Pragmatic Algorithm Research Articles

A pragmatic algorithm to select appropriate antiseizure medications in patients with epilepsy.

Antiseizure medications (ASMs) are the first-line treatment for epilepsy. Many ASMs are available; this offers the opportunity to improve therapy by tailoring it to individual characteristics, but also increases the possibility of healthcare professionals making inappropriate treatment choices. To assist healthcare professionals, we developed a pragmatic algorithm aimed at facilitating medication selection for individuals whose epilepsy begins at age 10years and older. Utilizing available evidence and a Delphi panel-based consensus process, a group of epilepsy experts developed an algorithm for selection of ASMs, depending on the seizure type(s) and the presence of relevant clinical variables (age, gender, comorbidities, and comedications). The algorithm was implemented into a web-based application that was tested and improved in an iterative process. The algorithm categorizes ASMs deemed to be appropriate for each seizure type or combination of seizure types into three groups, with group 1 ASMs considered preferred, group 2 considered second line, and group 3 considered third line. Depending on the presence of relevant clinical variables, the ranking of individual ASMs is adjusted in the prioritization scheme to tailor recommendations to the characteristics of the individual. The algorithm is available on a web-based application at: https://epipick.org/#/. The proposed algorithm is user-friendly, requires less than 2minutes to complete, and provides the user with a range of appropriate treatment options from which to choose. This should facilitate its broad utilization and contribute to improve epilepsy management for healthcare providers who desire advice, particularly those who lack special expertise in the field.

Pragmatic Clinical Perspective on Biologics for Severe Refractory Type 2 Asthma

Patients with severe refractory asthma present a challenging clinical conundrum for practicing clinicians. Biologics that target key mediators in the type 2 inflammation cascade, including IL-4, IL-5, IL-13, and IgE, can be effective strategies for these patients. However, with various biologics available, choosing the optimal one for a particular patient becomes a nuanced decision. We propose a pragmatic algorithm that identifies the optimal biologic class for patients who have specific type 2 disease endotypes. Patients with eosinophilic endotypes fare well with anti-IL-5(rα) medications, comprising mepolizumab, benralizumab, and reslizumab because they have been shown to reduce exacerbations in severe eosinophilic asthma by approximately 50%. In patients with fractional exhaled nitric oxide-high endotypes, anti-IL-4rα such as dupilumab is deemed to be most effective and has demonstrated a 47% reduction in asthma exacerbations. For patients with severe uncontrolled allergic asthma, anti-IgE (omalizumab) is effective and has been shown to confer a 25% reduction in asthma exacerbations. Type 2 comorbidities including chronic rhinosinusitis with nasal polyps, atopic dermatitis, chronic idiopathic urticaria, and eosinophilic esophagitis are important to bear in mind before the prescription of biologics. Further head-to-head studies are indicated to compare biologics in patients with mixed endotypes according to peripheral blood eosinophils, fractional exhaled nitric oxide, and allergic status. The evidence strongly supports endotype-driven prescribing of biologics to achieve clinically relevant outcomes in severe refractory asthma and related comorbidities.

Diagnostic Performance of MRI in the Detection of Renal Lipid-Poor Angiomyolipomas: A Systematic Review and Meta-Analysis.

