differing levels of behavior and knowledge that are interlinked to observe, understand, and change phenomena.
Systems thinking is the process of understanding how factors influence one another within a whole. In nature, examples of systems include ecosystems, in which various elements—such as air, water, movement, plants, and animals—work together to ensure that organisms within the system survive; without such cooperation, they would perish. In organizations, systems consist of people, processes, and structures that work together to make an organization healthy or unhealthy. Structures may consist of a physical plant or devices.
For systems of care, improvements in process may include increased use of effective drugs, devices, or procedures; decreased use of ineffective interventions; delivery of the same interventions with fewer resources; or improved organizational culture. Systems of care are distinct from health systems and are intended to meet the health care needs of patients with one or more specific clinical disorders (e.g., out-of-hospital cardiac arrest [OHCA], ST-elevation myocardial infarction [STEMI], heart failure, stroke, or trauma). According to the World Health Organization, health care systems meet the health care needs of a target population by providing 1) a financing mechanism, 2) a well-trained and adequately paid workforce, 3) reliable information on which to base decisions and policies, and 4) well-maintained facilities and logistics to deliver quality medicines and technologies.
Why Systems of Care Are Needed
Significant and important regional variations are found in process and outcome for a variety of cardiovascular conditions including cardiac arrest, STEMI, and heart failure. Moreover, patients with coronary artery disease (CAD) receive the recommended quality of care only 68% of the time. These differences in quality of care received mainly reflect disparities in access to or use of quality health care by demography or geography rather than differences in patient choice or risk. Some of these differences are associated with the characteristics of the hospital, such as urban location, teaching status, or safety net status. However, variations in outcome tend not to be associated with large differences in protocols or processes of care, such as the use of rapid response teams, hospitalists, clinical guidelines, and medication checks. Instead, hospitals in the top or bottom tier of risk-standardized mortality rates after myocardial infarction (MI) have been shown to differ substantially in terms of their organizational goals and values, senior management involvement, staff presence and expertise in care for patients with the condition of interest, communication and coordination among relevant groups, and problem solving and learning. A particular challenge in health care in the United States is the fragmented nature of our overall health care system. Overcoming this fragmentation is a theme that is the focus of much of health care reform, and it is a key element of the rationale for implementing and maintaining systems of care.
Interestingly, research showing that a given therapy is effective does not directly result in the use of that therapy in practice. Dissemination is the transfer of research results to decision makers to change the behavior of patients or providers so as to improve health. Effective implementation must include the identification of barriers to use of evidence and includes a strategy to actively overcome them. Dissemination and implementation interventions used to date have had variable degrees of success in various clinical conditions.
Patients who have an acute cardiovascular event in the out-of-hospital setting are transported to multiple physical locations for treatment delivered by diverse health care providers. These patients require time-sensitive interventions that must be continuously available so they can be quickly delivered to an eligible patient. Few hospitals are able to provide primary percutaneous coronary intervention (PCI) 24 hours a day, 7 days a week for STEMI or OHCA. In addition, hospital-based providers often infrequently treat patients who have an acute cardiovascular event such as OHCA—given the low rates of occurrence in areas with low population density, as well as low initial field resuscitation in their communities—so systems of care to address such events are rarely optimized. Identification that a patient has had an acute cardiovascular event outside of the hospital can improve both processes and outcomes of care by triaging a patient to a facility capable of providing quality care and by notifying the receiving facility immediately, so they can begin to prepare to provide timely care even before the patient has arrived.
Multiple examples may be found throughout the field of medicine of the positive correlation between greater provider experience or greater procedure volume and better patient outcome. These include the care of patients with conditions that require time-sensitive intervention, including in-hospital and out-of-hospital cardiac arrest, OHCA alone, and traumatic injury ; it also includes patients hospitalized with STEMI and those with STEMI who undergo primary angioplasty. However, the relationship between volume and outcome is a complex one. Procedural volume appears to be a surrogate marker for multiple patient, physician, and health care factors that have an impact on outcome but are difficult to quantify individually. Transferring patients from institutions with limited facilities to those able to provide time-sensitive interventions may have a salutary effect on outcomes by increasing the volume of patients treated by the receiving physicians and hospital.
Experience to Date with Cardiovascular Systems of Care
ST-Segment Elevation Myocardial Infarction
Survival following MI can be improved by reperfusion therapies that open the occluded infarct-related coronary artery. The earlier reperfusion is achieved, the greater the survival benefit. Moreover, reperfusion by primary PCI, as long as it is done in a timely fashioned by experienced centers, is more effective than fibrinolytic therapy to open occluded coronary arteries and improve survival. In some countries, time has been saved by providing fibrinolytic therapy in the ambulance, at least for patients who called emergency medical services (EMS) for assistance. Moreover, simply obtaining a 12-lead electrocardiogram (ECG) in the prehospital setting by EMS can result in an earlier diagnosis and faster treatment once the patient arrives at the hospital. Multiple randomized trials conducted mainly in Europe have demonstrated that outcome can be improved among patients who come to a non-PCI center in need of coronary intervention by transferring them to a PCI center for primary PCI, as long as it is done in an organized and rapid fashion.
