J Joshi, T Best
lean, patient care, pdca, quality improvement, root cause analysis
J Joshi, T Best. Quality Improvement Methods for Medical Handoffs. The Internet Journal of Medical Education. 2009 Volume 1 Number 2.
Objectives: We describe the application of fundamental quality improvement methods to medical handoffs in order to improve resident learning and clinical training.Methods: We select six pervasive quality improvement methods that each highlight a particular aspect of overall process improvement. The six methods are the 7-S Framework, Plan-Do-Check-Act, Define-Measure-Analyze-Improve-Control, Lean, Root Cause Analysis, and the Prioritization Matrix.Results: We describe how each method can be used to improve the efficiency and effectiveness of a patient care handoff. Examples are provided throughout to help depict each method in the handoff context. Our exposition is written so as to introduce the methods to a practitioner charged with improving a medical handoff, or a medical educator providing a course on healthcare process improvement.Conclusions: The medical field recognizes the importance of studying other disciplines in order to optimize key processes in medicine. We present six quality improvement methods not universally acknowledged in the medical field that are particularly applicable to improving medical handoffs. By contextualizing these methods into the medical handoff, we show both how to improve the overall process and emphasize the value in implementing concepts from other disciplines as part of the residency training curriculum.
In the ever changing residency environment, clinical departments seek to optimise patient care while simultaneously avoid overworking medical residents. Beginning July 2003, the American Council of Graduate Medical Education (ACGME) developed the duty hour restrictions as one of the means to reduce resident fatigue. Since then, studies conducted on work hour limitations demonstrate positive results on resident performance.1-5
Despite the preponderance of positive benefits, the change has produced some potential and unintended consequences.6 Simulated models estimate increased financial costs and decreased resident teaching time.7,8 While such analysis may be difficult to ascertain9; one certainty is an increase in handoff care from one resident to another.
A foundation in clinical practice, the handoff of care involves understanding and conveying the patient’s medical conditions, detailing a plan of action, and communicating this complex set of information efficiently. Limited resident work hours affects the continuity of care and makes a successful handoff all the more imperative.10 Maintaining a sense of continuity is a major concern among resident physicians11, and resolving it begins by improving the handoff experience. With new requirements set to begin in July 2011, recognizing how to improve the handoff experience becomes critical.12 Many published studies advocate modeling the handoff into a standard protocol.13-19
Given this need, we seek to contextualize quality improvement (QI) tools into developing the handoff model. In other industries, such quality improvement tools focus on optimizing a firm’s capabilities and resources to best produce a product or service. Similarly, the same QI measures can be used to develop an optimal standardised handoff model. We assume the perspective of a QI team focused on improving the handoff through six fundamental quality improvement models: 7-S Framework, Plan-Do-Check-Act (PDCA), Define-Measure-Analyze-Improve-Control (DMAIC), Lean Principles, Root Cause Analysis (RCA), and Prioritization Matrix.
Each model is derived from a core principle that addresses a specific aspect of quality improvement. We arrange the models to address the specific complexity in the order in which the issue typically arises. The 7-S Framework allows the QI team to appreciate how the new handoff will affect the department’s intricate structure. PDCA and DMAIC establish models for implementing a QI plan of action. Lean Principles address waste reduction initiatives fostering an efficiency-driven mentality. RCA define the nature and frequency of variations and problems inherent in common handoff models. Finally, the Prioritization Matrix sets up an objective, conflict-resolution model essential in prioritizing the many values and principles a department seeks in a handoff model.
Applying these principles serves as an educational tool to convey improvement opportunities and to recognize potential flaws inherent in the handoff process. These models structure the handoff experience by demonstrating its intricacies and educate residents on how to improve the transition of care.
All new handoff programs face conflicts that arise from competing values and varying levels of motivation.20 Anticipating such issues will prevent them from derailing the change. The 7-S framework21,22 addresses this by developing a basic understanding of the handoff environment. The 7-S breaks down a work system into seven elements: strategy, structure, systems, skills, staff, style, and shared values.21
The team first identifies which of the seven elements is central to the change under consideration. It then establishes links with and among the remaining six elements to create a network depicting the overall repercussions of the change.21,22,23,24 In general,
While some conflicts are resolved through clarification of change details and intent, many persist without an adjustment to the original design.25 For example, suppose the team suggests a patient handoff from a surgical resident to a resident in the recovery suite. Nurses may object to this change, given their current leadership of post-op patient management. In this case it may be best to compromise and split recovery management into tasks for both disciplines. In this example, nurse recovery management could be referred to as a
With the 7-S model, one shows the potential changes, and limits thereof, allowing a department to gauge a handoff’s efficacy through comprehensive understanding. This ultimately determines the proper course of action to be implemented. Once the environment is understood, the team will need a method to design and implement a new handoff model.
