Healthcare Quality Improvement: The Need for a Macro-Systems Approach

Abstract

While the structure of healthcare systems evolved out of the need to address acute conditions, the function of healthcare systems evolved to primarily address chronic conditions. The healthcare delivery system organically developed to respond to “one-off” acute illness or injury. Subsequently, healthcare delivery systems grew into legacy systems that evolved into complex systems over time. Healthcare delivery for acute conditions tends to utilize a specific part or form of the healthcare delivery system. In contrast, healthcare delivery for chronic conditions forces patients to seek care over time between different places or healthcare entities. Because of the self-contained structural organization of these healthcare delivery systems, they were not designed to provide coordinated, integrated, and longitudinal care over time and place. Consequently, today’s complex legacy healthcare delivery system requires significant improvement in the quality of care delivered to patients, especially those with chronic conditions. As a complex and legacy system, the most appropriate approach to improve the quality of delivered care is through a re-design quality improvement process, rather than a new system design process. In this paper, we describe the conceptual framework for quality improvement (QI) and the current micro and macro level approaches to quality improvement. We applied the current quality improvement approaches to the QI conceptual framework. We identified the limitations in current quality improvement processes in complex healthcare systems at the macro-level, pointing to the need for macro-systems approaches to healthcare quality improvement.

I. Introduction

While the structure of healthcare systems evolved out of the need to address acute conditions, the function of healthcare systems evolved to primarily address chronic conditions. From a structural perspective, the healthcare delivery system organically developed to respond to “one-off” acute illness or injury [1]. The focus of the system is on the urgency to diagnose and cure an individual’s acute (short-term) episode [2]. During the mid-19th century, the biomedical model of care viewed the body as a machine and disease as the breakdown in the machine [3]. This reductionist view of the physical body led to the organization of healthcare delivery systems by organ and organ systems into specialities and departments. However, advancement in science and medicine has led to a shift in the needs of the population towards chronic (long-term) care. This chronic disease burden is growing and the current healthcare delivery system is facing an unprecedented chronic disease burden. Healthcare delivery systems developed into decentralized islands that treat patients with chronic conditions by passing them along between these islands as healthcare providers independently deliver care for each disease or condition they address [4]. In 2010, chronic conditions accounted for 86% of all healthcare spending [5], 81% of all hospital admissions, 91% of all prescriptions filled; and 76% of all physician visits in the United States [6]. Consequently, healthcare systems evolved from entities that primarily treat trauma or infection, for patients with acute conditions, to entities that deliver preventative, curative, palliative, and end-of-life care for patients with chronic conditions.

Subsequently, healthcare delivery systems grew into large legacy systems that evolved into complex systems over time [7]–[10]. Healthcare delivery for acute conditions tends to utilize a specific part or form of the healthcare delivery system, such as the emergency department, or an outpatient clinic. In contrast, healthcare delivery for chronic conditions forces patients to seek care over time between different places or healthcare entities, such as departments, hospitals, clinics, centers, etc. Because of the self-contained structural organization of these healthcare systems, they were not designed to provide coordinated, integrated, and longitudinal care over time and place. Consequently, intricate changes in the functional utilization of the healthcare delivery system has led to complexities and emergent behaviors. Unfortunately, patients are now left to manage their continuity of care. This is a problem because the lack of coordination can lead to safety concerns from disparate pharmaceutical prescribing. and segregated models of care also lead to loss of critical information and gaps of service. Consequently, healthcare delivery systems today have been accepted as complex system [7]–[10]. Here, we define complex systems as systems composed of multiple interacting components that produce dynamic and emergent behaviors that cannot be understood by independently examining the individual components alone [11]. Furthermore, given the cost of the hard and soft infrastructure of the healthcare delivery system, it continues as an expensive legacy system. Here, we use the term legacy to describe a system composed of very expensive technical resources, such as the distributed soft infrastructures of hospitals and clinics, and human resources, such as highly trained providers and specialists that can require over a decade of training.