Although only an estimated 1500 of more than 5000 acute care hospitals in the United States have primary PCI capability, the majority of the population lives within 60 minutes of a PCI-capable hospital, and many non-PCI hospitals are within a 30- to 60-minute transfer time to a PCI center. Between 30% and 50% of patients with STEMI arrive at hospitals via EMS, and traditionally patients have been taken to the nearest hospital irrespective of PCI capability. Therefore, coordination of how patients flow through EMS, non-PCI centers, and PCI centers is crucial to optimize delivery of efficient care ( Figure 5-1 ).
These developments—early fibrinolytic therapy, primary PCI, prehospital diagnosis, and interhospital transfer for primary PCI—established the need for regional systems of care to ensure that all eligible patients receive reperfusion in the form of primary PCI, if it can be done in a timely fashion by an experienced center. Opportunities to develop systems of care that address each element of the care process with multidisciplinary stakeholders were addressed in a conference organized by the American Heart Association (AHA) and summarized in manuscripts that included an executive summary. Development of regional systems of coordinated care between EMS and networks of hospitals has become a strong (class I) recommendation in the 2009 American College of Cardiology (ACC)/AHA STEMI guideline update. Several experiences, summarized below, illustrate opportunities to improve care of STEMI through systems development and improvement ( Table 5-1 ).
|STEMI Receiving Center Networks||EMS triage to PCI centers||10 U.S. regions||Prehospital ECGs, prehospital diagnosis, use prehospital diagnosis to activate the catheterization lab, and direct transport to primary PCI center||Linking EMS diagnosis and triage to PCI centers can result in excellent D2B and first medical contact to device times|
|Primary PCI Network||EMS triage to PCI center||University of Ottawa Heart Institute and surrounding EMS and hospitals||Prehospital ECGs, prehospital diagnosis, use prehospital diagnosis to activate the catheterization lab, and direct transport to primary PCI center||EMS triage to primary PCI center results in much faster overall care than interhospital transfer|
|MHI Level 1 Protocol||Regional transfer network||Minneapolis Heart Institute and 30 non-PCI centers within 210-mile radius||Non-PCI centers, ambulance and helicopter transport systems, primary PCI center||Transfer for primary PCI can be done in 90 to 100 minutes first-door-to-device in hospitals out to 60 miles, and within about 120 minutes (with initial treatment with half-dose fibrinolytic therapy) out to 210 miles, with excellent clinical outcomes.|
|Mayo Clinic Protocol||Regional transfer network||St. Mary’s Primary PCI Center and 28 non-PCI centers up to 150 miles away||Non-PCI centers, ambulance and helicopter transport systems, primary PCI center||Transfer first-door-to-device was achieved at a median of 116 minutes with initial use of fibrinolytic therapy for patients within 3 hours of symptom onset, with excellent outcomes|
|RACE program||Statewide STEMI systems||Five regions across North Carolina, including EMS, 10 primary PCI centers, 55 non-PCI centers||EMS, non-PCI centers, primary PCI centers||All times (including door to needle, D2B, transfer first door to device) substantially improved, with median of 106 minutes first door-to-device for centers routinely transferring|
Prehospital Diagnosis, Catheterization Laboratory Activation, and Transport to Primary Percutaneous Coronary Intervention Centers
Integrated EMS and hospital care enables treatment to be “moved forward,” such that the initial diagnosis and initiation of the reperfusion process can occur in the prehospital setting by trained EMS providers. Experience was reported in 10 regions, including Los Angeles, where prehospital ECG diagnosis was used to activate cardiac catheterization laboratories. Door-to-balloon (D2B) times were approximately 60 minutes, and first-medical-contact-to-device times averaged less than 90 minutes, an achievement that demonstrated what an integrated system can accomplish. A similar program in Ottawa, Canada, has shown impressive reductions in time to reperfusion with prehospital diagnosis and direct transfer to a PCI center.
Regional Transfer Protocols
Although randomized controlled trials (RCTs) have shown that transferring patients from non-PCI centers for primary PCI, rather than administering fibrinolytic therapy, can improve outcomes, it has been difficult to accomplish a door-to-device time of less than 120 minutes, as was achieved in the trials. Using standardized protocols with data collection and feedback, efficient transfer for primary PCI for hospitals in a 60-mile radius was achieved in the regions of Minneapolis and Olmsted County, Minnesota, with door-to-device times in the 90- to 100-minute range and excellent clinical outcomes. Similar protocols have been successfully implemented in other communities.
