The fatigue journey
Brian Nicholas
Principal Adviser Workforce Planning & Reporting, Department of Corrections
James Scoon
Analyst, Workforce Planning & Reporting Team, Department of Corrections
Author biographies:
Brian has had over 15 years’ experience at the department. He is passionate about data-driven analytics and taking innovative approaches to deliver workforce data solutions (such as the fatigue monitoring tool explored in this article).
James began working as an intern at Corrections in 2012 while studying Commerce at Victoria University. James’ work focuses on the visualisation and presentation of data to create meaningful insights.
Introduction
The Department of Corrections’ work involves high risk activities. To work safely, our staff must be fully alert. For frontline custodial staff especially, the need to be able to continually assess other people’s behaviours and manage their own responses is critical. Impairment from fatigue is a significant factor that reduces the ability to respond effectively and safely to challenging situations. In our line of work, this poses a major safety risk. If people are fatigued, poor decisions are made and people can get hurt. We have a shared responsibility to manage fatigue risks for the safety of staff and the offenders in our care.
To provide greater visibility of staff at risk of fatigue, we have developed a fatigue risk management tool. Launched in September 2016, the fatigue monitoring tool for custodial staff monitors staff at risk of fatigue by applying scores based on their roster patterns. Fatigue scores are calculated based on a number of triggers including shift type (early/day/late/night), number of consecutive shifts and hours worked, opportunity for sleep (during day or night), offender-facing opportunity by shift type (e.g. lower on night shift), and travel time to work. With greater visibility of work patterns, informed decisions can be made to better manage the health and safety risks of fatigue.
The case for managing fatigue
Workplace fatigue is a physiological state of reduced mental or physical performance capability (IATA, ICAO, IFALPA, 2015). It negatively impacts reaction times, the ability to concentrate and the ability to assess risks. The main causes of fatigue are sleep loss, extended wakefulness, working and sleeping at suboptimal times in the body’s natural (circadian) clock cycle, and workload (mental or physical activity) (Gander, 2016). Ultimately, fatigue is a hazard; it impairs a person’s alertness and ability to perform work safely and effectively (Ministry of Transport, NZ, 2016).
The core problem with shift work is that it requires trying to override the body clock’s preference for sleep at night (Gander, 2016). This means our staff working night shifts have to work through times in the body clock cycle when they are least functional and most prone to making errors. They then have to try to sleep when the clock is gearing up the brain and body to be awake. Similarly, staff working early shifts are waking against the body’s natural sleep cycle.
As fatigue is a factor that impairs the ability to make effective decisions, it poses a significant safety risk which under the Health and Safety at Work Act 2015, the department has a duty to mitigate (WorkSafe NZ, 2016). The associated risks of driver fatigue also present a major risk for commuting staff and was a factor in 14 percent of fatal crashes in New Zealand in 2015 (Ministry of Transport, NZ, 2016).
The fatigue challenge
The difficulty for prison management was limited visibility of timely information about who was available to work, who had already worked long or demanding shift schedules and, critically, who had not taken sufficient time off for rest and recovery. As a result, common practice was to call on the same officers with a propensity to accept overtime or to cover gaps in the roster. This compounded the hours worked by some staff and their likelihood of being affected by fatigue. Under health and safety legislation, there are no excuses for not being aware of work schedules that are bringing about fatigue in our staff. This drove the need for a tool which gave visibility of at-risk staff by evaluating likely fatigue levels associated with our roster patterns.
How the tool works: Fatigue triggers
We began by evaluating the factors in the custodial work environment that would likely contribute to fatigue accumulation. These are called “fatigue triggers”:
- Shift type (early/day/late/night)
- Number of consecutive shifts worked
- Amount of overtime and call-back hours worked
- Opportunity for sleep around shift type (during day or night)
- Travel time to and from work
- Relative activity rate of each shift
- An additional weighting for new staff as they adjust to working in a prison.
Diagram 1:
Fatigue triggers
How the tool works: Data sources
To bring these triggers together in a meaningful way we needed reliable and valid data sources. The key to bringing this tool to life was to focus on individuals’ work patterns. Using averages by unit or site hides the reality for staff and any work patterns which should be raising red flags. This was an important philosophy while developing the tool; because it is an individual who drives home after a tiring shift and it is an individual who has to make safe decisions when dealing with offenders and difficult situations.
Fortunately, analysing individual fatigue work patterns was straightforward, with all frontline custodial staff hours being recorded centrally in the department’s rostering software, Click Roster: the key data driver of the fatigue tool. This gives us access to everyone’s shift patterns, visibility of overtime or call-back hours and whether rostered days off are taken.
SAP (Systems Applications and Product) is the second key data source for the tool. SAP identifies new staff and provides information on daily mileage for staff commuting to work (allowing us to factor additional scores for staff with longer commuting distances).
How the tool works: Understanding a typical day
Understanding the opportunity for sleep depending on shift type was a key breakthrough in developing the tool. To achieve this, we surveyed Corrections officers at every site, giving an insight to the average number of hours sleep staff get by shift type, time spent preparing for work, and time spent socialising, or with family after work. This gave a practical basis for assessing the impact of shifts on people’s daily lives.
One of the main findings from the survey was that staff working early shifts tend to have less sleep than any other shift type. Not only do they wake during the body’s “low point” in the natural sleep cycle, but they tend to follow regular evening routines and go to bed at normal times, in effect burning the candle at both ends.
During this process, we were fortunate to work with Professor Philippa Gander, Director of the Sleep-Wake Research Centre at Massey University, who helped validate the approaches we used, particularly around understanding the body’s natural sleep cycles and the impact of sleep loss on performance. From this collaboration, we added weighted penalties to shifts based on their impact on natural sleep opportunity (i.e. night and early shifts). This formed a logical basis for applying fatigue scores to the different shift types.
The fatigue scores and how they are reduced
With good data on individuals’ roster patterns and mileage, and an understanding of the impact shift types have on sleep opportunity, we were able to create a tool which analyses individual staff data held in Click Roster and SAP looking back over the past 90 days, and scheduled work patterns 28 days ahead. Points are applied based on the fatigue triggers an individual’s work pattern hits. These data sets are updated daily providing accurate, real-time scores. The scores then fit within a three-tier matrix system as follows:
The intention of this is to ensure fatigue scores are proportionate, only flagging those who have worked more demanding schedules with reduced opportunity for rest and recovery. An individual’s fatigue score is only offset by taking rostered days off, which reduces fatigue scores by 50 percent per day.
To put this in context, staff working regular Monday to Friday day shifts with weekends off would score very low on the scale. Conversely, staff working seven successive night shifts for example, or continually working rostered days off would begin accumulating a higher score and moving in to the “Watch List” category. From a clean slate of zero points, it would take 19 day shifts to reach the “Alert List”.
Fatigue reports to managers
Every week since late September 2016, all prison directors have received a report showing staff whose work patterns have given them, or are forecast to give them, a fatigue risk score within the “Watch List” and “Alert List” categories. The key message for managers is simply to make sure people are well rested and taking enough time away between shifts. Managers have used these reports to better allocate overtime and call-back hours and to have conversations about managing workloads and wellbeing with their staff.
Example fatigue metrics
 |
Impact of the fatigue tool
When we first ran the fatigue tool over the data sets, 93 staff were identified with scores on the Watch List or Alert List. Just two weeks after introduction of the tool, a change in behaviour saw a significant number of those initially on the “fatigue list” (on Watch or Alert tiers) come off it after having an opportunity to rest. Sixteen weeks later, on 31 January 2017, there had been a 76% reduction in total staff flagged on the fatigue list. This was a reduction from 93 to 22 staff, and a decrease in total fatigue scores from 36,842 to 7,729. Ultimately, with greater visibility of staff work patterns, informed decisions are being made to better manage the health and safety risks of fatigue.
Weekly trending, staff on fatigue list with regional breakdown:
Fatigue model validation
There are important caveats around the fatigue tool. Primarily, this tool only provides insight to work-related fatigue. We cannot control what staff do when they leave the site. Additionally, as some of the assumptions used to link work patterns to fatigue levels are not currently evidence-based, the fatigue tool is designed to be customised and refined. As our understanding of fatigue risks grows, including research from Professor Gander and business feedback, we will be able to recalibrate the weightings of the fatigue triggers so the scores produced are valid. Additionally, we are developing a sleep self-assessment tool to give staff instant feedback on their fatigue risks based on the quantity and quality of their recent sleep. Data from this will help validate weightings applied to the fatigue matrix.
With continued review and development of the fatigue tool and the use of technology, the profile of fatigue data will be made more accessible and relevant to all staff in real time, enabling us to take greater responsibility for managing fatigue risks.
References
Gander, PH. (2016). Heuristics for Effects of Shifts on Waking Function. Wellington: Massey University.
IATA, ICAO, IFALPA. (2015). Fatigue Management Guide for Airline Operators. Second Edition 2015. Montreal: International Air Transport Association.
Ministry of Transport, NZ. (2016). Fatigue: Crash Facts. Wellington: Ministry of Transport.
WorkSafe NZ. (2016). Healthy Work: WorkSafe’s Strategic Plan for Work-Related Health 2016 to 2026. Wellington: WorkSafe.