Specific Diagnostic Procedures Research Articles

A Novel Tool for the Analysis and Detection of Copy Number Variants Associated with Haemoglobinopathies.

Several types of haemoglobinopathies are caused by copy number variants (CNVs). While diagnosis is often based on haematological and biochemical parameters, a definitive diagnosis requires molecular DNA analysis. In some cases, the molecular characterisation of large deletions/duplications is challenging and inconclusive and often requires the use of specific diagnostic procedures, such as multiplex ligation-dependent probe amplification (MLPA). Herein, we collected and comprehensively analysed all known CNVs associated with haemoglobinopathies. The dataset of 291 CNVs was retrieved from the IthaGenes database and was further manually annotated to specify genomic locations, breakpoints and MLPA probes relevant for each CNV. We developed IthaCNVs, a publicly available and easy-to-use online tool that can facilitate the diagnosis of rare and diagnostically challenging haemoglobinopathy cases attributed to CNVs. Importantly, it facilitates the filtering of available entries based on the type of breakpoint information, on specific chromosomal and locus positions, on MLPA probes, and on affected gene(s). IthaCNVs brings together manually curated information about CNV genomic locations, functional effects, and information that can facilitate CNV characterisation through MLPA. It can help laboratory staff and clinicians confirm suspected diagnosis of CNVs based on molecular DNA screening and analysis.

Management of marginal zone lymphomas.

Marginal zone lymphomas (MZLs) represent about 7% of B-cell non-Hodgkin lymphomas and include 3 different subtypes-namely, extranodal (EMZL), nodal, and splenic (SMZL). The initial assessment requires specific diagnostic and staging procedures depending on organ-related peculiarities. In particular, although positron emission tomography/computed tomography was not initially recommended, recent data have reassessed its role in the routine staging of MZL, especially when only localized treatment is planned or there is a suspicion of histologic transformation. Recent findings have improved the risk stratification of MZL patients, highlighting the association of early progression after frontline therapy with worse overall survival. A significant fraction of MZL cases may be related to specific bacterial (ie, Helicobacter pylori in gastric EMZL) or viral infections (hepatis C virus), and in the earlier phases of disease, a variable percentage of patients may respond to anti-infective therapy. Involved-site radiotherapy has a central role in the management of localized EMZL not amenable to or not responding to anti-infective therapy. Although rituximab-based treatments (bendamustine- rituximab in advanced EMZL or rituximab monotherapy in SMZL) have demonstrated favorable results, the current therapeutic scenario is predicted to rapidly change as emerging novel agents, especially Bruton's tyrosine kinase inhibitors, have demonstrated promising efficacy and safety profiles, leading to their approval in the relapsed setting. Moreover, a large variety of novel agents (phosphatidylinositol 3-kinase inhibitors, chimeric antigen receptor T-cells, bispecific antibodies) are being tested in MZL patients with encouraging preliminary results.

Idiopathic pulmonary fibrosis and intestinal disorders: An observational study

Background and aimIdiopathic pulmonary fibrosis (IPF) is a chronic respiratory disease characterized by a progressive decline in lung function and a specific histopathologic pattern defined as usual interstitial pneumonia. Early diagnosis and new therapeutic protocols have contributed to a reduction in disease progression. Thus, some patients may develop extrapulmonary diseases including malignancies and chronic pathologies. The aim of this study was to investigate the frequency of intestinal disorders such as polyps, colorectal carcinoma (CRC), and chronic inflammatory bowel disease (IBD) in patients with IPF. MethodsFrom the database of 189 patients with IPF (148 males, 78.3 %; 41 females, 21.7 %) residing in the district of Modena, we identified 44 patients (36 males, 81.8 %; 8 females, 18.2 %) with a histologically confirmed intestinal disease. ResultsIntestinal polyps were detected in 41 cases (93.2 %), of which 4 were associated with CRC and 1 with IBD; 1 patient had only CRC (2.3 %), and 2 patients had only IBD (4.5 %). Both males and females developed bowel disease, but males seemed to have a higher number of polyps and high-grade adenomas with a predisposition to malignant transformation. ConclusionsAs patients with IPF may present with intestinal diseases that can evolve into malignancies in some cases, they should undergo appropriate follow-up and targeted colorectal screening. Thus, colorectal pathologies should not take a back seat. These preliminary results encourage further research to select suitable patients for specific diagnostic and therapeutic procedures in order to prolong survival and improve the quality of life.

