Over the past decade, quality improvement (QI) projects have become. mainstay of neonatal care, with an increasing number of publications devoted to this topic. Due to. long-standing history of participation in data collection, data benchmarking, and collaborative learning opportunities, neonatal intensive care units (NICUs) are well positioned for QI projects [1, 2].

Whereas most healthcare practitioners are familiar with evidence-based medicine and randomized controlled trials, historically less attention has been paid to the measure of processes and their role in day-to-day activities in medical care. Borrowing liberally from manufacturing and other industries, medical professionals have started to embrace the theory and methodology of QI science by applying this knowledge to their daily work in improving the care of infants and families in their respective NICUs.

Model for Improvement

Using the model developed by Langley, Nolan, and others [3], the Institute for Healthcare has proposed the Model for Improvement,. tool frequently utilized as. template for quality improvement projects in diverse medical settings. This model’s approach promotes rapid testing for rapid learning when applied to most processes in health care. This Model for Improvement, shown in Figure 1, has been adopted by healthcare systems to improve both processes and outcomes and can be used in. variety of settings in both small and large scale projects.

Fig. 1
figure 1

The Model for Improvement is. structural framework that teams utilize to improve different health care processes. Its simple approach allows for rapid tests of change and significant improvement in care delivery and outcomes.

Full size image

The model that Nolan developed is divided into three questions and. fourth section which is. tool for testing changes. The initial question asks about the aim of the project (What are we trying to accomplish?). This aim must be specific in scope as the more specific the aim is, the more likely improvement is to occur. An important action in QI projects is to reiterate the aim frequently to ensure that the multidisciplinary team members are striving for the same goal, thus avoiding drift from the original aim. Aim drift occurs when the team intentionally or unintentionally strays from the stated goal of. project by changing the outcome measure; for example reducing the goal from. 50% reduction in. specific outcome measure to. lower percentage (20–30% reduction).

Development of an aim statement allows the team to establish. structured approach and. common mental model for designing. work plan by systematically measuring performance and identifying opportunities for improvement, while communicating results of current progress to stakeholders.. useful tool for establishing an aim statement is by utilizing the SMART (Specific, Measureable, Achievable, Realistic and Timely) acronym [4]. This tool allows for the aim statement to be concise in scope and provides. system for tracking of progress during the QI project.

Nolan’s second question (How will. know if. change leads to an improvement?) details the important step of measurement. Without measurement, it is difficult to have improvement in. process, and planning for specific metrics in this step is paramount to the success of the quality improvement project. Measurement can be. very fulfilling step of. QI project as it allows for data on. specific project to be collected, collated and presented back to the team. Measurement does not need to be complex or cumbersome and can be as simple as. handwritten chart. The most important facet of measurement collection is having contemporaneously collected data available for review by the team, allowing feedback about the new change that was tested.

As the measurement data is assessed, it leads to Nolan’s final question (What changes could we make that we think will result in improvement?). This is the main question at the heart of quality improvement; can the team make. change in. process to create an advantageous result? Since most team members have. basic understanding of how the system works, it is important that the team theorizes about which changes can lead to improvement. This is. similar concept to the idea of. hypothesis in the design of. scientific experiment as this is. wise guess by the team, using evidence-based approaches, to produce. system change that leads to. desired improvement.

The PDSA (Plan-Do-Study-Act) cycle is. tool (Fig. 1) that allows teams to rapidly test process changes in. clinical setting. Although the teams start with their plan and end with their action, the PDSA cycles are. not. singular exercise, but rather. continuous circuit of sequential cycles that involve rapid tests of change to refine. process. This cycle allows for development of. change concept (Plan), implementation of the new process or test in the desired setting (Do), followed by monitoring and data collection of results (Study), and finally final step of integrating the learning generated, leading to acceptance or rejection of the change concept that was introduced (Act) [5].

The PDSA cycle enables teams to rapidly implement changes, measure and learn from the results and then decide if the introduced process change should be accepted or discarded. Each PDSA cycle results in new knowledge about the process change, allowing for refined predictions about the change which is being tested. The accumulated knowledge from these tests of change may lead to the overall goal of system change, therefore obtaining the desired outcome of the quality improvement initiative.

Initially, the planning stage is often the most labor intensive and time consuming component of. new PDSA cycle. In subsequent iterations of the PDSA cycle, the planning stage may take less and less time as only small adjustments may be required from data gathered in previous cycles. The second step of the cycle (Do) is the experimental phase as proposed changes are now implemented in clinical settings and those results are measured and collected for the next phase in the cycle (Study). The study section concentrated on the gathering of knowledge or data about the process and does not require. multitude of data points, as learning may occur rapidly with very few cases. Once learning is done with the contribution of the data collected, the team then analyzes the results, leading to the next (Act) step in the cycle. This step is not. discrete end to the process, but. continuous step that allows the teams to use the previous PDSA cycle to accept or reject the proposed change with the idea that multiple PDSA cycles may be required to further refine. plan. The goal of PDSA cycle is to continue to retest the changes and collect more data to reach the final goal of. sustainable system change resulting in the desired outcome.

