Study design and clinical setting
This is a retrospective cohort study done at Tikur Anbessa Specialized Hospital (TASH) in Addis Ababa, Ethiopia to assess the incremental effect of DM on incident HF in patients with IHD.TASH is an institution where specialized clinical services are rendered to the whole nation. TASH offers general internal medicine and cardiology services together with other services. The adult cardiology clinic at TASH provides specialized cardiac care to patients referred from hospitals and health institutions from all regions of the country.
The study population was recruited until November 30, 2015, and each case was followed up for 24 months. The incidence of new cases with HF was recorded among patients with and without DM. Patients with IHD, DM or HF were identified based on the treating Physician’s final diagnosis. Those with IHD and diabetes were taken as exposed and those with IHD but no diabetes were taken as unexposed.
Study population
The target population consists of adult patients (age of 18 years and above) with IHD. We followed our cohort of IHD patients to measure their outcomes. The primary outcome of interest was incident HF. Study participants were considered exposed if the duration of DM before a diagnosis of HF was three or more months so that the diagnosis of DM is established. Patients diagnosed with IHD and with no history of HF at baseline were included. Patients with a history of HF or who were using furosemide at baseline were excluded. There were a total of 306 patients with IHD who fulfilled the inclusion criteria, 96 of them diabetic and 210 non-diabetic. The calculated power of the study to detect difference was 93.7%.
Operational Definitions
IHD was defined based on information from patient records where the attending physician (Cardiologist) made the diagnosis. Diabetes Mellitus was defined as a confirmed history of this diagnosis or the prescription of glucose-lowering drugs based on information from patient records. HF was defined as the presence of Physician documented diagnosis of HF and/or prescription of a loop diuretic. Hypertension was defined based on physician-registered diagnosis on the medical record. Patients were diagnosed to have hypertension if systolic blood pressure ≥ 140 mm Hg, diastolic blood pressure ≥ 90 mm Hg, or taking anti-hypertensive medications.
Smoking status was classified as current smokers or non-smokers based on reports in the history record or risk factors of IHD listed on the medical record. Anemia diagnosis was made if the hemoglobin level was <12 g/dL. A nephropathy diagnosis was made if there was a trace or above proteinuria.
Data Collection
Data were collected by trained nurses from the medical records of the study participants using a pre-tested structured questionnaire. A total of 306 medical records of patients newly diagnosed with IHD until November 30, 2015, were identified to determine the incidence of HF in IHD with or without DM. Subjects were followed from the date of enrolment until they were diagnosed to have HF or November 30, 2017 (study exit date). Individuals with established HF before the date of enrolment were excluded. Those who died or lost to follow up before the study exit date were censored during analysis.
IHD patients diagnosed with DM anytime during the study period were labeled as exposed and those with no DM, until less than 3 months before the study exit were labeled as unexposed.
The questionnaire contained demographic variables (age, sex), smoking, diagnosis and date of diagnosis of IHD, diagnosis, and date of diagnosis of DM, diagnosis, and date of diagnosis of HF, diagnosis, and functional classification of HF based on New York Heart Association (NYHA) classification, blood pressure level, full blood count, fasting blood glucose, lipid profile, urea, and plasma creatinine, and electrolytes. Data on co-morbidities (anemia, dyslipidemia, HTN, and chronic kidney disease), medications, and hospital admission were collected. Information from resting 12-lead electrocardiography and transthoracic echocardiography were collected. In the case of participants with missing information, the available data was taken for analysis.
Data Management and analysis
Data were entered into an electronic database using SPSS version 23, and then data analysis was done. The proportion of HF in diabetic IHD patients and non-diabetic IHD was calculated. Comparisons were done within age groups. Categorical variables were reported as frequencies (%) and continuous variables reported using mean (standard deviation) or median (inter-quartile range). Pearson’s Chi-square test for independence (with Yates Continuity Correction) was used to compare the difference in proportions of HF between diabetic and non-diabetic IHD patients.
In patients diagnosed to have HF further analysis was done to compare differences between patients with diabetic IHD and non-diabetic IHD. The independent-sample t-test was used to compare the difference between the two groups in ejection fraction (LVEF), left atrial (LA) size, left ventricular diastolic dimension and demographic variables like age. Multivariable logistic regression analysis was conducted to estimate the effect of DM on HF adjusted for potential confounding variables such as age, sex, hypertension, smoking, dyslipidemia, nephropathy, LVEF, and LA size. The Kaplan-Meier procedure was used to estimate the survival curves and Cox regression analysis was used to calculate the hazard ratios for predictors of HF. In cases of missing data, the ‘Exclude cases pairwise’ option was used during analysis. A 5% significance level was adopted for all tests and all tests were 2-sided.
Ethical Considerations
Ethical clearance was obtained from the Institutional Review Board of the College of Health Sciences, Addis Ababa University. Informed consent was waived by the
IRB due to the retrospective nature of the study with no more than minimal risk. Eligible medical records were evaluated further for inclusion and exclusion criteria.