Profile and management of acute coronary syndromes at primary- and secondary-level healthcare facilities in Cape Town

Summary

Background

Little is known about the clinical profile and management of patients with acute coronary syndromes (ACS) in the South African public sector.

Methods

We conducted a retrospective study of patients presenting with ACS to a secondary-level healthcare facility in Cape Town during a one-year period to study the clinical profile and management of these patients.

Results

Among the 214 patients in this cohort, 48 (27.5%) had ST-segment elevation myocardial infarction (STEMI), 43 (24.7%) had non-ST-segment elevation myocardial infarction and 83 (47.7%) unstable angina pectoris. We identified high rates of >12-hour delays in first medical contact after symptom onset (46%) and inaccurate ECG diagnosis of STEMI (29.2%), which were associated with low rates of thrombolysis (39.6%). High rates of non-adherence and ACS recurrence were also observed.

Conclusions

To address the local challenges in ACS management highlighted in this study, we propose the development of a regional referral network prioritising access to expedited care and primary reperfusion interventions in ACS.

Cardiovascular disease (CVD) is a global health issue and the leading cause of mortality worldwide.^1,2 In sub-Saharan Africa, CVD was historically caused by non-ischaemic pathologies, such as rheumatic heart disease and cardiomyopathy.³ However, due to factors such as the rapid urbanisation of less affluent communities, the spectrum of CVD is changing and increasingly appears to resemble that of industrialised countries. In this regard, developing countries such as South Africa are faced with an increasing incidence of ischaemic heart disease (IHD).^3-8 In the Western Cape, IHD has consistently contributed to high rates of cardiovascular mortality since 2000.⁹ It is predicted that, in sub-Saharan Africa, coronary artery disease will continue to escalate within the next decade,⁶ and that CVD will ultimately supersede human immunodeficiency virus/acquired immunodeficiency syndrome as the leading cause of morbidity and mortality.^5,8,10

Rapid urbanisation is associated with a change in lifestyle, leading to a sharp rise in risk factors associated with cardiac disease in sub-Saharan Africa.¹¹ In South Africa, eight modifiable risk factors contribute to 90% of the population’s attributable risk of myocardial infarction.⁵ These risk factors include hypertension, diabetes mellitus, tobacco use, high lipoprotein ApoB/ApoA ratio, abdominal obesity, unhealthy diet, increased psychosocial stress and physical inactivity. Of these, obesity was found to be the most prevalent risk factor, affecting one-third of men and half of women in South Africa.³ A significant proportion of individuals with obesity (particularly women) also had concomitant risk factors associated with the metabolic syndrome, including hypertension, diabetes and elevated serum cholesterol levels.^3,8,10

Where primary prevention fails, morbidity and mortality of patients with acute coronary syndromes (ACS) can be reduced through early implementation of guideline-directed therapy, including medical and invasive revascularisation strategies.^9,12-14 However, in the larger South African context, multiple challenges exist that hinder the enforcement of these guidelines. These include, among others, a lack of patient awareness and insufficient resource allocation to healthcare, with resultant delays in initiation of treatment.^4,9,15-19

Meel et al. reported that only 16% of patients in South Africa made use of ambulance services. Furthermore, the prolonged time between symptom onset and the first call for medical assistance inevitably resulted in delayed first medical contact (FMC). This is a major concern, as 32.6 million (63%) of the total population live within 120 minutes of centres equipped to perform percutaneous coronary intervention (PCI), but are unable to access these facilities.^15,17,20 Indeed, 48 (77%) of these PCI facilities are privately owned and are not accessible to the majority of the population.^17,19,20 Stassen et al. reports that only 14 state-owned PCI facilities are available to those without medical aid (79.9%).^18-20

As state-owned PCI centres have limited capacity, the majority of South African patients with ACS are treated at primary- and secondary-level healthcare facilities in the public sector, where PCI is not available. Administrative delays at healthcare facilities and initial triaging errors further delay the initiation of appropriate medical care. Furthermore, the immediate medical management at non-PCI centres is primarily initiated by junior doctors without specialist or expert input.²¹ On all accounts, these reasons for delays in treatment were more evident and prolonged compared to that reported in the international literature.¹⁵

