  |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
10-October-2003, Vol 116 No 1183
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Adverse events in New Zealand public hospitals II:
preventability and clinical context
Peter Davis, Roy Lay-Yee, Robin Briant, Wasan Ali, Alastair
Scott and Stephan Schug
The subject of patient safety has received increasing
attention in New Zealand. Thus, a number of public inquiries into hospital care
have highlighted quality issues, most recently at Christchurch and Gisborne
hospitals.1,2 Furthermore, in the 1990s the
Ministry of Health investigated the introduction of various indicators of the
safety performance of hospitals, such as a risk-adjusted mortality, readmission
and complication indices.3 Despite the evident
activity of public authorities, however, there has – with some notable
exceptions4 – been little systematic
research in this area.
An important advance has been the development of
standardised procedures capable of eliciting from the audit of medical records a
generic patient safety outcome – the adverse event – that has
reasonable measurement properties, is susceptible to epidemiological analysis,
and has clinical relevance.5 A further
development has been the increasing application of these standardised procedures
within a quality, rather than a medicolegal, framework. With this evolution of
available methodology and focus has come greater interest in the causation and
preventability of adverse events.6
In recently published work the application of this
standardised methodology to hospital admissions in New Zealand public hospitals
demonstrated that 12.9% (850 from 6579 admissions) were associated with an
adverse event.4 The objective of the current
investigation is to assess these same adverse events as to their preventability,
with a special focus on incidents of ‘in-hospital’ origin. As
further assistance to the determination of causation, information is provided on
the broad clinical context of adverse events. As in the previous publication in
this series, only qualitative and indicative judgements are made from the
data.
MethodsThe survey population was
defined as all patients admitted in calendar year 1998 to 20 general hospitals
with more than 100 beds (excluding day, psychiatric and rehabilitation-only
cases). A random sample of 13 hospitals was then selected following
stratification by hospital type and geographical area. The sampling frame for
each hospital was a list of all eligible admissions in that hospital, ordered by
admission date. Approximately 575 admissions were randomly selected from each
hospital, resulting in 6579 medical records for assessment. Experienced and
specially trained nurse screeners and physician reviewers, using a standardised
protocol, then carried out a two-stage retrospective review of those records.
Data collection took place over a period of three weeks at each hospital.
Detailed information on the sample design, availability of routine
administrative data, and method of primary data collection is reported in an
earlier paper.4
The degree to which adverse events could be judged to
be preventable was a key aspect of this study. Preventability, and possible
means of preventing recurrence, were assessed by medical reviewers. A structured
set of ten preparatory questions was designed to assist reviewers in this task.
The questions were based on: consensus (ie, wide professional acceptance, about
diagnosis and treatment); case complexity; appropriateness of management;
comorbidity; and urgency and risk of
management.7
Definitions An
adverse event (AE) was operationally defined as (1) an unintended injury, (2)
resulting in disability, and (3) caused by healthcare management rather than the
underlying disease process. Each of these criteria needed to be fulfilled. To be
included in the study an AE had to be related to, or have occurred during, the
sampled admission. An AE that occurred during the sampled admission was defined
to be an ‘in-hospital’ event. An AE that occurred, for example, in a
community setting and resulted in the sampled admission was defined as an
‘outside-hospital’ event.
Disability was defined as temporary, lasting up to a
year, or permanent impairment of function, or death. The level of disability was
assessed from details documented in the medical record. Preventability of an AE
was assessed as an error in healthcare management due to failure to follow
accepted practice at an individual or system level.
In assessing impact on patients of AEs, three criteria
are drawn upon in the tables that follow: proportion of hospital stay affected;
extra bed days attributable to the AE; and effect on patient health status (for
example, permanent disability or death).
‘System’ was defined in two different
contexts according to the study protocol. First, an AE was classified into
‘clinical areas’, including whether it was a result of ‘system
error’ (ie, due to inadequacies of equipment/supplies, communication,
training/supervision, staffing, scheduling, or functioning of hospital services;
or to lack of or failure to implement a protocol). Second, in relation to areas
in which effort could be directed to prevent AE recurrence, ‘system
areas’ encompassed policies/protocols, information and communication
issues, and organisation management/culture.
ResultsInformation on the attribution of
preventability to AEs by medical reviewers is outlined in Table 1. For nearly
40% of cases reviewers could detect no evidence of preventability. For over one
third of events, however, the judgement of preventabilty was greater than 50:50.
