The importance of better understanding error and safety in the community setting is
widely accepted, with recent calls to promote efforts and improve resources in this
area of research (Jacobson, Elwyn et al. 2003). The measurement of patient safety
events in primary care is a relatively under-researched area and it is well recognized
that there are large gaps in the research describing patient safety in ambulatory
settings (Hammons, Piland et al. 2003). Attitudes towards embracing safety event
measurement have improved in recent years, however there remains a substantial
amount of work to be done before common standards can be recommended, despite
recent calls in the scientific literature for national and international systems
(Runciman, Williamson et al. 2006).
This thesis describes the Threats to Australian Patient Safety (TAPS) study, which
aimed to create a secure anonymous web-based error reporting system suited to the
Australian general practice setting, and then describe and quantify the errors reported
by a representative random sample of Australian general practitioners.
The study was made possible with the support of funding from a National Health and
Medical Research Council project grant, and also gained support from NSW Health
and the Commonwealth Department of Health and Aging in the form of granting
qualified privilege and providing essential Medicare data under legal instrument.
The study methodology involved the development of a database management system
which created an electronic method for managing and analysing a wide variety of
features related to large numbers of anonymously reported errors from Australian
general practice. A representative random sample of 84 general practitioners (GPs)
from New South Wales (NSW) participated in the study, with over 400 errors
reported in a 12 month period.
The key messages arising from the TAPS study were:
• GPs embraced anonymous patient safety event reporting using a secure
website, with the majority of study participants making reports
• New findings from this study on the incidence of reported error in general
practice were published in the scientific literature, which will help guide the
design of future error reporting systems
• A new taxonomy to describe reported error from GPs was developed as part of
this study and published in the scientific literature, with the view of allowing
future self-coding of reported patient safety events by GPs
The TAPS study presented the first calculations known worldwide of the incidence of
reported error in a general practice setting using a representative random sample of
general practitioners. It was found that if an anonymous, secure, web-based reporting
system was provided, approximately 2 errors were reported by general practitioners
per 1000 patients seen per year (Makeham, Kidd et al. 2006).
In addition, the study created a simple descriptive general practice based error
taxonomy, entitled the TAPS taxonomy (see Appendix 10) (Makeham, Stromer et al.
2007), and was the first study to test the reproducibility of the application of such a
tool using a group of general practitioners. The TAPS taxonomy developed as part of
this study was found to have a good level of inter-coder agreement.
With respect to the underlying causes of errors, the TAPS study found that the
majority of reported patient safety events were errors related to the processes of health
care (70%), rather than errors related to the knowledge and skills of health
Most errors reported in the TAPS study had the direct involvement of a patient (93%
of error reports). Overall the reporting general practitioners were very familiar with
these patients, who were on average 52 years old, and more often female (56%).
Around one quarter of the errors reported was associated with patients being harmed.
Reports containing events related to processes of health care were associated less with
harm than those containing events related to the knowledge and skills of health
The patients in errors associated with patient harm reported in the TAPS study were
on average older than patients in reports where no harm was known to have occurred
(58 years versus 50 years respectively). There was no statistically significant
difference found between these groups with respect to gender or ethnicity, including
people from Non-English speaking backgrounds or Aboriginal and Torres Strait
Islander (ATSI) peoples, although the association with the latter group approached
Cases of patient death were reported in 8 of 415 errors reported in the TAPS study
(2%), and more often involved events relating to the knowledge and skills of health
professionals than events relating to the processes of health care compared to reports
not involving a known patient death.
In support of suggestions in the scientific literature about the importance of
anonymity as a feature of an error reporting system, a feedback interview found that
an anonymous reporting system was a factor which made participants more likely to
report error events, with two thirds of participants agreeing that anonymity made them
more likely to participate in reporting. The majority of participants found the
reporting process easy to undertake, and took approximately 6 minutes to send a
The study provided a self directed learning educational activity for participating
general practitioners that was approved for 30 group 1 Quality Assurance and
Continuing Education points by the Royal Australian College of General Practitioners
An important practical outcome of the TAPS study was that it highlighted a
systematic error relating to immunisation failures with meningococcal vaccines which
was reported to relevant organisations including NSW Health, the RACGP and the
manufacturer involved, which was addressed with educational materials for GPs being
distributed and communication in Australian Family Physician.
There are further analyses that could be undertaken using the TAPS data to improve
our understanding of the errors reported, such as further statistical analyses using
techniques such as building a model with multiple regression to determine significant
factors that contribute to different error types. This work was beyond the scope of the
TAPS study aims, but is part of further research recommendations.
In addition, future studies should address aspects of patient safety and reported error
that it would not be possible to capture from the perspective of the reporting GP.
Rather than one taxonomy which describes the reported errors from the GP’s
perspective in the way that the TAPS taxonomy does, it may be useful to develop a
series of interlinked taxonomies that are directed to the needs of differing
constituencies, such as the organisation providing health funds or the health insurer,
the health regulators and legislators, and the patients or their significant others.
The assessment of potential and actual harms sustained by patients involved in
reported errors is a further area of patient safety research that is difficult to
comprehensively assess, and existing reporting systems in the literature, whilst
addressing this from the reporter’s perspective, require further work to improve the
accuracy by which harm is measured and correlated with other data sets such as those
managed by health insurers, and the experiences of people who are the subject of the
The TAPS study presents a number of new findings about the nature of error and
threats to patient safety that arise in the Australian health care environment, reported
by a representative sample of general practitioners, and it is hoped that these will be
useful to all stakeholders in the health care setting, from clinicians, through to policy
makers, and most importantly the patients who are the subject of the potentially
preventable harms and near misses that are highlighted in this thesis