The majority of data on ACS in South Africa is anecdotal, highlighting the need for further research in this area.^3,6 Additionally, international guidelines are primarily established for the developed world, where resource allocation is less constrained than in developing countries.^6,11 The paucity of local data impedes our ability to adequately respond to the best window of opportunity for concerted action against the emerging epidemic and public health burden of IHD.^3,6,11 As a result, several authors have called for further investigation into ACS in South Africa.^{1,3,6,11,18,22} The rationale for this study was, therefore, to describe the profile, clinical presentation and management of patients with ACS treated at primaryand secondary-level healthcare facilities in Cape Town, South Africa, in the hope of improving outcomes of these patients.

Methods

We conducted a retrospective folder review of all patients presenting to the emergency unit at New Somerset Hospital (secondary-level healthcare facility) with a suspected diagnosis of ACS between 1 January and 31 December 2016. Repeat presentations for suspected ACS were also reviewed.

The study was formally approved by the Human Research Ethics Committee (HREC) of the University of Cape Town, South Africa (HREC ref no 312/2017) and complied with the Declaration of Helsinki.

Potential patients were identified from the registers in the emergency unit at New Somerset Hospital. Patient folders were subsequently retrieved and reviewed to identify those who had a confirmed diagnosis of ACS.

• ST-segment elevation myocardial infarction (STEMI): patients with acute ischaemic chest pain and persistent ST-segment elevation for > 20 minutes. ECG criteria included:^{12 ,23}

  • – new ST-segment elevation at the J point in at least two contiguous leads ≥ 0.1 mV in all leads other than V2–V3; where men ≥ 40 years V2–V3 ≥ 0.2 mV; men < 40 years V2–V3 ≥ 0.25 mV; and women V2–V3 ≥ 0.15 mV

  • – presumed new left bundle branch block: defined as QRS duration ≥ 120 ms with small initial deflection followed by deep S wave with nadir within 70 ms of onset of QRS complex in V1 or V2; broad, notched, or slurred R waves in leads I, aVL, V5 and V6; absent q waves in leads I, V5, and V6.

• Non-ST-segment elevation myocardial infarction (NSTEMI): patients with acute ischaemic chest pain without persistent ST-segment elevation on ECG, but with myocyte necrosis present as proved by elevated cardiac biomarker (e.g. troponin T, troponin I, CK-MB). ECG criteria could include any of the following:^13,23

  • – transient ST-segment elevation (< 20 minutes)

  • – transient or persistent ST-segment depression

  • – T-wave inversion – flat T waves or pseudo-normalisation of T waves

  • – normal ECG.

• Unstable angina pectoris (UAP): patients with a history consistent with chest pain typical of angina at rest or minimal exertion, but with a negative cardiac biomarker result (myocardial ischaemia in the absence of myocyte necrosis).^13,23Data collection included patients’ demographic data (age, gender), co-morbidities and chronic treatment prior to the index presentation, clinical presentation (including symptoms and time to FMC), troponin values and the ECG diagnosis made by the managing physicians at the primary-level healthcare centres and the secondary healthcare facility. Acute treatment [thrombolysis, low-molecular weight heparin (LMWH), antiplatelet therapy] at primary-level healthcare centres (as recorded in the referral letters) and the secondary healthcare facility was documented, as well as discharge medication from the secondary healthcare facility.

Local ACS protocol refers to the pharmacotherapy administered to patients as appropriate for their diagnosis. For NSTEMI this included dual antiplatelet therapy (DAPT), including aspirin and clopidogrel in addition to LMWH (enoxaparin). In patients with STEMI this refers to thrombolysis (streptokinase or alteplase, depending on local availability) DAPT and LMWH, as appropriate. Outcomes included survival and referral for PCI at a tertiary healthcare facility (Groote Schuur Hospital).

