Keywords
DKA, Diabetes ketoacidosis, Mortality, East Africa, predictors
Diabetes ketoacidosis occurs at an estimated rate between 4.6 and 8 episodes per 100 patient-years in individuals with diabetes. Up to 47% of diabetes-related ketoacidosis deaths have been reported in East Africa, which is an unacceptably high rate. Socioeconomic factors, including cost and the unpredictability of insulin supplies, as well as social, cultural, and economic factors, contribute to the difficulty in controlling ketoacidosis in patients in Africa.
To conduct systematic review and meta-analysis on mortality and its predictors among diabetic ketoacidosis patients in East Africa: A 2025 Systematic review and meta-analysis.
A Comprehensive search of MEDLINE (via PubMed), the Cochrane Database of Systematic Reviews, UpToDate, EMBASE, and NICE (National Institute for Health and Care Excellence) was performed to identify studies reporting mortality and its predictors in adult patients with diabetes ketoacidosis in East Africa. A Random effect- model was applied for pooled diabetes ketoacidosis mortality incidence. The quality of the studies was assessed using the Newcastle-Ottawa scale. The evaluation of publication bias revealed no evidence of bias.
The final analysis included 2,260 patients from 11 studies. The pooled incidence of mortality among patients with diabetic ketoacidosis in East Africa was 12.02% (95% CI: 5.99–18.04; I2 = 89.9%; p < 0.001). Comorbidity (OR = 2.15; 95% CI: 1.56–2.74) and sepsis (OR = 1.77; 95% CI: 1.60–1.95) were found to be significant predictors of diabetic ketoacidosis related mortality.
The overall incidence of mortality among adult patients with diabetic ketoacidosis in East Africa remains high. The presence of comorbidity and sepsis increases mortality among adult patients with diabetic ketoacidosis.
DKA, Diabetes ketoacidosis, Mortality, East Africa, predictors
According to the International Diabetes Federation (IDF), the prevalence of diabetes was estimated to be 9.3% (463 million) in 2019 and is projected to reach 10.2% (578 million) by 2030 and 10.9% (700 million) by 2045.1 Globally, half a billion people have diabetes, and this figure represents a 25% increase in the number of diabetes by 2030, and 51% increase by 2045.1 Diabetes-related deaths have increased significantly, accounting for approximately five million deaths over the past ten years.
The life-threatening acute consequences of diabetes mellitus include diabetic ketoacidosis (DKA), hyperosmolar hyperglycemic state (HHS), and hypoglycemia associated with treatment.1,2 DKA is the most common and potentially life-threatening acute complication of DM that progresses rapidly to life treating illness; thus, it requires immediate medical intervention.3
In the US, 30 million diabetic patients suffer from diabetic ketoacidosis.4 There are an estimated 4.6–8 episodes of DKA for every 100 years of diabetes. The majority of DKA cases occurred in the 18–44 age range (56%). The next most common age range is 45–65 years (24%).5
Over the past two–three decades, the prevalence of DKA has dramatically increased in Africa, leading to serious economic issues with increased drug use and bed occupancy.6,7 Patients in Africa face significant challenges in managing diabetes and preventing ketoacidosis due to socioeconomic factors, such as the cost and unpredictable nature of insulin supplies, as well as social, cultural, and economic factors.8–11 Poorly controlled diabetes mellitus associated with DKA, which can lead to neurologic complications and death.3,12
In Africa, deaths due to diabetic ketoacidosis are disproportionately high and vary markedly, ranging from 7.5% in Zambia13 to 47.8% in Rwanda,14 with intermediate rates of 15.1% in Ethiopia15 and 29.8% in Kenya.16 Notably, such variations were also observed within individual countries. These findings highlight the need for a systematic review and meta-analysis to generate a strong conclusion that can guide both clinical practice and policy-making.
Several factors have been identified as significant predictors of DKA-related mortality, including smoking, infection, presence of comorbidities, severe hypokalemia, inadequate potassium replacement, insulin administration errors, sepsis, inappropriate fluid administration, age, severity of DKA, renal failure, aspiration, and altered mental status.13,17–21
Although several studies have been conducted in various East African countries, the findings are inconsistent ranges from 1.3%-47.8%. Such variability may pose challenges for clinical decision making and draw sound conclusions. Therefore, this study aimed to determine the incidence of DKA-related mortality and its predictors, with the result that the finding is intended to support clinicians and stakeholders in patient outcomes.
