This story was originally published by Data-Smart City Solutions
Every morning, sanitation workers across Louisville start their day with a regimen of stretching. In the offices of the city’s Public Works department, even those who spend most of their days behind a desk and computer monitor begin their day with the same routine. Stretching is nothing new. But this is, of course, an integral part of Louisville’s innovation agenda.
Public works is a strenuous job. Tippers, those who ride along on back of garbage and recycling trucks, must tote unwieldy cans and bins while minding curbs, potholes, traffic, and other obstacles. Strains and sprains of backs, knees, and ankles come with the territory.
As Keith Hackett, Louisville’s assistant director of Public Works, said, “We don’t have the luxury of working in a controlled environment. Our responsibility is to reduce the possibility of injury.”
To that end, in January 2012, Public Works began benchmarking its rate of OSHA recordable injuries – significant injuries that cause employees to miss time, limit their duties, or worse. Finding a rate of over 30 percent in the first month, officials in Public Works and Performance Improvement suspected they had significant room to improve. But they did not know how much improvement to reasonably expect. This made Public Works an ideal pilot case for performance improvement.
Louisville’s Office of Performance Improvement (OPI) builds on Mayor Greg Fischer’s philosophy that the center of the government is not (and should not be) a dispensary of departmental data, giving affirmations to departments for their good behavior or handing down orders to change.
Rather, the central office aims to empower departments to better interpret their own data, diffusing analytical capacities to the various departments. To that end, each of 20 departments has identified a dedicated data analyst to act as a liaison with the central office. As part of the LouieStat program, OPI supports participating departments with coaching and consulting, and regularly gathers working groups of counterparts across departments. There are some enterprise Key Performance Indicators (KPI) that are tracked across the city, but analysts within each department are also responsible for creating and tracking their own KPIs and drafting supporting documents for regular data forums.
Stephen Goldsmith has written on how OPI has linked performance reporting to strategy to transform Louisville’s operational culture. Initiatives in Public Works and Louisville Metro Emergency Medical Services (LMEMS) provide compelling evidence of this transformation.
In 2012, to help set benchmarks for itself, a team within Public Works surveyed the landscape of comparable public works departments and private solid waste management companies. They were shocked to discover Louisville’s injury rates were an order of magnitude higher than the cohort of comparable providers, and some of the providers had zero OSHA recordable injuries.
The department identified several KPIs, including OSHA Recordable Injury Rate, Lost Time Injury Rate, Hours Not Worked, and Overtime Hours Paid. By the summer of 2012, the department implemented a toolbox of recommendations to help Public Works hit its ambitious improvement goals in these KPIs.
The most visible change recommended from that toolbox is the regular stretching regimen, but other changes have also focused on injury prevention and mitigation as well as institutionalizing a culture of safety.
Perhaps the most effective policy change was a modified duty policy. Before July 2012, employees who suffered an injury severe enough to prevent them from doing their normal field work - a trash picker spraining an ankle, for example - would not be allowed to return to work in any capacity until he or she was well enough to return to normal duty. Under the modified duty policy, someone with a twisted ankle who is unable to work in the field can return to work in an administrative capacity, doing clerical work during his or her convalescence (or even completing additional safety training).
The response from employees has been positive, too. Employees want to get back to work; they would rather be out in the field than laid up at home in bed. Getting back to work quickly gives workers the benefit of drawing their full pay, rather than a lower sick pay. This policy change has the added benefit of encouraging returning workers to remain active during their healing process, rather than staying on their couch at home, sedentary, speeding their recoveries.
This anecdotal evidence is backed up by the numbers. OSHA recordable injuries have fallen steeply, from a pre-program high of 31 percent in January 2012 to a recent low of 15 percent, where the rate has held steady since November. Hours lost to work related injuries and illness have also fallen, from a 2011 peak of 4,000 hours per month (about five percent of working hours) to a low in February 2014 of 608 hours (about one percent).
The overall lost time to injury rate is falling, too, with the latest month meeting the goal of 6 percent or better. The next task is to maintain this gain over the 12-month running average.
One way to ensure such long-term gains is to promote a culture of safety. Many new hires enter public works in entry-level positions, and may have no experience in the strenuous line of work they are about to begin. For new hires to do things the right way from the get-go, it is crucial they understand they are joining a culture that values safety. The department has signaled this with the creation of the Accident Review Committee (ARC). The Committee is composed of representatives from the union, management, and safety teams and review all accidents and injuries. They give recommendations to departmental management on how to prevent future recurrences, and management pass on the recommendations to their staff.
As important as tone-setting is, Hackett also sees continued training through promotions as crucial: “When a tipper gets promoted to driver he has more things to be aware of. Now that you are driving you are responsible for the safety of your tippers - not only how you drive, what you do, your surroundings, but making sure your tippers know their surroundings. As you get promoted your safety sensitivity has to be promoted as well.”
Since the successful pilot period, the performance improvement measures taken by Public Works have spread to many other city departments and agencies. Several have adopted a similar modified duty policy and created bodies like the ARC. Louisville Metro Emergency Medical Services (LMEMS) has taken on ambulance turnaround time.
Turnaround time, also called downtime, refers to the amount of time it takes from when an ambulance unloads a patient at a hospital until the crew becomes available to respond to another service call. Emergencies do not wait for the “right” crew to respond. If the preferred crew (one closest to the emergency and with plenty of time left on shift) is unavailable, another crew will be called to make the run. In practice, long turnaround times by one crew mean another will work overtime to compensate - sometimes up to another hour on top of a twelve or even sixteen-hour shift.
