  |
|
|
||||
|
||||||
| Journal of the New Zealand Medical Association,
21-July-2006, Vol 119 No 1238
|
||||||
|
Community hospital versus tertiary hospital comparison in the
treatment and outcome of patients with acute coronary syndrome: a New Zealand
experience
Eng Wei Tang, Cheuk-Kit Wong, Peter Herbison
Information derived from randomised clinical trials are used
to formulate recommended treatment guidelines1,2 for the management of acute
coronary syndrome (ACS). Registry studies published locally3–5 and
overseas6–8 have demonstrated the underuse of evidence-based therapies
(both medications and revascularisation) in ‘real-life’ high-risk
ACS patients. The lower rate of revascularisation in community hospitals could
be due to the lack of facilities, necessitating transfer of patients to tertiary
centres.
Although revascularisation is generally beneficial for
high-risk ACS patients,3,4 the GRACE registry9 showed similar short-term (6
months) outcome between patients admitted into community hospitals for whom
revascularisation required transferral and patients admitted directly into
tertiary centres.
There is currently no local data comparing the treatment and
outcome of ACS patients admitted into a community hospital (staffed by general
physician without catheterisation facilities) versus a tertiary teaching
hospital (with cardiologists and catheterisation laboratory). However this
information is crucial to both the public and health professionals /
policymakers when we strive to offer an equitable healthcare service to all New
Zealanders.
Hence the aim of this study was to compare the management
(use of evidence-based medications and revascularisation) and mortality of
patients with ACS managed in a community hospital versus those managed in a
tertiary hospital.
MethodsThis is a retrospective registry study including all
consecutive patients with ACS admitted into the coronary care units (CCUs) of
two related centres in New Zealand, including the tertiary hospital in Dunedin,
Otago Province (Dunedin Hospital) and the regional hospital in Invercargill,
Southland Province (Invercargill Hospital) from the years 2000–2002
inclusive. Dunedin Hospital served as the referral centre for Invercargill
Hospital during the study. Stable patients were transferred from Invercargill to
Dunedin via ambulance, a 3-hour journey. Unstable patients were retrieved via
helicopter (Dunedin Hospital has a helipad).
All patients had ACS as their discharge diagnosis and
were above 18 years of age. Patients having ACS precipitated by significant
non-cardiac comorbidity, trauma, or surgery were excluded. The first admission
with ACS was used for analysis if readmissions with ACS were present. Patients
initially admitted to Invercargill Hospital and subsequently transferred to
Dunedin Hospital were categorised as Invercargill patients. This study protocol
was in accordance with local hospital research guidelines.
All clinical data were collected by a research
physician, which includes:
Death was defined as all-cause mortality
during hospitalisation and over a 1-year follow-up period. Information on the
deaths was obtained from medical records and the national death registry.
Definition of
ACS—Patients with ACS were classified into:
Statistical
analysis—Statistical analysis was carried out in STATA v8 software.
Data are presented as mean with 95% confidence intervals, or median with
interquartile ranges or actual numbers with percentages as appropriate.
Chi-squared tests were used to compare proportions, and t-tests or Mann-Whitney
U tests for continuous data. The test was double-sided and considered to be
statistically significant at α<0.05.
Multivariable analysis was performed using Poisson
regression with robust standard errors for patients who died in hospital and
Cox’s proportional hazards regression for hospital survival up to 1-year
to examine the association between admission to Invercargill Hospital vs Dunedin
Hospital and mortality during in-hospital stay, 6-months, and 1-year follow-up.
These models were adjusted for age, sex, history of
hypertension, diabetes mellitus, history of ischaemic heart disease, history of
coronary artery bypass graft, history of stroke, heart rate, systolic blood
pressure, Killip class and cardiac arrest on presentation, maximum troponin
elevation, renal impairment, and the subset of acute coronary syndrome.
Further models were run with adjustments for the use of
evidence-based medications in the first 24 hours of admission and the use of
coronary angiography, as well as those already in the model.
ResultsPatients—From
1 January 2000 to 31 December 2002, 843 patients and 299 patients were admitted
into the tertiary (Dunedin Hospital) and community (Invercargill Hospital) CCUs
respectively (Table 1).
There was no difference in age, sex—and history of
hypertension, diabetes, smoking, dyslipidaemia, cerebral vascular disease, or
peripheral vascular disease—between patients admitted to Dunedin Hospital
versus patients admitted to Invercargill Hospital.
