Introduction: Night shift work is associated with an increased risk of accidents, adverse health
effects, and psychosocial impairments. Consequently, various recommendations exist
regarding the maximum duration of night shifts, and similar parameters. To a lesser
extent, comparable research and guidelines are available for evening and morning shifts.
Currently, to study the effects of shift timing, and to develop guidelines for scheduling,
classifications of shift types are used broadly. While types of shifts are relatively
standardized in certain industries, e.g.., heavy industry, hospitals, shifts in other
sectors is way more divers with hundreds or thousands of different shift times. Typically,
legal definitions of night shifts are not helpful either. They often rely on timeframes,
such as requiring at least some hours of work within a predefined night period (e.g.,
Austria's law NSchG defines a night as 6 hours between 22:00 – 6:00). However, these
definitions exclude working patterns that have similar consequences as traditional
night shifts, particularly in service industries. Furthermore, using such definitions,
small variations in shift timing can lead to significant classification inconsistencies.
For example, under a definition where a night shift requires work between 23:00 and
05:00, a shift from 23:00 to 04:55 would not qualify, despite having nearly identical
physiological and psychosocial consequences as one ending at 05:00. Given these limitations,
there is a clear need for a more nuanced classification system that avoids discontinuities,
accurately reflects the diversity of real- world working hours, and facilitates the
application of existing research findings to a broader range of shift patterns.
Methods: As part of the Accident Risk Calculator / Risikorechner 2019 project, we developed
software to estimate the degree to which a given shift (regardless of start and end
time) resembles the consequences of a classical night, evening, morning, or day shift.
This model accounts for its impact on sleep and psychosocial well-being. To enable
further testing and comparative analysis, we have reimplemented the 2019 model as
an open-source Python module, ensuring proper documentation and accessibility for
broader research applications.
Results: The developed module allows for the classification of diverse working hours. It is
designed for straightforward application across different industries and working time
structures.
Conclusion: We invite other research groups to test and refine this initial version. Collaborative
improvements will contribute to a more comprehensive classification framework that
better captures the complexity of modern work schedules and their effects on health
and safety. Support: The projects was partly funded by the German Federal Institute
for Occupational Safety and Health (BAuA).