Examples of synthetic cases of varying preventability are provided in Figure
1.
Table 1. Adverse events (AEs): attribution of
preventability
Figure 1. Examples of high- and low-preventability
adverse events
*INR = international normalised ratio; this is a test
for fine-tuning warfarin dose
Taking ‘greater than 50:50’ as the study measure
of preventability (termed ‘high preventability’), Table 2 examines
the demographic and clinical distribution of preventable AEs. While the
occurrence of preventable AEs appeared to be higher for older patients, for
cases arising outside a public hospital, and for certain diagnostic categories,
these differences were not striking, and there seemed to be little systematic
variation by gender, ethnic group, or deprivation of area of
residence.
Table 2. High preventability of adverse events (AEs),
by age group, gender, ethnic group, area deprivation score, major diagnostic
category (MDC), and location of occurrence
*preventability score 4 to
6
†NZDep96 quintiles were derived from patient domicile codes as a measure of residential area deprivation; quintile 5 represents the highest level of deprivation; 7 cases could not be coded ‡principal diagnosis was classified according to 25 Major Diagnostic Categories (MDCs) derived from Australian AN-DRG 3.1; ordered according to % high preventability In order to provide more focus for the analysis of
preventability and causation, in Table 3 the data set is restricted to those
events that occurred in hospital and that were preventable (ie, any evidence of
preventability). The number of events in the analysis approximately halved
– from 850 to 413 – producing an occurrence rate of 6.3%. This rate
appeared to increase with age, as did the impact of these events (longer
hospital stay and higher levels of permanent disability and death). Variations
in occurrence and impact were evident by diagnostic category, although these
were not notably consistent across criteria. Thus injuries had high event
occurrence, but relatively low impact. The major diagnostic categories of
digestive, kidney and urinary, respiratory, and nervous system had high impact
on patient outcome, but no consistent level of effect on either event occurrence
or extended hospital stay.
Table 3. Occurrence and impact of preventable
in-hospital adverse event (AE) rates, by age group and major diagnostic category
(MDC)
*AEs that occurred inside public hospital,
preventability score 2 to 6
†extra bed days associated with an adverse event that were spent in the study hospital during one or more admissions; 5 cases had missing data ‡principal diagnosis was classified according to 25 Major Diagnostic Categories (MDCs) derived from Australian AN-DRG 3.1; ordered according to AE rate The issue of clinical context is taken up in more detail in
Table 4, with AEs broadly classified by both clinical area and specialty.
Furthermore, greater focus on aetiology is provided by presenting analysis of
preventable in-hospital events as well. Data on AE ‘mentions’
– including instances in which more than one area was affected – are
also addressed.
Considering first the distribution of AEs by specialty,
about 60% were associated with surgery, both for all events and for those
classified as ‘preventable in-hospital’. Most events were classified
in either medicine or surgery. Operative and system event mentions were typical
of surgery, while system and drugs were characteristic of medicine. However,
medicine had a wider range of affected areas, with therapy, diagnosis and
procedure also identified. Overall, both for all events and for those classified
as preventable in-hospital, operative and system areas were most frequently
mentioned. However, compared with all AEs, system problems were more prominent
for the in-hospital analysis, as were errors in therapy and diagnosis.
Table 4. Distribution of all adverse events (AEs) and
preventable in-hospital AEs: clinical areas by specialty
*operative = related to an operation or occurred during
the 30-day post-operative period; system = defective equipment or supplies,
equipment or supplies not available, inadequate reporting or communication,
inadequate training or supervision of doctor/other personnel, delay in provision
or scheduling of services, inadequate staffing, inadequate functioning of
hospital services, no protocol/failure to implement protocol or plan, other;
therapy = correct diagnosis but inappropriate or delayed treatment; other =
falls, fractures, obstetrics, neonatal, or
anaesthesia
†includes all surgical specialties plus anaesthiology and obstetrics ‡includes all medical specialties plus psychiatry and paediatrics §includes dentistry/oral surgery, dietary, hospital physical plant, midwifery, nursing, pharmacy, occupational therapy, podiatry, transportation support services, speech/language therapy ║a ‘mention’ refers to an instance where a reviewer identified a particular clinical area. Clinical areas are mutually exclusive except that ‘system’ may be identified alone or in addition. Note that the total number of mentions is greater than the total of AEs ¶AEs that occurred inside public hospital, preventability score 2 to 6 Table 5 assesses impact – as reflected in patient
outcome and extended stay in hospital – preventability and location, by
specialty and clinical area. Considering all AEs by specialty, incidents in
medicine tended to have a greater impact on patient outcomes, and were also more
preventable. Compared with surgery, however, a much lower proportion of these
took place in a public hospital. For preventable in-hospital events, a similar
impact profile to that for all AEs was evident.
