Utility of TaqMan Array Cards for detection of acute febrile illness etiologies in patients suspected of viral hemorrhagic fever infections Authors: Gloria Grace Akurut, Luke Nyakarahuka, Shannon Whitmer, Dianah Namanya, Kilama Kamugisha, Sophia Mulei, Jimmy Baluku, Alex Tumusime, Jackson Kyondo, Ketan Patel, Julius J. Lutwama, Trevor Shoemaker, Joel Montgomery, John D. Klena, Stephen Balinandi Published: 2025 ABSTRACT IntroductionDue to the difficulty in attributing a causative agent to acute febrile illnesses (AFI), multi-pathogen diagnostic tools should be prioritized in low-resource settings. A previously developed AFI-TaqMan Array Card (AFI-TAC), capable of detecting 26 pathogens within 2 h of nucleic acid extraction, was evaluated in Uganda.MethodsA cross-sectional retrospective study design was employed and utilized 182 viral hemorrhagic fever (VHF)-negative samples collected from Uganda, DRC, South Sudan and Kenya during routine surveillance from August 2018- March, 2019. These samples were tested on AFI-TAC targeting 17 viral, 8 bacterial and 3 protozoal pathogens known to cause fever. Patients with a body temperature of ≥ 38 °C, were bleeding, and had any other febrile symptoms were included. Previously confirmed VHF positive samples were used for assay verification.ResultsOverall, 7 pathogens were detected in 59 samples (32.42%) as follows: Plasmodium spp. (n = 49, 26.92%), non-typhoidal Salmonella (n = 3, 1.65%), Yellow Fever (YF) virus (n = 2, 1.10%), Salmonella enterica serovar typhi (n = 2, 1.10%), Leptospira spp (n = 1, 0.55%), Streptococcus pneumoniae (n = 1, 0.55%) and Rickettsia spp. (n = 1, 0.55%). Final outcome (alive vs. dead) as abstracted from case report forms differed significantly by pathogen category (p = 0.002) was significantly associated with assay positivity. We compared outcome across pathogen categories using a chi-square/Fisher’s exact test, as appropriate, reporting p-values (Table 2). Cough was the only clinical symptom significantly associated with Plasmodium infection (p = 0.016).ConclusionThe TAC is a feasible, readily adoptable diagnostic tool for use in Uganda and other sub-Saharan countries, particularly when incorporated into the national testing algorithm for differential diagnosis during AFI outbreaks and surveillance.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12879-025-12146-0. Background Viral hemorrhagic fevers (VHF) are acute febrile illnesses (AFI) generally characterized by fever with body temperatures ≥ 38 °C which may persist for longer than 72 h. Other symptoms commonly associated with AFI include headache, cough, vomiting, diarrhea, fatigue, and convulsions [1–5]. Establishing a causative agent based solely on clinical presentation remains a challenge due to the non-specific clinical presentation of AFI, making it difficult to infer a particular symptom to a specific disease. Without laboratory diagnostic testing, treatment is often delayed or ineffective. This is especially important in countries that have experienced multiple and often simultaneous outbreaks of diseases caused by VHFs such as filoviruses, Rift Valley Fever virus (RVFV), and /or Crimean Congo Hemorrhagic Fever virus (CCHFV) that can spread rapidly through human and animal populations [2, 6–9]. These pathogens are of significant public health concern and their detection and control is important for global health security and socio-economic stability. In Uganda, because malaria caused by Plasmodium falciparum is a major cause of AFI, many patients with fever are placed on antimalarial prophylaxis in the absence of laboratory diagnostic testing, thus contributing to antimalarial resistance, or broadly antimicrobial resistance [10–12]. In addition, the approach of symptomatic treatment in hospitals has been associated with the under-diagnosis of other AFIs known to exist in Uganda that include brucellosis, leptospirosis, rickettsia, chikungunya, dengue fever and orthopox viral infections. Diagnosis of a VHF is highly dependent on the direct detection of the causative virus by culture or the presence of viral nucleic acids in the sample from a suspected case. Although the humoral immune response can be utilized, it is not practically used for diagnosis in seclusion as it is only useful after several days of infection and more specifically in recovering patients [1–5, 13]. Traditional methods of clinical microbiology such as isolation, which requires BSL-4 laboratory environment, and acute- or convalescent-phase serology are slow and cumbersome, and values from common laboratory parameters like urinalysis, chemistry, and hematology are not specific, although they guide diagnosis. Molecular diagnosis of infectious diseases using nucleic acid amplification techniques (NAAT) such as the polymerase chain reaction (PCR) can be designed to be highly sensitive and specific to detect infectious agents. To maximize its effectiveness, NAAT methods are being designed to target more than one infectious agent in an assay; however, multiplex methods are often less sensitive compared to singleplex approaches [14]. TaqMan Array Cards (TAC) are microfluidic test cards designed to perform multiple real-time PCR specific to several different pathogens (depending on the TAC configuration) using primers and probes preloaded into the microwells of a 384-well plate. Between 1 and 8 samples can be tested in parallel against 12 to 384 assay targets [9, 14–17]. When analytical validation using clinical specimens was compared between AFI TAC performance and individual PCR assays, a sensitivity of 88% and specificity of 99% was observed [17]. Thus, the multi-pathogen detection card is convenient for screening AFIs during outbreaks or routine surveillance, consequently reducing the time taken to ascertain the cause of illness, and therefore prophylaxis. Whereas Next Generation Sequencing (NGS) has the potential to revolutionize pathogen detection, its cost and technical complexity pose challenges for smaller laboratories particularly in resource-limited environments. In Uganda, routine VHF surveillance is conducted year-round at the Uganda Virus Research Institute (UVRI) [1–5, 18–20]. Although the majority of samples submitted from symptomatic patients test negative for a VHF, the acute and fulminant nature of these illnesses necessitates the identification of the underlying etiologies. A multi-pathogen detection assay for differential diagnosis of acute febrile illnesses reduces the time taken to detect numerous etiologies which can lead to timely public health interventions [7, 9, 15, 21]. This study utilized the AFI-TAC assay to detect alternative etiologies for illnesses in patients whose samples tested negative for VHFs. Materials and methods Study design and population This cross-sectional retrospective study utilized 182 archived samples submitted to the VHF laboratory at UVRI from August 2018 to March 2019. During this period there was also enhanced surveillance for VHFs in Uganda following an outbreak of Ebola virus disease in Eastern Democratic Republic of Congo from which spillover cases were later detected in western Uganda [4, 18]. Samples were eligible if they had previously tested negative by real-time RT-PCR for Marburg virus, Ebola viruses (Sudan, Bundibugyo, Zaire), Crimean-Congo hemorrhagic fever virus (CCHFV) [1–5], and additionally with a myriad of other systemic and constitutional VHF signs and symptoms. From the case report forms, samples from we further selected patients who presented with a sudden onset of fever (≥ 38 °C), headache, fatigue, anorexia, abdominal pain, vomiting and hemorrhage (from any site), were selected for this study. Six known positive samples (RVFV = 5 and CCHFV = 1), obtained from routine surveillance, were used to verify the TAC assay characteristics. In addition, to measure the integrity of the respective samples, the TAC assay used in this study contained a target for 18 S ribosomal RNA. All tested samples amplified for the 18 S (with mean Ct ≤ 20) confirming that there was good quality RNA. Sample handling and TAC testing Archived specimens stored at − 80 °C were retrieved once for a single controlled thaw at 4 °C overnight prior to extraction. For inactivation in a high-containment laboratory (HCL), 100 µL of sample (or nuclease-free water as a negative control) was added to 400 µL lysis buffer containing 2 µL carrier RNA, incubated 10 min, transferred to sterile cryogenic vials, and moved to a BSL-2 laboratory. RNA was extracted using the MagMAX RNA Isolation Kit (ThermoFisher Scientific, CA, USA) per manufacturer instructions on an automated MagMAX 15-well bead-based extractor (Applied Biosystems), with elution in 