This systematic review and meta-analysis was conducted in accordance with the PRISMA guidelines used to report observational studies.22 The authors conducted a comprehensive search of the literature published between March 2008 and June 2025 across several databases, including MEDLINE (via PubMed), NICE (National Institute for Health and Care Excellence), UpToDate, EMBASE, and the Cochrane Database of Systematic Reviews. Ongoing systematic reviews were assessed using PROSPERO. PubMed Medical Subject Headings (MeSH) and text terms such as “diabetes keto acidosis”, “mortality”, and “hyperglycemic emergency”, and/or related factors were searched, either separately or in combination. Hand searches of reference lists were conducted, alongside adapted searches for other databases and forward citation searches using Web of Science and Google Scholar. The search strategy also included gray and unpublished literature (Supplementary Table 1).23
This systematic review and meta-analysis included cross-sectional, prospective cohort, and retrospective follow-up studies reporting mortality and its predictors among adult patients with diabetic ketoacidosis in East Africa. Studies that did not address diabetes ketoacidosis mortality or its predictors or studies that lacked sufficient data for analysis were excluded. Only complete manuscripts published in English were included.
Based on predefined study selection criteria, two authors (N.A. and A. G.) independently reviewed the studies for eligibility. The retrieved studies were exported to EndNote version 8 reference managers to remove possible duplicate studies. After screening the titles, abstracts, and keywords, the full texts of all potentially relevant studies were retrieved. Studies reporting mortality and its predictors among adult patients with diabetic ketoacidosis in East Africa were included.23 Additionally, any differences among the authors regarding the selection and inclusion of papers will be discussed and addressed ( Figure 1).
This systematic review and meta-analysis included studies reporting mortality and its predictors among adult patients with diabetic ketoacidosis. However, studies that did not report mortality due to diabetic ketoacidosis or its associated factors or insufficient data for analysis were excluded. The search focused on the study population, regardless of intervention, comparison, or outcome. Studies published in English were also included, whereas studies without abstracts and full texts, as well as qualitative studies, were excluded.
The Newcastle-Ottawa Scale was used to assess the quality of the included studies.24 Two authors (N.A and A.G) evaluated the quality of the studies before selecting them for analysis. Observational studies were assessed in three categories: selection (four questions), comparability (two questions), and outcome of interest (three questions), each of which received one point if the study met the criteria, except for the comparability of research groups, where two points were awarded if the study was adjusted for age, sex, or both, and for other confounding factors (Supplementary Table 2). The included studies had average quality in this systematic review and meta-analysis. Disagreements were resolved by consensus among the authors and collective re-evaluation of the relevant articles.
The primary objective of this systematic review and meta-analysis was to estimate the pooled mortality related to diabetic ketoacidosis among adult patients with diabetes in East Africa. The secondary outcome was the identification of factors that predict the occurrence of mortality among adult patients with diabetic ketoacidosis. Two independent authors (A. Z. and M. F.) extracted data using a predetermined data collection Microsoft Excel spreadsheet controlled by two others (A. G. and N. A.). The extracted data included the name of the first author, year of publication, study design, sample size, number of study populations, odds ratio, incidence of mortality related to diabetes ketoacidosis, diabetes ketoacidosis severity, patient age, sex, hospital length of stay, insulin administration, potassium supplementation, infection, and presence of comorbidities. In certain cases, the study’s first and final authors were emailed to obtain information on any ambiguities or missing reports. One month later, a second email was sent if there was no response.23
We computed the weighted mortality of diabetic ketoacidosis with corresponding 95% CI using the random-effect model described by Desrimonian and Liard.25 The Meta and Forest Plot programs in STATA version 17 were used for statistical pooling and quantitative data analysis of the meta-analyses and forest plots. Heterogeneity within the group was assessed using the I2 test, which determines the proportion of total variation among trials. Subgroup analysis was employed to address heterogeneity by defining the study design, sample size, study, and country of the study. A sensitivity test was conducted to assess the contribution of single studies to the pooled incidence of mortality due to diabetic ketoacidosis. Publication bias was assessed by visual inspection of the funnel plot using the Meta bias Egger test and Meta-trim fill analysis.23
Following title and abstract review, 3207 of the 3278 records retrieved from various databases were excluded. Finally, 51 studies were screened for full-text review, and 11 articles met the inclusion criteria and were included in the final analysis ( Figure 1).
In this systematic review and meta-analysis, 11 studies (n = 2260) were included: eight retrospective and three cross-sectional studies that aimed to estimate the pooled incidences of mortality among patients with DKA. Six studies from Ethiopia, two from Kenya, and one each from Tanzania, Rwanda, and Zambia contributed data for analysis. Overall, 192 deaths occurred among 2260 hospitalized patients with diabetic ketoacidosis ( Table 1).
