Skip to main content

Post-surgery survival and associated factors for cardiac patients in Ethiopia: applications of machine learning, semi-parametric and parametric modelling



Living in poverty, especially in low-income countries, are more affected by cardiovascular disease. Unlike the developed countries, it remains a significant cause of preventable heart disease in the Sub-Saharan region, including Ethiopia. According to the Ethiopian Ministry of Health statement, around 40,000 cardiac patients have been waiting for surgery in Ethiopia since September 2020. There is insufficient information about long-term cardiac patients’ post-survival after cardiac surgery in Ethiopia. Therefore, the main objective of the current study was to determine the long-term post-cardiac surgery patients’ survival status in Ethiopia.


All patients attended from 2012 to 2023 throughout the country were included in the current study. The total number of participants was 1520 heart disease patients. The data collection procedure was conducted from February 2022- January 2023. Machine learning algorithms were applied. Gompertz regression was used also for the multivariable analysis report.


From possible machine learning models, random survival forest were preferred. It emphasizes, the most important variable for clinical prediction was SPO2, Age, time to surgery waiting time, and creatinine value and it accounts, 42.55%, 25.17%,11.82%, and 12.19% respectively. From the Gompertz regression, lower saturated oxygen, higher age, lower ejection fraction, short period of cardiac center stays after surgery, prolonged waiting time to surgery, and creating value were statistically significant predictors of death outcome for post-cardiac surgery patients’ survival in Ethiopia.


Some of the risk factors for the death of post-cardiac surgery patients are identified in the current investigation. Particular attention should be given to patients with prolonged waiting times and aged patients. Since there were only two fully active cardiac centers in Ethiopia it is far from an adequate number of centers for more than 120 million population, therefore, the study highly recommended to increase the number of cardiac centers that serve as cardiac surgery in Ethiopia.

Peer Review reports


CVD was responsible for approximately 19.1 million deaths worldwide in 2020. Sub-Saharan Africa had the highest CVD-related mortality rates in 2020. It was estimated that 244.1 million people worldwide would live with ischemic heart disease alone in 2020, not including other heart diseases [1]. In the United States, one person dies from cardiovascular disease every 33 s [2].

Cardiovascular diseases are the number one cause of death [3]. Approximately one-third of CVD patients will require surgical or interventional care at some point [4]. However, about 6 billion people worldwide, primarily in low- and middle-income countries, lack access to cardiac surgery [5]. Congenital heart disease, rheumatic heart disease, and ischemic heart disease account for the lion’s share of the global CVD burden that necessitates surgical intervention [6].

Even though cardiovascular disease is the primary cause of mortality and morbidity, approximately, about 6 billion people on the globe lack access to timely, safe, and affordable cardiac surgery [7]. In low and middle-income countries, access to such services is also disproportionately low [4].

Furthermore, the risk factors for non-communicable diseases are increasing unacceptably rapidly in low- and middle-income countries, including Africa [8,9,10]. Chronic diseases, including CVD, are rising in low- and middle-income countries, posing a public health crisis [11, 12]. Aside from the increased burden of CVDs in LMICs, there is a scarcity of data on CVD risk factors [12].

Africa faces several challenges due to its CVD burden, including a lack of high-quality data, competing priorities, financial constraints, diagnostic, limited skill sets, and management challenges [13]. Africa has over one billion people and contributes significantly to the global CVD burden [14]. In 2013, CVD was responsible for an estimated one million deaths in Sub-Saharan Africa alone, accounting for 5.5% of all global CVD-related fatalities and 11.3% of all deaths in Africa [15]. In 2019, CVD was responsible for over one million deaths in Sub-Saharan Africa alone [16].

Estimating the CVD burden in Sub-Saharan African countries, including Ethiopia, is difficult [17]. According to the World Health Organization, around 30% of Ethiopians died from non-communicable diseases in 2014, with CVD accounting for 9% [18]. According to a systematic review conducted in Ethiopia, the prevalence of CVD ranges from 7.2 to 24% [19].

Due to a lack of workforce, infrastructure, and resources, Ethiopia has limited cardiac procedures, with long wait lists [6]. Ethiopia is a low-income country in East Africa, with approximately 120 million people [20]. Low-income countries are estimated to require 300 to 400 cardiac operations per million yearly population [7]. Despite the high demand for cardiac interventions, Sub-Saharan Africa has a scarcity of such services [21, 22]. There is still a scarcity of literature on cardiac surgery in Ethiopia [6].

According to recent studies, the elderly population over 60 is expected to grow from 962.3 million to 2080.5 million between 2017 and 2050 [23]. The heart-related figure is not known in developing countries [24]. There is little data on heart disease prevalence, especially in developing countries [25,26,27]. In recent years, the Ethiopian population has seen a significant epidemiological shift. The primary causes of illness and death change from infectious to noncommunicable diseases [28]. There is a deficiency in developing and executing novel solutions to lower the risks associated with cardiovascular illnesses because there is a mistaken perception that cardiovascular disease is solely a problem of the prosperous and industrialized world. There is also a tendency to associate heart disease mainly with a sedentary lifestyle and hyper-nutrition [29, 30]. Heart disease is neglected by health economists and experts [31].