All ECGs were de-identified and independently analysed by two physicians with special interest in electrocardiography, who were blinded to the ECG diagnosis made by the managing physicians and the acute treatment that patients received. In the case of discrepant diagnoses, the physicians met, discussed and adjudicated the ECG diagnoses. Patients were classified as having STEMI, NSTEMI and UAP, based on the expert ECG interpretation and troponin values, in line with international guidelines for these diagnoses.^12,13,23

Statistical analysis

Data were collected on Research Electronic Data Capture (REDCap version 9.5.13), a secure electronic database hosted by the University of Cape Town,²⁴ before being exported to Stata (version 14.2, StataCorp, College Station, TX, USA) for statistical analysis. Descriptive statistics were used to summarise data. Continuous variables are summarised as means with standard deviations (SD) for parametric data or median with interquartile range (IQR) for non-parametric data. Categorical variables are expressed as frequencies and percentages. Categorical variables were compared with chi-squared, or Fisher’s exact test, where appropriate. Continuous variables were compared using either the Student’s t-test (parametric data) or Wilcoxon ranksum test (non-parametric data). Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated using univariable and multivariable regression analyses to measure the association between risk factors and outcomes. A p-value < 0.05 was interpreted as statistically significant.

Results

As demonstrated in the study flow (Fig. 1), 214 patients presented to New Somerset Hospital between 1 January and 31 December 2016 with a suspected diagnosis of ACS. Of these, 174 fulfilled the diagnostic criteria for ACS. Included in the analysis are 48 (27.5%) patients with STEMI, 43 (24.7%) with NSTEMI and 83 (47.7%) with UAP.

Fig. 1.

Study flow.

As shown in Table 1, the study population had a male preponderance (59.2%) and a median age of 59.5 (IQR 50–68) years. Patients with STEMI (55 years) were younger than those presenting with NSTEMI (63 years) and UAP (60 years, p = 0.022). The majority of the STEMI group were men (77.1%), while NSTEMI and UAP were equally preponderant among both genders.

Table 1. Baseline characteristics (including demographics, co-morbidities and medication, point of first medical contact and outcome) as classified by diagnosis.

Characteristics	Total (n = 174)	STEMI (n = 48)	NSTEMI (n=43)	UAP (n=83)	p-value
Demographics
Age, median (IQR)	59.5 (50-68)	55.0 (47-64)	63.0 (53-72)	60.0 (50-68)	0.028
Male gender, n (%)	103 (59.2)	37 (77.1)	21 (48.8)	45 (54.2)	0.010
Co-morbidities, n (%)
Hypertension	118 (67.8)	28 (58.3)	29 (67.4)	61 (73.5)	0.20
Diabetes mellitus	59 (33.9)	13 (27.1)	16 (37.2)	30 (36.1)	0.50
Dyslipidaemia	50 (28.7)	13 (27.1)	8 (18.6)	29 (34.9)	0.15
Smoking history	84 (48.3)	34 (70.8)	15 (34.9)	35 (42.2)	< 0.001
Previous myocar- dial infarction	53 (30.5)	13 (27.1)	13 (30.2)	27 (32.5)	0.81
Retroviral disease	7 (4.0)	2 (4.2)	3 (7.0)	2 (2.4)	0.46
Family history of ACS	12 (6.9)	2 (4.2)	5 (11.6)	5 (6.0)	0.34
Respiratory (COPD/PTB)	14 (8.0)	3 (6.2)	2 (4.7)	9 (10.8)	0.42
Number of co-morbidities, median (IQR)	2 (1-3)	2 (1-3)	2 (1-3)	2 (1-3)	0.47
Medication, n (%)
On treatment	101 (58.0)	20 (41.7)	25 (58.1)	56 (67.5)	0.016
Defaulted treat- ment	73 (43.7)	28 (62.5)	18 (41.9)	27 (33.7)	0.006
Point of first medical	contact, n (%)
Primary level	48 (27.6)	26 (54.2)	8 (18.6)	14 (26.9)	<0.001
Secondary level	126 (72.4)	22 (45.8)	35 (81.4)	69 (83.1)	< 0.001
Outcome, n (%)
Thrombolysis	20 (11.49)	19 (39.58)	0	1 (1.20)	0.000
Local ACS proto- col	104 (59.77)	26 (54.17)	34 (79.07)	44 (53.01)	0.012
Survival	171 (98.27)	46 (95.83)	43 (100)	82 (98.79)	0.026
Demised	3 (1.7)	2 (4.2)	0 (0.0)	1 (1.2)	0.026
Referred for PCI	15 (7.5)	5 (10.4)	10 (23.3)	0	< 0.001

NSTEMI, non-ST-segment elevation myocardial infarction; STEMI, ST-segment elevation myocardial infarction; UAP, unstable angina pectoris; ACS, acute coronary syndrome; COPD, chronic obstructive pulmonary disease; PTB, pulmonary tuberculosis; PCI, percutaneous intervention.

