Eliminating waste in health care delivery has been a long-time driver of our efforts to improve the quality of care we provide. Previous work to identify waste demonstrated that more than 50% of care delivery can be considered waste.1 That same study estimated that an average of 35% of waste at the front line is recoverable. Although we know that waste in care delivery exists, identifying the opportunities to eliminate waste at the front line has proven to be a challenge.

Observing workflows with time-and-motion study techniques based on Lean concepts has been an effective tool that we have used to identify opportunities to reduce waste. Previous work done by Wallace and Savitz1 used concepts from the Toyota Production System to identify categories of waste in health care and then conducted structured observations of frontline workers to identify the most frequent categories of waste occurring at the front line. Through their observations, they were able to estimate the costs associated with recoverable waste. In our current environment where cost containment is a large national focus, such work is an important component of waste elimination efforts to address not only the quality of care but also the cost of care.

Historically, a barrier to this type of workflow observation has been the time and resource intensity required to observe, document, code, aggregate, and analyze the data produced from the Lean observations. The process required trained observers with a background in the clinical area being observed to manually record, time, and track the workflow of caregivers. Often, these observations have been paper-based with manual time-keeping-requiring large amounts of post-observation time to code and enter the data into a usable database. More recently, the observations have been done on computers, which has minimized some of the post-observation data entry work but still required highly trained observers to accurately capture the required data and later clean and format the data for analysis. The laborious process for Lean workflow observation led us to question the value of using Lean observation techniques to identify sources of waste across our system if the process itself was so laborious and wasteful. While Lean observation was considered to be the best technique for this work, it would be impossible to get the system-wide support needed to tackle waste reduction at our front line with such a labor-intensive process to do so.

In a system such as ours, with rich data sources related to patient care, we knew the information available through the electronic health record and the electronic data warehouse could not provide the data required for this work-the systems were not designed to do this. We had to design a system that was easy to use by minimizing the data collection burden and make it easy to aggregate and analyze the data. When Drs Wallace and Savitz did their initial work to estimate the cost of waste, they developed a computer-based data collection tool that made the documentation and categorization of observed frontline activities easier, therefore making the data reporting and aggregating easier. Following the development of the computer-based data collection tool, and with the rapidly expanding use of mobile electronic products such as tablets and smartphones, we began work to design a tablet-based tool based on the development and refinement of the computer-based tool in 2013.*

The design of the tablet-based tool focused on 3 key goals aimed at simplifying the observation and data collection process. First, we wanted to minimize the use of free-text data entry during the observation. This was intended to minimize the data entry requirements during the observation and, most importantly, maximize the computability of the collected data immediately following the observation. This led to developing customizable picklists of common activities performed on the basis of the setting of the observations being done. Before beginning the observations, the picklists can be compiled by the study team, which can then be exported to all tablets being used in the study. The tool allows for manual entry of any activities not included on the list. Any common activities that emerge during the observations that were not included on the original picklist can easily be added and exported to all tablets used in the study.

Second, we wanted to be able to instantaneously provide initial feedback to the frontline workers being observed. The original work done by Drs Savitz and Wallace included post-observation debriefings with the observed workers. These debriefings were well received and noted as an important part of the observation work, as it provided immediate feedback and value for participating in the observation. By focusing on minimizing free-text data entry in the tablet-based tool, the application can generate summary charts and tables of the data collected during the observation that can be immediately reviewed postobservation.

Third, the application needed to support commonly used tools and programs for data aggregation and analysis. Although the summary charts and tables produced in the application for each individual observation are of interest to those being observed, the power of the observations come from the aggregated data of the entire study. We recognized the need to have the data collected by the table-based application easily exported to computer-based tools and software commonly used for data analyses. This critically important requirement of the tool led to another key feature. By design, we wanted the tool to be functional in the absence of Internet connectivity-we wanted the tool to be usable in any setting. However, we needed the tool to easily export the collected data to a centralized computer or database. As a result, all data are automatically saved locally to the tablet and then exported in a preferred format such as .csv to a shared drive such as Dropbox or a common e-mail address once Internet connectivity is available.

Initial internal testing and use of this new, tablet-based tool were met with positive feedback and used as a part of our CMMI-funded Hospital Engagement Work (HEN). The tool eliminates the need to manually track time during the observations. Previously, observers were required to document timing of activities using a stopwatch and manual documentation. Now, the application automatically time-stamps all recorded activities-including duration of activity as well as time of day. It even accounts for interruptions to activities that are an important category of waste. This allows the observer to focus on the observation instead of tracking timing of activities. In addition, training of observers has been streamlined. By relying on the tested format of the tool and using study-specific picklists, training can focus on observation content as opposed to standardized documentation and observation background. Finally, data aggregation and analyses can occur shortly after the observation-allowing for waste elimination activities to occur more quickly. The data format also makes it easier to link the observation data to other data sources such as time and hourly rates to conduct costing analyses.

Nursing and patient flow teams at Intermountain have used the tool for structured observations to identify and reduce waste at the front line. Several of our medical-surgical nursing units had a goal to increase the amount of time that nurses spend with their patients. To accomplish this goal, nurses and their managers needed to first understand how much time they were spending with patients. Second, they wanted to understand the barriers that kept nurses from spending more time at the bedside. To help facilitate this goal, these units partnered with our Continuous Improvement and nursing leadership teams to perform time-and-motion studies on each of these units.

The nursing leadership team hired a part-time data manager (with no prior continuous improvement training) to perform time-and-motion studies in these units. The data manager used the Lean Observer application on an iPad to perform observations in each medical-surgical unit at Intermountain's largest hospitals. Over a 1- to 2-week period, the data manager spent a total of 16 hours on each unit observing 8 nurses, 2 hours at a time. After completing the observations for a given unit, the data manager used the application to summarize the observations and e-mailed observation results to unit managers.

The results of these observations provided actionable data for each unit manager and team of nurses. For example, observations showed that nurses were spending a lot of time traveling from patient rooms to the supply room in one particular unit (especially when compared with travel times in other units). As a result of this information, the nurse manager worked with the facility to place several supply carts throughout the unit, which dramatically reduced the amount of travel time for these nurses. The team then created a process to ensure that these carts were continuously stocked with the most frequently used supplies.

In general, the time-and-motion studies facilitated by the application did not uncover 1 or 2 glaring inefficiencies; rather, the application helped unit managers see that there are many small areas where improvements could be made. Information from these time-and-motion studies are still being used to help unit managers and nurses identify inefficiencies, continuously improve, and, ultimately, spend more time with patients.

With a shared goal to improve the quality and reduce the cost of health care, we wanted to make the tool available to you. It is designed to be customizable for use in a variety of settings and systems. The Health Quality Lean Observer app is available for iPad as a free download on the Apple App Store. We invite you to explore the tool and look for ways it may support your work to eliminate waste and continually improve the quality of care in your system.

REFERENCE