The report from Cardiac Center Ethiopia shows that the number of cardiac disease patients increases dramatically from time to time; this is why many patients require cardiac surgery. According to the Ministry of Health statement, around 40,000 cardiac patients have been waiting for surgery in Ethiopia since September 2020. According to the Cardiac Center Ethiopia report, more than 7000 patients seek surgery at Ethiopia’s cardiac center [32].

Limited information about the current burden status of cardiac disease in Ethiopia is available. Therefore, this study provided an up-to-date assessment to evaluate the cardiac burden after surgery. There is no sufficient report on cardiac disease, specifically on the survival status of patients after surgery in Ethiopia. The objective was to determine the main predictors of post cardiac surgery survival in Ethiopia.

Method and participants

Study area and period

Ethiopia is a nation in the horn of Africa. It is bounded to the north by Eritrea and Djibouti, to the northeast by Somaliland, to the east by Somalia, to the south by Kenya, to the west by South Sudan, and the northwest by Sudan. Ethiopia has more than 120 million people, making it the 12th most populated country globally and the second most populous in Africa.

The government has made enormous expenditures in the public health sector, which has resulted in improved health outcomes. Nonetheless, the Ethiopian government commitment or tangible effort to establish fully functional cardiac centers is limited. This study was focused on cardiac surgery patients and those registered for follow-up treatments in Ethiopia. The study was conducted at Cardiac Centre Ethiopia and Elouzeir Cardiac Center, which actively provided cardiac surgery during the study period. A non-governmental organization owns Cardiac Center Ethiopia, Children Heart Fund Ethiopia, and currently provides Ethiopia’s most significant cardiac surgery. The center started to provide surgical management for cardiac patients in 2009. The study was conducted on the patients’ follow-up charts from 2012 to 2023. The following figure shows one of the study areas called cardiac center Ethiopia (Fig. 1).

Fig. 1
figure 1

Cardiac center Ethiopia

Study design

The study was carried out through a retrospective cohort study design. The source of the population was cardiac patients whose follow-ups were between 2012 and 2023 in Ethiopia.

Study population

The study population was heart patients who had cardiac surgery between 2012 and 2023 in Ethiopia. Two cardiac centers in Ethiopia serve as cardiac surgery. All cardiac centers were part of the study. In the study area, 1520 patients had cardiac surgery. All 1520 patients were included in the study.

Data collection technique

The data were collected using structured questionnaires which was developed by the study (see Supplementary file 1) through an online data collection platform (Kobo tool box). BSc nurses and medical doctors participated in the data collection process. Data collection started after ethical clearance was obtained from Bahir Dar University. The permission was obtained from health provider institutions.

Data quality assurance method

The questionnaire was evaluated in randomly selected charts and participants; as a result, the pre-test was conducted to check data quality. Trained health professionals performed the data collection. Data consistency was supervised and reviewed by the investigator every day.

Variable under the study

Outcome variable: The outcome variable for survival analysis (Alive/censored, Death) outcome. The predictor variables: - The predictor variables are summarized as indicated in (Fig. 2).

Fig. 2
figure 2

Predictor variables framework

Data processing, importance variable selection and analysis techniques

After data collection, the collected data exported in to excel. Descriptive statistics characterise the study population in terms of relevant variables. The descriptive result is stated in the form of tables. Test of the association performed using chi-square test of association. Survival mean time comparison was done using t test. Different machine learning algorithms were considered. Feature selection, auto data preparation, and analysis were performed using IBM SPSS Modeler software. Python and Stata version 13 was used also for survival machine learning, semi parametric and parametric regression model comparison, analysis and model adequacy. From all 28 variables, 12 were selected using the machine learning variable selection method. The study used a feature selection procedure, ranking the importance of inputs relative to a specified target. Among all predictor variables, 12 variables were included in the model (Fig. 3).

Fig. 3
figure 3

Important variable extraction output

Machine learning for survival outcome

The most important feature of machine learning algorithms is their unique opportunity to know the surrounding environment from data input, even without assistance [33,34,35,36,37]. As such applying machine learning algorithm is preferable than traditional models [38,39,40]. A machine learning models trained on labelled data sets learn and grow more accurately over time. An algorithm would be trained with the baseline values even using labelled by humans, and the machine would learn to recognise patterns or something else on its own. Today, the most common type is supervised machine learning. The study employed four machine learning algorithms for comparison and generalisation. In the final model heart disease type, rheumatic disease type, NYHA class, age, weight, SPO2, INR, ejection fraction, duration of stay after surgery, hemoglobin level, creatinine value, waiting time for surgery variables were considered. After applying feature selection and machine learning algorithm comparison, we select the most important predictors for time to death prediction. Python jupyter notebook was implemented to perform machine learning survival modelling.

In this survival analysis the study implements different types of machine learning algorithms such as random survival forest (RSF), survival tree, support vector machine, and gradient boosted survival.

Parametric and semi parametric models often have a well-defined mathematical form with interpretable parameters, allowing for a clear understanding of how each variable contributes to the model’s predictions. This interpretability can be crucial in domains were explaining and understanding the underlying relationship between variables is essential. As such the study also implements parametric survival analysis for detail interpretations to deal the impact of each predictor on the outcome.

