➢ The manufacturing industry has supplied many quality-improvement methodologies that have been successfully utilized in health-care delivery, such as Plan-Do-Study-Act (PDSA) cycles, Total Quality Management, Six Sigma, and Lean.
➢ Many tools of quality improvement, such as PDSA cycles and DMAIC (Design-Measure-Analyze-Improve-Control) of the Six Sigma method, are similar to the scientific method that is familiar to clinicians.
➢ Correct identification of the sources and types of process variation within a system is paramount for process improvement.
➢ Reduction in process variation via standardization and reinforcement of process protocols leads to improved process outcomes.
➢ Quality-improvement projects should define a clear governance structure to maintain project timeliness and completion.
There is a long history of the application of quality-improvement (QI) methods in health care1,2. Many of these methods originated in the manufacturing industry. When applied in health-care settings, QI projects have led to reduced costs, reduced infection rates, and reduced start-time delays3. The purposes of the present article are to review the history of QI methods in the manufacturing industry, to discuss current and historical implementations of QI in health care, and to outline various QI methods that can be used to improve the quality of an orthopaedic surgery practice. The goal is to provide an introductory reference for providers who intend to perform QI projects in their own practice, with a focus on improving quality while decreasing costs.
History of Quality Improvement
The development of QI processes originated in the manufacturing industry. Three prominent figures contributed most to the development of these processes: Walter Shewhart (1891 to 1967), W. Edwards Deming (1900 to 1993), and Joseph M. Juran (1904 to 2008).
Walter Shewhart was a physicist, engineer, and statistician. In 1918, he was hired by the inspection engineering department of Western Electric Company. By 1924, he had developed a system to identify 2 distinct types of variation in the manufacturing process of telephone hardware, which he called assignable cause (often called special cause) and chance cause (common cause)4. Chance cause refers to random variation that may occur as a result of forces that are constantly active within the system itself. Chance-cause variation could result from measurement errors, poor machine maintenance, substandard raw materials, inappropriate procedures, poor working conditions, and normal wear and tear. These causes of variation produce a repeatable, stable distribution of variation over time. Special cause refers to a variation in a system that is unpredictable and the result of an inherent change in the system itself. Examples of special-cause variation could include a computer crash, a machine malfunction, an absent technician or member of the operating room staff, and errors resulting from untrained workers. A single source of special-cause variation can often be identified as the sole reason for abnormally high or low observed process values and may be easier to identify than chance-cause variation. Shewhart proposed that systems possessing only chance-cause variation were in “statistical control.” To determine if a variation was due to chance cause or special cause, he developed the “Control Chart,” which plots variation in outcome over time (Fig. 1). Shewhart proposed that any point on the graph that differs from the mean expected outcome (i.e., the center line of the control chart) by >3 standard deviations (SD) has an increased likelihood of being caused by assignable cause or a non-expected source of variation (99.73% of data points lie within 3 SD).
In addition, the Western Electric Rules of statistical process control outline 3 additional rules for detecting assignable (or special) causes of variation (Fig. 1). If any of these 4 situations arise, a system may be reflecting newly introduced special-cause variation.
In medicine, control charts have been used to analyze the impact of QI interventions in time series data by plotting infection rates before and after QI projects. Control charts provide a way to determine the efficacy of a given intervention5 and also a way to monitor postoperative complication rates over time6. In the study by Duclos et al., for example, the use of control charts allowed for the identification of two sources of special-cause variation when infection rates nearly doubled: the absence of a surgeon and the performance of procedures during renovation of operating rooms6. Control charts also offer a real-time data graphic for the clinician to appreciate outcomes with a visual pattern. In these ways, control charts are useful tools for improving surgical care.
While at Western Electric and later at Bell Telephone Laboratories, Shewhart worked with 2 other prominent figures in QI history: W. Edwards Deming and J.M. Juran, both of whom would further develop his work.
