Cardiac care in Model 3 hospitals

The National Review of Specialist Cardiac Services is due to report later this year. This review was established “to achieve optimal patient outcomes at population level with particular emphasis on safety, quality and sustainability of the services that patients receive by establishing the need for an optimal configuration of a national adult cardiac service”.

In 2018, with this review in mind, I examined aspects of acute cardiovascular care at the 17 Model 3 hospitals in Ireland. I collated latest activity data from HIPE, inpatient mortality data from NOCA and conducted a survey of colleagues. I hope this analysis will help us to plan for safe, effective and sustainable cardiovascular care for those who live in the catchment areas of these hospitals.

Organisation of acute cardiac care in Ireland

Smaller (Model 3) hospitals admit unselected medical patients and have onsite EDs and critical care facilities. From a cardiovascular perspective, they provide cardiac diagnostics and CCU-level care but not onsite coronary intervention. An effective hospital network is essential to ensure that patients who present to Model 3 hospitals can access appropriate coronary intervention in a timely manner. 

National cardiovascular planning in Ireland began with Building Healthier Hearts, published in 1999. It established that “Each hospital admitting patients with acute cardiac problems should have: An appropriately trained physician; a cardiac investigation area with appropriate staff and facilities and echocardiography facilities”. Resourcing followed these plans and 17 new cardiology posts were established and echocardiography services were rolled out to every acute hospital. These appointments and diagnostic services form the bedrock for current acute cardiovascular care in Model 3 hospitals.

Irish hospitals have, since 2013, been organised into groups configured around existing teaching hospitals. The process is protracted and the hospital groups remain on a non-statutory footing. This system may see further change following the recent plan to reorganise health services into six regional bodies. It has been envisaged, however, that strategies such as national models of care in conjunction with shared clinical pathways underpinned by the hospital groups will drive quality in patient care. Additionally, it is hoped the hospital groups can facilitate cross-site consultant appointments. The development of such shared appointments between the larger Model 4 and 3 hospitals is the sole policy response to the unattractiveness of consultant posts at Model 3 hospitals.

Reflecting this thinking in 2010, Changing Cardiovascular Health was published. The authors of this strategy set out “to improve clinical effectiveness and efficiency by way of organisational change’ in the process ‘developing new cardiovascular models of care and the introduction of clinical pathways”. It recommended a network-based approach to ensure access to cardiovascular care. The strategy was published at a time of economic crisis and the authors “envisaged that the development of new emergency models of care and new clinical networks will have to take place within current resources.” 

While this report had a broad scope, the key achievement was the development and roll-out of the Optimal Reperfusion Strategy. This strategy established a model of care for STEMI with PPCI as the preferred approach. A heart failure model of care was also published but has had less impact. Other specialty services, eg. electrophysiology, advanced cardiac imaging and structural heart disease have not been planned nationally. Such services that exist within the public sector are inadequate for population needs and any development that has occurred has been piecemeal.

In terms of hospital organisation, a decision was taken to develop PPCI services at selected Model 4 hospitals. This was essential to get the PPCI programme running on a cost-neutral basis and offered acceptable population coverage. However this decision added significantly to the workload of already stretched cardiologists at these PPCI hospitals and was problematic from the perspective of ensuring equitable geographic distribution of cardiovascular care. The existing Model 4 hospitals are not strategically located with population coverage in mind. There is a particular concentration of Model 4 hospitals in Dublin.  

This arrangement is not optimal for delivering on the promise of care close to home and presents a challenge to deliver tangible clinical benefit for many patients from time-critical care arrangements such as PPCI for myocardial infarction. Additionally, the significant number of model 3 hospitals that are located a considerable distance from Model 4 hospitals limits the practicality of cross-site appointments, the opportunity for effective cross-site consultant team-work and the development of shared clinical pathways. 

Outcomes, efficiency and activity

Approximately 50% of the total caseload in acute cardiology is delivered at Model 3 hospitals. In 2016 this included 4,666 emergency admissions with a principal discharge diagnosis of ischaemic heart disease, 2,803 with heart failure, 3,166 with atrial fibrillation and 9,474 with chest pain. 

There has been a steady rise in admissions to Irish hospitals for evaluation of chest pain. These admission now account for 2% of total admissions, accounting for a greater number of admissions than acute complications of ischaemic heart disease and heart failure combined. Overall in-hospital mortality from acute myocardial infarction in Ireland is falling. An analysis of NOCA data shows the 30-day mortality rates are similar whether patients present to a Model 3 (In hospital 30 day SMR 5.15) or a Model 4 (SMR 5.96) hospital. Equally, outcomes in acute decompensated heart failure are similar when compared between the two hospital groups. 

An analysis of HIPE data shows that the average length-of-stay in Model 3 and Model 4 hospitals is comparable. The length of stay for acute ischaemic heart disease and complex chest pain is longer in the Model 3 hospitals, reflecting the need to transfer patients for invasive testing. The difference in average length of stay however is modest, suggesting that inter-hospital transfer of cardiac patients is being achieved without unreasonable delay. The length of stay is shorter by three days for patients with decompensated heart failure and by 0.6 days for patients with atrial fibrillation in Model 3 compared to Model 4 hospitals. 