Background Lipid-poor angiomyolipomas (AMLs) are challenging to differentiate from other renal lesions at imaging and often necessitate biopsy or surgery. If sufficiently accurate, MRI may play a role as a replacement test for biopsy. Purpose To perform a systematic review to evaluate the diagnostic performance of MRI for lipid-poor AMLs in patients with renal masses. Materials and Methods A systematic review of MEDLINE, EMBASE, Scopus, the Cochrane Library, and the "gray literature" (conference proceedings) was performed without language restriction through July 18, 2019, with the assistance of a health sciences librarian. Original articles with more than 10 patients evaluating the diagnostic performance of MRI, with histopathologic findings used as the reference standard, for the diagnosis of lipid-poor AMLs in patients with renal masses were included. Studies including AMLs with macroscopic fat and studies with insufficient data were excluded. Patient, clinical, MRI, and diagnostic performance parameters were independently acquired by two authors. Meta-analysis was performed by using a random-effects or bivariate mixed-effects regression model depending on the number of studies. Risk of bias of individual studies was evaluated by using Quality Assessment of Diagnostic Accuracy Studies-2. Results Twenty-three studies with 2196 patients and 25 contingency tables were included. The pooled sensitivity, specificity, and area under the receiver operating characteristic curve were 83% (95% confidence interval [CI]: 72%, 90%), 90% (95% CI: 84%, 94%), and 0.93 (95% CI: 0.91, 0.95), respectively. Considerable variability was present for several variables, including MRI parameters; however, subgroup analysis did not identify MRI sequence or field strength as sources for variability. All studies were at high risk of bias for index test domain because no reported thresholds were prespecified. Conclusion MRI shows promising accuracy for detecting lipid-poor angiomyolipomas (area under the receiver operating characteristic curve, >0.9), indicating a potential role as a replacement test for biopsy in selected patients. Studies evaluating MRI accuracy with a pragmatic algorithm and prespecified threshold may be helpful to confirm this potential role in the management pathway. © RSNA, 2020 Online supplemental material is available for this article.

Practical Link Adaptation Algorithm With Power Density Offsets for 5G Uplink Channels

This letter proposes a pragmatic link adaptation algorithm considering power density offsets (PDOs) for next-generation uplink wireless channels. The proposed algorithm consists of PDO calculation between a physical uplink shared channel and its associated sounding reference signal, key channel state metric generation, and modulation and coding scheme (MCS) adaptation with respect to the PDO. Scaling is applied to estimated channel matrices based on multiple reference PDO points to generate corresponding reference mutual information (MI) values, followed by interpolation or extrapolation to obtain the adapted MI and ultimately MCS. The proposed algorithm has low complexity in terms of hardware implementation, while yielding satisfactory block error rates and throughput for a wide range of PDOs as shown by simulation results.

Defining major depressive disorder cohorts using the EHR: Multiple phenotypes based on ICD-9 codes and medication orders

BackgroundMajor Depressive Disorder (MDD) is one of the most common mental illnesses and a leading cause of disability worldwide. Electronic Health Records (EHR) allow researchers to conduct unprecedented large-scale observational studies investigating MDD, its disease development and its interaction with other health outcomes. While there exist methods to classify patients as clear cases or controls, given specific data requirements, there are presently no simple, generalizable, and validated methods to classify an entire patient population into varying groups of depression likelihood and severity. MethodsWe have tested a simple, pragmatic electronic phenotype algorithm that classifies patients into one of five mutually exclusive, ordinal groups, varying in depression phenotype. Using data from an integrated health system on 278,026 patients from a 10-year study period we have tested the convergent validity of these constructs using measures of external validation, including patterns of psychiatric prescriptions, symptom severity, indicators of suicidality, comorbidity, mortality, health care utilization, and polygenic risk scores for MDD. ResultsWe found consistent patterns of increasing morbidity and/or adverse outcomes across the five groups, providing evidence for convergent validity. LimitationsThe study population is from a single rural integrated health system which is predominantly white, possibly limiting its generalizability. ConclusionOur study provides initial evidence that a simple algorithm, generalizable to most EHR data sets, provides categories with meaningful face and convergent validity that can be used for stratification of an entire patient population.

Practical Evaluation and Management of Insomnia in Parkinson's Disease: A Review.

ABSTRACTBackgroundInsomnia is one of the most common nonmotor features of Parkinson's disease (PD). However, there are few practical guidelines for providers on how to best evaluate and treat this problem.Methods and FindingsThis review was developed to provide clinicians with a pragmatic approach to assessing and managing insomnia in PD. Recommendations were based on literature review and expert opinion. We addressed the following topics in this review: prevalence of insomnia in PD, sleep–wake mechanisms, theoretical models of insomnia, risk factors, assessment, pharmacologic and nonpharmacologic treatments. Insomnia treatment choices may be guided by PD severity, comorbidities, and patient preference. However, there is limited evidence supporting pharmacotherapy and nonpharmacologic treatments of insomnia in PD.ConclusionsWe provide a pragmatic algorithm for evaluating and treating insomnia in PD based on the literature and our clinical experience. We propose personalized insomnia treatment approaches based on age and other issues. Gaps in the existing literature and future directions in the treatment of insomnia in PD are also highlighted.