Dr. W. Edwards Deming was an electrical engineer who was introduced to Shewhart in 1927. He believed that Shewhart’s model of source variation in manufacturing could be applied to management. He developed the Shewhart cycle (now evolved into the Plan-Do-Study-Act [PDSA] or Plan-Do-Check-Act [PDCA] cycle) as a method for quality improvement and developed a management system known as Deming’s System of Profound Knowledge (SoPK). The SoPK outlines 4 core content areas that a manager must master in order to reduce waste and increase quality and profitability in a system. These 4 areas are (1) appreciation for a given system (understanding the relationships between all system components), (2) knowledge of variation (distinguishing between common-cause and special-cause variation), (3) theory of knowledge (a way to learn about a system, observe past results, hypothesize, make predictions about the future, test theories, and check results, leading to continual analysis and process improvement [e.g., PDSA cycles, which resemble the scientific method]), and (4) psychology (how people in a system relate to one another)7. Deming taught his theories extensively in Japan after World War II and is often credited for much of Japan’s post-war economic success.
It is imperative that QI project leaders distinguish between sources of common-cause variation and special-cause variation. If, for example, an x-ray technician routinely increases operative time during orthopaedic procedures by improperly operating the fluoroscopy machinery, an appropriate solution to this source of special-cause variation might include replacing or appropriately training the technician. However, if the fluoroscopy machine malfunctions because of design properties associated with wear and tear (common-cause variation), training or replacing the technician would be futile and would result in increased process cost and decreased efficiency. In this case, procedural and design improvements such as updating the machine technology could be implemented to lower the rate of common-cause variation.
J.M. Juran was a Romanian-American electrical engineer and lawyer who became the chief of Industrial Engineering at Western Electric. Juran is often remembered for his classic 1951 Quality Control Handbook, in which he outlined a method for quality improvement in cross-functional management (with team members having different specialties, or functions) known as the Juran Trilogy. The Juran Trilogy consists of the 3 managerial practices that optimize systems and performance outcomes. The Trilogy includes Quality Planning (identifying the customer’s needs and quality goals), Quality Control (reducing the difference between current performance as compared with quality goals), and Quality Improvement (defining QI projects, creating QI teams, and maintaining improvements made)8. In addition to these principles, Juran also developed the Pareto Principle (the 80-20 rule), which states that 80% of problems in a system arise from 20% of the system. This understanding allows a focused identification of problems within systems for maximum improvement.
History of Quality Improvement in Health Care
Medicine has a rich history of developing new methods of improvement. Ignaz Semmelweis (1818 to 1865) was a Hungarian obstetrician who identified cadaverous materials as the offending agents responsible for high puerperal fever mortality rates at a hospital in which medical students were performing autopsies. He noted a 90% decrease in mortality rates after instituting a policy of hand-washing with chlorinated lime solution9.
Ernest Codman (1869 to 1940), a shoulder surgeon from Massachusetts General Hospital (MGH), developed the “End-Result System.”1 He proposed that surgeons should keep track of every patient’s symptoms, diagnosis, operative outcomes, and complications. Codman designed the End-Result System as “The common sense notion that every hospital should follow every patient it treats, long enough to determine whether or not the treatment has been successful, and then to inquire, ‘If not, why not?’ with a view to preventing similar failures in the future.” Codman proposed the use of outcome measures, rather than seniority, as a gauge of surgical ability and a basis for career promotion10. At the time, Codman’s ideas were ridiculed and opposed. He resigned his post at MGH in 1911 to start his own hospital, the Codman Hospital, where he could implement the End-Result System.