Semi parametric (cox) survival

Survival analysis investigates and models the time it takes for events to occur. Survival analysis is mostly associated with death-related events, but the scope of survival analysis is much broader. The distribution of survival times is the focus of survival analysis [41]. Let T stand in for survival time. The instantaneous risk of death at time t, assuming survival to that point, can be expressed as follows: \( {h}_{i}\left(t\right)=\text{exp}\left(\alpha +{\beta }_{1}{x}_{i1}+{\beta }_{2}{x}_{i2}+\cdots +{\beta }_{k}{x}_{ik}\right)\)

Parametric survival

Most common Parametric survival analysis; Exponential, Gompertz, Generalized gamma, Weibull, log normal and log logistic were considered for model fit. The study compares the most fitted model for the data.

Model comparison

Parametric and semi parametric model comparison was conducted using AIC and BIC such that the one with smallest AIC and BIC was considered as the best fit. Based on model fit, the model with the smallest AIC is goes to Gompertz over other semiparametric and parametric survival models. As such the model fitted with Gompertz survival analysis (Table 1). Comparing the jagged line with the reference line, we observe that the Gompertz survival model fits the data (Fig. 4).

Table 1 Model comparison
Fig. 4
figure 4

Cox Snell model fit


In this study, 1520 participants were included with an average age of 24 years and an average weight of 42 kg, with average baseline systolic and diastolic blood pressure of 107 and 66, respectively. The mean oxygen saturated value was also 95%. The average value of the post-surgery ejection fraction was 59%. The Average waiting time in the cardiac center after surgery was seven days. T-test used to compare average values. There was a significant mean difference between average censored and average death for age, weigh, SPO2, INR, cardiac center duration after surgery, hemoglobin and creatine value predictors (Table 2).

Table 2 Post-cardiac surgery patients’ descriptive analysis for continuous variables mean comparison through t-test

Of the total 1520 patients, 122 (8%) patients died. From all patients who had cardiac surgery, Female (842), patients from rural area (692), Heart disease type valvular (874), congenital (593), other (53), patients with family history (145), patients with heart failure status(20), patients with smoking habit (4), Cardiac rhythm normal sinus rhythm (49), number of surgery more than one (25), Joint heart case Combination(177). From chi-square test of association, there was a significant association between outcome variable and heart disease type, rheumatic disease status, family history, NYHA class, and heart failure status (see Table 3).

Table 3 Frequency and chi-square test of association analysis of post-cardiac patients in Ethiopia

The log-rank test showed a significant association between survival outcome status (Death, Alive) and heart disease type, rheumatic disease status, and NYHA class (See Table 4).

Table 4 Survival experience comparison of post-cardiac patients in Ethiopia

The mean cross-validation score represents the average performance of the model across multiple cross-validation folds, while the model score is the performance on a specific test set. The mean cross-validation score provides an estimate of how well the model is expected to generalize to new, unseen data. It is computed by averaging the performance scores obtained from each fold of the cross-validation process. From a 10-fold cross validation and model test result, the study observe that the best survival model is random survival forest as compared to other algorithms such as Support Vector Machine, Gradient Boosted and survival tree. In classification tasks, a score of 0.89 could indicate a high level of correct predictions, where the model is accurately classifying 89% of the instances. The mean cross-validation score represents the average performance across all the folds. A score of 0.83 indicates that, on average, the model correctly predicts the outcome or class label for approximately 83% of the instances in the dataset. As such the interpretation is performed through random survival forest (see Table 5).

Table 5 Table average mean cross validation and model score

From random forest survival, the graph highlights the significance of various predictor variables in the selected model. It emphasizes the importance of certain features, namely SPO2, Age, time to surgery waiting time, length of health facility leave date, and creatinine value. These variables have a significant impact on the model’s performance and play a crucial role in predicting the time to event outcome. Specifically, SPO2 contributes approximately 42.55% to the model, Age contributes around 25.17%, time to surgery waiting time contributes approximately 11.82%, length of health facility leave date contributes around 8.27%, and creatinine value contributes approximately 12.19% (Fig. 5).

Fig. 5
figure 5

Random survival forest variable importance

From the Gompertz regression, of saturated oxygen, age, ejection fraction, length of cardiac center leave day, creatine value, and waiting time to surgery were statistically significant predictors at a 95% confidence interval.

For cardiac patients under post-surgery treatment, as age increased by one year, the mortality hazard increased by 69.2%. ((HR = 1.692, 95% CI: 1.306–2.192, p < 0.001). With a single rise of saturated oxygen, the hazard of mortality was reduced by 57.5% (HR = 0.425, 95% CI: 0.359–0.503, P < 0.001). Regarding ejection fraction, for a single increase of ejection fraction, the mortality hazard decreased by 25.3% (HR = 0.747, 95% CI: 0.629–0.887, P = 0.001). Regarding cardiac center duration, for a single day increase of leave day in a cardiac patient, the mortality hazard was decreased by 38.5% (HR = 0.615, 95% CI: 0.497–0.762, P < 0.001). About hemoglobin value, for a single unit increase in hemoglobin level, hazard of mortality was decreased by 13.7% (HR = 0.863, 95% CI: 0.741–1.005, P = 0.057). As a unit increase of creatinine value for cardiac patients, the hazard of mortality increased by 39% (HR = 1.364, 95% CI: 1.165–1.597, P < 0.001). Another most important predictor was waiting time, as a day prolonged time to surgery or waiting time for cardiac surgery, the mortality hazard was increased by 59.2% (HR = 1.592, 95% CI: 1.327–1.911, P < 0.001) given the other conditions constant (Table 6).