Iatrogenic intra-arterial injection in the upper limb: A pragmatic guide for the on-call vascular surgeon

Background: Peripheral venous cannulation is one of the most commonly performed procedures to establish venous access in the hospital setting. Inadvertent arterial cannulation is a rare event but can have serious consequences including ischemia and limb loss, especially if medication is administered through the device. Vascular surgeons should have an understanding of the management of this potential complication Methods: We reviewed the risk factors, mechanism, pathophysiology, and management options for inadvertent upper limb arterial cannulation and injection. Results: Inadvertent arterial cannulation and administration of medication has a reported incidence of 1 in 3440 cases. Several risk factors were identified including difficult venous access, difficulty in communication, and aberrant vascular anatomy. Both arterial injury from the misplaced cannula and injection of medication can cause ischemic events through a number of mechanisms. We recommend that patients sustaining an inadvertent arterial injury should have an urgent clinical assessment and be managed according to the degree of upper limb ischemia as per the Rutherford classification. We provide a pragmatic management algorithm to aid the vascular surgeon when encountering such a situation. Conclusion: Iatrogenic arterial injury is a rare but potentially serious complication of peripheral venous cannulation. Prompt recognition and management of ischemic complications is essential for favorable outcomes. Our pragmatic management guide should aid the vascular surgeon in managing this situation.

Data-driven Development of ROTEM and TEG Algorithms for the Management of Trauma Hemorrhage: A Prospective Observational Multicenter Study.

Developing pragmatic data-driven algorithms for management of trauma induced coagulopathy (TIC) during trauma hemorrhage for viscoelastic hemostatic assays (VHAs). Admission data from conventional coagulation tests (CCT), rotational thrombelastometry (ROTEM) and thrombelastography (TEG) were collected prospectively at 6 European trauma centers during 2008 to 2013. To identify significant VHA parameters capable of detecting TIC (defined as INR > 1.2), hypofibrinogenemia (< 2.0 g/L), and thrombocytopenia (< 100 x10/L), univariate regression models were constructed. Area under the curve (AUC) was calculated, and threshold values for TEG and ROTEM parameters with 70% sensitivity were included in the algorithms. A total of, 2287 adult trauma patients (ROTEM: 2019 and TEG: 968) were enrolled. FIBTEM clot amplitude at 5 minutes (CA5) had the largest AUC and 10 mm detected hypofibrinogenemia with 70% sensitivity. The corresponding value for functional fibrinogen (FF) TEG maximum amplitude (MA) was 19 mm. Thrombocytopenia was similarly detected using the calculated threshold EXTEM-FIBTEM CA5 30 mm. The corresponding rTEG-FF TEG MA was 46 mm. TIC was identified by EXTEM CA5 41 mm, rTEG MA 64 mm (80% sensitivity). For hyperfibrinolysis, we examined the relationship between viscoelastic lysis parameters and clinical outcomes, with resulting threshold values of 85% for EXTEM Li30 and 10% for rTEG Ly30.Based on these analyses, we constructed algorithms for ROTEM, TEG, and CCTs to be used in addition to ratio driven transfusion and tranexamic acid. We describe a systematic approach to define threshold parameters for ROTEM and TEG. These parameters were incorporated into algorithms to support data-driven adjustments of resuscitation with therapeutics, to optimize damage control resuscitation practice in trauma.

