Prevalence and pattern of echocardiographic abnormalities among patients on haemodialysis at an urban hospital in Central Uganda

Globally, chronic kidney disease (CKD) is a major cause of morbidity and mortality, affecting almost 700 million people, as of 2017, and with a steady increase in mortality between 1990 and 2017 [1]. The prevalence of CKD in sub-Saharan Africa (SSA) is around 14% [2]; it is estimated that SSA will contribute more than 70% of the world’s CKD burden by the year 2030. In Uganda, the prevalence is 15.2% [3]. Cardiovascular disease (CVD) is the most common


INTRODUCTION
stage renal disease (ESRD). Left ventricular hypertrophy (LVH) and diastolic dysfunction are the two cardiac abnormalities most implicated as risk factors for cardiac death in haemodialysis patients [5,6]; echocardiography provides a non-invasive assessment for these and other abnormalities of cardiac structure and function [7].
A study from Cameroon among patients on haemodialysis found the prevalence of cardiac abnormalities to be 80%, with LVH being the most prevalent at 60% [8]. In Uganda, a previous study in Mulago Hospital found LVH in 54.4% of patients with CKD [9]; however, this study excluded patients on haemodialysis.
Over the past ten years, there have been great strides in the establishment of both public and private haemodialysis centres in Uganda but there remains a paucity of data on cardiovascular complications in these patients. We therefore sought to establish the prevalence of LVH, diastolic dysfunction and other echocardiographic abnormalities among patients on chronic haemodialysis in Mulago National Referral Hospital.

METHODs
This single-centre, cross-sectional study was carried out at Mulago National Referral Hospital in Kampala, Uganda. This has the only public dialysis unit in Uganda, and provides two subsidised, 4-hour dialysis sessions per week to each patient. In addition to serving the Ugandan population, the unit accommodates patients from the neighbouring eastern Democratic Republic of Congo and the Republic of Southern Sudan. In total, the dialysis unit has about 90-100 patients. We enrolled 80 consecutive patients aged 18 years and above from November 2017 to March 2018. Patients were eligible for inclusion if they had been on chronic haemodialysis for at least 3 months.
Baseline characteristics were collected by reviewing the patients' charts and conducting patient interviews using a pretested questionnaire. Data included age, sex, comorbidities, date of initiation of dialysis, distance travelled to access dialysis and drug treatment history for hypertension.

blood pressure measurements and laboratory tests
Blood pressure (BP) was measured before dialysis while the patient was lying down. Measurements were taken twice in each arm (except for patients with an arteriovenous fistula access on the arm) using a Welch Allyn DS66 Trigger Aneroid machine (Welch Allyn, USA). The higher reading was recorded. Pre-dialysis blood samples were drawn for white cell count, haemoglobin, serum albumin, serum creatinine, serum calcium and phosphate, and a fasting lipid profile. Blood counts were performed using a Sysmex XN 1000 CBC machine (Sysmex, USA) whereas all chemistry tests were conducted on serum using a Roche cobas ® 6000 analyser (Roche Diagnostics, USA).
Results for HIV, hepatitis B and C serology were abstracted from the patients' charts. Creatinine was measured using the Jaffe method traceable to an isotope dilution mass spectrometry method [10] and estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI formula without use of the coefficient for African Americans [11]. Serum albumin was measured using capillary zone electrophoresis.
Uncontrolled hypertension was defined by systolic BP >140 mmHg or diastolic BP >90 mmHg in a patient on anti-hypertensive medication [12]. Anaemia was defined as a haemoglobin concentration < 11 g/dL. Dyslipidaemia was defined per ACC/AHA guidelines as any of: i) HDL <35 mg/dL, ii) LDL >100 mg/dL or iii) total cholesterol >190 mg/dL [13]. Hypoalbuminaemia was defined as serum albumin <35 g/L. four-chamber, and parasternal long and short axis) were used to evaluate patients for pericardial effusion, which was graded as mild (<10 mm ), moderate (10-20 mm) and large (>20 mm) as per the European Society of Cardiology guidelines [17]. Heart valve anatomy and function were assessed and these parameters together with those of chamber function were used to grade valvular lesions. echocardiographic abnormalities in patients on haemodialysis in Uganda LVH was defined as interventricular septum and/or left ventricular posterior wall diameter >11 mm [18]. Systolic dysfunction was defined by evidence of an EF <50% [19]. Left ventricular diastolic dysfunction was defined as more than two of the following: i) E/E' >14, ii) TR velocity >2.8 m/s, iii) LA volume index > 34 mL/m 2 and iv) septal/medial e-velocity <7/10 m/s in a patient with an ejection fraction >50%, according to algorithm A of the 2016 ASE/EACVI guidelines [16].