Table 6 Gompertz post cardiac patients’ survival multivariable result in Ethiopia


The death prevalence of cardiac patients after surgery was 122(8%). The finding is similar to the previously conducted research, which is more or less similar to that found in most other studies [42,43,44,45]. After cardiac surgery, the significant predictors for heart patients were creatine value, length of post-cardiac surgery stay in the cardiac center, ejection fraction, saturated oxygen (SPO2), time to surgery, and age. The study was designed to identify the responsible predictors of post cardiac surgery mortality.

Age is one of the significant predictor variables for heart disease death, and this result is aligned with other previously conducted studies [46,47,48], consistent with the investigation. It is because when the patients age increases the possibilities of complication and other comorbidity also increases. And the responsiveness of the treatments also decreases when age increases. Physiological aging of the heart as a major causative predictor in the manifestation and onset of cardiac in aging due to increased inflammation and oxidative stress [49].

Other studies [50,51,52] also showed that creatinine is essential in predicting mortality after heart surgery. The study demonstrated the feasibility of using serum creatinine as an outcome indicator in post-cardiac surgery [53]. The critical role of creatinine as a strong predictor has been incorporated in the different mortality risk scores currently used for cardiac surgery patients [54,55,56]. Creatinine is a chemical waste product. This test is done to see how well your kidneys work. Creatinine is removed from the body entirely by the kidneys. Elevated creatinine level signifies impaired kidney function or kidney disease. As the kidneys become impaired for any reason, the creatinine level in the blood will rise due to poor clearance of creatinine by the kidneys. Abnormally high levels of creatinine thus warn of possible malfunction or failure of the kidneys. When the heart is no longer pumping efficiently it becomes congested with blood, causing pressure to build up in the main vein connected to the kidneys and leading to congestion of blood in the kidneys, too this might facilitate mortality of cardiac patients.

Saturated oxygen was the main predictor for predicting death [57]. Reduced tissue oxygen saturation may be associated with poor postoperative outcomes in cardiac surgery patients [58]. Management and understanding of patient care depend on oxygen saturation. The level of oxygen within the body is tightly regulated as hypoxemia can have a wide variety of acute adverse effects on different organ systems depending on the degree of hypoxia. These include the brain, heart, and kidneys. The blood can’t deliver enough oxygen to your organs and tissues if it has low oxygen levels (hypoxia). If it persists for a long time, it can damage your heart and brain. When hypoxemia occurs acutely, it can be fatal. Such a situation might increase the risk of mortality.

Left Ventricular Ejection Fraction (LVEF) indicates the efficiency of the ventricle and is regarded as an optimal marker of LV function. Left ventricular ejection fraction has been considered among the strongest predictors of clinical outcomes after cardiac surgery [59, 60]. Left Ventricular Ejection Fraction had a significant association with mortality [61]. The ejection fraction low value was the main death predictor, which agrees with [50, 62,63,64,65]. Variables that have often been shown to predict mortality ejection fraction were predictive in this population [47]. Ejection fraction is an indicator of heart strength. It measures the amount of oxygen-rich blood pumped out to the body with each heartbeat. A low ejection fraction is an indicator that the heart can’t plumb enough blood and this further leads to its failure. A low ejection fraction number can be an indicator of heart failure and may not have symptoms at first but can lead to a variety of symptoms, like shortness of breath. This might be the case to drive mortality.

Hemoglobin value was related to worse outcomes. This value is probably related to the patient’s previous comorbidities-related issue, contributing to a worse death outcome in agreement with a study [45]. Lower hemoglobin value has been identified as a predictor of poor short- and long-term outcomes in a nonoperative setting in the general and elderly population [66, 67], in patients with coronary artery disease [68, 69], and in patients with congestive heart failure [70]. Recently, several studies addressed pre-operative Lower hemoglobin as a predictor of poor short-term outcomes after cardiac surgery [71,72,73,74,75]. Pre-operative lower hemoglobin values during cardiopulmonary bypass have been identified as significant risk factors for blood transfusion during cardiac surgery [76]. Lower hemoglobin value leads to anemia. The anemia itself can worsen cardiac function, both because it causes cardiac stress through tachycardia and increased stroke volume, and because it can cause reduced renal blood flow and fluid retention, adding further stress to the heart.

The study on hospital mortality found that increased length of postoperative hospital stay after cardiac surgery is associated with an increased likelihood of in-hospital mortality [77], which is inconsistent with the study; the difference may arise from the study follow-up period and study design. another reason might be since patients are from a limited economic setting, the required money for hospitality might be difficult to afford as such patients might withdraw from treatment, especially for private treatments too early which increases the risk of mortality.