S.no | Author name | Year | Study country | Study design | Study population | Number of deaths | Incidences of mortality |
---|---|---|---|---|---|---|---|
1. | Sairabanu S et al26 | 2023 | Kenyia | Retrospective | 150 | 2 | 1.3% |
2. | Shabani I et al27 | 2017 | Tanzania | Retrospective | 29 | 7 | 24.1% |
3. | François R et al14 | 2021 | Rwanda | Cross-sectional | 23 | 11 | 47.8% |
4. | Kassaye D et al15 | 2018 | Ethiopia | Retrospective | 357 | 54 | 15.1% |
5. | Getu T et al20 | 2021 | Ethiopia | Retrospective | 225 | 27 | 12% |
6. | Temesgen D et al28 | 2022 | Ethiopia | Retrospective | 452 | 23 | 5.1% |
7. | Otieno C et al16 | 2010 | Kenyia | Cross-sectional | 47 | 14 | 29.8% |
8. | Derse T et al19 | 2023 | Ethiopia | Cross-sectional | 357 | 14 | 3.9% |
9. | Mekonen G et al18 | 2022 | Ethiopia | Retrospective | 387 | 17 | 4.4% |
10. | Tekeste P et al29 | 2021 | Ethiopia | Retrospective | 153 | 17 | 11.1% |
11. | Kakusa M et al13 | 2016 | Zambia | Retrospective | 80 | 6 | 7.5% |
Diabetes ketoacidosis related-mortality among adult patients in East Africa ranges from 1.3%-47.8%. The random effects model analysis in this study revealed that the incidence of mortality among DKA patient in East Africa found to be 12.02% (95% CI: 5.99–18.04; I2 = 89.9%; p < 0.001) ( Figure 2). Subgroup analysis was conducted in the stratified study countries. The findings showed that the incidence of mortality among DKA found to be 8.18% (4.9, 11.46) in Ethiopia and 14.78 (13.1,42.6) in Kenya (Supplementary figure 1).
Publication bias was assessed by visual inspection of the meta-funnel plot, which showed asymmetrical distribution (Supplementary figure 2). The Egger regression test revealed p < 0.0001, suggesting the presence of publication bias (Supplementary figure 3). Finally, a meta-trim-fill analysis was conducted, and no imputation was performed.
Sensitivity analysis: Furthermore, a leave-one-out sensitivity analysis was conducted to identify the potential source of heterogeneity in the incidence of mortality among adult DKA patients in East Africa. This sensitivity analysis demonstrated that the study does not depends on a single study. In conclusion, the mortality rate among adult patients with diabetic ketoacidosis are found to be 12.02% (95% CI: 5.99–18.04; I2 = 89.9%; p < 0.001).
Several risk factors are significantly associated with treatment outcomes of diabetes ketoacidosis among patients with diabetes in East Africa. Different studies from East Africa have revealed that smoking, urinary tract infection, comorbidities, severe hypokalemia,20,29 inadequate potassium replacement, insulin administration error, sepsis, inappropriate fluid administration,15,19,29 age, severity of DKA,18 presence of renal failure, aspiration, altered mental status,13 severity of DKA, newly diagnosed diabetes,26,28,29 and presence of medical insurance,14 are factors contributing to diabetes ketoacidosis treatment outcomes.
Eight variables were selected for analysis in 11 studies. In this thesis, the presence of sepsis, comorbidities, potassium supplementation, and time of diabetes diagnosis were pooled. The pooled estimate of this systematic review and meta-analysis showed that the presence of sepsis and comorbidities was a significant predictor of mortality among adult patients with diabetes ketoacidosis. Patients with diabetes ketoacidosis who had any comorbidities had a higher mortality rate 2.15 (95% CI; 1.56, 2.74), than those DKA patients without any comorbidities (supplementary Figure 4). Additionally, the mortality rate was 1.77 (95% CI: 1.60, 1.95), greater for DKA patients with a sepsis diagnosis than for DKA patients without a sepsis diagnosis (Supplementary figure 5).
Patients with DKA, receive appropriate care should recover fully in less than a day.30,31 Prolonged DKA occurrence results in thromboembolic problems, cerebral edema, cardiac arrhythmia, hypoglycemia, hypokalemia,32 renal failure, coma, cerebral edema, cardiac arrhythmia,33,34 lung edema with respiratory failure, and death.35,36 Recovery is associated with prompt diagnosis and timely initiation of DKA treatment. This, in turn, reduces hospital bed occupancy, the duration of hospital stay, and the quantity of material and human resources required to manage it and its complications.
The purpose of this systematic review and meta-analysis was to evaluate mortality and its predictors in adult patients with DKA. As a result, the pooled incidence DKA related mortality among diabetes patient was found to be 12.02%. The findings of this study were lower than study conducted in 25.9%.37 The possible reasons for this discrepancy might be differences in the subjects of the included studies. A previous systematic review and meta-analysis was conducted among DKA patients who were diagnosed with SARS-CoV-2,37 while in the current review, the study subjects were all DKA patients, regardless of their comorbidities. In fact, the presence of comorbidities such as SARS-CoV-2 worsens the treatment outcome of DKA.38–40 In addition, the authors did not rule out the possibility that mortality was due to DKA or SARS-CoV-2.