statistical analysis
Data were double entered into EpiData version 3.1 and exported to STATA version 14.0 (STATA Corp, Texas, USA) for analysis. Continuous variables were reported as mean ± standard deviation (SD) for normally distributed data or median with interquartile ranges (IQR) for nonnormally distributed variables. Categorical variables were expressed as frequency (percent). The primary outcome was the prevalence and patterns of echo abnormalities and these are presented as frequency (percentage) or proportions. The secondary outcome, demographic and clinical factors associated with LVH and diastolic dysfunction, was evaluated by logistic regression. Those factors with P <0.2 on bivariable analysis were included in the multivariable regression model. Odds ratios and 95% confidence intervals were used to quantify associations and results were considered statistically significant when the P value was <0.05.

Ethical approval
The Makerere University School of Medicine Research and Ethics Committee approved the study (reference number 2017-135).

REsULTs
We screened 92 patients and twelve were excluded for various reasons (Figure 1). A total of 80 patients, aged 19 to 78 years, were enrolled in the study. The patient characteristics are summarized in Table 1. The mean age was 49 years, 66% were male and the median duration on dialysis was 9.5 months. One-third of the patients had to travel >50 km to access haemodialysis. All had hypertension with approximately three-quarters having uncontrolled blood pressure despite taking antihypertensive medications.
Baseline echocardiographic parameters used to define structural abnormalities (LVH, systolic and diastolic dysfunction) were similar in male and female participants ( Table 2).
Seventy-four patients (93%) had at least one cardiac echo abnormality and 24 (30%) had at least three abnormalities (Table 3). We found that 68% had LVH and 64% had left ventricular diastolic dysfunction. Most patients had both LVH and diastolic dysfunction ( Figure 2). Although these abnormalities were more common in older patients, there was still a high prevalence in the group <45 years ( Figure 3).
Thirty-five patients (44%) had mild to moderate valvular lesions (Table 3), including 37% with mitral regurgitation, 26% with both mitral and tricuspid valve regurgitation, 17% with both mitral and aortic valve regurgitation, 17% with echocardiographic abnormalities in patients on haemodialysis in Uganda

DIsCUssION
We have found a high prevalence of echocardiographic abnormalities among Ugandan patients on haemodialysis.
echocardiographic abnormalities in patients on haemodialysis in Uganda   Diastolic dysfunction was the second-most common cardiac abnormality in our study and its prevalence was comparable to what has been reported previously [6,20].
Hickson et al. [20] found the prevalence of diastolic dysfunction ≥grade 2 as high as 78% at the start of haemodialysis. The risk factors for diastolic dysfunction are mostly similar to those for LVH, including hypertension and volume overload [7,24]; however, LVH itself is one of the biggest risk factors for diastolic dysfunction [25] and our study confirms data from previous literature that LVH and diastolic dysfunction often coexist.
LVH and diastolic dysfunction are independently associated with cardiovascular death among patients on haemodialysis [5]. Barberato [9]. This study highlights the fact that dialysis improves uraemic pericarditis among patients with ESRD.
Only 10% of our patients had systolic dysfunction, comparable to the 13% reported by Hickson et al. [20]. As regards valvular lesions, our findings are similar to those of previous studies in terms of prevalence and type of valve lesions. Several studies have reported that these lesions are more frequent in patients with ESRD on haemodialysis and that they occur at a younger age compared to the general population [20]. Aortic valve calcification with stenosis has been found to be common in dialysis patients [27]. This was not the case in our study and may be explained by the fact that ours were young patients with a relatively short period on dialysis.
Our study has some limitations. First, it is a relatively small, single-centre study; however, it was conducted at the largest dialysis centre in Uganda and more than 90% of the available patients were included. Second, given the crosssectional design of the study, the longer-term outcomes and complications of the observed cardiac abnormalities could not be evaluated. Third, echocardiography is not as echocardiographic abnormalities in patients on haemodialysis in Uganda  We also recommend that the government allocate additional resources and provide the standard three haemodialysis sessions per week to contribute to better cardiac and overall outcomes for these patients.

Acknowledgements
We thank the Mulago National Referral Hospital dialysis unit and echo laboratory staff for their assistance in this study. We also acknowledge the financial contribution provided by the Rainer Arnold Senior House Officers Teaching Support (RASHOTS) project.
echocardiographic abnormalities in patients on haemodialysis in Uganda