Results of survey of consultant cardiologists

In April 2018 I surveyed 17 Model 3 hospitals. Twenty-three consultant cardiologists work predominantly at a Model 3 hospital, 16 of whom participate in a general internal medicine roster. One Model 3 hospital does not have a single consultant cardiologist, while 11 hospitals have just one. Female cardiologists account for 13% of all cardiologists at Model 3 hospitals. 

There is just one SpR currently occupying a post approved for high intensity cardiology training at a Model 3 hospital. Four consultants are due to retire by 2023 and there are insufficient SpRs currently pursuing dual training in cardiology/GIM to meet this need (two of 47 trainees).

There was also evidence of unmet clinical need; in seven hospitals (n = 15 responded) patients presenting with common cardiology presentations generally have their in-hospital care delivered by the admitting team without direct clinical input (ie. take- over care or in-patient consultation) from a cardiologist. Cardiac CT was available in five of 17 hospitals.

A geographic pattern emerged in cardiology staffing and aspects of acute care. I used distance from a PPCI centre, as set out in the Herity report, as a measure of remoteness for Model 3 hospitals. Nine Model 3 hospitals are located more than 60 minutes average blue-light travel time from a PPCI centre. These hospitals have similar activity levels to the remaining eight Model 3 hospitals and account for approximately 25-30% of the national caseload in acute cardiovascular care. These hospitals tend to have a single cardiologist (seven of nine has one cardiologist, one has two cardiologists and one currently has no cardiologist in post). None of these hospitals employ a cardiology SpR. Conversely, of the eight remaining hospitals closer to a PPCI centre, four had more than one cardiologist. In addition, cardiologists in this group of Model 3 hospitals were less likely to participate on a GIM roster (63% vs 100%) and more likely to have sessional commitments in Model 4 hospitals. Patients in these hospitals with acute myocardial infarction were more likely to have their care routinely taken over by cardiology (57% vs 25%).

Conclusions

Roughly half of national acute cardiovascular medicine activity is accessed through Model 3 hospitals. Outcomes and length of stay are comparable to those achieved at Model 4 hospitals. This care is reliant on a small number of cardiologists who typically work without cardiology consultant or SpR colleagues and participate in a general internal medicine (GIM) roster. 

The lack of SpR involvement in cardiology in Model 3 hospitals is a missed opportunity for patients to benefit from the expertise of this group of doctors. Trainees lose out on the opportunity to experience care provision in a sector that delivers a significant proportion of the national cardiology caseload. This arrangement also means perceptions gained from sources other than direct personal experience of working in a Model 3 hospital are informing trainees’ decisions as they consider their consultant careers. 

It is notable that over the lifetime of cardiology SpR training in Ireland (1998-2017), only 13 of the 90 doctors who achieved certification did so with dual accreditation in GIM and cardiology. There are now insufficient SpRs with dual accreditation in training to address upcoming consultant retirements.

A significant proportion of acute cardiovascular care is delivered by physicians with a specialty interest other than cardiology. While many such physicians may have considerable skill in cardiology, this arrangement does not align with national models of care. Additionally, we may see a gradual reduction in the level of cardiovascular competence and confidence of GIM physicians in practice. Training for medical SpRs in GIM provides limited exposure to acute cardiovascular medicine.

An argument can be made for cardiac CT angiography to be rolled out, as ECHO was in the 2000s. Emergency discharges with a diagnosis of chest pain now exceed emergency discharges for heart failure and acute myocardial infarction combined. A national chest pain policy encompassing cardiac imaging offers improved accuracy of diagnosis, greater care close to home and potentially reduced and more appropriate demand for urgent invasive cardiac angiography. 

Sustaining cardiologist-delivered cardiac care into the future in Model 3 hospitals requires action to support the appointment and retention of cardiologists and senior trainees at these hospitals. A policy of shared consultant appointments can help with recruitment, born out of the experience of Model 3 hospitals within commuting distance of large hospitals. It facilitates cardiologists to practise general cardiology and pursue a subspecialty, a work balance that mirrors that in Model 4 hospitals. The effectiveness of this policy of shared appointments is undermined, however, by the geographical distribution of Model 4 hospitals. A significant number of Model 3 hospitals are situated distant from such hospitals and have yet to benefit from shared appointments. 

To sustain acute cardiology care in Model 3 hospitals we must address the unattractiveness of consultant posts and facilitate a richer, more attractive training experience for SpRs. The review of cardiology services aims to achieve an optional configuration of service to deliver optimal patient outcomes. To date, reconfiguration in Ireland has focused on centralising services and opportunities for specialty practice and training at larger hospitals, which were not strategically planned with population coverage in mind. 

There is an opportunity now as we plan specialty services in critically underdeveloped areas of cardiology such as advanced cardiac imaging, electrophysiology and clinical pharmacology, to plan with national needs in mind. 

This means developing these services at a small number of hospitals outside major urban centres. In this way, the pull effect of specialty cardiology practice can be used to attract cardiologists and SpRs to regions where the continuance of general cardiology care is threatened.