A Theory-Based Ferritin-Procalcitonin Ratio Differentiates COVID-19 Pneumonia vs Bacterial Pneumonia

A Theory-Based Ferritin-Procalcitonin Ratio Differentiates COVID-19 Pneumonia vs Bacterial Pneumonia | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A Theory-Based Ferritin-Procalcitonin Ratio Differentiates COVID-19 Pneumonia vs Bacterial Pneumonia Leland Shapiro, Jorge L Salinas, Guillermo Rodriguez-Nava, Sa Shen, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5581463/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Rapid and inexpensive biomarker-based clinical instruments that can diagnose infectious diseases are desired, but developing clinical instruments has proved challenging. Proliferation of large clinical databases and expansive computational capability risks uncovering spurious associations that cannot be reproduced. Objectives : We present an approach to biomarker instrument creation that may enhance clinical applicability. We prospectively derived a biomarker instrument from a theoretical model of infection pathogenesis. Our theory-derived ferritin/procalcitonin (ferritin/PCT) ratio was designed to differentiate Coronavirus Disease 2019 (COVID-19) pneumonia from bacterial pneumonias. Materials and Methods : We assessed this ratio in over 30,000 patients in the TrinetX global database containing over 200 million patients. Results : Ferritin/PCT was significantly increased in COVID-19 pneumonia patients compared to bacterial pneumonia pateints. Ferritin/PCT accuracy for separating pneumonia due to COVID-19 vs Pneumococcus was assessed by calculating area under Receiver Operating Characteristic curve, which revealed a value of 0. 812. Conclusions : The ferritin/PCT ratio may have clinical use for differentiating COVID-19 pneumonia vs Pneumococcal pneumonia. Calculating the ferritin/PCT ratio is easy, rapid, and inexpensive. Clinical utility in resource-poor locations is an especially attractive application. Moreover, the conceptual model of infection pathogenesis that underlies this ratio may have broad applicability to differentiate other viral from bacterial infections. Pneumonia Infection pathogenesis COVID-19 Ferritin Procalcitonin Ferritin-Procalcitonin ratio Biomarker Theory Figures Figure 1 Introduction Rapid, inexpensive diagnosis of Coronavirus Disease 2019 (COVID-19) pneumonia and differentiation from bacterial pneumonia was needed early in the COVID-19 pandemic and remains essential in low-resource settings. We devised a novel diagnostic index consisting of the ferritin/procalcitonin (ferritin/PCT) ratio that was based on a model of pathogenesis that differentiates viral from bacterial infections ( 1 ). Our theoretical model proposes that viral infections like COVID-19 are characterized by relatively high levels of cell death or cytolysis and concomitant low amounts of inflammation. Conversely, bacterial infections like Pneumococcal pneumonia are characterized by relatively low degrees of cytolysis and high amounts of inflammation. Prior investigation suggests the most accurate readily available measure that can quantify cytolysis is serum ferritin ( 2 ). Inflammation can be quantified using blood levels of PCT, which is a readily-available established biomarker for inflammation that is significantly increased in bacterial infections compared to viral infections ( 3 – 5 ). In a prior report we showed ferritin/PCT could differentiate COVID-19 from bacterial pneumonias ( 1 ). In this report we assessed ferritin/PCT capacity to differentiate pneumonia caused by COVID-19 vs pneumococcus in a much larger and diverse validation cohort. A separate goal of this study was to assess a specific approach to biomarker instrument creation. Successful validation of the ferritin/PCT ratio would provide powerful support for our approach to biomarker development that emphasizes theoretical rationale prior to assessing statistical association between proposed biomarkers and diseases in clinical datasets. In contrast, we believe biomarker construction based on unrestrained mathematical evaluation of datasets presents high risk for generating random (non-causal) associations ( 6 – 10 ). In such cases of ransacking data, biomarker performance has high risk for failing to replicate in prospective clinical assessment. Materials and Methods In the TrinetX database ( https://trinetx.com/clinical-trial-design-optimization/#s_0 ), we identified COVID-19 pneumonia patients by requiring both diagnostic ICD-10 codes and a positive genetic or antigen test for the presence of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) March 2020 through August 2023. Patients with Streptococcus pneumoniae (bacterial) pneumonia were identified by requiring both diagnostic ICD-10 codes and a positive respiratory sample culture or body fluid specific antigen test May 2009 through July 2023. Patients with concomitant COVID-19 and pneumococcal pneumonia were excluded. The ferritin/PCT ratio was calculated using values for these 2 molecules in clinical blood samples obtained within seven days from recorded pneumonia diagnosis. The ferritin/PCT ratio numerical values for pneumonia caused by SARS-CoV-2 vs pneumococcal pneumonia were compared. Usefulness of the ferritin/PCT ratio as a separator of pneumonia caused by SARS-CoV-2 vs Pneumococcus was assessed by comparing mean ferritin/PCT values for COVID-19 and Pneumococcus . We assessed accuracy of ferritin/PCT for differentiating COVID-19 pneumonia vs pneumococcal pneumonia by generating the Receiver-Operating Characteristic (ROC) curve for the data and calculated the area under the ROC curve. A separate ROC curve was constructed that was adjusted for age, gender, ethnicity, race, marital status, region, and Charleson comorbidity index and the area under this ROC was calculated. Using the adjusted ROC curve, the optimal numerical cut point for the ferritin/PCT ratio was determined and the sensitivity and specificity for differentiating COVID-19 and Pneumococcal pneumonia was determined for this cut point. Results Table 1 shows clinical characteristics, ferritin, procalcitonin, and ferritin/PCT ratio values in patients with COVID-19 pneumonia or pneumococcal pneumonia. Mean serum ferritin determinations for COVID-19 patients and Pneumococcal pneumonia patients were 1,052. 