A high prevalence of traditional risk factors for coronary artery disease was observed in the overall study cohort (systemic hypertension 67.8%, smoking 48.3%, diabetes mellitus 33.9%, dyslipidaemia 28.7%). Of these, most patients (64.2%) had at least two risk factors for coronary artery disease. Patients with STEMI were more likely to have a smoking history than those with NSTEMI or UAP (70.8 vs 34.9 vs 42.2%, p < 0.001). Almost a third of all patients (30.5%) had a prior diagnosis of coronary artery disease.

Two-thirds of STEMI patients (62.5%) reported non-adherence to their chronic medication prior to presentation. This was significantly higher compared to the other groups (62.5 vs 41.9 vs 33.7%, respectively, p = 0.006).

As expected, most patients (93.1%) presented with chest pain. A quarter of patients (28.7%) reported dyspnoea at the time of primary assessment. As depicted in Fig. 2, patients had multiple overlapping symptoms. There was no significant difference in clinical presentation between the three cohorts. Fig. 3A depicts the time from onset of symptoms to FMC. Across the entire cohort, almost half (46%) of all patients presented more than 12 hours after symptom onset. This was consistent among all three groups.

Fig. 2.

Clinical presentation of (A) all patients with ACS, (B) STEMI, (C) NSTEMI and (D) UAP patients.

Fig. 3.

Time between symptom onset and FMC of (A) all patients with ACS, and (B) patients with STEMI and the proportion of patients receiving thrombolytic therapy over time.

Fig. 3B illustrates the cumulative proportion of STEMI patients that received thrombolysis based on time between symptom onset and FMC. Only 19 (39.6%) STEMI patients received thrombolysis, and 26 (54.1%) received dual antiplatelet therapy and LMWH in hospital. As expected, delayed presentation (> 12 hours) to an appropriate healthcare facility was associated with not receiving thrombolysis (OR 0.07, 95% CI 0.01–0.37).

ECG diagnostic accuracy for STEMI was 70.8%. Anterior STEMI (10/48, 20.8%) was the most frequently missed diagnosis. Others included inferior STEMI (3/48, 6.3%) and presumed new left bundle branch block (1/48, 2.1%). None of the patients with a missed STEMI diagnosis received thrombolysis. Furthermore, one patient with UAP, originally incorrectly diagnosed as STEMI, received thrombolysis. The majority of patients presenting with NSTEMI (79.1%) were treated acutely with antiplatelet therapy and LMWH.

Table 2 elaborates on the discharge treatment of all patients with ACS. Overall, 72% received a statin, 66.1% received aspirin, 59% a beta-blocker and 61% either an angiotensin converting enzyme inhibitor or angiotensin receptor blocker. As expected, clopidogrel was more frequently prescribed to patients with STEMI and NSTEMI than those with UAP (72.9 and 55.8%, respectively, vs 8.4%, p < 0.001). Similarly, those with STEMI and NSTEMI received beta-blockers more commonly than patients with UAP (79.2 and 60.5%, respectively, vs 47%, p = 0.001).

Table 2. Acute management and discharge medication of patients with ACS.