More patients admitted to Dunedin Hospital had a history of
ischaemic heart disease (49.2% vs 40.5%, p=0.010), previous coronary artery
bypass graft (10.1% vs 2.0%, p<0.0005) and renal impairment (23.5% vs 15.4%,
p=0.003).
There were a higher proportion of patients with STEMI (56.5%
vs 33.3%) or NSTEMI (34.5% vs 26.1%, p<0.0005 for both) amongst patients
admitted to Invercargill Hospital. Furthermore, there was a higher incidence of
cardiac arrest (12.7 vs 7.3%, P<0.004) and clinical heart failure (24.1% vs
17.8%, p=0.039).
Amongst patients with non-ST-elevation acute coronary
syndrome (NSTEACS), those admitted into Invercargill more frequently had
presenting ECG showing ischaemic ST depression (p=0.006).
Management—In
the first 24-hours, patients with STEMI were as likely to receive aspirin,
beta-blockers, statins, and thrombolysis in both centres. However, the use of
statins in both hospitals was low (21.1% vs 17.2%, p=0.312).
In patients admitted into Dunedin Hospital, percutanous
coronary intervention (PCI) or coronary artery bypass graft (CABG) during the
index admission was more frequently performed (46.4% vs 16.6%, p<0.0005;
Table 2). In addition, patients in Dunedin Hospital were more likely to be
discharged on statins (66.4% vs 42.3%, p<0.0005) or an ACE-inhibitor (60.2%
vs 49.7%, p=0.040; Table 2) than patients admitted to Invercargill Hospital,
despite having a similar incidence of dyslipidaemia (60.4% vs 60.7%, p=0.940),
clinical heart failure (Killip class >1) on admission (20.7% vs 18.9%
p=0.737) and echocardiographic measurements of left ventricular systolic
dysfunction (p=0.772; Table 2).
Table 1. Baseline characteristics of ACS patients
 *Age and creatinine levels
were reported as mean and 95% confidence intervals
Table 2. Evidence-based medicine and revascularisation
procedures
 Table 3. Baseline characteristics of NSTEACS patients
from the community hospital (Invercargill Hospital) transferred and not
transferred to the tertiary centre (Dunedin Hospital)
 *Age and creatinine levels
were reported as mean and 95% confidence intervals
For patients with non-ST-elevation acute coronary syndrome
(NSTEACS), there were no difference in the use of aspirin, beta-blockers, and
heparin during the first 24 hours amongst the two centres. For patients admitted
to Dunedin Hospital, the use of statins was more frequent both in the first 24
hours of admission (37.0% vs 23.1%, p<0.003) and on discharge (64.4% vs
40.7%, p<0.0005), despite similar rates of dyslipidaemia (62.4% vs 67.7%,
p=0.303). More patients with NSTEACS received revascularisation (PCI/CABG)
during index admission in Dunedin Hospital (48.9% vs 18.5%, p<0.0005).
Patients with NSTEMI admitted into Dunedin Hospital were
more likely to receive clopidogrel (27.4% vs 0%, p<0.0005) and glycoprotein
IIbIIIa (20.1% vs 8.7%, P=0.008), which parallel the higher revascularisation
rate: PCI (37.8% vs. 14.1%, P<0.0005) and CABG (19.8% vs. 8.3%,
p=0.008).
Demographics of NSTEACS
patients transferred to a tertiary centre—Compared to patients not
transferred to Dunedin Hospital, transferred patients were younger (60 vs 68,
p=0.003); with a lower incidence of diabetes mellitus (15.2% vs 27.8%, p=0.144);
more clinically stable with a lower heart rate on hospital admission (71 vs 82
beats/minute, p=0.030); and a lower incidence of clinical heart failure (12.1%
vs. 37.1%, p=0.024, Table 3). Their mean creatinine level was also
significantly lower
(87 vs 110 μmol/L, p=0.012).
Clinical
outcome—The mortality rate for ACS is significantly higher for
patients admitted into Invercargill Hospital during index admission, at 6 months
and at 1 year (Table 4). At 1-year, there was a 10.0% absolutely mortality
difference between the two hospitals (12.1% vs 22.1%, p<0.0005). The adjusted
and unadjusted hazard ratios for death are shown in Table
4.