9 ng/mL and 803. 5 ng/mL, respectively (p < 0. 001). Mean serum PCT in COVID-19 pneumonia patients and pneumococcal pneumonia patents were 1. 5 ng/mL and 10. 3 ng/mL, respectively (p < 0. 001). Ferritin/procalcitonin ratio significantly differed between the two diagnoses with mean values 5,914. 4 for COVID-19 pneumonia and 1,439. 2 f or pneumococcal pneumonia (p < 0. 001). Figure 1 shows unadjusted and adjusted ROC curves for diagnosing COVID-19 pneumonia vs bacterial pneumonia. Area under the ROC curve adjusted for age, gender, ethnicity, race, marital status, region, and Charleson comorbidity index was 0. 812. Using the Liu method, the optimal ferritin/PCT value for differentiation is 1,158. 14, with sensitivity and specificity 0. 73. Table 1 Clinical Information and Outcomes in Patients with COVID-19 Pneumonia or Pneumococcal Pneumonia COVID-19 pneumonia (N = 29,348) Pneumococccal Pneumonia (N = 1,921) p-value Demographic Characteristics Age* (sd) 62. 2 (16.2) 58. 9 (17.9) < 0.001 Sex Male Female 13,383 (45.6%) 15,963 (54.4%) 1,007 (52.4%) 914 (47.6%) < 0.001 Race White Black Unknown Native Hawaiian American Indian Asian 20,100 (68.5%) 4,731 (16.1%) 2,569 (8. 8%) 505 (1.7%) 114 (0.4%) 1,329 (4.5%) 1,223 (63.7%) 398 (20.7%) 197 (10.3%) 14 (0.7%) 32 (1.7%) 57 (3.0%) < 0.001 Associated Diseases Heart failure 9,787 (33.4%) 872 (45.7%) < 0.001 COPD 7,669 (26.1%) 816 (42.8%) < 0.001 CKD 10,084 (34.4%) 748 (39 2%) < 0.001 Liver disease 1,983 (6.8%) 271 (14.2%) < 0.001 DM2 13,243 (45.1%) 723 (37.9%) < 0.001 Neoplasm 5,836 (19.9%) 531 (27.8%) < 0.001 HIV 228 (0.8%) 103 (5.4%) < 0.001 Laboratory Information WBC x 10 − 3 /µL 8.5 (9.2) 12. 6 (10.0) < 0.001 Platelets x 10 − 3 /µL 218.2 (100.8) 238. 3 (155.5) < 0.001 Hemoglobin (g/dL) 12.6 (2.5) 10. 5 (2.3) < 0.001 Lymphocytes x 10 − 3 /µL 5.4 (92.2) 7. 0 (80.3) 0.755 Creatinine (mg/dL) 1.5 (2.1) 1. 7 (2.1) < 0.001 ALT (U/L) 46.3 (140.9) 64. 2 (315.0) < 0.001 AST (U/L) 62.8 (207.2) 91. 9 (410.8) < 0.001 CRP (mg/L) 84.7 (78.7) 110. 8 (111.1) < 0.001 PCT (ng/mL) 1.5 (13.1) 10. 3 (32.6) < 0.001 Ferritin (ng/mL) 1,052.9 (1985.9) 803. 5 (2187.6) < 0.001 Ferritin/PCT 5,914.4 (9274.1) 1,439. 2 (3206.2) < 0.001 Location of Care and Follow-up ‘Follow-up (days) 540 9 (494.8) 931. 0 (849.2) < 0.001 Hospitalization 7,485 (25.5%) 621 (32.3%) < 0.001 ICU admission 7,599 (25.9%) 643 (33.5%) < 0.001 Mortality at 1-year 7,232 (26.9%) 455 (30.7%) 0.001 *All values reported as mean with standard deviation (sd) COPD = chronic obstructive pulmonary disease CKD = chronic kidney disease CRP- C-reactive protein DM2 = type 2 diabetes HIV = human immunodeficiency virus infection ICU = Intensive Care Unit Ln = natural logarithm PCT = procalcitonin U = units WBC = white blood count Discussion This report confirms significant discriminatory power for the ferritin/PCT instrument to differentiate COVID-19 pneumonia from bacterial pneumonia in retrospective observational databases ( 1 ). This report is the most convincing and accurate to date due to the large number of subjects studied and the diversity of subjects. Early in the COVID-19 pandemic, there was need to diagnose pneumonias caused by SARS-CoV-2 vs bacterial etiologies rapidly. Without diagnostic tests that could identify SARS-CoV-2, up to 72% of COVID-19 patients were administered empiric broad-spectrum antibiotic therapy ( 11 ). This kind of antibiotic overuse can incur adverse effects and increase costs of care. To address the clinical problem of diagnosing COVID-19 vs bacterial infection at the beginning of the COVID-19 pandemic, we created a novel diagnostic biomarker instrument using a theory-based (mechanistic) approach. Two considerations motivated pursuing a theory-based approach. First, little clinical information was available early in the pandemic that could be used to statistically link biomarkers to diagnosis. Second, we were concerned about the problem of false-positive associations that can arise from unrestrained assessment of databases using powerful computational algorithms ( 6 – 10 ). We additionally constrained candidate theoretical mechanisms to those capable of immediate translation to clinical usefulness using readily available “off-the-shelf” biomarkers. Based on understanding of microbial pathogenesis, we created a theory of infection that focused on 2 characteristics that could potentially differentiate viral from bacterial etiologies. These 2 characteristics were cytolysis and inflammation. Specifically, we theorized that viral infections were characterized by large amounts of host cell destruciuon and relatively low levels of inflammation. In contrast, our model proposed lower amounts of cytolysis and higher levels of inflammation in cases of bacterial infections. Regarding cytolysis, viral infections often produce host cell destruction due to the necessary replication of inside cells and cell lysis to disseminate and initiate fresh infections ( 12 ). Bacteria have no such requirement. Clinical application of this component of our theory required a readily available marker for cell lysis. Surprisingly, circulating ferritin can be used to quantify cytolysis since ferritin has no established non-lytic mechanism to exit cells ( 2 ). Ferritin is often mischaracterized as a typical acute-phase reactant. Assessment of background knowledge in developing our theory revealed that SARS-CoV-2 replication induced cytolysis in vitro ( 13 , 14 ). Also, clinical observations showed elevated ferritin associated with COVID-19 diagnosis and disease severity ( 15 ). These observations suggested COVOD-19 causes cytolysis and ferritin quantifies the amount of cytolysis in cases of COVID-19. As predicted, high ferritin levels in COVID-19 patients were observed in our study, and significantly lower ferritin levels were observed in pneumococcal infections (Table 1 ) . To quantify inflammation, we chose PCT as the best readily available indicator. Procalcitonin is increased by prototypical cytokine mediators of inflammation tumor necrosis factor-alpha (TNF) and interleukin-1 beta ( 16 , 17 ). This explains why PCT mirrors the level of inflammation. Increased PCT in bacterial compared to viral infections has been widely reported ( 18 – 20 ). Therefore, our theoretical model predicted lower PCT concentrations (inflammation) in COVID-19 compared to bacterial infections. This prediction was borne out in our study, as shown in Table 1 . The above considerations led us to compare cytolysis and inflammation in a single biomarker instrument since there is theoretical rationale to believe they should differ in viral compared to bacterial infections. We focused on a proposed inverse relationship between cytolysis (high) and inflammation (low) in viral infections. An interesting theoretical rationale may explain an inverse relationship between (high) cytolysis and (low) inflammation in viral infections. Cytolysis liberates intracellular substances that directly suppress inflammation ( 21 ). Corroborating clinical data in Fan et al show that hyperferritinemic (cytolytic) sepsis associated with depressed inflammation, as indicated by reduced ex vivo whole blood endotoxin-stimulated TNF (the prototype pro-inflammatory cytokine) ( 22 ). Therefore, COVID-19-associated cytolysis may directly suppress inflammation and explain a relatively high cytolysis/inflammation ratio in COVID-19 pneumonia. In contrast to infections caused by viruses, our pathogenesis theory predicts bacterial infections are characterized by lower cytolysis and reduced ejection of intracellular anti-inflammatory substances. This should permit increased inflammation to be generated in bacterial infections. Translating our theory-based model of infections caused by viruses vs bacteria into a practical biomarker instrument, we created the ratio of ferritin/PCT. The model predicts that viral infections are characterized by relatively high ferritin and low PCT, whereas bacterial infections should be characterized by relatively low amounts of ferritin and high levels of PCT. As anticipated, we observed higher ferritin/PCT in viral COVID-19 infections than ferritin/PCT in bacterial infections (Table 1 ). Our previous results ( 1 ), along with confirmatory data shown in this report, support the internal and external validity of the theory-derived ferritin/PCT ratio. These corroborating analyses also suggest that biomarker creation using a theory-driven approach holds promise to mitigate uncovering chance-derived relationships. Theory-guided biomarker discovery may enhance the likelihood of developing biomarker instruments that reflect true disease pathogenesis ( 6 , 8 – 10 ). The accuracy of a diagnostic instrument is quantified by calculating the area under the ROC curve ( 23 ). Ferritin/PCT compares favorably with other diagnostic biomarkers used in infectious diseases clinical practice ( 24 ). For example, the accuracy of blood QuantiFERON ® testing to predict subsequent incident tuberculosis shows an area under the ROC curve of 0. 8 − 0. 82 ( 25 ), compared to 0. 812 for ferritin/PCT in this report. A systematic review and meta-analysis showed that levels of PCT alone cannot be used to differentiate viral from bacterial causes of pneumonia, with an area under the ROC curve of 0. 73 ( 5 ). Limitations of this report and prior assessment of ferritin/PCT to diagnose viral vs bacterial causes of pneumonia ( 1 ) include assessment using retrospectively derived databases. Prospective validation that assesses clinical utility in diverse patient populations is needed. Positive features of this report include confirmation of ferritin/PCT performance in a very large clinical database. We point out that this retrospective analysis may underestimate the clinical utility of ferritin/PCT in discriminating between COVID-19 pneumonia and Streptococcus pneumoniae pneumonia since ferritin and PCT were obtained at the treating clinicians' discretion. Disparate acquisition timing for ferritin and PCT may result in suboptimal biomarker performance. This is a limitation associated with using retrospective data. If ferritin and PCT are obtained simultaneously, the ferritin/PCT ratio may outperform our results as a discriminator between COVID-19 and Pneumococcal pneumonia. Another limitation of our study is the inability to test the accuracy of ferritin/PCT in infections caused by different strains of SARS-CoV-2, since this information is unavailable in the database we used. We summarize by noting some special strengths associated with this report. Since ferritin/PCT is derived from a pathogenesis theory that generally differentiates viral and bacterial infections, ferritin/PCT values may differentiate viral from bacterial infections regardless of specific microbial etiology. It seems reasonable ferritin/PCT can be used to differentiate viral from bacterial infections different from COVID-19 pneumonia and pneumococcal pneumonia. We are pursuing this line of investigation. Prospective studies can also assess the ratio for use in antibiotic stewardship measures in settings with a low likelihood of bacterial infections. Conclusion This report supports the concept that sound theoretical understanding of disease can be leveraged to discover novel diagnostic instruments with a high likelihood of clinical utility. We show that a theory-derived biomarker ratio ferritin/PCT can differentiate COVID-19 pneumonia vs pneumococcal pneumonia. This biomarker instrument is inexpensive, rapidly calculated and readily-accessible. It may have special use in low resource settings and during subsequent pandemics where rapid differentiation between viral and bacterial infection is needed in the absence if a definitive diagnostic test. Abbreviations COVID 19-Coronavirus Disease 2019 CKD chronic kidney disease CRP C-reactive protein COPD chronic obstructive pulmonary disease DM2 type 2 diabetes ICU intensive care unit PCT procalcitonin ROC receiver-operating characteristic SARS CoV-2-Severe Acute Respiratory Syndrome Coronavirus 2 TNF tumor necrosis factor alpha HIV human immunodeficiency virus infection Ln natural logarithm U units WBC = white blood count Declarations Ethics approval and consent to participate : Human Participants Statement- “This retrospective study is exempt from informed consent. The data reviewed is a secondary analysis of existing data, does not involve intervention or interaction with human subjects, and is de-identified per the de-identification standard defined in Section §164. 