Abstract: With the increased emphasis within the pharmaceutical industry on business productivity through the dual application of Lean Six Sigma and disciplined project management methodologies, there is a need to provide a harmonized project life cycle roadmap for productivity project teams. This paper compares the Lean Six Sigma Define Measure Analyze Improve Control (DMAIC) project life cycle with that suggested by the Project Management Institute (PMI) in the form of a guide to Project Management Book Of Knowledge (PMBOK ) 1. The paper illustrates areas of commonality and suggests a possible harmonization in the form of DMAI 2 C 2. The objective is to assist productivity improvement project teams create a project roadmap that leverages the best of both methodologies. The reader is advised that while DMAIC and PMBOK methodologies are purposely generic so they can be applied to a wide range of projects beyond those of interest to the pharmaceutical industry, this paper is limited in its scope to projects aimed at business process productivity improvement typical to pharmaceutical product development and prior to manufacturing of an approved drug product. Introduction: Lean Six Sigma (LSS) and the PMBOK agree that a project has (i) a distinct life cycle with clearly defined start and end, and (ii) it follows a prescribed logical progression from start to end. However, each methodology prescribes a different set of phases to prescribe a project s life cycle. These phases are outlined in Tables I and II, which may be found at the end of the paper. The intent of this paper is to suggest an integration of the objectives of the different phases into a harmonized life cycle roadmap to assist productivity improvement project teams create a project roadmap that leverages the best of both methodologies. Before the methodologies are compared and harmonization is suggested, it is necessary to acknowledge that pharmaceutical R&D processes are unique. The evolution of a drug product requires the integration of diverse scientific and business disciplines, over a very long time line, and under a highly regulated environment. Recent economic pressures have pushed the pharmaceutical sector to leverage lower cost regions and, in doing so, further complicated the workflow by adding global touch points for knowledge and data transfer. It is this uniqueness that is often the principal difficulty in introducing LSS to a pharmaceutical R&D organization. However, as this paper will illustrate, harmonization of DMAIC and PMBOK principles will enable LSS practitioners within the pharma industry to more efficiently achieve project success. Discussion: Lean Six Sigma projects typically employ the Define Measure Analyze Improve Control (DMAIC) life cycle. These are described briefly in Table I. The PMBOK 2 employs the Initiating Planning Executing Monitoring&Controlling Closing life cycle. These are described briefly in Table II. It should be noted that Tables I and II are intended solely as a brief overview for readers who may not be versed in either topic. The tables are not meant to be a substitute for formal training and several outstanding text books available in the open literature. Since the intended audience for this paper is LSS practitioners, the proposal expands the classical DMAIC methodology by borrowing and integrating select PMBOK process groups and knowledge areas and suggesting a 1 PMBOK and PMP are registered trademarks of the Project Management Institute (PMI) 2 This paper relies on the Fourth edition of the PMBOK 1
new life cycle: DMAI 2 C 2. The following matrix is an overlay of the resulting DMAI 2 C 2 with PMBOK. Each of the project life cycle phases are described below. PMBOK Project Life Cycle Suggested Lean Six Sigma Project Lifecycle Phases Define Measure Analyze Improve Implement Control Close Initiating X Planning X X Executing X X X X Monitoring & Controlling X X X X X X X Closing X Define is that phase of the LSS project during which the process that requires productivity improvement is defined in order to provide a scope/boundary for the project team. Since DMAIC evolved from discrete manufacturing, typically (but not always), the nature of the problem has been narrow and specific, such as the yield from a CNC machine or the throughput of a manufacturing cell 3. Analogous pharmaceutical product development examples may include unacceptable throughput from a syringe filler or unacceptable yield from a label machine used to manufacture clinical samples 4. However, when applying LSS to improve the productivity of an R&D business process, such as the effort around the planning of a clinical study or the effort around the process of authoring and submitting an IND application, the classical Define takes a whole new and broader meaning. Therefore, it is necessary to expand the classical Define phase to borrow selected knowledge areas associated with the Initiating and the Planning process groups which together offer some 22 knowledge areas (see Table II) 5. It is suggested that LSS practitioners initially approach the PMBOK knowledge areas as a checklist of items to consider while authoring the project Charter. As different areas are determined to be applicable to their project, the LSS practitioner may wish to take the time to familiarize him/herself with the knowledge areas leveraging the resources available on line from PMI and others. It should be noted that LSS and PMBOK agree in two very important respects. First, both require a project Charter without which a project should not proceed (although the template/content of the Charter differs between the two methodologies). Second, both require the project leader to identify and engage the project Stakeholders whose support is critical to project success. An area where the two differ is the scope of the project plan. The PMBOK methodology requires that a very detailed and thorough project plan be written and, in fact, this importance is demonstrated by a distinct Initiating phase. The LSS methodology requires a charter 3 A good reference book on this topic is The Goal by Eliyahu Goldratt. 4 This is another example of the uniqueness of pharmaceutical product development. Whereas R&D does not manufacture drug products, they are required to manufacture products that can be used in clinical trials and, therefore, such processes are of interest to productivity teams since they impact both direct and indirect costs. 5 We use the term selected in acknowledgement of the reality that pharmaceutical R&D processes range from finite processes through multi functional workflow processes that stretch the classical intent of DMAIC. This is where the judgment and experience of the project lead becomes important to selecting an appropriate subset of knowledge areas. 2
without providing much guidance regarding project planning. Therefore, It is suggested that, at a minimum, a LSS project charter include: The SIPOC of the process under study A delineation of what is and what is not in scope for the team A delineation of the relevant KPI s and their associated to be values relative to the baseline A listing of stakeholders and project team members A delineation of the project lead and the project sponsor A timeline for the project including key decision gates, and, when the project s duration is measured in months (rather than days to weeks), a timeframe for creating a suitable project plan Measure is that phase of the LSS project during which data describing the as is performance is collected. This is a critical element/characteristic of a LSS project. The principle here is that without data showing the existing capabilities (baseline), it is impractical for a project team to eventually demonstrate to management that the solution the team has come up with has in fact improved the situation. However, whereas it is common for discrete manufacturing processes to continuously measure output (e.g., throughput, yield, cost, etc), this is not true for most R&D business processes. Enterprise Resource Planning (ERP) systems are most often tailored to meet the needs of Finance and Purchasing and do not lend themselves to providing meaningful (actionable) business process productivity information to R&D managers. And, while there may be timesheet systems in place, the fact is that other than anecdotal statements, there are seldom historical data of the Key Performance Indicators (KPIs) of interest to R&D process managers responsible for achieving high productivity. Therefore, LSS productivity teams find that too often they need to plan, develop, and implement an appropriate data collection project just so they can establish a factual baseline against which future changes in process can be measured and compared. This is the reason why the matrix shows an overlap between Measure and both the Planning and the Executing phases of PMBOK. Since the Measure phase may require a staff to collect data and to create process maps, a more detailed planning of the activities in this phase may be warranted. Specifically, and as suggested by the PMBOK under Planning, details around team size, roles and responsibilities, timeline, and deliverables should be developed and discussed with the project sponsor to set expectations. In addition, and depending on the magnitude of the process under consideration and whether historical performance data is readily available, the Measure phase could be a mini project in its own right with a significant Executing component that requires application of PMBOK s Monitoring & Controlling knowledge areas. A cautionary note is in order here. As Table II illustrates, the PMBOK offers a lengthy (and evolving) list of knowledge areas (tasks) that the project team should perform during the Planning and the Executing phases. Here, once again, the experience and judgment of the LSS project leader (or the LSS PMO) is critical to selecting the appropriate subset of knowledge areas consistent with the scope/magnitude of the process under study. Not every LSS project requires the application of every PMBOK knowledge area. Therefore, as was suggested earlier, the LSS practitioner should review the PMBOK knowledge areas as a checklist of items to consider when first authoring the project Charter. As different knowledge areas are determined to be applicable to their 3
project, the LSS practitioner may wish to take the time to familiarize him/herself with the knowledge areas leveraging the resources available on line from PMI and others 6. Analyze is that phase of the LSS project during which the team analyses the data collected during Measure in order to understand as is performance and inform root cause analysis. The most critical aspect of this phase is to ensure that the team s focus is on root cause(s) and not symptoms. Too often, project teams are made up of scientists with their creative nature who use this phase to architect new solutions and to decide how best to employ a beloved 7 solution instead of spending their time on disciplined process mapping, value stream mapping, and root cause analysis. An effective practical approach to Analyze is the Kaizen event as long as the majority of the participants are process actors who actually perform the as is work 8. Improve is that phase of the LSS project during which the team formulates alternative solutions. Having completed the root cause analysis, the team s creative juices are directed at coming up with process changes to mitigate/eliminate the root causes. This is most often accomplished during the Kaizen event and it may require additional participants with skills in information systems, automation, and/or process re design. An integral component of this phase is a formal management stage gate to ensure that the process stakeholders are aligned with the proposed changes before these are implemented. The Communication knowledge area of PMBOK is critical to the success of this phase. Implement is the phase of the LSS project during which the team implements the approved process changes. Depending on the magnitude of the process undergoing change, Implement is often a project in its own right, typically requiring development and staffing of a work breakdown structure to develop processes, to design and implement systems and infrastructure changes, to create procedures and associated training, and to complete a rollout with minimal adverse impact to the productivity and quality of on going business. It is suggested that if the duration of the implementation phase is measured in months (rather than days weeks), a more detailed project plan be written and the knowledge areas of the PMBOK be consulted and selectively adopted and employed by the team. Sponsors should note that at the start of Implement (i.e., after the Improve Stage Gate), the sponsor may wish to replace the project lead, who has brought value to the team through his/her Lean Six Sigma skills and expertise, with a project management professional having PMBOK skills and expertise. Control This phase is unique to Lean/Six Sigma projects in several regards. First, it introduces the concept of process owner, which is not commonly/broadly used within Pharma R&D. Without an accountable process owner, the long term success of process changes are less likely to be recognized since no one is held 6 This is an area where an effective PMO will, over time, compile lessons learned and develop multiple templates that LSS project leads can use to get a jump start on which PMBOK knowledge areas apply to what types of LSS projects. 7 The term beloved is used to capture such euphemisms as technology, automation, outsourcing, etc. solutions that management determined are the correct approach. If this is the case, the M and A and I phases are not needed. 8 It cannot be overemphasized that the success of a LSS project seeking productivity gains is directly related to the Kaizen participants. Process actors, the individuals performing the tasks day in/day out, are preferred over low to mid level managers who lack an understanding/appreciation of the process pain points. 4
accountable for monitoring and sustaining the changes. Second, LSS introduces the control chart concept that is used most often by Six Sigma practitioners in the discrete manufacturing domain. While control charts can come in many forms, the basic tenet of providing the process owner with a dashboard by which to routinely gauge productivity trends 9 is as valid for R&D processes as it is for discrete manufacturing. The challenge is to determine what KPIs to track and what system to put in place to collect the necessary data without undue burden to the process actors. Finally, the Control phase normally marks the end of the LSS project in that at the end of this phase, the LSS project team is disbanded and the implemented process changes are rolled into routine operations. This is where the classical definition of DMAIC needs to expand to add the Closing phase from PMBOK. Close This phase is unique to PMBOK, but is very much applicable to LSS productivity projects. It is even more relevant now that many in the pharmaceutical industry are pursuing the concept of a learning organization. By leveraging the PMBOK knowledge areas associated with Closing, the organization (LSS PMO) can develop a meaningful and value add archive of data from completed productivity projects and, in doing so, improve the efficiency of planning and executing future productivity projects. Finally, the matrix shows that the Monitoring & Controlling process group of PMBOK touches every phase of DMAI 2 C 2. This is because DMAI 2 C 2 is the project, and the LSS project lead has to monitor and control every aspect of each of the D.M.A.I.I.C.C phases. Summary: With the increased emphasis within the pharmaceutical industry on business productivity through the dual application of Lean Six Sigma (LSS) and disciplined project management methodologies, this paper offers an expanded definition of the classical DMAIC by incorporating aspects of the PMI s PMBOK methodology and creating the DMAI 2 C 2. It is expected that the suggested harmonization in the form of DMAI 2 C 2 will assist productivity improvement project teams create a project roadmap that leverages the best of both LSS and PMBOK methodologies. Specifically, it is suggested that LSS practitioners approach the PMBOK knowledge areas as a checklist of items to consider when authoring the LSS project Charter. The reader is reminded that while the LSS DMAIC and the PMI s PMBOK methodologies are purposely generic so they can be applied to a wide range of projects beyond those of interest to the pharmaceutical industry, this paper is limited in its considerations to projects aimed at business process productivity improvement typical to pharmaceutical product development prior to manufacturing of an approved drug product. About the author Michael Herskovitz obtained his certification in Six Sigma Black Belt in 2007 and his certification as Project Management Professional in 2010. During the past 12 years, Michael directed business process simplification projects in discrete manufacturing and in Pharma R&D. 9 gains/losses over time 5
TABLE I Overview of Lean Six Sigma DMAIC Key objective 10 Key deliverable Define Measure Analyze Improve Control Collect data that Evaluate the describes the data that was performance or collected with productivity of emphasis on the as is identifying root process causes Establish the scope and purpose of the project Project charter 12 Process flow map, Value Stream Map, asis performance measures 13 A team consensus around root causes in priority order Based on the analysis, evaluate alternative changes to the as is process to mitigate the adverse root causes A Benefit Effort chart of proposed changes along with a suggested to be process flow 14. A project plan delineating the implementation of the approved changes. A pilot or other form of agreed upon validation that the changes will yield the desired effect. Establish a mechanism (dashboard) by which the process owner is able to monitor the process going forward to ensure the productivity are realized and the organization does not drift into past practices 11 Control dashboard. Formal transfer of the process from the Lean sigma team to the process owner. Closure of the project. 10 The intent of this table is not to provide an extensive detail on DMAIC. Instead, key objectives and deliverables are described 11 The classical intent of Control is to demonstrate that the changes made to a manufacturing process returned a process that was out of control (variations, quality, inconsistencies) back into control. For Lean, the need for a dashboard remains valid, although the mechanism and data may differ. 12 The template and content of a project charter are key to project success 13 In typical productivity improvement projects, the data of interest includes cycle time, lead time, first pass yield, throughput, level of effort, cost 14 Typically, the proposed changes are the outcome of a Kaizen event and are subject to formal Stage Gate review and approval. Only the approved changes are authorized for implementation 6
TABLE II OVERVIEW OF PMBOK 15 PROJECT MANAGEMENT PROCESS GROUPS AND KNOWLEDGE AREAS Initiating Planning Executing Monitoring& Controlling Closing Key objective 16 Develop the Charter Develop a detailed project management plan Direct and manage the project following the approved plan. The principle is plan the work and work the plan. Monitor progress against plan, and perform integrated change control Archive project information Knowledge Areas Develop Project Charter. Identify Stakeholders. There are 20 knowledge areas, including: collect requirements, define scope, create WBS, define project activities, sequence and timeline, estimate budget, plan communication, and project risk management. There are 8 knowledge areas with emphasis on quality assurance, team staffing and leadership, and communication with stakeholders There are 10 knowledge areas, including change control, verify scope, report performance, and monitor and control project risk Close project, including any procurements 15 Fourth edition 16 The intent of this table is not to provide an extensive detail on PMBOK. Instead, key objectives and deliverables are described 7