Article Text

Download PDFPDF

Safety and feasibility of hospital discharge 2 days following primary percutaneous intervention for ST-segment elevation myocardial infarction
  1. Daniel A Jones1,2,3,
  2. Krishnaraj S Rathod1,
  3. James Philip Howard1,
  4. Sean Gallagher1,3,
  5. Sotiris Antoniou1,
  6. Rodney De Palma1,3,
  7. Oliver Guttmann1,
  8. Samantha Cliffe1,
  9. Judith Colley1,
  10. Jane Butler1,
  11. Eileen Ferguson1,3,
  12. Saidi Mohiddin1,2,3,
  13. Akhil Kapur1,3,
  14. Charles J Knight1,3,
  15. Ajay K Jain1,3,
  16. Martin T Rothman1,3,
  17. Anthony Mathur1,2,3,
  18. Adam D Timmis1,2,3,
  19. Elliot J Smith1,3,
  20. Andrew Wragg1,2,3
  1. 1Department of Cardiology, Barts and the London NHS Trust, London, UK
  2. 2Department of Clinical Pharmacology, William Harvey Research Institute, Queen Mary University, London, UK
  3. 3NIHR Cardiovascular Biomedical Research Unit, London Chest Hospital, London, UK
  1. Correspondence to Dr Andrew Wragg, Department of Cardiology, London Chest Hospital, Bonner Road, Bethnal Green, London E2 9JX, UK; andrew.wragg{at}


Aim Primary percutaneous coronary intervention (PPCI) produces more effective coronary reperfusion and allows immediate risk stratification compared with fibrinolysis. We investigated the safety and feasibility of very early discharge at 2 days following PPCI in selected low-risk cases.

Methods This was a prospective observational cohort study of 2779 patients who underwent PPCI between 2004 and 2011. Patients meeting the following criteria were deemed suitable for very early discharge; TIMI III flow, left ventricle (LF) ejection fraction >40%, and rhythmic and haemodynamic stability out to 48 h. Higher-risk patients who did not fulfil these criteria were discharged later according to physician preference. All patients were offered outpatient review by a multidisciplinary team. Endpoints included 30 day readmission rates and major adverse cardiac events (MACE) out to a median of 2.8 years (IQR range: 1.3–4.4 years).

Results 1309 (49.3%) PPCI patients met very early discharge criteria, of whom 1117 (85.3%) were actually discharged at 2 days. 620 (23.4%) were discharged at 3 days, and 916 (34.5%) >3 days after admission (median 5, IQR: 4–8) days). Patients discharged at 2 days were younger, and had lower rates of diabetes, renal dysfunction, multivessel coronary artery disease, previous myocardial infarction, and previous coronary artery bypass surgery, compared with patients discharged later. 30-day readmission rates for non-MACE events were 4.8%, 4.9% and 4.6% for patients discharged 2 days, 3 days and >3 days after admission, respectively. MACE rates were lowest in patients discharged at 2 days (9.6%, 95% CI 4.7% to 16.6%) compared with patients discharged at 3 days (12.3% 95% CI 6.0% to 19.2%) and >3 days (28.6% 95% CI 22.9% to 34.7%, p<0.0001) after admission.

Conclusions Our data suggest that discharge of low-risk patients 2 days after successful PPCI is feasible and safe. Over 40% of all patients with ST-elevation myocardial infarction may be suitable for early discharge with important implications for healthcare costs.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


International guidelines recommend discharge within 4 days of admission for patients with uncomplicated acute myocardial infarction (MI).1 ,2 This recommendation is based on studies that predate primary percutaneous coronary intervention (PPCI).3–,5 With more effective coronary reperfusion, earlier discharge from hospital has been suggested.6 ,7 Two prospective randomised trials have confirmed that it is safe and cost-effective to discharge low-risk patients within 3 days of PPCI,8 ,9 but the potential for discharge even earlier following MI has not been fully explored. Although discharge prior to day 3 has been adopted in some institutions,10 there remains a paucity of outcome data. In the present study, we report outcomes from a large cohort of consecutive patients undergoing PPCI who were managed according to an active discharge protocol with specific criteria allowing discharge at 2 days. This study does not attempt to investigate the length of hospital stay on outcome. The aim of this study however, was to investigate the safety and feasibility of discharge 2 days following PPCI in selected low-risk patients.


This was a prospective observational study of 2980 consecutive patients undergoing PPCI for ST-segment elevation MI (STEMI) in a single high-volume centre from January 2004 to July 2011. The London Chest hospital is the tertiary heart attack centre for the northeast region of London, and takes all patients with STEMI for primary PCI in an unselected manner. This includes patients with cardiogenic shock and postcardiac arrest, including intubated and ventilated patients. The hospital serves a well developed network of six local district general hospitals covering a population of 1.6 million people and includes close working with the local London ambulance services. There were 2779 patients (93.2%) with complete electronic records that were suitable for analysis. A further 126 patients died in hospital (none of which were proposed for 2 day discharge) leaving 2653 patients who were discharged and available for analysis (figure 1). A sensitivity analysis confirmed that the time to discharge in the 201 patients excluded due to incomplete records was not significantly different compared with the 2653 patients who comprised the study cohort.

Figure 1

Flow diagram of patients treated by PPCI between January 2004 and July 2011.

Standard PPCI protocol included preloading with 300 mg aspirin, 300 mg or 600 mg clopidogrel, and GPIIb/IIIA inhibitors (predominantly intravenous abciximab) unless contraindicated. All patients who received abciximab had baseline full blood counts and repeats at 4 h. Thrombectomy catheter use was at the discretion of the operator and increased significantly over the study time period. Likewise, local guidelines advised routine use of bare metal stents for PPCI, but use of drug eluting stents also increased over the study period. Data entered into the clinical database at the time of PPCI included patient characteristics (age, sex, previous MI, coronary revascularisation and stroke, smoking status, hypertension, diabetes mellitus, hypercholesterolaemia, peripheral vascular disease, New York Heart Association functional class, estimated glomerular filtration rate, left ventricular ejection fraction, cardiogenic shock), procedural factors (indication, target vessel, number of diseased vessels, use of intravascular ultrasound (IVUS)/pressure wire, stent characteristics, antiplatelet therapy), and procedural complications (MI, emergency coronary artery bypass grafts, arterial complications, aortic/coronary dissection, side branch occlusion, heart block arrhythmias requiring direct current (DC) cardioversion). Successful primary PCI result was defined as final TIMI (thrombolysis In myocardial infarction) flow grade 3 and residual stenosis <20% in the infarct-related artery at the end of the procedure. Decisions regarding revascularisation of non-culprit lesions was also at the discretion of the operator as there were no local guidelines. Subsequent revascularisation decisions were based on the original angiographic findings±ischaemia assessment and symptoms.

Very early discharge (2 days) was suggested for patients meeting the following criteria:

  • TIMI III flow in the infarct-related artery.

  • LV ejection fraction >40% (assessed by left ventriculogram at the time of procedure, or echocardiogram within the first  24 h).

  • Rhythmic and hemodynamic stability.

  • Absence of heart failure.

  • Absence of severe comorbidity.

The criteria were derived from pre-existing risk scores used to identify low-risk patients following PPCI, specifically the Zwolle and Grace scores.11 ,12 The ultimate timing of discharge for all patients remained at the discretion of the treating cardiologist. We would like to emphasise that the discharge criteria were for guidance, and there was not a formal decision-making algorithm. After discharge, all patients were offered follow-up 1 week after discharge in a heart attack clinic staffed by a multidisciplinary team that included rehabilitation, heart failure and psychological specialists. Major adverse cardiac events (MACE) were defined as death, recurrent MI (defined as ‘new ischaemic pain with new ST-elevation, or ischaemic ECG changes and further elevation of enzymes (increase of creatine kinase-MB to ≥2 times the reference value or rise in troponin T >30 ng/l), whether treated with further revascularisation therapy or not’), stroke and target vessel revascularisation. These MACE rates were adjudicated by three independent physicians who were not involved in the procedure and were unaware of the patients’ length of hospital stay. All-cause mortality data was recorded as of 11 August 2011 from the UK Office of National Statistics. A retrospective data quality audit of 100 randomly selected medical records established that 94.8% of data fields, including complications, were entered correctly into the database. Drug therapy on discharge was recorded prospectively with 30-day readmissions, postdischarge complications and further revascularisation procedures entered retrospectively from the electronic patient record and cardiac surgical database.


The data were collected as part of a mandatory national cardiac audit, and all patient-identifiable fields were removed prior to analysis. The local ethics committee advised that formal ethical approval was not required.

Statistical analysis

Continuous variables are presented as mean±SD, categorical variables as percentages. Categorical variables were compared among groups using the χ2 test or Fisher's exact test when appropriate, whereas, continuous variables were compared with the analysis of variance test. Kaplan-Meier curves were used to represent survival and cumulative incidence of events over follow-up, with the log rank test used for evidence of a statistically significant difference between the groups. Time was measured from the first admission for a procedure to outcome. We used SPSS V.18 for Mac for all analyses.


A total of 42% of all patients with STEMI treated by PPCI were discharged at 2 days based on simple discharge criteria. Of the 2653 patients in this study, 1327 (49.8%) were immediately stratified to a provisional 2-day discharge plan. At 2 days, 1309 patients (98.6%) were actually suitable for discharge, as 18 patients did not satisfy the 2-day discharge criteria. Of these 1309 patients, 1117 (85.3% of those suitable) were actually discharged at 2 days (figure 1); 150 suitable patients were delayed due to timing issues (admission at unsociable hours (eg, 1:00), and 42 (3.2%) patients suitable by criteria for 2-day discharge were judged not clinically fit for discharge at 2 days after admission. Breakdown of the reasons for delayed discharge are shown in table 1. The 150 patients delayed due to timing issues were discharged on day 3. Of the 42 patients not suitable for discharge at 2 days, all remained inpatients for >3 days. Of these patients provisionally selected for very early discharge, but not suitable at 2 days, six (0.5%) were due to a MACE event, seven (0.5%) due to rhythm disturbances, and five due to heart failure. All these events occurred in the first 24 h after PPCI. There were no inpatient deaths in the cohort of patients initially proposed for 2-day discharge.

Table 1

Reasons for prolonged admission in patients initially proposed for 2-day discharge

Of the remaining patients not discharged at 2 days, 470 (17.8%) patients were discharged 3 days after admission, and 856 (32.3%) >3 days after admission (median 5, IQR: 4–8) days).


Patients discharged 2 days after admission were younger, and more commonly male, with lower rates of diabetes, multivessel disease, previous  MI, and previous coronary bypass surgery compared with patients discharged later. Peak troponin was lower, and estimated ejection fraction higher in patients discharged 2 days after admission, none of whom had cardiogenic shock compared with 9.4% of patients discharged after >3 days (table 2). Prescription rates for secondary prevention drugs were similar, regardless of the timing of discharge (table 3).

Table 2

Baseline characteristics according to length of stay group

Table 3

Patient medication on discharge by length of stay

Procedural characteristics and outcomes

Radial access for PPCI was used more commonly in patients discharged 2 days after admission, among whom anterior infarction and multivessel disease was less common compared with patients discharged later. Procedures were successful in all patients discharged after 2 days, and no inhospital MACE was recorded. This contrasts with patients discharged >3 days after admission who had high rates of arterial complications (hematoma 3.1%, dissection 0.9%, false aneurysm 0.6%), major bleeding events (gastrointestinal bleeding 4.2%, retroperitoneal bleeding 0.9%) and inhospital MACE (recurrent MI 2.4%, target vessel revascularisation 3.1%, stroke 2.7%) (table 4).

Table 4

Procedural characteristics according to length of stay


Patient follow-up was for a median of 2.8 years (IQR range: 1.3–4.4 years). Readmission rates for non-MACE events (including heart failure, troponin negative chest pain syndromes and chest infections) in the first 30 days were 4.8%, 4.9% and 4.6% for patients discharged 2 days, 3 days and >3 days after admission, respectively. Kaplan-Meier analysis showed that MACE rates during follow-up among patients discharged 2 days after admission (7.5% 95% CI 3.2% to 17.7%) were similar to patients discharged after 3 days (12.2% 95% CI 6.1% to 21.5% p=0.07), but significantly lower than patients discharged after >3 days (31.1% 95% CI 23.1% to 35.3% p≤0.0001) The pattern was similar for all MACE components including all-cause mortality, recurrent MI, and target vessel revascularisation (figures 24).

Figure 2

Kaplan-Meier curves showing cumulative probability of major adverse cardiac events (MACE) after PCI according to length of stay. This figure is only reproduced in colour in the online version.

Figure 3

Kaplan-Meier curves showing cumulative probability of all-cause mortality after PPCI according to length of stay. This figure is only reproduced in colour in the online version.

Figure 4

Kaplan-Meier curves showing cumulative incidence of (A) target vessel revascularisation and (B) myocardial infarction after primary percutaneous coronary intervention (PPCI) by length of stay. This figure is only reproduced in colour in the online version.


This is the first large observational study exploring the safety and feasibility of discharge from hospital 2 days after PPCI for STEMI. The results show that low-risk patients can be easily identified with simple clinical criteria. These low-risk patients can be discharged very early at 2 days in a manner that is both feasible and safe. Our data suggests that up to 40% of STEMI patients may be suitable for very early discharge at 2 days. However, we would like to emphasise that this study does not attempt to investigate the direct effect of length of hospital stay on outcome in a causative manner. Rather, this study presents data regarding the feasibility and safety of an early discharge policy.

Current guidelines recommend that length of stay should be determined by final infarct size, the presence of comorbidities and patient demographics,2 patients who are asymptomatic with small infarcts being suitable for early discharge, particularly after successful and uncomplicated PPCI. Our own criteria for very early discharge reflect these guideline recommendations: only patients who had undergone successful PPCI and were rhythmically and haemodynamically stable with well preserved left ventricular function being candidates for discharge after 2 days. However, while previous studies have confirmed that such patients are at low risk of recurrent cardiovascular events,5 ,8 ,11–,14 this is the first study that confirms the feasibility and safety of discharging low-risk patients from hospital after only 2 days.