514(a) of the HIPAA Privacy Rule. The process by which the data is de-identified is attested to through a formal determination by a qualified expert as defined in Section §164. 514(b)(1) of the HIPAA Privacy Rule. This formal determination by a qualified expert refreshed on December 2020. ” Data Availability Statement : Data analyses available upon reasonable request to corresponding author. All primary data were extracted from the TrinetX database (https://trinetx. com/clinical-trial-design-optimization/#s_0). Consent for Publication : not applicable. Funding : Leland Shapiro is supported by a grant from the Emily Foundation for Medical Research. Author contributions : LS, AFH-M. and SS developed ideas expressed in the manuscript. JLS, GR-N, SS, PC, AGB, and DBC performed data extraction and statistical analyses. LS, AFH-M, and CF-P wrote the first draft of the manuscript and LS wrote the final draft. All authors proofread, edited, and approved the final manuscript text. Disclosure of interest statement : Dr. Shapiro reports grant support from the Emily Foundation for Medical research. Emily Foundation provided no direct support for this work and had no participation in conduct of this study or writing of the manuscript. All other authors report no actual or potential competing financial interests or personal relationships that could appear to influence this work References Gharamti AA, Mei F, Jankousky KC, Huang J, Hyson P, Chastain DB, et al. Diagnostic Utility of a Ferritin-to-Procalcitonin Ratio to Differentiate Patients With COVID-19 From Those With Bacterial Pneumonia: A Multicenter Study. Open Forum Infect Dis. 2021;8(6):ofab124. Kell DB, Pretorius E. Serum ferritin is an important inflammatory disease marker, as it is mainly a leakage product from damaged cells. Metallomics. 2014;6(4):748–73. Becze Z, Molnar Z, Fazakas J. The molecular basis of procalcitonin synthesis in different infectious and non-infectious acute conditions. J Hum Virol Retrovirol. 2016;3(2):1–6. Hamade B, Huang DT. Procalcitonin: Where Are We Now? Crit Care Clin. 2020;36(1):23–40. Kamat IS, Ramachandran V, Eswaran H, Guffey D, Musher DM. Procalcitonin to Distinguish Viral From Bacterial Pneumonia: A Systematic Review and Meta-analysis. Clin Infect Dis. 2020;70(3):538–42. Calude CS, Longo G. The deluge of spurious correlations in big data. Found Sci. 2017;22(3):595–612. Huang S. The Tension Between Big Data and Theory in the Omics Era of Biomedical Research. Perspect Biol Med. 2018;61(4):472–88. Richman JT, Roberts R. Assessing spurious corelations in big search data. Forecasting. 2023;5(1):285–96. Smith G. The AI delusion. First edition. ed. Oxford: Oxford University Press; 2018. p. 249. Smith G, Cordes J. The phantom pattern problem: the mirage of big data. First edition. ed. Oxford ; New York, NY: Oxford University Press; 2020. vii, 227 pages. Lingas EC. Empiric Antibiotics in COVID 19: A Narrative Review. Cureus. 2022;14(6):e25596. Heaton NS. Revisiting the concept of a cytopathic viral infection. PLoS Pathog. 2017;13(7):e1006409. Kurian SJ, Mathews SP, Paul A, Viswam SK, Kaniyoor Nagri S, Miraj SS, et al. Association of serum ferritin with severity and clinical outcome in COVID-19 patients: An observational study in a tertiary healthcare facility. Clin Epidemiol Glob Health. 2023;21:101295. Zhu Z, Shi J, Li L, Wang J, Zhao Y, Ma H. Therapy Targets SARS-CoV-2 Infection-Induced Cell Death. Front Immunol. 2022;13:870216. Kaushal K, Kaur H, Sarma P, Bhattacharyya A, Sharma DJ, Prajapat M, et al. Serum ferritin as a predictive biomarker in COVID-19. A systematic review, meta-analysis and meta-regression analysis. J Crit Care. 2022;67:172–81. Christ-Crain M, Muller B. Procalcitonin in bacterial infections–hype, hope, more or less? Swiss Med Wkly. 2005;135(31–32):451–60. Linscheid P, Seboek D, Nylen ES, Langer I, Schlatter M, Becker KL, et al. In vitro and in vivo calcitonin I gene expression in parenchymal cells: a novel product of human adipose tissue. Endocrinology. 2003;144(12):5578–84. Azijli K, Minderhoud TC, de Gans CJ, Lieveld AWE, Nanayakkara PWB. Optimal use of procalcitonin to rule out bacteremia in patients with possible viral infections. J Am Coll Emerg Physicians Open. 2022;3(3):e12621. Gendrel D, Raymond J, Coste J, Moulin F, Lorrot M, Guerin S, et al. Comparison of procalcitonin with C-reactive protein, interleukin 6 and interferon-alpha for differentiation of bacterial vs. viral infections. Pediatr Infect Dis J. 1999;18(10):875–81. Lubell Y, Blacksell SD, Dunachie S, Tanganuchitcharnchai A, Althaus T, Watthanaworawit W, et al. Performance of C-reactive protein and procalcitonin to distinguish viral from bacterial and malarial causes of fever in Southeast Asia. BMC Infect Dis. 2015;15:511. Horner E, Lord JM, Hazeldine J. The immune suppressive properties of damage associated molecular patterns in the setting of sterile traumatic injury. Front Immunol. 2023;14:1239683. Fan Z, Kernan KF, Qin Y, Canna S, Berg RA, Wessel D, et al. Hyperferritinemic sepsis, macrophage activation syndrome, and mortality in a pediatric research network: a causal inference analysis. Crit Care. 2023;27(1):347. Zou KH, O'Malley AJ, Mauri L. Receiver-operating characteristic analysis for evaluating diagnostic tests and predictive models. Circulation. 2007;115(5):654–7. Hwang H, Hwang BY, Bueno J. Biomarkers in Infectious Diseases. Dis Markers. 2018;2018:8509127. Gupta RK, Kunst H, Lipman M, Noursadeghi M, Jackson C, Southern J, et al. Evaluation of QuantiFERON-TB Gold Plus for Predicting Incident Tuberculosis among Recent Contacts: A Prospective Cohort Study. Ann Am Thorac Soc. 2020;17(5):646–50. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5581463","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":391050312,"identity":"1ce95f56-2931-40d4-b028-fb02ade2b6df","order_by":0,"name":"Leland Shapiro","email":"data:image/png;base64,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","orcid":"","institution":"University of Colorado Anschutz Medical Campus","correspondingAuthor":true,"prefix":"","firstName":"Leland","middleName":"","lastName":"Shapiro","suffix":""},{"id":391050313,"identity":"c939477c-db91-4d8d-a77a-478e0f60e92a","order_by":1,"name":"Jorge L Salinas","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Jorge","middleName":"L","lastName":"Salinas","suffix":""},{"id":391050314,"identity":"0f0662dc-13c7-4b71-981d-8a5f7c27ef2a","order_by":2,"name":"Guillermo Rodriguez-Nava","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Guillermo","middleName":"","lastName":"Rodriguez-Nava","suffix":""},{"id":391050315,"identity":"09dd9c31-4eb2-46d7-a47b-fcb6cbcdddeb","order_by":3,"name":"Sa Shen","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Sa","middleName":"","lastName":"Shen","suffix":""},{"id":391050316,"identity":"6aa69df7-eacc-41ad-9e6b-1f41ec790758","order_by":4,"name":"Daniel B. Chastain","email":"","orcid":"","institution":"UGA College of Pharmacy, SWGA Clinical Campus, Phoebe Putney Memorial Hospital","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"B.","lastName":"Chastain","suffix":""},{"id":391050317,"identity":"4e62200f-7e7a-4c7f-9996-9da80b4d580c","order_by":5,"name":"Carlos Franco-Paredes","email":"","orcid":"","institution":"Colorado State University","correspondingAuthor":false,"prefix":"","firstName":"Carlos","middleName":"","lastName":"Franco-Paredes","suffix":""},{"id":391050318,"identity":"0ba34a38-dab0-4e3e-a38f-b054d39321bd","order_by":6,"name":"Patrick Connelly","email":"","orcid":"","institution":"University of Colorado Boulder","correspondingAuthor":false,"prefix":"","firstName":"Patrick","middleName":"","lastName":"Connelly","suffix":""},{"id":391050319,"identity":"bc7b0d30-439a-433c-a233-90514d6fdf98","order_by":7,"name":"Alfonso G. Bastias","email":"","orcid":"","institution":"University of Colorado Boulder","correspondingAuthor":false,"prefix":"","firstName":"Alfonso","middleName":"G.","lastName":"Bastias","suffix":""},{"id":391050320,"identity":"5e3c55c0-4191-4e69-b417-954dae8b998c","order_by":8,"name":"Sias Scherger","email":"","orcid":"","institution":"University of Nebraska Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Sias","middleName":"","lastName":"Scherger","suffix":""},{"id":391050321,"identity":"149ff97e-4649-461b-933d-dfab6e01be66","order_by":9,"name":"Andrés F. Henao-Martínez","email":"","orcid":"","institution":"University of Colorado Anschutz Medical Campus","correspondingAuthor":false,"prefix":"","firstName":"Andrés","middleName":"F.","lastName":"Henao-Martínez","suffix":""}],"badges":[],"createdAt":"2024-12-04 16:38:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5581463/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5581463/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":71604653,"identity":"b3a2288d-faff-46be-bbce-3a0cf56e52f2","added_by":"auto","created_at":"2024-12-17 06:06:55","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":23667,"visible":true,"origin":"","legend":"\u003cp\u003eReceiver Operating Characteristic curves for ferritin/PCT ratio cutoff values in unadjusted (red) and adjusted (blue) analyses for predicting COVID-19 pneumonia. \u0026nbsp;Sensitivity represents true positive rate and 1-specificity depicts false positive rate. \u0026nbsp;Area under curves are shown in parentheses. \u0026nbsp;Diagonal line shows diagnostic accuracy equal to chance alone.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-5581463/v1/1b03dbe4a5fa76cbf81c245e.png"},{"id":71606055,"identity":"532ef67e-92a8-41b7-88fa-aa8ea1899b19","added_by":"auto","created_at":"2024-12-17 06:14:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":537690,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5581463/v1/f80342f7-4d39-4cb3-8d4c-2f0c6e3802b9.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Theory-Based Ferritin-Procalcitonin Ratio Differentiates COVID-19 Pneumonia vs Bacterial Pneumonia","fulltext":[{"header":"Introduction","content":" \u003cp\u003eRapid, inexpensive diagnosis of Coronavirus Disease 2019 (COVID-19) pneumonia and differentiation from bacterial pneumonia was needed early in the COVID-19 pandemic and remains essential in low-resource settings. We devised a novel diagnostic index consisting of the ferritin/procalcitonin (ferritin/PCT) ratio that was based on a model of pathogenesis that differentiates viral from bacterial infections (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Our theoretical model proposes that viral infections like COVID-19 are characterized by relatively high levels of cell death or cytolysis and concomitant low amounts of inflammation. Conversely, bacterial infections like Pneumococcal pneumonia are characterized by relatively low degrees of cytolysis and high amounts of inflammation. Prior investigation suggests the most accurate readily available measure that can quantify cytolysis is serum ferritin (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Inflammation can be quantified using blood levels of PCT, which is a readily-available established biomarker for inflammation that is significantly increased in bacterial infections compared to viral infections (\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). In a prior report we showed ferritin/PCT could differentiate COVID-19 from bacterial pneumonias (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). In this report we assessed ferritin/PCT capacity to differentiate pneumonia caused by COVID-19 vs pneumococcus in a much larger and diverse validation cohort.\u003c/p\u003e \u003cp\u003eA separate goal of this study was to assess a specific approach to biomarker instrument creation. Successful validation of the ferritin/PCT ratio would provide powerful support for our approach to biomarker development that emphasizes theoretical rationale prior to assessing statistical association between proposed biomarkers and diseases in clinical datasets. In contrast, we believe biomarker construction based on unrestrained mathematical evaluation of datasets presents high risk for generating random (non-causal) associations (\u003cspan additionalcitationids=\"CR7 CR8 CR9\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). In such cases of ransacking data, biomarker performance has high risk for failing to replicate in prospective clinical assessment.