90 µL; eluates were kept at 4 °C for immediate analysis. TaqMan Array cards, first described by [17] with modification; added targets include; Lassa virus, Oropouche virus, Mayaro virus, Burkholderia, Orientia, Salmonella suis, Streptococcus pneumonia, Plasmodium vivax and P. falciparum. (Supplemental File; excel 1). Each AFI-TAC tested a total of six patient samples, plus a positive and a negative control. Mastermix was prepared to include 1x Ag-Path-ID 2X rt PCR buffer and the enzyme mix (Applied Biosystems ThermoFisher Scientific, California, USA). Into each microfuge tube, 80 µL of master mix and 20 µL of the extracted RNA (brief vortex and spin); 100 µL of the mix was loaded into the inlet port, cards were centrifuged at 2,400 x g for 2 min, sealed and run on a VIIA7 Real-time PCR system (Life Technologies Carlsbad, California, USA) following cycling conditions: 45 °C for 20 min; 95 °C for 10 min; 40 cycles of 95 °C for 15s, and 60 °C for 1 min. Positivity was defined as cycle threshold (Ct) ≤ 35with acceptable amplification curves. Quality control included endogenous 18 S rRNA (sample adequacy) and exogenous MS2/PhHV (process) controls; runs were accepted only when controls performed as expected. Data capture and analysis The data were organized in Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) and analyzed in R (RStudio). We summarized demographics and pathogen frequencies with descriptive statistics. Bivariate associations between occupation, age, sex, and disease status were assessed using Pearson’s χ² test or Fisher’s exact test, as appropriate, with two-sided α = 0.05. Because Plasmodium infection was relatively common, effect sizes for symptom–infection were expressed as prevalence ratios (PRs). We fit generalized linear models with a log link to estimate adjusted PRs for Plasmodium in relation to clinical symptoms, attempting log-binomial regression first and using Poisson regression with robust (HC0) standard errors if the log-binomial model was unsuitable. Adjusted analyses were limited to Plasmodium a priori due to adequate events; other pathogens had insufficient case counts for stable multivariable modeling. For completeness, crude PRs from 2 × 2 tables are reported with Katz 95% CIs (continuity correction applied when necessary). We evaluated basic model diagnostics (over-dispersion and influential observations) for the Plasmodium models. Results Socio-demographic characteristics of study participants A total of 182 samples collected from August 2018 to March 2019 met the inclusion criteria. Of the 182 samples, 64.29% were from males, the median age of the patients was 27 years (range 0–80 years) and 79.67% were adults. The geographical distribution of samples selected for this study was as follows; Uganda (n = 149; 81.7%)), Democratic Republic of Congo (n = 21; 11.8%), South Sudan (n = 7; 3.76%), Kenya (n = 1; 0.54%), and n = 2; (2.15%) had no recorded information on country of origin. The baseline characteristics of samples and patients included in this study have been summarized in Table 1. Table 1Sociodemographic profile of patients and distribution of detected pathogensCharacteristicN = 182SexMale117 (64.29%)Female65 (35.71%)Age < 18 years 37 (20.33%) ≥ 18 years 145 (79.67%)Region Uganda 149 (81.9%) Democratic Republic of Congo 21 (11.8%) South Sudan 7 (3.76%) Kenya 1 (0.54%) No country indicated 4 (2.15%)Pathogen detected Plasmodium P.falciparum (34/49)49 (26.9%) Salmonella 3 (1.65%) Yellow fever virus 2 (1.10%) S. typhi 2 (1.10%) Rickettsia 1 (0.55%) Leptospira 1 (0.55%) S. pneumoniae 1 (0.55%) Negative 123 (67.58%)Occupation Farmer 64 (35.16%)Child/pupil52 (29.57%) Other 53 (29.12%) No occupation recorded 13 (7.14%)Living Status Dead 42 (23.08%) Alive 140 (76.92%) Table 1Sociodemographic profile of patients and distribution of detected pathogensCharacteristicN = 182SexMale117 (64.29%)Female65 (35.71%)Age < 18 years 37 (20.33%) ≥ 18 years 145 (79.67%)Region Uganda 149 (81.9%) Democratic Republic of Congo 21 (11.8%) South Sudan 7 (3.76%) Kenya 1 (0.54%) No country indicated 4 (2.15%)Pathogen detected Plasmodium P.falciparum (34/49)49 (26.9%) Salmonella 3 (1.65%) Yellow fever virus 2 (1.10%) S. typhi 2 (1.10%) Rickettsia 1 (0.55%) Leptospira 1 (0.55%) S. pneumoniae 1 (0.55%) Negative 123 (67.58%)Occupation Farmer 64 (35.16%)Child/pupil52 (29.57%) Other 53 (29.12%) No occupation recorded 13 (7.14%)Living Status Dead 42 (23.08%) Alive 140 (76.92%) Sociodemographic profile of patients and distribution of detected pathogens Plasmodium P.falciparum (34/49) Pathogens detected by AFI-TAC Overall, TAC detected 7 pathogens in 59 samples (32%) and 123 samples were negative as shown in Table 1. The major etiologies detected were Plasmodium spp (n = 49, 26.9%), and particularly of these were determined to be Plasmodium falciparum species (69%, 34 out of 49). Other etiologies detected include; Salmonella spp. (n = 3; 1.65%), Yellow Fever virus (n = 2; 1.1%), S. typhi (n = 2; 1.1%), Streptococcus pneumoniae (n = 1, 0.55%), Leptospira spp (n = 1, 0.55%), and Rickettsia spp (n = 1, 0.55%). No coinfections were detected. Final outcome (alive vs. dead) differed significantly by pathogen category (p = 0.002). All cases with Leptospira spp., Rickettsia spp., Salmonella typhi, YF virus, and S. pneumoniae were fatal (CFR 100%), whereas CFRs were 16% for Plasmodium and 33% for other Salmonella infections (Table 2). There was a statistical significance between case fatality and having AFI related disease (p = 0.002). When clinical signs and symptoms were analyzed for their association with Plasmodium infections, no statistical significance was observed other than the high association between cough and plasmodium infections (p = 0.016) as shown in Table 3. Table 2Association between pathogens detected and different demographic parameters (category)Category 1OverallN = 182Lp2n = 1Pl3n = 49Rk4n = 1Sal5n = 3S.p6n = 1S.th7n = 2YF8n = 2p-valueSex%%%%%%%%0.9 F 9 360430330500 M 10 64100571006710050100Age0.9 < 18 2002700000 ≥ 18 8010073100100100100100Occupation0.3 Farmer 381003800100050 Child/ pupil 31040033000 Other 3102210067010050Final outcome0.002 Dead 2301610033100100100 Alive 7710084067000Region0.4 North 110190010000 West 7110062106701000 Central 180190330001Demographic parameters including; sex, age, occupation, region2Leptospira spp 3 Plasmodium spp 4 Ricketssia spp 5 Salmonella enterica 6 Streptococcus pneumoniae 7 Salmonella typhi 8Yellow fever virus9Female10Male Table 2Association between pathogens detected and different demographic parameters (category)Category 1OverallN = 182Lp2n = 1Pl3n = 49Rk4n = 1Sal5n = 3S.p6n = 1S.th7n = 2YF8n = 2p-valueSex%%%%%%%%0.9 F 9 360430330500 M 10 64100571006710050100Age0.9 < 18 2002700000 ≥ 18 8010073100100100100100Occupation0.3 Farmer 381003800100050 Child/ pupil 31040033000 Other 3102210067010050Final outcome0.002 Dead 2301610033100100100 Alive 7710084067000Region0.4 North 110190010000 West 7110062106701000 Central 180190330001Demographic parameters including; sex, age, occupation, region2Leptospira spp 3 Plasmodium spp 4 Ricketssia spp 5 Salmonella enterica 6 Streptococcus pneumoniae 7 Salmonella typhi 8Yellow fever virus9Female10Male Association between pathogens detected and different demographic parameters (category) 1Demographic parameters including; sex, age, occupation, region 2Leptospira spp 3 Plasmodium spp 4 Ricketssia spp 5 Salmonella enterica 6 Streptococcus pneumoniae 7 Salmonella typhi 8Yellow fever virus 9Female 10Male Table 3Adjusted prevalence ratios (PR) for clinical characteristics associated with plasmodium infectionCharacteristic1PR295% CI1/np-valueFever1.370.48–3.930.556Bleeding0.730.39–1.360.327Vomiting1.730.94–3.190.076Diarrhea0.790.41–1.540.490Anorexia0.660.36–1.200.170Abdominal pain0.930.50–1.710.811Chest pain1.480.78–2.830.233Muscle pain1.110.60–2.060.732Headache1.310.68–2.520.412Cough0.370.16–0.820.015Sore throat1.280.51–3.190.596Conjunctivitis0.520.14–1.890.322Skin rash0.990.38–2.600.9861Clinical characteristics/ symptoms2Prevalence ratio3Confidence interval Table 3Adjusted prevalence ratios (PR) for clinical characteristics associated with plasmodium infectionCharacteristic1PR295% CI1/np-valueFever1.370.48–3.930.556Bleeding0.730.39–1.360.327Vomiting1.730.94–3.190.076Diarrhea0.790.41–1.540.490Anorexia0.660.36–1.200.170Abdominal pain0.930.50–1.710.811Chest pain1.480.78–2.830.233Muscle pain1.110.60–2.060.732Headache1.310.68–2.520.412Cough0.370.16–0.820.015Sore throat1.280.51–3.190.596Conjunctivitis0.520.14–1.890.322Skin rash0.990.38–2.600.9861Clinical characteristics/ symptoms2Prevalence ratio3Confidence interval Adjusted prevalence ratios (PR) for clinical characteristics associated with plasmodium