Our study revealed the prevalence of diabetic ketoacidosis among admitted DM patients is low 29 (16.3%) which is similar to the studies in Saud Arabia (18–22%) (22), and in Nigeria (12.2%) (23). But the result is different from previous studies done in Ethiopia (40% at Hawassa university comprehensive specialized hospital (24), 73.9% at Jimma Medical Center, Southwest Ethiopia(25), 48.3% at public hospitals in northwest Ethiopia(26)), 46.7% in 12 different Arab countries(27) and 48.7% at a district-level public hospital in Cape Town (28). The variation might be due to the outbreak of COVID-19 in that patients were at home and always follow their treatment.
In this study, DKA prevalence is more common in type 2 DM patients (45%) than in type 1 DM (13%). This finding is different from the previous studies done in Hawassa university comprehensive specialized hospital (24) the prevalence of DKA was 28.7% in type 1 DM and 11.28% in type 2 DM and also in Colorado American university, 25–30% in type 1 and 4–29% in type 2 DM (29).
Our study also revealed that the prevalence of DKA is higher in males 24 (83%). This result is in line with previous studies done in Bahrain there was a higher prevalence of DKA in males 131 (58.5%) than in females 93 (41.1%)(2) and Hawassa that DKA was more common in males. In this study, the prevalence of DKA in newly diagnosed DM and known DM was 41% and 59% respectively. This current finding is in line with studies conducted in Debre Tabor General Hospital Ethiopia, the prevalence of DKA in type one diabetes and type 2 was 43% and 57% respectively(30).
We have revealed in our study that complication of DKA has developed more in the age category 31 – 45 type 2 DM with DKA tend to be older at presentation between 46 to 60 years old, 7 (54%), than type 1 DM with age category 14 to 30 years old, 3 (75%). This result is contrary to previous studies from our country that complications of DKA are developed between the 15-34 years category(31).
Our study shows a history of missing drug 9 (31%) was the most frequent precipitating factor followed by community-acquired pneumonia 6 (21%), and more than one factor 5 (17%) which statistically significant difference between no precipitating factors, p-value 0.000 (95%CI 2.18; 6.91). An institutional-based retrospective follow-up study in other parts of Ethiopia similarly reported that drug discontinuation had high odds of developing DKA (AIRR = 2.91, 95% CI = 2.02–4.22)(32). The study conducted in a tertiary care hospital, in Mysuru, has demonstrated infections in 57 (52%) and poor compliance to antidiabetic treatment in 23 (21%) being the most common cause of DKA(33). A recent study conducted in Pakistan showed that infections (36.5%) and inadequate insulin doses (22.5%) were the predisposing factors frequently revealed(34). But a study conducted in 2020 identified different precipitating factors for DKA including socioeconomic disadvantage, adolescent age (13–25 years), female sex, high HbA1c, previous DKA, and psychiatric comorbidities(35).
We found the mean random blood sugar (480.9 ± 89.2) in DKA complication comparable in all the underlying types of DM, with a tendency to be higher in type 2 DM (492.7± 74.9). The result is consistent with the reports at a university hospital in Damascus in 2015, 478±166(36), and also comparable with the results from the study(11).
The study conducted as a prospective, observational design in a tertiary hospital from July-December 2018 revealed that DKA patients presented with severe vomiting (32.2%), abdominal pain (27.9%), and depressed mental state (DMS) (26.8%)(34). But our study on contrary reported the most common clinical features in DKA were Polysymptoms (polydipsia, polyuria) with or without weight loss (44.8%), fever, headache, and cough (17.2%), GI symptoms (vomiting, diarrhea, abdominal pain) (13.8%), and few at follow up (6.9%). Similar to our study polyuria (96.9%), polydipsia (92.4%), weight loss (16.8%), vomiting (8.0%), and abdominal pain (8.9%) were the most common symptoms on presentation of the DKA patients at the Medical ward of Shashemene Referral Hospital, Ethiopia(12). This result is also not similar to the cross-sectional study conducted at Dilla referral hospital reported dry mouth (30.2%) followed by altered mentation (27.8%) as a common presentation of DKA(37).
Missing drugs (31%), and infection (45%) commonly community-acquired pneumonia (69.2%), urinary tract infection (23.1%), and malaria (7.7%) are the most identified precipitating factors for admitted DKA in our study. This finding agreed with studies conducted at North Wollo and Waghimra zone public hospitals, Amhara region, Northern Ethiopia(38), and a retrospective study conducted at Debre Tabor General Hospital from June 1 to June 30 of 2018(31). Another study done in Pakistan from August 2019 to February 2020 found that the most common precipitating factors as infections (69.0%) and non-compliance to treatment (53.5%) which supports our study(39).
All (100%) patients admitted with DKA diagnosis were improved and discharged with a mean length of hospital stay of greater than three days than others (two days). This result is lower than a study done in Debre Tabor General Hospital, Ethiopia around five days(31), sub-Himalayan region nine days(40), and in KwaZulu-Natal in which the estimated hospital stay was around nine days(41)
Limitations of the Study
The challenge was the missing information due to poor documentation of all necessary patient data. Since the study was based on secondary data, some important risk factors like education status and patient income were not available, and unable to analyze them.