Optimizing emergency services with lean Six Sigma

Black belt project levels staffing, cuts response time and answers more calls

Optimizing emergency services with lean Six Sigma

By Casey Bedgood

In 2013, Navicent Health’s emergency medical services was a well-established provider in the state of Georgia with a strong focus and reputation for clinical excellence. However, it was experiencing a variety of operational inefficiencies, including an outdated pay system and a mismatch between staffing levels and call volume. This let the local competitor snap up numerous calls per month, costing Navicent revenue.

Navicent Health EMS is a fast-paced, 911 hospital-based, advanced life support provider licensed by the state of Georgia as both an ambulance service and neonatal transport service. Its 911 call volume exceeds 40,000 calls per year in its four-county service area, which includes urban, rural and suburban geographical response zones. The ambulance service dates back, most believe, to the early 1900s when Navicent Health was a small community medical center called the Macon Hospital.

The service provider has more than 100 staff members, a centralized communication center and staffs up to 20 ambulances per day for emergency responses, along with various community assignments. In addition, Navicent EMS transports nonemergent interfacility patients across the state and beyond. The EMS service provides advanced life support to all patients suffering any and all medicaltraumatic emergencies, ranging from neonate to the oldest geriatric.

Navicent’s management decided to tackle the EMS service problem by deploying a lean Six Sigma team for a black belt project, adding administrative changes to optimize the system. In 2014, Navicent Health’s first black belt led this interdisciplinary group of EMS leaders, administrators and human resources leaders down the path of continuous improvement by way of DMAIC (define, measure, analyze, improve, control) to reduce variation and optimize other operational key performance indicators (KPIs).

The emergency services optimization black belt project Sigma to improve critical-to-quality elements drastically, including emergency response times, turnaround times, customer satisfaction, cost of poor quality and wasted motion.

In addition, this black belt project saved a significant amount of money and optimized the system. The project positively impacted tens of thousands of customers across the multicounty service area, significantly reduced operational costs, improved employee satisfaction and reduced various types of unwarranted variation.

Define

The EMS system was experiencing a number of issues, including an outdated pay and staffing system structure where staff worked 24-hour shifts in all service areas. The pay system and staffing model were no longer competitive based on industry standards related to cost, service and utilization.

The pay system was inconsistent with current industry trends for a number of reasons. First, staff members were paid a premium rate (time and a half) for each hour of paid annual leave, which resulted in staff being paid more money per hour for each hour of paid annual leave than for each hour worked. Second, staff members earned an overtime rate after eight worked hours in each shift. So at Navicent, an employee was earning 16 hours of overtime in each 24-hour shift.

Obviously, in some instances this meant that yearly salaries climbed to rates that were excessive compared to industry benchmarks.

The location where those staff members worked made a difference in their utilization, which had high levels of variation. Urban staff members who worked 24-hour shifts experienced excessive call volumes, ranging from 16 to 24 calls per shift on average.

This imposed high safety and risk factors and obviously led to excessive response times, delaying service to patients. Crews became fatigued and were not satisfied, which led to excessive turnover and poor service to customers.

One key metric is unit hour utilization. In the urban area, it exceeded 0.5 and often was above that, whereas the industry standard is 0.4 dependent on the service area. Rural staff averaged two to five calls per shift, which resulted in underutilization. In those areas, the unit hour utilization levels were below 0.2. A reasonable industry rural standard is 0.2 or greater per unit for that metric.

Navicent’s deployment plan was flawed, continuously pulling and pushing ambulances among the four-county service area. This resulted in excessive fatigue, prolonged emergency response times, wasted motion, reworks and excessive fleet expenses.

These inefficiencies, of course, came with a financial cost. Navicent’s EMS system gave away dozens of emergency calls per month to the local competitor due to these problems. The suboptimal staffing capacity across the system delayed service, lost revenue and resulted in unfavorable public relations.

The black belt team set out to improve emergency response times by 10 percent, optimize utilization (namely unit hour utilization) of resources to industry standards, eliminate paying excessive wages, improve the quality of care and provide safer services to the communities served – all within six months.

