Impaired glomerular filtration rate, high grade albuminuria and associated factors among adult patients admitted to tertiary Hospital in Ethiopia

The study was conducted at Jimma University Medical Center (JUMC) between November 1, 2016 and April 30, 2017. JUMC is one of the oldest public hospitals in the Ethiopia located 352 km southwest of Addis Ababa. Currently, it is the only teaching and referral hospital in the southwestern part of the country, providing services for about 15 million peoples. Approximately 17,000 inpatients, 218,000 outpatient visits, 14,000 emergency cases and 6000 deliveries in a year coming to the hospital.

Institutional cross-sectional study was conducted to recruit adult patients admitted to Jimma University Medical Center during the study period. However, only adults with at least one test result for Renal Function Test (RFT) and urinalysis were included in this study. Patients with possibility of functional proteinuria, and those with established pre-eclampsia/eclampsia were excluded from study.

The sample size was calculated using a formula for estimation of single population proportion taking prevalence of impaired eGFR and albuminuria to be 50%, margin of error of 5%, and using 95% confidence level. Finally, considering the expected patient loss from the study as 10%, the final sample size was calculated to be 422. Consecutive sampling method was used to recruit those study participants.

Data were collected from patients and their record using structured questionnaire which was subdivided into four parts; socio-demographic characteristics, medical history, physical examination and laboratory findings, was developed in English with modification from Screening and Early Evaluation of Kidney Disease (SEEK) study [17]. All laboratory tests were done following the standard procedures recommended by the manufacturer. Serum creatinine was determined using fully automated machine (Horiba Ltd., Japan) and ABX Pentra Creatinine 120 CP reagent using Jaffe’s reaction by laboratory technicians. Quality control test was done for all tests and results were used for analysis only for those who passed the test. Moreover, all laboratory tests at the hospital are validated continuously by external quality assurance done by Ethiopian Public Health Institute. Data was collected by trained nurses and medical interns with supervision by internal medicine resident from Jimma University Medical Center. During preparatory stage, the questionnaires was carefully designed and pre-tested on 5% of study population to minimize errors. Based on the result of pretest, revision was made on few contents and concepts of the questionnaire. Data collectors and supervisors were trained for two days prior to data collection period. The collected data was checked for consistency and completeness immediately at the end of the interview.

Acute kidney injury was defined in the study as an increase in serum creatinine (SCr) by > = 0.3 mg/dl within 48 h; and/or an increase in SCr to > = 1.5 times baseline, which is known or presumed to have occurred within the prior 7 days; and/or urine volume < 0.5 ml/kg/h for 6 h. Chronic kidney disease was defined as abnormal serum creatinine, eGFR and/or proteinuria which persisted for > 3 months or eGFR < 60 ml/min in patients with ultrasonographic finding of shrunken kidney and/or normocytic normochromic anemia. On the other hand, alcohol use problem was defined as drinking alcohol with ‘yes’ response at least to one of the CAGE assessment question. Smoking status was measured as a person who had used at least 100 cigarettes during their lifetime and who at the time they participated in the study, reported smoking every day or some of the day. Elevated serum creatinine was measured as serum creatinine level > 1.01 mg/dl for female and > 1.25 mg/dl for male using NHANES-III cut off point for blacks and end stage renal disease/chronic kidney failure as eGFR< 15 ml/min using steady state serum creatinine [18]. Impaired estimated glomerular filtration rate (eGFR) was measured in our study as eGFR less than 60 ml/min using one of the three creatinine based equations; Cockcroft-Gault (CG), Modification of Diet in Renal Disease (MDRD-4) or Chronic Kidney Disease Epidemiology (CKD-EPI). High grade albuminuria was also measured as dipstick proteinuria > = 1+ excluding functional proteinuria. Khat chewing (Catha edulis) an evergreen plant that is extensively cultivated in the highlands of Ethiopia and surrounding countries, was also measured. A patient was considered a chewer if s/he respond ‘yes’ to the question ‘Did you ever chew khat?”

Collected data was checked for completeness by principal investigator. Serum creatinine based glomerular filtration rate was estimated for all study participants by CG, MDRD-4 and CKD-EPI equations using QxMD calculator with correction for black race. Urinalysis of all study subjects were revised; after excluding possible causes of functional albuminuria, dipstick proteinuria of > = + 1 was taken as albuminuria. Finally, the data was entered in to the computer using EpiData software and after verification; it was exported to SPSS (IBM SPSS Statistics for Macintosh, Version 20.0. Armonk, NY) for analysis. Descriptive statistics like percentages, means, medians, standard deviations and ranges were used to describe findings.