	medication	of patients	with ACS
Medication and management	Total (n=174)	STEMI (n=48)	NSTEMI (n=43)	UAP (n=83)	p-value
Discharge medication,	n (%)
Statin	127 (72.99)	41 (85.4)	34 (79.1)	52 (62.65)	0.011
Aspirin	115 (66.09)	38 (79.2)	33 (76.7)	44 (53.01)	0.002
Clopidogrel	66 (37.93)	35 (72.9)	24 (55.8)	7 (8.43)	0.000
Warfarin	9 (5.17)	4 (8.3)	2 (4.7)	3 (3.61)	0.494
Beta-blocker	103 (59.20)	38 (79.2)	26 (60.5)	39 (46.99)	0.001
Calcium channel	42 (24.14)	7 (14.6)	11 (25.6)	24 (28.92)	0.176
blocker
ACEI/ARB	107 (61.49)	34 (70.8)	29 (67.4)	44 (53.01)	0.085
Nitrate	64 (36.78)	20 (41.7)	16 (37.2)	28 (33.73)	0.661
Diuretic	60 (34.48)	15 (31.2)	15 5 34.9)	30 36.14)	0.849
Acute management,	n (%)
Thrombolysis	20 (11.49)	19 (39.58)	0	1 (1.20)	0.000
Local ACS protocol	104 (59.77)	26 (54.17)	34 (79.07)	44 (53.01)	0.012

ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; ACS, acute coronary syndrome.

Fifteen patients (8.6%) with ACS were referred for urgent PCI at a tertiary-level healthcare facility. These included 10 (23.26%) patients with NSTEMI (of whom seven were referred for post-infarct angina, one with ventricular tachycardia, one with pulmonary oedema and one for a NSTEMI after recent coronary artery bypass graft) and five (10.42%) with STEMI (of whom four patients had persistent ST-elevation and ongoing chest pain despite thrombolysis and one patient had pulmonary oedema).

The in-hospital death rate at the secondary healthcare centre was 1.7%. Of those discharged, 47 patients (27.0%) re-presented to the emergency unit with a repeat episode of ACS within the first year after the index admission [9 (5.1%) with STEMI; 11 (6.32%) with NSTEMI; 27 (15.5%) with UAP].

Discussion

We aimed to define the demographic and clinical profile of patients presenting with ACS to primary- and secondary-level facilities within the public healthcare sector of the Western Cape. We evaluated contextual factors that impacted on time to FMC and the subsequent effects thereof on in- and out-patient management.

We found a high prevalence of risk factors traditionally associated with the development of coronary artery disease, and considerable rates of non-adherence to chronic medication. This study also highlights significant delays in time to FMC. Furthermore, our results demonstrate high rates of missed STEMI diagnoses on the ECG. Both factors contributed to lower rates of thrombolysis at both the index secondary-care facility, as well as the primary referral facilities. Although the in-patient mortality rate was low, a significant number of patients were re-admitted with recurrent episodes of ACS within one year of discharge.

When comparing the spectrum of ACS presentations within this cohort to those of other local⁶ and international registries,^25-28 we identified a higher proportion of patients diagnosed with STEMI than those with NSTEMI. This was not anticipated nor in keeping with the contemporary literature. This could be the result of poor control of cardiovascular risk factors, as demonstrated by high non-adherence rates in this population.

This study shows that the demographic profile of patients with ACS treated in public healthcare facilities in the Western Cape was different to studies performed in the developed world. Patients enrolled in our study were younger than those from the Euro- Heart Survey of Acute Coronary Syndromes (EHS-ACS-II) and GRACE registry,^26-28 but were of a similar age to those included in the ACCESS registry, which predominantly represented patients treated in the South African private sector.⁶ In keeping with the findings of local and international registries, our study demonstrated a male preponderance in the STEMI group.^6,25-28

Patients with STEMI were found to be younger than those with NSTEMI, which correlates with the findings of the ACCESS registry locally⁶ and EHS-ACS-II and GRACE internationally.^25-28 In our cohort the most common presenting complaint was chest pain, which is in keeping with the findings of a local study performed by Geyser et al., which showed high rates of admission in patients presenting with chest pain secondary to cardiovascular causes.²⁹

There was a high prevalence of co-morbidities associated with coronary artery disease throughout the cohort, without any identifiable differences between the STEMI, NSTEMI and UAP groups, which relates to the findings of local and international registries.^6,25-28,30 Despite high rates of previously diagnosed risk factors, a large proportion of patients in this cohort, particularly those with STEMI, reported non-adherence to their chronic medication prior to admission for ACS.