Even with this structured approach, use of the PDSA cycle does not guarantee success with. quality improvement project. Common errors which occur with PDSA cycles include lack of detail in planning, failure to predict PDSA results beforehand, lack of plan for data collection, not engaging team members in data analysis and failure to use PDSA findings to plan. new or follow up test. The PDSA cycle must not be thought of. singular experience, but similar to. set of stepping stones to cross. body of water, multiple PDSA cycles are needed for teams to reach their goals [6, 7].

Certain quality improvement tools, adapted from the manufacturing industry have started to be used more and more frequently to improve health care. The tools are simple to use with graphical interfaces that make the collected data easy to interpret and have been shown to be very useful in solving problems in healthcare delivery processes.

Process mapping/flow charts

Process mapping, also known as. flow chart, is. pictorial demonstration of the sequential steps involved in. process [8, 9]. It may be. simple or complex tool, as each and every step of. procedure needed to deliver care (e.g., which steps are involved when an infant receives surfactant therapy) is documented. Simple maps, also known as high level process maps, show how the process works in just. few steps. The purpose is to provide quick and easy insights into what the process does, without getting into the details of how it’s done. This simple map can be useful when communicating. process to leadership, but it does not encompass the details of each process step. Complex process maps include significantly more detail than simple maps and may be required if the team is planning on streamlining an entire comprehensive process. Creation of. complex map requires more team members and front-line participants in order to record the required details of each discrete step.

In order to be useful, process maps should not be created by. single individual, but rather by input of. multidisciplinary team. The various members of the healthcare team provide an understanding of how each step may be influenced by the preceding or subsequent steps in. process. By using process maps, the team is able to visualize how each member performs. step in. certain procedure, allowing for an improved understanding of each other’s roles and responsibilities during clinical situations. The value of the map created in. multidisciplinary meeting allows for discussion, and more importantly, understanding and appreciation of each other’s roles in discrete steps of the process. This establishes. baseline knowledge of the process for all team members and once completed, the team proceeds with identification of steps that are wasteful and do not add value to the overall process. The identification of important steps that add value to. process and the elimination of steps that do not add value are tenets of Lean management. Lean management is. systematic method for removing waste from processes without impacting productivity. Although Lean management has roots in manufacturing, it has been rapidly adopted by healthcare systems [10].

Creation of. process map does not need to be elegant or sophisticated; most successful process maps are created by pieces of paper representing each step that are then attached to the walls of. meeting room (Fig. 2). The use of this tool allows participants to discover that some of the steps in the procedure may involve delays in care, unnecessary/repetitive work or communication errors; all steps that add little value and may be able to be modified or even eliminated. Further refinement of the process by the team can serve as. basis for designing. new and improved process.

Fig. 2
figure 2

Process maps are. graphic display of steps, events and operations that constitute. specific process. This map demonstrates the 16 discrete steps that an improvement team outlined in the process of administering surfactant to. preterm infant with respiratory distress. The red circles denote steps in which surfactant may be administered during this process.

Full size image

Pareto chart

The Pareto principle was initially described by management consultant Joseph Juran as he observed that for many events, approximately 80% of the effects come from 20% of the causes [11]. It was named for Italian economist Vilfredo Pareto, based upon his observation that 85% of the overall wealth in Milan was concentrated in 15% of the population. This was then adopted by accident prevention practitioners as hazards could be addressed in. systematic order as targeted interventions to eliminate the more common causes of injury will be more successful than randomly targeted interventions.

Juran adapted this principle of separating out the vital few causes of an event from the trivial many, as. majority of organizational effects resulted from just. few causes. This became the basis for the Pareto chart (Fig. 3),. bar graph in which causative factors for defects in. process are placed in descending order of frequency, allowing for the team to concentrate their efforts on the factors that have the greatest impact. The horizontal axis of the chart contains the categories of the problem identified and the vertical axis contains the frequency of the measurement. Simply,. large proportion of quality problems are often created by. small number of causes, allowing for. more focused approach to prioritize one’s quality improvement efforts. Focusing on the most common issues allows the team to achieve the greatest improvement [12].

