Detailed Notes on Shift Work, Health, and

Safety

I. The Effects of Shift Work

The effects of shift work are supported by evidence on the relationship

between working at night/on a shift work pattern and serious medical
disorders (Knutsson, 2003 review).

A. Specific Disorders and Conditions

 Gastrointestinal Disease:
o Significantly more common in shift workers than in day
workers.
o Common symptoms: constipation, abdominal pain, and
diarrhea during night shifts.
o Shift workers also experience elevated peptic ulcer risk (risk
for ulcers of the small intestine is also reported to be doubled).
o Risk groups: taxi and truck drivers, factory workers, printers, and
security guards.
 Cardiovascular Disease (CVD):
o Shift work is a potential risk factor for CVD.
o Physical factors involved: sleep deprivation, chemicals,
noise, and vibration, as well as psychosocial factors like
stress and the organization of work schedules.
 Cancer (Breast Cancer):
o Some research suggests an increased risk of breast cancer
in women who work night shifts.
o Studies conducted with nurses, flight attendants, and
radio/telegraph operators.
o Exposure to other carcinogens was not controlled for.
o Possible link: low levels of melatonin (a hormone that
regulates the sleep-wake cycle), but evidence is inconclusive.
 Diabetes and Other Metabolic Disturbances:
 o Increased concentrations of certain substances in the
body (including potassium, uric acid, glucose, and cholesterol)
occur during night work, which may be related to increased
metabolic disturbances.
o Studies show shift workers tend to have higher Body Mass
Index (BMI), though this evidence is inconclusive.
o Evidence suggests the increased chances of developing
diabetes if you work shifts.
 Pregnancy:
o Shift work shows relationships between shift work and low birth
weight as well as shift work and premature birth.
o Reported an increased risk of miscarriage among shift workers.
 Exacerbation of Existing Disorders:
o Normal biological processes follow a circadian rhythm, which
can be interrupted or interfered with by shift work.
o This may make taking medicines more complex, as
medication timings interfere with the internal body clock.

B. Other General Health Effects

 Sleep Deprivation and Disrupted Circadian Rhythms: These are

consistent results from laboratory demonstrations.
 Increased Cognitive Errors: Associated with the disruption of
circadian rhythms.
 Life Span: A Danish study reported no significant difference in the
relative death risk for shift workers.

II. Research Methods and Causality

A. Correlation vs. Causation

 Correlation: A relationship between two things (e.g., shift work and

gastrointestinal disorders). It does not show that working shifts
causes an increase in these disorders.
 Causal Relationship: To establish this, researchers must conduct
experiments.
o Example: An experimental group might work night shifts for a
time, while a control group works normal day shifts. Significant
 differences in the incidence of gastrointestinal disorders between
the groups would suggest a causal link.

B. Difficulties in Establishing Causation (Gold's Research)

 Quasi-experimental design: Studies comparing two naturally

occurring groups (e.g., shift workers vs. day workers) are quasi-
experimental because the researcher cannot randomly allocate
participants to the groups.
 Confounding variables: Differences between the groups (other than
the work schedule) may explain the results.

C. Recommendations for Better Research

 Combine research findings (e.g., laboratory experiments into short-

term sleep deprivation/cognitive errors) with a longitudinal study to
track accidents and errors and correlate these with shift work patterns.

III. Health and Safety: Accidents at Work

A. Human Error Theory

 Traditional view: Accidents are caused by human errors (person-

centred explanation).
 Errors can be reduced by identifying and targeting interventions
at those individuals most likely to make errors.
 Critique: Human error theory is often a blame-oriented approach
that can create fear and inhibit employees from reporting issues. It
may be unfair to solely blame the individual.
 Alternative: Blaming individuals may be emotionally satisfying but it
is potentially more convenient than trying to address the wider
system.

B. System Model of Errors

 System model (Nagel, 1988) / Operator-machine systems

theory:
 o Errors are caused by systems in which humans play a small
part.
o Argues that some individuals are more prone to error than
others, but this model suggests the most errors are those made
by people carrying out tasks that are more difficult or where
mistakes are most noticeable (i.e., those that result in
damage or injury).

IV. Safety Study Example: Gold et al. (1992) on Nurses

A. Study Design

 Participants: 878 registered nurses.

 Method: Self-administered questionnaire between June and
September 1986.
 Variables measured:
o Quality of sleep.
o Use of alcohol.
o Use of prescription or non-prescription medication.
o Sleeping aids.
o Whether they worked various shifts or adhered to the same shift.
o The previous two weeks and the following week on day, evening,
or night shifts.
 Shift Groups:
o Day and evening shift work (grouped together as the shift
from day to evening work has not been shown to disrupt
circadian rhythms).
o Night shifts only (worked only night shifts).
o 'Rotators' (changed shift patterns frequently).
 Average Age of Sample: 33.9 years.

B. Key Findings

 Anchor Sleep (Quality):

o 92% of day/evening nurses obtained regular anchor sleep.
o 63% of night nurses and none of the rotators obtained
regular anchor sleep throughout the month.
 o Anchor sleep disruption was experienced by 49% of the
day/evening nurses and 94% of the rotators.
 Nodding Off ('Fallen Asleep'):
o Night nurses were 1.8 times more likely to report poor
quality sleep than day/evening workers and rotators.
o Night nurses were 2.8 times more likely to report poor quality
sleep than day/evening workers.
o Night nurses and rotators were twice as likely to use
medications to help them sleep.
o Nodding off (at work or while driving):
 Night nurses: 33%
 Rotators: 32%
 Day/evening nurses: 2.7%
 In contrast, 20% of rotators and 2% of day/evening
nurses reported any instances of nodding off on day or
evening shifts.
o Night nurses and rotators were 3.9 times and 3.6 times as
likely as day/evening nurses to nod off while driving to/from
work in the previous year.
 Accidents and Near-Misses:
o Measured in the past year: driving accidents, medication errors,
job procedural errors, and work-related personal injuries that
could be attributed to sleepiness.

V. Issues and Debates

 Nature vs. Nurture: The material raises the question of whether shift
work interferes with the natural functioning of the body (nature) or
whether the effects are influenced by environmental and work
organization factors (nurture).

VI. Key Terms

 Cardiovascular: The system that includes the heart and blood

vessels.
  Near-miss: An event with the potential to result in injury or damage,
but that did not actually cause damage or injury.

Research Methods Notes: Shift Work and

Health

I. Distinguishing Correlation and Causation

 Correlation: A correlation simply describes a relationship between

two variables (e.g., between working shifts and an increased chance of
gastrointestinal disorders).
o Crucial Note: Correlation does not prove causation. Showing
a correlation does not prove that shift work causes the increase
in disorders.
 Causal Relationship: To demonstrate that shift work causes a health
effect, researchers must conduct experiments.
o Ideal Experiment: An experimental group would work
changing shifts for a period of time, while a control group
would work normal day shifts.
o Outcome: Significant differences in the incidence of disorders
between the two groups would suggest a causal relationship.
o Practicality: This type of research is often difficult and raises
ethical issues, making it unlikely to be conducted.

II. Difficulties in Drawing Causal Conclusions (Gold's

Research)

 Quasi-Experimental Design: Research like Gold's is described as

quasi-experimental.
o This is because the study compares two naturally occurring
groups (shift workers vs. day workers), thereby creating an
independent variable without manipulation.
o Limitation: The researcher was not able to randomly
allocate participants to these different groups.
 o Implication: There may be many differences between the
groups (confounding variables) that could explain the results,
making it difficult to definitively draw causal conclusions.

III. Recommendations for More Effective Research

 Combining Research Findings: The research findings supported by

laboratory experiments into short-term sleep deprivation and cognitive
errors should be combined with other methods.
 Longitudinal Study: A longitudinal study is recommended to track
accidents and errors over time and correlate these with specific shift
work patterns.
o This approach is suggested as potentially more effective for
understanding the long-term effects of shift work.

Notes: Accidents at Work and Systems

Failure

I. Analysis of Accidents (The Three Mile Island Incident)

The Three Mile Island nuclear power plant accident in 1979 is a famous
example of system errors with near-catastrophic consequences.

A. The Event

 An employee shut down an alternate feedwater pipe and then

went off duty without turning this back on.
 The reactor started to overheat, and warning sirens began to
sound, but no one at the power plant knew the problem.
 It took 2 hours and 18 minutes to find the fault.
 Further delay could have led to a large-scale release of radioactive
material.
B. Errors and Failures

 Human Error: The initial failure to switch the feedwater pipe back on.
o Psychologists categorized this as an error of omission—doing
something that should not have been done (shutting off the
feedwater pipe).
o Contrast: An error of commission is doing something that
should have been done but wasn't.
 Design/System Failures: It is inappropriate to simply blame the
operator; the system was designed poorly.
o No warning system was in place to tell anyone that the pipe
had been turned off.
o A relief valve also failed to open, but no warning was given.
o Fault Identification: Operators had to scan over 1,600
gauges to find the fault.
o Poor Display Design: Different colors and names were used to
represent safety in different systems and on some places a color
represented safety whereas in another part of the system it
represented danger.
o The incident revealed significant failings in the design of
the systems as well as the safety procedures.

II. Human Error Categories

Psychologists classify human errors into different types that can lead to
accidents at work:

 Error of Sequence: Actions are carried out in the wrong order.

 Error of Omission: An action is skipped or repeated (e.g., the
Three Mile Island feedwater pipe).
 Error of Timing: An action is performed too late or too early.

III. System Design and Cognitive Human Factors

Since the incident, there has been an emphasis on the development of

'human factors' expertise.
  Goal: To ensure that the design of machine controls and displays
reflects our knowledge and understanding of human cognition.
 Good Design Examples:
o Making sure that displays are clear and easy to interpret.
o Ensuring operators are not expected to maintain vigilance
for too long (e.g., air traffic controllers).
o Providing and practicing tested emergency procedures.

IV. Reducing Accidents at Work: Token Economy Example

 Study: Fox et al. (1987) study on workers' behaviors.

 Method: Investigated the use of a token economy to reward workers
for not having accidents or injuries for a specified amount of time.
o A token economy is a type of operant conditioning.
 Setting: Two associated product processing plants in two open pit
mines in the USA.
 Findings: Before the study, the number of days lost from work due to
injuries on the job in one of the mines was over eight times the
national average for all mines and three times the national average
at the other mine.
o In the five years preceding this study, two people had been
killed and a third person had suffered a permanent disability.
o The two settings were similar in many ways (wages, tools,
procedures, size of injuries).
o Injuries had occurred in all areas of both mines but were
particularly associated with the use and maintenance of
heavy equipment.

Detailed Notes: Health and Safety in the

Workplace

I. Models of Accident Causation

There are two primary models for understanding the cause of workplace
accidents, which lead to different approaches to safety management.
A. Human Error Theory (Person-Centred Approach)

 Focus: Unsafe acts (errors and procedural violations) of individuals at

the "sharp end" (front-line workers).
 View of Error: Errors arise primarily from aberrant mental
processes such as:
o Forgetfulness or inattention.
o Poor motivation, carelessness, negligence, or recklessness.
 Core Idea: Accidents are caused by human errors.
 Countermeasures: Directed at reducing unwanted variability in
human behaviour. Methods include:
o Targeting individuals most likely to make errors.
o Disciplinary measures, retraining, and appeals to
fear/compliance (posters, more procedures).
 Critique:
o Often a blame-oriented approach that can discourage
employees from reporting issues.
o Ignores the wider system and context; isolates unsafe acts
from their systemic causes.
o It's often more convenient for management to blame
individuals than to fix system flaws.

B. System Model of Errors (System Approach / Operator-Machine

Systems Theory)

 Focus: The conditions under which individuals work, and the

upstream systemic factors that lead to errors.
 View of Error: Errors are expected (humans are fallible) and are
seen as consequences rather than causes.
 Core Idea: Accidents are caused by systems failures (Nagel, 1988),
in which humans play only a small part.
 Countermeasures: Directed at changing the conditions under
which humans work and building system defenses to avert errors or
mitigate their effects.
o Focus on the person, team, task, workplace, and institution.
o Identifying and fixing recurrent error traps.
 Swiss Cheese Model (Reason, 1990): High-technology systems
have multiple layers of defenses (e.g., engineering, people,
 procedures). Failures only occur when "holes" in many layers
momentarily line up, creating a "trajectory of accident opportunity."
o Active Failures: Unsafe acts committed by people (e.g., a
momentary slip).
o Latent Conditions (Failures): Problems inherent in the system
(e.g., poor design, inadequate training) that can lie dormant until
they combine with an active failure.

II. The Three Mile Island Accident (1979)

This incident is a classic case study illustrating the failures of the System
Model and the importance of Human Factors.

 The Initial Event: The accident began with an Error of Omission—

an employee shut down an alternate feedwater pipe and left without
turning it back on.
 The System Failures: The crisis escalated due to system and design
flaws:
o Stuck Valve: A relief valve failed to close, but instrumentation
only showed the 'close' signal was sent, not the valve's actual
position, leading to massive coolant loss.
o Poor Display Design: Operators had to scan over 1,600
gauges to find the fault. Different systems used conflicting
colour codes for safety/danger.
o Inadequate Training: Operators misunderstood key pressure
readings and their training led them to shut off emergency
cooling pumps to prevent the pressurizer from filling up, which
was the opposite of what was needed.
 Psychological Error Categories:
o Error of Omission: Failing to perform a necessary action (like
turning on the feedwater pipe).
o Error of Commission: Performing a task incorrectly or doing
something that should not have been done (like turning off the
emergency cooling).
o Error of Sequence/Timing: Carrying out actions in the wrong
order or at the wrong time.
  Result: The incident emphasized the need for 'human factors'
expertise in designing machine controls and displays to reflect human
cognitive abilities (e.g., clear displays, managing vigilance).

III. Accident Reduction Strategy: Token Economy

 Study: Fox et al. (1987) on open-pit mining facilities (uranium and

coal).
 Context: The mines had a very poor safety record (one mine had
eight times the national average for days lost due to injuries). Injuries
were particularly associated with heavy equipment.
 Method: Implemented a token economy, which is based on operant
conditioning.
o Reinforcer: Employees earned trading stamps (tokens) which
could be exchanged for prizes at redemption stores.
o Targeted Behaviors (Reinforcement): Stamps were given to
workers for:
 Not having a lost-time injury for a specified period.
 Being in a work group where no one had a lost-time
injury.
 Making safety suggestions or performing acts that
prevented accidents.
o Punishment: Workers lost stamp awards if they or their
group were injured, caused equipment damage, or failed to
report accidents.
 Findings: The token economy resulted in large, long-term
reductions in the number of lost worker days and lost-time injuries at
both mines. The cost savings from reduced accidents far exceeded
the cost of running the program.

You can learn more about the Three Mile Island accident and the safety
culture changes it brought about by watching Moments in NRC History: Three
Mile Island - March 28,1979.