Avedis Donabedian (1919 to 2000) was a Lebanese physician at the School of Public Health at the University of Michigan. Donabedian’s frequently cited classic paper entitled “Evaluating the Quality of Medical Care” suggests that the practice of medicine can be evaluated through an analysis of 3 components of care: (1) clinical outcomes, (2) the process of care delivery (comprising all actions included in the delivery of care, such as the appropriateness of medical actions, the acceptability of care to patients, continuity of care, etc.), and (3) the structure of care (e.g., the adequacy of facilities/equipment, medical/support staff qualifications, administrative structure, fiscal organization, etc.)11. Donabedian’s work was novel in that he suggested using more than outcomes as a measure of medical care. For example, outcomes after treatment may not fairly represent the appropriateness of medical care delivered if treatment is still unlikely to produce optimal health, such as when survival is used as an outcome after nonfatal trauma that results in a chronic condition despite treatment (e.g., limb amputation)11. Furthermore, even when outcomes are an appropriate measure, limitations still exist that must be considered. If the effects of treatment are delayed, possibly by decades, results may be unavailable when they are needed for comparison. Thus, some outcomes may be an inconvenient or inadequate measure of the quality of care provided, although appropriate (e.g., relearning to walk after severe orthopaedic trauma, which could take many years, may not immediately or consistently reflect successful orthopaedic care). The process of analyzing the quality of care by drawing information from structure, process, and outcomes has come to be known as the Donabedian model.
Tools for Quality Improvement
The manufacturing industry has developed many techniques to improve the quality of production that can be applied to health care. These tools have been used to improve product design, to reduce production costs, to increase system efficiency, and to decrease costs associated with correcting for defective products. Four of the most common techniques used in health care include PDSA cycles, Total Quality Management (TQM), Six Sigma, and Lean. Health-care systems have adopted these techniques in attempts to improve the quality of many aspects of medical care12-15.
PDSA is a systematic approach to learning about a process that leads to continuous quality improvement. Dr. W. Edwards Deming first popularized PDSA cycles in the early 1950s as previously described. PDSA is a rephrasing of the scientific method, which can be written as Hypothesize-Experiment-Evaluate-Action based on observations (or Plan-Do-Study-Act)16. van Tiel et al. successfully utilized PDSA cycles to improve compliance with postoperative infection-control guidelines after cardiothoracic surgery by taking the following steps12.
• Plan: Define modifiable risk factors for wound infection and develop a strategy to prevent new cases.
• Do: Collect baseline data on compliance and carry out a planned strategy utilizing infection-control indices for improvement.
• Study: Analyze the outcomes and delineate what part of the strategy failed to achieve the desired goal of improved compliance with each of the described indices.
• Act: Identify and adopt strategies to improve the initial lack of compliance and repeat the cycle for reassessment.
This method proposes a framework for the continual improvement of systems through the repetitive iterations of the PDSA cycle and with implementation of system changes after each cycle (Fig. 2). The applications of the cycle are vast and can easily be applied to orthopaedic and surgical care2,12.
In completing a PDSA project, it is important to consider 3 primary questions17: (1) What are we trying to accomplish, (2) How will we know that a change is an improvement, and (3) What changes can we make that will result in improvement?
By defining clear goals, measurable metrics for improvement, and many possible solutions, PDSA cycles can be an effective tool for identifying and altering specific process weaknesses for continual improvement of process outcomes and efficiency.
Total Quality Management (TQM)
Total Quality Management (TQM) is a management system built on the work of Deming and Juran that prioritizes customer expectations and a team-based approach to deliver improved products or services to customers and to increase customer satisfaction18. A central concept in TQM management is the integration of all employees into the common goal of quality improvement. The work of Deming has been adapted for use in health care by the Institute of Healthcare Improvement (IHI), Cambridge, Massachusetts19. It relies on 3 primary elements: (1) considering the needs of both patients and the practice, (2) establishing a culture of quality improvement with special emphasis on communication and teamwork, and (3) forming small multidisciplinary teams to complete specific QI projects. These team-based improvements can be accomplished through the use of analytical tools such as flowcharts, statistical charts, and check sheets as well as through the use of process techniques such as nominal groups, brainstorming, and consensus-forming to produce communication between team members and improved decision-making13,18.