State Systems for ST-Segment Elevation Myocardial Infarction CARE
A statewide system in which each hospital (PCI and non-PCI) and EMS system has a standardized protocol for STEMI care has resulted in substantial improvements in care in North Carolina in the Reperfusion of Acute MI in Carolina Emergency Departments (RACE) program. The state used a single data collection tool (ACTION Registry—Get With The Guidelines) and common protocols, which have enabled communities to receive standardized, quality care. Support by professional societies, including the ACC, and sponsorship by the AHA have been powerful tools to coordinate care in communities with more than one primary PCI center, where competition has been a barrier to collaboration. Nearly 90% of patients brought by EMS to PCI centers have prehospital ECGs, and EMS frequently takes patients to the nearest PCI center, rather than to closer, non-PCI centers. As is happening in many STEMI systems of care, coordination of cardiac arrest care is being incorporated into the system.
Mission: Lifeline Program to Improve ST-Segment Elevation Myocardial Infarction CARE
Mission: Lifeline is an AHA program designed to improve STEMI outcomes through integration of regional systems of care (see www.heart.org/HEARTORG/HealthcareProfessional/Mission-Lifeline-Home-Page_UCM_305495_SubHomePage.jsp ). Using elements from the demonstration projects above, it has established criteria based on evidence of best care to guide development and improvement of STEMI systems of care. With nearly 600 systems registered in 2011, and more than half the U.S. population covered by Mission: Lifeline systems, this national program has become the standard for development of regional systems of STEMI care. Many hospitals now have received recognition for their accomplishments, and an accreditation plan is being implemented. The impact of participation in Mission: Lifeline on processes of care or patient outcome remains to be determined.
Compliance with recommended processes of care is low and variable among patients with heart failure in an outpatient setting. Hospitalization for heart failure is frequent, debilitating, and costly. Readmission after hospitalization for heart failure also occurs frequently. To date, no published evidence exists for or against the effectiveness of an interconnected care delivery system for patients with heart failure in a geographically contiguous region. But multiple randomized trials and other observational experiences have assessed interconnected strategies to reduce rates of initial admission or readmission for heart failure and to reduce mortality rates. Because of the heterogeneous nature of these interventions, it is infeasible either to pool the results of the trials or to tease out which component of the intervention was effective. Nonetheless, these trials demonstrated that interventions that use follow-up by a specialized multidisciplinary team in a clinic or nonclinic setting reduce heart failure hospitalizations, all-cause hospitalizations, and mortality rate. Interventions used enhanced patient self-care activities to reduce heart failure hospitalizations and all-cause hospitalizations but not mortality rate. Interventions that used automated telephone contact only, with advice to see a primary care physician in the event of deterioration, failed to reduce heart failure hospitalizations, all-cause hospitalizations, or mortality rate.
As of February 2011, a few regions of the United States have implemented cardiac resuscitation systems of care to improve outcomes after OHCA (e.g., Arizona, Maryland, parts of Minnesota, New York, Ohio, Texas, and Virginia). Those that exist usually developed ad hoc, without comprehensive evidence-based criteria, common standards, or dedicated reimbursement. Some institutions have designated themselves as resuscitation centers of excellence that may or may not be part of a regional system. Some regions have attempted to create a system of care by transporting patients resuscitated in the field from OHCA only to hospitals capable of inducing hypothermia, but other regions have been unable to do so. EMS providers and physicians have recently begun to work collaboratively in Arizona, Minnesota, North Carolina, Pennsylvania, and Washington to improve processes and outcomes after OHCA by sharing knowledge, tools, and techniques as part of the Heart Rescue Project. Some of these cardiac resuscitation systems link to STEMI systems of care. To date, there is no published evidence for or against the effectiveness of such programmatic interventions on the structure, process, or outcome of cardiac resuscitation, because regional systems of care for OHCA have not been evaluated formally. Therefore, we summarize evidence of the effectiveness of interconnected field and hospital-based interventions for OHCA.
Several cities have implemented multiple interconnected changes to EMS care for patients with OHCA, and most have reported improved outcomes compared with historical controls. Because of the observational nature of these studies, it is not possible to ascertain which component of the intervention was effective. Components that may be important include 1) emphasis on improved chest compressions; 2) reduced pauses for rhythm analysis; 3) use of a single, rather than stacked, shock sequence; or 4) use of devices intended to improve venous return.
Moreover, several groups have implemented multiple interconnected changes to hospital care for patients resuscitated from OHCA. All have reported improved outcomes compared with historical controls. Again, due to the observational nature of these studies, it is not possible to ascertain which component of the intervention was effective. Components that may be important include 1) delivery of therapeutic hypothermia to selected comatose patients, 2) coronary angiography when there is a high degree of suspicion of an acute ischemic trigger, 3) early hemodynamic stabilization of the patient with the ability to effectively treat rearrest, 4) reliable prognostication, and 5) cardiac electrophysiologic assessment and treatment before discharge.