A total of 422 patients were included in the study. About half of the participants (50.9%) were male. The mean age of the study participants was 45.37 (SD = 18.49); 29.4% were 60 years and older. One hundred ninety one (45.3%) couldn’t read and write while those who had completed college and above were less than 10 %. About 60% of them were rural residents and 38.4% of them were farmers (Table 1).

Majority of patients (~ 85%) have only one urine analysis which is both microscopic and dipstick examination of random urine sample. For those having urine analysis more than once, we have used the last one in our study. Majority (89.8%) of the study participants were medical patients. Of this 71.3% had non-communicable disease (NCD) as one of their major diagnosis including 12.8% patients who were diagnosed with AKI and/or CKD during their hospital stay. The rest of patients (10.2%) were surgical patients including 10 (2.4%) who were post-operation state; however, all laboratory data was determined before surgery.

Forty one percent of study participants had BP measurement above optimal at admission; of which 27% were previously diagnosed to have hypertension. Forty nine (11.6%) were not checked for hypertension previously of which 4 (8.2%) were diagnosed to have hypertension during admission. Fifty four (12%) were known diabetic patients. One hundred forty nine (35.3%) study participants had used Khat; of which about 15.9% were using it at the time of this study. About one third (33.4%) of study participants were engaged in regular physical exercise or their work involve significant physical activities. The study revealed that about 6 % (5.9%) of patients were obese and about 6 % (6.2%) of study subjects were HIV patients (Table 2 and Fig. 1).

One hundred seventeen (27.7%) of the study subjects had elevated serum creatinine level; 138 (32.7%), 81 (19.2%) and 82(19.4% had impaired estimated GFR by CG, MDRD-4 and CKD-EPI equations respectively. The study revealed that about 12 % (12.3%) of the study participants had albuminuria. Moreover, 21.7 and 32.1% of the study participants with impaired eGFR using CG and MDRD-4 equation respectively had albuminuria (Table 3).

Logistic regression analysis was conducted to see the association between dependent and independent variables. Accordingly, on bivariate analysis to look for candidate variables for final model; age (COR = 3.6; 95%CI: 2.15–5.9), sex (COR = 2.1; 95%CI: 1.4–3.16), residence (COR = 2.01;95%CI:1.30–3.1), educational status (COR = 0.46; 95% CI: 0.31–0.7), occupational status (COR = 0.55:95%CI:0.36–0.83), Khat chewing (COR = 1.7;95% CI:1.11–2.56), hypertension (COR = 2.3;95%CI:1.53–3.5), and physical exercise (COR = 1.6;95%CI:1.02–2.54) were associated with impaired eGFR using CG equations (Table 4).

After multivariable logistic regression analysis using Backward LR, old age (AOR = 2.376, 95%CI 1.38–4.1), male gender (AOR = 1.61, 95% CI:1.03–2.5), rural residence (AOR = 1.9, 95%CI:1.18–3.0), and elevated BP (hypertension) (AOR = 1.97, 95%CI:1.14–3.4) were independently associated with impaired eGFR computed using CG equation (Table 5). On the other hand, age (COR = 1.87;95%CI: 1.04–3.36), sex (COR = 2.1;95%CI:1.26–3.5), residence (COR = 2.53;95%CI: 1.45–4.42), occupation (COR = 0.564;95%CI: 0.35–0.92) and hypertension (COR = 3.6;95%CI: 2.1–6.1) were associated with impaired eGFR computed by MDRD-4 equation on bi-variate analysis (Table 6). After multivariable logistic regression analysis; male gender (AOR = 2.1, 95%CI: 1.2–3.7), rural residence (AOR = 2.95;95%CI: 1.6–5.6), doing moderate exercise (AOR = 2.3; 95%CI: 1.12–4.7), and hypertension (AOR = 2.6; 95%CI:1.4–4.9) were associated independently with impaired eGFR using MDRD-4 equation (Table 7).

On bi-variate analysis being obesity (COR = 3.04; 95%CI: 1.2–7.7), having history of known hypertension (COR = 3.2; 95%CI: 1.7–5.9), and diabetes mellitus (COR = 3.1; 95%CI: 1.5–6.1) were associated with albuminuria (Table 8). After multi-variate logistic regression analysis, we found that diabetes mellitus (AOR = 2.8, 95%CI:1.3–5.8) and hypertension (AOR = 6.303, 95%CI: 3.1–12.99) were strongly associated with high grade albuminuria (Table 9).

The study showed that the mean SCr was 1.83 mg/dl whereas the mean of eGFR by CG, MDRD-4 and CKD-EPI of study subjects were 78.47, 109.4 and 97.34 respectively. The three serum creatinine based equations performance correlate better for low eGFR (Figs. 2-3).