Magnitude and associated factors of anemia among AZT based HAART experienced adult HIV patients at University of Gondar Comprehensive Specialized Referral Hospital, Northwest, Ethiopia, 2019: a retrospective cohort study

Background Anemia is the most common hematologic abnormalities in AIDS patients usually associated with disease progression and poor clinical outcomes. Zidovudine (AZT), which is one of the nucleoside reverse transcriptase inhibitor drug families of the first line antiretroviral therapy regimen for HIV/AIDS patients, causes anemia due to early long-term of higher-dose therapy. This study was aimed to assess the magnitude and associated factors of anemia among AZT containing HAART experienced adult HIV/ADIS patients at University of Gondar Comprehensive Specialized Referral Hospital, northwest, Ethiopia, 2019. Methods A retrospective cohort study was conducted among a total of 320 adult AZT based HAART experienced HIV/AIDS patients from January 2016 to December 2018. Systematic random sampling technique was used to select the patients’ charts. All required data for this study were extracted from patients’ medical charts. Data were coded, cleared and entered into Epi Info version 3.5.3, and transformed to SPSS version 20 for analysis. Descriptive statistics, bivariable and multivariable logistic regression models were fitted to identify associated factors of anemia and P-value < 0.05 was considered as statistically significance. Results A total of 320 adult AZT based HAART experienced HIV/AIDS patients’ charts were assessed. Of the total patients, 198 (61.9%) were females and 133 (41.6%) were within the age range of 35–45 years. More than half, 237(76.9%) of the patients were from the urban area and 186 (58.1%) were on WHO clinical stage III at the baseline. The prevalence of anemia was 50% (95% CI 44.7–55.0%), 44.1% (95% CI 38.4–50.0%), 35.6% (95% CI 30.3–40.6%), 40% (95% CI 34.4–45.6%), 40.6% (95% CI 35.0–46.3) and 39.1% (95% CI 33.4–44.1%) at baseline, 6 months, 12 months, 18 months, 24 months and 30 months of follow-up period, respectively. The overall prevalence of anemia was 41.6%. Anemia had significant association with WHO clinical stage and base line Hgb values. Conclusions A significant number of participants were anemic in this study. WHO clinical stage and baseline Hgb value were the contributing factors for anemia among these patients. Therefore, anemia needs an immediate intervention on associated factor to improve the anemic status and living condition of HIV patient. Supplementary Information The online version contains supplementary material available at 10.1186/s12879-021-06712-5.