Six Sigma is a management system, developed by Motorola in 1986, that is centered on reducing variability within work processes through reducing error rates to 3.4 defects per million opportunities (DPMO) (i.e., 99.99967% of process values will be between −3 and 3 SD [σ] of a process mean). The 5-step methodology used in the Six Sigma system is known as DMAIC (Define, Measure, Analyze, Improve, Control).
• Define: Identify problems, goals, resources, project scope, and a timeline for project completion. Useful tools for teams in the Define step include a project charter, Supplier-Input-Process-Output-Customer (SIPOC) analysis, and process maps, which can be created to overview the entire process in its current state.
• Measure: Collect baseline performance data. Histograms, run charts, and Pareto charts are often useful. Pareto charts are a type of vertical bar graph with independent variables plotted in order of decreasing relative frequency from left to right. Often a point-to-point line graph will be superimposed to display cumulative frequency. These charts provide visual identification of the most relevant variables to be addressed by the QI team.
• Analyze: Identify and remove true root causes of variation from a process. An Ishikawa fishbone diagram/root-cause analysis (Fig. 3) and process-mapping tools may be useful to help identify potential sources of system variation. In the creation of a fishbone diagram, the “6M” methodology can be considered to analyze variation arising from various process components: Mother Nature (patient factors), manpower (registered nurse and physician factors), measurement (data collection), material (equipment, sutures, graft materials, etc.), methods (process organization and flow), and machines (surgical tools, computer systems).
• Improve: A plan for the elimination of top causes of system variation should be established and implemented. Standardization of current nonstandard procedures and simplification of bureaucratic processes are often effective process improvements. After removal of root causes of variation, the analyses of process outcomes data should show decreased process variation.
• Control: Validation of the new process technique must be monitored through process outcomes analysis. Statistical analysis that compares before-intervention and after-intervention outcomes data should be performed. Quality gains should be standardized to maintain the benefit produced from the project.
In one study, Frankel et al. used a Six Sigma approach to increase patient safety through a reduced rate of catheter-related bloodstream infections in a surgical intensive-care unit14. Using the DMAIC process, they identified root causes of system variation and established an action plan for improvement that included the creation of new standard operating procedures, the production of an instructional videotape for incoming residents, the use of antibiotic-coated catheters, a measurement system to track catheter dwell time, insertion of all catheters in the presence of a surgical intensive care unit (SICU) attending surgeon, assembly of standardized catheter kits, and standardization of dressing changes and types. In a two-year period, the rate of catheter-related bloodstream infections rates was reduced from 11 infections per 1000 to 1.7 infections per 1000 (p < 0.0001).
In addition, the Six Sigma method has been shown to be effective in many surgical settings by reducing operating room turnaround time20, improving adherence to perioperative antibiotic prophylaxis procedures21, and improving surgical outcomes22.
Lean manufacturing emerged from the Toyota Production System (TPS) in the 1990s and is aimed at reducing waste (muda in Japanese), or non-value-adding activities. Waste processes include overproduction, waiting time, unnecessary transportation, non-value-adding processes, inventory, motion, and costs of quality (scrap, rework, and inspection)23. Tools that can be used to apply Lean production theories include value stream mapping to visualize a process overview and to reduce process excess and bottlenecks. Yousri et al. showed that value stream mapping led to improved orthopaedic outcomes in patients with femoral neck fractures by lowering overall and 30-day mortality by 9.3% (p = 0.002) and 5% (p = 0.034), respectively, through the elimination of non-value-adding steps in the process of orthopaedic care24.
The utilization of 5S principles is also useful in Lean methodology. 5S is an organizational technique that is used to eliminate wasteful components in a workspace. The 5S pillars include the following25:
• Sort: Eliminate whatever is not needed from the workspace.
• Straighten: Organize what remains for easy selection for work use.