In the case of clinical area, system issues were rated
highly preventable for all events, while operative incidents were not. It should
be noted that only about one half of drug-related events occurred inside a
public hospital. There was, however, no marked pattern of impact by clinical
area. For preventable in-hospital events, operative incidents had a higher than
average impact on patient outcomes, although not markedly so for attributable
bed days.
Table 5. Impact, preventability, and location of
occurrence of adverse events (AEs), by specialty and clinical area
*extra bed days associated with an adverse event that
were spent in the study hospital during one or more admissions (11 and 5 cases
respectively had missing
data)
†preventability score 4 to 6 ‡system may be identified alone or in addition to another clinical area §AEs that occurred inside public hospital, preventability score 2 to 6 The potential for prevention is an essential aspect of the
assessment of the preventability of AEs. This was determined by study reviewers
considering likely ‘areas of effort’ for preventing the recurrence
of incidents, and these data are reported in Table 6. Alongside these areas are
considered impact, preventability, and location. Nearly one third of all events
were seen as having a system component, followed by deficiencies in consultation
with colleagues, and education. In the case of preventable in-hospital events,
nearly one half were affected by system issues.
On impact criteria, only ‘consultation’ stood
out for patient outcomes, and problems with ‘resources’ for extended
stay. This was the case both for all events and for preventable in-hospital
incidents. For all events, most areas of effort showed high levels of
preventability and occurred inside a public hospital.
Table 6. Prevention of recurrence of all adverse events
(AEs) and preventable in-hospital AEs: areas of effort by impact and
preventability
*the percentage for each area is the percentage of all
AEs; more than one area could be identified so the percentages do not add to
100%; system = policies/protocols, access to or transfer of information,
communication, discharge procedures/protocols, organisation management/culture,
record-keeping, other; consultation = with specialists or peers; resources =
personnel, equipment/other physical resources,
other
†extra bed days associated with an adverse event that were spent in the study hospital during one or more admissions (11 and 5 cases respectively had missing data) ‡preventability score 4 to 6 §AEs that occurred inside public hospital, preventability score 2 to 6 DiscussionNearly 40% of AEs identified in
admissions records to public hospitals in New Zealand were judged by study
reviewers to be associated with a significant degree of preventability (Table
1). This result is consistent with findings for the United States
(27.6%),8 Australia
(51%),9 and the United Kingdom
(48%).10 Of all sampled admissions, 6.3% were
associated with preventable events of in-hospital origin, a level approximately
half that of the overall rate of occurrence (12.9%).
Older patients suffered more highly preventable events, a
higher rate of preventable in-hospital AEs, and experienced a greater level of
impact – as reflected in permanent disability or death and in extended
hospital stay (Tables 2 and 3). High preventability – and preventable
in-hospital event rates – also varied by diagnostic category, as did
patient impact. The vulnerability of older patients to preventable AEs has been
confirmed in a number of studies internationally. In the United States a
comparator study determined that proportionately more hospitalised elderly
patients experienced preventable AEs, and experienced permanent disability or
death as a result, than did the non-elderly.11
More generally, medical injuries were more common and more often preventable
among elderly hospitalised patients (of 65 years and over) than in younger
patients.12 Although increasing age was a
factor in the proportion of admissions resulting in serious disability or death
in the QAHCS (Quality in Australian Health Care Study), the proportion of AEs
with high preventabililty was not strongly associated with
age.9 In a British study of hospital patients,
those with AEs were older than patients who did not experience an
AE.10
The majority of all AEs occurred in surgery. These were
predominantly operative, were generally less preventable, and had lower impact
than events in medicine, and they occurred almost exclusively inside public
hospitals. Over one third of events were in medicine. Many of these were drug
related, but therapy and diagnosis were also implicated. These events had
greater impact, were more preventable, and a higher proportion occurred outside
a public hospital than events in surgery (Tables 4 and 5).