If patients resuscitated from cardiac arrest are to be preferentially transported to designated receiving hospitals, an interesting issue is how outcomes are related to the distance or duration of transport. Multiple observational studies in the pre-hypothermia and post-hypothermia eras demonstrate that transport time to the hospital was not significantly associated with survival to hospital discharge after OHCA. Interpretation of some of these studies is limited by their high rate of missing transport time data or low overall survival. In another multicenter observational study of patients with OHCA in North America, survival to discharge tended to be lower among those taken to the closest hospital compared with those transported to distant hospitals. But this study did not measure which hospital-based interventions patients received, although the distant hospitals were more likely to have PCI facilities, electrophysiology laboratories, more beds, higher patient volumes, and teaching hospital status. Collectively these studies suggest that it is feasible to bypass a less capable hospital after the patient’s circulation has been restored. No current studies define a safe journey time, the use of different modes of transport, or the role of secondary transfer to a regional center after initial care at a local hospital.
Quality Improvement Theory
Systems-based interventions make extensive use of feedback to providers. The development of the theory of providing feedback to workers to improve process and outcomes is attributed to W.E. Deming, Joseph Juran, and Armand Feigenbum. The underlying principle is that individuals reflect on their performance to encourage change in their behavior or in the system. Small incremental changes are applied throughout the work process. A strong dose-response association has been found between adherence to guidelines or performance measures and outcomes.
Multiple methods are used to improve personnel performance, work processes, and products in business or health care settings. These include total quality management, reengineering, rightsizing, restructuring, cultural change, turnaround, disruption and lean management, and other methods. None of these methods is consistently better than the others, and each of these improvement processes goes through a series of phases that usually requires time to achieve the intended change in process and outcome. Skipping steps creates the illusion of speed, but it does not lead to a satisfying result. Mistakes in any of the phases can reduce impact, momentum, and hard-won gains. Measurement of processes of care is necessary but not sufficient to achieving improved outcomes.
Four key barriers stand in the way of implementing change to improve processes and outcomes in an organization. The first barrier is lack of understanding that change is needed. For EMS agencies and hospitals that treat patients who have acute cardiovascular events, this need for change is driven by the large regional, intrahospital, and interhospital disparity in outcomes. The second barrier is resource limitations, which force organizations to change resource allocations. The third barrier is a lack of desire among individuals to make changes. A final barrier can be institutional politics.
A tipping-point approach to implementing change can be considered. Initial efforts to change should focus on local opinion leaders who have a disproportionate influence in the organization. For EMS agencies, such a leader could be the medical director, shift supervisor, or person responsible for training or quality assurance. Once such an individual is committed to change, that person’s achievements should be highlighted to encourage others to change also. In the unlikely event that individuals are not committed to change, consideration can be given to reassigning their duties. Lecturing on the need for change is unlikely to succeed, so the organization should seek to continuously experience the realities that make change necessary. For organizations responsible for care of patients with acute cardiovascular events, this includes monitoring survival to discharge and performing functional measures before or after discharge (e.g., ejection fraction for patients with STEMI, Minnesota Living with Heart Failure questionnaire for patients with heart failure, neurologic status based on modified Rankin score for patients resuscitated from cardiac arrest). Resources can be redistributed from activities that are high effort and low yield to those that are low effort and high yield. For resuscitation organizations, this might include shifting away from training and equipping field providers to obtain intravenous access to training providers and the public to deliver effective chest compressions. For hospitals, this might include shifting away from routine use of pulmonary artery catheters to improving organizational culture or training providers to counsel patients on daily weight measurement and to see their primary care physician in the event of clinical deterioration. Each organization will have different activities that require redistribution. Finally, a resuscitation organization should appoint a mentor who is highly respected, knowledgeable about those who support change and those who resist it, and able to devise strategies and build the coalitions necessary for change. The mentor can advise the change leader of what is happening at lower levels of the organization.
Process improvements are unlikely to be sustained without evidence that outcomes are also improved. This is especially true when financial resources are involved. The duration of the measurement period will vary with the process variable. A reasonable approach is periodic examination of process data (e.g., quarterly) and outcomes data (e.g., annually). During the measurement period, relevant providers should be given timely feedback on both process variables and outcomes. If process variables were not affected by the intervention, efforts should be made to determine why, and alternative approaches should be identified and implemented. If the intervention was successful, and benchmarks were achieved, the next weakest link should be addressed. A structured approach can be used to identify which links are weak and what leverage points can be modified to address them. Collectively, application of these theories and methods can be used to achieve sustained and important improvements in the process, outcome, and quality of cardiovascular care.