Long COVID-19 Myocarditis and Various Heart Failure Presentations: A Case Series.

(1) Background: Emerging data indicate that the ongoing COVID-19 pandemic may result in long-term cardiovascular complications, among which long COVID-19 myocarditis seems to be one of the most dangerous. Clinical presentation of cardiac inflammation ranges from almost asymptomatic to life-threatening conditions, including heart failure (HF) in different stages. (2) Methods: This is a retrospective case-series study that includes three adults with different clinical presentations of heart failure on grounds of myocarditis after initial COVID-19 infection. (3) Results: All patients had new-onset symptomatic HF of various severity: from a moderately reduced left ventricular ejection fraction in one patient to significantly reduced fractions in the remaining two. Moreover, complex ventricular arrhythmias were present in one case. All patients had confirmed past myocarditis in cardiac magnetic resonance. With optimal medical treatment, cardiac function improved, and the symptoms subsided in all cases. (4) Conclusions: In COVID-19 patients, long COVID myocarditis may be one of the severe complications of this acute disease. The heterogeneity in clinical symptoms and a paucity of specific diagnostic procedures expose the patient to the significant risk of misdiagnosing and further HF development.

Patterns of Childhood Tuberculosis Diagnosis in Hawassa University Comprehensive Specialized Hospital, Hawassa, Sidama Regional State, Ethiopia.

Because of the pauci bacillary nature of childhood tuberculosis and the difficulties in obtaining proper sputum samples from young children, diagnosing childhood tuberculosis (TB) is difficult. Childhood TB needs early identification and care since it advances swiftly to more advanced stages.This study was aimed to determine the patterns of all forms of childhood tuberculosis diagnosis at Hawassa University Comprehensive Specialized Hospital in Hawassa, Ethiopia. A retrospective cross-sectional study was conducted from February 1, 2017 to January 30, 2021 among 175 children diagnosed and treated for tuberculosis in the pediatric ward. Children medical charts and pediatrics ward logbook were used to extract pertinent data by structured checklists. SPSS version 23.0 was used for data entry and statistical analysis. Of 175 children, fever was the leading clinical symptoms and diagnosed in 166 (94.9%) children followed by weight loss (154, 88%), and cough (136, 77.7%). In twenty seven out of 88 (30.6%) children, gastric aspirate was positive for TB infection by Xpert MTB/Rif, while 3/40 (7.5%) were positive for TB using fine needle aspiration cytology (FNAC), 19/66 (28.8%) had suggestive TB by cerebrospinal fluid analysis (CSF), 10/29 (34.5%) were smear positive for TB and 70/162 (43.2%) were suspected for TB by chest X-ray. Despite recent breakthroughs in quick microbiological detection, such as Xpert MTB/Rif, this study revealed that more than half of the children, 89/175 (51%), were treated for TB diseases solely based on clinical criteria. This will significantly underestimate the true nature of the illness or disease and make them vulnerable to mistreatment. As a result, in order to appropriately treat the disease and manage patients in our settings, getting a microbiological diagnosis of childhood tuberculosis requires improvement, and we call for expanded availability and use of a more sensitive and specific diagnostic technique to circumvent these concerns.

Value-Based Healthcare, PROMs and Shared Decision-Making - How Are They Connected?