Waiting time was highly related and the most significant predictor of cardiac patients’ post-surgery mortality. However, some papers stated that the surgery outcome was not related to the waiting time [78]. Another study stated that Prolonged waiting was not associated with worse surgical outcomes [79]. Those studies contradicted the study just due to the study area with a lower waiting time unlike to developing country Ethiopia. In Ethiopia, the number of cardiac centers was too few as compared to the case. As of only two fully active cardiac centers for more than 120 million population. Another difference might be the coverage of cardiac disease type.


In conclusion, lower saturated oxygen, prolonged waiting time, aged, lower ejection fraction, short period cardiac center duration after surgery, and lower creatine values were responsible for time-to-death outcomes. Special attention is required for surgery patients under follow-up with those parameters. There are only two fully active cardiac centers in Ethiopia for more than 120 million people, at the end, the study highly and urgently recommended to increase the number of cardiac centers in Ethiopia. The study recommends that clinicians, the Ministry of Health, policymakers, and the general public raise awareness and develop policies for cardiac heart disease patients to facilitate better management and save lives.

Data availability

The data used in the current study is available under the corresponding author and can be attached on request.



New York Health Association


Oxygen Saturated


International Normalised Ratio




low and middle-income countries


  1. Heart Disease & Stroke Statistical Update Fact Sheet Global Burden of Disease [].

  2. Statistics NCfH. Multiple cause of death 2018–2021 on CDC WONDER database. Accessed Febr 2023, 2.

  3. Roth GA, Mensah GA, Johnson CO, Addolorato G, Ammirati E, Baddour LM, Barengo NC, Beaton AZ, Benjamin EJ, Benziger CP. Global burden of cardiovascular diseases and risk factors, 1990–2019: update from the GBD 2019 study. J Am Coll Cardiol. 2020;76(25):2982–3021.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Vervoort D, Swain JD, Pezzella AT, Kpodonu J. Cardiac surgery in low-and middle-income countries: a state-of-the-art review. Ann Thorac Surg. 2021;111(4):1394–400.

    Article  PubMed  Google Scholar 

  5. Vervoort D, Meuris B, Meyns B, Verbrugghe P. Global cardiac surgery: access to cardiac surgical care around the world. J Thorac Cardiovasc Surg. 2020;159(3):987–96. e986.

    Article  PubMed  Google Scholar 

  6. Argaw S, Genetu A, Vervoort D, Agwar FD. The state of cardiac surgery in Ethiopia. JTCVS Open; 2023.

  7. Zilla P, Yacoub M, Zühlke L, Beyersdorf F, Sliwa K, Khubulava G, Bouzid A, Mocumbi AO, Velayoudam D, Shetty D. Global unmet needs in cardiac surgery. Global Heart. 2018;13(4):293–303.

    Article  PubMed  Google Scholar 

  8. Organization WH. Cardiovascular diseases (CVDs): fact sheet 317. Media Centre World Health Organ. 2013;20(1):3.

    Google Scholar 

  9. Noncommunicable diseases [

  10. Akpa OM, Made F, Ojo A, Ovbiagele B, Adu D, Motala AA, Mayosi BM, Adebamowo SN, Engel ME, Tayo B. Regional patterns and association between obesity and hypertension in Africa: evidence from the H3Africa CHAIR study. Hypertension. 2020;75(5):1167–78.

    Article  CAS  PubMed  Google Scholar 

  11. Cappuccio FP, Miller MA. Cardiovascular disease and hypertension in sub-saharan Africa: burden, risk and interventions. Intern Emerg Med. 2016;11:299–305.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Owolabi MO, Akpa OM, Made F, Adebamowo SN, Ojo A, Adu D, Motala AA, Mayosi BM, Ovbiagele B, Adebamowo C. Data resource profile: cardiovascular H3Africa innovation resource (CHAIR). Int J Epidemiol. 2019;48(2):366–g367.

    Article  PubMed  Google Scholar 

  13. Marquez PV, Farrington JL. The challenge of non-communicable diseases and road traffic injuries in sub-Saharan Africa: an overview. 2013.

  14. Moran A, Forouzanfar M, Sampson U, Chugh S, Feigin V, Mensah G. The epidemiology of cardiovascular diseases in sub-saharan Africa: the global burden of diseases, injuries and risk factors 2010 study. Prog Cardiovasc Dis. 2013;56(3):234–9.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Mensah GA, Sampson UK, Roth GA, Forouzanfar MH, Naghavi M, Murray CJ, Moran AE, Feigin VL. Mortality from cardiovascular diseases in sub-saharan Africa, 1990–2013: a systematic analysis of data from the global burden of Disease Study 2013. Cardiovasc J Afr. 2015;26(2):S6.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Compare GBD. [

  17. Gemechu T, Mahmoud H, Parry EH, Phillips DI, Yacoub MH. Community-based prevalence study of rheumatic heart disease in rural Ethiopia. Eur J Prev Cardiol. 2017;24(7):717–23.

