Study selection
A total 163 published and 2 unpublished literatures were identified from several electronic databases and Addis Ababa digital library, respectively. Of the total identified studies, 45 duplicates papers were removed and 97 records were removed by reviewing titles and abstracts. The full text of the remaining 23 studies were assessed and screened for eligibility. Thirteen studies were excluded because studies were not conducted in Ethiopia, outcome was not chronic kidney and participants were not diabetic patients. Finally, 10 articles which scored seven and above on the JBI quality appraisal eligibility criteria were included in the systematic review and meta-analysis. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram to present the systematic review overview [26] (Figure 1).
Features of included studies
Of the total included studies, all were published articles except one study which was retrieved from gray literature [28]. Regarding to study design, eight studies were cross-sectional study design [16-19, 25, 28-30], one case-control [31], and one retrospective cohort study design [32]. The studies were conducted from 2010 [29] to 2019 [18, 19] in various regions of the country. An overall 3091 diabetic patients were participated in this study. The sample sizes ranged from minimum of 163 [19] to maximum of 435 [32] participants from study conducted in Addis Ababa. Of the ten studies included in the final analysis, four in Addis Ababa [19, 28, 30, 32], two in Oromia region [18, 29], Two in Amhara region [16, 17], One in SNNP region [25] and One in Tigrai region [31] (Table 1).
Table 1: Summary of included studies regarding burden and determinants of CKD among diabetic patients in Ethiopia, 2020.
S.N
|
Author
|
Year
|
Region
|
Study design
|
Study area
|
Method of estimating eGFR
|
CKD definition
|
Sample size
|
Response rate
|
No of CKD
|
Burden of CKD (95% CI)
|
1
|
Temesgen F, et al.[25]
|
2014
|
SNNP
|
Cross-sectional
|
Butajira hospital
|
MDRD
|
eGFR <60 ml/min/1.73 m2
|
214
|
100%
|
39
|
18.2 (13.0, 23.4)
|
2
|
Kidist R, et al. [16]
|
2017
|
Amhara
|
Cross-sectional
|
Felegehiwot hospital
|
not stated
|
eGFR <60 ml/min/1.73 m2
|
344
|
100%
|
39
|
11.3 (7.9, 14.6)
|
3
|
Kabaye K, et al. [18]
|
2019
|
Oromia
|
Cross-sectional
|
JUMC
|
MDRD
|
eGFR <60 ml/min/1.73 m2
|
208
|
100%
|
54
|
25.9 (20.0, 31.9)
|
4
|
Shawaneh D M. et al. [17]
|
2018
|
Amhara
|
Cross-sectional
|
UGH
|
MDRD
|
eGFR <60 ml/min/1.73 m2
|
229
|
100%
|
50
|
21.8 (16.4, 27.1)
|
5
|
Dawit W, et al.[29]
|
2010
|
Oromia
|
Cross-sectional
|
JUMC
|
not stated
|
eGFR <60 ml/min/1.73 m2
|
305
|
100%
|
48
|
15.7 (11.6, 19.8)
|
6
|
M Gizaw, et al.[30]
|
2015
|
AA
|
Cross-sectional
|
Black Lion Hospital
|
not stated
|
eGFR <60 ml/min/1.73 m2
|
418
|
100%
|
39
|
9.3 (6.5, 12.1)
|
7
|
Getahun C, et al. [19]
|
2019
|
AA
|
Cross-sectional
|
Black Lion Hospital
|
MDRD
|
eGFR <60 ml/min/1.73 m2
|
163
|
100%
|
39
|
23.9 (17.3, 30.4)
|
8
|
Meron M.[28]
|
2016
|
AA
|
Cross-sectional
|
AA
|
MDRD and Cockcroft Gault equation
|
eGFR <60 ml/min/1.73 m2
|
355
|
100%
|
68
|
19.1 (15.0, 23.2)
|
9
|
Solomon H,et al. [31]
|
2017
|
Tigrai
|
Case control
|
Ayider referral hospital
|
not stated
|
eGFR <60 ml/min/1.73 m2
|
420
|
100%
|
84
|
20.0 (16.1, 23.8)
|
10
|
Alemayehu H, et al.[32]
|
2018
|
AA
|
Retrospective follow up
|
St. Paul’s Hospital
|
Cockcroft-Gault equation
|
eGFR <60 ml/min/1.73 m2
|
435
|
100%
|
62
|
14.2 (10.9, 17.5)
|
AA: Addis Ababa; CKD:Chronic Kidney Disease; eGFR: Glomular Filtration Rate; JUMC: Jimma University Medical College;MDRD: Modification of Diet in Renal Disease; SNNP: Southern Nation, nationalities and peoples; UGH: University of Gondar Hospital
Burden of chronic kidney disease among diabetic patients in Ethiopia
In this meta-analysis, we found significant heterogeneity across studies (I2 = 84.6%, p = 0.00), which is an indicator to use random effect-model to estimate the pooled burden of chronic kidney disease among diabetic patients. The findings of original studies indicated that there were uneven and inconclusive burden of chronic kidney disease among diabetic patients in Ethiopia. From forest plot the largest burden was observed in study conducted in Jimma University Medical Center (JUMC), Oromia region 25.9 (95% CI: 20.0, 31.9) [18] while the smallest burden was reported in Black Lion hospital, Addis Ababa 9.3 (95% CI: 6.5, 12.1) [30].The pooled burden of chronic kidney disease among diabetic patients was 17.55 % (95%CI: 14.23–20.88) (Figure 2).