Value-Based Healthcare, PROMs and Shared Decision-Making - How Are They Connected? Abstract. The best possible healthcare delivery measured by the individually perceived value of a treatment at reasonable costs is the short formula of value-based healthcare (VBHC). Developed in 2006 by Harvard economists Michael Porter and Elizabeth Teisberg in their landmark book 'Redefining Healtcare' [1], this formula and its principles are currently under considerations as a comprehensive concept for transforming the healthcare system in many countries, including Switzerland. The aim is to focus on the value of a treatment from the patient's perspective as a measure for treatment quality. Standardized Patient-Reported Outcome Measurement (PROM) as a core element of the VBHC concept is the prime metric, which can only be optimized, when Shared Decision-Making (SDM) is an embedded part, thereby integrating the individually best choices regarding the 'value' (and harm!) of specific diagnostic and therapeutic medical procedures.

Artificial intelligence (AI) and recent advancements in periodontology

Periodontics is one of the important field in dentistry in which we come across various conditions that are related to oral and systemic health. Periodontal diagnosis is the pillar of successful treatment planning since prevention and proper treatment depend on the accuracy and precision of specific diagnostic techniques. Periodontics has evolved in recent years in the field of Nanotechnology, 3-D printing, stem cell therapy etc. Hence this article illustrates current practices and recent advancements in periodontology.

Malignant Breast Lesion Detection with MRI and its Accuracy

Objective: Using histology as the gold standard, this study aims to determine whether or not magnetic resonance spectroscopy can reliably diagnose malignant breast lesions. Study Design: Cross-sectional Place and Duration: Conducted at department of Radiology Central Park Medical College, Lahore. Duration was 6 months from October 2021 to March 2022. Methods: This research included 75 females. Malignancy-suspicious palpable lesions were included. Patients completed a consent form admitting the hazards before any data was gathered. The kinematics and morphology of dynamically enhanced magnetic resonance (MRS) images were studied. MRS's choline peak (Cho) was utilized to detect cancer. Single-voxel technique was employed to evaluate MRS' diagnostic accuracy in cancer. MRS and biopsies were compared. Data was analyzed using SPSS 24.0. Results: Malignant lesions were found in 65 (86.7%) of the patients by MRS, and in 58 (77.3%) of the cases according to the histological findings. Malignant lesions by MRS most frequently showed a ductal enhancement or a peripheral enhancement morphology. The results of MRS showed a 78% specificity, 85% accuracy, 90% sensitivity, 75% NPV, and 91% PPV. Conclusion: MRS must be employed because of its better specificity, sensitivity, and accuracy in characterizing breast lumps. MRS is a specific, sensitive, and effective diagnostic technique for breast cancer. Keywords: Accuracy, Magnetic resonance spectroscopy (MRS), Malignant breast lesion, Histopathology,

A Review of Diagnostic Techniques for Parkinson’s Disease

Parkinson's disease is the second most common progressive neurodegenerative disorder in older people. Idiopathic Parkinson's Disease is associated with risk factors such as aging, family history, pesticide exposure, and environmental chemicals (e.g., synthetic heroin use). It is caused by a pathophysiologic loss or degeneration of dopaminergic neurons in the substantia nigra of the midbrain and the development of neuronal Lewy Bodies. Its underlying causes are unknown. Despite significant advances in neurodegenerative disease research since James Parkinson's first medical description of Parkinson's disease in 1817, these disorders continue to pose significant diagnostic and treatment challenges. A valid diagnosis at early disease stages is critical because it can help accommodate differential prognostic and disease management approaches, elucidate reliable clinicopathological relationships ideally at prodromal stages, and would make it easier to evaluate innovative treatments in clinical trials. The pursuit of early diagnosis in Parkinson's disease and atypical Parkinsonian syndromes, however, is hampered by significant clinical and pathological heterogeneity, which can affect disease presentation and progression. Therefore, more accurate more specific diagnostic techniques are required to differentiate Parkinson’s disease from other neurological conditions.

​​Evaluation of Hematological and Serum Biochemical Alterations in Dogs Affected with Congestive Heart Failure

Background: Congestive heart failure (CHF) is common disorder of dogs that occur most frequently as a result of Dilated Cardiomyopathy (DCM), Degenerative Mitral Valve Disease and pericardial diseases. An early diagnosis can be carried out through the use of specific diagnostic techniques such as x-rays, echocardiography etc. The goal of this study was to see if dogs with CHF go through any substantial alterations in their hematology or serum biochemistry. Methods: The current investigation was conducted during the period April, 2020 to April 2021 at Veterinary Clinical Complex, Khanapara, Assam. Thirty four (n=34) dogs with CHF confirmed by echocardiography and eight (n=8) healthy dogs were considered for the study. Blood samples from all selected dogs were subjected to routine hematological and serum biochemical examination. Result: The hematological assessment revealed no significant alterations. During the course of the study, there were significant alterations in the serum biochemical parameters viz. LDH, BUN, creatinine, AST, ALT, total serum protein, albumin and albumin: globulin ratio. Findings of the study revealed that hematology and biochemistry were more useful in ruling out CHF or in identifying concurrent diseases than diagnosing CHF itself.

Graphene-based nanocomposite using new modeling molecular dynamic simulations for proposed neutralizing mechanism and real-time sensing of COVID-19

Abstract A new virus, the coronavirus (COVID-19), is causing serious respiratory infections in humans. Rapid, specific, and sensitive diagnostic techniques for early-stage detection of SARS-CoV-2 viral protein are developing as a necessary response for effective smart diagnostics, treatment optimization, and exploration of therapeutics with better effectiveness in the fight against the COVID-19 pandemic. Keeping the considerations mentioned above, we propose a new modeling graphene nanocomposite-based biosensing device for detecting COVID-19 at the site of the epidemic as the best way to manage the pandemic. It is important to address the problems of COVID-19 management. With the challenges and aspects of COVID-19 management in mind, we present in this review a collective approach involving electrochemical COVID-19 biosensing required for early-stage COVID-19 diagnosis and the direct interaction with viral surface glycoproteins and metal nanoparticles that can enter cells and neutralize viruses by interacting directly with the viral genome (ribonucleic acid), which identifies the COVID-19 spike protein and antiviral procedure including virus inactivation, host cell receptor inactivation, electrostatic entrapment, and physicochemical destruction of viral species by nucleotide ring opening. The interactions between the graphene composite and virus may be boosted by functionalization of the carbon surface and decoration of metallic components that enhance these interactions. Our proposed new modeling molecular dynamic simulation-based neutralizing mechanism and real-time detection of COVID-19 on graphene nanocomposite-based biosensors are suitable for point-of-care diagnostic applications, and this sensing platform can be modified for the early diagnosis of severe viral infections using real samples. For the potential application, the suggested one is the chemical reaction and bond breaking between the metallic component and molecule of COVID19 with computer simulation data.

Imaging Approach to Pulmonary Infections in the Immunocompromised Patient.

Pulmonary infections are the major cause of morbidity and mortality in immunocompromised patients and almost one-third of intensive care unit patients with pulmonary infections belong to the immunocompromised category. Multiple organisms may simultaneously infect an immunocompromised patient and the overwhelming burden of mixed infections further predisposes critically ill patients to acute hypoxemic respiratory failure. Notwithstanding that lung ultrasound is coming into vogue, the primary imaging investigation is a chest radiograph, followed by thoracic CT scan. This review based on our experience at tertiary care teaching hospitals provides insights into the spectrum of imaging features of various pulmonary infections occurring in immunocompromised patients. This review is unique as, firstly, the imaging spectrum described by us is categorized on basis of the etiological infective agent, comprehensively and emphatically correlated with the clinical setting of the patient. Secondly, a characteristic imaging pattern is emphasized in the clinical setting-imaging-pattern conglomerate, to highlight the most likely diagnosis possible in such a combination. Thirdly, the simulating conditions for a relevant differential diagnosis are discussed in each section. Fourthly, not only are the specific diagnostic and tissue sampling techniques for confirmation of the suspected etiological agent described, but the recommended pharmaco-therapeutic agents are also enumerated, so as to provide a more robust insight to the radiologist. Last but not the least, we summarize and conclude with a diagnostic algorithm, derived by us from the characteristic illustrative cases. The proposed algorithm, illustrated as a flowchart, emphasizes a diagnostic imaging approach comprising: correlation of the imaging pattern with clinical setting and with associated abnormalities in the thorax and in other organs/systems, which is comprehensively analyzed in arriving at the most likely diagnosis. Since a rapid evaluation and emergent management of such patients is of pressing concern not only to the radiologist, but also for the general physicians, pulmonologists, critical care specialists, oncologists and transplant surgery teams, we believe our review is very informative to a wide spectrum reader audience.

Hepatitis delta is a public health concern in the community setting: The role of prison health care units in limiting the spread of infection in general population

Hepatitis D virus (HDV) is a public health concern for its severe medical consequences. The aim of this pilot study was to evaluate the prevalence of the main infectious disease between the detainees of the Veneto Region. The study consisted of a survey carried between the heads of the prison healthcare staff. The data showed that on 2,119 detainees the prevalence of hepatitis B (HBV) is the 6% and that the most relevant prevalence of HBV was revealed in the East Europe detainees (Ex USSR countries, Albania, Bosnia, Kosovo, Macedonia), being the 39% of all HBcAb positive subjects. The survey also revealed as the population less protected by HVB vaccination was that comes from East Europe. Moreover, the study indicates that hepatitis D diagnosis is not well known by clinician working inside prisons and that it should be improved through specific diagnostic and therapeutic procedures. Our data indicate that the prison setting may play an important and crucial role in term of public health both in intercepting patients that need of treatment and in preventing actions able to minimize the risk of infection. In prison settings screening, linkage to care and harm reduction measures should be improved in order to protect the general population from the impact of the emerging infectious diseases, including HDV.

Sleeping Sickness: a congenital case after a possible parents sexual transmission

The authors identify a rare case of congenital infection by Trypanosoma brucei gambiense on a 21-year-old Afro-Brazilian boy, born in the USA, who was diagnosed at 16 months of age after returning to Portugal. At that moment, neither the patient nor his mother had ever been to Africa. He was admitted in our Paediatric Department under symptoms of a wasting syndrome, lymphadenopathies and intermittent fever. During the initial evaluation, he was diagnosed with anaemia, hypergammaglobulinemia and a positive serology for cytomegalovirus infection. A month later, the wasting syndrome had gotten worse and he was identified with daytime somnolence. At this point, his mother was diagnosed with HAT, which motivated specific diagnostic procedures in the child. A blood test revealed Trypanosomes, the CATT (Card Agglutination Test for Trypanosomes) was positive and a lumbar puncture was performed, which clearly established the diagnosis of late-stage HAT. Consequently, Electroencephalogram and brain MRI showed extensive brain damage and a Fundoscopic examination revealed spots of an atypical chorioretinitis, probably associated with the combined trypanosomal and cytomegalovirus infection. Treatment was established with intravenous eflornithine (DFMO). The patient tolerated the drug, which led to a gradual clinical improvement, including the complete regression of brain and retinal injuries. Two years after this treatment, parasitological cure was confirmed according to established criteria. At the moment, the patient remains well, and shows a normal psychomotor development. Diagnostic difficulties in pediatric sleeping sickness, including the elusive characteristics of the patient and his parents’ epidemiological history are highlighted.

Wave Intensity Analysis Combined With Machine Learning can Detect Impaired Stroke Volume in Simulations of Heart Failure.

Heart failure is treatable, but in the United Kingdom, the 1-, 5- and 10-year mortality rates are 24.1, 54.5 and 75.5%, respectively. The poor prognosis reflects, in part, the lack of specific, simple and affordable diagnostic techniques; the disease is often advanced by the time a diagnosis is made. Previous studies have demonstrated that certain metrics derived from pressure–velocity-based wave intensity analysis are significantly altered in the presence of impaired heart performance when averaged over groups, but to date, no study has examined the diagnostic potential of wave intensity on an individual basis, and, additionally, the pressure waveform can only be obtained accurately using invasive methods, which has inhibited clinical adoption. Here, we investigate whether a new form of wave intensity based on noninvasive measurements of arterial diameter and velocity can detect impaired heart performance in an individual. To do so, we have generated a virtual population of two-thousand elderly subjects, modelling half as healthy controls and half with an impaired stroke volume. All metrics derived from the diameter–velocity-based wave intensity waveforms in the carotid, brachial and radial arteries showed significant crossover between groups—no one metric in any artery could reliably indicate whether a subject’s stroke volume was normal or impaired. However, after applying machine learning to the metrics, we found that a support vector classifier could simultaneously achieve up to 99% recall and 95% precision. We conclude that noninvasive wave intensity analysis has significant potential to improve heart failure screening and diagnosis.

Peroxiredoxins as Markers of Oxidative Stress in IgA Nephropathy, Membranous Nephropathy and Lupus Nephritis

IgA nephropathy (IgAN), membranous nephropathy (MN), and lupus nephritis (LN) represent important causes of chronic kidney disease. They belong to the immune-mediated glomerulonephritis (GNs), and have distinct pathogenesis, distinct clinical courses, and variable responses to treatment. Therefore, specific diagnostic procedures are necessary for more effective patient management. Recently, a role for oxidative stress has been proposed in various renal disorders. Thus, molecules related to oxidative stress, such as 2-Cys-peroxiredoxins (PRDXs), may represent plausible candidates for biomarkers in renal pathologies. The aim of this study was to assess whether there are differences between individual GNs and healthy controls in the context of PRDXs serum concentration. We enrolled 108 patients with biopsy-proven IgAN (47), MN (26), LN (35) and 30 healthy age- and sex-matched controls. The serum concentrations of PRDX 1–5 were measured with ELISA assays and correlated with demographic and clinical data. The PRDXs’ concentration varied depending on the GN type. We also observed an association of PRDXs with lower estimated glomerular filtration rates, complement, hemoglobin, and body mass index. Our study indicates that individual PRDX can play roles in pathophysiology of selected GNs and that their serum concentrations may become useful as a new supplementary diagnostic markers in IgAN, MN as well as LN. The results of this study open a new avenue for prospective research on PRDXs in renal diseases.

Current status of the diagnosis of chronic pancreatitis by ultrasonographic elastography.

Chronic pancreatitis (CP) is pathologically characterized by the loss of exocrine pancreatic parenchyma, irregular fibrosis, cellular infiltration, and ductal abnormalities. Diagnosing CP objectively is difficult because standard diagnostic criteria are insufficient. The change of parenchymal hardness is the key factor for the diagnosis and understanding of the severity of CP. The ultrasonography (US) or endoscopic ultrasonography (EUS) elastography have been used to diagnose pancreatic diseases. Both strain elastography (SE) and shear wave elastography are specific diagnostic techniques for measuring tissue hardness. Most previous studies were conducted with SE. There are three methods of interpreting SE; the method of recognizing the patterns in SE distribution images in the region of interest, the method of using strain ratio to compare the hardness of adipose tissue or connective tissue with that of the lesion, and the method of evaluating the hardness distribution of a target by histogram analysis. These former two methods have been used primarily for neoplastic diseases, and histograms analysis has been used to assess hardness distribution in the evaluation of CP. Since the hardness of the pancreas increases with aging, it is necessary to consider the age in the diagnosis of pancreatic disorders using US or EUS elastography.

Commentary: Abdominal Ultrasound and Its Diagnostic Accuracy in Diagnosing Acute Appendicitis: A Meta-Analysis.

with the authors about the importance of the abdominal ultrasound, but we would like to pay attention to several important missing aspects in the article.First, in this meta-analysis (1), the author depicted that only studies adopting histopathology report as the reference standard were included. But the reference Finally, in this meta-analysis, the overall sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio were estimated at 77%, 60%, 2.62, and 0.45, respectively, demonstrating that abdominal ultrasound has high false-positive rate (40%) and should not be used for exclusion or inclusion of appendictis. But, in conclusion, the author demonstrated that abdominal ultrasound shows significant accuracy of diagnosis in patients with suspected acute appendicitis and is an effective diagnostic alternative to reduce the rate of unnecessary surgeries in acute appendicitis.We consider that the conclusion might be not appropriate. According to the overall results, patients suspected with appendictis should be referred to more sensitive and specific diagnostic procedures, such as computed tomography.

A Multi-Epitope Fusion Protein-Based p-ELISA Method for Diagnosing Bovine and Goat Brucellosis.

In recent years, the incidence of brucellosis has increased annually, causing tremendous economic losses to animal husbandry in a lot of countries. Therefore, developing rapid, sensitive, and specific diagnostic techniques is critical to control the spread of brucellosis. In this study, bioinformatics technology was used to predict the B cell epitopes of the main outer membrane proteins of Brucella, and the diagnostic efficacy of each epitope was verified by an indirect enzyme-linked immunosorbent assay (iELISA). Then, a fusion protein containing 22 verified epitopes was prokaryotically expressed and used as an antigen in paper-based ELISA (p-ELISA) for serodiagnosis of brucellosis. The multi-epitope-based p-ELISA was evaluated using a collection of brucellosis-positive and -negative sera collected from bovine and goat, respectively. Receiver operating characteristic (ROC) curve analysis showed that the sensitivity and specificity of detection-ELISA in diagnosing goat brucellosis were 98.85 and 98.51%. The positive and the negative predictive values were 99.29 and 98.15%, respectively. In diagnosing bovine brucellosis, the sensitivity and specificity of this method were 97.85 and 96.61%, with the positive and negative predictive values being identified as 98.28 and 97.33%, respectively. This study demonstrated that the B cell epitopes contained in major antigenic proteins of Brucella can be a very useful antigen source in developing a highly sensitive and specific method for serodiagnosis of brucellosis.

Diagnosing hospital bacteraemia in the framework of predictive, preventive and personalised medicine using electronic health records and machine learning classifiers

BackgroundThe bacteraemia prediction is relevant because sepsis is one of the most important causes of morbidity and mortality. Bacteraemia prognosis primarily depends on a rapid diagnosis. The bacteraemia prediction would shorten up to 6 days the diagnosis, and, in conjunction with individual patient variables, should be considered to start the early administration of personalised antibiotic treatment and medical services, the election of specific diagnostic techniques and the determination of additional treatments, such as surgery, that would prevent subsequent complications. Machine learning techniques could help physicians make these informed decisions by predicting bacteraemia using the data already available in electronic hospital records.ObjectiveThis study presents the application of machine learning techniques to these records to predict the blood culture’s outcome, which would reduce the lag in starting a personalised antibiotic treatment and the medical costs associated with erroneous treatments due to conservative assumptions about blood culture outcomes.MethodsSix supervised classifiers were created using three machine learning techniques, Support Vector Machine, Random Forest and K-Nearest Neighbours, on the electronic health records of hospital patients. The best approach to handle missing data was chosen and, for each machine learning technique, two classification models were created: the first uses the features known at the time of blood extraction, whereas the second uses four extra features revealed during the blood culture.ResultsThe six classifiers were trained and tested using a dataset of 4357 patients with 117 features per patient. The models obtain predictions that, for the best case, are up to a state-of-the-art accuracy of 85.9%, a sensitivity of 87.4% and an AUC of 0.93.ConclusionsOur results provide cutting-edge metrics of interest in predictive medical models with values that exceed the medical practice threshold and previous results in the literature using classical modelling techniques in specific types of bacteraemia. Additionally, the consistency of results is reasserted because the three classifiers’ importance ranking shows similar features that coincide with those that physicians use in their manual heuristics. Therefore, the efficacy of these machine learning techniques confirms their viability to assist in the aims of predictive and personalised medicine once the disease presents bacteraemia-compatible symptoms and to assist in improving the healthcare economy.