    Article  PubMed  Google Scholar 

  18. Organization WH. Noncommunicable diseases country profiles 2014. 2014.

  19. Misganaw A, Mariam DH, Ali A, Araya T. Epidemiology of major non-communicable diseases in Ethiopia: a systematic review. J Health Popul Nutr. 2014;32(1):1.

    PubMed  PubMed Central  Google Scholar 

  20. DataBank population estimates. and projections. [

  21. Mirabel M, Grimaldi A, Freers J, Jouven X, Marijon E. Access to cardiac surgery in sub-saharan Africa. Lancet. 2015;385(9968):606.

    Article  PubMed  Google Scholar 

  22. Forcillo J, Watkins DA, Brooks A, Hugo-Hamman C, Chikoya L, Oketcho M, Thourani VH, Zühlke L, du Toit H, Nghaamwa J. Making cardiac surgery feasible in African countries: experience from Namibia, Uganda, and Zambia. J Thorac Cardiovasc Surg. 2019;158(5):1384–93.

    Article  PubMed  Google Scholar 

  23. Chen J, Li W, Xiang MJJGH. Burden of valvular heart disease, 1990–2017: results from the global burden of Disease Study 2017. 2020, 10(2).

  24. Ljungberg J. Cardiovascular risk factors in aortic stenosis. Umeå universitet; 2018.

  25. Habte B, Alemseged F, Tesfaye DJE. The pattern of cardiac diseases at the cardiac clinic of Jimma University specialised hospital, south West Ethiopia. 2010, 20(2).

  26. Abdissa SG, Oli K, Feleke Y, Goshu DY, Begna DM, Tafese AJE. Spectrum of cardiovascular diseases among Ethiopian patients at tikur anbessa specialized university teaching hospital. Addis ababa. 2014;52(1):9–17.

    Google Scholar 

  27. Yadeta D, Guteta S, Alemayehu B, Mekonnen D, Gedlu E, Benti H, Tesfaye H, Berhane S, Hailu A, Luel AJHA. Spectrum of cardiovascular diseases in six main referral hospitals of Ethiopia. 2017, 9(2).

  28. Defo BKJGha. Demographic, epidemiological, and health transitions: are they relevant to population health patterns in Africa? 2014, 7(1):22443.

  29. Kothari, SSJAopc. Of history, half-truths, and rheumatic fever. 2013, 6(2):117.

  30. Mocumbi AOJCA. Rheumatic heart disease in Africa: is there a role for genetic studies? 2015, 26(2 H3Africa Suppl):S21.

  31. d’Arcy J, Prendergast B, Chambers JB, Ray S, Bridgewater B. Valvular heart disease: the next cardiac epidemic. In.: BMJ Publishing Group Ltd; 2011.

  32. Abera E. Analysis: with 7000 patients on waiting list Ethiopia’s Cardiac Center fights to stay open. Addis Standard. Addis Ababa: Addis Standard; 2019.

    Google Scholar 

  33. Alpaydin E. Introduction to machine learning, 3rd editio. ed. In.: The MIT Press; 2014.

  34. Sharma M, Kumar CJ, Talukdar J, Singh TP, Dhiman G, Sharma A. Identification of rice leaf diseases and deficiency disorders using a novel DeepBatch technique. Open Life Sci. 2023;18(1):20220689.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Sharma M, Kumar CJ. Improving Rice Disease diagnosis using ensemble transfer learning techniques. Int J Artif Intell Tools. 2022;31(08):2250040.

    Article  Google Scholar 

  36. Sharma M, Nath K, Sharma RK, Kumar CJ, Chaudhary A. Ensemble averaging of transfer learning models for identification of nutritional deficiency in rice plant. Electronics. 2022;11(1):148.

    Article  CAS  Google Scholar 

  37. Sharma M, Kumar CJ. Advances in Deep Learning-Based technologies in Rice Crop Management. Computer Vision and Machine Learning in Agriculture, volume 3. edn.: Springer; 2023. pp. 79–89.

  38. Bhadra S, Kumar CJ. Enhancing the efficacy of depression detection system using optimal feature selection from EHR. Comput Methods Biomech BioMed Eng. 2024;27(2):222–36.

    Article  Google Scholar 

  39. Bhadra S, Kumar CJ. An insight into diagnosis of depression using machine learning techniques: a systematic review. Curr Med Res Opin. 2022;38(5):749–71.

    Article  PubMed  Google Scholar 

  40. Kumar CJ, Das PR, Hazarika A. Autism spectrum disorder diagnosis and machine learning: a review. Int J Med Eng Inf. 2022;14(6):512–27.

    Google Scholar 

  41. Cox DRJJRSSSB. Regression models and life-Tables 1972, 34(2):187–202.

  42. Jamieson WE, Germann E, Ye J, Chan F, Cheung A, MacNab JS, Fradet GJ, Stanford EA, Bryson LA. Lichtenstein SVJTAots: Effect of prosthesis-patient mismatch on long-term survival with mitral valve replacement: assessment to 15 years. 2009, 87(4):1135–42.

  43. Appelbaum A, Kouchoukos NT, Blackstone EH, Kirklin JWJTAJC. Early risks of open heart surgery for mitral valve disease. 1976, 37(2):201–9.

