Our study highlighted how the war and siege of Tigray, Ethiopia in 2020–2022 affected children with diabetes and their care. Our study showed an increased number of total diabetes dmission during the war and siege as compared prewar. This could be because of the fact that only 27.5% of hospitals and 17.5% of health centers in the entire Tigray region were functional and the majority of those were dealing with limited functionality and lack of medical supplies including insulin3. Hence, many patients from various parts of Tigray were forced to go to Ayder Comprehensive Specialized Hospital which was the only relatively functional hospital with some supplies. In addition, Mekelle city was also a center for many internally displaced people19 resulting in more potential patients visiting Ayder Comprehensive Specialized Hospital.
A higher percentage of children with diabetes were also admitted due to DKA during the war and siege than before the start of the war. Infection like pneumonia and UTI was the commonest precipitating factor of DKA in overall admissions of our study but the infection rate was higher in those patients admitted during the war and siege (35% vs 46%). This may be due to the collapse of infection prevention services, decreased access to water and sanitation materials, and inadequate nutrition (3). In addition, during the war and siege, our study identified several significant precipitating factors for DKA which resulted from the collapse of the health care system and inadequate access to food & medicines.
Our study also revealed a significantly higher percentage of patients who develop complications of DKA such as cerebral edema, renal failure, and hypokalemia during the war siege than prewar. The DKA related mortality in our study was double compared to a study done in Ayder Comprehensive Specialized Hospital in 2019 (pre-war)20. Delayed presentation to health care as result of disrupted transportation, lack of security, damage to diabetic care services in other parts of Tigray, and lack of access to food and clean water during the war and siege period in the region contributed to negative outcomes in our patients. Similar to our study, the systematic review conducted by Gesesew H et al. highlighted destruction of health system followed by medical and humanitarian siege; resulting to interruption of health and social services, shortage of food and medicine and increased magnitude of illness.3
The percentage of admitted diabetic children with malnutrition was also alarmingly high in our study during the war and siege. Although we had a high percentage of patients with type-1 DM who were malnourished even before the war (25%) the percentage increased by 2.5-fold during war and siege (62%). The high food insecurity in Tigray due to the war and siege also contributed to raise the rate of malnutrition in children. For example, World Food Program reported that 91% of the population in Tigray required emergency assistance2, 5. Gesesew H et al. also reported fateful humanitarian crises including the use of food, medicine as a tool of war during the Tigray war.3 A case report from a Cameroon refugee camp also supported our finding demonstrating insecurity increased the rate of severe acute malnutrition among patient with type-1 DM14. In contrast to our study, a study from Yemen reported that 74.7% followed diabetic diet and 98% of the patients didn't change their dietary habits11.
Whilst our study attempted to demonstrate the impact of the war and siege on children with diabetes and their care, it has its own limitations. First, the study was conducted in a single hospital which affects the generalizability of the findings to the entire Tigray. That said, given that Ayder Comprehensive Specialized Hospital was the most well equipped, we feel outcomes elsewhere could even be worse. Second, the sample size was small where we were not able to apply inferential statistics (regression analysis). This deters from reporting independent predictors affecting the main diabetic outcomes among children in our study. Third, data were extracted using chart review only and were not collected prospectively, which limited the list of variables to the available records. Fourth, the demographic characteristics of the patients studied prewar versus during the war & siege are not the same (including gender and age), and making statistical adjustments is difficult due to the small sample size. That said, the reason the demographics are different might also be indicative of the selective effect of the war and siege on who goes to the hospital and who doesn’t.