Rapid Antimicrobial Susceptibility Testing Research Articles

Barriers to implementation of rapid identification and antimicrobial susceptibility testing technologies in the clinical microbiology laboratory: an American perspective

Abstract Clinical microbiology laboratories are responsible for confirming the aetiology of infectious diseases and providing antimicrobial susceptibility testing results. Traditional culture-based testing can be augmented by more rapid testing modalities to provide clinically actionable information as quickly as possible. Despite improvements in patient outcomes, many clinical microbiology laboratories are facing challenges to in-sourcing these technologies. Depending on a multitude of factors, including size, location and patient population served, these barriers may affect some laboratories and hospital systems to greater or lesser extents than others. It will be up to each individual facility to ascertain its ability to overcome barriers. To aid in this self-assessment, we present for thoughtful consideration a discussion of the barriers to implementation of rapid identification and antimicrobial susceptibility testing technologies, with specific attention to matters of financial cost, staff expertise, operational issues and stakeholder buy-in.

Retrospective evaluation of rapid genotypic ID and phenotypic AST systems on positive blood culture turnaround time and simulated potential impacts on bloodstream infection management

Abstract Background Bloodstream infections are linked to heightened morbidity and mortality rates. The consequences of delayed antibiotic treatment can be detrimental. Effective management of bacteraemia hinges on rapid antimicrobial susceptibility testing. Objectives This retrospective study examined the influence of the VITEK® REVEAL™ Rapid AST system on positive blood culture (PBC) management in a French tertiary hospital. Materials and methods Between November 2021 and March 2022, 79 Gram-negative monomicrobial PBC cases underwent testing with both VITEK®REVEAL™ and VITEK®2 systems. Results The study found that VITEK®REVEAL™ yielded better results than the standard of care, significantly shortening the time to result (7.0 h compared to 9.6 h) as well as the turnaround time (15 h compared to 31.1 h) when applied for all isolates. Conclusions This study implies that the use of VITEK®REVEAL™ enables swift adaptations of antibiotic treatment strategies. By considerably minimizing the turnaround time, healthcare professionals can promptly make necessary adjustments to therapeutic regimens. Notably, these findings underscore the potential of VITEK®REVEAL™ in expediting appropriate antibiotic interventions, even in less ideal conditions. Further studies in varied laboratory contexts are required to validate these encouraging outcomes.

Platinum nanoparticles-based electrochemical H2O2 sensor for rapid antibiotic susceptibility testing

With the increase of antimicrobial resistance, rapid antibiotic susceptibility testing (AST) to guide precise antibiotic administration has become increasingly important. However, current gold standard AST approaches tend to take up to 24–48 h. In this work, based on the nature of catalase-positive bacteria decomposing H2O2, we developed a rapid, portable, straightforward, and cost-effective phenotypic AST approach by detecting residual H2O2 using a Pt nanoparticles-based electrochemical sensor. The pulse current of the sensor exhibited a linear increase with rising H2O2 concentration, demonstrating a high sensitivity of ∼382.2 μA cm−2 mM−1. This approach showed superb diagnostic performance, with an area under the curve of 1 for 24 clinical samples of Escherichia coli and Staphylococcus aureus, with a total detection time of 60 and 45 min, respectively. Furthermore, the performance of the sensor showed no degradation even after 100 detections, promising a substantial reduction in AST costs. Overall, the proposed approach exhibited immense potential for diagnosing bacterial antibiotic resistance.

A comparative study of a rapid phenotypic antimicrobial susceptibility testing system directly from positive blood cultures to the disk diffusion and VITEK 2 methods

BackgroundSepsis is a life-threatening condition that impacts 49 million people annually and causes 11 million deaths worldwide. Surviving bloodstream infections (BSIs) depends on the rapid administration of effective antimicrobial treatment, underscoring a need for rapid antimicrobial susceptibility testing (AST). AimTo evaluate the performance of Quantamatrix's dRAST v2.5 system (Seoul South Korea) for AST directly from positive blood cultures as compared to the Disk-Diffusion (DD) and VITEK 2 methods. MethodsThe study included 191 positive blood cultures from clinical samples and spiked blood culture bottles. Following Gram staining and species-level identification, AST was performed by VITEK 2 and standard DD methods using CLSI (2021) interpretation. ResultsdRAST demonstrated very good AST performance for a Gram-negative isolate, and good performance for Gram-positive isolates, meeting CLSI criteria for the acceptance of a new method. Antimicrobials that were not considered verified compared to VITEK 2 and DD were cefazolin, ceftazidime, meropenem, and trimethoprim/sulfamethoxazole for Gram-negatives and clindamycin, erythromycin, penicillin, and oxacillin for Gram-positives. dRAST ESBL detection results were strongly correlated with the ESBL phenotypes obtained with other methods. Additional resistance mechanisms were in concordance with traditional tests. ConclusionsdRAST demonstrated good AST performance, meeting CLSI criteria for most relevant antibiotics. dRAST was associated with a significant reduction in time-to-results, labor, and the subjectivity of result analyses, making it a valuable addition to efforts supporting the treatment of patients with bacteremia.AST (antimicrobial susceptibility test), blood culture, dRAST, rapid methods, sepsis, turnaround time (TAT).

Evaluating the impact of rapid antimicrobial susceptibility testing for bloodstream infections: a review of actionability, antibiotic use and patient outcome metrics

Evaluating the impact of rapid antimicrobial susceptibility testing for bloodstream infections: a review of actionability, antibiotic use and patient outcome metrics

Evaluation of the Reveal® AST (SPECIFIC) for Antimicrobial Susceptibility Testing from Positive Blood Culture Spiked with Carbapenem-Resistant Isolates

As bloodstream infections and associated septic shock are common causes of mortality in hospitals, rapid antibiotic susceptibility testing (AST) performed directly on positive blood cultures is needed to implement an efficient therapy in clinical settings. We evaluated the Reveal® rapid AST system on a collection of 197 fully characterized carbapenem-resistant Enterobacterales, including 177 carbapenemase producers (CPE) spiked in blood culture bottles. The clinical categorization based on the Minimal Inhibitory Concentration (MIC) determination of eighteen antimicrobial molecules was compared to the clinical categorization based on the disk diffusion assay as a reference. The Reveal AST system provided results within a mean time to result of 5 h. Overall, the categorical agreement (CA) between the two techniques was 94.1%. The rates of very major errors (VMEs), major errors (MEs) and minor errors (mEs) were 3.8%, 3.7% and 5.6%, respectively. Imipenem was the antimicrobial with the lowest CA rate (78.7%), with rates of 15% VMEs and 10.7% MEs, but the performances were better when considering only the non-CPE category (CA of 89%). On this resistant collection of Enterobacterales with numerous acquired β-lactamases, the Specific Reveal assay proved to be useful for a rapid determination of AST compatible with a quick adaptation of the patient’s antimicrobial treatment.

Rapid Phenotypic and Genotypic Antimicrobial Susceptibility Testing Approaches for Use in the Clinical Laboratory.

The rapid rise in increasingly resistant bacteria has become a major threat to public health. Antimicrobial susceptibility testing (AST) is crucial in guiding appropriate therapeutic decisions and infection prevention practices for patient care. However, conventional culture-based AST methods are time-consuming and labor-intensive. Therefore, rapid AST approaches exist to address the delayed gap in time to actionable results. There are two main types of rapid AST technologies- phenotypic and genotypic approaches. In this review, we provide a summary of all commercially available rapid AST platforms for use in clinical microbiology laboratories. We describe the technologies utilized, performance characteristics, acceptable specimen types, types of resistance detected, turnaround times, limitations, and clinical outcomes driven by these rapid tests. We also discuss crucial factors to consider for the implementation of rapid AST technologies in a clinical laboratory and what the future of rapid AST holds.

Evaluation of the Vitek® Reveal™ system for rapid antimicrobial susceptibility testing of Gram-negative pathogens, including ESBL, CRE and CRAB, from positive blood cultures

Blood culture (BC) remains the reference diagnostic tool for bloodstream infections but is hampered by long turn-around time (TAT). This study evaluated the Vitek® Reveal™ (VR) system for rapid antimicrobial susceptibility testing (AST) with 72 cases of monomicrobial BCs (55 Enterobacterales, 12 Pseudomonas aeruginosa and 5 Acinetobacter baumannii), including isolates producing carbapenemases and/or extended-spectrum β-lactamases. VR returned AST results with a mean TAT of 5.4 h. Compared to a conventional workflow based on broth microdilution, VR exhibited essential agreement (EA) and category agreement (CA) >90 % in most cases, except with meropenem for Enterobacterales (CA, 85.5 %), piperacillin/tazobactam for P. aeruginosa (EA, 83.3 %), and trimethoprim/sulfamethoxazole for A. baumannii (CA and EA, 80 %). Bias exhibited an underestimation trend with ceftazidime/avibactam (−78.9 %) and ceftazidime (−50 %) for Enterobacterales and P. aeruginosa, respectively. Overall, VR appears an interesting tool to decrease TAT of the BC workflow, although further evaluation with some antibiotic-pathogen combinations would be warranted.

Membrane potential dynamics unveil the promise of bioelectrical antimicrobial susceptibility Testing (BeAST) for anti-fungal screening.

Rapid assessment of the efficacy of antimicrobials is important for optimizing treatments, avoiding misuse and facilitating the screening of new antimicrobials. The need for rapid antimicrobial susceptibility testing (AST) is growing with the rise of antimicrobial resistance. Here, we present bioelectrical AST (BeAST). Combining time-lapse microscopy and mathematical modeling, we show that electrically induced membrane potential dynamics of yeast cells correspond to the strength of growth inhibition. Furthermore, we demonstrate the utility of BeAST for assessing antimicrobial activities of novel compounds using biogenic silver nanoparticles.

Rapid determination of antibiotic susceptibility of clinical isolates of Escherichia coli by SYBR green I/Propidium iodide assay

Infections caused by pathogenic Escherichia coli are a serious threat to human health, while conventional antibiotic susceptibility tests (AST) have a long turn-around time, and rapid antibiotic susceptibility methods are urgently needed to save lives in the clinic, reduce antibiotic misuse and prevent emergence of antibiotic-resistant bacteria. We optimized and validated the feasibility of a novel rapid AST based on SYBR Green I and Propidium Iodide (SGPI-AST) for E. coli drug susceptibility test. A total of 112 clinical isolates of E. coli were collected and four antibiotics (ceftriaxone, cefoxitin, imipenem, meropenem) were selected for testing. Bacterial survival rate of E. coli was remarkably linearly correlated with S value at different OD600 values. After optimizing the antibiotic concentrations, the sensitivity and specificity of SGPI-AST reached 100%/100%, 97.8%/100%, 100%/100% and 98.4%/99% for ceftriaxone, cefoxitin, imipenem and meropenem, respectively, and the corresponding concordances of the SGPI-AST with conventional AST were 1.000, 0.980, 1.000 and 0.979, respectively. The SGPI-AST can rapidly and accurately determine the susceptibility of E. coli clinical isolates to multiple antibiotics in 60 min, and has the potential to be applied to guide the precise selection of antibiotics for clinical management of infections caused by pathogenic E. coli.

CS/PVP Hydrogel-Based Nanocapillary for Monitoring Bacterial Growth and Rapid Antibiotic Susceptibility Testing.

Infection with drug-resistant bacteria poses a significant threat to human health. Judicious use of antibiotics could reduce the likelihood of bacterial resistance, which can be evaluated through antibiotic susceptibility testing (AST). This paper focuses on the application of a needle-like nanocapillary tip filled with chitosan (CS)/polyethylene pyrrolidone (PVP) hydrogel based on its specific pH-sensitive properties. The gel-filled nanocapillary has the potential to be used for electrical pH detection with a sensitivity of 3.06 nA/pH and a linear range from 7.3 to 4.3. Such sensitivity for pH measurement could be extended for monitoring of bacterial (such as Escherichia coli and Streptococcus salivarius) growth because of the relationship between pH and bacterial growth. Bacterial growth curves obtained using the hydrogel-filled nanocapillary showed good agreement with the OD600 method. Moreover, this device could be applied for rapid AST for tetracycline and norfloxacin on E. coli with minimum inhibitory concentrations of 2 and 0.125 μg/mL, respectively. This study expands the application of the hydrogel-based nanocapillary for bacterial research by monitoring changes in pH values.

Blood culture-free ultra-rapid antimicrobial susceptibility testing.

Treatment assessment and patient outcome for sepsis depend predominantly on the timely administration of appropriate antibiotics1-3. However, the clinical protocols used to stratify and select patient-specific optimal therapy are extremely slow4. In particular, the major hurdle in performing rapid antimicrobial susceptibility testing (AST) remains in the lengthy blood culture procedure, which has long been considered unavoidable due to the limited number of pathogens present in the patient's blood. Here we describe an ultra-rapid AST method that bypasses the need for traditional blood culture, thereby demonstrating potential to reduce the turnaround time of reporting drug susceptibility profiles by more than 40-60 h compared with hospital AST workflows. Introducing a synthetic beta-2-glycoproteinI peptide, a broad range of microbial pathogens are selectively recovered from whole blood, subjected to species identification or instantly proliferated and phenotypically evaluated for various drug conditions using a low-inoculum AST chip. The platform was clinically evaluated by the enrolment of 190 hospitalized patients suspected of having infection, achieving 100% match in species identification. Among the eight positive cases, six clinical isolates were retrospectively tested for AST showing an overall categorical agreement of 94.90% with an average theoretical turnaround time of 13 ± 2.53 h starting from initial blood processing.

Evaluation of Vitek-2 and MicroScan ASTs directly from positive blood culture: A cost-effective 24-h reduction for Enterobacterales and staphylococci

The reduction of antimicrobial susceptibility testing (AST) time-to-result is a central need, especially in sepsis treatment. The current automated rapid ASTs are still too expensive for many laboratories.We aimed to evaluate three pre-treatment methods for a same-day inoculation on both automated AST platforms available in our laboratory. We tested 100 Enterobacterales or staphylococci positive bottles.We obtained good results with the different methods and instruments. In particular, Vitek-2 showed good performances with Enterobacterales AST when inoculated with bacterial pellet (96.6% categorical agreement - CA-, 93.3% essential agreement - EA). Also short-term incubation colonies for staphylococci AST had acceptable CA (94.2%), even if with 77.5% EA. MicroScan system for staphylococci AST with both short-term incubation and direct blood inoculation reached >95% CA, but 92.5% and 83.6% EA, respectively. On the other hand, Enterobacterales AST showed optimal performances only with bacterial pellet inoculation (97.6% CA). In fact, direct blood inoculation showed not acceptable parameters for several molecules.Both systems allow a 24-h reduction in time-to-result, by using the same instruments of routine activity after rapid and cheap pre-treatments.

Development of Rapid Disk Diffusion Device Using Laser Speckle Formation Technology for Rapid Antimicrobial Susceptibility Testing

The escalation of antimicrobial resistance (AMR) due to the excessive and inappropriate use of antimicrobials has prompted the urgent need for more rapid and effective antimicrobial susceptibility testing (AST) methods. Conventional AST techniques often take 16–24 h, leading to empirical prescription practices and the potential emergence of AMR. The study aimed to develop a rapid disk diffusion (RDD) method utilizing laser speckle formation (LSF) technology to expedite AST results. The study aimed to evaluate the performance of LSF technology in determining antimicrobial susceptibility. In this study, preclinical and clinical settings were established to compare the LSF technology with conventional disk diffusion (DD) methods to measure the inhibition zones. Preclinical experiments with different bacterial strains demonstrated more than 70% categorical agreement (CA) against most antimicrobials. Further, clinical experiments with multiple strains and antibiotics revealed CA ranging from 40 to 79%, while major and minor discrepancies were observed around 30% and 11%, respectively. These observations revealed high concordance between RDD and DD for multiple antimicrobials in multiple species. The results underscore the potential of RDD-based LSF technology for hastening AST procedures. The current study is marked by a unique equipment setup and analysis approach. Collectively, the suggested laser-based RDD showed greater potential than previously developed comparable methods. The proposed method and design have a higher application potential than formerly developed similar technologies. Together, the study contributes to the ongoing development of rapid AST methods.

Evaluation of Quantamatrix dRASTTM system for rapid antimicrobial susceptibility testing of bacterial isolates from positive blood cultures, in comparison with commercial Micronaut broth microdilution system

Antimicrobial susceptibility testing (AST) from blood culture (BC) may take several days, limiting the eventual impact on antimicrobial stewardship. Hence, rapid AST systems represent a valuable support in shorting the time-to-response. In this work, the Quantamatrix dRASTTM system (dRAST) was evaluated for rapid AST on 100 monomicrobial BCs (50 Gram-negatives and 50 Gram-positives), including several isolates with clinically relevant resistance mechanisms. AST results were provided in 6-hours, on average. Compared to Micronaut (Merlin) system based on broth microdilution, dRAST exhibited an overall categorical agreement of 92.5 %, essential agreement of 89.0 %, and mean bias of 15.9 %. Category overestimation (potentially leading to unnecessary high-dosage treatment or to exclude active agents) and category underestimation (potentially leading to underdosing or using ineffective agents) were observed in 4.3 % and 3.1 % of cases, respectively. Even though several issues were reported, results confirmed the potential contribution of dRAST to shorten the BCs clinical microbiology workflow and management.

Clinical Pilot of Bacterial Transcriptional Profiling as a Combined Genotypic and Phenotypic Antimicrobial Susceptibility Test.

Antimicrobial resistance is a growing health threat, but standard methods for determining antibiotic susceptibility are slow and can delay optimal treatment, which is especially consequential in severe infections such as bacteremia. Novel approaches for rapid susceptibility profiling have emerged that characterize either bacterial response to antibiotics (phenotype) or detect specific resistance genes (genotype). GoPhAST-R is a novel assay, performed directly on positive blood cultures, that integrates rapid transcriptional response profiling with detection of key resistance gene transcripts, thereby providing simultaneous data on both phenotype and genotype. Here, we performed the first clinical pilot of GoPhAST-R on 42 positive blood cultures: 26 growing Escherichia coli, 15 growing Klebsiella pneumoniae, and 1 with both. An aliquot of each positive blood culture was exposed to 9 different antibiotics, lysed, then underwent rapid transcriptional profiling on the NanoString® platform; results were analyzed using an in-house susceptibility classification algorithm. GoPhAST-R achieved 95% overall agreement with standard antimicrobial susceptibility testing methods, with the highest agreement for beta-lactams (98%) and the lowest for fluoroquinolones (88%). Epidemic resistance genes including the extended spectrum beta-lactamase bla CTX-M-15 and the carbapenemase bla KPC were also detected within the population. This study demonstrates the clinical feasibility of using transcriptional response profiling for rapid resistance determination, although further validation with larger and more diverse bacterial populations will be essential in future work. GoPhAST-R represents a promising new approach for rapid and comprehensive antibiotic susceptibility testing in clinical settings.

Rapid Antimicrobial Susceptibility Testing Using the MicroScan System: Performance Evaluation of a 4-Hour Bacterial Cultivation From Positive Blood Cultures.

A reliable and above all, rapid antimicrobial susceptibility test (AST) is required for the diganostics of blood stream infections (BSI). In this study, resistance testing using DxM MicroScan WalkAway (MicroScan) from a 4-h subculture is compared with the standard overnight culture (18-24h). Randomly selected positive blood cultures (PBC, n = 102) with gram-negative bacteria were included in the study. PBC were sub-cultured onto appropriate agar plates and AST by MicroScan was performed after 4 h of incubation and repeated after incubation for 18-24 h as standard. In a total of 1909 drug-strain pairs, the 4-h subculture approach showed a very high essential agreement (EA) (98.6%) and categorical agreement (CA) (97.1%) compared with the standard. The incidence of minor error (mE), major error (ME), very major error (VME), and adjusted very major error (aVME) was 1.1%, 0.4%, 12.9%, and 5.3%, respectively. In summary, the use of 4-h subcultures for resistance testing with the MicroScan offers a very reliable and easy to realize time saving when testing positive blood cultures with gram-negative bacteria.

ARTIFICIAL INTELLIGENCE APPLICATIONS IN CLINICAL MICROBIOLOGY LABORATORY

Artificial intelligence is becoming an increasingly important component of clinical microbiology informatics. Researchers, microbiologists, laboratorians, and diagnosticians are interested in AI-based testing because these applications have the potential to improve the turnaround time, quality, and cost of a test. Artificial intelligence which has gained importance in the laboratory, is used to support decision-making, identification and antimicrobial susceptibility testing with various technologies, image analyses, and MALDI-TOF-MS in medical microbiology and in infectious disease testing. Treatment of infections requires rapid and accurate identification and antimicrobial susceptibility testing. Modern artificial intelligence (AI) and machine-learning (ML) methods can now complete tasks with performance characteristic comparable to those of expert human operators. As a result, many healthcare fields combine these technologies, including in vitro diagnostics and, more broadly laboratory medicine, incorporate these technologies. These technologies are rapidly being developed and disclosed, but by comparison, their application so far has been limited. We need to further establish best practices and improve our information system and communications infrastructure to promote the implementation of reliable and advanced machine learning-based technologies. İnvolvement of the clinical microbiology laboratory community is essential to ensure that laboratory data is adequately accessible and thoughtfully incorporated into robust, safe and clinically effective ML-supported clinical diagnoses. This has been made possible by advances in modern computing and the widespread digitalization of health information. In this review, it is aimed to give information about to AI and its subfield of ML application areas in clinical microbiology laboratory.

Can flow cytometric measurements of reactive oxygen species levels determine minimal inhibitory concentrations and antibiotic susceptibility testing for Acinetobacter baumannii?

Current antimicrobial susceptibility testing (AST) requires 16-24 hours, delaying initiation of appropriate antibiotics. Hence, there is a need for rapid AST. This study aims to develop and evaluate the feasibility of a rapid flow cytometric AST assay to determine minimum inhibitory concentration (MIC) for carbapenem-resistant Acinetobacter baumannii (CRAB). Antibiotic exposure causes increased intracellular reactive oxygen species (ROS) in bacteria. We hypothesized that ROS can be used as a marker to determine MIC. We assessed three CRAB clinical isolates across fifteen antibiotics at various concentrations in a customized 96-well microtiter plate. The antibiotics assessed include amikacin, beta-lactams (ampicillin/sulbactam, aztreonam, cefepime, ceftolozane/tazobactam, doripenem, imipenem, meropenem, and piperacillin/tazobactam), levofloxacin, polymyxin B, rifampicin, trimethoprim/sulfamethoxazole, and tetracyclines (tigecycline and minocycline). These clinical CRAB isolates were assessed for ROS after antibiotic treatment. Increased ROS levels indicated by increased RedoxSensorTM Green (RSG) fluorescence intensity was assessed using flow cytometry (FCM). MIC was set as the lowest antibiotic concentration that gives a ≥1.5-fold increase in mode RSG fluorescence intensity (MICRSG). Accuracy of MICRSG was determined by comparing against microtiter broth dilution method performed under CLSI guidelines. ROS was deemed accurate in determining the MICs for β-lactams (83.3% accuracy) and trimethoprim/sulfamethoxazole (100% accuracy). In contrast, ROS is less accurate in determining MICs for levofloxacin (33.3% accuracy), rifampicin (0% accuracy), amikacin (33.3% accuracy), and tetracyclines (33.3% accuracy). Collectively, this study described an FCM-AST assay to determine antibiotic susceptibility of CRAB isolates within 5 hours, reducing turnaround time up to 19 hours.

Rapid, antibiotic incubation-free determination of tuberculosis drug resistance using machine learning and Raman spectroscopy

Tuberculosis (TB) is the world's deadliest infectious disease, with over 1.5 million deaths and 10 million new cases reported anually. The causative organism Mycobacterium tuberculosis (Mtb) can take nearly 40 d to culture, a required step to determine the pathogen's antibiotic susceptibility. Both rapid identification and rapid antibiotic susceptibility testing of Mtb are essential for effective patient treatment and combating antimicrobial resistance. Here, we demonstrate a rapid, culture-free, and antibiotic incubation-free drug susceptibility test for TB using Raman spectroscopy and machine learning. We collect few-to-single-cell Raman spectra from over 25,000 cells of the Mtb complex strain Bacillus Calmette-Guérin (BCG) resistant to one of the four mainstay anti-TB drugs, isoniazid, rifampicin, moxifloxacin, and amikacin, as well as a pan-susceptible wildtype strain. By training a neural network on this data, we classify the antibiotic resistance profile of each strain, both on dried samples and on patient sputum samples. On dried samples, we achieve >98% resistant versus susceptible classification accuracy across all five BCG strains. In patient sputum samples, we achieve ~79% average classification accuracy. We develop a feature recognition algorithm in order to verify that our machine learning model is using biologically relevant spectral features to assess the resistance profiles of our mycobacterial strains. Finally, we demonstrate how this approach can be deployed in resource-limited settings by developing a low-cost, portable Raman microscope that costs <$5,000. We show how this instrument and our machine learning model enable combined microscopy and spectroscopy for accurate few-to-single-cell drug susceptibility testing of BCG.