Background
The utilization of combined antiretroviral therapy (ART) was useful to alleviate the burden of human immunodeficiency virus (HIV) pandemic in resource restricted settings including Ethiopia. Combination ART decreases the mortality rate, progression of HIV to AIDS, expedient infections, and hospitalization due to HIV viral suppression [1]. In Ethiopia, many thousands of ART indigent individuals living with HIV are registered in ART because of the improved coverage of free antiretroviral medical aid program [2]. Per the revised 2018 national comprehensive HIV care and treatment guideline, the sequencing choices for most popular first, second and third-line ART regimens for adults and adolescents are: TDF + 3TC + DTG or TDF + 3TC + EFV, AZT + 3TC + ATV/r or LPV/r, and DRV/r + ABC + 3TC + EFVor NVP, respectively. Whereas for those pregnant, breastfeeding and childbearing age women, AZT + 3TC + EFV/ NVP, TDF + 3TC + ATV/r or LPV/r, and DRV/r + DTG + TDF + 3TC are the most popular first, second and third-line ART regimens, respectively. Whereas TDF + 3TC + EFV/NVP, AZT + 3TC + ATV/r or LPV/r and DRV/r + DTG + TDF + 3TC are the second alternatives for first, second and thirdline ART regimens, respectively. The last option is ABC + 3TC + EFV/NVP, AZT + 3TC + ATV/r or LPV/r and DRV/r + ABC + 3TC + EFVor NVP, correspondingly [3]. However, failure of ART and toxicity became the alarming drawback of long-run ART users [4].
Human immunodeficiency virus (HIV) infection is related to profound hematological abnormalities within which anemia being the key one often discovered among these patients [5]. Anemia has been estimated to vary from 30 to 95% among these patients [6,7]. In the early 1990s, the prevalence of anemia in HIV-positive patients reached 90%, then with the widespread adoption of highly active antiretroviral therapy (HAART) it became decreased to approximately 46% [8].
Anemia is a major health problem globally, which can cause a vicious impediment on quality of life, mortality, morbidity, and socioeconomic progress of individuals especially in the developing countries [9]. Anemia has been associated with disease progression and poor clinical outcomes in HIV/AIDS patients [10].
Even though the problem of anemia was no longer thoroughly understood, it become expected that approxi-mately60 to 80% of HIV/AIDS individuals sufferers of their late-degree of the sickness had been affected globally. Of these patients, around 22% is considered to be associated with the treatment regimens including ART provided to the patients. Moreover, it had been observed that females have been strangely more tormented by the situation in comparison to males. For example, in the US up to 71% extra incidence of anemia had been recorded among women than men [11].
The prevalence of anemia became increasing with disease advancement. For example, about 28% of HIV infected individuals in the pre AIDS stage of the disease develop anemia, whereas it could reach as high as 71% in the advanced or AIDS stage of the HIV disease [12]. On the other hand, anemia ranges from 23 to 58% in adult HIV patients on ART [13].
Anemia is a widely spread hematological abnormality in HIV/AIDS patients in developing countries. For example, 54% [14] and 60.6% [15] of HIV infected individuals were anemic in Egypt and Nigeria, respectively. Likewise, large segments of HIV infected patients have been suffering from anemia and anemia related complications in Ethiopia. For example, in Addis Ababa, 52.6% of HIV infected patients on ART were anemic in 2012 [16].
The causes of anemia in HIV infected individuals are multifactorial in origin. Infiltrations of the bone marrow by neoplasm or infection, use of myelosuppressive medications, decrease production of endogenous erythropoietin, hemolysis due to RBC autoantibodies and HIV infection itself are some of them [17][18][19][20][21][22][23].
On the other hand, nutritional deficiencies; most commonly, deficiencies in iron, folic acid, or vitamin B12 are also be causes of anemia. In patients with HIV disease, folic acid deficiency is generally caused by either dietary deficiency or jejunal pathology [23]. Vitamin B12 deficiency may result from malabsorption in the ileum or from gastric pathology caused by an array of infections or other conditions that affect the gastric mucosa in HIV-infected patients [24]. HIV medications contains zidovudine (AZT) and other drugs are considered as the causes of anemia in HIV/AIDS patients [8,25].
Zidovudine (AZT) is one of the nucleoside reverse transcriptase inhibitor drug families used together with other ART drugs to slow down the progression of HIV infection. Due to delay progression of infection, the immune system become improving which is characterized by an increase in helper lymphocyte cell count [26].
However, AZT is known to be associated with life threatening hematological toxicity like anemia due to early long-term of higher-dose therapy. AZT also causes bone marrow suppression, which causes anemia. Bone marrow toxicity appears to be more common in those patients with advanced disease and related to dose and duration of the treatment. All of these conditions were generally found to be reversible upon reduction of AZT dosages. They have been attributed to several possible causes, including transient depletion of mitochondrial DNA, sensitivity of the DNA polymerase in some cell mitochondria [27]. The depletion of thymidine triphosphate, oxidative stress, and reduction of intracellular l-carnitine or apoptosis of the muscle cells [28].
AZT is associated with a risk of hematological toxicity, and measuring hemoglobin is recommended before initiating ART and monthly at least for the next three months. Avoid use of AZT as first-line therapy for people with HIV if there is severe anemia at baseline (hemoglobin < 7.0 g/dl) [3]. AZT is known to be associated with myelotoxity that often presents with anemia. However, the magnitude of anemia and its associated risk factors among HIV patients on AZT based HAART are not well document in Ethiopia. Therefore, this study aimed to determine the prevalence and associated factors of anemia among adult HIV/AIDS infected patients on AZT containing HAART at the University of Gondar Comprehensive Specialized Referral Hospital, northwest Ethiopia.

Study design and setting
A retrospective cohort study was conducted at the University of Gondar specialized referral hospital ART clinic. The hospital is located in Gondar town, Central Gondar Zone, Amhara regional state, Ethiopia. Gondar town is located 727 km away from Addis Ababa, the capital city of Ethiopia and 180 km far from Bahir Dar city, the capital city of Amhara regional state in northwest direction. The town is situated at a latitude and longitude of 12°36'N 37°28'E with an elevation of 2133 m above sea level [29]. Currently, the hospital has been serving people from Central, North and Western Gondar Zone as well as the surrounding districts of Benshangul Region. Around 5 million dwellers of the North, West and Central Gondar Zone, and those of the neighboring district of Northwest Ethiopia have been seek medical services, including HIV care and treatment in this hospital. As of 2018, there had been almost 10,000 HIV/AIDS patients with ongoing follow-up in the ART clinic of the hospital, of which 5600 had been on combination AZT based HAART. This study was performed from February 01 to May 30, 2019.

Study population and sampling
A total of 320 AZT based HAART experienced adults HIV/AIDS patients' medical charts had been selected by a systematic random sampling approach and all of the essential data were collected. The single population proportion formula was used to determine the sample size by considering the formerly reported 25% prevalence of anemia [30], with 95% confidence interval (CI) and 5% margin of error. All adult HIV infected patients greater than or equal to 18 years old who have been taking AZT based HAART drugs in the clinic from January 2016 to December 2018 have been taken into consideration for the study population. Totally 5600 AZT based HAART experienced HIV/AIDS patients have been served in the ART clinic during the course of the study period. The participants' medical charts were selected at approximate sampling interval of 18 (5600/320) from their sequence of ART clinic visit. The first participant of the study was chosen by lottery method from the 18 order of adult HIV/AIDS patients. Afterwards, participants were enrolled in the study at every 18th interval till the overall sample size become achieved.

Inclusion and exclusion criteria
All adult HIV/AIDS patients aged greater than or equal to 18 years who were taking AZT based HAART drugs in the hospital from January, 2016 to December, 2018 were included in the study. Patients with incomplete record at base line and follow up, pregnant patients, patients with chronic kidney disease, rheumatic arthritis, cancer, tuberculosis, liver disease, diabetics, cardiovascular disease, burns and gastrointestinal diseases were excluded from the study.

Definitions of terms
The study outcome variables were defined based on World Health Organization (WHO) criteria [31]: anemia in men was defined as an adjusted Hgb levels lower than 13 g/dL (11-12.9 g/dL, 8-10.9 g/dl and < 8 g/dL for mild, moderate and severe anemia, respectively) and while it was 12 g/dL (10-11.9 g/dL, 8-10.9 g/dL and < 8 g/dL for mild, moderate and severe anemia, respectively) for adult non-pregnant women.

Data collection tool and process
First patients' charts were identified and retrieved using the patient registration number found in the registration log book. Then appropriate data extraction format (check list) was developed based on the objective of the study and pretested. All the available data on the patient's charts had been identified and those valuable once for the study had been collected by three nurses. The pretested data collection format contains sociodemographic data and clinical information such as unique ART numbers, date when treatment started, baseline CD-4 count, WHO clinical stage, ART drug combination, body mass index (BMI), opportunistic infections (OPIs), baseline and current Hgb levels, and AZT based HAART regimen.

Data quality control and management
The limpidity and completeness of the data collection format was pre-tested on randomly selected patients' charts before the actual data collection had been started. Corrections and modification were done on the data collection format based on available data and review of previous articles. Completed data collection tools were checked regularly for completeness of information. Intensive training had been given to the data collectors regarding the objective of the study and how to extract or collect data from patients' charts. Additionally, data collection was closely supervised.

Data processing and analysis
After data were checked and entered in to EPI Info version 3.5.3 statistical software, verification were made and then they had been exported to Statistical Package for Social Sciences (SPSS) version 20 for analysis. Data cleaning was performed to check for frequencies, accuracy, consistencies and missed values. Descriptive statistics of all the relevant data like sociodemographic variables, baseline CD-4 count, WHO clinical stage, BMI, OPIs, baseline and current Hgb levels, and AZT based HAART regimen were done and presented by tables. A binary logistic regression analysis and Generalized Estimating Equations (GEE) model have been used to fit a repeated measures logistic regression analysis to study the effect of AZT based HAART on patients and to determine the level of association between the independent variables and anemia. All the explanatory variables were analyzed by bivariable logistic regression model by the enter method. Then, those variables with P-value < 0.2 in the bivariable logistic regression model were fitted into multivariable logistic regression model and analysed by the backward LR to identify independently associated factors with anemia. The strength of statistical association was measured by adjusted odds ratios (AOR) and 95% confidence intervals (CI). P-value < 0.05 was considered as statistical significance. The incident outcome-survival analysis approach using a Kaplan-Meier curve was used to show time to anemia in the cohort, as well as by important factors such as WHO stage at baseline, binary and adjusted Cox regression model.

Baseline characteristics of the study participants
A total of 320 adult AZT based HARRT experienced HIV/AIDS patients' charts were assessed and all the valuable data were collected for this retrospective study. The majority, 198 (61.9%) of the study participants were females. Around 133 (41.6%) were within the age range of 35-45 years. Most, 237 (76.9%) of the patients were urban inhabitants and 186 (58.1%) of the patients were at WHO clinical stage III but only 66 (20.1%) were underweight. All most all of the patients, 297 (92.8%) had CD4 count greater than 200 cell/mm 3 . The mean Hgb value was 12.3 g/dl with 1.7 g/dl standard deviation (SD) whereas the mean CD4 cell count was 348.2 (SD = 223.9) cells/mm 3 (Table 1).

Factors associated with anemia prevalence
Generalized Estimating Equations (GEE) model have been used to fit a repeated measures logistic regression to study effects of AZT based HAART on adult HIV/AIDS patients. Anemia status was taken as a dependent variable and sex, residence, educational status, baseline WHO clinical stage, social drug use, OPI, ART regimen combination, age category, baseline BMI, baseline CD4 count, and duration of AZT treatment were taken as independent or intercept model (Table 5 and Additional file 1).
Univariable logistic regression analysis was done to determine the association of independent variables with the dependent variable at the end of the follow up period. Sex, age, residence, OPIs, ART regimen combination, CD4 count, social drug use, educational status and WHO clinical stage were analyzed. However, only sex, social drug use, WHO clinical stage and presence of baseline anemia have a p-value of less than or equals to 0.2 in univariable binary logistic regression analysis. These variables have been fitted to the final multivariable logistic regression model, but only sex, ART regimen combination and time of visit were remained statistically significant factors with a P-value of less than 0.05 (Additional file 2: Table S1).

Incident anemia
Among a total of 320 patients assessed, 160 (50.0%) were free of anemia at the baseline of AZT based HAART  (Table 6).
Our finding was also higher than similar study findings conducted abroad Ethiopia such as Cambodia 11.8% [36], Banaras Hindu University 16.2% [37], Odisha, India, 14.6% [38] and Indonesia, 29.6% [39]. The reason for this difference might be due to variation in the study population with respect to their clinical stage, sociodemographic factor and difference in their life style of study population [30] or it could be due to differences in their study design short duration follow up of their study.
However, the prevalence of anemia in our study was lower than in Goma, Democratic Republic of Congo, 69% [40] and Gujarat, western India, 75% [41]. Geographical variation, socio-economic difference, life style and living standards, knowledge and attitude towards the disease might be the reason for the discrepancy of prevalence of anemia among the study population. Also the sample size used for the research might be the possible reason for the difference in the prevalence of anemia. Additionally, the general magnitude/prevalence of HIV in Ethiopia is lower than other settings in East Africa or low and middle income countries globally [42]. Therefore, it could be the reason for the lower prevalence of anemia among the study participants in our study. The other possible explanation for the anemia prevalence difference might be due to the improved coverage (73%) of free ART program that helped hundreds of thousands of ART needy people living with HIV in Ethiopia enrolled in the HAART treatment program [43].
Furthermore, our result showed that microcytic anemia followed by macrocytic anemia was found to be a common type of anemia. The general characteristic of anemia manifested in this study was dimorphic population of RBCs. This might be due to the deficiencies of   iron and folic acid caused by either dietary deficiency or jejunal pathology [23] or due to malabsorption of vitamin B12 deficiency in the ileum or from gastric pathology caused by an array of infections or other conditions that affect the gastric mucosa in HIV-infected patients [24]. The microcytic anemia which is the most common type of anemia observed in this study might be due to anemia of chronic disease with low iron levels, iron deficiency, intestinal malabsorption, as well as poor intake [44]. Normocytic anemia in HIV infection and observed in a mixed nutritional problem of iron deficiency and folate or vitamin B12 deficiency, bone marrow infiltrative disease or opportunistic infection of the marrow by MAC, a destructive process such as hemolytic anemia, suppression of marrow by virus or therapy [44].
The other possible cause of the raised prevalence of macrocytic polychromic anemia might be due to the effect of drug therapy which is the common cause of megaloblastic or macrocytic anemia in HIV infected patients. Vitamin B12 and folate deficiency and liver disease also cause macrocytic anemia [44].
The 21 (16.8%) patients had an elevated RDW value in which commonly observed in HIV infected individual. Peripheral RBC changes such as anisocytosis, poikilocytosis, rouleaux formation, increased background staining are commonly observed in such patients [44]. WHO clinical stage has significant association with anemia prevalence in this study. WHO AIDS clinical stage II and III/IV were 1.82 times and 2.16 times more likely to be anemic than stage I, respectively. This was in line with a study conducted in Ethiopia [1], Bali Indonesia. This might be due to the presence of secondary infections [39]. Additionally, low baseline Hgb level was found the risk factor for developing anemia. Those anemic individuals at the baseline were 4 times more likely to develop anemia than the counterparts. Baseline anemia decreased survival, and increased disease progression in patients with HIV/AIDS. This was supported by different retrospective studies [32,45]. The fact that anemia is a reduced level of Hgb, it is not surprising to see patients with lower value of baseline Hgb level developed anemia. Low level of Hgb supported by the additive effect of AZT for anemia due to suppression of bone marrow precursor cells [34]. AZT is associated with myelotoxity that often presents with anemia. Intolerance to AZT induced gastrointestinal problems, are important barriers to adherence unless appropriate measures are taken. Anaemia and neutropaenia are the major types of toxicity of AZT use particularly in those patients with baseline anaemia or neutropaenia, CD4 count ≤ 200 cells/mm 3 . Myopathy, lipoatrophy or lipodystrophy, Lactic acidosis or severe hepatomegaly with steatosis are additional types of AZT toxicity among those with the following risk factors; BMI > 25 (or body weight > 75 kg), and prolonged exposure to nucleoside analogues. Additionally, AZT is associated with a risk of hematological toxicity, and measuring hemoglobin is recommended before initiating ART and monthly at least for the next three months. Therefore, patients with HIV/ADIS with severe anemia at baseline should start with the other more safer types (preferred) of first-line regimens [(TDF + 3TC + DTG (FDC) or (TDF + 3TC + EFV (FDC)] [3]. The adverse event that has been occurred in greater than 40% of our study participants was anemia. Therefore, switching to other safer ART drugs like combination is very crucial [46]. Male study participants were more at high risk of developing anemia than female participants. Males were more than 1.5 times more likely to develop anemia than females. This was in agreement with some studies [1]. However, it was different from many previous studies that reported a high prevalence of anemia in females compared to male HIV infected patients were significantly associated with anemia.
HIV/AIDS patients on AZT based HAART treatment experience the highest burden of anemia at the early phases of the therapy. At the first 6 months of therapy, they were 1.6 times more at risk of anemia than the other subsequent phases. This was supported by the Central Statistics of the Ethiopia [47] and Ferede G. et al. [48]. These two studies confirmed that, anemia mostly occur within 2-48 weeks of HAART treatment initiation, especially in the first 24 weeks.
AZT drug combination was another factor found associated with the prevalence of anemia in this study. Those patients on treatment of AZT + 3TC + NVP combination were 5.9 times at risk of developing anemia than the others. Although AZT + 3TC + NVP was the recommended HAART combination used in most of the hospitals, it was in contrast to our finding and the study from Ethiopia [1], Uganda [49] and Thailand [25].
In this retrospective cohort study, 106 of 160 AZT containing HAART experienced HIV adult patients developed anemia within the observation period of two and half years (30 months). Incidence of anemia this period of study was around 663 per 1000 patients, or an average of 265 per 1000 patients per year. This finding was lower than study conducted in St. Paul's and ALERT hospital Ethiopia, 353 per 1000 person years [50].
The imitation of this study was lack of relevant information like nutritional and socio economic status of the study participant since the information were collected from medical records retrospectively. Additionally, all patients on ART were not included in the research due to the resource limitations.