Clinical Outcome of Febrile Neutropenia and Associated Factors Among Adult Patients with Cancer Treated at Ethiopian Oncology Centers: A Retrospective Observational Study

Abstract

Introduction

Febrile neutropenia (FN) is a major cause of dose-limiting complications in cancer treatment that predisposes patients to serious infections. Despite advancements in therapies, including empirical and definitive antibiotics, FN remains a major morbidity and mortality issue among patients with cancer undergoing cancer treatment. Little is known about the 30-day all-cause mortality rates from FN in Ethiopia, particularly in the Northwest Ethiopia oncology centers.

Methods

This retrospective cross-sectional study was conducted via chart review without direct patient contact at two Northwest Ethiopia oncology centers. Adult patients diagnosed with cancer who developed FN and were treated at the two oncology centers between July 2017 and July 2021 were included in the study. Multivariable logistic regression was used to identify factors associated with 30-day all-cause mortality, with statistical significance determined at P < 0.05 and a 95% confidence interval (CI).

Results

A total of 405 patients with FN were included in the final analysis. The overall 30-day all-cause mortality was 20.7%. Age > 60 years [adjusted odds ratio (AOR) = 3.1, 95% CI (1.6–5.9), P = 0.009], low Multinational Association for Supportive Care in Cancer (MASCC) score [AOR = 4.8, 95% CI (2.5–9.1), P = 0.0001], hypoalbuminemia [AOR = 2.8, 95% CI (1.4–5.8), P = 0.026], lymphopenia [AOR = 4.9, 95% CI (2.9–6.5), P = 0.001], and elevated gamma-glutamyl transferase (GGT) [AOR = 3.5, 95% CI (1.5–4.7), P = 0.009] were significantly associated with 30-day all-cause mortality.

Conclusion

The 30-day all-cause mortality rate was high among patients with FN. Old age, low MASCC score, hypoalbuminemia, lymphopenia, and elevated GGT levels were found to be significantly associated with 30-day all-cause mortality. Healthcare providers should consider these factors in order to manage and mitigate the risks associated with the 30-day all-cause mortality. Further prospective studies are warranted to confirm our results and identify therapeutic strategies that can improve survival.

Key Summary Points

Why carry out the study?

Limited data are available regarding the clinical characteristics and treatment outcomes of patients with febrile neutropenia (FN) in Ethiopia, particularly, Northwest Ethiopia.

Adequate information on patient management practices and clinical outcome of FN could increase patient adherence to treatment and decrease both the length of hospitalization and mortality.

Providing information on the clinical outcome of FN and related complications could decrease hospitalization-related costs and enhance the quality of life and survival of patients with cancer.

What was learned from the study?

The overall 30-day all-cause mortality of patients with FN was 20.7%.

Multivariable analysis revealed that old age, low Multinational Association for Supportive Care in Cancer (MASCC) score, hypoalbuminemia, lymphopenia, and elevated gamma-glutamyl transferase levels were significantly associated with 30-day all-cause mortality.

The findings provide valuable insights on the risk factors associated with 30-day all-cause mortality in patients with FN; these insights should be considered by healthcare providers when managing this patient population as a means to mitigate these risks.

Introduction

Febrile neutropenia (FN) is a common complication of standard therapies administered to patients with cancer [1] that leads to treatment delay and dose modification. Modifications in dose intensity may affect treatment efficacy, frequently leading to hospital admissions with significant mortality due to infection-related complications [2, 3]. The incidence rates of FN vary according to patient-related risk factors, tumor type, treatment modality, and genetic susceptibility factors [4, 5]. It is estimated that FN occurs in 10–50% of patients with solid tumors and > 80% of patients with hematologic malignancies [6].

Patients with neutropenia develop ongoing active infections caused by various pathogens, mostly commonly by bacteria, viruses, and fungi, with bacteria being the predominant causal pathogen of FN [7, 8]. Fungal infections may be associated with mucositis, prolonged neutropenia, central venous catheters (CVC), high-dose steroid therapy (> 20 mg/day of prednisone for ≥ 28 days), prior antibiotic use, or numerous chemotherapeutic lines [9]. Significant costs are attributed to FN due to increased antibiotic use and unplanned hospital stays [10]. In one study carried out in the USA, patients with cancer who were neutropenic spent 3 days longer in the hospital with about $6000 additional costs compared to those admitted for other causes [11].

Patients falling into the high-risk FN categories continue to have high rates of death (9–12%) and serious complications (25–30%) [12]. Inappropriate initial antibiotic use has been associated with a much greater mortality rate [13]. A Multinational Association for Supportive Care in Cancer (MASCC) risk-index score of < 21, tachypnea, thrombocytopenia, high C-reactive protein, and prolonged neutropenia are strongly related to poor outcomes in patients with FN [14]. Age, cancer type, comorbidities, delayed antibiotic therapy, and laboratory abnormalities such as low serum albumin, anemia, and elevated lactate dehydrogenase (LDH) are risk factors for mortality from FN [15–18]. The risk of death was found to increase for those patients with multiple hospitalizations for FN [19].

To reduce mortality, healthcare professionals in different countries have undertaken a variety of different strategies and prevention measures, including FN self-management, antibiotic therapy, and granulocyte-colony-stimulating factor (G-CSF) prophylaxis, and developed guidelines [7, 20]. However, in Ethiopia, these strategies have not been well applied. In addition, the lack of both financial resources and medication supplies, presence of comorbid conditions, and medication non-adherence have complicated the management of patients with FN in Ethiopia [21, 22]. There are scarce data on treatment outcomes, including hospital stay, infection control during hospitalization, and mortality related to FN in Ethiopia, particularly in Northwest Ethiopia. In this study we determined the 30-day all-cause mortality of patients with FN in Northwest Ethiopia oncology centers. The findings of this study will contribute to the national effort on antimicrobial resistance prevention and expansion and also help to reduce mortality, shorten hospital stays, and enhance FN care.

Methods

Study Design

This study was a facility-based retrospective cross-sectional design in which data from the medical charts of patients treated at two major Northwest Ethiopia oncology centers from 15 July 2017 to 15 July 2021 were assessed. Data were collected between 30 June and 30 July 2021. Ethical approval (ethical approval code SoP/133/2021) was obtained from the ethical review committee of the Department of Clinical Pharmacy, School of Pharmacy, College of Medicine and Health Sciences, University of Gondar. Informed consent was waived by the review committee of the Department of Clinical Pharmacy, School of Pharmacy, College of Medicine and Health Sciences, University of Gondar because the study was conducted retrospectively and it was difficult to contact the patients directly. Permission to access patient medical records was obtained from the University of Gondar Comprehensive Specialized Hospital (UoGCSH) and Felege Hiwot Comprehensive Specialized Hospital (FHCSH) clinical directors. Confidentiality was maintained and sufficiently anonymized, and the study was conducted according to the Declaration of Helsinki and the International Council on Harmonization guidelines for good clinical practice.

Study Area and Settings

The study was conducted at UoGCSH and FHCSH, two of the larger comprehensive and specialized hospitals in Northwest Ethiopia. These hospitals serve more than 10 million people in the area and provide patient care and teaching services. FHCSH is in Bahir Dar city, the capital city of the Amhara regional state, which is located 565 km from Addis Ababa. UoGCSH is located in Gondar, 735 km away from Addis Ababa, the capital city of Ethiopia. These hospitals offer clinical services for cancer and other diseases. The BEZA associations of Switzerland collaborated with UoGCSH to develop the oncology clinic in January 2015 [23]. The FHCSH oncology clinic opened its doors in 2017 [24]. Currently, FHCSH and UoGCSH have 20 and 32 beds in oncology units, respectively.

Study Population

All adult patients (≥ 18 years) with FN attending the oncology clinics of UoGCSH and FHCSH during the study period who had 30-day outcomes documented and legible data in medical charts during the study period were included in the study. Patients who fulfilled the inclusion criteria were selected using the convenience sampling technique. These two hospitals treated 2824 patients with cancer between July 2017 and July 2021, of whom 451 patients developed FN. All patients with FN whose information was complete were included in the analysis. After removing patients lost to follow-up, 405 patients with FN were included in the final analysis (Fig. 1).

Fig. 1.

Flow chart diagram for selection of patients with febrile neutropenia (FN) at the two participating Northwest Ethiopia oncology centers (n = 405 patients)

Data Collection Tools and Procedures

Following a review of the relevant literature, the authors designed a structured questionnaire, which was then provided to qualified research assistants at the oncology clinics who were extracting and collecting data from the medical records, including sociodemographic, clinical, and treatment-related data [25–29]. FN-associated medical complications, including refractory hypotension, respiratory failure, renal failure, severe bleeding, altered mental status, arrhythmia, and allergic reactions to treatment, were adopted from previous studies [26, 27]. The albumin level is classified as hypoalbuminemia when its value is < 3 g/dL [30]. Age was categorized into two age categories ( ≤ 60 year and > 60 years) based on a previous study [31]. The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation 2021 was used to determine creatinine clearance, and renal failure was considered as if the estimated glomerular filtration rate (eGFR) was < 30 mL/min [7]. Lymphopenia, leukopenia, and thrombocytopenia were determined and classified according to the Common Terminology Criteria For Adverse events (CTCAE) version 5 (US National Cancer Institute [30]). Consequently, in conformity with the CTCAE, lymphopenia (cells/mm³) was categorized as grade 1 [mild; lower limit of normal value (LLN)-800/mm³], grade 2 (moderate; < 800–500/mm³), grade 3 (severe; < 500–200/mm³), grade 4 (life threatening; < 200/mm³), and grade 5 (death) [30]; thrombocytopenia (cells/mm³) was categorized as grade 1 (mild; LLN-75,000 mm³), grade 2 (moderate; < 75,000–50,000 mm³), grade 3 (severe; < 50,000–25,000/mm³), grade 4 (life threatening; < 25,000/mm³), and grade 5 (death) [30]; and leukopenia (cells/mm³) was categorized as grade 1 (mild; LLN-3000 mm³), grade 2 (moderate; < 3000–2000 mm³), grade 3 (severe; < 2000–1000/mm³), grade 4 (life-threatening; < 1000/mm³), and grade 5 (death) [30]. Elevated gamma-glutamyl transferase was defined as values > 2.5 × upper limit of the normal value as defined by CTCAE version 5 [30]. Neutropenia was defined as an absolute neutrophil count (ANC) < 500 neutrophils/mm³ [7]. Neutropenia severity was categorized based on the guideline of the Infectious Diseases Society of America as severe (100 to < 500 neutrophils/mm³) or profound (< 100 neutrophils/mm³) [7]. ANC (cells/mm³) was calculated using the following formula: ANC = [white blood cell × (neutrophil (%) + % band)]/100. The two participating Northwest Ethiopian oncology centers used the Eastern Cooperative Oncology Group Performance Status (ECOG PS) Scale to describe the functional status of patients. The ECOG PS of patients in this study was classified into bad (III–IV) and good (0–II) based on a previous study [28]. The risk of FN complications was evaluated using the MASCC score and was classified into a low MASCC score (MASCC score < 21 points) and a high MASCC score (score ≥ 21 points) [26]. The time of antibiotic administration was classified as early when patients received antibiotics within 1 h of FN onset, whereas antibiotic administration after 1 h was considered late [7]. Prolonged neutropenia is considered when neutropenia is persistent for > 7 days [7]. Body mass index (BMI) was measured as body weight in kilograms divided by the square of height measured in meters (kg/m²), and participants were classified into the categories underweight (< 18.5), normal weight (18.5–24.9), overweight (25–29.9), and obese (≥ 30) according to the World Health Organization definition [32]. The Charlson Comorbidity Index (CCI) score was calculated according to the original scoring system [33], with patients categorized into three groups: mild (CCI scores 1–2); moderate (CCI scores 3–4); and severe, (CCI scores ≥ 5).

Variables

The dependent variable was 30-day all-cause-mortality. Sociodemographic, clinical, laboratory results, and treatment-related characteristics were considered to be independent variables.

Operational Definitions

Febrile neutropenia: a single oral temperature measurement of > 38.3 °C or a temperature of 38.0 °C sustained for > 1 h, with an ANC of < 500 neutrophils/mm³ or an ANC that is expected to decrease to < 500 neutrophils/mm³ over the next 48 h [7].

Performance status: the degree of functioning of the patient as measured by their capacity for self-care, everyday activities, and physical exercise [34].

Lost to follow-up: the outcome status of patients diagnosed with FN after the start of antibiotics is unknown.

Data Analysis

The EpiData version 4.6.2 software package was used to enter the data, and the STATA version 17 software package (StatCorp LLC, College Station, TX, USA) was used for the analysis. Descriptive statistics, including frequency, percentage, mean, standard deviation (SD), median, and interquartile range (IQR) were used to summarize the patients’ sociodemographic, clinical, laboratory tests and treatment-related characteristics. The association between the independent variables and 30-day-all-cause mortality was examined using bivariable and multivariable logistic regression models. The Hosmer–Lemeshow goodness test was used to evaluate model fitness, and the model was found to be well fitted (P = 0.38). The highest variance inflation factor was < 5 after multicollinearity was examined. A P-value of 0.2 was used to select candidate variables for the multivariate logistic regression analysis. A 95% confidence interval (CI) and a P-value of 0.05 were used to determine statistical significance.

Results

Clinical and Sociodemographic Characteristics of the Patients

A total of 405 patients with FN were included in the final analysis, among whom more than three-quarters (75.3%) were female. The median age (IQR) of the patients was 50 (39–88) years. Nearly two-thirds of the patients (61.2%) had a poor performance status. More than half of the patients (51.9%) had hematologic malignancies, of which acute myeloid leukemia (AML) (13.6%) and acute lymphoblastic leukemia (ALL) (12.1%) were the most common. Among the solid malignancies identified, cervical cancer (8.1%) was the most common, followed by breast cancer (7.4%) and ovarian cancer (6.7%). Comorbidities were present in more than half of the patients (56.3%), of which psychiatric illness (28.1%) was the most common, followed by diabetes mellitus (DM) (21%) and ischemic heart disease (IHD) (14%). More than half of the patients (56.8%) had a moderate CCI scores. More than one-third of the patients (39.8%) had low MASCC scores. More than two-thirds of the patients (71.6%) had severe neutropenia (100 to < 500 neutrophils/mm³), and nearly two-thirds of the patients (61.7%) had a time to antibiotic administration (TTA) ≥ 1 h after the onset FN (Table 1).

Table 1.

Sociodemographic and clinical characteristics of patients with febrile neutropenia in the Northwest Ethiopia oncology centers

Variables	Category	Frequency, n (%)
Gender	Male	100 (24.7)
	Female	305 (75.3)
Age	Median (IQR)	50 (39–88)
	> 60	108 (26.7)
	≤ 60	297 (73.3)
Residence	Urban	94 (23.2)
	Rural	311 (76.8)
BSA (m2)	Median (IQR)	1.5 (1.4–1.6)
	≤ 2	382 (94.3)
	> 2	23 (5.7)
BMI (kg/m2)	Median (IQR)	21 (18.2–24.2)
	Normal	192 (47.4)
	Obesity	67 (16.6)
	Overweight	22 (5.4)
	Underweight	124 (30.6)
Temperature (°C) median (IQR)	38.9 (38.8–39)
ECOG Performance Status	Poor	248 (61.2)
	Good	157 (38.8)
Cancer stage	I	36 (8.9)
	II	89 (22.1)
	III	69 (17.1)
	IV	178 (44.2)
	Unknown	33 (7.7)
Solid malignancies	Breast	30 (7.4)
	Cervical	33 (8.1)
	GTN	24 (5.9)
	Ovarian	27 (6.7)
	Endometrial	23 (5.7)
	Colorectal	19 (4.7)
	Osteosarcoma	15 (3.7)
	Lung	17 (4.2)
	Pancreatic	7 (1.7)
Hematologic malignancies	AML	55 (13.6)
	ALL	49 (12.1)
	CLL	39 (9.6)
	MM	5 (1.2)
	HL	36 (8.9)
	NHL	26 (6.5)
Malignancies	Hematologic	210 (51.9)
	Solid	195 (48.1)
MASCC Score	High MASCC score	244 (60.2)
	Low MASCC score	161 (39.8)
Comorbidities	Yes	228 (56.3)
	No	177 (43.7)
ICU admission	Yes	98 (24.2)
	No	307 (75.8)
Types of comorbidities (total = 285)	DM	60 (21)
	Hypertension	29 (10.2)
	IHD	40 (14)
	Psychiatric illness	80 (28.1)
	Dyslipidemia	29 (10.1)
	Congestive heart failure	16 (5.6)
	AIDS	6 (2.1)
	Musculoskeletal	16 (5.6)
	Tuberculosis	9 (3.3)
CCI index score	Mild	35 (8.6)
	Moderate	230 (56.8)
	Severe	140 (34.6)
Treatment received	Surgery alone	137 (33.8)
	Radiotherapy alone	39 (9.6)
	Chemotherapy alone	143 (35.3)
	Chemotherapy + radiotherapy	52 (12.9)
	Chemotherapy + surgery	34 (8.4)
Neutropenia (neutrophils/mm3)	100 to < 500	290 (71.6)
	< 100	115 (28.4)
Thrombocytopenia (cell/mm3)	None	355(87.7)
	< LLN-75,000	10 (2.5)
	< 75,000-50,000	23 (5.7)
	< 50,000–25,000	17 (4.1)
Lymphopenia(cell/mm3)	None	325 (80.3)
	< LLN-800	16 (3.9)
	< 800–500	30 (7.4)
	< 500–200	34 (8.4)
Leucopenia (cell/mm3)	None	163 (40.2)
	< LLN-3000	72 (17.8)
	< 3000–2000	85 (21)
	< 2000–1000	85 (21)
Time since diagnosis (years)	Median (IQR)	2(1,3)
	< 1	146 (36)
	1–3	245 (60.5)
	> 3	14 (3.5)
Previous antibiotic use	Yes	137 (33.8)
	No	268 (66.2)
Duration of neutropenia (days)	Median (IQR)	8 (4,9)
	≥ 7	207 (51.1)
	< 7	198 (48.9)
Time from initiation of antibiotics (hour) after FN diagnosis	Median (IQR)	6(4,20)
	< 1	155 (38.3)
	≥ 1	250 (61.7)

AIDS Acquired immunodeficiency syndrome, ALL acute lymphoblastic leukemia, AML acute myelogenous leukemia, BMI body mass index, BSA body surface area, CCI Charlson Comorbidity Index, CI confidence interval, CLL chronic lymphocytic leukemia, DM diabetes mellitus, ECOG Eastern Cooperative Oncology Group, ESR erythrocyte sedimentation rate, FN febrile neutropenia, GTN gestational trophoblastic disease, HL Hodgkin lymphoma, ICU intensive care unit, IHD ischemic heart disease, IQR interquartile range, LLN lower limit of normal value, MASCC Multinational Association for Supportive Care in Cancer, MM multiple myeloma, non-Hodgkin lymphoma, WBC white blood cell

Association Between the Type of Cancer Therapy and 30-Day All-Cause Mortality

The patients with FN received different types of cancer therapy. The most commonly used cancer therapies were chemotherapy alone (143 patients; 35.3%) and surgery alone (137 patients; 33.8%). A chi-square (χ²) test showed that there was no statistical difference in the 30-day all-cause mortality among patients with FN who received different types of cancer therapies (P = 0.33) (Table 2).

Table 2.

Association between the type of cancer therapy and 30-day all-cause mortality among patients with febrile neutropenia treated in the Northwest Ethiopia oncology centers

Type of cancer therapy	Total, n (%)	Mortality, n (%)	No mortality, n (%)	P-value
Surgery alone	137 (33.8)	31 (36.9)	106 (33)	0.33
Radiotherapy alone	39 (9.6)	8 (9.5)	31 (9.7)
Chemotherapy alone	143 (35.3)	28 (33.3)	115 (35.8)
Chemotherapy + radiotherapy	52 (12.9)	14 (16.7)	38 (11.8)
Chemotherapy + surgery	34 (8.4)	3 (3.6)	31 (9.7)

Association Between Baseline Laboratory Values and 30-Day All-Cause Mortality

The association of patient laboratory values with 30-day all-cause mortality were evaluated using the Student’s t-test. The t-test revealed that there was a substantial association between the 30-day all-cause mortality and low albumin level (P = 0.0014), high GGT level (P = 0.001), and low lymphocyte count (P = 0.0001) (Table 3).

Table 3.

Association of laboratory values with 30-day all-cause mortality among patients with febrile neutropenia treated in the Northwest Ethiopia oncology centers

Laboratory tests (reference values)	Total (mean ± SD)	Mortality		P-value
		No (mean ± SD)	Yes (mean ± SD)
Hgb (12–16 g/dL)	12.3 ± 4.9	12.38 ± 5.4	12.1 ± 2.77	0.64
WBC (4.5–11 k/µL)	4.8 ± 2.6	5.995 ± 1.4	4.5 ± 2.3	0.22
PLT (150–450 k/µL)	269 ± 224	276 ± 237	243 ± 159	0.32
Lymphocytes (1–4.8 K/µL)	1.94 ± 1.2	2.1 ± 1.1	0.95 ± 0.3	0.0001*
CRP (0–0.5 mg/dL)	24.9 ± 8.6	22.7 ± 2.4	25.49 ± 1.8	0.43
ESR (2–30 mmh)	41.2 ± 3.5	40.47 ± 3.6	41.44 ± 2.9	0.079
eGFR (90–120 mL/min/1.73 m2)	93.6 ± 41.4	103.6 ± 42.4	89 ± 37.6	0.91
TP (6.4–8.3 g/dL)	6.1 ± 1.3	6.13 ± 1.28	6.3 ± 1.4	0.37
Albumin (3.5–5.5 g/dL)	4.1 ± 1.4	4.19 ± 1.3	2.8 ± 1.48	0.0014*
LDH (60–160 U/L)	896.38 ± 762.3	923.6 ± 143	792.1 ± 110	0.434
ALP (30–120 U/L)	79.1 ± 46.7	79.82 ± 45.8	76.35 ± 50.27	0.545
GGT (8–38 U/L)	39.4 ± 21.9	35.81 ± 21.5	53.2 ± 17.5	0.001*
AST (0–35 U/L)	35.5 ± 26.3	36.2 ± 26.1	32.8 ± 27	0.28
ALT 1 (0–40 U/L)	42.4 ± 34.5	41 ± 31.6	47.4 ± 43.8	0.13
Bil T (0.3–1.2 mg/dL)	1.17 ± 0.43	1.18 ± 0 .44	1.15 ± 0.43	0.58
Bil D (0.1–0.3 mg/dL)	0.81 ± 0.35	0.8 ± 0.36	0.82 ± 0.34	0.56
Na (135–145 mEq/L)	134.9 ± 8.86	135.1 ± 8 .8	134.4 ± 8.9	0.53
K (3.5–5 mEq/L)	3.89 ± 0.98	3.9 ± 0.99	3.76 ± 0 .97	0.15

ALT Alanine transferase, ALP alkaline phosphatase, AST aspartate transferase, Bil D bilirubin direct, Bil T bilirubin total, CRP C-reactive protein, eGFR estimated glomerular filtration rate, ESR erythrocyte sedimentation rate, GGT gamma-glutamyl transferase, Hgb hemoglobin, LDH lactate dehydrogenase, K potassium, Na sodium, PLT platelet count, SD standard deviation, TP total protein, WBC white blood cell count

*Significant association with 30-day all-cause mortality at P < 0.05

Results of Microorganism Culture

A total of 163 samples/specimens were collected for culture, of which 73 (44.8%) were culture positive. Of the collected specimens sent for culture, gram-positive bacteria (GPB) species 41(25.2%) were commonly identified, of which coagulase-negative Staphylococcus (CoNS)(20 samples; 12.3%) and Staphylococcus aureus (17 samples; 10.5%) were the most prevalent species. The most common gram-negative pathogens were Klebsiella pneumoniae (16 samples; 9.8%) and Escherichia coli (12 samples; 7.4%). Fungal species were rarely identified (3 samples; 1.8%) (Table 4).

Table 4.

Culture results among patients with febrile neutropenia treated in the Northwest Ethiopia oncology centers (n = 73 positive samples)

Microorganisms	Specific microorganism	Source of the specimen				Total n (%)
		Blood (n = 37)	Urine (n = 19)	Sputum (n = 10)	Swab (n = 7)
Gram-negative bacteria (n = 29/163 samples; 17.8%)	Klebsiella pneumoniae	8	4	3	1	16 (9.8)
	Escherichia coli	6	3	2	1	12 (7.4)
	Pseudomonas aeruginosa	1	0	0	0	1 (0.6)
Gram-positive bacteria (n = 41/163 samples; 25.2%)	Staphylococcus aureus CoN	12	4	3	1	20 (12.3)
	Staphylococcus aureus	9	6	1	1	17 (10.5)
	Streptococcus species	1	0	0	1	2 (1.2)
	Enterococcus species	0	1	1	0	2 (1.2)
Fungi (n = 3/163 samples; 1.8%)	Aspergillosis	0	1	0	1	2 (1.2)
	Candida	0	0	0	1	1 (0.6)

CoN Coagulase-negative

Sources of Infections and Complications

Among the 405 patients diagnosed with FN in the study, the most common infection category was clinically documented infections (CDI; 44%), followed by fever of unknown origin (FUO; 38%), and microorganism-documented infections (MDI; 18%). The most common site of infection was the respiratory tract (14.3%), followed by skin and soft tissue (13.4%). Among all complications (17.5%), respiratory failure was the most common (24.2%), followed by severe bleeding (3.9%) (Table 5).

Table 5.

Sources of infections and complications among the 405 patients with febrile neutropenia treated in the Northwest Ethiopia oncology centers

Infection category	Category	n (%)
Microorganism-documented (n = 73 patients; 18%)	Respiratory tract	58 (14.3)
	Urinary tract	8 (2)
	Gastrointestinal tract	7 (1.7)
Clinical-documented (n = 178; 44%)	Urinary tract	41 (10.1)
	Gastrointestinal tract	43 (10.6)
	Skin and soft tissue infections	54 (13.4)
	Mucositis	30 (7.4)
	Othersa	10 (2.5)
Fever of unknown origin (n = 154; 38%)		154 (38%)
Total complications (n = 71; 17.5%)	Respiratory failure	34 (8.4)
	Severe bleeding	16 (3.9)
	Arrhythmia	2 (0.5)
	Acute kidney injury	10 (2.5)
	Altered mental status	2 (0.5)
	Refractory hypotension	7 (1.7)

^aCholecystitis, abdominal abscess

Management of FN

The most frequently administered antibacterial therapy at the two oncology centers during the study was ceftazidime + vancomycin (124 patients; 30.7%), followed by cefepime + vancomycin (65; 16.1%). Combination therapy (325 patients; 80.3%) was the most common treatment type of therapeutic regimen, followed by monotherapy (80; 19.7%). The most common single antibacterial therapy was cefepime (31 patients; 7.6%). Prophylactic antibacterial agents were used in 91 (22.5%) of the patients, of whom ciprofloxacin was used in 11.1% of cases. Antifungal therapy was indicated for one-third of the patients with FN (137; 33.8%). Filgrastim was administered to more than one-half of the patients (229; 56.5%) (Table 6).

Table 6.

Management of febrile neutropenia among 405 adult patients with cancer treated in the Northwest Ethiopia oncology centers

FN management	Specific treatment	n (%)
Anti-bacterial used (n = 405)	Ceftazidime + vancomycin	124 (30.7)
	Cefepime + vancomycin	65 (16.1)
	Vancomycin + meropenem	47 (11.6)
	Ceftazidime + gentamicin	46 (11.4)
	Ceftazidime + ciprofloxacin	43 (10.6)
	Cefepime	31 (7.6%)
	Ceftazidime	22 (5.4)
	Meropenem	18 (4.4)
	Ceftriaxone	9 (2.2)
Anti-bacterial therapeutic regimen (n = 405)	Combination therapy	325 (80.3)
	Monotherapy	80 (19.7)
Antibacterial prophylaxis (n = 91) (22.5%)	Ciprofloxacin	45 (11.1)
	Levofloxacin	31 (7.7)
	Cotrimoxazole	15 (3.7)
Antifungal therapy	Yes	137 (33.8)
	No	268 (66.2)
Antiviral therapy	Yes	52 (12.5)
	No	353 (87.2)
Filgrastim use	Yes	229 (56.5)
	No	176 (43.5)

FN Febrile neutropenia

Outcomes of Patients with FN

Among the 405 patients with FN included in the final analysis, 30-day all-cause mortality occurred in 84 (20.7%) patients; the remaining 321 (79.3%) patients were alive at 31+ days. The 30-day all-cause mortality was higher among patients with a low MASCC score (15.5%) than those with a high MASCC score (5.2%). The overall median hospital stay was 11 days. Patient with a low MASCC score had a longer hospital stay (13 days) than those with high MASCC score (9 days) (Table 7).

Table 7.

Outcomes of the 405 adult patients with febrile neutropenia treated in the Northwest Ethiopia oncology centers

Outcome categories	Low MASCC score (%)	High MASCC score (%)	Total number of patients (%)
Alive at 31+ days	98 (24.2)	223 (55.1)	321 (79.3)
Mortality within 30 days	63 (15.5)	21 (5.2)	84 (20.7)
Long hospital stay (> 5 days)	173 (42.7)	131 (32.4)	304 (75.1)
Short hospital stay (≤ 5 days)	7 (1.7)	94 (23.2)	101 (24.9)
Median length of hospital stay (days)	13	9	11

MASCC Multinational Association for Supportive Care in Cancer

Factors Associated with 30-Day All-Cause Mortality

Bivariable and multivariable logistic regression analyses were conducted to identify factors associated with 30-day all-cause mortality. Following the bivariable logistic regression analysis, we performed multivariable logistic regression analysis on those variables with P-values < 0.2. In the multivariable logistic regression analysis, older age (> 60 years) was associated with a 3.1-fold increased risk of 30-day all-cause mortality compared with younger age (≤ 60 years)  [AOR = 3.1, 95% CI (1.6–5.9), P = 0.009].

A low MASCC score was associated a 4.8-fold increased risk of 30-day all-cause mortality compared with a high MASCC score [AOR = 4.8, 95% CI (2.5–9.1), P = 0.0001]. Hypoalbuminemia was associated a 2.8-fold increased risk of 30-day all-cause mortality in comparison with patients without hypoalbuminemia [AOR = 2.8, 95% CI (1.4–5.8), P = 0.026].Lymphopenia was associated a 4.9-fold increased risk of 30-day all-cause mortality compared with patients without lymphopenia [AOR = 4.9, 95% CI (2.9–6.5), P = 0.001]; and increased GGT was associated with a 3.5-fold increased risk of 30-day all-cause mortality compared with normal GGT  [AOR=3.5, 95% CI (1.5–4.7), P = 0.009] (Table 8).

Table 8.

Factors associated with the 30-day all-cause mortality of the 405 patients with febrile neutropenia treated in the Northwest Ethiopian oncology centers

Variables	Category	Mortality		Bivariable analysis		Multivariable analysis
		Yes (n = 84)	No (n = 321)	Crude odds ratio (95% CI)	P-value	Adjusted odds ratio (95% CI)	P-value
Age (years)	> 60	36	72	2.6 (1.6, 4.3)	0.001	3.1 (1.6, 5.9)	0.009*
	≤ 60	48	249	1	Ref	1	Ref
Malignancy type	HM	50	160	1.5 (0.9, 2.4)	0.11	1.8 (0.93, 3.5)	0.067
	Solid	34	161	1	Ref	1	Ref
Comorbidity	Yes	63	165	2.8 (1.6, 4.9)	0.011	0.5 (0.11, 2.5)	0.41
	No	21	156	1	Ref	1	Ref
ECOG Performance Status	Poor	56	131	2.9 (1.7, 4.8)	0.0001	0.96 (0.47, 1.9)	0.91
	Good	28	190	1	Ref	1	Ref
LOH	Longer	77	227	4.5 (2, 10)	0.0001	1.2 (0.46, 3.3)	0.67
	Shorter	7	94	1	Ref	1	Ref
ICU admission	Yes	25	73	1.4 (1.5, 4.3)	0.182	1.4 (0.73, 2.6)	0.32
	No	59	248	1	Ref	1	Ref
MASCC score	Low MASCC score	63	98	6.8 (3.9, 11.8)	0.001	4.8 (2.5, 9.1)	0.0001*
	High MASCC score	21	223	1	Ref	1	Ref
TTA	Late	66	184	2.7 (1.5, 4.8)	0.001	1.4 (0.69, 2.9)	0.33
	Early	18	137	1	Ref	1	Ref
Neutropenia severity	Profound	29	86	1.4 (0.9–2.4)	0.16	1.6 (0.8–3.3)	0.21
	Severe	55	235	1	Ref	1	Ref
Duration of the neutropenia	Long	35	172	0.6 (0.4–1.1)	0.06	1.5 (0.75, 2.7)	0.34
	Short	49	149	1	Ref	1	Ref
Hypoalbuminemia	Yes	52	140	2.1 (1.3, 3.4)	0.003	2.8 (1.4, 5.8)	0.026*
	No	32	181	1	Ref	1	Ref
Lymphopenia	Yes	40	40	6.4 (3.7–10.1)	0.0001	4.9 (2.9, 6.5)	0.001*
	No	44	281	1	Ref	1	Ref
Category of infection	MDI	22	51	2.2 (1.1, 4.3)	0.017	0.89 (0.38, 2.1)	0.81
	CDI	37	141	1.35 (0.78, 2.4)	0.29	0.98 (0.48, 2.1)	0.96
	FOU	25	129	1	Ref	1	Ref
GGT	Elevated	68	134	5.9 (3.3, 10.7)	0.0016	3.5 (1.5, 4.7)	0.009*
	Normal	16	187	1	Ref	1	Ref

CDI Clinically documented infections, CI confidence interval, ECOG Eastern Cooperative Oncology Group, FUO fever of unknown origin, GGT gamma-glutamyl transferase, HM hematologic malignancy, ICU Intensive care unit, LOH length of hospital stay, MASCC Multinational Association for Supportive Care in Cancer, MDI microorganism-documented infections, Ref reference, TTA time to antibiotic administration

*Significant difference from reference value at P < 0.05

Discussion

The aim of this study was to determine the 30-day mortality rate of FN and associated factors at two oncology centers in Northwest Ethiopia. In those cancer patients who developed FN, 18% of patients had MDI, 44% had CDI, and 38% had FUO. This result is comparable with that reported in a previous study which identified CDI and MDI in 46% and 11% of patients with FN [35], but they differ from those reported in a study in Thailand that found 44.0%, 45.1%, and 11.0% of patients with FN experienced FUO, MDI, and CDI episodes, respectively [36]. Research carried out in Turkey found that 28.5% of FN cases were MDI [37], and a study in Israel found that 67% of FN episodes were FUO [25]. These differences might be correlated with the healthcare setting, availability of laboratory setups, and differences in healthcare professional knowledge among different countries. The fewer MDI cases in our study may be due to the lack of availability of better diagnostic methods.

The availability of evidence-based therapy has found to be significant for reducing mortality and morbidity in patients with FN. Overall, the distribution of affected sites in our study was comparable with findings from other studies on FN, which includes the respiratory tract, gastrointestinal tract, skin and soft tissue, and oral cavity [7, 38]. Of the specimens sent for culture, GPB species (41 specimens; 25.2%) were commonly identified, of which CoNS was the most prevalent, which is consistent with the results of an earlier study conducted in Addis Ababa, Ethiopia [21]. The most common gram-negative pathogens in the present study were K. pneumoniae and E. coli, which is comparable with that of a study carried out in Addis Ababa reporting that the most common gram-positive pathogens were CoNS and the most common gram-negative bacteria (GNB) species were K. pneumoniae and E. coli [39]. In our study, GPB were more commonly isolated than GNB. The results of our study are consistent with those of a study conducted in Tanzania, which found that Staphylococcus aureus (60%) and CoNS (40%) were the main bacterial isolates [40], and also with those of a study conducted in India that reported the most common GNB was K. pneumonia, followed by E.coli [41] and with a Turkish study that found GPB were the commonly isolated bacterial species [37]. In contrast, our results are contradictory to those of a number of previous studies that reported GNB were the most common isolates [25, 35, 42]. The most likely reason for these differences is that the prophylactic use of antimicrobials in our study, which inhibited the gram-negative infections, led to a resultant shift towards increased GPB species. In addition, the judicious use of intravenous lines and several cases of chemotherapy-induced mucositis in our study might have increased the frequency of gram-positive organisms, suggesting that injury to the skin and mucous membranes could have increased the risk of infections caused by the patient’s own endogenous flora. Among the isolated microorganisms, fungi isolates (1.8%) were detected rarely, which is consistent with a study conducted in India (2%) [42]. The number of fungal species in our study was lower than that in studies conducted in Addis Ababa, Ethiopia (9.1%) [39] and Turkey (15%) [37]. This difference might be due to the difference in the use of fungal prophylaxis, duration of neutropenia, and availability of diagnostic algorism among different settings [43].

In our study, combination therapy (80.3%) was the most common treatment type, with ceftazidime + vancomycin (30.7%) being the most frequently administered combination antibacterial therapy, followed by cefepime + vancomycin (16.1%). Monotherapy was administered to 19.7% of patients, with cefeprime being the most common single therapy (7.6%). This result is comparable with those of a previous study conducted in Addis Ababa [21]. The mortality rate of patients with FN appears to be lower when early and proper antimicrobial therapy using broad-spectrum antibiotics is administered [44, 45]. When FN is treated empirically, an antibiotic that is active against Pseudomonas as well as effective against Enterobacteriaceae that produce extended-spectrum beta-lactamases may be the primary option [25]. For the treatment of patients with FN, the Infectious Diseases Society of America (IDSA) has developed clinical practice guidelines that advocate for the timely delivery of empirical broad-spectrum antibiotic therapy [7]. The extended-spectrum cephalosporins (cefepime or ceftazidime), carbapenems (imipenem or meropenem), and the antipseudomonal penicillin (piperacillin–tazobactam or ticarcillin–clavulanate) are the mainstay of empiric monotherapy or combination therapy among patients with FN [7].

In our study, prophylactic antibacterial agents were administered to 22.5% of patients, of whom 11.1% received ciprofloxacin. A study conducted in Palestine found that levofloxacin was the most commonly used prophylactic agent for bacterial prophylaxis among patients with FN [46]. Other studies have shown that fluoroquinolone prophylaxis effectively reduced the incidence of gram-negative infections [47, 48]. A meta-analysis study also showed that antibiotic prophylaxis reduced the mortality of patients with neutropenia receiving cytotoxic chemotherapy [49]. The IDSA guidelines issued in 2010 recommend that antibiotic prophylaxis should be considered in all patients at high risk of FN [7].

A G-CSF, filgrastim, was indicated in more than half of the patients in our study (56.5%), suggesting some overuse, as current guidelines mostly recommend its administration before the establishment of complications in patients at high risk [3, 7, 50]. However, recent guideline recommendations indicate that G-CSF should be considered among patients with fever and neutropenia who are at high risk for infection-associated complications or who have prognostic factors predictive of poor clinical outcome [50]. In one study, the administration of G-CSF with antibiotics significantly decreased the length of hospital stay and promoted faster recovery from the neutropenic phase [51].

We found that the overall 30-day all-cause mortality of FN was 20.7%, which is comparable with those reported in a study conducted in Texas (17.3%) [52], in Italy (20.3%) [53], and in Brazil (24.5%) [54]. The overall 30-day all-cause mortality of FN in the present study is higher than that reported in studies conducted in Canada (3.8%) [55], in the USA (9.5%) [12], in Iran (5.3%) [29], in Thailand (14%) [56], and in Korea (12%) [57]. The low mortality rate in these countries may be due to better medical and pharmaceutical care, as well as better education and awareness of patients regarding FN treatment. The 30-day all-cause mortality in our study is also higher than that reported in a study of patients with FN in Israel (15.8%) [25]. Of the 405 patients with FN is our study, 39.8% had low MASCC scores, which increases the risk of death; in comparison, a study in Israel reported that 35% of patients with FN had low MASCC scores. Another possible reason for a higher 30-day all-cause mortality of patients with FN in our study might be that nearly two-thirds of our study participants (61.2%) had a poor performance status, which is a main risk factor of all-cause mortality.

A chi-square test showed that there was no statistical difference in the 30-day all-cause mortality among patients with FN who received different types of cancer therapy. Patients who are neutropenic and undergoing surgery, especially emergency surgery, may experience greater rates of morbidity and mortality than individuals who are not neutropenic [58]. The risk of surgical site infections and other infections is increased when neutropenia is present [59]. Chemotherapy and radiation therapy can increase the odds of developing FN, a dangerous side effect of cancer treatment, and they may also indirectly increase the death rate of patients with FN. Although these therapies may not directly contribute to FN mortality, they can weaken a patient’s immune system, resulting in them being more vulnerable to infections, which can result in life-threatening complications and, in some circumstances, death [60]. In addition, the incidence of FN varies due to different risk factors, such as patient-related risk factors, disease-related risk factors, and genetics-related risk factors [5]. Of these risk factors, nearly two-thirds of our study participants (61.2%) had a poor performance status, which is main risk factor of FN [5]. More than three-quarters of the study participants (75.3%) were female, another risk factor of FN [5]. Laboratory abnormalities, including low serum albumin and low lymphocyte count were also present in the patients in our study; these two laboratory parameters are significant factors for the incidence of FN and FN-related hospitalizations [5]. Therefore, the incidence of FN and FN-related hospitalization and associated mortality is not only due to cancer therapy but also might be due to patient-related, disease-related, and genetics-related risk factors.

A study conducted in Denmark among patients admitted to the intensive care unit (ICU) reported that the mortality rate of FN increased by up to 38% following ICU admission [61], possibly because the odds of dying is significantly high in these patients. In comparison, among our  study population, the proportion of patients admitted to the ICU was smaller (24.2%). A study conducted in Turkey found that the mortality rate of patients with FN admitted to the Emergency Department was 32% [27], which is also higher than that in our study, possibly due to 36% of the patients having bacteremia and 43% having a low MASCC scores, which are known risk factors for mortality. In addition, 52.5% of the patients with FN in the Turkish study developed complications, whereas a smaller percentage of patients in our study had complications (17.5%). Low MASCC score, various complications, and bacteremia are known risk factors associated with mortality among patients with FN [27]. In a study carried out in Thailand, the overall mortality rate among patients with FN was 28% [62]; this higher mortality rate may be attributed to the fact that only patients with hematologic malignancies were included in the study, with most of the patients included having acute leukemia. Patients with acute leukemia may have bone marrow failure and receive intensive chemotherapy. Therefore, patients with prolonged neutropenia are at a high risk of FN. In contrast, our study included patients with both hematologic and solid malignancies. The mortality rate in our study was lower than that in a study conducted in Addis Ababa, Ethiopia (34.1%) [39], possibly due to the higher number of invasive aspergillosis cases in the Addis Ababa study, which is a significant mortality risk factor.

The overall mortality rate in our study was lower than that in a study conducted in Pakistan (20.7% vs. 26%) [63]. One possible factor explaining this difference may be that the study conducted in Pakistan included patients with profound neutropenia, severe sepsis comorbidity, and a performance status ≥ 3. In contrast, our study included all patients regardless of neutropenia, sepsis, comorbidity, or performance status. In addition, most patients in the Pakistan study were diagnosed with GNB, in contrast to our study in which GPB were more common. GNB are usually associated with higher case fatality rates than gram-positive infections, and the risk is further increased if effective treatment is delayed due to antimicrobial resistance [64, 65].

The overall mortality rate in our study was lower than that reported in a study conducted in India (20.7% vs. 31.81%) [41], possibly because the study conducted in India reported a higher number of multidrug-resistant (MDR) (48.01%) and extensively drug-resistant (XDR) (29.76%) bacteria species among the patients. A systematic review reported that before the empirical use of antibiotics, the mortality rate for patients with FN could reach 90% [66]. A French study involving neutropenic patients with severe sepsis or septic shock reported a hospital mortality rate of 43–58.7% [67]. On the other hand, our study did not include patients with shock, which increases the mortality rate. In addition, Pseudomonas aeruginosa (25%) was the predominant microorganism isolated from the culture in another French study, and this bacterium was most lethal after the first blood culture was performed [68].

In our study, we found that old age, low MASCC scores, hypoalbuminemia, lymphopenia, and elevated GGT level were significant factors of 30-day all-cause mortality. Patients aged > 60 years were more vulnerable to 30-day all-cause mortality than those in the younger age group. The role of advanced age as a predictor of mortality in our patients is consistent with the results of a study conducted in Iran [29]. One possible explanation is that older patients are more vulnerable to the adverse effects of cancer therapy, such as toxicity and/or infection and complications. An additional explanation may be a certain reluctance on the part of physicians to treat elderly patients as aggressively as younger patients [69]. In addition, advanced age is a well-known risk factor for nosocomial infection, which is an independent predictor of 30-day mortality, and the rate of colonization and subsequent infection by resistant bacteria may be higher in older hospitalized patients [70].

A low MASCC score was significantly associated with the 30-day all-cause mortality in our study population. It is well established that a MASCC scores < 21 is associated with a high risk of mortality [3]. This finding in the present study is consistent with a study conducted in India which found that a MASCC score < 21 was significantly associated with mortality [41]. Upon admission, more than one-third of patients fit well with a low MASCC scores (39.8%), showing a morality rate of 15.5%. It is possible that a low MASCC scores may lead to different complications, such as bacteremia and invasive fungal infections, which are known risk factors for increased mortality among patients with FN [27]. Upon admission, MASCC scores stratification is an important tool for patient monitoring and better patient management.

We found a significant association between the 30-day all-cause mortality and hypoalbuminemia. This result is in agreement with that of another study that found hypoalbuminemia to be associated with a higher likelihood of mortality and serious sequelae [25, 71]. A common metric for evaluating the nutritional status of patients with cancer is serum albumin. Low serum albumin levels can indicate disease severity and are a reliable indicator of prognosis. Malnutrition and inflammation also inhibit the synthesis of albumin [72, 73]. Hypoalbuminemia is not merely a sign of malnutrition, but it also represents a complicated interaction between the inflammatory state, hepatic function, and glomerular filtration [74]. According to the authors of a Serbian study, patients with FN who have low serum albumin levels are more likely to die or experience serious consequences [71]. Stratification according to serum albumin levels may promote more intensive monitoring and early intervention for high-risk patients. Our results are also consistent with those of another study reporting that albumin was a useful predictor of mortality among hospitalized patients and that low albumin level was associated with a high mortality rate among patients with sepsis [75]. A study carried out in Israel found that low serum albumin level at admission was associated with long-term mortality, while normalization of albumin levels prior to discharge corresponded to a lower mortality rate than hypoalbuminemia prior to discharge [76].

In our study, lymphopenia was a significant factor for 30-day all-cause mortality, which is consistent with results from previous studies [25, 77]. According to earlier research, early lymphopenia following cytotoxic chemotherapy is a risk factor for FN and may indicate susceptibility to the hematologic toxicity of chemotherapeutic drugs since they cause lymphopenia before they cause neutropenia [25, 77]. The possibility that lymphocytes contribute to the recovery of normal hematopoiesis following cytotoxic treatment is another explanation. Normal neutrophil counts cannot be restored because a reduction in lymphocyte counts decreases cytokine production [31].

High GGT level was significantly associated with 30-day all-cause mortality. The findings of earlier studies [25, 52] showed that a correlation between abnormal liver enzyme levels and 30-day all-cause mortality following FN are in agreement with those found in the present study. The enzyme GGT facilitates glutathione metabolism, which is essential for antioxidant activities. Higher serum GGT was found to be an independent predictor of death at all time periods and of both overall and disease-specific mortality, with a dose–response relationship in the general population [78]. The underlying mechanism and clinical implications of GGT prediction warrant further investigation.

Strength and Limitations

This study has several strengths. It was conducted in two large comprehensive specialized hospitals with oncology centers, with a relatively large sample size. To the best of our knowledge, this is the first study which identified variables associated with the 30-day all-cause mortality in patients with FN and its management in Northwest Ethiopia oncology centers.

Despite these advantages, there are also limitations. Although this was a multicenter study, it was a retrospective study. The number of positive blood cultures was low, and it was difficult to assess the association between the appropriateness of FN management and 30-day all-cause mortality.

Conclusion

Patients with FN had a significant 30-day all-cause mortality rate. Our study found that one in five patients with FN died. Combination antibiotic therapy was the common empiric treatment approach for FN, with ceftazidime + vancomycin being the most frequently used therapy. The most common monotherapeutic regimen was cefepime. We found that 30-day all-cause mortality was associated with older age, low MASCC score, hypoalbuminemia, lymphopenia, and elevated GGT levels. Healthcare professionals should consider these factors in order to manage and mitigate the risks associated with the 30-day all-cause mortality. Further prospective studies are warranted to confirm our results and identify therapeutic strategies that can improve survival.

Author Contributions

Samuel Agegnew Wondm: conceptualization, data curation, formal analysis, investigation, methodology, project administration, supervision, visualization, writing—original draft, writing—review, and editing. Getachew Yitayew Tarekegn: formal analysis, investigation, project administration, visualization, writing, review, and editing. Fisseha Nigussie Dagnew: methodology, visualization, and writing—original draft, writing—review, and editing. Samuel Berihun Dagnew: formal analysis, project administration, writing—original draft, writing—review, and editing. Tilaye Arega Moges: methodology, supervision, software, and writing—original draft. Tirsit Ketsela Zeleke: formal analysis, investigation, project administration, writing—review, and editing. Mekdes Kiflu: investigation, methodology, supervision, writing—original draft, writing—review, and editing. Wubetu Yihunie Belay: formal analysis, methodology, supervision, writing—original draft, writing—review, and editing. Bantayehu Addis Tegegne: methodology, project administration, visualization, writing—review, and editing. Ashenafi Kibret Sendekie: formal analysis, project administration, visualization, writing—review, and editing. Eyayaw Ashete Belachew: methodology, project administration, visualization, writing—review, and editing. Fasil Bayafers Tamene: formal analysis, methodology, project administration, visualization, writing, review, and editing. All authors have read and approved the final version of the manuscript.

Funding

The author(s) received no financial support for the research, authorship, or publication of this article.

Data Availability

The datasets used in this study are available from the corresponding author upon reasonable request.

Declarations

Conflict of Interest

Samuel Agegnew Wondm, Getachew Yitayew Tarekegn, Fisseha Nigussie Dagnew, Samuel Berihun Dagnew, Tilaye Arega Moges, Tirsit Ketsela Zeleke, Mekdes Kiflu, Wubetu Yihunie Belay, Bantayehu Addis Tegegne, Ashenafi Kibret Sendekie, Eyayaw Ashete Belachew, and Fasil Bayafers Tamene have nothing to disclose.

Ethical Approval

Ethical approval (ethical approval eode SoP/133/2021) was obtained from the ethical review committee of the Department of Clinical Pharmacy, School of Pharmacy, College of Medicine and Health Sciences, University of Gondar. Informed consent was waived by the review committee of the Department of Clinical Pharmacy, School of Pharmacy, College of Medicine and Health Sciences, University of Gondar because the study was conducted retrospectively and it was difficult to contact the patients directly. Permission to access patient medical records was obtained from UoGCSH and FHCSH clinical directors. Confidentiality was maintained and sufficiently anonymized, and the study was conducted according to the Declaration of Helsinki and the International Council on Harmonization guidelines for good clinical practice.