[Thesis]. Manchester, UK: The University of Manchester; 2018.
Despite the increased focus on medication safety in the last two decades, it is difficult
to know if anticipated improvements are occurring as medication safety is not routinely
measured in healthcare settings. To address this issue within the UK National Health
Service (NHS), the Medication Safety Thermometer (MedsST) was developed in 2013 with
the aim of enabling organisations to collect routine data to monitor medication safety
and related improvement over time. Guidance about use of the MedsST was developed,
but a knowledge gap existed about how it had been implemented into practice, used
to monitor medication safety and to facilitate improvements. This programme of research
aimed to examine these issues by conducting four related studies.
The initial part of the programme of research explored how the MedsST has been designed,
developed and implemented nationally. Study One investigated the design, development
and national implementation of the MedsST and found that measuring harm from medication
errors is complex and requires several steps to measure individual errors, triggers
of harm and actual harm. Improvement science methods, particularly Plan-Do-Study-Act
cycles, were found to be useful for developing complex systems. Study Two was a qualitative
study that explored how the MedsST had been implemented within individual organisations.
This study found that all staff involved with the MedsST understood what the tool
was and why measurement was vital for facilitating improvements in medication safety.
However, less understanding existed about how MedsST data could be used for improvement.
Several issues with the MedsST implementation were also highlighted such as it being
unsuitable for use in primary care settings.
The second part of the programme of research investigated how MedsST data were used
or could be used for learning about and improving medication safety. Study Three was
a qualitative study that used quantitative MedsST data to find out how data had been
used for improving medication safety in hospitals. The study found that only a small
amount of data had been used for improvement, and that this was often at ward-level.
Although some improvement had occurred, communication about improvements was poor
and most data remained not viewed and unused. Study Four was quantitative and used
nationally aggregated MedsST data to determine the prevalence, nature and predictors
of patients experiencing medication administration omissions in hospitals, as an exemplar
of how MedsST data could be used to learn about medication safety issues. It was found
that 30% of patients experienced omissions (95% confidence interval [CI] 29-30) (excluding
valid clinical reasons). The rates found were similar to that of previous research,
reiterating that omissions are a substantial problem.
The research presented in this thesis demonstrated that there had been success in
the implementation of the MedsST for data collection, but limited success in terms
of using collected data for learning and improvement purposes. In order for more improvements
in medication safety to occur, more work needs to be done within the NHS to successfully
implement a system of data collection, review and use of MedsST data as a holistic
system. This thesis has provided specific recommendations to increase engagement with
this holistic system, and for healthcare organisations and researchers to benefit
from collected MedsST data with the aim of improving medication safety.