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003eIn the TrinetX database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://trinetx.com/clinical-trial-design-optimization/#s_0\u003c/span\u003e\u003cspan address=\"https://trinetx.com/clinical-trial-design-optimization/#s_0\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), we identified COVID-19 pneumonia patients by requiring both diagnostic ICD-10 codes and a positive genetic or antigen test for the presence of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) March 2020 through August 2023. Patients with \u003cem\u003eStreptococcus pneumoniae\u003c/em\u003e (bacterial) pneumonia were identified by requiring both diagnostic ICD-10 codes and a positive respiratory sample culture or body fluid specific antigen test May 2009 through July 2023. Patients with concomitant COVID-19 and pneumococcal pneumonia were excluded. The ferritin/PCT ratio was calculated using values for these 2 molecules in clinical blood samples obtained within seven days from recorded pneumonia diagnosis. The ferritin/PCT ratio numerical values for pneumonia caused by SARS-CoV-2 vs pneumococcal pneumonia were compared. Usefulness of the ferritin/PCT ratio as a separator of pneumonia caused by SARS-CoV-2 vs \u003cem\u003ePneumococcus\u003c/em\u003e was assessed by comparing mean ferritin/PCT values for COVID-19 and \u003cem\u003ePneumococcus\u003c/em\u003e. We assessed accuracy of ferritin/PCT for differentiating COVID-19 pneumonia vs pneumococcal pneumonia by generating the Receiver-Operating Characteristic (ROC) curve for the data and calculated the area under the ROC curve. A separate ROC curve was constructed that was adjusted for age, gender, ethnicity, race, marital status, region, and Charleson comorbidity index and the area under this ROC was calculated. Using the adjusted ROC curve, the optimal numerical cut point for the ferritin/PCT ratio was determined and the sensitivity and specificity for differentiating COVID-19 and \u003cem\u003ePneumococcal\u003c/em\u003e pneumonia was determined for this cut point.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTable \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e shows clinical characteristics, ferritin, procalcitonin, and ferritin/PCT ratio values in patients with COVID-19 pneumonia or pneumococcal pneumonia. Mean serum ferritin determinations for COVID-19 patients and \u003cem\u003ePneumococcal\u003c/em\u003e pneumonia patients were 1,052. 9 ng/mL and 803. 5 ng/mL, respectively (p\u0026thinsp;\u0026lt;\u0026thinsp;0. 001). Mean serum PCT in COVID-19 pneumonia patients and pneumococcal pneumonia patents were 1. 5 ng/mL and 10. 3 ng/mL, respectively (p\u0026thinsp;\u0026lt;\u0026thinsp;0. 001). Ferritin/procalcitonin ratio significantly differed between the two diagnoses with mean values 5,914. 4 for COVID-19 pneumonia and 1,439. 2 \u003cstrong\u003ef\u003c/strong\u003eor pneumococcal pneumonia (p\u0026thinsp;\u0026lt;\u0026thinsp;0. 001). Figure \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e shows unadjusted and adjusted ROC curves for diagnosing COVID-19 pneumonia vs bacterial pneumonia. Area under the ROC curve adjusted for age, gender, ethnicity, race, marital status, region, and Charleson comorbidity index was 0. 812. Using the Liu method, the optimal ferritin/PCT value for differentiation is 1,158. 14, with sensitivity and specificity 0. 73.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab1\" style=\"width: 504px;\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eClinical Information and Outcomes in Patients with COVID-19 Pneumonia or Pneumococcal Pneumonia\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth style=\"width: 133.84px;\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003eCOVID-19 pneumonia\u003c/p\u003e\n \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;29,348)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003ePneumococccal Pneumonia\u003c/p\u003e\n \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;1,921)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth style=\"width: 480px;\" colspan=\"4\" align=\"left\"\u003e\n \u003cp\u003eDemographic Characteristics\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eAge* (sd)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e62. 2 (16.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e58. 9 (17.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eSex\u003c/p\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e13,383 (45.6%)\u003c/p\u003e\n \u003cp\u003e15,963 (54.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1,007 (52.4%)\u003c/p\u003e\n \u003cp\u003e914 (47.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eRace\u003c/p\u003e\n \u003cp\u003eWhite\u003c/p\u003e\n \u003cp\u003eBlack\u003c/p\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003cp\u003eNative Hawaiian\u003c/p\u003e\n \u003cp\u003eAmerican Indian\u003c/p\u003e\n \u003cp\u003eAsian\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e20,100 (68.5%)\u003c/p\u003e\n \u003cp\u003e4,731 (16.1%)\u003c/p\u003e\n \u003cp\u003e2,569 (8. 8%)\u003c/p\u003e\n \u003cp\u003e505 (1.7%)\u003c/p\u003e\n \u003cp\u003e114 (0.4%)\u003c/p\u003e\n \u003cp\u003e1,329 (4.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1,223 (63.7%)\u003c/p\u003e\n \u003cp\u003e398 (20.7%)\u003c/p\u003e\n \u003cp\u003e197 (10.3%)\u003c/p\u003e\n \u003cp\u003e14 (0.7%)\u003c/p\u003e\n \u003cp\u003e32 (1.7%)\u003c/p\u003e\n \u003cp\u003e57 (3.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 480px;\" colspan=\"4\" align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eAssociated Diseases\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eHeart failure\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e9,787 (33.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e872 (45.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eCOPD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e7,669 (26.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e816 (42.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eCKD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e10,084 (34.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e748 (39 2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eLiver disease\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e1,983 (6.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e271 (14.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eDM2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e13,243 (45.