  44. Murana G, Alfonsi J, Savini C, Mariani C, Coppola G, Lo Coco V, Pilato E, Pacini D, Di Bartolomeo RJIC. On-X mitral valve replacement: a single-centre experience in 318 patients. Surg T. 2018;27(6):836–41.

    Google Scholar 

  45. Moreira JL, Barletta PHAAS, Baucia JAJBJCS. Morbidity and mortality in patients undergoing mitral valve replacement at a cardiovascular surgery referral service: a retrospective analysis. 2021, 36:183–91.

  46. Barons MJ, Turner S, Parsons N, Griffiths F, Bethell H, Weich S, Thorogood M. Fitness predicts long-term survival after a cardiovascular event: a prospective cohort study. BMJ open. 2015;5(10):e007772.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Weintraub WS, Clements SD Jr, Crisco LV-T, Guyton RA, Craver JM, Jones EL, Hatcher CR Jr. Twenty-year survival after coronary artery surgery: an institutional perspective from Emory University. Circulation. 2003;107(9):1271–7.

    Article  PubMed  Google Scholar 

  48. Lemaire A, Soto C, Salgueiro L, Ikegami H, Russo MJ, Lee LY. The impact of age on outcomes of coronary artery bypass grafting. J Cardiothorac Surg. 2020;15(1):1–8.

    Article  Google Scholar 

  49. Curtis AB, Karki R, Hattoum A, Sharma UC. Arrhythmias in patients ≥ 80 years of age: pathophysiology, management, and outcomes. J Am Coll Cardiol. 2018;71(18):2041–57.

    Article  PubMed  PubMed Central  Google Scholar 

  50. Pieri M, Belletti A, Monaco F, Pisano A, Musu M, Dalessandro V, Monti G, Finco G, Zangrillo A, Landoni G. Outcome of cardiac surgery in patients with low preoperative ejection fraction. BMC Anesthesiol. 2016;16(1):1–10.

    Article  Google Scholar 

  51. Roques F, Nashef S, Michel, PJTJohvd. Risk factors for early mortality after valve surgery in Europe in the 1990s: lessons from the EuroSCORE pilot program. 2001, 10(5):572–7; discussion 577.

  52. Khan MF, Khan MS, Bawany FI, Dar MI, Hussain M, Farhan S, Fatima K, Hamid K, Arshad MH. Aziz MJGJoHS: predictors of mortality in patients undergoing mitral valve replacement. 2016, 8(3):37.

  53. Hou J, Shang L, Huang S, Ao Y, Yao J, Wu Z. Postoperative serum creatinine serves as a prognostic predictor of cardiac surgery patients. Front Cardiovasc Med. 2022;9:740425.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Nilsson J, Algotsson L, Höglund P, Lührs C, Brandt J. Comparison of 19 pre-operative risk stratification models in open-heart surgery. Eur Heart J. 2006;27(7):867–74.

    Article  PubMed  Google Scholar 

  55. Geissler HJ, Hölzl P, Marohl S, Kuhn-Régnier F, Mehlhorn U, SuÈdkamp M, de Vivie ER. Risk stratification in heart surgery: comparison of six score systems. Eur J Cardiothorac Surg. 2000;17(4):400–6.

    Article  CAS  PubMed  Google Scholar 

  56. Badreldin A, Doerr F, Kroener A, Wahlers T, Hekmat K. Preoperative risk stratification models fail to predict hospital cost of cardiac surgery patients. J Cardiothorac Surg. 2013;8(1):1–7.

    Article  Google Scholar 

  57. Zhong Z, Yuan X, Liu S, Yang Y, Liu F. Machine learning prediction models for prognosis of critically ill patients after open-heart surgery. Sci Rep. 2021;11(1):1–10.

    Article  CAS  Google Scholar 

  58. Sanders J, Toor IS, Yurik TM, Keogh BE, Mythen M, Montgomery HE. Tissue oxygen saturation and outcome after cardiac surgery. Am J Crit Care. 2011;20(2):138–45.

    Article  PubMed  Google Scholar 

  59. Arroyo JP, Schweickert AJ. Back to basics in physiology: fluids in the renal and cardiovascular systems. Academic; 2013.

  60. Marwick TH. Ejection fraction pros and cons: JACC state-of-the-art review. J Am Coll Cardiol. 2018;72(19):2360–79.

    Article  PubMed  Google Scholar 

  61. Kurniawaty J, Setianto BY, Supomo, Widyastuti Y, Boom CE. The effect of low preoperative ejection fraction on mortality after cardiac surgery in Indonesia. Vasc Health Risk Manag 2022:131–7.

  62. Remenyi B, ElGuindy A, Smith SC, Yacoub M, Holmes DRJTL. Valvular aspects of rheumatic heart disease. 2016, 387(10025):1335–46.

  63. Lam B-K, Chan V, Hendry P, Ruel M, Masters R, Bedard P, Goldstein B, Rubens F. Mesana TJTJot, surgery c: the impact of patient–prosthesis mismatch on late outcomes after mitral valve replacement. 2007, 133(6):1464–73. e1463.