Plan-Do-Check-Act (PDCA) and Define-Measure-Analyze-Improve-Control (DMAIC)
After understanding the environment, the first task for any handoff improvement team is to select a method for designing and implementing a new handoff model. Two methods are
To understand PDCA, think of the
Next, the team
For example, a new handoff intends to improve the quality of inpatient trips from the general wards to the Catheterization (Cath) Lab. The goals are to improve the completeness of documentation and the efficiency of transfer. Instead of attempting to design a handoff model that immediately corrects every issue, the team can implement PDCA to sequentially address each problem. The team can also design a template to improve documentation by ensuring completeness before the transfer request is sent to the Cath Lab. After this initial pilot, the team notices that although documentation has improved, transfers are now taking longer. The second PDCA cycle should identify the cumbersome components of the handoff, and test a new program which improves efficiency.
While PDCA is effective in certain situations, it has limitations. The frequency of PDCA cycles are limited by the time it takes to develop a revised handoff. If each pilot needs extensive review, practicing PDCA can be inefficient: due to the rapid plan phase, initial pilots underperform and time is wasted waiting to review each subsequent pilot.27 Further, the success of PDCA is particularly dependent on whether the
The DMAIC model imposes a rigor of analysis that is not explicit in PDCA.34,35 In the
Immediately following implementation, DMAIC requires the team to sustain the improvements by using tools to
Both iterative models establish a paradigm for controlled, sustainable improvement. The PDCA initiates a series of quick cycles to effectively address individual revisions. The DMAIC initiates more deliberate, in depth revisions tested through rigorous measures. Determining the optimal model for a department depends on how the new model will change from the original handoff, and how involved the department will be in extensively monitoring the changes.27,33
A “Lean” Perspective
Regardless of the improvement method chosen, the team will identify certain components of the handoff that are prone to error, delay, or redundancy. In general, such issues can be classified as
Lean principles are rooted in the seven categories of Lean “waste”37,41:
To illustrate what handoff components would fit in each category, consider the following examples:
The defect itself (e.g. During a shift change handoff, a caregiver forgetting to mention that a verbal order was recently given to remove the patient’s urinary catheter),
The time and medical supplies spent making the defect (e.g. the time wasted providing the original verbal order)
The remedy for the defect (e.g. the time spent providing the new verbal order, or even the time and medical supplies spent caring for a UTI that results from the excessive catheter use).
Some forms of waste are impossible to remedy.42,44 A patient in need of an MRI unavoidably needs to travel to the MRI suite due to the machine’s size and expense. In other cases, desirable variation is mistakenly identified as waste. Assuring a new cancer patient multiple, “redundant” times that many treatment options are available for their condition isn’t wasteful, it is appropriate bedside manner.
Defining waste is one among the many lessons in Lean philosophy. Lean techniques can also enhance the way a team views a Process Map.37-46 It is productive to rank each step in the process by the amount of waste and begin addressing the worst. One of the best ways to evaluate waste is to determine the
In Figure 2, we demonstrate how one would depict value-added time and effectiveness on a process map; specifically the surgery-to-recovery handoff at a hypothetical academic institution.
Lean fosters a working environment where continual learning and inquiry thrive. New sources of waste will continue to challenge any process, but Lean initiatives will lead to the resolution of such issues. Such a culture is critical to a successful, sustainable patient handoff.
Root Cause Analysis
Root cause analysis (RCA) is a problem solving method that identifies undesirable variation in a specific event or process, and uncovers the root causes of those problems.47 Since a handoff model’s success is determined by its ability to handle the variations of clinical care, RCA is a valuable tool for identifying the underlying causes of problematic variation.47,48,49
In our context, a QI team starts by documenting a handoff process map. Next, note the inherent fluctuations in each step (e.g. the delays in surgery start time). Once each source of variation is identified, the team separates that variation which is desirable from that which is problematic or unnecessary.48,50,51 A helpful guide for distinguishing between desirable and problematic variation is our discussion of Lean waste in the previous section.
Next, the team identifies what contributes to the problems. The process proceeds onward until the basic root causes are clearly identified. Some of the underlying factors are broad, systemic issues, and deserve thorough investigation. Others are specific and corrective action is easier to implement.
The team presents RCA results in a variety of ways. We discuss three formats: a Tree Diagram, Pareto Chart, and a Control Chart.49,51-54
Often, root problems stem from a few defined sources, so the Tree Diagram is a natural method.49 It organizes each event individually and branches out to display its causes. A problematic event can have many branches before leading to the root cause.
Alternatively, Pareto Charts are not organized in a cause-and-effect diagram52; rather they indicate how often each problematic event occurs. As such, they require data collection to populate.
A Control Chart isolates one or more error types in a Pareto Chart and plots performance over time.30 The chart alerts the team to significant, undesirable trends in a given performance metric.