On the Orientation Average Based on Central Orientation Density Functions for Polycrystalline Materials

The present work continues the investigation first started by Lobos et al. (J. Elast. 128(1):17–60, 2017) concerning the orientation average of tensorial quantities connected to single-crystal physical quantities distributed in polycrystals. In Lobos et al. (J. Elast. 128(1):17–60, 2017), central orientation density functions were considered in the orientation average for fourth-order tensors with certain index symmetries belonging to single-crystal quantities. The present work generalizes the results of Lobos et al. (J. Elast. 128(1):17–60, 2017) for the orientation average of tensors of arbitrary order by presenting a clear connection to the Fourier expansion of central orientation density functions and of the general orientation density function in terms of tensorial texture coefficients. The closed form of the orientation average based on a central orientation density function is represented in terms of the Fourier coefficients (referred to as texture eigenvalues) and the central orientation of the central orientation density function. The given representation requires the computation of specific isotropic tensors. A pragmatic algorithm for the automated generation of a basis of isotropic tensors is given. Applications and examples are presented to show that the representation of the orientation average offers a low-dimensional parametrization with major benefits for optimization problems in materials science. A simple implementation in Python 3 for the reproduction of all examples is offered through the GitHub repository https://github.com/mauricio-fernandez-l/centralODF-average.

A Pragmatic Clinical Algorithm for the Prediction of Severe Neonatal Morbidity in Preterm Infants

Background: Severe neonatal morbidity (SNM) in preterm infants is now increasingly recognised as a major public health concern due to its strong correlation with not only neonatal mortality but also longer term disability. Therefore, we aimed to investigate predictors for severe neonatal morbidity (SNM) in preterm infants and to develop a predictive model for SNM utilising a range of maternal, pregnancy, intrapartum and neonatal variables. Methods: We compared SNM across demographic, antepartum, intrapartum, gestational age and birthweight centile categories. We performed robust univariable and multivariable binomial logistic regression analyses to identify major predictors. We developed a predictive model using the multivariable regression output and evaluated its internal validity using the K-fold cross validation technique. Model performance parameters were used to evaluate the diagnostic accuracy of the final model. Results: Of the total preterm infants 74·3% (4785/6440) had SNM. The final adjusted analysis showed that antepartum haemorrhage (aOR 1·18, 95% CI: 1·02-1·36), preterm premature rupture of membranes (aOR 1·37, 95% CI: 1·21-1·55), very preterm birth (aOR 3·79, 95% CI: 3·33-4·30), moderate preterm birth (aOR 4·17, 95% CI: 3·50-4·97), instrumental delivery (aOR 1·40, 95% CI: 1·20-1·62), elective caesarean section (aOR 1·46, 95% CI: 1·29-1·66), caesarean section for non-reassuring fetal status (aOR 2·08, 95% CI: 1·75-2·47), BW <5th centile (aOR 1·79, 95% CI: 1·36-2·37) and BW 5th-<10th centile (aOR 1·52, 95% CI: 1·25-1·84) were significant predictors for SNM. The final model had a sensitivity of 71·8% and specificity of 76% with a positive predictive value of 89·6%. Conclusion and relevance: Our study demonstrates that antepartum haemorrhage, premature rupture of membrane, gestational age, birth weight, operative birth and specific indications for operative birth are significant predictors of SNM. Clinicians may use identified risks to counsel mothers and priority setting in clinical decision making process to optimise neonatal outcomes. Funding Statement: Research support from the University of Queensland and the Mater Foundation. Declaration of Interests: The authors report no conflicts of interest. Ethics Approval Statement: The study was approved by the institution’s Human Research Ethics Committee (HREC/18/MHS/46).

A Pragmatic Algorithm for Attack Prevention on Semantic Web Services

In this digital age, semantic Web services have emerged as services that are used in every field. Today various types of services like e-commerce, social networking, and online payment of various services, etc., are used by the user through the Semantic Web. The Semantic Web has emerged as a powerful medium for using online services. The use of semantic Web services greatly simplified the life of the common citizen. The database is used to hoard information related to various services provided by Semantic Web Services. It turns out to be indispensable that when a user communicates with the database then there should not be any interference in it. The expansion of semantic Web services led various assailants to start making various efforts to stop these services. The network assail for instance Man-in-middle or DoS assail has adverse effect on the performance of the Web service. This paper insinuates a principle server of ARP algorithm that uses IP-MAC trussing approach on the network to prevent Man-in-the-middle and DoS assail within a entreaty succumbed to the Web service. Aftermath revelations that after implementing the algorithms, the number of packet loss reduces drastically and enhancing the overall performance of Web service. The staple objective of this paper is how the services bestowed by the Web service can be protected from assailant.

Algorithm-based preventive strategies for avoidance of residual neuromuscular blocks

Postoperative residual neuromuscular block (train of four ratio <0.9) is an outcome-relevant problem in balanced anesthesia, which increases postoperative morbidity and mortality. Implementation of medium and short-acting muscle relaxants, quantitative neuromuscular monitoring and pharmacological reversal of muscle relaxation reduce the incidence of residual neuromuscular block. The question arises whether this is suitable to create apragmatic algorithm integrating these three individual methods to reduce paralysis-associated complications? A selective literature search was carried out in PubMed and guidelines of national specialist societies were searched using special terms. The incidence of residual neuromuscular block varied among the studies but was as high as 93%. Using calibrated acceleromyography it is possible to identify a residual relaxation with anegative predictive value of 97% (95% confidence interval, CI 83-100%). Reversal by administration of the acetylcholinesterase inhibitor neostigmine reduces the incidence of residual neuromuscular block to 15.4%, in combination with calibrated acceleromyography to 3.3%. Reversal with sugammadex can nearly completely eliminate residual neuromuscular block. Quantitative neuromuscular monitoring and pharmacological reversal can be suitably integrated into astage-based, pragmatic treatment algorithm. The algorithm-based concept of quantitative neuromuscular monitoring and pharmacological reversal using neostigmine and sugammadex enables residual neuromuscular block to be treated within 10 min before extubation of the patient. Ongoing educational programs are essential for implementation of modern muscle relaxant management concepts. Quantitative neuromuscular monitoring should be mandatory for all patients receiving neuromuscular blocking drugs. It would be desirable that German-speaking societies for anesthesiology make appropriate recommendations in guidelines.

The logical and pragmatic algorithm of inferring euphemistic implicatures in the English and Ukrainian languages

Abstract. This article focuses on euphemistic implicatures in the English and Ukrainian languages inferred from the logical and pragmatic aspects. The general scientific methods, pragmatic analysis , m ethods of formal logics and model creating are used in order to infer euphemistic implicatures. A logical and pragmatic model is chosen as tertium comparationis of the research . The logical structure of euphemistic implicatures in English and Ukrainian can be modeled in a manner that is isomorphic for the two contrasted languages under analysis. A new logical and pragmatic algorithm designed for inferring euphemistic implicature has been elaborated. The logical operations of implication and conjunction have been applied to infer implicatures from euphemisms in the English and Ukrainian languages.

Sphincter repair or Sacral Nerve Modulation: Still debatable?

Fecal incontinence (FI) is a debilitating condition with negative consequences to patients. It is recognized as a “quality of life” illness. Sphincter tears resulting often from difficult childbirth or from surgical trauma, are well-identified causes of FI. When conservative measures fail to resolve FI symptoms, a surgical treatment is recommended on the basis of a comprehensive pelvic floor work-up. A sphincter tear is frequently found on endoanal ultrasonography. The best way to surgically approach a patient with FI related/associated with a sphincter lesion is still debatable. International guidelines are inconsistent regarding the role of either sacral neuromodulation (SNM) or anal sphincteroplasty (anal sphincter repair) (AS) in patients with anal sphincter defect. Decision making for an individual patient often relies on expert opinion and personal experience due to the poor quality of the few published studies. Currently the presence of a sphincter defect is no longer considered a contraindication for sacral nerve modulation (SNM) which has several advantages. SNM is a minimally invasive procedure with very low morbidity, its results can be accurately predicted with a test phase, and are sustained with long-term placement of the implant. Nevertheless, AS has clearly a role to play, for instance in young female patients reluctant to get an implanted nerve stimulator early in life and/or in case of a cloacae-like deformity as a sequela of a post-obstetrical 4th degree tear. While waiting for prospective studies directly comparing SNM and AS for various types of FI, we propose in this paper a pragmatic treatment algorithm based on the most recently published studies and recommendations for the management of sphincter defect related/associated FI.

Development of an algorithm to discriminate between plasmid- and chromosomal-mediated AmpC β-lactamase production in Escherichia coli by elaborate phenotypic and genotypic characterization.

ObjectivesAmpC-β-lactamase production is an under-recognized antibiotic resistance mechanism that renders Gram-negative bacteria resistant to common β-lactam antibiotics, similar to the well-known ESBLs. For infection control purposes, it is important to be able to discriminate between plasmid-mediated AmpC (pAmpC) production and chromosomal-mediated AmpC (cAmpC) hyperproduction in Gram-negative bacteria as pAmpC requires isolation precautions to minimize the risk of horizontal gene transmission. Detecting pAmpC in Escherichia coli is challenging, as both pAmpC production and cAmpC hyperproduction may lead to third-generation cephalosporin resistance.MethodsWe tested a collection of E. coli strains suspected to produce AmpC. Elaborate susceptibility testing for third-generation cephalosporins, WGS and machine learning were used to develop an algorithm to determine ampC genotypes in E. coli. WGS was applied to detect pampC genes, cAmpC hyperproducers and STs.ResultsIn total, 172 E. coli strains (n=75 ST) were divided into a training set and two validation sets. Ninety strains were pampC positive, the predominant gene being blaCMY-2 (86.7%), followed by blaDHA-1 (7.8%), and 59 strains were cAmpC hyperproducers. The algorithm used a cefotaxime MIC value above 6 mg/L to identify pampC-positive E. coli and an MIC value of 0.5 mg/L to discriminate between cAmpC-hyperproducing and non-cAmpC-hyperproducing E. coli strains. Accuracy was 0.88 (95% CI=0.79–0.94) on the training set, 0.79 (95% CI=0.64–0.89) on validation set 1 and 0.85 (95% CI=0.71–0.94) on validation set 2.ConclusionsThis approach resulted in a pragmatic algorithm for differentiating ampC genotypes in E. coli based on phenotypic susceptibility testing.

Development and validation of a pragmatic natural language processing approach to identifying falls in older adults in the emergency department

BackgroundFalls among older adults are both a common reason for presentation to the emergency department, and a major source of morbidity and mortality. It is critical to identify fall patients quickly and reliably during, and immediately after, emergency department encounters in order to deliver appropriate care and referrals. Unfortunately, falls are difficult to identify without manual chart review, a time intensive process infeasible for many applications including surveillance and quality reporting. Here we describe a pragmatic NLP approach to automating fall identification.MethodsIn this single center retrospective review, 500 emergency department provider notes from older adult patients (age 65 and older) were randomly selected for analysis. A simple, rules-based NLP algorithm for fall identification was developed and evaluated on a development set of 1084 notes, then compared with identification by consensus of trained abstractors blinded to NLP results.ResultsThe NLP pipeline demonstrated a recall (sensitivity) of 95.8%, specificity of 97.4%, precision of 92.0%, and F1 score of 0.939 for identifying fall events within emergency physician visit notes, as compared to gold standard manual abstraction by human coders.ConclusionsOur pragmatic NLP algorithm was able to identify falls in ED notes with excellent precision and recall, comparable to that of more labor-intensive manual abstraction. This finding offers promise not just for improving research methods, but as a potential for identifying patients for targeted interventions, quality measure development and epidemiologic surveillance.

Detection of complicated ectopic pregnancies in the hospital discharge database: A validation study.

ObjectiveComplicated ectopic pregnancies with severe bleeding (CEPSB) are life-threatening situations and should be considered maternal near-miss cases. Previous studies have found an association between severe maternal morbidity secondary to CEPSB and substandard care. Almost all women with CEPSB are hospitalized, generating administrative and medical records. The objective of this study was to propose a method to measure the validity of the hospital discharge database (HDD) to detect CEPSB among hospital stays in two gynecological units.MethodsWe included all hospital stays of women who were 18–45 years old and hospitalized for acute pelvic pain or/and metrorrhagia in the two hospitals. The HDD was compared to medical data (gold standard). Two algorithms constructed from the International Classification of Disease (ICD-10) and Common Classification of Medical Procedures (CCAM), were applied to the HDD: a “predefined algorithm” according to coding guidelines and a “pragmatic algorithm” based on coding practices. Sensitivity, specificity and positive likelihood-ratios were calculated. False negatives and positives were analyzed to describe coding practices.ResultsAmong 370 hospital stays included, 52 were classified as CEPSB cases. The “predefined algorithm” gave a sensitivity of 23.1% (95% CI: 11.6–34.5) and a specificity of 99.1% (95% CI: 98.0–100.0) to identify CEPSB. The “pragmatic algorithm” gave a sensitivity of 63.5% (95% CI: 50.4–76.5) and a specificity of 94.7% (95% CI: 92.2–97.5) to identify CEPSB. Coding errors (77.6%) were due to misuse of diagnosis codes and because complications were not coded.ConclusionHDD is not reliable enough to detect CEPSB due to incorrect coding practices. However, it could be an ideal tool to monitor quality of care if a culture in data quality assessment is developed to improve quality of medical information.

Predictors of Unlikely Bacterial Pneumonia and Adverse Pneumonia Outcome in Children Admitted to a Hospital in Central Vietnam.

Pneumonia is the leading cause of antibiotic use and hospitalization in Vietnam. There is a need for better prediction of unlikely bacterial pneumonia and adverse pneumonia outcome in order to guide hospital admission and improve rational antibiotic use. All children under 5 admitted with pneumonia (per clinician assessment) to the Da Nang Hospital for Women and Children were prospectively enrolled. Children were classified as having likely or unlikely bacterial pneumonia and followed for outcome assessment. A Bayesian model averaging approach was used to identify predictors of unlikely bacterial pneumonia and adverse pneumonia outcome, which guided the development of a pragmatic management algorithm. Of 3817 patients assessed, 2199 (57.6%) met World Health Organization (WHO) pneumonia criteria. In total, 1594 (41.7%) children were classified as having unlikely and 129 (3.4%) as having likely bacterial pneumonia. The remainder (2399; 62.9%) were considered to have disease of uncertain etiology. Factors predictive of unlikely bacterial pneumonia were no fever, no consolidation on chest radiograph, and absolute neutrophil count <5 × 109/L at presentation, which had a negative predictive value (NPV) for likely bacterial pneumonia of 99.0%. Among those who met WHO pneumonia criteria, 8.6% (189/2199) experienced an adverse outcome. Not having any WHO danger sign or consolidation on chest radiograph had an NPV of 96.8% for adverse pneumonia outcome. An algorithm that screens for predictors of likely bacterial pneumonia and adverse pneumonia outcome could reduce unnecessary antibiotic use and hospital admission, but its clinical utility requires validation in a prospective study.

New sequential combinations of non-invasive fibrosis tests provide an accurate diagnosis of advanced fibrosis in NAFLD

Advanced liver fibrosis is an important diagnostic target in non-alcoholic fatty liver disease (NAFLD) as it defines the subgroup of patients with impaired prognosis. The non-invasive diagnosis of advanced fibrosis is currently limited by the suboptimal positive predictive value and the grey zone (representing indeterminate diagnosis) of fibrosis tests. Here, we aimed to determine the best combination of non-invasive tests for the diagnosis of advanced fibrosis in NAFLD. A total of 938 patients with biopsy-proven NAFLD were randomized 2:1 into derivation and validation sets. All patients underwent liver stiffness measurement with vibration controlled transient elastography (VCTE) and blood fibrosis tests (NAFLD fibrosis score, Fibrosis-4 [FIB4], Fibrotest, Hepascore, FibroMeter). FibroMeterVCTE, which combines VCTE results and FibroMeter markers in a single test, was also calculated in all patients. For the diagnosis of advanced fibrosis, VCTE was significantly more accurate than the blood tests (area under the receiver operating characteristic curve [AUROC]: 0.840 ± 0.013, p ≤0.005). FibroMeter was the most accurate blood test (AUROC: 0.793 ± 0.015, p ≤0.017). The combinatory test FibroMeterVCTE outperformed VCTE and blood tests (AUROC: 0.866 ± 0.012, p ≤0.005). The sequential combination of FIB4 then FibroMeterVCTE (FIB4-FMVCTE algorithm) or VCTE then FibroMeterVCTE (VCTE-FMVCTE algorithm) provided an excellent diagnostic accuracy of 90% for advanced fibrosis, with liver biopsy only required to confirm the diagnosis in 20% of cases. The FIB4-FMVCTE and VCTE-FMVCTE algorithms were significantly more accurate than the pragmatic algorithms currently proposed. The sequential combination of fibrosis tests in the FIB4-FMVCTE and VCTE-FMVCTE algorithms provides a highly accurate solution for the diagnosis of advanced fibrosis in NAFLD. These algorithms should now be validated for the diagnosis of advanced liver fibrosis in diabetology or primary care settings. The evaluation of liver fibrosis is mandatory in non-alcoholic fatty liver disease (NAFLD), as advanced fibrosis identifies the subgroup of patients with impaired prognosis. FibroMeterVCTE is a new fibrosis test combining blood markers and the result of vibration controlled transient elastography (VCTE) into a single diagnostic test. Our results show that FibroMeterVCTE outperforms other blood fibrosis tests and VCTE alone for the diagnosis of advanced fibrosis in a large multi-centric cohort of 938 patients with biopsy-proven NAFLD. Sequential algorithms using a simple blood test or VCTE as a first-line procedure, then FibroMeterVCTE as a second-line test accurately classified 90% of patients.

PCN242 APPROACHES TO PRAGMATIC METHODOLOGY IN ONCOLOGY RETROSPECTIVE STUDIES

Increasing focus has been placed on real-world data (RWD) to address applicability of clinical trial results to a real-world population. However, in 2018, ASCO identified several gaps that affect the utility of RWD for comparative analyses in oncology patients. We describe over-arching gaps and propose practical methodological handling strategies. Gaps common to oncology RWD are: 1) patient risk-stratification based on variables not linked to diagnoses (ICD)-9/-10 coding (ie, cytogenetics/disease stage) [Case 1, below], and 2) identification of treatments, including determination of lines of therapy (LOT), as many agents are given in combination regimens [Case 2]. Besides these gaps, inherent bias within non-randomized studies further compound comparison validity and RWD utility. Utilizing proxies and employing a patient-level match for confounding variables ensures appropriate comparisons between oncology patient groups [Case 3]. Case 1, risk stratification and non-coded variables: In EHR data, identification of a “high-risk” population of patients (eg, a rare hematologic malignancy) is challenging as risk-stratification is based on unavailable data within an EHR. To overcome this, risk is calculated as a composite score of existing data supplemented with population-based estimates. Case 2, LOT identification: In EHR/claims data, LOT determination is critical for outcomes evaluations but difficult to ascertain. In an example of myeloma patients, pre-validated algorithmic approach was undertaken based on national treatment/monitoring guidelines and clinical input. Case 3, comparability of study groups in oncology: Outcomes between cohorts are influenced by patient-level variables impacting treatment choice and monitoring. Utilizing proxies and employing a patient-level match for confounding variables, two matching methods were compared. While propensity score matching retained more patients, a direct match more conservatively ensured confounding covariates are matched across cohorts. Challenges associated with comparative analyses in retrospective oncology studies can be overcome by leveraging pragmatic algorithms collaboratively derived from variables available in RWD.