Wouldn't We All Like to be a Little More LEAN?

There are several business based tools and systems that are designed for deconstructing a complicated process and analyzing that process for waste and efficiency. ¹ Examples of these tools include Six Sigma, which focuses on assessing mistakes so that the process that facilitates errors can be changed and Toyota Lean. The Lean system is based on generating efficiencies via eliminating waste. Lean has been adopted by many health care organizations in an attempt to control costs and improve the quality of health care delivered to the patient. ² There are several key components of “Lean thinking” that make it ideally suited as a philosophy to optimize translational research practices. Lean requires data driven decision‐making. Changes to a process and/or organization cannot be made on anecdotal evidence. The Lean approach teaches members of the organization how to systematically collect and analyze data to address a specific question and then act on that data. In essence, Lean tools train users in the scientific method. Lean is a communications tool. All members of the organization working on a project using Lean methods learn to talk the same language. They are brought into the project team as equals. Everyone is encouraged to question and trouble shoot the process. All voices are heard equally. Lean requires consensus on action from all parties involved in the process. As the process is deconstructed and reconstructed, representatives of every stakeholder who uses or oversees the process contribute to its regeneration. For this reason, dramatic changes may be better institutionalized since there is a large personal investment by team members in success. Finally, Lean is continuous process improvement. Data collection, questioning, and optimization of the process or organization never end. Lean promotes a shift out of more traditional “top down” cultures. The philosophy encourages challenging the status quo and teaches people to facilitate change from the bottom up. This is the type of culture that is needed to support the complexities of translational and clinical research.

The Clinical and Translational Science Awards (CTSA) support and promote the transformation of translational and clinical research and the cultures that support this type of research. In the business world it has been observed that transformation often fails due to a set number of critical mistakes; (1) lack of urgency, (2) no coalition to drive transformation forward, (3) lack of vision, (4) poor communication, (5) lack of empowerment of all participants in the process of change, (6) no short term wins to keep people motivated, (7) loss of momentum, and (8) lack of institutionalization of the new process or culture. ³ Tackling transformation using the Lean system can help avoid some of these key errors. We have been using the Lean approach to develop and revamp research structures at our institutions including re‐designing our Clinical Research Center (CRC). We describe a case study of the use of the Lean approach to decrease study initiation time across our CRC units.

Our initial work started with a survey. We polled 262 CRC users, both investigators and staff at our University and Children’s hospitals, and asked a series of questions about CRC services and usability. One of the lowest ranked services was the process for application to use the CRC. On a scale of 1 (very unsatisfactory) and 5 (very satisfactory), 66% of respondents stated their satisfaction with the application process was a 3 or 4. We created a Rapid Process Improvement Workshop (RPIW) to review and trouble‐shoot the CRC application and study initiation process. During an RPIW, a team is assembled that will interact very intensively over a short period of time (1 week). The goal is to change the broken process immediately. A diverse team of stakeholders is selected for such projects as decisions made at the RPIW need to be final and all parties must come to a consensus during the workshop. Our Rapid Process Improvement (RPI) team consisted of Scientific Advisory Committee (SAC) members (review and approve the studies), investigators and research coordinators (users of the process), hospital and academic officials (oversee the CRC and may have to approve or support process change), and CRC staff (review and implement studies). The team started by mapping the current steps to initiate a study on the CRC (Current State). This exercise demonstrated the application/study initiation process was more much complex than anyone had realized. There were 41 steps to get an application to approval and enrollment that consisted of an internal unit specific review, scientific review, and a study implementation meeting. Six of these steps included handoffs to staff or scientific reviewers that could increase the wait time for the investigators by days or weeks. Data were collected on randomly chosen applications and it was determined that the average time from application to initiation of a protocol in the CRC was close to 60 days.

