In our study the prevalence of Efavirenz based regimen induced central nervous system adverse effect was 182(52.8%), which is similar with other research done in Kenya and America.(16,19) But higher than Ambo referral hospital study (35.7%) The difference might be due to Ambo study include only review of patient charts, include patients who takes the drug for more than 6 months, and include patients on non efavirenz based ART regimen.
The most commonly reported CNS adverse effect in our study was vivid dreams (29.5%), confusion (23.4%), insomnia (23.3%), somnolence (21%), and headache (19.9%) but Kenyan study found that the most common CNS adverse effect was vertigo (dizziness) (25.5%), nightmares (13.6%), somnolence (drowsiness) (9.5%) and insomnia was the least (1%) this might be due to difference in socio-demographic and genetic difference among the participants.(16) But similar with the research done in Ghana, Ethiopia and Uganda (13,20,21,) at which headache and insomnia were the most prevalent neuropsychiatric adverse effect.
In our research Age, gender, economic status, marital status, employment status and number of children, were not significantly associated with the occurrence of CNS side effect in multivariate analysis. Economic status was associated with the occurrence of CNS ADRs during treatment with an efavirenz based regimen in bivariate analysis, but when carrying out multivariate analysis it became insignificant. This finding is not in support of different studies in the world. (22–26)
Most patients were at the age of 25–45 years in this study which was similar with the research done in Nigeria which identified young adults in the productive age of 25 to 45 years (78%) were those mostly infected with HIV and accessing ART. This finding was presented with a 2015 report reflected this age group as the most at risk of HIV infection. (27)
Variables with p values less than 0.3 during bivariate analysis entered into the multivariate analysis like economic status, efavirenz based regimen, stage, concomitant medical condition, alcohol use but only alcohol use and stage became significantly associated. BMI did not have association with the occurrence of CNS side effect in our study. More than half of the patient who experience the adverse effect was on BMI range (18.5–25) and, but other studies shown that BMI enhance the incidence of the CNS adverse effect.(12,28,16)
As TDF + 3TC + EFV serves as the first line regimen in Ethiopia 89.6% of patients who experienced CNS adverse effect took this regimen and remaining took AZT + 3TC + EFV(10.4%), which showed no association with the occurrence of CNS adverse effect which is similar with Kenyan research. (16) But different from Ambo referral hospital study in Ethiopia and Nigerian study. (12, 23) The difference might be sample size and the method difference, because the data only comes from chart review, so it is more susceptible to bias.(12)
In our research only alcohol use and stage of disease became significantly associated with the occurrence of CNS adverse effect. stage I RVI (p = 0.001, AOR = 0.006) which means stage I RVI patients were 99.4% less likely to experience CNS adverse effect than stage IV RVI patient. Stage II RVI (p = 0.001, AOR = 0.017) can be interpreted as 98.3% less likely to experience CNS side effect than stage IV patients. Stage III RVI (P = 0.01, AOR = 4.13) which showed patients in stage III experienced 4.13 times more CNS side effect than stage IV patients. the least reaction was seen in stage1 which was in line with what was found in Bahr Dare Felege- Hiwot referral hospital at which The risks of ADRs in WHO clinical stage II, III and IV were much higher than in Stage 1. (29) Most ADR were occurred in stage1 and 2 according to the research done in Nigeria the difference might be due to hence some clinical side effect may overlap with symptoms of HIV/AIDS in symptomatic patients (WHO stage 3 or 4) and might be missed, potentially resulting in under-reporting of adverse drug reactions in symptomatic patients. (30)
In this study tuberclosis (17.3%) is the leading medical condition happened prior to starting the ART and the majority of the patient didn’t have prior medical condition (82.7%). There is no statistically significant asociation of the occurance of CNS adverse effect with prior medical condition but Nigerian research shown high proportion of medical condition is accounted by tuerclosis then followed by meningitis, and pnemonia had association with the occurrence of side effect. The difference observed was due to ahigh occurance of infectious disease in Nigeria comapared to Ethiopia. (3)
Analysis of socio-behavioral characteristics among the respondents revealed that a higher proportion (39.9%) used alcohol and few participants smoked cigarettes (8.1%) whereas no one used narcotic drugs. Smoking didn’t have significant effect on the occurrence of CNS adverse effect but alcohol use significantly associated with the occurrence of CNS adverse effect (p = 0.004 AOR = 4.450) which means patients who drink alcohol had 4.45 times more likely to encounter CNS adverse effect than patients who do not drink alcohol.
Limitation of the study
Patient may experience recall bias and being cross sectional study was difficult to confirm cause and effect of the CNS adverse effect and Efavirenz and, additionally it is single center study, so generalization for other area may be difficult.