The safety of very early discharge in this low-risk group was reflected in the survival analysis, which showed that cumulative MACE rates during follow-up were consistently lower for patients discharged after 2 days compared with patients who were retained longer in hospital because they did not fulfil the early discharge criteria. Although other criteria could be used to stratify patients to early discharge, such as baseline renal function, this was not included in our criteria which focused on clinical information available at the time of the procedure. Interestingly, other studies have demonstrated that length of stay after acute MI may be partly related to non-medical factors. This is manifest in the wide variations between countries, in time to discharge. For example, patients are discharged after 2–4 days in the USA, 3–6 days in Canada, but remain as inpatients for as long as 6–11 days in Germany.15 This wide variation in discharge policy suggests the need for more outcome data on the safety and feasibility of early discharge as presented in this manuscript.

Among patients meeting criteria for 2-day discharge, but retained an extra 24 h, a trend towards increased MACE was observed, but there were no MACE events between days 2 and 3, and no significant increase in MACE out to 30 days, speculating no clear advantage to the extra day inpatient stay. As expected, patients retained >3 days had substantially higher inhospital MACE rates justifying the more cautious approach to discharge. Meanwhile, among those patients appropriate for 2-day discharge, the event rate between days 1 and 2 was very low, suggesting that opportunities for even earlier discharge could be investigated in the future.

The results of this study demonstrate that more than 40% of all patients undergoing PPCI can be safely discharged after only 2 days. This has important implications for healthcare costs which, as previous investigators have emphasised, are closely related to the duration of hospital stay.8 Reductions in costs for a policy of discharge that is applicable to over 40% of all patients with STEMI, coupled with low event rates during follow-up, is likely to be highly cost-effective. Implications for service provision are also potentially favourable in settings where bed availability may be a limiting factor in meeting the volume requirements of heart attack centres applying invasive management strategies across the spectrum of acute coronary syndromes.

It might be argued that such a short inpatient stay might restrict the opportunities for the initiation of secondary prevention treatments and the delivery of education and lifestyle counselling. It has been suggested that shorter hospital stays may be associated with higher readmission rates at 30 days.15 We found no evidence however, of a care detriment in terms of secondary prevention treatment for patients discharged after 2 days. An essential component of our discharge model was early multidisciplinary outpatient care, which appeared to compensate adequately for education and counselling, such that readmission rates for non-MACE events in the first 30 days following discharge were not increased among patients discharged after 2 days.

Limitations to our study include its observational design, and we cannot be sure that patients selected for discharge after only 2 days might not have fared even better with a more prolonged hospital stay. We were able to analyse only 93% of individuals who had PPCI during the study period due to absence of a complete dataset, although we detected no difference in outcomes in these patients compared with the study group. Although all patients were offered follow-up at our centre, it is likely that some events may have been missed from the analysis by patients presenting to other hospitals, which we are unable to quantify. However, we feel that our endpoint of mortality is based on reliable national statistics, although as it is all-cause mortality, it is possible that outcomes between the groups were affected by other disease processes.


Our data suggest that discharge of low-risk patients 2 days after successful primary PCI is feasible and safe. Over 40% of all patients with STEMI may be suitable for very early discharge based on simple clinical criteria, with important implications for healthcare costs.


Interventional colleagues at the London Chest Heart attack centre including Andrew Archbold, Andrew Deaner, Mark Westwood, Simon Kennon, Raj Amersey, David Wald, John Hogan, Rex Dawson and Duncan Dymond. Cardiac specialist nurses from the heart failure and cardiac rehabilitation teams. The Clinical psychology team. The nursing staff of the cardiac catheterisation laboratory, coronary care unit and Riviere ward at the London Chest hospital. Cardiology and emergency medicine nursing and medical colleagues from hospitals within the NE Thames cardiac and stroke network. Mark Whitbread and crews from the London ambulance service.



  • Contributors DJ designed data collection tools, monitored data collection, wrote and undertook the statistical analysis plan, drafted and revised the paper. KSR/JH/SG/OG/RdP cleaned and analysed the data, assisted with the statistical analysis, drafted and revised the paper. SA/SC/JC/EF/JB/SM/AK/CK/AKJ/MTR/AM were part of the clinical teams that cared for the patients. They also developed the clinical pathways, developed our data collection systems, and critically revised the paper. ADT/EJS assisted in study design, drafted and revised the paper. AW was responsible for the conception and initiation of the project, designed data collection tools, analysed the data, and drafted and revised the paper.

  • Competing interests Prof MTR: Professor of interventional cardiology, Barts and the London and Vice-President Medical Affairs Coronary and Renal denervation since January 2010.

  • Provenance and peer review Not commissioned; externally peer reviewed.