Meta regression was computed to see underlying sources of heterogeneity using sample size and year of publication, but none of them showed a statistically significance presence of heterogeneity. Moreover, to minimize potential heterogeneity, subgroup analysis was conducted based on the region where the studies were conducted. Its result showed the highest burden in Oromia region while the smallest was seen in Addis Ababa (Figure 3).
To see for the presence of publication bias, graphical funnel plot and Egger’s test at 5% significance level were computed (Figure 5). The asymmetric funnel plot indicates the presence of publication bias. In addition, Egger’s and Begg’s tests showed statistically significant presence of publication bias (p = 0.001, 0.006), respectively (Table 2).
Table 2: Meta regression using sample size and year of publication to observe related heterogeneity on burden of CKD among diabetic patients in Ethiopia, 2019
. Variables
|
Coefficients
|
p-value
|
Publication Year
|
0.0344531
|
0.309
|
Sample size
|
0.0008549
|
0.058
|
Sensitivity analyses of the studies were done to test the effect of a single study on the pooled result of remaining studies using random effect model. We found no strong suggestion for influence of individual study on remaining studies (Figure 4).
Determinants of chronic kidney disease among diabetic patients in Ethiopia
Sex and chronic kidney disease among diabetic patients
To see the effect of sex on chronic kidney disease among diabetic patients, six studies were included in meta-analysis [17-19, 28, 31, 32]. The pooled result showed that, there was no a statistically significant association between sex of the patients and chronic kidney disease among diabetic patients in Ethiopia (OR=0.89, 95%, CI: 0.70, 1.14) (Figure 6).
Hypertension and Chronic kidney disease among diabetic patients
To observe the pooled effect of hypertension history on chronic kidney disease, five studies were selected in final meta-analysis [17, 19, 28, 31, 32]. Since, high heterogeneity was observed (I2 = 81.7, P=value < 0.001), a random effects model was used to report the effect of hypertension on chronic kidney disease. Three studies [17, 19, 31] showed the presence of a statistically significant association between hypertension and chronic kidney disease, while two studies [28, 32] did not show a statistical significance association. The finding discovered that the odds of developing chronic kidney disease was 2.65 times more likely among hypertensive patients than non-hypertensive patients. (OR=2.65, 95%, CI: 1.38, 5.09) (Figure 7).
Types of DM and chronic kidney disease among diabetic patients
To compute the effect of types of DM on chronic kidney disease, three studies were selected for meta-analysis [17, 28, 31]. A random effects model was used to estimate the pooled effect of types of DM on chronic kidney disease (I2 =73.1, P-value < 0.024). Two of the included studies showed a statistically significance association [17, 31] while one of the study did not indicated a statistically significant between types of DM and chronic kidney disease [28]. The pooled result of the analysis revealed type I DM decrease the odds of developing chronic kidney disease by 67% as compared to type II DM (OR=0.33, 95%, CI: 0.14-0.76) (Figure 8).
Family history of kidney disease and chronic kidney disease among diabetic patients
Two studies were selected to observe effect of family history of kidney disease on chronic kidney disease among diabetic patients [18, 28]. Both studies showed no statistically significant association between family history of kidney disease and chronic kidney disease among diabetic patients (OR=1.06, 95% CI: 0.56, 2.00) (Figure 9).
Alcohol consumption and chronic kidney disease among diabetic patients
Four studies were included in final meta-analysis to see the effect of alcohol consumption on chronic kidney disease among diabetic patients [17, 28, 31, 32]of which, a single study showed significance association between alcohol consumption and chronic kidney disease [31]. The pooled result showed that there was no significance effect of consuming alcohol on developing kidney disease (OR=0.98, 95% CI: 0.47, 2.06) (Figure 10).
BMI and chronic kidney disease among diabetic patients.
To identify the association between BMI and chronic kidney disease, four studies were selected for meta-analysis [17, 19, 28, 31]. One study showed statistically significant association [31] and three studies revealed no significant association between BMI and chronic kidney disease. The pooled finding uncovered no significant association between BMI and chronic kidney disease among diabetic patients (OR=1.70, 95% CI: 0.70, 4.14) (Figure 11).
Duration of DM and chronic kidney disease among diabetic patients
Three studies were identified to see the effect of duration of patient stayed with DM on developing chronic kidney disease [17, 18, 28]. The pooled finding figures out that the duration of the patients stayed with DM were significantly associated with development of chronic kidney disease among diabetic patients. Being diabetic patients for less than 10 years decrease the odds of developing chronic kidney disease by 49% as compared to diabetic patients who stayed with DM for 10 years (OR=0.51, 95%, CI: 0.34-0.77) (Figure 12).