Amongst the subgroup of ACS, patients with NSTEMI (but not
STEMI or unstable angina) in Invercargill Hospital have a significantly higher
in-hospital, 6-months, and 1-year mortality and risk of dying compared to
patients admitted into Dunedin Hospital (Table 5 and Table 6).
Adjusting for patient’s baseline characteristics
(including cardiac arrest on arrival) and use of evidence-based medications
during the first 24 hours, the higher risk of death remains for patients
admitted to Invercargill Hospital without catheterisation facilities (Table 4).
After further adjusting for angiography and/or
revascularisation, the higher risk of death for Invercargill patients remains
significant. Patients first admitted there were at a 202% increased risk of
death at 1-year (adjusted hazard ratio 3.02, 1.60–5.71).
Table 5. Mortality figures (percentages) for ACS
patients at different follow-up periods after being admitted into a tertiary
(catheterisation laboratory/cardiologists) vs a community (general physicians)
hospital
 Table 6. Adjusted hazard ratio (95% confidence
interval) for patients first admitted into a community hospital vs a tertiary
hospital, according to final ACS diagnosis
 DiscussionIn this study, patients with ACS admitted into a tertiary
hospital(Dunedin Hospital) have significantly lower in-hospital, 6-months, and
1-year mortality (absolute mortality difference of 4.3%, 9.5%, and 10.0%
respectively) compared to patients admitted into a community
hospital(Invercargill Hospital).
This difference in survival outcome cannot be accounted for
by the baseline patient characteristics because our multivariable regression
model adjusted for multiple important prognostic factors. These included age and
sex; history of hypertension, diabetes mellitus, ischaemic heart disease,
coronary artery bypass graft, and stroke; heart rate and systolic blood pressure
on admission; Killip class and cardiac arrest on arrival; maximum troponin
elevation; renal impairment; and the subset of acute coronary syndrome.
We further adjusted our model for the use of evidence-based
medicine in the first 24 hours and found an over two-fold increase in 1-year
mortality in patients admitted to a community hospital.
Invercargill Hospital is a community hospital serving the
Southland districts without any onsite cardiologist or catheterisation
facilities. Referrals for angiography are made to the attending cardiologist in
Dunedin Hospital before patient transfer (a 3-hour ambulance journey) can be
organised.
After adjusting for the use of angiography and
revascularisation, the adjusted hazard ratio for 6-month and 1-year mortality
remains significantly higher in patients initially admitted into Invercargill
Hospital than patients admitted into Dunedin Hospital. This difference in
mortality is particularly marked in the NSTEMI subgroup.
Published New Zealand guidelines on the management of NSTEMI
in 20051 recommends the use of an early invasive strategy for patients with
NSTEMI (positive troponins) showing high risk features including dynamic ST
changes, patients with diabetes, patient with continuous or recurrent ischaemic
symptoms at rest or on exertion despite medical therapy and patients with
clinical evidence of left ventricular failure.
Selecting high-risk patients who would benefit most from
revascularisation is important to maximise the cost-benefit gains from invasive
intervention. A meta-analysis of seven major randomised trials on routine versus
selective invasive approach for patients with NSTEACS10 found a better outcome
with patients managed with a routine invasive approach, in those with positive
cardiac markers such as troponins and those managed with a more contemporary
treatment strategy (studies published after 1999). There was, however, an early
hazard with a routine invasive strategy leading to higher in-hospital mortality
and non-fatal myocardial infarction.
The benefit of routine revascularisation is seen after
discharge, resulting in fewer subsequent deaths or myocardial infarction.
Overall, there was a benefit after a mean follow-up period of 17 months from
hospital admission (the time of randomisation to trials), including a
non-statistically significant 8% relative reduction in death, a significant 18%
reduction in combined death and non-fatal myocardial infarction, and a one-third
reduction in severe angina and rehospitalisation.
Interestingly, the current study also demonstrated a bigger
survival difference at 6 months and 1 year between the two centres at hospital
discharge. This may also reflect the delayed benefit from the more frequent use
of revascularisation in Dunedin Hospital.
Recently the ICTUS11 study found routine invasive
revascularisation strategy (98% catheterisation rate, 76% rate of in-hospital
PCI or CABG) to confer no additional benefit over a selective invasive (53%
catheterisation rate, 40% PCI or CABG) strategy in NSTEMI patients. Of note, the
Dunedin rate of revascularisation for patients with NSTEMI (71% catheterisation,
58% PCI or CABG) represented a midway between routine and selective
revascularisation. In centres with catheterisation facilities in the GRACE9
registry, the average revascularisation rate (PCI or CABG) is 48%.