infection 1Clinical characteristics/ symptoms 2Prevalence ratio 3Confidence interval As part of a quality control process, samples previously positive for individual real-time PCR (RVF, n = 5 and CCHF, n = 1) were included for assay verification and there was 100% agreement as they also tested positive by AFI-TAC. Similarly, samples that initially tested positive for YF virus by TAC were further found positive by individual real time RT-PCR. Discussion Multiplex pathogen detection tools like TAC PCR are valuable for diagnosing multiple infections for effective disease surveillance [9, 14, 16]. Because these systems are open, they can be custom-designed to investigate possible causes of syndromic disease (e.g., acute febrile illness) or for priority diseases within a country or region and environmental sample testing [14]. TAC systems present advantages including minimal need for skilling personnel reduced handling time and thus improving overall cost of clinical diagnosis, ease of data interpretation, cards can be made available through bulk purchase since the shelf life is about 2 years under refrigeration [14]. One of the TAC system, is the possibility to identify co-infections. From this data, no co-infections were identified contrary to other studies [9, 16]. Prior to this study, after completion of the VHF testing algorithm (using singleplex assays for Ebolavirus, Sudan Ebolavirus, Bundibugyo Ebolavirus, Marburg virus, Crimean-Congo Hemorrhagic fever virus and Rift Valley fever virus) at UVRI [1–5], no further testing would be performed on a sample, leaving behind several questions on the course of public health actions that needed to be made. In this study, of the 182 VHF-negative samples tested, at least one organism was identified in 59 (32.42%) of the samples by AFI-TAC assay, practically meaning that a diagnosis of the infectious cause of disease can be made for 1 of every 3 VHF-negative cases. The most prevalent pathogen was Plasmodium spp., which was not unexpected as most of these are known to cause malaria fever that for a long time is a challenge for differential diagnosis of VHFs in sub-Saharan Africa [6, 9, 11, 16, 22]. Interestingly, two cases of YF virus detected by this study triggered an outbreak investigation of this etiology in Masaka district, Central Uganda, by the Ministry of Health following confirmation by singleplex PCR. From this study, YF virus infection and mortality was associated with occupations such as hunting and farming whose activities predispose people to mosquito bites. Furthermore, YF virus cases were all male similar to the outbreak in northern Uganda that found males were more likely than females to acquire the YF virus [23]. This predisposition of males towards YF virus infection was linked to economic activities undertaken by males around forests. Both non-typhoidal and typhoidal Salmonella enterica were detected, (3; 1.65%) and (2; 1.10%) respectively, as well as Leptospira (1; 0.55%), and Rickettsia (1, 0.55%). Finally, a single case of Streptococcus pneumoniae, was found. Similar studies from Tanzania and Kenya found Plasmodium spp. to be the most common pathogen detected (using the identical AFI card while Leptospira, Salmonella enterica and Rickettsia were detected in 1% of cases [6, 9, 21]. For the detection of AFI using this card, blood samples are recommended [17], however for some pathogens such as S. pneumonia and Rickettsia blood is not the most ideal sample. In as much as lung biopsies, pleural fluid and blood cultures are recommended for S. pneumonia detection, a bacterial load suitable for blood culture would be reliably detected by PCR which is considered for sensitive. Although eschar biopsies or swabs are recommended for the detection of Rickettsia, it has been detected in blood [6, 9, 16, 21, 24]. In this study several pathogens of outbreak potential and concern to Uganda’s public health system were identified. However, the detection of high consequence pathogens rarely reported in Uganda such as West Nile virus, Bartonella, and Coxiella was not made. To assure that the assay would have detected a variety of low frequency but high impact pathogens, this study included six RVFV and CCHFV samples (RVFV, n = 5 and CCHFV, n = 1) with a range of Ct values (25 to 28). All the 6 samples were detected as positive for their