There are many factors that can add to turnaround time, including the time it takes the crew to bring a patient through registration and triage to an emergency room bed or another point of care, such as the cardiac lab or obstetrics delivery floor. This also includes the time it takes the crew to complete the electronic patient care report and to brief the ER triage nurse on the patient’s condition. These tasks are essential to the process. On the hospital side, there are several elements that also impact on this process, including availability of beds, cardiac monitors and nursing staff to perform triage. There are other reasonable uses of time as well, including crew members taking time at the hospital to eat or use the bathroom. And, in a less defined fashion, downtime at the hospital may also provide an opportunity for rest and social interaction, something which might be reasonable if there were not a queue of potentially life-threatening 911 calls getting ready to be dispatched.
To better understand potential inefficiencies in the system, Jordan Mudd, LMEMS Personnel Coordinator, began tracking how much time crews spent at each step of the process in the ER, measuring those elements which were under the control of the crew versus the hospital staff. Managing the time spent in triage, for example, would be the responsibility of hospitals, not LMEMS. But after collecting data over a few days, Mudd noticed that the crews he observed personally were faster than the average crew, which suggested there was indeed room for improving turnaround time on the EMS side.
Drawing on about six months worth of data, the average turnaround time across all ambulance trips was about 38 minutes. The average time also varied from one hospital to another, suggesting that some hospitals might be able to cut inefficiencies from their process of receiving patients. Of about 200 daily ambulance transports, an average of 115 crews did not leave the hospital within 30 minutes.
After gathering the benchmark data, LMEMS Command Staff initially mandated that crews should be at a hospital no longer than 30 minutes, with a goal of reducing the average turnaround time below that threshold.
Louisville Metro EMS is responsible for an area of about 400 square miles. On any given shift, there are two supervisors on duty in the field. Needless to say, it is difficult to monitor such a wide coverage area with such a small workforce. Management relies on employees to do the right thing, and to check in with dispatch to update their location and status. If a supervisor happens to see an idle crew at a hospital, he could direct them to make themselves available, but supervisors cannot monitor all crews all the time. Rather than hire more supervisors, both EMS and city leadership sought improvement through better performance informed by data.
LMEMS already had a source of valuable data in its Computer Aided Dispatch (CAD) system. Dispatchers log calls in the CAD, which creates a record of critical time intervals for each 911 call from the time it is received to the time the ambulance goes en route and the time it arrives on-scene. The department saw opportunities to make better use of the system’s potential to offer real-time monitoring and longer-term feedback.
One additional supervisor who is typically located at headquarters and has exclusive responsibility for scheduling also has an extra monitor to review crews’ real-time location data. This allows the supervisor to contact crews who have been at a hospital for a long time, and the crews can provide detailed feedback (for example, “triage has a big backup”). Using the real-time location data, the supervisor can compare the turnaround speed of crews at the same hospital successively or concurrently, and notice bottlenecks even if the crews don’t report them first.
The department generates daily reports from the CAD showing all ambulance turnaround times longer than 30 minutes. The reports also allows LMEMS to compare these runs, providing detailed information on the kind of medical emergency involved, to which hospitals patients were transported, and more. The automated data is supplemented by feedback from crews, who are now required to submit an explanation for their delay to a dedicated email address any time they take longer than 30 minutes to go available from the hospital. Supervisors get reports detailing their crews’ explanations, determine if they are reasonable, and then review these incidents with their crews.
Sending the message that excessively slow turnaround times are detrimental to both overall system function and individual patient care is only part of the plan. The department also showcases employees who are getting the job done well. The regular LouieStat meetings provide one highly visible way for LMEMS leaders to showcase their employees’ accomplishments to other agencies and the mayor’s staff. Internally, the department holds monthly training sessions, opportunities to present employees with both useful criticism and praise while equipping them for more future improvement.
Thus far, the results of this initiative have been overwhelmingly positive. Since the beginning of this initiative, Mudd’s observations have shown most crews lowered their turnaround times to approximately 22-25 minutes, with more serious cases taking longer. From the previous high of an average of 115 daily turnaround times taking more than 30 minutes, the rate has fallen to about 31 instances per day of turnaround times in excess of 30 minutes. This means that more than 80 percent of daily ambulance runs are meeting LMEMS’ turnaround time goal.
With faster turnarounds, more crews are available to respond to emergency medical calls and fewer personnel must be held on duty past the end of their shifts, meaning less spending on overtime and fewer runs made by fatigued crews. According to Theresa Reno-Weber, Louisville’s Chief of Performance & Technology, this improvement in turnaround times was equivalent to adding one to two new ambulances and five full-time staff to the street each day, which would cost about $1.4 million.
Mudd says that this project is just the first step in a much longer process. As LMEMS considers the next steps, there are the natural incremental improvements, like increasing the percent of crews turned around within 30 minutes from 80 to 90, or reducing the threshold from 30 minutes to 25. Deeper investigations may reveal more significant improvements. The next step may be to expand the investigation of long runs. Are there certain kinds of emergencies that frequently lead to longer turnaround times? How can hospital staff reduce the time they require from ambulance crews without sacrificing the quality of their care?
These questions are essential to Louisville’s innovation agenda. Mayor Fischer’s mantra, known within the administration as “The Job,” breaks down the mission of government into three categories: daily work, continuous improvement, and breakthroughs. Many of the efforts from the Office of Performance Improvement fit in the second category, as LouieStat and performance indicators help drive initiatives like those by Public Works and LMEMS. As departments continue to ask how they can better accomplish their goals - just as LMEMS asks deeper questions about how its work relates to that of the hospitals - they may be able to reframe their programs and develop entirely new processes to benefit their communities.