Fig. 3
figure 3

Pareto chart. http://www.cec.health.nsw.gov.au/quality-improvement/improvement-academy/qualityimprovement-tools/pareto-charts

Full size image

Ishikawa/fishbone diagram

Another important tool in quality improvement is the Ishikawa, or fishbone diagram (Fig. 4). These diagrams, created by Kaoru Ishikawa, are. visual representation of the many causes for. specific problem or effect, with the design of the diagram representing the skeleton of. fish. This tool categorizes potential causes of. problem in successive detail in order to identify the root cause [13]. The fishbone diagrams are typically constructed from the right side to the left, with each large “rib” of the fish branching out to include smaller bones containing more details.

Fig. 4
figure 4

An Ishikawa or Fishbone diagram demonstrates the relationships between potental causes of. presenting problem. This diagram details all the potential causes that the team identified that resulted in delays of bloodwork values in. hospital setting.

Full size image

Construction of the diagram starts with the problem statement, or effect, at the head of the “fish” and the categories of the major causes as the “ribs” of the fish attached to the backbone,. horizontal line drawn from the problem statement. The “rib” portion of the diagram is usually populated by the following categories: methods (process), materials (information), man (human factors), environment and equipment. Contributing factors or root causes to each category are then attached to each category “rib” with some contributing factors possibly fitting into more than one category. Creation of this tool benefits the quality improvement process by identifying multiple causes that may contribute to. specific effect. This graphic representation displays the relationship of the causes to the effect and to each other, thereby further focusing the team to the area of improvement [14].

Key driver diagram

A key driver diagram (KDD) is. quality improvement tool that teams use to increase their odds of success during their journey of enacting process or system change. This diagram is. pictorial representation (Fig. 5) of the relationship between the aim statement of the project and the primary drivers that contribute directly to achieving the aim and the secondary drivers that influence the primary drivers. Similar to the Fishbone diagram, creation of the KDD allows the entire multidisciplinary team to have. shared view of the proposed change in the system that is being evaluated. The KDD is. representation of the cause and effect in this system and what changes will likely result in the desired effect [15].

Fig. 5
figure 5

This key driver diagram outlines the steps that. hospital-based team has identified to improve the percent of eligible premature infants with respiratory distress who receive surfactant within two hours of birth. The primary driver are the components that contribute to the project aim with the secondary drivers and change concepts aligning with each primary driver.

Full size image

Development of. KDD allows for the project to have. visual representation of. structured improvement roadmap. Construction of these diagrams detail the project planning by keeping. focus on factors that allow the team to realize the project aim. This diagram organizes the improvement aim, the key drivers (the “what” needed to accomplish that aim), and the interventions (the “how” [change concepts], also known as secondary drivers) into. learning and communication framework.

In Figure 5, the primary aim is to administer surfactant within 2 h of age to 100% of inborn infants ≤1500 g and/or <32 weeks gestation who met criteria to reduce the incidence of chronic lung disease (CLD). The primary drivers (early selective surfactant administration, team effort and parental involvement) are the system components that contribute directly to the project aim and are the key elements that must change in order to accomplish the goal detailed in the aim statement. The secondary drivers are actions or interventions (development of criteria for surfactant requirement, staff education) that can be tested in order to affect the primary drivers and are necessary for the success of the primary drivers. The secondary drivers are influenced by specific change ideas that are put into practice where the change can actually occur. These primary and secondary drivers are the key leverage points is the system that is being studied and each driver should be measureable and align with specific process measures of the project. The KDD should also indicate the causal relationship between secondary drivers, primary drivers and the aim statement.

Related to the secondary drivers are the change ideas (literature review, obtaining blood gas by 90 min of life) which are actionable elements that can be tested and measured on the system with the expressed outcome of accomplishing the aim statement. The change ideas that are proposed and the tests of change that accompany these change ideas fit in well with the PDSA method that was described earlier. The PDSA model for testing ideas allows for the prediction, testing, learning and revision (if needed) of the change ideas, leading to the discovery of knowledge for achieving improved system outcomes.

Conclusion

This brief review of quality improvement tools for the clinician was intended as an introduction to structured quality improvement work. Utilization of the IHI Model for Improvement lends. structured framework for teams to follow when designing. QI project. The addition of process mapping, Pareto charts, Ishikawa diagrams and key driver diagrams provides teams with powerful tools to understand processes, organize root causes of inefficient processes and construct. plan for implementation of. QI project.

By virtue of working in neonatal intensive care units, the multidisciplinary team structure lends itself to collaboration and cooperation in the daily care of patients. Whether it is local participation in quality improvement work at the unit level, system level or participation in national quality improvement systems (Vermont-Oxford, Mednax Center for Research, education, quality, safety, etc.), most neonatal health providers have some exposure to quality improvement initiatives. As care models continue to evolve and continuous quality improvement becomes standard work in the neonatal intensive care units, integration of these tools into daily routines will provide the best care possible for the infants and families, ensuring improved outcomes in the years to follow.