The project scope related to staffing redesign was focused on the urban service area of Bibb County, and the pay system conversion was intended to be applied to staff in all service areas. The critical-to-quality focus was on emergency response times, unit hour utilization per unit and cost savings.

Measure

Training for continuous improvement

Navicent Health is a 6,000-employee health system/academic medical center that serves 800,000 residents in Central and South Georgia with 830 beds for medical, surgical, rehabilitation and hospice purposes in more than 30 locations.

The medical system began its continuous improvement journey in 2013 when management created a continuous improvement cost center. The center included a chief quality officer who also was a master black belt.

The quality officer taught cohorts of lean green belts internally in classes accredited by the Institute of Industrial and Systems Engineers. Early in the following year, Navicent began its first Six Sigma black belt cohort. Fifteen leaders completed the rigorous training and successfully executed an internal black belt project within one year of class completion to earn their black belts.

Since the inception of the program, Navicent Health has trained nearly 700 lean green belts and two cohorts of Six Sigma black belts for a total of 26. In addition, Navicent invested in 23 other staff members by training them in a change acceleration process course sponsored by a partnership with GE.

Most recently, Navicent Health created its Center for Disruption and Innovation, which focuses on enterprisewide continuous improvement, research, project management and commercialization. This center is led by the chief strategy and innovation officer. 

Training for continuous improvement

To measure the current state and future improvements, the team focused on the following operational KPIs: emergency response times, turnaround times, unit hour utilization, out-of-chute times and the hourly call volume demand analysis.

The response, turnaround and out- of- chute times were measured on a fractile basis, with 90 percent considered the minimum acceptable standard. Also, due to state-imposed zoning requirements, Navicent Health EMS was required to give emergency calls to the local competitor if no ambulances were available within two minutes of receiving the emergency request.

These calls given to the competitor represent reworks and revenue loss for the service, as shown in Figure 1. The figure also indicates that the actual KPI measures were underperforming to Navicent Health’s goal in all categories.

Analyze

After the actual KPIs were measured and compared to goal, control charts and a histogram were configured to analyze whether the overall processes related to emergency response times and calls given to the local competitor were in or out of control and capable.

The process was in control and capable (i.e., sigma level > 3) for emergency response times, but it was not optimized based on the KPI measures noted in Figure 1. The process for calls given to the local competitor was out of control and not capable (i.e., sigma level < 3). This resulted in delayed service, reworks, revenue loss and dissatisfied patients.

In order to analyze the hourly call volume demand for ambulances, a demand analysis was built to measure hourly demand for each hour of the day each day of the week using the 90th percentile and averages.

As noted in Figure 2, the call volume, represented by the blue vertical bars, was temporal in nature, as indicated by peaks and valleys; whereas the staffing curve shown by the red line was static in nature.

Figure 2 is an accurate reflection of the current state, at that time, of operations for each day of the week. During peak hours of the day the EMS system did not have enough ambulances on duty to manage high peak volumes, and during nonpeak hours (midnight to 6 a.m.) the system had too many ambulances on duty for the actual call volume.

This led to excessive costs. These structural inefficiencies directly contributed to prolonged emergency response times, suboptimal levels of service, crew fatigue and revenue loss as emergency calls were given to the local competitor.

The process map in Figure 3 shows that under the current call-taking/callassignment process required by state zoning rules, nearly 50 percent of the process steps are nonvalue-added. Thus, when calls were given to local competitors the current process was riddled with delays, waste and reworks, which added layers of inefficiencies between the supplier and consumer.

The takeaway is that if the system was optimized and staffed properly, then the nonvalue-added steps in red on Figure 3 would become nonfactors as fewer calls would be given away to the competitor.

Improve/implement

Once the KPIs were measured and the analysis identified significant operational gaps, the team focused on several initiatives that were designed to improve the situation.

All EMS staff members were converted to the 40-hour pay system, which paid fair market rates competitive with EMS industry standards. Overtime was paid only after 40 hours of work in a workweek for all areas.