Used together, the three tools allow one to visualize inefficiencies along the entire handoff process map and then focus on variations at each individual step.
In the Lean section, we reviewed part of a hypothetical handoff process map. Applying RCA to this map starts by interviewing the relevant staff. After the interviews, the team could use a Tree Diagram to visually depict the issues (shown in Figure 3). Notice that each branch displays a sequential series of logic that eventually leads to a root cause.48 The root causes suggest an intuitive plan for improvement.
A Pareto Chart could also help in our example. Assume we institute a survey to assess the frequency of the five root causes. After administering the survey, we obtain the following frequencies: (1), 40%; (2), 25%; (3), 20%; (4), 10%; (5), 5%;. The Pareto Chart inserted (Figure 4, left) demonstrates the frequency of each variation relative to each other.
To understand a particular variation source more thoroughly, the team could develop a Control Chart. The chart plots performance over time for a specific step. Abnormal variation is observed outside designated upper and lower “control limits”. Control limits are a function of the data defined by a confidence interval as a function of the mean, variation (standard deviation), and sample size.36,53 Consider the first (1) problem cited in our Pareto Chart on the left side of Figure 4 that we hypothetically monitored for thirty independent events. The control limit (Figure 4, right) was set about a ninety-five percent (95%) confidence interval with a mean of six minutes and a standard deviation of two minutes.
Used together, the Tree Diagram, Pareto Chart, and Control Chart visualize inefficiencies along the entire handoff process map and then focus on variations at each individual step. A QI team can study improvement on a more complex level, effectively specifying where the improvements are needed and how well the improvements work. Each model complements one another to organically synthesize a complete understanding of the handoff improvement.
Implementing a quantitative method to complement PDCA and DMAIC assists in prioritizing the department’s values and the desired goals of the handoff initiative.54
Quantitative Prioritization specifies the scope of improvement and organizes the values within a department.54,55 In the handoff model, these are the aspects that improve the new design (e.g. staff-to-staff communication, patient education, regulatory compliance). This model prioritizes each dimension based on a number ranking that represents the appropriate importance level.55 Orthogonal to the handoff dimensions are the department’s, or stakeholder’s, values. Such values can be solicited with a variety of methods, including surveys. The goals of the department are similarly determined and given an importance score. By conducting this exercise, the team aligns the department’s values with the dimensions of the handoff to find the best balance.56,57
Figure 5 gives an example of the end result. The inner square represents the strength of correlation between any two items. For example, a nine was assigned to the cell where “reduce errors” and “extended personal communication” intersect. Assigning a value of nine implies that stakeholders believe that the two components are strongly correlated. By applying numerical values, we can quantify the aggregate correlations. Note that the 9, 6, or 3 valuation is arbitrary; it simply assesses the relative strength of correlation in this example.54,55 Each stakeholder has unique values and each team member assesses a handoff dimension’s importance differently. It is left to the team to use a sub-method such as the survey to properly assess the overall strength of correlation.
Lastly, the team adds across each row and column to determine the sum total of each correlation. We then multiply the sum with the preference ranking to obtain a total score.54,55 Items prioritized as “one” would be multiplied by a factor of “three”, just as items prioritized as “three” would be multiplied by a factor of “one.” The total score helps the QI team determine what to emphasize in merging the handoff dimensions with the department values.
Quantitative Prioritization allows comprehensive, collaborative input from all stakeholders in a value weighted manner. Such collaboration can motivate members within a department to help implement the handoff and help ameliorate potential conflicts that arise during implementation.
Medical programs must develop innovative methods to incorporate the trends in medical handoffs into their curricula. Knowledge brokering key concepts from other disciplines elucidates opportunities to improve the handoff transition and to educate why such improvements are necessary. Medicine must integrate new disciplines, such as Industrial Engineering QI, to best prepare residents for medical practice because these methods discordantly reflect how one perceives handoff transitions. And this perspective provides a different latitude that is both implementable and educational.
These QI principles represent an overview of the analyses necessary to overcome common impediments to successfully implementing handoff quality improvement. These models are among the many one can utilize to study handoff improvements; and a focus on why these principles lead to success is more important than rote memorization of these principles. We provide unique insight from these models to best understand how one can apply each model in a specific context. We arranged each model in the order that they are likely to be useful to a QI team innovating a new handoff model.
A medical department seeking to educate its residents on handoff improvements should structure the training along this framework. Presenting handoffs as a complex field ripe with principles from many disciplines allows the residents to appreciate the complexity in handoffs and to systematically learn the improvement methods. These models, when combined, integrate into a framework conducive for resident learning.
Each model is described in brief with suggestions for when it can be most efficiently implemented in the handoff process.
We would like to acknowledge Dr. Vineet Arora of the University Of Chicago Pritzer School Of Medicine and Dr. David Mayer of the University of Illinois, Chicago for their educational guidance and support during the drafting of this manuscript.