These results are consistent with the findings from the
QAHCS,9 in which the majority of AEs occurred
in surgery, with half of these being operative, and such events being less
preventable and associated with lower impact. The converse held for AEs in
medicine, among which a high proportion (nearly 16%) occurred outside hospital.
A similar pattern was apparent in research in the UK, where the majority of AEs
detected were in surgery (and orthopaedics), were less preventable, but led to
more extra bed days in hospital.10 The findings
from the US studies also supported these results: most AEs occurred in surgery,
were mainly operative, had lesser impact, and were less preventable than those
occurring in medicine (where adverse drug events were
predominant).13,14
System-related issues were an important clinical area common
to both surgery and medicine in this investigation. These showed high levels of
preventability. They were also reflected in the assessments of prevention of
recurrence, where one third of all events and half of preventable in-hospital
incidents were judged to have system involvement (Table 6). In the Australian
study, system errors produced higher proportions of AEs with permanent
disability and high preventability, and were responsible for 16% of all
events.9 Similarly, in the Harvard Medical
Practice Study, one third of the AEs due to negligent care, and a similar
proportion associated with severe disability, were linked to systems and related
problems.14
The strengths and limitations of the general methodology
being applied in this study have been discussed
elsewhere.4,7 It should also be noted that,
despite the demonstrated weakness of the study instrument in measuring AE
rates,15 there were consistency and
plausibility in key areas. Thus, reviewers were more likely to attribute
preventability to events that were judged to show relative professional
consensus, inappropriate management, substantial deviation from accepted
management, or lack of acceptability of treatment; all conditions plausibly
related to the assessment of
preventability.7
The results of the investigation reported here establish
that a significant proportion of admissions to public hospitals in New Zealand
are associated with preventable AEs, that about half of these have their origins
inside hospital, and that a high proportion are system related. This indicates
that the issue of medical error is as significant in hospital practice in New
Zealand as it is elsewhere. The substantial costs of such incidents have since
been quantified.16
While the study establishes the epidemiology and impact of
preventable AEs in New Zealand public hospitals, it provides only an initial
indication of required remedial action, and the judgements from the data have
been qualitative rather than definitive. Thus, in assessing the prevention of
recurrence, study reviewers identified as potential areas for improvement
‘system’ factors, followed in importance by the need for
consultation with colleagues, and also for education. Clearly, such judgements
are indicative only, and more focused investigations are required in order to
better identify the potential for quality improvement. An area in which
substantial progress has been made internationally is in reducing adverse drug
events.17 Infection control is another area of
considerable potential, where the impact of hospital-acquired infections can be
assessed for further development of preventive and control strategies. Various
studies report that nosocomial infections are common and are associated with
substantial morbidity, high costs,18,19 and
mortality.20 There is potential here for
research to reduce in-hospital AEs.
In summary, just over 5% of admissions to public hospitals
in New Zealand are associated with a preventable AE of in-hospital origin. This
study establishes a baseline and provides indications of the potential for
quality improvement, but few clear-cut signposts for intervention. Nevertheless,
older patients are an area of obvious concern, as is the potential for reducing
AEs occurring outside hospitals.
Author information:
Peter Davis, Professor, Department of Public Health and General Practice,
Christchurch School of Medicine and Health Sciences, University of Otago; Roy
Lay-Yee, Analyst; Robin Briant, Clinical Director, School of Population Health,
Faculty of Medical and Health Sciences, University of Auckland; Wasan Ali,
Assistant Research Fellow, NZHTA, Department of Public Health and General
Practice, Christchurch School of Medicine and Health Sciences, University of
Otago; Alastair Scott, Professor, Department of Statistics, University of
Auckland; Stephan Schug, Professor, Department of Anaesthesia, University of
Western Australia, Perth, Australia
Acknowledgements:
Work on this study was funded by the Health Research Council of New
Zealand. We thank the 13 New Zealand hospitals that participated in the study
and Dr David Richmond (Chair) and members of the study’s Advisory and
Monitoring Committee. We also thank Sandra Johnson and Wendy Bingley, our
medical review and data processing teams, and hospital records staff.
Correspondence:
Professor Peter Davis, Department of Public Health and General Practice,
Christchurch School of Medicine, University of Otago, PO Box 4345, Christchurch.
Fax: (03) 364 0425; email: peter.davis@chmeds.ac.nz
References:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|