In our study, patients with STEMI were more likely to have a smoking history than those with NSTEMI or UAP. This correlates with the findings of Kenelly, who studied young patients with STEMI in Cape Town, and found that 85% had smoked cigarettes within the last year.³¹

As a matter of great concern, nearly half of all patients were able to access appropriate treatment for their ACS only more than 12 hours after onset of their symptoms. The reasons for these delays in FMC are multi-factorial, and could possibly be attributed to pre-hospital factors (such as poor patient education and overburdened emergency medical services), and hospital factors (such as prolonged triage times, resource limitations and insufficient ECG diagnostic proficiency).^4,9,15-19 As expected, those with delayed presentation were unlikely to receive thrombolytic therapy at the non-PCI centres in this study. This is in keeping with the findings of Chetty et al., who observed that late presentation with STEMI was the most common contraindication to thrombolysis.³²

The establishment of structured rapid referral networks for STEMI management has been identified as a key factor in improving rates of reperfusion and mitigating mortality in patients with STEMI.^12-14 Despite the challenges that are imposed by resource constraints in developing countries, several countries have been successful in implementing such systems to improve outcomes in STEMI.^33-35 Implementation of the Latin American Telemedicine Infarct Network was associated with improved rates of reperfusion and reduced mortality rates from STEMI in Brazil.³³ Similarly, implementation of the Tamil Nadu-ST-Segment Elevation Myocardial Infarction programme was associated with improvements in one-year survival after STEMI in India.³⁴

In Cape Town, ACS is treated according to a hub–spoke model, where patients present to their nearest primary or secondary healthcare hospitals. In the case of STEMI, thrombolytic therapy can be instituted in these healthcare facilities and patients are referred to tertiary centres for failed thrombolysis or whenever haemodynamically or electrically unstable.

As the ECG is central in the diagnosis of STEMI, it is worrying that almost a third of STEMIs were missed by the managing physicians, and as a result, none of these patients received thrombolytic therapy. However, ECG proficiency in this study was better than that reported by Mabuza et al. locally³⁶ and in a recent meta-analysis including 78 international studies.³⁷ This highlights the need for continuous ECG education globally, particularly in the context of ACS.

The low in-patient mortality rate in this study (1.7%) is in line with the ACCESS (2.6%) and EHS-ACS-II registries (4.0%).^6,28 However, we did identify a concerning rate of recurrence of ACS within the first year of the index diagnosis. This may be explained by poor access to healthcare services for appropriate follow up and interventions, such as PCI, as well as high rates of non-adherence to prescribed guideline-directed therapy. High rates of recurrence will continue to place additional strain on the already overburdened public healthcare services.

Limitations

There were several limitations of this study. The retrospective nature of this study limited access to detailed and complete clinical data, such as the timing of STEMI diagnosis and implemented reperfusion interventions, as well as associated in-hospital outcomes. As a result, we were unable to assess the effect of variables such as time from FMC to ECG diagnosis, and door-to-reperfusion time. We acknowledge that these variables would serve as more reliable indicators of quality of care in this context and we therefore recommend that a prospective database of STEMI referral and management data be implemented to facilitate further assessment and optimisation of STEMI management. Despite the inclusion of all patients who presented to the emergency unit with ACS, the sample size remained small. As a result, sub-analyses in this study should be interpreted with caution and the generalisability of the study findings is limited.

Conclusions

To the best of our knowledge, this study is among the first to report the profile of patients with ACS treated in the South African public healthcare sector. This study highlights several challenges faced in the management of ACS in this context, such as high prevalence of cardiovascular risk factors and rates of non-adherence to chronic medication, significantly delayed FMC, insufficient ECG diagnostic proficiency, the absence of a dedicated ACS referral network, as well as high rates of recurrence of ACS within one year of the index presentation. Thrombolysis is often not offered to those with STEMI, due to delayed presentation or missed ECG diagnosis.

To address the challenges identified in our study, we propose the establishment of an improved regional referral network for ACS in our public health system. This network should prioritise early identification of patients requiring reperfusion and facilitate rapid referral from primary- and secondarylevel healthcare facilities to PCI-capable tertiary centres. The goal of this network would be to improve patient access to appropriate primary reperfusion interventions. In addition, we recommend improved patient education and optimisation of medical education directed at improving ECG diagnostic ability and management of ACS.