Latest / Safe As Podcast / Ep 3: Learning Teams vs Root Cause Analyses

Transcript

Ep 3: Learning Teams vs Root Cause Analyses · Safe As Podcast

• 0:06
  We all want safer workplaces marked by continual risk
• 0:10
  reduction. We investigate incidents, hoping
• 0:13
  to fix the problem for good. But sometimes these
• 0:16
  investigations feel like Groundhog Day, investigating the
• 0:19
  same things, finding the same things, and fixing apparently
• 0:24
  the same things for good. Relying on investigations isn't
• 0:27
  a proactive cycle for learning and improvement in the first
• 0:30
  instance. Since we're doing it.
• 0:32
  What if our standard approach only scratches the surface,
• 0:36
  leaving the core or even systemic issues untouched?
• 0:39
  Is there a better way? Good everyone.
• 0:42
  I'm Ben Hutchinson, This Is Safe has podcast dedicated to the
• 0:45
  thrifty analysis of safety, risk and performance research.
• 0:49
  Visit safetyinsights.org for more research.
• 0:52
  Today's study from Robbins ET al, 2021 is titled Evaluation of
• 0:57
  Learning Teams versus Root Cause Analysis for Incident
• 1:00
  Investigation in a large United Kingdom National Health Service
• 1:04
  Hospital. It was published in the Journal
• 1:06
  of Patient Safety. It compared 22 conventional root
• 1:09
  cause analysis investigations against 22 learning team
• 1:13
  approaches. All the specific root cause
• 1:16
  analysis methodologies weren't detailed.
• 1:18
  The researchers compared their reports and interviewed staff
• 1:21
  familiar with both approaches. So what's a learning team?
• 1:25
  It's defined as a facilitated conversation between those that
• 1:28
  do the work and those that design the work to share
• 1:31
  operational intelligence between the two groups and improve
• 1:34
  system design. Really not dissimilar to pre and
• 1:37
  postmortems or quality circles. So what did they find?
• 1:41
  Learning team investigations delivered significantly more
• 1:43
  actions, averaging 7 1/2 actions per incident compared to just 3
• 1:48
  1/2 per instant for RCAS recalls analysis?
• 1:52
  Even better, 57% of the learning team actions were systems
• 1:56
  focused versus just 30% for the root cause analysis.
• 2:00
  They observed clear differences between the learning teams and
• 2:03
  the root cause analysis in personal involvement,
• 2:05
  discussion, content, cultural aspects, challenges, and
• 2:10
  outcome. Let's explore some of these.
• 2:12
  So if you're personal involved. The interviewees noted that the
• 2:15
  learning teams drive different outcomes because they were
• 2:18
  designed from the outset to be inclusive events involving all
• 2:21
  people involved with the processes that have been
• 2:23
  evaluated. This brought in better
• 2:25
  representation from the whole team, not just those directly
• 2:28
  involved in the incident. Conversely, people lower the
• 2:32
  organisational hierarchy were less often invited to the RCA
• 2:36
  investigations, even though their expertise was highly
• 2:39
  valued within the learning team framework For the content of
• 2:42
  discussions, what people discussed in the learning teams
• 2:46
  consistently differed from the root cause analysis.
• 2:49
  Learning teams focus more on normal and holistic operational
• 2:52
  factors, with less reliance on instance specific details.
• 2:56
  What they mean is how does work normally get done?
• 2:59
  What are the things that normally make work difficult or
• 3:01
  successful? In contrast, the root cause
• 3:04
  analysis investigations were seen as focused around a single
• 3:07
  specific incident, with significant time spent
• 3:10
  establishing exactly what went wrong.
• 3:13
  It was also felt that the root cause analysis often focused
• 3:16
  only on the last barrier to fail, while learning things
• 3:19
  looked more holistically at other barriers and factors
• 3:21
  influencing operational work, The Rcas also focus less on the
• 3:26
  deeper sort of wires within the broader context environment.
• 3:30
  Learning things offered more flexibility for complex themes
• 3:32
  and multiple perspectives, while the root cause analysis tended
• 3:36
  to oversimplify many factors into just a few causal factors.
• 3:40
  This flexibility in learning teams allowed for better
• 3:43
  transfer of lessons to other areas and more systems focused
• 3:46
  actions cultural areas so interviewees observed a distinct
• 3:51
  cultural difference. Learning teams felt more
• 3:53
  supportive and open to learning, with less fear of interpersonal
• 3:57
  risk. They also favoured less
• 3:59
  technical themes, allowing broader participation.
• 4:02
  In contrast, the root cause analysis investigations were
• 4:06
  believed to foster more blame and focus on individual
• 4:09
  mistakes. People described an environment
• 4:12
  of fear and guilt where they were less likely to speak truth
• 4:15
  to power. Some challenges.
• 4:18
  While generally positive about the learning teams, the study
• 4:21
  emphasised they aren't suitable for every situation.
• 4:24
  They're one tool among many. Recourse analysis are well
• 4:27
  established and familiar, unlike sort of the unknown nature of
• 4:31
  learning teams at this point. Also, staff lowering the
• 4:34
  hierarchy then it harder to get time to attend the learning
• 4:38
  teams and the overall time investment for learning teams
• 4:40
  could be higher. On the positive side, people who
• 4:44
  attended the first learning team session were often highly
• 4:46
  engaged and interested in the second.
• 4:49
  Learning teams were seen to challenge the status quo more
• 4:51
  often too, seeking innovative solutions.
• 4:55
  They also focus more on the coalface, the frontline
• 4:58
  operations. While the root cause analysis
• 5:00
  were seen to reflect more of the senior management's perspective
• 5:03
  more the work has imagined. So let's unpack some of the
• 5:06
  findings more. In this sample, learning teams
• 5:09
  delivered more systems focused actions of higher quality,
• 5:12
  addressing broader issues compared to the root cause
• 5:14
  analysis that were compared, which seemed to have a narrower
• 5:17
  focus on singular incidents, but importantly regarding which
• 5:21
  techniques suit which occasion. The paper reemphasizes that
• 5:25
  learning teams aren't universally superior.
• 5:27
  Other sorts of investigation techniques may be more suitable
• 5:30
  for a specific incident, less likely to recur outside of its
• 5:33
  unique context. Also, for a repeat event, if a
• 5:37
  comprehensive learning team was already done, the recourse
• 5:40
  analysis might be more efficient to capture any changes since the
• 5:43
  last learning team. They conclude that there may not
• 5:47
  actually be one right investigation method.
• 5:50
  Incidents lie on a spectrum, and therefore judgement needs to be
• 5:54
  applied to establish which investigative methods to choose.
• 5:58
  So what do we make of the findings?
• 6:00
  Well, like any single study, we shouldn't make too much of the
• 6:03
  findings, but rather look for a body of evidence.
• 6:06
  It does provide some interesting implications though about
• 6:09
  expanding our adaptive tool boxes.
• 6:12
  Like they know it's less about one single technique being
• 6:15
  wholesale superior, but rather fitting the more appropriate
• 6:18
  approach to that situation. Learning teams, despite these
• 6:21
  findings, might still focus too much on lower level individual
• 6:25
  factors compared to other systems based methods, but of
• 6:29
  course nothing says they can't be combined since the learning
• 6:33
  teams is more of an approach rather than a specific
• 6:36
  technique. Also consider how root cause
• 6:39
  analysis in this sample, we're more focused on preventing
• 6:41
  yesterday's accident instead of trying to prevent tomorrow's.
• 6:45
  Think about it, yesterday's accident has already happened,
• 6:47
  let's focus on tomorrow. Finally, consider the Trojan
• 6:51
  Horse strategy. If people can't get permission
• 6:54
  to attend the learning teams because they're not seen as
• 6:57
  legitimate yet, then just schedule and a root cause
• 6:59
  analysis and then interject the learning team philosophies into
• 7:02
  that. What are some limitations?
• 7:05
  Importantly, learning teams weren't randomly assigned, but
• 7:09
  that was selected based on the perceived utility of using that
• 7:12
  technique, which could introduce some level of bias.
• 7:15
  Also, interviews were mainly with those sharing learning
• 7:19
  teams, potentially introducing some bias, so all of them still
• 7:23
  had extensive root cause analysis experience.
• 7:26
  Finally, no followu was possible to assess the long term changes,
• 7:30
  if any. That's it on Safe as I'm Ben
• 7:33
  Hutchinson and hoe you found this useful.