vrijdag 2 september 2011

Hypothesis tested? Lessons and results from a daily ops meeting

In January the RTT section of Radiation Oncology started having daily ops meetings. We were so inspired by our lean exchange with Akzo Nobel. The aim was to identify problems through process monitoring. Or better: monitoring deviations from the expected outcome of the process. These deviations, so we thought, were symptoms of underlying causes. If our improvement activities to address these underlying causes were successful, we would see less deviations from expected process outcomes. Such was our hypothesis. Two things went wrong from the outset, however, so we were not able to test our hypothesis right away.

No copy paste
The first lesson we learned was that simply copying the ops meeting we had seen at Akzo Nobel generated information which did not tell us anything about our own process. Deviations at Akzo were mainly caused by operational availability of machines. In radiation oncology deviations are in the timely execution of a process step in the pre-treatment process (right first time – on time). This puts the starting date of radiation treatment at jeopardy. Both to patients as to our staff any delay to starting patient treatment is deemed as highly undesirable. We will do almost anything to make that target date. This leads to searching for (and organizing) detours, shortcuts and rush jobs. So within the first few weeks of our ops meeting we changed focus form capacity to throughput time.

Shifting our focus to throughput times, the RTT sections’ management started to take over work which was up to then being done by paramedics themselves. And especially the work identified as ‘waste’. Pursuing case files form colleague workers, expediting and managing work arounds. As such though, management started to show a genuine interest in some of the activities that did not add any value to the pre-treatment work for patients. We got a good view of our own ‘hidden department’ . But after a few weeks lab paramedics began to wonder about us being in such tight control. After all, we were on their turf.
And so in March, we started a second ops meeting, designed specifically for our pre-treatment processes. At half past eight in the morning various stations are represented and start working their case-at-risk-of-being-late-load through a standard "decision tree" they created themselves. Any problem is then escalated to the departments central ops meeting at nine.
Hypothesis: the results
Both the ops meetings have undoubtedly brought progress. Through a daily cycle we can speed up decision making and take immediate action. Cooperation and communication, as confirmed participants, have been improved. The amount of emails sent between members has decreased significantly.
But how about the original hypothesis? Do we know which underlying causes to address and are there indeed fewer cases handled at the meeting?

The graph shows the results between March and July. A tremendous result. The number of cases that arrives late at one of those stations has almost been halved!
Therefore we might conclude that the pre-treatment ops meeting is very effective. But… is that assumption correct? Weren’t we to identify deviations and search for root causes? Which root causes did we the remove? The ops meeting team stated when reporting the data: "Conclusions from these figures are as yet difficult to pull (are they a true result of our meeting?)." And: "Most of the root causes are beyond the reach of pre-treatment sections and require a multidisciplinary approach "
Undoubtedly, through working the cases systematically, we were able to manage expectations form other departments better then before. So there is some result form influencing and teaching others. But the team is absolutely right. And supplies as with a great ‘lean lesson learned’. The hypothesis is not proven! Shorter lead times can only be realized by focusing on the process. Not the individual elements. And therefore requires a multidisciplinary approach in which hypotheses are tested through countermeasures. And that does npot only fit in perfectly well in an academic setting. It also fits into a true lean culture of improvement.
The focus on throughput is now considered one of the main objectives for the entire department radiotherapy seen. A team of 18 staff from all disciplines has only just started to short-term turnaround for a group of radiotherapy patients to half.

1 opmerking:

  1. The knowledge exchange between Erasmus MC and AkzoNobel is very interesting. Here is the account of the visit at Sassenheim

    Roberto Priolo
    Editor, Lean Management Journal


Real Time Web Analytics