What were the steps in the process that accounted for this unacceptable delay? We evaluated the course of the 198 applications that were submitted to our University and Children’s CRCs over the 12 months preceding the RPIW ( Figure 1 ). Only 15% were approved outright, 79% were “approved with contingencies” that required the investigator to make a response back to the review committee (sometimes multiple times), and 7% were not approved due to an incomplete application or serious flaws perceived by the reviewers. Eventually, however, 98% of all applications submitted were approved and initiated on the respective units. The RPI team focused on dissecting the scientific approval step as the source of the greatest wait time/waste. The SAC recommendations from several applications were reviewed and it was determined that; (1) there was “scope creep” of the SAC comments with Institutional Review Board (IRB) responsibilities resulting in requests for information that were redundant with the IRB application, (2) many comments by the SAC were scientific suggestions that did not fundamentally change the content of the protocol and were often additive to the work proposed (thus potentially unfunded), and (3) many of the protocols had already undergone an external scientific review either by an NIH study section or national academic committee. Indeed, 41% of applications for adult clinical studies and 13% of pediatric applications had a previous high quality academic review and an assigned “priority score.” Using this and other data collected during the workshop, the RPI team was able to now map the “Ideal State” of the CRC application process ( Figure 2 ), that is, create a shared vision of what the process should be. The group worked to discard steps that were of no value to the customer or client, which they defined as the research subject (with the investigators and their staff coming in as a close second). The entire process was decreased from 41 to 28 steps and the 6 handoffs were decreased to 2. The creation of the “Ideal State” required each stakeholder participating to compromise and give up some control over the process. In the end, all members of the RPI team needed to agree on the final product. If successful, reaching the “Ideal State” could reduce by half the length of time to approval and study start‐up. Moreover, the number of contingency acceptances could be reduced by 80%.

Figure 1.

Impact of the RPI Workshop on the CRC application process. Applications are either approved (white bars), approved with contingency (gray bars), or not approved (black bars). Data were shown for the 90 studies evaluated 12 months prior to the RPI and the 80 studies evaluated 12 months after the RPI. Data were shown as the percent of applications.

Figure 2.

The “Ideal State” for protocol review and acceptance. The graphic displays the people involved in the application process (multi‐colored vertical bar) and the steps required to complete the process (boxes). Application handoffs and wait times are shown by the arrows that deviate from the main work flow path.

With the “Ideal State” of the application process mapped, the RPI team made several recommendations that were immediately initiated. A lengthy application document was discarded and the investigators were asked to submit only their protocol, completed IRB application, a copy of any external scientific review, and a short form describing their use of the CRC for review by CRC staff and the SAC. The form was reconfigured to be an online submission with click boxes to reduce need for writing text. The data collected before and during the RPI was reviewed with all SAC members and delineation of SAC scope of work, as different from IRB scope of work, was discussed and clarified. The review committee agreed that studies would be approved outright unless there were serious flaws that impacted the ability to get useful data or put research subjects at risk. Finally, those studies that were previously critiqued by an external scientific group could be administratively approved by the Medical Director of the individual research unit if that person felt the reviews adequately assessed the protocol. Such studies could become activated quickly after application submission. Other suggestions included convening the protocol implementation meeting prior to the review, since nearly all protocols were accepted eventually, so that staff could assist in the application process and perform their assessment with the investigator or their research staff present to answer questions. Moreover, all files would now be maintained electronically with the investigator never having to supply any paper copies. These files could then be used by regulatory as well as clinical research staff on the CRC housed on a secure server. Figure 1 shows the results of these improvements. The approval rate of the 80 applications received in the 12 months after the workshop increased six‐fold, to 82%. Only 12% of applications were approved with contingencies that would require a re‐review and 6% of the applications were not approved. The average time from application to study initiation decreased nearly 50%.

In this example, there clearly was a tangible benefit to the application of continuous process improvement to research practices at our institutions. The changes happened quickly and for both investigators and staff it was a significant win. Investigator satisfaction with the application process greatly increased in subsequent surveys and, in general, the ability to maintain the “Ideal State” has been achieved in the long term. There have been, however, substantial intangible benefits to this approach. The process of creating a diverse team to work together 8–9 hours a day over a 1 week period created a deeper understanding for each team member of how the CRC functioned which led to different conversations about other aspects of clinical research and the development of additional projects, many well outside the walls of the CRC. Each team member contributed equally to the solution during facilitated discussion that resulted in a greater likelihood of maintaining the new structure. Finally, participants were motivated by data driven decision making and eager to see if the recommendations they had made resulted in any impact. These intangible benefits, relationship building, fearless communication, and enthusiasm for change are essential elements for a dynamic service oriented research culture. This type of culture will be able to adapt to and support the ever‐changing landscape of translational research.

We are using Lean tools to transform and modernize our resources and to create an adaptable infrastructure that will be flexible enough to take on any research related challenge. Our experience with the Lean philosophy, however, has encouraged us to explore the tools for other purposes such as the development of projects that would create team science or defining the “Ideal State” of a disease specific research process and create a plan to achieve that ideal state. Solving complex scientific problems could benefit from tools that simplify and clarify the process. In our goal to eradicate the diseases that most impact all of us, we need every tool in the toolkit.