The significantly higher mortality in Invercargill patients
admitted with NSTEMI (Table 5) is concerning. The in-hospital revascularisation
rate was low (26% catheterisation, 22.4% PCI or CABG) but similar to the
findings from the New Zealand Acute Coronary Syndrome audit performed in 20023
(25% angiography, 11.2% PCI or CABG). However, even after adjusting for transfer
and angiography, the hazard ratio (risk of dying) remains increased.
We found that transferred patients from Invercargill
Hospital were generally lower-risk ACS patients (younger, clinically stable, and
with fewer premorbidities), than those not transferred. These lower-risk
patients would be expected to benefit less from revascularisation, at least
within the one-year follow-up. However, numerous practical issues requires
consideration when transferring high-risk ACS patients, including the need for
sufficient medical escorts, the possible reluctance of medical staff in Dunedin
Hospital to accept patients with significant comorbidities (renal insufficiency
as a contraindication for angiography, severe non-reversible airways disease)
and non-medical logistic reasons (bed availability).
The delay in angiography because of the time required for
transfer may also have impacted negatively on survival, although the optimal
timing for coronary intervention post-NSTEACS is yet to be determined.12,13
The fact that higher-risk ACS patients (who potentially
would derive a greater absolute benefit from angiography) were less likely to
receive these procedures is not unique to New Zealand. This had been shown in a
large survey of 158,831 elderly Medicare patients with acute myocardial
infarction in United States in the mid 1990s.14 In the contemporary CRUSADE
Registry15,16—17,926 patients from 248 United States hospitals with
angiographic facilities—interventions again were more likely to be offered
to younger patients, those under cardiology care and those without heart
failure, renal dysfunction, or ischaemic ECG changes.
Further analysis of CRUSADE registry found high-risk ACS
patients managed conservatively solely with medical therapy had higher
in-hospital mortality.17 It may also be argued that interventions, even when
performed in lower risk patients, will confer prognostic benefit over a longer
time period after the first 1 or 2 years.
Interestingly, patients with interventions were more often
prescribed evidence-based medicine (Table 3), findings similar to those in the
United States.14 The use of evidence-based medicine in ACS patients should be
encouraged even if intervention/revascularisation were not offered.
Study
limitations—This is a retrospective registry study. Although a
vigorous multivariable logistic regression was used incorporating multiple known
prognostic predictors, there were possibly other relevant factors (which we did
not record or adjust for) which may account for the discrepancy in mortality
observed between the two centres.
As this study included only patients admitted into the CCUs
of the two hospitals, issues like bed availability and CCU admission policies
might have influenced our results.
The only documented endpoint was all-cause
mortality—secondary endpoints (such as myocardial infarction, stroke,
major bleeding, recurrent angina, or rehospitalisation) were not studied.
Furthermore, we did not study management (use of medications and
revascularisation) after hospital discharge which could have affected 6-month
and 1-year mortality.
ConclusionThere was a disparity in ACS outcome between community and
tertiary hospitals in New Zealand, even after adjustment has been made for
baseline characteristics, use of evidence-based medicine on arrival, and
transfer for angiography. The use of evidence-based medicine in all ACS patients
should be encouraged even if revascularisation was not offered.
Author information:
Eng Wei Tang, Senior Research Fellow; Cheuk-Kit Wong, Associate Professor in
Medicine; Department of Cardiology, Dunedin School of Medicine, University of
Otago (and Dunedin Public Hospital), Dunedin; Peter Herbison, Statistician
and Associate Professor, Department of Preventive and Social Medicine, Dunedin
School of Medicine, University of Otago, Dunedin. (Eng Wei Tang is now a
consultant cardiologist in Palmerston North Hospital.)
Acknowledgements: Dr
EW Tang received support from The Cardiac Society of Australia and New Zealand /
MSD Fellowship as well as partial support from the University of Otago Frances G
Cotter Scholarship.
Correspondence: Dr
Cheuk-Kit Wong, Associate Professor in Medicine, Department of Cardiology,
Dunedin School of Medicine, University of Otago, Dunedin Public Hospital,
Dunedin. Fax: (03) 474 7655;
email: cheuk-kit.wong@healthotago.co.nz References:
|
|
||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|