Prevalence and factors associated with concomitant bacteremia among adults admitted with severe malaria at Kayunga Regional Referral Hospital, Uganda

Abstract

Background

Malaria-bacteremia co-infection significantly increases mortality and the risk of ICU admission. Diagnostic overlap with bacterial infections often results in misdiagnosis, impacting outcomes. While pediatric data exists, adult studies in Uganda are limited. This study aimed to determine the prevalence, bacterial isolates, and associated factors of bacteremia in adults with severe malaria at Kayunga Regional Referral Hospital.

Methods

A cross-sectional study enrolled 207 adults with severe malaria. Blood samples were cultured, and isolates tested for antimicrobial susceptibility. Sociodemographic, clinical, and laboratory data were collected using structured tools. Logistic regression in SPSS version 26 was done to determine the significant factors. The outcome predicted was the presence of bacterium. P < 0.05 was considered significant.

Results

Of the 207 participants, 14.5% had bacteremia. Central nervous system (CNS) symptoms, low peripheral oxygen saturation (SPO2), hyperparasitaemia, and leucocytosis were significantly associated with bacteremia. Salmonella typhi (33.3%), Staph aureus (30%), and Streptococcus spp. (16.7%) were the most common isolates. Ciprofloxacin and penicillin derivatives showed strong coverage.

Conclusion

The prevalence of bacteremia among patients with malaria was high, seen in over one of every seven patients with malaria. Malaria patients with CNS symptoms, low peripheral oxygen saturation, malaria hyperparasitaemia and leucocytosis should be considered to be at high risk for bacteremia. If bacteria co-infection is suspected among patients with malaria, in the absence of culture and sensitivity results, a combination of ciprofloxacin and a penicillin can be considered since these two can provide an acceptable cover of the most common isolates, yet readily available in our resource limited setting.

Introduction

The 2024 World Malaria Report [1] highlights increasing global cases of malaria (263 million in 2023) but declining deaths (597,000 in 2023), with the WHO African Region bearing the highest burden (95% of cases/deaths), especially children under 5, while facing challenges like insecticide resistance [1]. Uganda continues to bear a high malaria burden, particularly in regions like Kayunga, where intermittent rains create ideal breeding grounds for malaria vectors [2, 3].

Severe malaria remains a major cause of hospitalization in Africa according to the 2024 World Malaria Report. While most studies in East Africa have focused on children, adult data are limited. For instance, studies in Kenya and Tanzania reported bacteraemia prevalence rates of 11.7% and 9.3%, respectively, among children with severe malaria [4, 5].

Co-infection with malaria and bacterial pathogens is associated with poor outcomes, including increased mortality and prolonged hospitalization. However, overlapping clinical features between malaria and bacterial infections, coupled with limited diagnostic capacity, often lead to underdiagnosis and mismanagement [6].

In Uganda, poor hygiene and sanitation may further elevate the risk of bacteremia in malaria patients [7]. Despite this, limited research exists on bacteremia among adults with severe malaria. This study therefore aimed to assess the prevalence, bacterial profile, and associated factors of bacteremia in adults with severe malaria at Kayunga Regional Referral Hospital.

Methods

Study design

This cross-sectional study was conducted among adults diagnosed with severe malaria at Kayunga Regional Referral Hospital (KRRH). Blood samples were collected for bacteremia testing, and bacterial isolates were identified along with their antimicrobial susceptibility pattern.

Study setting

Kayunga Regional Referral Hospital, located in Kayunga District, serves as a referral hospital for neighboring areas. It has a 254-bed capacity and multiple departments, including internal medicine. The hospital also functions as a teaching site for Kampala International University and is equipped to perform culture and sensitivity tests.

Study population

The study population consisted of adults aged ≥ 18 years who were admitted to the emergency department or inpatient medical wards with severe malaria. Screening for malaria was based on the presence of compatible clinical features, including fever or history of fever, chills, headache, jaundice, altered mental status, prostration, or signs of organ dysfunction. Malaria diagnosis was confirmed using both a histidine-rich protein 2 (HRP-2) rapid diagnostic test (Standard Diagnostics Bioline Malaria Ag P.f) and Giemsa-stained peripheral blood smear microscopy for parasite detection and quantification. Severe malaria was defined as malaria with any severity feature including presence of one or more features such as impaired consciousness, prostration, respiratory distress, shock, acute kidney injury, jaundice, severe anemia (Hb < 7 g/dL), hyperparasitemia (> 250,000 parasites/μL), or evidence of bleeding or disseminated intravascular coagulation in line with the Who criteria.

Eligibility criteria

Adults with confirmed severe malaria who consented to participate were included for participation, while patients who had taken antibiotics within the two weeks prior to presentation were excluded.

Sample size

Using OpenEpi and findings from Chau et al. [8] a sample size of 188 was calculated based on a reported risk ratio of 8.1 for bacteremia in patients with > 20% parasitemia. With an additional 10% to account for non-response, the final sample size was 207.

Sampling technique

Participants were enrolled consecutively until the target sample size was achieved.

Data collection tools

Data were collected using a structured questionnaire in both English and the local language. Sociodemographic and clinical data were extracted from medical records. Diagnostic tools included a rapid HIV test, HIV testing was performed using the Uganda National HIV Testing Algorithm, consisting of Determine™ HIV-1/2 as the screening test and STAT-PAK^® for confirmation. Uni-Gold™ HIV was used as the tie-breaker where results were discordant. Littman Class III stethoscope was used for auscultation, digital thermometer used to take temperature, and Sysmex XN-1000 hematology analyzer used for blood count. Blood for culture was drawn using sterile technique and processed in both aerobic (BD Peds Plus/F) and anaerobic (VersaTREK Redox) bottles.

Measurement of study variables

Dependent variables

Bacteraemia Presence of viable bacteria in blood, confirmed by positive blood culture.

Physiological and laboratory variables were classified using standard cut-off values. Body mass index (BMI) was categorized using WHO criteria [9]: underweight (< 18.5 kg/m²), normal (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obese (≥ 30 kg/m²) [9] Blood pressure was classified following American Heart Association 2017 guidelines [10]. Respiratory rate was categorized as tachypnea if > 20 breaths/min, and pulse rate as tachycardia if > 100 beats/min. Peripheral oxygen saturation (SpO₂) was considered low if < 94%. Glasgow Coma Scale (GCS) was considered reduced if < 15. Laboratory variables were classified using site-standard reference ranges.

For blood culture, 8–10 mL of venous blood was aseptically collected before antibiotic administration and inoculated into a single aerobic BACT/ALERT^® FA Plus bottle. The bottles were incubated in an automated BACT/ALERT^® VIRTUO Culture System for up to 5 days. The instrument automatically flagged positive bottles, which were then subjected to Gram staining, subculture on blood agar, chocolate agar, and MacConkey agar, and incubated at 35–37 °C. Organism identification was performed using standard biochemical tests and API^® identification systems. Suspected Salmonella isolates underwent serogrouping and serotyping using commercial antisera to differentiate Salmonella Typhi from non-typhoidal serovars.

Antimicrobial susceptibility testing (AST) was performed by the Kirby-Bauer disk diffusion method according to the Clinical and Laboratory Standards Institute (CLSI) M100, 2023 guidelines. Cefoxitin disc testing was used for Methicillin-resistant Staphylococcus aureus (MRSA) detection. Where azithromycin or ceftriaxone resistance was detected in Salmonella Typhi, minimum inhibitory concentration (MIC) confirmation was performed using E-test strips (bioMérieux).

Internal quality control included E. coli ATCC 25922, S. aureus ATCC 25923, and Salmonella Typhi ATCC 9992. The laboratory participates in an ongoing external quality assessment scheme supervised by the Uganda National Health Laboratory Services.

Culture significance was determined based on clinical context, organism type, and growth characteristics. Contaminants were defined as typical skin flora (e.g., coagulase-negative staphylococci, Corynebacterium spp., or Bacillus spp.) isolated in a single bottle without clinical evidence of infection and were excluded from analysis.

Bacterial profile Refers to the identified organisms and their antibiotic susceptibility.

Independent variables

Age The length of time a person has lived, typically measured in years from birth.

Sex A biological classification (male, female, or intersex) based on physical characteristics such as chromosomes and reproductive anatomy.

Level of Education The highest degree or formal schooling a person has completed (e.g., high school, bachelor’s degree, PhD).

Occupation A person’s job, profession, or primary work role (e.g., teacher, engineer, nurse).

All are key demographic variables used in research, surveys, and policy-making.

Convulsions: Sudden, uncontrolled muscle contractions (seizures) often with jerking movements and loss of consciousness.

Respiratory Distress: Difficulty breathing, characterized by rapid breathing, gasping, or labored breaths.

Jaundice: Yellowing of the skin and eyes due to high bilirubin levels (liver dysfunction or excessive red blood cell breakdown).

CNS Symptoms Symptoms affecting the central nervous system (brain/spinal cord), such as confusion, headaches, seizures, or paralysis.

Hepatomegaly Abnormal enlargement of the liver.

Splenomegaly Abnormal enlargement of the spleen.

Data quality control

A pretest of the questionnaire was conducted to ensure clarity. Interviews were conducted in the local language to reduce recall bias. All instruments used were regularly calibrated. The principal investigator reviewed completed questionnaires daily and provided training and supervision to data collectors. A physician oversaw clinical aspects.

Data management and analysis

Completed forms were verified at the point of collection. No post-collection changes were made. Data were securely stored in password-protected files. Final datasets were cleaned, coded, and analyzed using IBM SPSS version 26.0 (Armonk, NY: IBM Corp). The prevalence of bacteremia was determined by dividing the number of participants who had the growth by the total number of participants. A frequency and percentage were used to express it. Binary logistic regression was used to investigate the factors linked to bacteremia. We reported both the unadjusted (crude) odds ratios along with their respective confidence intervals (CI) and the adjusted odds ratios. In the multivariable model, a variable was considered significant if P is less than 0.05. The proportion of the most common bacterial isolates that were sensitive to specific antibiotics were calculated as fractions of sensitive organisms for a specific antibiotic over the total number of participants with the bacteremia.

Results

Characteristics of the study participants

A total of 207 adults with severe malaria were enrolled. Over half (51.7%) were aged 18–45 years, and males slightly outnumbered females (51.2% vs. 48.8%). Most participants (69.1%) had symptoms lasting more than three days before presentation. Hyperparasitaemia was observed in 38.2%, and 25.6% had leucocytosis. Comorbidities (HIV, diabetes mellitus, sickle cell disease and tuberculosis) were classified as previously diagnosed and on treatment based on documented medical records; participants without known history underwent confirmatory testing where applicable (HIV rapid test; fasting glucose for diabetes; review of laboratory and radiological records for tuberculosis and sickle cell disease). Detailed characteristics are presented in Table 1.

Table 1.

Characteristics of study participants

Characteristic	Frequency	Percentage
Age (years)
18–45	107	51.7
46–65	44	21.3
66 +	56	27.0
Sex
Male	106	51.2
Female	101	48.8
Residence
Urban	58	28.0
Rural	149	72.0
Marital status
Married	78	37.7
Single	117	56.5
Widow	12	5.8
Religion
Christian	177	85.5
Muslim	30	14.5
Education level
primary	73	35.3
Secondary	110	53.1
Tertiary	24	11.6
Occupation
Formal employment	8	3.9
Peasant	63	30.4
Business	53	25.6
Student	75	36.2
Other	8	3.9
Smoking
No	202	97.6
Yes	5	2.4
Alcohol use
No	172	83.1
Yes	35	16.9
Chronic illness
No	148	71.5
Yes	59	28.5
HIV
No	183	88.4
Yes	24	11.6
Diabates
No	187	90.3
Yes	20	9.7
Sickle cell disease
No	199	96.1
Yes	8	3.9
Tuberculosis
No	198	95.7
Yes	9	4.3
Duration of symptoms
≤ 3	64	30.9
4 +	143	69.1
CNS symptoms
No	112	54.1
Yes	95	45.9
Respiratory symptoms
No	64	30.9
Yes	143	69.1
Jaundice
No	125	60.4
Yes	82	39.6
BMI category
Normal	124	59.9
underweight	9	4.3
Overweight	67	32.4
Obese	7	3.4
Blood pressure
Normal	150	72.5
Low	12	5.8
Elevated	14	6.8
High	31	15.0
Respiratory rate
Normal	27	13.0
Tachypnea	180	87.0
Pulse rate
Normal	97	46.9
Tachycardia	110	53.1
SPO2
Low	52	25.1
Normal	155	74.9
GCS
< 15	114	55.1
Normal	93	44.9
Hepatomegaly
No	181	87.4
Yes	26	12.6
Splenomegaly
No	171	82.6
Yes	36	17.4
Hyperparasitaemia
No	128	61.8
Yes	79	38.2
Hemoglobin
Severe anemia	10	4.8
Moderate anemia	93	44.9
Mild anemia	39	18.8
Normal	65	31.4
WBC
Normal	154	74.4
Leucocytosis	53	25.6
Platelets
Thrombocytopenia	34	16.4
Normal	173	83.6

Prevalence of bacteraemia among adults presenting with severe malaria

Among the 207 patients with severe malaria enrolled into the study, only 30 had bacterial growth, showing an incidence of 14.5% with a corresponding 95% confidence interval of 9.7–19.3% as shown in Fig. 1 below.

Fig. 1.

Prevalence of bacteraemia among adults presenting with severe malaria

Factors associated with bacteraemia among adults presenting with severe malaria

Variables with p-values < 0.2 from bivariate analysis were considered for multivariable analysis. These included age category, sex, diabetes mellitus, sickle cell disease, tuberculosis, duration of illness, central nervous system (CNS) symptoms, SPO2, Glasgow Coma Scale (GCS), hyperparasitaemia, and leucocyte count. Full bivariate results are presented in Table 2 below.

Table 2.

Bivariable analysis of factors associated with bacteraemia among adults presenting with severe malaria

Characteristic	No bacteremia N = 177	Bacteremia N = 30	Bivariable analysis
			cOR	95% CI	P value
Age (years)
18–45	98 (55.4)	9 (30.0)	Ref
46–65	38 (21.5)	6 (20.0)	1.719	0.573–5.159	0.334
66 +	41 (23.2)	15 (50.0)	3.984	1.614–9.830	0.003
Sex
Male	87 (49.2)	19 (63.3)	1.787	0.804–3.972	0.154
Female	90 (50.8)	11 (36.7)	Ref
Residence
Urban	52 (29.4)	6 (20.0)	Ref
Rural	125 (70.6)	24 (80.0)	1.664	0.643–4.308	0.294
Education level
Primary	61 (34.5)	12 (40.0)	1.377	0.354–5.359	0.644
Secondary	95 (53.7)	15 (50.0)	1.105	0.293–4.165	0.882
Tertiary	21 (11.9)	3 (10.0)	Ref
Smoking
No	173 (97.7)	29 (96.7)	Ref
Yes	4 (2.3)	1 (3.3)	1.491	0.161–13.819	0.725
Alcohol use
No	146 (82.5)	26(86.7)	Ref
Yes	31 (17.5)	4(13.3)	0.725	0.236–2.225	0.573
Chronic illness
No	125(70.6)	23(76.7)	Ref
Yes	52(29.4)	7(23.3)	0.732	0.296–1.810	0.499
HIV
No	157 (88.7)	26 (86.7)	Ref
Yes	20 (11.3)	4 (13.3)	1.208	0.382–3.818	0.748
Diabates
No	162 (91.5)	25(83.3)	Ref
Yes	15 (8.5)	5(16.7)	2.160	0.722–6.465	0.169
Sickle cell disease
No	172 (97.2)	27(90.0)	Ref
Yes	5(2.8)	3(10.0)	3.822	0.863–16.921	0.077
Tuberculosis
No	171 (96.6)	27 (90.0)	Ref
Yes	6 (3.4)	3 (10.0)	3.167	0.747–13.421	0.118
Duration of symptoms
≤ 3	60 (33.9)	4(13.3)	Ref
4 +	117 (66.1)	26(86.7)	3.333	1.112–9.991	0.032
CNS symptoms
No	110 (62.1)	2 (6.7)	Ref
Yes	67 (37.9)	28 (93.3)	2.985	1.304–9.603	< 0.001
Respiratory symptoms
No	55 (31.1)	9 (30.0)	Ref
Yes	122 (68.9)	21 (70.0)	1.052	0.453–2.445	0.906
Jaundice
No	105 (59.3)	20 (66.7)	Ref
Yes	72 (40.7)	10 (33.3)	0.729	0.322–1.649	0.448
BMI category
Normal	110 (62.1)	14 (46.7)	Ref
underweight	7 (4.0)	2 (6.7)	2.245	0.424–11.889	0.342
Overweight	54 (30.5)	13 (43.3)	1.892	0.831–4.304	0.229
Obese	6 (3.4)	1 (3.3)	1.310	0.147–11.687	0.809
Blood pressure
Normal	132 (74.6)	18 (60.0)	Ref
Low	10 (5.6)	2 (6.7)	1.467	0.297–7.236	0.638
Elevated	10 (5.6)	4 (13.3)	2.933	0.832–10.339	0.294
High	25 (14.1)	6 (20.0)	1.760	0.636–4.871	0.276
Respiratory rate
Normal	22 (12.4)	5 (16.7)	Ref
Tachypnea	155 (87.6)	25(83.3)	0.710	0.246–2.046	0.526
Pulse rate
Normal	83 (46.9)	14 (46.7)	Ref
Tachycardia	94 (53.1)	16 (53.3)	1.009	0.465–2.192	0.982
SPO2
Low	38 (21.5)	14 (46.7)	3.201	1.435–7.137	0.004
Normal	139 (78.5)	16 (53.3)	Ref
GCS
< 15	89 (50.3)	25 (83.3)	4.944	1.811–13.498	0.002
Normal	88 (49.7)	5 (16.7)	Ref
Hepatomegaly
No	153 (86.4)	28 (93.3)	Ref
Yes	24 (13.6)	2 (6.7)	0.455	0.102–2.036	0.303
Splenomegaly
No	148 (83.6)	23 (76.7)	Ref
Yes	29 (16.4)	7 (23.3)	1.553	0.610–3.956	0.356
Hyperparasitaemia
No	118 (66.7)	10 (33.3)	Ref
Yes	59 (33.3)	20 (66.7)	4.000	1.760–9.090	0.001
Hemoglobin
Severe anemia	8 (4.5)	2 (6.7)	1.375	0.254–7.449	0.712
Moderate anemia	82 (46.3)	11 (36.7)	0.738	0.293–1.855	0.518
Mild anemia	32 (18.1)	7 (23.3)	1.203	0.417–3.471	0.732
Normal	55 (31.1)	10 (33.3)	Ref
WBC
Normal	138 (78.0)	16(53.3)	Ref
Leucocytosis	39 (22.0)	14(46.7)	3.096	1.390–6.894	0.006
Platelets
Thrombocytopenia	29 (16.4)	5 (16.7)	Ref
Normal	148 (83.6)	25 (83.3)	0.980	0.346–2.770	0.969

cOR Crude odds ratio, CI Confidence interval, HIV  Human immunodeficiency virus, BMI Body mass index, CNS central nervous system, SPO2 peripheral oxygen saturation, GCS Glasgow coma scale, WBC white blood count.

During multivariable analysis, the predictor determinants related to bacteremia were presence of CNS symptoms (aOR = 2.5449, CI = 1.456–6.998, P < 0.001), having low peripheral oxygen saturation (aOR = 4.389, CI = 1.591–12.106, P = 0.004), presence of malaria hyperparasitaemia (aOR = 3.816, CI = 1.166–12.494, P = 0.027) and presence of leucocytosis (aOR = 2.472, CI = 1.963–6.342, P = 0.046).

Bacteria isolates associated with bacteraemia among adults presenting with severe malaria

Among the 30 participants in whom bacteremia growth was observed, the most common organism isolated was Salmonella typhi, accounting for 33.3% of the isolates, followed by Staph aureus accounting for 30.0% and Streptococcus SPP in 16.7% (Fig. 2). Among the five streptococcal isolates recovered, two were identified as Streptococcus pneumoniae based on optochin susceptibility and bile solubility testing, one was Streptococcus pyogenes (Group A), and one was Streptococcus agalactiae (Group B) (Table 3). The remaining isolate was classified as viridans group streptococcus and considered a contaminant since it was isolated from a single culture bottle and lacked corresponding clinical features of true infection (Table 4).

Fig. 2.

Bacteria isolates associated with bacteraemia among adults presenting with severe malaria

Table 3.

Multivariable analysis of factors associated with bacteraemia

Characteristic	Multivariable aOR	95% CI	P-value
Age (years)
18–45 (Ref)	Ref
46–65	1.136	0.203–2.666	0.640
66 +	1.788	0.517–6.178	0.359
Sex
Male	1.504	0.552–4.096	0.425
Female (Ref)	Ref
Diabetes
No (Ref)	Ref
Yes	1.597	0.109–3.276	0.553
Sickle cell disease
No (Ref)	Ref
Yes	1.401	0.504–8.591	0.144
Tuberculosis
No (Ref)	Ref
Yes	1.091	0.118–8.336	0.593
Duration of symptoms
≤ 3 days (Ref)	Ref
> 3 days	2.334	0.640–8.506	0.199
CNS Symptoms
No (Ref)	Ref
Yes	2.5449	1.456–6.998	< 0.001
SPO2
Normal (Ref)	Ref
Low	4.389	1.591–12.106	0.004
GCS
Normal (Ref)	Ref
< 15	1.255	0.349–4.510	0.072
Hyperparasitaemia
No (Ref)	Ref
Yes	3.816	1.166–12.494	0.027
Leucocytosis
No (Ref)	Ref
Yes	2.472	1.963–6.342	0.046

cOR Crude Odds Ratio, aOR Adjusted Odds Ratio, CI Confidence Interval, CNS Central Nervous System, SPO2 Peripheral Oxygen Saturation, GCS Glasgow Coma Scale, WBC White Blood Cell Count.

Table 4.

Susceptibility patterns of bacterial isolates in severe malaria patients with bacteremia

Antibiotic	Salmonella typhi (N = 10)	Staph aureus (N = 9)	Streptococcus spp (N = 5)	Klebsiella pneumoniae (N = 3)	E. coli (N = 2)	Proteus vulgaris (N = 1)
Ciprofloxacin	S 90.0% I 10.0%	S 55.6% I 22.2% R 22.2%		S 66.7% I 33.3%		S 100.0%
Levofloxacin	S 90.0% I 10.0%		S 80.0% I 20.0%
Doxycycline		S 77.8% I 22.2%
Penicillin			S 80.0% I 20.0%
Imipenem				S 100.0%	S 100.0%	S 100.0%
Chloramphenicol	S 70.0% I 30.0%				S: 100.0%	S: 100.0%
Trimethoprim-sulfamethoxazole				I 33.3% R 66.7%		R 100.0%
Gentamycin		R 100.0%		I 33.3% R 66.7%	I 50.0% R 50.0%	R 100.0%
Ceftriaxone	S 70.0% R 30.0%			S 100.0%	S 50.0% I 50.0%	I 100.0%
Amikacin		I 77.8% R 22.2%		I 33.3% R 66.7%	I 50.0% R 50.0%
Vancomycin			I 100.0%
Cefepime			S 100.0%
Azithromycin	S 60.0% R 40.0%	S 77.8% I 22.2%
Clindamycin		S 77.8% I 22.2%	S 100.0%
Cefoxitin		S 88.9% I 11.1%			S 50.0% I 50.0%
Tetracycline			I 20.0% R 80.0%
Minocycline		R 100.0%
Oxacillin		I 77.8% R 22.2%
Cefalexin					S 50.0% I 50.0%
Ampicillin				I 66.7% R 33.3%	I 50.0% R 50.0%
Ceftazidime-Tazobactam				S 66.7% I 33.3%	I 50.0% R 50.0%
Erythromycin			S 80.0% I 20.0%

S (Susceptible): The bacteria are sensitive to the antibiotic and can be treated effectively.

I (Intermediate): The antibiotic may be effective at higher doses or in specific body sites.

R (Resistant): The bacteria are not inhibited or killed by the antibiotic at normal doses.

Discussion

This study aimed to determine the prevalence, bacterial profile, and factors associated with concomitant bacteremia among adults admitted with severe malaria at Kayunga Regional Referral Hospital. Our findings provide insights relevant to clinical management and align with, or differ from, previously published literature.

Prevalence of bacteremia

Among the 207 patients enrolled, bacteremia was identified in 30 individuals, yielding a prevalence of 14.5%. This is relatively high compared to the pooled prevalence of 7.6% reported in a meta-analysis by Wilairatana et al. [11], and far higher than 0.3% reported in a retrospective study among malaria patients in Sweden [11]. This discrepancy may be attributed to differences in socioeconomic factors, healthcare access, and nutritional status, as noted by Ricci [12], who emphasized the role of poverty and undernutrition in infectious disease susceptibility.

Our findings are comparable to those by Hanson et al. [13] and Nyein et al. [14], who reported bacteremia prevalence rates of 10% and 13%, respectively, among hospitalized malaria patients. The similarity may be due to comparable clinical settings and study populations. Conversely, Ukaga et al. [15] reported a much higher prevalence (35.2%), possibly due to historical differences in economic conditions and the predominance of Salmonella infections linked to poor hygiene.

Factors associated with bacteremia

Multivariable analysis identified four independent predictors of bacteremia: presence of CNS symptoms, low peripheral oxygen saturation, malaria hyperparasitemia, and leukocytosis.

CNS symptoms increased the odds of bacteremia by over 2.5 times. Donnelly et al. [16] explain that falciparum-infected erythrocytes adhere to vascular endothelium, causing microvascular occlusion, hypoxia, and GI epithelial damage that permits bacterial translocation. Our study found that 93.3% of bacteremic patients had CNS symptoms, aligning with this mechanism. Additionally, White [17] associated falciparum malaria with seizures even in uncomplicated cases.

Low oxygen saturation increased the odds of bacteremia by over four times. White [18] and Adebola et al. [19] discuss how malaria-induced anemia and respiratory complications such as aspiration and airway obstruction contribute to hypoxia. In our study, 70% of bacteremic participants had respiratory issues, with 83.3% exhibiting tachypnea.

Hyperparasitemia was significantly associated with bacteremia (OR ≈ 4). According to Takem et al. [20], hemolysis in hyperparasitemia elevates iron levels and impairs neutrophil function, promoting bacterial proliferation. Chau et al. [8] found higher rates of bacteremia in patients with parasitemia > 20%.

Leukocytosis was also associated with a twofold increase in the odds of bacteremia. Babatunde and Adenuga [21] suggest that while neutrophils combat malaria via phagocytosis and ROS production, malaria parasites can suppress antimicrobial responses, heightening susceptibility to secondary infections such as non-typhoidal Salmonella.

Bacterial profile

Salmonella typhi was the most frequently isolated organism (33.3%), followed by Staphylococcus aureus (30.0%) and Streptococcus spp. (16.7%). This is consistent with the findings of Wilairatana et al. [11], who reported similar predominant isolates in a meta-analysis of bacteremia in malaria patients.

Other studies also support our findings. For instance, Piyaphanee et al. [22] reported two cases of Salmonella bacteremia in P. vivax-infected patients in Thailand. Bhattacharya et al. [23] found that Gram-negative organisms, including S. Typhi, were prevalent among febrile patients in Kolkata. Similarly, Park et al. [24] observed high rates of Salmonella Typhimurium and S. Enteritidis in febrile African adults with malaria.

Differences in bacterial isolates have also been observed. For example, Chau et al. [8] reported varied pathogens including K. pneumoniae and H. influenzae, while Ukaga et al. [15] found a higher prevalence of Gram-negative organisms like Pseudomonas and Klebsiella. These discrepancies may result from local microbial ecology or different pathophysiological pathways, such as variations in immune response and intestinal permeability [16].

Antibiotic susceptibility patterns

Salmonella typhi showed high sensitivity to ciprofloxacin and levofloxacin. Staphylococcus aureus showed moderate susceptibility to doxycycline, azithromycin, and clindamycin but was resistant to gentamicin and minocycline. Streptococcus spp. was highly susceptible to clindamycin and cefepime but resistant to tetracycline.

These results differ from Akinyemi et al. [25], who reported reduced fluoroquinolone susceptibility among Salmonella in Lagos. Popoola et al. [26] also found high multidrug resistance in Salmonella and Staphylococcus isolates among febrile Nigerian patients, 68.5% of whom were children. Egbe and Enabulele [27] highlighted the efficacy of ceftriaxone and ceftazidime against Klebsiella spp., which was not commonly isolated in our setting. Such variations in susceptibility patterns are likely due to geographic differences and evolving antibiotic stewardship practices.

Strengths and limitations

This is the first study, to our knowledge, to assess concomitant bacteremia among adults with severe malaria in Uganda. However, its single-center design and short study period may limit the generalizability of the results.

Conclusion and recommendations

Bacteremia was present in over one in seven patients with malaria. Key associated factors included CNS symptoms, low peripheral oxygen saturation, hyperparasitaemia, and leucocytosis. Salmonella typhi, Staphylococcus aureus, and Streptococcus spp. were the most common isolates. Salmonella typhi was highly sensitive to ciprofloxacin and levofloxacin; S. aureus showed moderate sensitivity to doxycycline, azithromycin, and clindamycin but was resistant to gentamicin.

Patients with severe malaria should be assessed for bacteremia risk. Those with the above clinical features should be prioritized for empirical antibiotic treatment. In settings without culture access, ciprofloxacin combined with a penicillin may offer adequate empirical coverage for the most likely bacterial pathogens.

Author contributions

FDA, MJ, and AMH conceived and designed the study. FDA, AK, AAS, AGH, and AAY contributed to data collection and curation. MJ, AE, and TH conducted data analysis and interpretation. MA, MJ, HKB, ABM and AE were involved in drafting and critically revising the manuscript. All authors read and approved the final version of the manuscript.

Funding

This research received no specific funding from any public, commercial, or not-for-profit organizations.

Data availability

The datasets generated and analyzed during the study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethical clearance for this study was obtained from the Research Ethics Committee of Kampala International University (Ref No: KIU-2024-552). Written informed consent was obtained from all participants before enrollment.

Consent for publication

All participants provided written informed consent for the publication of this study and any accompanying data.

Competing interests

The authors declare no competing interests.