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e723 (37.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eNeoplasm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e5,836 (19.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e531 (27.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eHIV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e228 (0.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e103 (5.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 480px;\" colspan=\"4\" align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eLaboratory Information\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eWBC x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e/\u0026micro;L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e8.5 (9.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e12. 6 (10.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003ePlatelets x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e/\u0026micro;L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e218.2 (100.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e238. 3 (155.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eHemoglobin (g/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e12.6 (2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e10. 5 (2.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eLymphocytes x 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e/\u0026micro;L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e5.4 (92.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e7. 0 (80.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e0.755\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eCreatinine (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e1.5 (2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e1. 7 (2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eALT (U/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e46.3 (140.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e64. 2 (315.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eAST (U/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e62.8 (207.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e91. 9 (410.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eCRP (mg/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e84.7 (78.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e110. 8 (111.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003ePCT (ng/mL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e1.5 (13.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e10. 3 (32.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eFerritin (ng/mL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e1,052.9 (1985.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e803. 5 (2187.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eFerritin/PCT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e5,914.4 (9274.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e1,439. 2 (3206.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 480px;\" colspan=\"4\" align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eLocation of Care and Follow-up\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lsquo;Follow-up (days)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e540 9 (494.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e931. 0 (849.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eHospitalization\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e7,485 (25.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e621 (32.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eICU admission\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e7,599 (25.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e643 (33.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 133.84px;\" align=\"left\"\u003e\n \u003cp\u003eMortality at 1-year\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 133.16px;\" align=\"left\"\u003e\n \u003cp\u003e7,232 (26.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 167px;\" align=\"left\"\u003e\n \u003cp\u003e455 (30.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\" align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 480px;\" colspan=\"4\" align=\"left\"\u003e\n \u003cp\u003e*All values\u0026nbsp;reported\u0026nbsp;as mean\u0026nbsp;with standard deviation (sd)\u003c/p\u003e\n \u003cp\u003eCOPD\u0026thinsp;=\u0026thinsp;chronic obstructive pulmonary disease\u003c/p\u003e\n \u003cp\u003eCKD\u0026thinsp;=\u0026thinsp;chronic kidney disease\u003c/p\u003e\n \u003cp\u003eCRP- C-reactive protein\u003c/p\u003e\n \u003cp\u003eDM2\u0026thinsp;=\u0026thinsp;type 2 diabetes\u003c/p\u003e\n \u003cp\u003eHIV\u0026thinsp;=\u0026thinsp;human immunodeficiency virus infection\u003c/p\u003e\n \u003cp\u003eICU\u0026thinsp;=\u0026thinsp;Intensive Care Unit\u003c/p\u003e\n \u003cp\u003eLn\u0026thinsp;=\u0026thinsp;natural\u0026nbsp;logarithm\u003c/p\u003e\n \u003cp\u003ePCT\u0026thinsp;=\u0026thinsp;procalcitonin\u003c/p\u003e\n \u003cp\u003eU\u0026thinsp;=\u0026thinsp;units\u003c/p\u003e\n \u003cp\u003eWBC\u0026thinsp;=\u0026thinsp;white blood count\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis report confirms significant discriminatory power for the ferritin/PCT instrument to differentiate COVID-19 pneumonia from bacterial pneumonia in retrospective observational databases (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). This report is the most convincing and accurate to date due to the large number of subjects studied and the diversity of subjects. Early in the COVID-19 pandemic, there was need to diagnose pneumonias caused by SARS-CoV-2 vs bacterial etiologies rapidly. Without diagnostic tests that could identify SARS-CoV-2, up to 72% of COVID-19 patients were administered empiric broad-spectrum antibiotic therapy (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). This kind of antibiotic overuse can incur adverse effects and increase costs of care. To address the clinical problem of diagnosing COVID-19 vs bacterial infection at the beginning of the COVID-19 pandemic, we created a novel diagnostic biomarker instrument using a theory-based (mechanistic) approach. Two considerations motivated pursuing