• Shine: Clean the work area as an inspection technique; prevent deterioration.
• Standardize: Schedule regular cleaning and maintenance.
• Sustain: Make 5S a way of life and perform internal audits to maintain uniformity.
Lean thinking has been used in health care to improve the efficiency of the orthopaedic surgery scheduling process15, to standardize examination rooms (via 5S techniques), to improve perioperative antibiotic therapy for patients undergoing surgery26, to decrease patient length of stay in examination rooms, to increase face-to-face physician-patient time, to increase the number of patients seen, and to increase and sustain patient and physician satisfaction27. In addition, Lean and Six Sigma methodologies have been combined into the Lean Six Sigma (LSS) model, which has been used widely in orthopaedics to reduce hospital length of stay following trauma28, hip fracture29, and joint replacement30.
In 2009, at the Richard L. Roudebush Veterans Affairs (VA) Medical Center in Indianapolis, Indiana, a program was established in which LSS principles were used to decrease length of stay, to improve overall efficiency, and to reduce non-VA costs in the setting of total joint replacement surgery30. This model, known as Vision-Analysis-Team-Aim-Map-Measure-Change-Sustain (VA-TAMMCS), closely follows the fundamental DMAIC model of Six Sigma and aims to reduce non-value-adding steps as in Lean. Through this initiative, the medical center successfully reduced length of stay by 36% and completely eliminated non-VA health costs.
In 2008, at the Mayo Clinic in Rochester, Minnesota, an institutional initiative was established in which a multidisciplinary team was used to improve operating room efficiency across multiple surgical subspecialties with use of the LSS methodology31. A detailed value stream map of patient flow from preoperative evaluation to patient discharge was developed to identify bottlenecks in patient care. Five work streams that contributed most to the flow of surgical patients were identified. Within each of the 5 work streams, non-value-adding steps were identified and solutions to improve efficiency were developed with use of the DMAIC process of Six Sigma. Through this initiative, the institution was able to significantly (p < 0.05) improve operating room on-time starts and also to reduce the number of cases starting after 5 p.m. Furthermore, the institution was able to decrease nonoperative time in between cases and to gain financially with increases margin per operating room per day.
Before beginning a QI journey, it is crucial to establish a clear governance structure, regardless of the type of QI methodology used. A clear governance structure allows project members to understand when and how decisions will be made. Without such a structure, members of the project may become frustrated because the project seems to “flounder” without any clear leader or approval to move forward.
What Is a Governance Structure?
A governance structure includes several aspects of a project, including the teams of individuals who will develop recommendations and implement changes, the process by which these recommendations will be communicated and approved, and the forum for results to be presented and celebrated. The project team members must have a clear understanding of each of these 3 aspects of their project. This understanding will ensure that project teams remain focused on making tangible progress on a recurring basis and are regularly updating key project stakeholders. The governance structure also ensures that stakeholders recognize the importance of the project as a priority of senior leadership. Regardless of QI methodology (DMAIC, TQM, or other), the establishment of a governance structure to define a clear plan for how a QI project will be organized and how decisions will be made is paramount to successful project implementation.
The methodologies of quality improvement described here have all proved to be useful in health-care settings. Table I lists some tools that may be used when implementing one of these QI programs. These are only a few of the many useful tools for improving the quality of medical care. Although QI techniques will continue to evolve, many of these processes build on the basic framework of the scientific method: observe, experiment, analyze. The fundamental forces driving these developments in industry are the same as in health care: the need for increased quality through increased systems efficiency (as opposed to increased spending), and the desire to lower associated production costs. Different QI techniques are appropriate for addressing different medical problems. Physicians should select a QI method best suited to their particular process of inquiry, collect data, analyze sources of system variation, implement change, and repeat the cycle as needed for maximum and continuous quality improvement.
Investigation performed at the Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, NY
Disclosure: No funding was required to complete this review study. The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated