Laboratory-Confirmed Malaria and Associated Factors Among Febrile Adults Attending Screening Points in Selected IDP Camps, Mogadishu, Somalia: A Cross-Sectional Study

Abstract

Background

Malaria remains a public health concern in humanitarian settings, where displacement-related living conditions, environmental exposure, limited preventive measures, and disrupted healthcare may increase vulnerability. However, evidence on laboratory-confirmed malaria and associated factors among febrile adults in Somali IDP camps is limited. This study assessed malaria positivity and associated factors among febrile adults attending screening points in selected IDP camps in Mogadishu, Somalia.

Methods

A camp-based cross-sectional study was conducted from 15 January to 14 February 2026 among febrile adults attending screening points in six selected IDP camps in Mogadishu. Of 1761 participants enrolled, 1688 with complete questionnaire and laboratory data were analyzed. Data were collected using an interviewer-administered questionnaire, and malaria was confirmed by microscopy of stained thick and thin blood smears. Frequencies, percentages, and cross-tabulations summarized participant characteristics and malaria positivity. Associated factors were identified using modified Poisson regression with robust variance, reporting adjusted prevalence ratios (aPRs) and 95% confidence intervals (CIs). Mixed-effects sensitivity analysis accounted for camp-level clustering.

Results

Among 1688 participants, 488 tested positive for malaria (28.91%; 95% CI: 26.76–31.14%). Plasmodium falciparum accounted for 441 (90.37%) infections and P. vivax for 47 (9.63%); no mixed infections were detected. Although 60.66% reported household ITN availability, consistent use was limited. Malaria positivity was higher in Samaraac, Gulow, Gal’ad, Dulqad 2, and Xaqdhowr than in Bilan. No other sociodemographic, household, preventive, or healthcare-seeking factors were independently associated with positivity. Sensitivity analysis produced broadly similar findings.

Conclusion

Almost three in every ten febrile adults attending screening points in six selected IDP camps in Mogadishu tested positive for malaria, with P. falciparum predominating and variation observed between camps. Malaria control efforts in these selected IDP camps should prioritize improving consistent ITN use and consider the observed camp-level variation when targeting prevention activities, particularly in camps with higher malaria positivity.

Introduction

The WHO African Region accounts for the overwhelming majority of global malaria morbidity and mortality, reflecting persistent challenges related to poverty, climate variability, weak health systems, insecticide resistance, population mobility, and gaps in access to preventive, diagnostic, and treatment services.^1,2 Malaria transmission remains heterogeneous across the region because of differences in ecological conditions, health system capacity, population movement, and implementation of malaria prevention and control interventions.^2,3

Humanitarian settings represent an important context for malaria prevention and control because displacement, overcrowding, inadequate living conditions, and reduced access to vector-control and diagnostic services may increase exposure to malaria and complicate timely detection and treatment.^4,5 Population movement, disrupted healthcare access, and operational constraints may further challenge surveillance and implementation of malaria control measures.^4,5 However, the relationship between displacement and malaria risk is context-dependent and influenced by local transmission intensity, environmental conditions, population characteristics, and access to effective prevention and healthcare interventions.⁴

Somalia continues to experience malaria transmission, with substantial geographical variation influenced by climatic conditions, ecological factors, population movement, and health system capacity.^6–8 Recent global malaria estimates continue to identify Somalia as one of the malaria-endemic countries in the WHO Eastern Mediterranean Region, with transmission affected by climatic variability, humanitarian instability, and challenges in sustaining malaria prevention and control interventions.^6,7 Malaria transmission in Somalia is characterized by seasonal and spatial heterogeneity, with increased risk in areas affected by rainfall patterns, flooding, population displacement, and limited access to health services.^6,9 The country has experienced recurrent humanitarian challenges, including conflict, climate-related shocks, and population displacement, which have placed pressure on health systems and affected access to essential health services and disease surveillance capacity.^7,10 Mogadishu hosts one of the largest concentrations of internally displaced persons (IDPs) in Somalia, with many people displaced by conflict, climate-related shocks, food insecurity, and other recurrent humanitarian crises.^11,12 Evidence from Somalia indicates that IDPs experience substantial health vulnerabilities, including barriers to healthcare access, inadequate living conditions, and increased vulnerability to preventable health conditions.^11–13

Although information on malaria-control and management interventions specifically implemented within IDP camps is limited, malaria control in Somalia is guided by national strategies involving parasitological diagnosis, prompt and effective case management, vector-control interventions, insecticide-treated net (ITN) promotion, and malaria surveillance activities.⁹ These activities are complemented by community engagement approaches aimed at improving prevention practices, early care-seeking, and linkage to malaria services. However, the availability, coverage, and utilization of these interventions may vary across settings due to differences in healthcare access, health system capacity, environmental conditions, and operational challenge.¹⁴ Effective malaria control in humanitarian settings requires an integrated approach combining timely diagnosis, appropriate treatment, vector control, community engagement, and surveillance systems capable of detecting changes in transmission patterns and guiding targeted responses.^6,15

Despite the fact that malaria affects individuals across age groups, adults living in humanitarian settings may have different patterns of malaria exposure and healthcare access related to mobility, occupational activities, healthcare-seeking behavior, and availability of prevention services.^4,16 Previous malaria studies conducted in Somalia have provided important epidemiological information; however, available evidence has largely focused on specific populations, including pregnant women, hospital-based patients, or individuals with confirmed malaria infection.^17–19 Although these studies provide valuable insights into malaria epidemiology in Somalia, available evidence remains limited regarding laboratory-confirmed malaria positivity and associated factors among febrile adults living in IDP camps. To date, no study has specifically evaluated this population using community-based fever screening approaches in humanitarian displacement settings in Mogadishu. Therefore, this cross-sectional study aimed to assess laboratory-confirmed malaria and associated factors among febrile adults attending community-based screening points in six selected IDP camps in Mogadishu, Somalia.

Methods

Study Design, Period and Setting

A camp-based cross-sectional study was conducted from 15 January to 14 February 2026 among febrile adults attending community-based screening points in selected internally displaced persons (IDP) camps in Mogadishu, Somalia. Mogadishu, the capital city of Somalia, hosts the country’s largest population of internally displaced persons (IDPs). According to the October 2022 Camp Coordination and Camp Management (CCCM) Cluster IDP Site Verification Exercise, 2410 verified IDP sites were identified across the Dayniile and Kahda districts, accommodating approximately 1,347,075 individuals in 273,181 households. Of these, Dayniile contained 1302 IDP sites hosting 739,137 individuals, while Kahda contained 1108 IDP sites hosting 607,938 individuals. In this study, an IDP site refers to a location identified and inhabited by internally displaced households according to the Camp Coordination and Camp Management (CCCM) Cluster framework. The term IDP camp is used interchangeably to refer to the selected IDP sites where internally displaced households reside and where community-based screening points were established.²⁰ The selected camps (Bilan, Samaraac, and Gulow in Dayniile District; and Gal’ad, Dulqad 2, and Xaqdhowr in Kahda District) were among the larger IDP settlements in Mogadishu and were selected because of their large displaced populations and feasibility for implementing community-based screening activities.

Study Population and Eligibility Criteria

The study population comprised adults aged 18 years and above who were current residents of the selected IDP camps and who either reported a history of fever within the preceding 24–48 hours or had a measured axillary temperature of ≥37.5°C at the time of screening. Participants were recruited among adults presenting to the community-based screening points in the selected IDP camps and meeting the eligibility criteria. Adults were excluded if they had signs or symptoms suggestive of severe malaria or any other severe illness requiring immediate referral or treatment, were unable to respond to the interview, declined to provide informed consent or a blood sample for malaria testing, or had received anti-malarial medication within the preceding four weeks to reduce the potential influence of recent antimalarial treatment on parasite detection.^21,22

Sample Size Determination

The sample size was determined using Cochran’s single population proportion formula for a cross-sectional study:

where n is the required sample size, Z is the standard normal value at a 95% confidence level (1.96), p is the expected malaria positivity proportion, and d is the margin of error. Since no previous estimate of malaria positivity among febrile adults attending IDP camps in Mogadishu was available, a conservative estimate of 50% was assumed. Using a 95% confidence level (Z = 1.96), an expected malaria positivity proportion of 50% (p = 0.50), and a margin of error of 3% (d = 0.03), the initial sample size was calculated as follows:

n = (1.96² × 0.50 × (1 − 0.50)) / (0.03²)

n = (3.8416 × 0.50×0.50) / 0.0009

n = 1067 participants

Because participants were recruited from multiple IDP camps, a design effect of 1.5 was applied to account for potential clustering, resulting in an adjusted sample size of:

1067 × 1.5 = 1601 participants

Finally, a 10% allowance was added to account for potential non-response or incomplete data:

1601 × 1.10 = 1761 participants

Therefore, the final required sample size was 1761 participants.

Sampling Method and Procedure

A purposive, camp-based sampling approach with consecutive recruitment of eligible attendees was used. Dayniile and Kahda districts were purposively selected because they contain large concentrations of internally displaced persons. Within these districts, six IDP camps were purposively selected based on their large displaced populations and feasibility for implementing community-based fever screening activities. Within the selected IDP camps, temporary community-based fever screening points were established as fixed temporary locations for the study period. Community mobilizers informed camp residents about the availability of fever screening services, and adults experiencing current fever or a recent history of fever were able to voluntarily attend the screening points. At each screening point, trained data collectors screened all presenting adults for eligibility. Axillary body temperature was measured using a digital thermometer, and individuals meeting the predefined eligibility criteria were invited to participate.

Because reliable camp-specific estimates of eligible febrile adults were unavailable, the sample was not allocated proportionally according to the total resident population of each camp. Instead, consecutive sampling was employed, whereby every eligible adult presenting to a screening point during the data collection period was invited to participate until the required overall sample size was achieved. Therefore, the study population consisted of febrile adults who attended the screening points during the study period in the selected IDP camps. To prevent duplicate enrolment, participants were asked whether they had previously participated in the study. Recruitment records were also checked using each participant’s study identification number, initials, age, sex, and camp of residence. Individuals identified as having been previously enrolled were excluded from repeat participation.

Data Collection Instrument and Procedure

Data were collected using a structured interviewer-administered questionnaire and a laboratory result recording form. The questionnaire was developed based on previously published studies.^16,23–30 The questionnaire was comprised of four sections: socio-demographic characteristics, housing characteristics, malaria-related knowledge, and malaria prevention and care-seeking practices. The socio-demographic characteristics section captured sex, age group, marital status, educational status, occupational status, and IDP camp of residence. The housing characteristics section collected information on household size, duration of residence in the IDP camp, availability of screened windows, type of shelter, wall material, presence of stagnant water around the residence, and presence of damp areas in the locality. The malaria-related knowledge section assessed participants’ understanding of malaria transmission, symptoms, prevention, and treatment using structured items. The malaria prevention and care-seeking practices section assessed malaria prevention behaviors, access to malaria prevention information, insecticide-treated net availability and use, environmental practices, and healthcare-seeking behaviors. The complete questionnaire items are provided as supplementary material (Supplementary File S1. Questionnaire). The main outcome of interest was malaria test positivity, categorized as positive or negative. Independent variables included selected socio-demographic characteristics, housing characteristics, malaria-related knowledge items, and malaria prevention and care-seeking practice items.

To ensure content validity, the questionnaire was reviewed by two physicians with experience in malaria diagnosis and management and two public health professionals with experience in malaria epidemiology. The questionnaire was pretested among 88 respondents, representing approximately 5% of the calculated sample size, who were recruited from the same six selected IDP camps included in the study. The pretest assessed the clarity, relevance, and cultural appropriateness of the questionnaire items. Minor revisions were made based on feedback from the pretest to improve comprehension and contextual suitability. The final questionnaire was developed after integrating evidence from previous studies, expert feedback, and pretest findings to ensure contextual relevance and clarity before implementation. Participants involved in the pretest were not included in the final analysis. Internal consistency was assessed separately for each multi-item domain. Cronbach’s alpha coefficients were 0.856 for malaria-related knowledge and 0.895 for malaria prevention and care-seeking practices, indicating good internal consistency for both domains.³¹ The questionnaire was administered through face-to-face interviews using a paper-based format by 12 trained data collectors. After eligibility screening, informed verbal consent was obtained from participants before any study procedures. Participants who provided consent were then interviewed using the study questionnaire, followed by collection of the blood specimen for malaria testing. Each participant was assigned a unique study identification number, which was used to link questionnaire information with the corresponding laboratory result recording form. Data collection was conducted from Saturday to Thursday between 8:00 AM and 4:00 PM throughout the study period.

Operational Definitions of Key Variables

Screened Windows

The presence of windows covered with mesh, netting, or similar screening material intended to reduce mosquito entry into the household.

Damp Areas in the Locality

The self-reported presence of naturally occurring or man-made wet/damp areas near the participant’s living environment, including swamps or wet ground, drainage areas.

Availability of Insecticide-Treated Nets in the Household

The self-reported presence of at least one insecticide-treated net in the participant’s household.

Frequency of Insecticide-Treated Net Use

Self-reported frequency of sleeping under an insecticide-treated net, categorized as always, sometimes, or never.

Regular Use of Mosquito Repellents

Self-reported regular use of mosquito repellent when outdoors or during periods of mosquito exposure.

Care-Seeking for Malaria Symptoms

Self-reported history of seeking medical help or visiting a healthcare facility when experiencing symptoms suggestive of malaria, such as fever or chills.

Self-Medication for Fever or Chills

Self-reported use of medication for fever or chills without prior consultation with a qualified healthcare provider.

Educating Others About Malaria Prevention

Self-reported sharing of malaria prevention information or advice with family members, neighbors, or community members.

Blood Specimen Collection

Following the interviews, approximately 2 mL of peripheral venous blood was collected from each participant into a labeled ethylenediaminetetraacetic acid (EDTA) tube by trained medical laboratory technologists using standard aseptic procedures. The 2 mL volume was collected as a standardized EDTA venous blood specimen volume to ensure adequate sample availability for malaria film preparation, repeat examination when required, and laboratory quality assurance procedures.

Blood specimens were stored in temperature-controlled cool boxes and transported under refrigerated conditions (2°C–8°C) to the Medical Laboratory Department of Al Hayat Medical University within 2 hours of collection for laboratory analysis. Samples showing clotting, leakage, inadequate volume, contamination, or labeling discrepancies were excluded from analysis. Participant confidentiality was maintained throughout specimen collection, laboratory processing, and data analysis.

Laboratory Investigation

Microscopic examination of Giemsa-stained thick and thin blood films was used to detect Plasmodium parasites and identify the infecting species. Microscopy is a widely used reference method for malaria diagnosis and enables parasite detection and species identification.³²

Malaria Microscopy

For each participant, separate thick and thin blood films were prepared on two labeled glass slides using approximately 6 µL and 2 µL of blood, respectively. Blood films were prepared as soon as possible after specimen collection to preserve parasite morphology. The thin blood film was fixed in absolute methanol for 10–15 seconds, while the thick blood film was left unfixed. Both films were stained with 10% Giemsa stain for 10 minutes. Prepared slides were examined after staining without prolonged storage. The thick blood film was used for parasite detection, whereas the thin blood film was used for species identification, including assessment of mixed Plasmodium infections.

After staining, the blood films were examined using an Olympus CX23 light microscope (Olympus, Japan) at ×1000 magnification. Microscopists were blinded to participants’ questionnaire responses. Each set of thick and thin blood films was examined independently by two trained medical laboratory technologists, with a third senior medical laboratory technologist available to resolve any discordance regarding malaria positivity or Plasmodium species identification. A blood film was considered positive when one or more malaria parasites were detected by microscopic examination of the stained blood films. A blood film was considered negative when no malaria parasites were detected after examination of 200 oil-immersion fields. Plasmodium species identification was based on findings from the corresponding thin blood film.³³

Clinical Referral and Management

All participants were informed of their malaria test results. Participants with positive microscopy results were referred to the nearest designated health facility for clinical assessment and treatment in accordance with the national malaria treatment guidelines. Participants with signs of severe malaria or other urgent conditions were referred immediately without waiting for completion of the study procedures.

Data Processing and Analysis

Data cleaning and preparation were conducted in Microsoft Excel version 16 to ensure completeness, consistency, and accuracy, after which the cleaned dataset was exported to Stata version 17.0 (StataCorp LLC, College Station, TX, USA) for analysis. Variables were coded, categorized where appropriate, and labeled before analysis. Descriptive statistics, including frequencies and percentages, were used to summarize participant characteristics. Cross-tabulations were performed to describe the distribution of malaria positivity across socio-demographic characteristics, housing characteristics, malaria-related knowledge, and malaria prevention and care-seeking practice variables.

Malaria-related knowledge was assessed using 10 structured questionnaire items covering malaria transmission, symptoms, prevention methods, and treatment-related knowledge. Responses reflecting correct knowledge were assigned a score of 1, whereas incorrect or “don’t know” responses were assigned a score of 0. Malaria prevention and care-seeking practices were assessed using 9 structured questionnaire items. Responses considered favorable were assigned a score of 1, whereas unfavorable responses were assigned a score of 0. For items with three response options (“always”, “sometimes”, and “never”), only responses indicating consistent adherence to the recommended behavior (“always”) were scored as 1, while “sometimes” and “never” were scored as 0. Negatively worded practice items, including self-medication for malaria symptoms without consulting a qualified healthcare provider, were reverse-coded so that responses reflecting the recommended practice (ie, not self-medicating) received a score of 1. Composite scores were calculated within each domain by summing individual item scores. Participants with scores at or above the observed domain-specific mean were classified as having good knowledge or good prevention and care-seeking practices, whereas those scoring below the mean were classified as having poor knowledge or poor prevention and care-seeking practices. This empirical mean-based approach has been applied in KAP studies when validated external thresholds are unavailable and the objective is to classify participants relative to the distribution of scores within the study population.^34,35 Although internal consistency of the domains was assessed, reliability alone does not establish validity; therefore, categorized knowledge and practice scores were interpreted within the context of the study population.

Given that malaria positivity was relatively common in the study population, associations between potential predictors and malaria positivity were assessed using modified Poisson regression models with robust variance estimation to directly estimate prevalence ratios (PRs) and 95% confidence intervals (CIs). Bivariable modified Poisson regression analyses were initially conducted for individual socio-demographic, housing, malaria-related knowledge, and prevention and care-seeking practice variables. Variables with a global p-value <0.25 in bivariable analysis were considered candidates for inclusion in the multivariable model. The multivariable modified Poisson regression model was then fitted to estimate adjusted prevalence ratios (aPRs) and 95% CIs. IDP camp of residence was included in the multivariable model as a fixed effect due to the observed clustering of malaria positivity by camp. Statistical significance was defined as a p-value <0.05. As a sensitivity analysis, a mixed-effects Poisson regression model was fitted with IDP camp of residence specified as a random intercept to account for potential within-camp correlation. The consistency of estimates between the primary model and the sensitivity analysis was assessed to evaluate the robustness of the findings. Multicollinearity was assessed before model fitting, and variance inflation factors (VIFs) were examined to identify potential collinearity among included predictors. VIF values below 5 were considered acceptable, indicating absence of problematic multicollinearity.

Results

Participant Flow

A total of 1870 adults presented to the community screening points in the selected IDP camps. Of these, 109 were excluded during eligibility screening, including 84 participants with severe illness requiring urgent referral, 13 who had received antimalarial medication within the previous four weeks, and 12 who declined to participate. Consequently, 1761 eligible adults provided informed consent and were enrolled. Of these, 73 participants were excluded from the final analysis because complete questionnaire information or corresponding laboratory results required for analysis were unavailable. Specifically, 31 participants had incomplete questionnaire records, while 42 participants had unavailable laboratory results due to documented sample quality and processing issues, including clotting (11), leakage (8), contamination (6), and labelling discrepancies (17).

Socio-Demographic Characteristics and Malaria Test Results Among Febrile Adult Patients

Most participants were female, aged 25–34 years, married, had informal education, and were unemployed. Overall, malaria positivity was relatively similar by sex, although it was slightly higher among females than males (29.75% vs 26.42%). Positivity varied across age groups, with the highest proportion observed among participants aged 35–44 years (30.28%) and the lowest among those aged ≥45 years (26.45%). Positivity was also higher among participants with informal or primary education than among those with secondary or tertiary education. Across the six selected IDP camps, positivity was highest in Xaqdhowr (44.06%) and Gulow (43.28%), compared with the other camps (Table 1).

Table 1.

Socio-Demographic Characteristics and Malaria Test Results Among Febrile Adult Patients (N = 1688)

Items	Categories	Malaria Test Result		Total n (%)
		Negative, n (%)	Positive, n (%)
Sex	Female	888 (70.25)	376 (29.75)	1264 (74.88)
	Male	312 (73.58)	112 (26.42)	424 (25.12)
Age group	18–24	261 (72.50)	99 (27.50)	360 (21.33)
	25–34	485 (70.09)	207 (29.91)	692 (41.00)
	35–44	251 (69.72)	109 (30.28)	360 (21.33)
	≥45	203 (73.55)	73 (26.45)	276 (16.35)
Marital status	Single	216 (70.13)	92 (29.87)	308 (18.25)
	Married	949 (71.03)	387 (28.97)	1336 (79.15)
	Divorced	18 (75.00)	6 (25.00)	24 (1.42)
	Widowed	17 (85.00)	3 (15.00)	20 (1.18)
Educational status	Informal	922 (70.71)	382 (29.29)	1304 (77.25)
	Primary	199 (70.07)	85 (29.93)	284 (16.82)
	Secondary	60 (78.95)	16 (21.05)	76 (4.50)
	Tertiary	19 (79.17)	5 (20.83)	24 (1.42)
Occupational status	Employed	456 (70.81)	188 (29.19)	644 (38.15)
	Unemployed	744 (71.26)	300 (28.74)	1044 (61.85)
IDP camp of residence	Bilan	311 (89.37)	37 (10.63)	348 (20.62)
	Samaraac	166 (75.45)	54 (24.55)	220 (13.03)
	Gulow	135 (56.72)	103 (43.28)	238 (14.10)
	Gal’ad	199 (64.19)	111 (35.81)	310 (18.36)
	Dulqad 2	229 (80.07)	57 (19.93)	286 (16.94)
	Xaqdhowr	160 (55.94)	126 (44.06)	286 (16.94)

Notes: Percentages in the Total column are calculated using the total number of participants (N = 1688) as the denominator. Percentages in the Negative and Positive columns are row percentages calculated within each category.

Housing Characteristics and Malaria Positivity Among Febrile Adult Patients

Most participants lived in larger households, temporary shelters, and structures with plastic-tarpaulin walls, and many reported no screened windows, stagnant water around the residence, and damp areas in the locality. Malaria positivity was higher among participants living in households with four or more members than among those living in households with 1–3 members (29.21–29.37% vs 23.00%). Positivity was slightly higher among participants who had lived in the IDP camp for more than three years (29.27% vs 28.59%), those with screened windows (30.87% vs 27.64%), and those living in temporary shelters (29.32%) or structures with plastic-tarpaulin walls (29.57%). Positivity was similar among participants with and without stagnant water around the residence (28.99% vs 28.57%) and was slightly lower among those reporting damp areas in the locality (28.46% vs 29.59%) (Table 2).

Table 2.

Housing Characteristics and Malaria Positivity Among Febrile Adult Patients (N = 1688)

Items	Categories	Malaria Test Result		Total n (%)
		Negative, n (%)	Positive, n (%)
Number of people living in your household	1–3 members	77 (77.00)	23 (23.00)	100 (5.92)
	4–6 members	606 (70.79)	250 (29.21)	856 (50.71)
	7+ members	517 (70.63)	215 (29.37)	732 (43.36)
Duration of residence in the IDP camp	1–3 years	637 (71.41)	255 (28.59)	892 (52.84)
	>3 years	563 (70.73)	233 (29.27)	796 (47.16)
Availability of screened windows in the household	Yes	459 (69.13)	205 (30.87)	664 (39.34)
	No	741 (72.36)	283 (27.64)	1024 (60.66)
Type of shelter	Permanent	86 (74.14)	30 (25.86)	116 (6.87)
	Semi-permanent	215 (71.67)	85 (28.33)	300 (17.77)
	Temporary	899 (70.68)	373 (29.32)	1272 (75.36)
Material used for the wall	Cement	6 (75.00)	2 (25.00)	8 (0.47)
	Corrugated iron sheet	58 (72.50)	22 (27.50)	80 (4.74)
	Plastic Tarpaulin	955 (70.43)	401 (29.57)	1356 (80.33)
	Wood	181 (74.18)	63 (25.82)	244 (14.45)
Presence of stagnant water around the residence	Yes	980 (71.01)	400 (28.99)	1380 (81.75)
	No	220 (71.43)	88 (28.57)	308 (18.25)
Presence of damp areas in the locality	Yes	724 (71.54)	288 (28.46)	1012 (59.95)
	No	476 (70.41)	200 (29.59)	676 (40.05)

Notes: Percentages in the Total column are calculated using the total number of participants (N = 1688) as the denominator. Percentages in the Negative and Positive columns are row percentages calculated within each category.

Overall Malaria-Related Knowledge and Prevention and Care-Seeking Practices Among Febrile Adult Patients

Composite scores were calculated for malaria-related knowledge and malaria prevention and care-seeking practices. Detailed distributions of participants’ responses to the individual knowledge and malaria prevention and care-seeking practice items, together with malaria positivity according to each item, are presented in (Supplementary Tables S1 and S2).

The malaria-related knowledge mean score was 6.20 ± 2.80. Overall, 892 (52.84%) had good knowledge and 796 (47.16%) had poor knowledge. The malaria prevention and care-seeking practice mean score was 4.61 ± 2.20. Overall, 896 participants (53.08%) had good prevention and care-seeking practices, while 792 (46.92%) had poor practices.

Malaria Positivity Among Febrile Adult Patients

Overall, 488 of 1688 participants tested positive for malaria, corresponding to a malaria test positivity of 28.91% (95% CI: 26.76–31.14%). Among the positive cases, 441 (90.37%) were identified as Plasmodium falciparum and 47 (9.63%) as Plasmodium vivax. No mixed P. falciparum/P. vivax infections were identified. The remaining 1200 participants tested negative, representing 71.09% of the study population. No discordant microscopy results requiring third-reader review were identified.

Factors Associated with Malaria Positivity Among Febrile Adult Patients

Bivariable modified Poisson regression was initially conducted for 32 individual variables covering socio-demographic characteristics, housing characteristics, malaria-related knowledge, and malaria prevention and care-seeking practices. Variables with a global p-value <0.25 were considered candidates for the multivariable model. Eight variables met this criterion: sex, educational status, IDP camp of residence, wall material, availability of screened windows, household availability of insecticide-treated nets (ITNs), regular use of mosquito repellents, and ever having sought medical advice for malaria symptoms. In bivariable analysis, IDP camp of residence showed the strongest association with malaria positivity, with higher crude prevalence ratios observed among participants from all five camps compared with Bilan.

Multicollinearity was assessed before model fitting using variance inflation factors (VIFs). No evidence of problematic multicollinearity was identified among the variables included in the multivariable model (mean VIF = 1.17). A mixed-effects sensitivity analysis incorporating camp as a random intercept produced estimates for the individual covariates that were broadly similar to those from the primary model, indicating that the main findings were generally consistent after accounting for potential within-camp correlation. However, this sensitivity analysis should be interpreted cautiously because only six camps were included (Supplementary Tables S3 and S4).

After adjustment using modified Poisson regression with robust variance and including camp of residence as a fixed effect, differences in malaria positivity between camps remained evident, whereas the associations for the other candidate variables were attenuated and were not statistically significant. Composite knowledge and prevention/care-seeking practice scores were not included in the final multivariable model because their component variables were assessed individually, and inclusion of both composite scores and individual items could introduce redundancy and complicate interpretation. The final adjusted model included 1688 participants with complete data for all included variables. Compared with participants from Bilan, malaria positivity remained higher among those from Samaraac (aPR 2.33, 95% CI 1.60–3.41), Gulow (aPR 4.02, 95% CI 2.87–5.64), Gal’ad (aPR 3.34, 95% CI 2.38–4.68), Dulqad 2 (aPR 1.86, 95% CI 1.27–2.72), and Xaqdhowr (aPR 4.10, 95% CI 2.95–5.71) (Table 3).

Table 3.

Factors Associated with Malaria Positivity Among Febrile Adult Patients

Items	Categories	Malaria Negative, n (%)	Malaria Positive, n (%)	PR (95% CI)	aPR (95% CI)	p-value
Sex	Male	312 (73.58)	112 (26.42)	Ref	Ref	—
	Female	888 (70.25)	376 (29.75)	1.126 (0.941–1.348)	1.108 (0.929–1.322)	0.254
Educational status	Informal	922 (70.71)	382 (29.29)	Ref	Ref	—
	Primary	199 (70.07)	85 (29.93)	1.022 (0.839–1.244)	0.994 (0.814–1.214)	0.953
	Secondary/Tertiary	79 (79.00)	21 (21.00)	0.717 (0.486–1.058)	0.719 (0.496–1.042)	0.081
IDP camp of residence	Bilan	311 (89.37)	37 (10.63)	Ref	Ref	—
	Samaraac	166 (75.45)	54 (24.55)	2.309 (1.574–3.385)	2.332 (1.595–3.410)	<0.001
	Gulow	135 (56.72)	103 (43.28)	4.070 (2.904–5.705)	4.021 (2.868–5.639)	<0.001
	Gal’ad	199 (64.19)	111 (35.81)	3.368 (2.399–4.727)	3.339 (2.384–4.677)	<0.001
	Dulqad 2	229 (80.07)	57 (19.93)	1.875 (1.278–2.750)	1.859 (1.269–2.723)	0.001
	Xaqdhowr	160 (55.94)	126 (44.06)	4.144 (2.975–5.772)	4.101 (2.949–5.705)	<0.001
Material used for wall	Other wall material	245 (73.80)	87 (26.20)	Ref	Ref	—
	Plastic Tarpaulin	955 (70.43)	401 (29.57)	1.129 (0.925–1.376)	1.073 (0.882–1.307)	0.481
Availability of screened windows	No	741 (72.36)	283 (27.64)	Ref	Ref	—
	Yes	459 (69.13)	205 (30.87)	1.117 (0.961–1.299)	1.107 (0.956–1.283)	0.174
ITNs available in household	No	460 (69.28)	204 (30.72)	Ref	Ref	—
	Yes	740 (72.27)	284 (27.73)	0.903 (0.776–1.050)	0.908 (0.773–1.065)	0.235
Regularly use mosquito repellents	No	505 (69.37)	223 (30.63)	Ref	Ref	—
	Yes	695 (72.40)	265 (27.60)	0.901 (0.776–1.047)	0.931 (0.784–1.106)	0.417
Ever sought medical advice for malaria symptoms	No	423 (69.12)	189 (30.88)	Ref	Ref	—
	Yes	777 (72.21)	299 (27.79)	0.900 (0.772–1.048)	0.969 (0.814–1.155)	0.726

Notes: PRs are from separate bivariable modified Poisson regression models with robust variance estimation. aPRs and p-values are from the multivariable modified Poisson regression model with robust variance estimation.

Abbreviations: PR, prevalence ratio; aPR, adjusted prevalence ratio; CI, confidence interval; Ref, reference category.

Discussion

The 28.91% malaria test positivity observed among febrile adults in Mogadishu IDP camps is comparable to, although generally lower than, findings reported from other displacement settings in malaria-endemic regions. A study among patients in the Ardamata IDP camp in Sudan found a substantially higher malaria positivity of 61.1% using molecular detection methods, with P. falciparum identified as the predominant species.¹⁶ The higher positivity reported in Sudan may reflect differences in diagnostic sensitivity, transmission intensity, seasonality, and the inclusion of patients with suspected malaria from a highly endemic displacement environment. A recent systematic review and meta-analysis of malaria among displaced populations in African refugee camps reported a pooled malaria prevalence of 35.93%, with estimates ranging widely from 3.93% to 67.93%, further demonstrating the variability of malaria transmission across humanitarian settings.³⁶ More recently, a study among IDPs in Nigeria has also documented considerable malaria burden, including a malaria positivity of approximately 30.5% by rapid diagnostic testing and 42.0% by microscopy among displaced populations in Abuja.³⁷ Taken together, the malaria positivity identified in Mogadishu IDP camps falls within the range reported from other African displacement settings and reinforces the need for sustained malaria surveillance, timely diagnosis, vector-control interventions, and improved access to preventive services among displaced communities.

The predominance of P. falciparum observed in this study is consistent with the broader epidemiology of malaria in sub-Saharan Africa, where P. falciparum remains the predominant malaria parasite species associated with the highest burden of disease.³ Humanitarian crises, population displacement, limited access to health services, environmental exposure, and challenges in maintaining prevention interventions can increase vulnerability to malaria among displaced populations. However, malaria transmission intensity varies considerably between settings depending on ecological conditions, seasonality, population characteristics, vector dynamics, and the availability of prevention and treatment services.³ Therefore, comparisons between displaced populations should consider differences in context and study design. The malaria positivity observed in this study adds evidence from an understudied humanitarian population in Somalia. Previous studies have highlighted that internally displaced populations may experience increased vulnerability to malaria because of densely populated living conditions, disrupted access to healthcare services, limited availability of prevention interventions, and environmental factors that facilitate malaria transmission.^5,36 In Somalia, malaria remains an important public health concern, with displaced and vulnerable populations facing additional challenges related to malaria prevention, diagnosis, treatment access, and continuity of health services.³⁸

After adjustment for potential confounders, none of the individual socio-demographic, household, malaria-prevention, or healthcare-seeking factors examined in this study remained independently associated with malaria positivity. Similar findings have been reported in studies among displaced populations, where individual-level characteristics were not consistently associated with malaria outcomes after multivariable adjustment. These findings suggest that malaria positivity in humanitarian settings may be influenced by broader contextual factors, including environmental conditions, population mobility, local transmission patterns, and access to prevention and healthcare services, which were not fully captured in the present study.^16,36

The most notable finding was the variation in malaria positivity between IDP camps. In the adjusted analysis, participants from Samaraac, Gulow, Gal’ad, Dulqad 2, and Xaqdhowr had higher malaria positivity compared with participants from Bilan. However, this finding should not be interpreted as evidence that camp residence itself causes malaria infection. The observed differences may reflect variation in environmental conditions, vector exposure, population characteristics, timing of recruitment, healthcare access, prevention coverage, or other unmeasured camp-level factors. Such spatial heterogeneity in malaria transmission has been documented in different settings, where ecological conditions, vector-related factors, and differences in prevention and health-service access contribute to variation in malaria risk.³⁶ Since camp-level explanatory factors were not directly measured and only six purposively selected camps were included, these findings should be interpreted cautiously. The mixed-effects sensitivity analysis produced broadly similar estimates after accounting for possible within-camp correlation, suggesting consistency of the observed pattern while acknowledging uncertainty related to the limited number of clusters.

Housing conditions reflected important vulnerabilities among participants. Many participants lived in temporary shelters and structures constructed with plastic tarpaulin, and a substantial proportion reported limited household protection measures, including absence of screened windows. Although malaria positivity was slightly higher among some participants with less protective housing characteristics, these variables were not independently associated after adjustment. Housing conditions may influence malaria exposure through mosquito entry, reduced protection from vectors, and environmental exposure; however, their effects may depend on broader factors such as vector density, local ecology, and access to malaria prevention interventions.³⁹ Malaria-related knowledge showed both strengths and important gaps. Although approximately half of participants achieved good overall knowledge scores, gaps remained regarding malaria transmission and prevention. Previous studies have shown that malaria knowledge does not always translate into consistent preventive practices, as adoption of recommended behaviors may be influenced by access to prevention tools, socioeconomic conditions, and contextual barriers.^14,40 In this study, individual knowledge variables were not independently associated with malaria positivity, suggesting that knowledge alone may not explain malaria infection patterns among febrile adults in IDP camps.

The findings related to insecticide-treated nets (ITNs) highlight an important prevention gap. Although 60.66% of households reported availability of ITNs, consistent use was limited, with only 10.43% reporting always using ITNs. This difference between availability and regular use suggests that access alone may not guarantee consistent protective behavior. In humanitarian settings, sustained access, appropriate utilization, and replacement of ITNs remain important components of malaria prevention strategies.^14,39 The observed ITN availability among this displaced population indicates continued need to strengthen and monitor vector-control coverage. Care-seeking practices demonstrated both engagement with formal healthcare services and ongoing challenges. While many participants reported seeking medical advice or visiting healthcare facilities when experiencing malaria symptoms, self-medication was also common. Delays in seeking appropriate care may occur because of factors such as financial barriers, distance, availability of medicines outside health facilities, previous treatment experiences, and perceptions of illness severity.¹⁵ Malaria positivity among nearly three in every ten febrile participants underscores the importance of integrating timely diagnosis and treatment with vector-control interventions in IDP camps. WHO guidance recommends an integrated emergency malaria response combining risk assessment, diagnosis, case management, vector control, and operational coordination.⁴¹ Such approaches are relevant for strengthening malaria prevention and control planning in vulnerable displaced populations.

Strengths and Limitations

This study has several strengths. It included laboratory-confirmed malaria testing among a relatively large sample of febrile adults and assessed multiple domains relevant to malaria risk, including demographic characteristics, housing conditions, knowledge, prevention practices, and healthcare-seeking behaviors. The use of modified Poisson regression with robust variance estimation allowed direct estimation of prevalence ratios for a relatively common outcome. Furthermore, the sensitivity analysis accounting for camp-level clustering strengthened interpretation of the findings.

Several limitations should be considered. First, participants were febrile adults who voluntarily attended screening points in six purposively selected IDP camps; therefore, selection bias may have occurred because attendance could have been influenced by community mobilization activities, screening schedules, accessibility, and individual healthcare-seeking behavior. In addition, adults with signs or symptoms suggestive of severe malaria or other severe illness requiring urgent referral were excluded from participation. This exclusion may have reduced the inclusion of individuals with more severe malaria presentations and could have influenced the estimated malaria positivity; therefore, findings should be interpreted as representing febrile adults able to attend community-based screening points and should not be generalized to asymptomatic individuals, severely ill adults unable to attend screening points, or all residents of the selected IDP camps. Second, the study was limited to adults, and children were not assessed. This was because, at the time of the study, another team from our institution (Al Hayat Medical University) was conducting a parallel assessment of malaria positivity among febrile children in IDP camps; therefore, the present study focused on the adult population. Nevertheless, children remain an important malaria-risk group, and the findings should not be generalized to pediatric populations. Third, the use of the observed domain-specific mean to categorize knowledge and prevention/care-seeking practice scores represents a data-dependent threshold and may limit external comparability, although this approach was used because validated context-specific cut-offs were unavailable. Furthermore, although the internal consistency of these domains was assessed, reliability alone does not establish validity; therefore, the categorized scores should be interpreted within the context of the study population. Fourth, although internal quality control was in place for microscopy through independent double reading and third senior laboratory technologist resolution of discordance, external quality assurance was not feasible in our setting and therefore was not performed, which may have limited the reliability of the microscopy results. Fifth, the cross-sectional design limits assessment of temporal relationships and prevents causal conclusions. Sixth, camp-level factors such as vector density, rainfall patterns, environmental characteristics, and service availability were not directly measured, limiting interpretation of differences between camps. Additionally, although a mixed-effects sensitivity analysis was conducted to account for potential within-camp correlation, the small number of clusters (six camps) may have limited the precision of cluster-level variance estimates; therefore, these findings should be interpreted cautiously. Finally, data collection occurred during a limited study period and may not capture seasonal variation in malaria transmission.

Conclusion

This study demonstrates that malaria remains an important finding among febrile adults attending malaria screening points in selected internally displaced person (IDP) camps in Mogadishu, with Plasmodium falciparum accounting for the majority of confirmed infections. Malaria positivity varied between camps, suggesting that local contextual factors may contribute to differences in malaria positivity patterns in displacement settings; however, these differences should be interpreted cautiously because relevant camp-level determinants were not directly measured. Individual socio-demographic characteristics, housing factors, malaria-related knowledge, and prevention practices were not independently associated with malaria positivity after adjustment, highlighting the complexity of malaria risk in humanitarian environments. The observed malaria positivity, variation between camps, and inconsistent use of ITNs highlight the importance of continued malaria surveillance, efforts to improve consistent use of available ITNs, and further assessment of camp-specific factors to inform context-appropriate malaria prevention and control activities. Future research using longitudinal designs and direct measurement of environmental, vector-related, and health-service factors is needed to better understand the drivers of malaria transmission in IDP settings.

Funding Statement

No funding was obtained for this study.

Abbreviations

CCCM, Camp Coordination and Camp Management; CDC, Centers for Disease Control and Prevention; EDTA, Ethylenediaminetetraacetic acid; IDP, Internally displaced persons; ITNs, Insecticide-treated nets.

Data Sharing Statement

All data supporting the findings of this study are available from the corresponding author upon reasonable request.

Ethics Approval and Consent to Participate

This study was reviewed and approved by the Research Ethics Committee (REC) of Al Hayat Medical University (Approval No. AHREC/10/January/2026) and was conducted in accordance with the ethical principles of the 1964 Declaration of Helsinki and its subsequent amendments. Permission to conduct the study was obtained from the directors or responsible representatives of the selected IDP camps before initiating data collection. After eligibility assessment, trained data collectors explained the study objectives, procedures, potential benefits, potential risks or inconveniences, and participants’ rights in the Somali language to ensure adequate understanding. Verbal informed consent was obtained from eligible participants before any study procedures were conducted, including questionnaire administration and blood specimen collection. The Research Ethics Committee approved the use of verbal informed consent instead of written consent because some participants were reluctant to sign documents, as written agreements are commonly associated with legal matters in the local context. Participants were informed that participation was voluntary and that they could refuse participation or withdraw at any stage without any consequences. Following verbal consent, the data collector documented the participant’s agreement by marking the designated consent checkbox on the questionnaire, and the verbal consent process was witnessed by the trained interviewer. Participant confidentiality was maintained throughout the study, and no personal identifiers were collected or included in the analysis.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no conflicts of interest.