  64. Rankin JS, Hammill BG, Ferguson TB Jr, Glower DD, O’Brien SM, DeLong ER, Peterson ED. Edwards FHJTJot, surgery c: determinants of operative mortality in valvular heart surgery. 2006, 131(3):547–57.

  65. Rahimtoola SHJJotACoC. Choice of prosthetic heart valve in adults: an update. 2010, 55(22):2413–26.

  66. Sarnak MJ, Tighiouart H, Manjunath G, MacLeod B, Griffith J, Salem D, Levey AS. Anemia as a risk factor for cardiovascular disease in the atherosclerosis risk in communities (ARIC) study. J Am Coll Cardiol. 2002;40(1):27–33.

    Article  PubMed  Google Scholar 

  67. Izaks GJ, Westendorp RG, Knook DL. The definition of anemia in older persons. JAMA. 1999;281(18):1714–7.

    Article  CAS  PubMed  Google Scholar 

  68. Sabatine MS, Morrow DA, Giugliano RP, Burton PB, Murphy SA, McCabe CH, Gibson CM, Braunwald E. Association of hemoglobin levels with clinical outcomes in acute coronary syndromes. Circulation. 2005;111(16):2042–9.

    Article  CAS  PubMed  Google Scholar 

  69. Reinecke H, Trey T, Wellmann J, Heidrich J, Fobker M, Wichter T, Walter M, Breithardt G, Schaefer RM. Haemoglobin-related mortality in patients undergoing percutaneous coronary interventions. Eur Heart J. 2003;24(23):2142–50.

    Article  CAS  PubMed  Google Scholar 

  70. Tang Y-D, Katz SD. Anemia in chronic heart failure: prevalence, etiology, clinical correlates, and treatment options. Circulation. 2006;113(20):2454–61.

    Article  PubMed  Google Scholar 

  71. Zindrou D, Taylor KM, Bagger JP. Preoperative haemoglobin concentration and mortality rate after coronary artery bypass surgery. Lancet. 2002;359(9319):1747–8.

    Article  CAS  PubMed  Google Scholar 

  72. Magovern JA, Sakert T, Magovern GJ Jr, Benckart DH, Burkholder JA, Liebler GA, Magovern Sr GJ. A model that predicts morbidity and mortality after coronary artery bypass graft surgery. J Am Coll Cardiol. 1996;28(5):1147–53.

    Article  CAS  PubMed  Google Scholar 

  73. Kulier A, Levin J, Moser R, Rumpold-Seitlinger G, Tudor IC, Snyder-Ramos SA, Moehnle P, Mangano DT. Impact of preoperative anemia on outcome in patients undergoing coronary artery bypass graft surgery. Circulation. 2007;116(5):471–9.

    Article  PubMed  Google Scholar 

  74. Cladellas M, Bruguera J, Comín J, Vila J, De Jaime E, Martí J, Gomez M. Is pre-operative anaemia a risk marker for in-hospital mortality and morbidity after valve replacement? Eur Heart J. 2006;27(9):1093–9.

    Article  PubMed  Google Scholar 

  75. Karkouti K, Wijeysundera DN, Beattie WS. Risk associated with preoperative anemia in cardiac surgery: a multicenter cohort study. Circulation. 2008;117(4):478–84.

    Article  PubMed  Google Scholar 

  76. Scott DA, Tung H-MA, Slater R. Perioperative hemoglobin trajectory in adult cardiac surgical patients. J Extracorpor Technol. 2015;47(3):167.

    Article  Google Scholar 

  77. Mazzeffi M, Zivot J, Buchman T, Halkos M. In-hospital mortality after cardiac surgery: patient characteristics, timing, and association with postoperative length of intensive care unit and hospital stay. Ann Thorac Surg. 2014;97(4):1220–5.

    Article  PubMed  Google Scholar 

  78. Badakhshan A, Arab M, Gholipour M, Behnampour N, Saleki S. Heart surgery waiting time: assessing the effectiveness of an action. Iran Red Crescent Med J 2015, 17(8).

  79. Ray AA, Buth KJ, Sullivan JA, Johnstone DE, Hirsch GM. Waiting for cardiac surgery: results of a risk-stratified queuing process. Circulation. 2001;104(suppl1):I–92.

    Google Scholar 

Download references


The authors must forward their acknowledgment to all the health staff at the two Cardiac centers. Data collectors shall also owe our heartfelt thanks for their cooperation in the success of this research work.


There is no agent/institution that funded this research.

Author information

Authors and Affiliations



Melaku Tadege, Awoke Seyoum Tegegne and Zelalem G. Dessie, all authors made a significant equal contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Corresponding author

Correspondence to Melaku Tadege.

Ethics declarations

Ethics approval

There was no direct human involvement in the study. Ethics approval was granted and the informed consent was waived by the Bahir Dar University Science College’s ethical committee. And permission from the health institutions was taken. All methods were performed in accordance with the relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Tadege, M., Tegegne, A.S. & Dessie, Z.G. Post-surgery survival and associated factors for cardiac patients in Ethiopia: applications of machine learning, semi-parametric and parametric modelling. BMC Med Inform Decis Mak 24, 91 (2024).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: