MCRU 2U Workflow Analysis

Final Report

Submitted To:

Cyndi Bower

Administrative Director

Michigan Institute for Clinical Research

1500 E. Medical Center Dr.

Ann Arbor, MI 48109

Margaret McCammon

Research Translation Specialist

Michigan Institute for Clinical Research

1500 E. Medical Center Dr.

Ann Arbor, MI 48109

Amanda Silva

Central Lean Coach and Coordinator

University of Michigan Health System

2101 Commonwealth Blvd, Suite A

Ann Arbor, MI 48105

Chris Kafantaris

Industrial Engineering Fellow and Coordinator

Program & Operations Analysis

2101 Commonwealth Blvd, Suite A

Ann Arbor, MI 48105

Professor Mark Van Oyen

Associate Professor

University of Michigan Industrial and Operations Engineering Department

1205 Beal Ave.

Ann Arbor, MI 48109

Submitted By:

Team 9

Joseph Beck, Emily Lozon, Shair Malik

Date Submitted: December 10, 2013

Table of Contents

Executive Summary...... 1

Project Goals...... 1

Methods...... 1

Findings...... 2

Process Variation...... 2

Volume through Processes...... 2

Conclusions...... 3

Conclusions from Process Variation...... 3

Conclusions from Volume through Processes...... 3

Recommendations...... 3

Short Term Recommendations...... 3

Long Term Recommendations...... 3

Introduction...... 4

Background...... 4

Key Issues...... 6

Goals and Objectives...... 6

Project Scope...... 6

Methodology...... 6

Observations and Time Studies...... 7

Research for Workflow and Staff Assignment and Allocation...... 7

MCRU 2U Employee Survey...... 8

Historical Data Set Analysis...... 8

Methods for Data Analysis...... 9

Findings and Conclusions...... 10

Process Variation...... 10

High Variation in Process Steps...... 10

Variation in Protocol Locations...... 12

Variation of Staff Performance and Opinions...... 12

Conclusions from Process Variation...... 15

Volume through Processes...... 15

Time Range Analysis...... 16

Location Analysis...... 17

Protocol Analysis...... 18

Conclusions from Volume through Processes...... 19

Recommendations...... 20

Short Term Recommendations...... 20

Long Term Recommendations...... 21

Expected Impact or Outcome...... 21

Understanding Process Variability...... 21

Understanding Volume through Processes...... 22

Appendix 1: Time Studies Data Collection Sheet...... 23

Appendix 2: MCRU Survey...... 24

Appendix 3: Protocol Trends...... 26

Appendix 4: Further Survey Results...... 29

Appendix 5: Average Man Time...... 33

List of Figures and Tables

Figure E-1: MCRU 2U Workflow Process...... 2

Figure 1: Trigger Process for Mobile Clinical Research Unit...... 5

Figure 2: MCRU 2U Workflow Process...... 10

Figure 3: Interval A has Highest Average Process Time and Spread...... 11

Figure 4: Extended MCRU 2U Workflow Process...... 11

Figure 5: Over Half of MCRU 2U Staff Does Not Take Backpack...... 13

Figure 6: MCRU 2U Backpack is Disorganized and Heavy...... 14

Figure 7: MCRU 2U Staff Agree Walking Takes a Good Portion of Service Time....15

Figure 8: Tuesday and Thursday Receive Highest Number of Calls...... 16

Figure 9: Five Protocols Contribute 31% of Mobile Visits...... 18

Figure 4-1: Most Employees go on 3 to 5 Mobile Visits per Week...... 29

Figure 4-2: MCRU 2U Staff Agree that Number of Calls Vary Daily...... 30

Figure 4-3: MCRU 2U Staff Believe Assigned Scheduling will Interfere with Work...31

Figure 4-4: MCRU 2U Staff Disagree that there is enough staff to meet MCRU 2U Demand 31

Figure 4-5: MCRU 2U Staff are Neutral Regarding MCRU 2U Interfering with Daily Work 33

Figure 4-6: MCRU 2U Staff Disagree that large portion of their job involves MCRU 2U33

Table 1: Duration by Location and Process Step (minutes)...... 12

Table 2: Average Call Frequency by Time...... 17

Table 3: Average Call Frequency by Location...... 17

Table 4: Trends for the Top Five Serviced Protocols...... 19

Table 5: Expected Man Time and Staffing Requirements at 80th percentile (minutes)...... 20

Table 3-1: Protocol Frequency Data...... 26

Table 5-1: Average Man Time (minutes)...... 33

1

Executive Summary

The Michigan Clinical Research Unit (MCRU) provides clinical staff and resources to perform human clinical trial research protocols at the University of Michigan. The Administrative Director reported that an increase in demand (from 15 to 2700 visits since 2007) has led to a chaotic and inefficient model for one of their core clinical services, the Mobile Clinical Research Team (MCRU 2U). Workers within MCRU report inefficiency and high variability in the workflow process, which have resulted in unpredictability in demand, demand exceeding staff capacity, and research protocols that interrupt the daily duties of nurses and medical assistants. To address these problems, workers within MCRU 2U asked an IOE 481 student team from the University of Michigan to conduct a series of observations and analyses to help understand current workflow and recommend an improved workflow. The team was also asked to determine if the current model can be scaled to fit current and future demand or if MCRU 2U needs to be offered as a separate service with more allocated resources.

Project Goals

To help sustain and improve the current MCRU 2U model, the student team achieved the following tasks:

  • Conducted a series of studies to understand the current workflow and demographic trends within MCRU 2U
  • Identified sources of inefficiencies, redundancy, disorganization, and variation

Methods

The team performed four types of tasks to evaluate and improve the current MCRU 2U workflow process.

  • Performed Time studies. First, the student team shadowed10 mobile visits to understand time between processes and to understand process variability. The team calculated average duration and standard deviation for each process step.
  • Administered MCRU employee surveys. The team developed and administered a survey to 9 MCRU 2U staff to quantify the opinions of the staff regarding MCRU 2U. The student team further analyzed surveys to reflect on process variability from staff perspective.
  • Analyzed historical dataset. The team analyzed demographic trends from 1590 mobile visits dated between June 2012 and May 2013. The team received this information from the Research Translation Specialist and used it to answer the following questions:
  • What day of the week receives a majority of MCRU 2U calls? What is the variability in calls on these days?
  • What times are a majority of the calls being received?
  • What location is being serviced the most based on time of day?
  • What location is being serviced the most based on day of week?
  • What protocols are serviced the most?
  • What protocols are serviced on the same day within 30 minutes (back-to-back)? How often?
  • Developed recommendations for improving the workflow. The team created a list of steps that MCRU 2U could take to understand their demand better and minimize process variability.

Findings

Analysis of the time studies, surveys, and historical data set were divided into two main categories: Process Variation and Volume through Process.

Process Variation

Analysis of time studies highlights certain problems within the MCRU 2U workflow process. Process steps in Figure E-1 were grouped into different intervals to help identify these problems. Figure E-1 showsthat traveling from MCRU to the site and traveling back to MCRU after the completion of a service accounted for 7 and 9 minutes on average, respectively. Intervals A1 and B2 have high standard deviations, at 6 and 5 minutes, respectively. The high standard deviations means that there is high variation from the time a call is received to when supplies are collected and from the start of a service to the end of a service.

Figure E-1: MCRU 2U Workflow Process

October 9-19, 2013; N = 10; Time Studies

Volume through Processes

Findings from the historical data set show that Tuesdays and Thursdays had the highest average number of calls, at 7.32 and 7.10 calls, respectively. In addition, Thursday had the highest variability in calls while Monday had the lowest. Analysis of different time blocks showed that Tuesdays between 11:00 and 13:00 had highest average number of calls at 2.84 calls. Finally, analysis of different locations indicated that Mott Hospital, on Thursdays, had the highest average number of calls at 3.09 calls.

Conclusions

The following section provides the conclusions formed after analysis and evaluation of the findings, which are based on Process Variation and Volume through Processes

Conclusions from Process Variation

Analysis of the process variation data showed that there is high variation from the time a call is received to the time staff starts their service, specifically when the calls come from the University Hospital and Mott Hospital. In addition, the amount of time MCRU 2U staff waits to start a service upon arrival at the site needs to be shortened. Finally, walking constituted over 40% of the total process time and needs to be reduced.

Conclusions from Volume through Processes

Analysis of the demographic trends showed that MCRU 2U should schedule more staff more Tuesdays and Thursdays, specifically between 09:00 and 13:00. MCRU 2U should plan to service the Taubman Center primarily on Mondays, Tuesdays, and Wednesdays, and Mott Hospital on Thursdays and Fridays. Finally, analysis of all protocols showed that 11% of all protocols are contributing to 50% of total mobile visits from June 2012 to May 2013.

Recommendations

The student team developed a list of short term and long term recommendations to help MCRU 2U increase the predictability of their process and meet demand.

Short Term Recommendations

The short term recommendations are actions MCRU 2U can implement in the near future. The team recommends that MCRU 2U implement guidelines and organizational methods that help control the variable demand and predict future calls. This can be done by making the following changes:

  • Screen investigators during the initial call
  • Add storage units at the most frequently served locations, such as the Taubman Center and Mott’s Hospital
  • Develop new organizational method for backpack
  • Develop nurse assignments based on demographic trends
  • Regulate calls by establishing guidelines with PIs

Long Term Recommendations

The long term recommendations focus on continual improvement and sustainability of the MCRU 2U process.

  • Clean and validate data regularly so it is easier to understand for all users
  • Record and input data using a standardized method (create organized excel template, discuss with employees the importance of recording accurate measurements)
  • Regularly analyze (every 3-4 months) trends between times, locations, and protocols to understand workflow and to assess the need for potential improvements
  • Perform further studies to analyze demand and potential areas of improvements

Introduction

The Michigan Clinical Research Unit (MCRU) is a unit within University of Michigan Health Services (UMHS) that provides clinical staff and resources to perform human clinical trial research protocols at the University of Michigan (UM). Many of these research protocols take place at UM’s Cardiovascular Center (CVC). In 2007, MCRU implemented a Mobile Clinical Research Support Team (MCRU 2U) as an addition to its core services. MCRU 2U staff provides clinical research support to clinical investigators. Instead of having participants come into the CVC to perform these protocols, MCRU 2U staff goes to the participant. Since its implementation in 2007, the service has grown from 15 to 2700 visits annually through word of mouth.

The Administrative Directorof the Research Innovation Core reported that an increase in demand for mobile visits has led to a chaotic and no longer efficient model. Workers within this unit report inefficiency and disorganization in the workflow process. Inefficiencies result in high variation and unpredictability in the workflow process, demand exceeding staff capacity, and research protocols that interrupt the daily duties of nurses and medical assistants.

To address this problem, workers within MCRU 2U asked an IOE 481 student team from the University of Michigan to conduct a series of tests to analyze the current workflow and to provide an improved workflow. After conducting observations, interviews, and further analyses, the team has determined if the current model can be scaled to fit current and future demand, or if MCRU 2U needs to be offered as a separate service with more allocated resources. This report details the student team’s data collection and analysis methods, and presents the team’s findings, conclusions and recommendations.

Background

MCRU, located in the Cardiovascular Center (CVC), is a unit within UMHS and the Michigan Institute for Clinical and Health Research (MICHR) that provides the clinical staff and resources necessary to perform human clinical research protocols at the University of Michigan. MCRU also serves as an institutional resource, allowing investigators to perform pilot studies. Investigators who are federally, philanthropically, and industrially sponsored are serviced by MCRU.

One of the core clinical services provided by MCRU is the mobile service, MCRU 2U. MCRU 2U was implemented in 2007 with hopes of providing improved access to clinical research support services with little increase in resources. In addition, the mobile service hoped to extend the practice of clinical research outside the MCRU, increase enrollment, and decrease dropout rates.

The clinical staff within the MCRU 2U team provides services and support to clinical investigators. The mobile team services investigators by responding to specialized protocols, which include PK blood draws, ECGs, etc.

The mobile services are offered through three routes: (1) walk-in, (2) just-in-time, and (3) prescheduled services. In the current process, mobile calls come in via telephone, and the responder gathers required information for the visit. After determining what clinical staff will perform the service, the clinical staff member gathers supplies and walks to the service site. After performing the service, the staff member returns to the CVC, delivers the sample and participant data to the laboratory, and documents the visit. Figure 1 details this workflow process.

Figure 1: Trigger Process for Mobile Clinical Research Unit

Adapted From: MCRU JIT Process, Administrative Director and Research Translation Specialist

The inefficiencies begin to surface when high demand for the services occur. MCRU 2U has limited staff and resources and attempting to meet the service demand can lead to disorganization and chaos. For example, there are demographic trends in which certain days, times, or sites will have low demand for the service, but other times will have high demand. The staff who perform these services must also perform their daily duties, which include servicing walk-in participants. The daily duties of the staff can be interfered with on a day with multiple protocol calls. The high amount of variation and unpredictability has prompted MCRU 2U to ask the IOE 481 team to look at demographic trends, inefficiencies and wasted work, and to offer a current and future workflow analysis.

The essential goal was to quantify the current process and identify potential areas of improvement. In addition, the team determined if the existing model can be scaled to meet current and future demand. This project helped identify these key issues and provide recommendations for an improved workflow.

Key Issues

The following key issues are driving the need for this project:

  • Demand has led to inefficiencies
  • Inefficiencies result in high variation and unpredictability in the workflow process
  • Protocol calls interrupt the daily in-clinic workflow for nurses and medical assistants

Goals and Objectives

To help sustain and improve the current model within the Mobile Clinical Research Unit, the student team achieved the following tasks:

  • Conducted a series of studies to understand the current workflow and demographic trends within the MCRU
  • Identified sources of inefficiencies, redundancy, disorganization, and variation

With this information, the team developed recommendations to:

  • Determine if the current model is sustainable or if additional resources are needed
  • Increase efficiency
  • Eliminate redundancy
  • Increase predictability

Project Scope

This project included only the MCRU 2U workflow process. The protocol process begins when MCRU 2U is notified of a protocol call and ends when the MRCU 2U staff member returns to the clinic, gives participant data to the lab, and documents the visit. The project scope included the 55 out of the 245 active protocols that require the mobile services of the MCRU 2U.

Any task not connected to the protocol process was not included in this project. Specifically, the team did not study research protocols not serviced by MCRU 2U, or activities associated with departments outside of MCRU 2U. The project did not focus on details regarding participant information nor did it focus on the details involving specific protocol services. The project did not focus on complex protocols such as EKGs.

Methodology

The team developed recommendations based on a select group of simple protocols. This adjustment in focus reduced variation in the workflow process and allowed the team to recommend changes to a more specific function of the MCRU 2U.

Observations and Time Studies

The team observed the MCRU 2U workflow process prior to the time studies to develop a data collection sheet (see Appendix 1) that matched the MCRU 2U workflow process. The team utilized this data collection sheet and a stopwatch to record start and end times for the steps in the workflow process depicted in Figure 1. Data was collected for two weeks. The team observed several staff members and performed 10 time studies on MCRU 2U mobile visits.

Mobile visit was further divided into the following steps:

  • Arrive at site
  • Start Service
  • End Service
  • Leave site

These steps were chosen because the team wanted to observe the activities before a protocol and the steps following a protocol to understand process variability. Specifically, the team focused on preparation and travel time to a site location to standardize the start and end of each step in the workflow process. Team members utilized the data collection sheet to record staff name, route type, site location, date and time, and notes for each step in the workflow process.

Each member of the student team conducted time studies individually. Data was collected from a random sample of nine MCRU staff. A student followed either a Registered Nurse (RN) or a Medical Assistant (MA) to a location site and recorded times while observing the process.

Research for Workflow and Staff Assignment and Allocation

The team conducted a literature search and each team member consulted three documents. The team searched through past IOE 481 reports and gathered ideas and methods, which have been incorporated into this project.