respective pathogens by TAC with almost identical Ct values that also ranged from 26 to 29. This showed that the performance of TAC is comparable to singleplex PCR. Our internal laboratory standard operating procedures use a Cutoff value of ≤ 35 for a positive result for both assays. During the course of the study, due to the nature of the VHF program at the UVRI [4, 20], samples known to be positive for other pathogens other than Ebola virus, Marburg virus, Crimean Congo Hemorrhaghic Fever virus, Rift Valley fever virus and YF virus were not easy to get and include. To validate the other pathogens, we had to rely on the developer’s results [17]. This study found Plasmodium infections to be the most prevalent, similar to findings from studies conducted in Tanzania, Kenya and Bangladesh using a similar approach [6, 9, 16, 21]. A study on metagenomic next-generation sequencing of samples from pediatric febrile illness in Tororo, Uganda found the most prevalent AFI etiology to be Plasmodium falciparum at 51.1% [25]. As much as P.vivax targets were on the card, 15/49 samples that tested positive for Plasmodium still remain unspeciated. The rarity and low prevalence of P. vivax in sub-Saharan Africa might owe to the fact that there is a wide range of Duffy negative alleles in the population making them more resistant to P. vivax [26]. Although targets for P.malariae or P.ovale were not included on the card used in this study, they have been documented to exist in Uganda and could account for the 30.58% unspeciated plasmodium positive samples in this study [11, 12]. At 95% confidence level, cough was the only clinical characteristic statistically associated with Plasmodium infections. Similarly, complicated Malaria has been associated with lung damage, altered pulmonary functions such as acute respiratory distress with signs such as cough [27]. In this study, all mortalities were associated with an infection. Although, this could correspond to delays in diagnosis or misdiagnosis culminating to inadequate patient support and therefore more severe forms of illness, it cannot be deduced from this study that all cases infected with the involved etiologies do die in the general population. Due to the small number of samples (≤ 3) that tested positive for typhoidal and non typhoidal S. enterica, S. pneumonia, Rickettsia, Leptospira and YF virus, it was difficult deduce conclusions about associations between clinical presentation and infection. In this study, samples from males were more than those that had been collected from females. We did not investigate the factors responsible for this observation. However, previous studies in Uganda have shown gender differences in health seeking behaviors, with more males preferring to seek healthcare only when their disease conditions become severe [28, 29]. Similarly [9, 16, 21], reported a predominance of male in their findings. Viral hemorrhagic fevers are more frequently diagnosed in male patients [1–5]. All samples used in this study were from severely ill-patients, given the clinical signs and symptoms used for their selection, thus perhaps the main reason for this observation. Similarly, the higher number of samples from middle-aged adults (with a median age of 27 years) compared to other age groups in this study, is simply a reflection of work life behaviors within the traditional communities in Uganda, as this is also the most-active age group with perhaps more predisposing factors for disease infections [28–30]. Despite the above encouraging findings, some of the identified limitations of the TAC assay include the need to test 8 samples per run to avoid waste of the TAC plate. In situations where the sample volumes are low plus the necessity to adhere to set result turnaround times, the assay becomes expensive. On the other hand, if sample volumes are very high, testing 6–8 samples may be a limitation to meeting the turnaround time. The PCR products cannot be used for any downstream analysis like sequencing. In addition, the need for a single efficient annealing temperature for the different gene fragments between the targeted etiologies makes optimization of the TAC assay very cumbersome to achieve [14, 31]. Having the targets already on the card presents a challenge of no possibility for manipulation. Another important limitation of our study is that all testing was performed on blood samples, yet not all pathogens included in the AFI-TAC panel are optimally detectable from blood alone. While most viral agents and parasitic infections such as Plasmodium are reliably identified from blood, certain bacterial pathogens including Streptococcus pneumoniae and Rickettsia spp. are often more sensitively detected from respiratory, cerebrospinal fluid, or tissue samples. Reliance on blood-only testing may therefore underestimate the true burden of these pathogens in our cohort. This highlights the importance of appropriate sample selection when applying TAC for AFI surveillance and underscores the need for complementary specimen types in future studies. One of the challenges of using the AFI-TAC is that the pathogen targets are predetermined on the card and cannot be altered by the end user. In the version applied here, assays targeted pan-Plasmodium, P. falciparum, and P. vivax, but not P. malariae or P. ovale. This limitation likely contributed to the 15/49 Plasmodium-positive samples that could not be speciated. Both P. malariae and P. ovale are known to circulate in Uganda, usually at low density and often as mixed infections with P. falciparum, but their true burden remains poorly defined [11, 12]. In addition, some of the unspeciated detections occurred at high Ct values, suggesting low parasite density may also have contributed to failed speciation. Future iterations of the AFI-TAC for AFI surveillance in Uganda should therefore consider inclusion of P. malariae and P. ovale targets and the need to quantify parasite burden in both mixed and mono-infections. From this study, it is eminent to include the TAC card testing as routine for all VHF negative testing done in country. The unspeciated plasmodium positive samples need to be tested for other Plasmodium species such as Plasmodium malariae and Plasmodium ovale. The reason for omitting other plasmodium species on the card remains unknown but could most likely be linked to low parasite density. Although they have been detected as mixed infection with P. falciparum, the actual burdens of each remains to be established. Based on the commonly circulating AFI pathogens in Uganda, the AFI card needs further customization to include etiologies such as COVID-19, Monkeypox virus, and measles with clear guidance from the Ministry of Health. In addition, a TAC assay targeting travelers could be designed for cross boarder surveillance especially for pathogens that are not endemic to Uganda like Lassa fever virus and Nipah virus disease that, could otherwise be imported by international travelers from countries where these etiologies are known to exist or be endemic. While the strategy of duplication of pathogen targets reduces the number of pathogens being detected thus making the assay even more expensive, it could increase the efficiency of the assay. Conclusion Although the TAC cards are expensive, with the platform and reagents costing ~ 10,000 and between 50 and 100 Usd respectively [6], for the number of pathogens tested and the number of samples that can be tested per card, strategic use of the cards could be a cost-effective way to perform passive surveillance for a range of pathogens that are of public health importance to Uganda. In addition, since the primers and probes are lyophililized, shipment can be done at room temperature, the cards can be reliably used for up to two years of storage under refrigeration. Furthermore, the use of the TAC does not need specialized training as long as personnel have been trained on the use of real time PCR which reduces the cost [14]. The TAC system provides flexibility of testing for multiple pathogen targets within a short period of time and without the need to purchase and maintain multiple kits, positive controls and reference materials. Coupled with confirmatory testing at the appropriate reference laboratory for presumptive-positive samples identified using TAC, the system could be effective at more quickly recognizing and responding to outbreaks involving suspected high consequence pathogens compared to singlex-PCR procedures. This study has therefore proved that the AFI- TAC testing modality is of great value and can be adopted to Uganda for AFI surveillance and outbreak investigations. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1