Urban staff members were provided alternate rural pay rates so they would be paid properly if they chose to work overtime shifts with lower call volumes on 24-hour shifts. The rural staffers also were given alternate urban pay rates so they would be compensated properly if they chose to work overtime shifts in the urban area with higher call volumes on 12-hour shifts.

Paid annual leave was no longer paid at the time-and-a-half pay rate. Shift differentials were modeled after inhospital compensation, which paid for nights, weekends and other hard-to-fill shifts. This alleviated staffing shortages and improved service levels. One year after implementation, the EMS system was approximately $800,000 favorable to budget for salaries.

EMS staffers in the urban area of Bibb County with excessively high call volumes were converted from 24-hour shifts to 12-hour shifts. They worked a staggered rotation of seven total shifts per 14-day pay period with days off between shift sequences.

All staffers were polled prior to the shift changes and were allowed to submit shift preferences that would work best for their families and lifestyles. More than 90 percent of the staff members received their first preference, which significantly helped satisfy the workforce.

Detailed unit hour utilization reports per crew for all service areas were built with goals of 0.2 for rural areas (24-hour shifts) and 0.4 for urban areas (12-hour shifts).

All rural area crews’ unit hour utilization was increased to and maintained at 0.2, while in urban areas utilization was reduced to and maintained at approximately 0.4. This ensured that crews were adequately productive while reducing overutilization, fatigue and other things that led to a dissatisfied workforce. In turn, service levels and response times increased.

Staffing levels for all areas were adjusted based on specific hourly demand analysis data using the 90th percentile, along with a staffing buffer of one to two ambulances per hour based on the service area.

Temporal demand was addressed with temporal staffing structures to ensure more ambulances were available during peak call hours, leaving fewer available during nonpeak call hours. This helped improve and maintain emergency response times.

Geographical software analysis was conducted to determine the best place to locate EMS substations in all counties in the service area. This determination was based on historical call volume data and risk analysis. This statistically placed the on-duty ambulances in areas with the greatest probability of call locations, which contributed to faster response times and higher levels of service.

Figure 4 details the summary of results after implementation, which shows significant improvements systemwide in emergency response times (12 percent improvement), unit hour utilization (30 percent reduction in the urban area), call volume variation per ambulance (50 percent improvement), out-of-chute times (40 percent improvement) and employee call outs (70 percent reduction).

The improvements saved more than 37,000 minutes in emergency response times, and in this industry, time equals lives. The lean Six Sigma team used test of hypothesis (paired comparison small sample) to test the improvement of emergency response times post-implementation, which also is summarized in Figure 4.

Control

After implementation and verification of the successful results, the changes have been controlled by monitoring real-time dashboards and analytics for emergency response times per county, out-of-chute times per crew, turnaround times per crew, unit hour utilization per crew and hourly call volume demand analysis to ensure optimization is achieved and maintained.

Figure 5 shows the current performance of emergency response times over one-year postimplementation. This is evidence of sustainability.

Lessons learned

The Navicent Health lean Six Sigma black belt team learned a number of great lessons that can be applied in any enterprise.

First, designing real-time dashboards and utilization reports were essential to implementing and sustaining positive change, as these statistics guided the team each step of the process. Next, the team learned early on that change is much easier and more likely to be sustained if customers (both external and internal) are engaged and have a voice in the process from inception through implementation.

Finally, well-defined goals and key performance indicators (KPIs) that can be measured in real time are essential to realizing long-term sustainable change over large geographic areas.

Casey Bedgood is a black belt leadership fellow in Navicent Health’s Center for Disruption and Innovation. In 2014, Bedgood became Navicent’s first internally trained Six Sigma black belt via the Institute of Industrial and Systems Engineers. He also was the administrative leader for the Navicent Health spent the previous 16 years serving Navicent Health in various EMS roles, including as an emergency medical technician, cardiac technician, paramedic, field training officer, manager and operations manager/ director. Bedgood is certified as both a Georgia and National Registry paramedic. He earned an AAS in paramedic technology, a BBA magna cum laude from Mercer University and an MPA from Georgia College and State University. He also is an IISE lean green belt, is CAP trained via GE and is a member of the American College of Healthcare Executives.