In this finding, the mortality of DKA among adult D. M was lower Compared to that study conducted in India (18.9%).41 The possible reasons for this discrepancy might be the differences in the study settings. In India, the study was conducted among DKA patients admitted to the ICU with and without NIV.41 The mortality of patients admitted to the ICU is higher than that of other inpatients.42
The current systematic review revealed that the mortality of DKA among adult patients was higher than the study conducted in Japan (3.3%).43 A possible reason for this discrepancy might be differences in socio-economic factors such as infrastructure, trained human power, and resources. Several factors predict DKA-related mortality in Africa, including availability of medical care, access to insulin, and other forms of treatment, in addition to social, cultural, and ethnic issues.11
The findings of this systematic review and meta-analysis revealed that comorbidities and the presence of sepsis were significant predictors of mortality among patients with diabetic ketoacidosis. Patients with diabetes ketoacidosis who had any comorbidities had a higher mortality rate than those with DKA without any comorbidities, which is consistent with the results of studies conducted in Kenya Nairobi26 and Jimma Ethiopia.44 A possible reason for this might be comorbid conditions such as heart disease and renal failure causing multi-organ dysfunction, as patients with both comorbidities and DKA had a higher risk of death.
Patients with diabetic ketoacidosis who were diagnosed with sepsis had a higher mortality rate than those with DKA without sepsis, which is supported by a study conducted in Kenya Nairobi,26 Jimma, Ethiopia,44 Addis Ababa, Ethiopia,19 Hiwot-fana, Ethiopia,29 and Saudi Arabia.45 This may be because sepsis results in the production of counter-regulatory hormones, such as adrenal and cortisol, which in turn decrease the effect of insulin on the regulation of glucose.46
These findings are of key importance to clinicians, hospital decision makers, and health policymakers. Prioritizing the early management of high-risk patients, ensuring adequate resources for sepsis and comorbidity management, and implementing standard care pathways can improve outcomes. The study also helps synthesize previous conflicting results, delivering a strengthened evidence base to inform clinical practice and regional health policies.
The cumulative mortality rate among patients with diabetic ketoacidosis in East Africa remains high. Sepsis and comorbidities are independent predictors of mortality and highlight the need for increased clinical suspicion in patients who present with these risk factors. Early identification and targeted treatment of comorbidities and sepsis are essential to prevent in-hospital mortality.
All data used in this systematic review and meta-analysis are openly available in Zendo under a CC0 1.0 Universal license.
The data set includes full data extraction sheet with study characteristics, sample sizes, outcomes, and other variables used for pooling estimates on Mortality and its predictors in adult diabetic ketoacidosis patients in East Africa, along with the PRISMA checklist for this review. The data extraction dataset can be accessed at: https://doi.org/10.5281/zenodo.16917476.47
The PRISMA checklist for this review is also available at Zenodo: https://doi.org/10.5281/zenodo.1691803348
All supplementary files are available at https://doi.org/10.5281/zenodo.16932536 .49 This includes:
• Supplementary figure 1: Forest plot showing subgroup analysis of study stratified by study countries on incidences of mortality among DKA patient in East Africa.
• Supplementary figure 2: Meta funnel plot showing publication bias on studies incidences of mortality among adult DKA patient in East Africa.
• Supplementary figure 3: Meta bias, egger test showing publication bias on studies incidences of mortality among adult DKA patient in East Africa.
• Supplementary figure 4: Forest plot showing the presences of association between comorbidities and mortality of adult DKA East Africa.
• Supplementary figure 5: Forest plot showing the presences of association between sepsis and mortality of adult DKA East Africa.
• Supplementary table 1: Search strategies used for one of the data bases.
• Supplementary table 2: Quality of studies assessed with New Castle-Ottawa scale.
Views | Downloads | |
---|---|---|
F1000Research | - | - |
PubMed Central
Data from PMC are received and updated monthly.
|
- | - |
Provide sufficient details of any financial or non-financial competing interests to enable users to assess whether your comments might lead a reasonable person to question your impartiality. Consider the following examples, but note that this is not an exhaustive list:
Sign up for content alerts and receive a weekly or monthly email with all newly published articles
Already registered? Sign in
The email address should be the one you originally registered with F1000.
You registered with F1000 via Google, so we cannot reset your password.
To sign in, please click here.
If you still need help with your Google account password, please click here.
You registered with F1000 via Facebook, so we cannot reset your password.
To sign in, please click here.
If you still need help with your Facebook account password, please click here.
If your email address is registered with us, we will email you instructions to reset your password.
If you think you should have received this email but it has not arrived, please check your spam filters and/or contact for further assistance.
Comments on this article Comments (0)