Consequently, today’s complex legacy healthcare delivery system requires significant improvement in the quality of care delivered to patients, especially those with chronic conditions. Over the last several decades, authoritative reports by the United States Institute of Medicine and others have pointed to the critical importance of improving the quality of the care provided to patients [12]. This is because despite significant advancements in medicine, it is only the delivery of these advancements to patients that dictates the level of positive impact on patient health outcomes [13]. The quality of delivered care requires improvement for several reasons. To name a few, these include safety issues, lack of evidence-based medicine, coordination, and lack of care that is concordant with patient preferences. These issues are compounded when multiple systems and sectors interact, as is common in chronic care delivery [14], [15]. The Institute of Medicine (IOM) defined quality as “the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge.” [16] We describe quality following the United States Institute of Medicine’s six domains of quality [16], where quality is expected to be safe, effective, patient-centered, timely, efficient, and equitable, see Figure 1.

Fig. 1.

The six domains of healthcare quality described by the United States Institute of Medicine “Crossing the Quality Chasm” Report [16].

As a complex and legacy system, the most appropriate approach to improve the quality of delivered care is through a re-design quality improvement process, rather than a new system design process [17]. Healthcare delivery systems are socio-technical systems that include expensive human and technical resources that are part of a large and expensive soft and hard infrastructure. Similar to other complex and legacy systems, such as the transportation or energy systems, the design literature realistically describes changes to these systems through a re-design process, rather than a process that removes the current systems for the sake of a completely new system design process [18]–[20].

A. Novel Contribution

In this paper, we describe a conceptual framework for quality improvement (QI). We also detail the current micro and macro level approaches to quality improvement from healthcare, engineering, and business. We then apply current quality improvement approaches to the QI conceptual framework to identify the advantages and limitations in current quality improvement processes. In particular, we use these limitations to identify the needs for a macro-systems approach to healthcare quality improvement that can satisfy the QI conceptual framework.

B. Paper Outline

First, we detail the underlying conceptual framework for quality improvement in Section II. Next, we describe the current approaches to quality improvement from healthcare, engineering, and business, at the micro-level in Section III and the macro-level in Section IV. We apply the current quality improvement approaches to the conceptual framework to examine the advantages and limitations of these approaches for micro-level QI in Section V and for macro-level QI in Section VI. In Section VII, we present the need for a macro-systems approach to quality improvement to address the identified limitations in Section VI. Finally, we conclude in Section VIII.

II. A Conceptual Framework for Quality Improvement

Quality improvement is a process of deciding what and how to improve the quality of a system, product, or service. Quality has become a ubiquitous term. The Oxford Languages defines quality as “the standard of something as measured against other things of a similar kind; the degree of excellence of something” or “a distinctive attribute or characteristic possessed by someone or something.” The field of systems engineering defines quality attributes as non-functional requirements used to evaluate the performance of a system [21]. These quality attributes are generally described as the ‘ilities’ [21]. In healthcare, the most accepted quality attributes are described by the Institute of Medicine by six domains of safety, effectiveness, patient-centeredness, timeliness, efficiency, and equitability [16].

At its core, quality improvement stems from the field of quality engineering; a discipline of engineering concerned with the principles and practice of product and service quality assurance and control [22]–[24]. The conceptual model for quality control in engineering stems from control systems. The practice uses sensors and detectors to measure the output performance of a system being controlled; these measurements are fed into a controller to make decisions about the need for corrective action and to issue the appropriate change signals to provide corrective feedback to implement a change that achieves the desired performance. This control-feedback loop is abstracted and serves as the basis for the conceptual framework. Figure 2 visualizes the conceptual framework, which can be described by the following steps.

Fig. 2.

Quality Improvement Conceptual Model. A system provides output information, y, as input to the (1) Measurement Function, H, to produce quality information, z, as input to the (2) Decision Function, G, to produce the change ideas, u, that then serve as input to the (3) Implementation Function, F.

1. A system outputs information (y) to be used as input to the Measurement Function. The Measurement Function applies a measure function, H to the output information, y, to produce the quality information, z. Therefore, H(y)=z.

2. The Decision Function applies a decision function, G, to the quality information, z, to produce a set of change ideas, u. Therefore, G(z)=u.

3. The Implementation Function applies an implement function, F, based on the change ideas, u, to produce new system output information, y. Therefore, F(u)=y.

Interestingly, in healthcare, quality improvement conceptual frameworks describe quality improvement as an intervention, based on improvement science [25], [26]. These frameworks focus on the deeper levels of barriers and facilitators of change and change management. In other words, these frameworks focus on the actions needed to implement the change ideas, which leads to a focus on the “socio” side of a socio-technical system. This focus stems from the widely-held definition of quality improvement in healthcare: “Quality improvement (QI) consists of systematic and continuous actions that lead to measurable improvement in health care services and the health status of targeted patient groups.” [27]

In the next sections, we detail the quality improvement approaches from healthcare, engineering, and business at both the micro and macro levels.

III. Micro-Level Approaches to Healthcare Quality Improvement

It is worth noting that the phrase healthcare delivery system can be used to describe any level of a system, given a system boundary. In this paper, we use the phrase healthcare delivery system to denote large hospitals comprised of many departments, large centers of multiple hospitals, or sets of hospitals and clinics that make up larger groups. Examples of these healthcare systems in the US include Kaiser Permanente, or the Mayo Clinic Health System. In this section, micro-level quality improvement approaches refer to efforts focused on a single clinic or department or a very small tractable set of clinics, as a much smaller and more bounded part of a healthcare delivery system. For example, the emergency department (ED) and intensive care unit (ICU) of a hospital.

The most successful quality improvement efforts have been at the micro-level. At a small-scale, these efforts tend to address acute episodes, where patient health is recovered predominantly during a visit. For example, an emergency room provider setting a dislocated shoulder. Micro-level quality improvement efforts to improve healthcare quality exist across the fields of healthcare, engineering, and business. We briefly highlight these efforts. At the micro-level, many of the methods used in Healthcare have roots from Engineering and Business.

A. Micro-Level Healthcare Quality Improvement Approaches

Prior to 1960, the history of quality improvement in healthcare can be described by a disparate set of unrelated events rather than an organized effort [28]. In brief, this history includes clinically focused efforts for improvement by Florence Nightingale [29], Clara Barton [30], Abraham Flexner [31], Ernest Codman [32], and several others [28], [33].

However, the history of quality improvement typically begins with a mention of Walter Shewhart. He developed statistical process control (SPC) charts in 1931. [34] He also believed that management and production practices need to be continuously evaluated and adopted or rejected. Shewart was considered the grandfather of the quality movement, while his understudies W. Edwards Deming and Joseph Juran are sometimes referred to as the fathers of the quality movement.

W. Edward Deming was an American engineer and statistician whose work was first embraced in the 1950s in Japan [21]. The wide application of his techniques led to unprecedented quality and productivity levels, which lowered costs and boosted global demand for Japanese products. More than two decades later, his work would become more recognized in the United States. He developed The Deming cycle of improvement, or Deming wheel (plan-do-check-act [PDCA] or plando-study-act [PDSA]), which was adapted from Shewhart’s work [35].

Joseph Juran’s primary contribution to quality was his adaptation of the Pareto principle to quality. According to his principle, 80 percent of defects are caused by 20 percent of problems, and quality improvement should therefore focus on the 20 percent to gain the most benefit. Also, Juran’s Quality Handbook [22], first published in 1951, remains a standard reference for quality today. Juran’s (1986) quality trilogy [36] also informed the roots of quality programs, such as Six Sigma.

The systematic approach introduced by Deming in Japan became known as Total Quality Management (TQM). The TQM methodology relies on the plan-do-check-act (PDCA) cycle to manage processes and, when problems arise, statistical tools to solve them. In the early 1980s, Drs. Berwick and Batalden, prominent champions of TQM, translated the principles into terms more familiar to healthcare and first coined the term “Continuous Quality Improvement” (CQI), which is composed of several QI tools including Cause and Effect Diagrams, Failure Modes and Effects Analysis, Run Charts and Control Charts, and Plan-Do-Study-Act rapid-cycle testing [37]. Their work led to the formation of the Institute for Healthcare Improvement (IHI) in the 1990s. Dr. Batalden was the founding Chairman of the Board of IHI and Berwick the President and CEO. The IHI remains a leading nonprofit organization in encouraging healthcare improvement. Quality improvement continues to be adapted, and this type of QI is often referred to as clinical microsystems approach to quality improvement [38]. As the name suggests, this body of quality improvement focuses on the micro-level of a system.

B. Micro-Level Engineering and Business Quality Improvement Approaches applied to Healthcare

While much of the implemented quality improvement efforts come from within healthcare, most quality improvement approaches originated predominantly from other fields such as industrial engineering and systems engineering, and operations research from engineering and business. Shewart, Deming, and Juran’s work paved the way for the development of the previously mentioned total quality management (TQM), coined by the U.S. Navy and defined mainly by academics. The theories and tools of TQM were codified for implementation into a program called Six Sigma, which specifically defines a metric or goal of less than 3.4 defects per million. Six Sigma also developed certification programs and was based mainly in industry. Other strategies evolved from these to create the Just-in-Time (JIT) inventory management strategy, which lead to Lean production to eliminate waste. All the above are programs or frameworks for performance improvement, and each has a slightly different focus, tools, and techniques associated with it. At its core, the focus is on change management at the micro-level of a system.

More recently, a more quantified approach to improve management practice focus on patient flow logistics, capacity planning and appointment scheduling. With a focus on the system rather than the parts, the field of industrial engineering and operations research (IEOR) utilizes mathematical modeling approaches of discrete-event simulations, and linear programming [39] to improve healthcare performance measures, such as reducing patient wait times in the emergency room.

IV. Macro-Level Approaches to Healthcare Quality Improvement

Macro-level quality improvement approaches describe large healthcare system attempts to improve care. It is important to improve quality at this level because the macro-level is much more likely to capture quality of care for patients with chronic conditions across departments and care settings.

A. Macro-Level Healthcare Quality Improvement Approaches

While micro-level quality improvement follows a bottom-up approach, macro-level improvement follows a top-down approach. While attempts to improve healthcare quality date back to the late 1800s, a quality improvement movement to audit large amounts of healthcare data stemmed out of the introduction of the United States federal health insurance program for the aged and disabled in the 1960s. In 1965, the US Congress passed legislation to establish the Medicare and the Medicaid programs [40]. To assess and direct care of these programs, the US Congress established a set of conditions entitled “Conditions of Participation,” which required staff credentials, 24-hour nursing services, and utilization review [41]. Utilization Review Committees were established as a form of effective monitoring using administrative claims data; yet, success was limited. “The lack of effectiveness was retrospectively attributed to an absent association between the review process and the identification of ways to improve care.” [42]

Efforts of utilization review spawned significant quality measurement efforts. During this time, Avedis Donabedian, often considered the father of quality measurement, described quality in terms of a model that relies upon the elements of structure, process, and outcomes to examine the quality of care delivered. He defined the structures of health care as the physical and organizational aspects of care settings (e.g., facilities, equipment, personnel, operational and financial processes supporting medical care, etc.). The processes of patient care sit in the middle of the diagram because they rely on the structures to provide resources and mechanisms for participants to carry out patient care activities. Finally, processes are performed to improve patient health in terms of promoting recovery, functional restoration, survival and even patient satisfaction, or, in other words, outcomes [43]. It is worth a pause to note the similarity of Donabedian’s model elements to those found in systems architecture. Elements of structure correspond to elements of system form, while elements of process. Similarly, system resources perform system processes for the patient operand to improve patient health, which can be used as a performance measure.

Notable efforts to improve quality continued and include the 1994 National Surgery Quality Improvement Project (NSQIP), which developed a program of data collection of both risk adjustment and outcome data, to benchmark care facilities within [44]. In addition, in 1999, the National Quality Forum (NQF), a nonprofit organization, was established with a mission to improve the quality of US healthcare. NQF endorsement became the “gold standard” for healthcare performance measures [45]. Newer efforts, such as the Quality Reporting and Hospital Value-Based Purchasing program by CMS, incentivize quality improvement by combining quality reporting with hospital value-based purchasing. Hospitals would be able to earn incentive payments either based on their performance or on their performance improvement relative to their starting baseline. Programs, such as those mentioned above, continue to emerge in further efforts to improve quality. Clearly, these efforts focus on a macro-level of a system.

B. Macro-Level Engineering and Business Quality Improvement Approaches

Efforts for quality improvement from the engineering and business fields began with micro-level bottom-up approaches. Very few non-healthcare fields, if any, incorporate an audit of services provided, described above for healthcare. The closest need and efforts for standardization across the nation are undertaken through the development of standards, primarily by the International Organization for Standardization (ISO).

Macro-level QI approaches are far less extensive or developed than micro-level QI approaches. Extending engineering and business quality improvement into the macro-level is typically viewed as intractable because of the loss of detailed data and the loss of qualitative observable organizational or process information [46]. However, a model-based systems engineering approach to modeling healthcare delivery utilization across many settings has been proposed [47]–[49].

V. Applying Micro-Level Approaches to the QI Conceptual Framework

While the names of micro-level frameworks may differ between healthcare and engineering and business, these QI approaches are actually very similar and stem from the same roots and foundations. When applying the QI conceptual framework, from Section II, to micro-level quality improvement, all three steps can be matched to various aspects of micro-level QI approaches.

First, the Measurement Function of the conceptual framework aligns closely with the concepts of statistical process control and run, and control charts. It is clear that measurement is a key aspect of micro-level quality improvement.

Second, the Decision Function of the conceptual framework aligns closely with the plan and study parts of the plan-do-study-act cycle. In the plan part of the study, the decision function is closely tied to the concept of which measures are chosen to support a specific aspect of quality to improve or a specific aspect of performance to understand. In the study part of the cycle, the decision function supports comparing results to predictions and ideating over the learnings.

Third, the Implementation Function of the conceptual framework aligns well with the do and act part of the plan-do-study-act cycle. The do part incorporates implementing the change idea, while the act part also incorporates implementing changes. In addition, the implementation function is closely tied to the ability to change. Organizational concepts of management, practices, and leader champions are also critical to implementing changes.

Therefore, applying micro-level QI approaches clearly fulfill all three steps of the QI conceptual framework. Consequently, it is not surprising that some of the most successful healthcare QI efforts occur at the micro-level.

VI. Applying Macro-Level Approaches to the QI Conceptual Framework

Given the significant difference between macro-level QI approaches from Healthcare and those found in Engineering and Business, we will apply the QI conceptual framework to each field separately.

A. Macro-Level Healthcare

Applying the QI conceptual framework to macro-level QI, highlights significant gaps in all three steps. In Figure 3, we briefly describe the limitations at each step. Each of these three limitations are described in turn.

Fig. 3.

Macro-Level Quality Improvement approach applied to Quality Improvement Conceptual Model. (1) Hospital system provides and, which generates claims; (2) Measurement function measures healthcare delivery from claims and produces quality measures; and (3) Hospital decision-makers attempt to use quality measures to make decisions on actions to take to improve care quality and quality measure values.

First, the Measurement Function applied to the macro-level QI produces aggregate and static quality measures, z, as values. More specifically, the measure function applies aggregated descriptive statistical methods to produce many healthcare-delivery and quality measures for individual system components. Such a strategy assumes the healthcare system is a simple system, where the overall quality of care can be elucidated from the subset of individual components. On the contrary, as we described in the introduction, healthcare delivery is complex because the nonlinear interactions of its components produce an output that is greater than the sum of its parts [50]. Thus, the healthcare delivery system produces dynamic and emergent behaviors that no subset of the elements have, and therefore, cannot be understood by independently examining the individual components alone. Thus, the current quantitative approach is mismatched to the system complexity level found at the macro-level [11], [51]. Consequently, this problem leads to the first limitation. Limitation #1. The macro-level QI measurement approach is static and aggregated and, therefore, mismatched to the dynamic and complex healthcare delivery system.

To highlight this limitation, we describe a palliative care example detailing emergent behaviors in the healthcare delivery system. The addition of a palliative care provider to a cancer team can shift a patient’s care trajectory or may prime the care system to behave differently. More specifically, during a palliative care visit, a patient may decide that hospice care is more aligned with their wishes, and shift their course of treatment to avoid life-sustaining therapies, such as intubation. In a less direct manner and depending on the level of integration of palliative care into the healthcare system, the existence of palliative care may prime an intensive care unit (ICU) provider to refer a patient to palliative care. In doing so, the patient may elect hospice care before the need for aggressive treatment arises. These scenarios and many others, affected by the level of integration of palliative care, collectively lead to patients being referred to hospice at different times and settings during their care trajectory. Thus, the behavior of referral to hospice is an emergent property of the full system and cannot be deduced by the existence of palliative care or ICU intensity measures.

Second, the Decision Function applied to macro-level QI produces no clear way to combine the plethora of independent quality values to support a decision of changes needed. Instead, the sheer number of measures has led to measure fatigue [52], [53]. While quality measures indicate quality issues; they do not provide actionable information as to how, what, and where to make changes and for which patients [52], [54]–[57]. The literature acknowledges that scalar measures are difficult to combine to deduce the behavior of a system [58] and consequently cannot provide information on the proper levers to change this behavior. This is because the aggregation process inherently leads to loss of critical information to support the decision of what changes may be needed. Instead, decision-makers are left to exert cognitive and operational effort to extract other new information directly from their system to decide on the changes needed. Consequently, this problem leads to the second limitation. Limitation #2. The macro-level QI measurement approach produces quality measures that do not provide sufficient detail to serve as actionable feedback to the decision function.

By way of an example, this is akin to capturing the dynamic heart rate signal by aggregating, in time, values for the different parts of the heart–left and right atrium and ventricles–and asking a cardiologist to provide a treatment solution without having the constructed dynamic heart rate signal. The founder of the study of healthcare quality, Dr. Donabedian, himself foreshadowed these limitations when he wrote: “The use of simple indices in lieu of more complex measures may be justified by demonstrating high correlations among them. But, in the absence of demonstrated causal links, this may be an unsure foundation upon which to build.” [59]

Third, the Implementation Function lacks the clear change ideas as input to this function. Furthermore, in some instances, on-the-ground QI human resources that understand the organizational processes and have buy-in from front-line providers may not exist to identify and execute an implementation plan. In the case that QI teams do exist, they do not have the change ideas they need, or the data sources they typically would use to make the decision to produce these change ideas. Instead, the organization may decide on a set of quality measures to address, one measure at a time, by now applying micro-level QI approaches. Limitation #3. The macro-level QI measurement approach does not provide sufficient detail to provide information as to what changes to implement.

For example, in an effort to provide patient-concordant wishes of dying at home, hospitals have made changes to healthcare delivery, which resulted in a smaller proportion of decedents dying in hospitals [60]. Yet, the underlying actions taken affected other parts of the system and led to unintended increases in intensive care and care transitions [60].

Overall, macro-level healthcare approaches essential fail at fulfilling the QI Conceptual Framework at each of the steps. It is therefore not surprising that there is minimal empirical evidence to support that the macro-level QI approach has triggered meaningful improvements in healthcare delivery systems [53]–[55].

B. Macro-Level Engineering and Business

With a focus on the system rather than the parts, the fields of Systems Engineering, Industrial Engineering and Operations Research, utilize approaches to elucidate dynamic system behavior. In particular, these fields utilize discrete-event simulations to understand the dynamic behavior of healthcare systems because healthcare delivery is a discrete-event system. While, the medical field is more accustomed to natural continuous-time systems that produce continuous-time varying signals such as heart rate or sleep-cycles, many built systems, such as healthcare delivery, are discrete-event systems characterized by discrete, not continuous, values. For example, healthcare visits are in discrete numbers 1, 2, but cannot be 1.5. Additionally, the state of a discrete-event system changes based on events and not time. For instance, healthcare services are rendered only in the event that a patient enters the system and not because a specific amount of time has passed. While discrete-event simulation has been applied at a micro-level to address primarily short-term acute episodes, it has not been deemed intractable to address long-term chronic episodes [46].

Consequently, extending micro-level approaches to the macro-level are limited both technically and conceptually. From a technical perspective, the mathematical formulations used to analyze the acute care setting fundamentally focus on transportation and optimize the system by identifying delays related to transporting a set of patients through the healthcare delivery system as they receive care, just as you would transport widgets through a manufacturing system as they are built up. On the contrary, transportation is not a fundamental issue in chronic care, thus rendering a transportation formulation unusable. From a conceptual perspective, moving patients faster through an acute visit does not fundamentally examine nor address the longitudinal care for patients with chronic disease. Many have described the value of being able to describe a discrete-event simulation using macro-level administrative claims data for chronic care, [61], however, very few have been able to formulate methods to utilize administrative claims “big data” to dynamically model chronic care delivery. [49]. Consequently, this problem leads to the fourth limitation. Limitation #4. Traditional approaches to understand the dynamic behavior of a system apply to acute care, but such approaches are not easily formulated for chronic care. Therefore, there is no large evidence base of macro-level engineering and business quality approaches to apply to the QI conceptual model.

VII. The Need for a Macro-Level Systems Approach to Quality Improvement

Healthcare has invested significantly into the macro-level approach to quality improvement. Yet, the field recognizes that a fundamental change in the approach to quality measurement is needed [62], if this approach is to support quality improvement, and actually lead to improved quality. At the same time, there is also recognition that, conceptually, many of the quality challenges that confront healthcare need to be solved at the level of entire systems [14], [63]. Furthermore, it has been established that multiple ill-coordinated micro-level QI projects may, accordingly, degrade rather than improve the ability to achieve improvements across healthcare as a whole [63]. The recognition of the importance of addressing the whole system stems back to statements made by W. Edward Deming when he said: “Management’s job is to optimize the whole system.” [64]

Therefore, there needs to be an approach to macro-level QI that addresses the four limitations identified in the previous section. The underlying difficulty arises from the healthcare QI approach of quantifying a complex healthcare system using simple singular measurements of the parts (Limitations 1–3), and the intractable engineering and business QI approach to modeling and computationally formulating a discrete-event dynamic healthcare system model at the macro-system level (Limitation 4). A model-based systems engineering approach to modeling healthcare delivery utilization across many settings has been described [47]–[49]. Its systems engineering modeling efforts have elements that address Limitations 1–3, and the quantification of the model through discrete-event simulation has elements that address Limitation 4.

Consequently, there is a clear need to incorporate systems thinking, systems architecture, and systems engineering to develop a measurement strategy that can develop a macro-level system model that can capture emergent behaviors in the complex healthcare delivery system. The healthcare literature has recognized the value of systems theory and complexity theory to improve healthcare systems [7], [65]–[69]. There is also the need to quantify these system models using dynamics [70] that can describe the discrete-event based healthcare system [71].

VIII. Conclusion

In conclusion, this paper described a conceptual framework for quality improvement and detailed the current micro- and macro-level approaches to quality improvement from healthcare, engineering, and business. We applied these current quality improvement approaches to the conceptual framework and identified the limitations of current quality improvement approaches for macro-systems across healthcare, engineering, and business. We then highlighted the need for systems modeling and quantification to support macro-systems healthcare quality improvement.