a theory-based approach. First, little clinical information was available early in the pandemic that could be used to statistically link biomarkers to diagnosis. Second, we were concerned about the problem of false-positive associations that can arise from unrestrained assessment of databases using powerful computational algorithms (\u003cspan additionalcitationids=\"CR7 CR8 CR9\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). We additionally constrained candidate theoretical mechanisms to those capable of immediate translation to clinical usefulness using readily available \u0026ldquo;off-the-shelf\u0026rdquo; biomarkers. Based on understanding of microbial pathogenesis, we created a theory of infection that focused on 2 characteristics that could potentially differentiate viral from bacterial etiologies. These 2 characteristics were cytolysis and inflammation. Specifically, we theorized that viral infections were characterized by large amounts of host cell destruciuon and relatively low levels of inflammation. In contrast, our model proposed lower amounts of cytolysis and higher levels of inflammation in cases of bacterial infections.\u003c/p\u003e \u003cp\u003eRegarding cytolysis, viral infections often produce host cell destruction due to the necessary replication of inside cells and cell lysis to disseminate and initiate fresh infections (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Bacteria have no such requirement. Clinical application of this component of our theory required a readily available marker for cell lysis. Surprisingly, circulating ferritin can be used to quantify cytolysis since ferritin has no established non-lytic mechanism to exit cells (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Ferritin is often mischaracterized as a typical acute-phase reactant. Assessment of background knowledge in developing our theory revealed that SARS-CoV-2 replication induced cytolysis \u003cem\u003ein vitro\u003c/em\u003e (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). Also, clinical observations showed elevated ferritin associated with COVID-19 diagnosis and disease severity (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). These observations suggested COVOD-19 causes cytolysis and ferritin quantifies the amount of cytolysis in cases of COVID-19. As predicted, high ferritin levels in COVID-19 patients were observed in our study, and significantly lower ferritin levels were observed in pneumococcal infections (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. To quantify inflammation, we chose PCT as the best readily available indicator. Procalcitonin is increased by prototypical cytokine mediators of inflammation tumor necrosis factor-alpha (TNF) and interleukin-1 beta (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). This explains why PCT mirrors the level of inflammation. Increased PCT in bacterial compared to viral infections has been widely reported (\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). Therefore, our theoretical model predicted lower PCT concentrations (inflammation) in COVID-19 compared to bacterial infections. This prediction was borne out in our study, as shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe above considerations led us to compare cytolysis and inflammation in a single biomarker instrument since there is theoretical rationale to believe they should differ in viral compared to bacterial infections. We focused on a proposed inverse relationship between cytolysis (high) and inflammation (low) in viral infections. An interesting theoretical rationale may explain an inverse relationship between (high) cytolysis and (low) inflammation in viral infections. Cytolysis liberates intracellular substances that directly suppress inflammation (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). Corroborating clinical data in Fan \u003cem\u003eet al\u003c/em\u003e show that hyperferritinemic (cytolytic) sepsis associated with depressed inflammation, as indicated by reduced \u003cem\u003eex vivo\u003c/em\u003e whole blood endotoxin-stimulated TNF (the prototype pro-inflammatory cytokine) (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e). Therefore, COVID-19-associated cytolysis may directly suppress inflammation and explain a relatively high cytolysis/inflammation ratio in COVID-19 pneumonia. In contrast to infections caused by viruses, our pathogenesis theory predicts bacterial infections are characterized by lower cytolysis and reduced ejection of intracellular anti-inflammatory substances. This should permit increased inflammation to be generated in bacterial infections. Translating our theory-based model of infections caused by viruses vs bacteria into a practical biomarker instrument, we created the ratio of ferritin/PCT. The model predicts that viral infections are characterized by relatively high ferritin and low PCT, whereas bacterial infections should be characterized by relatively low amounts of ferritin and high levels of PCT. As anticipated, we observed higher ferritin/PCT in viral COVID-19 infections than ferritin/PCT in bacterial infections (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003cb\u003e).\u003c/b\u003e\u003c/p\u003e \u003cp\u003eOur previous results (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e), along with confirmatory data shown in this report, support the internal and external validity of the theory-derived ferritin/PCT ratio. These corroborating analyses also suggest that biomarker creation using a theory-driven approach holds promise to mitigate uncovering chance-derived relationships. Theory-guided biomarker discovery may enhance the likelihood of developing biomarker instruments that reflect true disease pathogenesis (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe accuracy of a diagnostic instrument is quantified by calculating the area under the ROC curve (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Ferritin/PCT compares favorably with other diagnostic biomarkers used in infectious diseases clinical practice (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). For example, the accuracy of blood QuantiFERON\u003csup\u003e\u0026reg;\u003c/sup\u003e testing to predict subsequent incident tuberculosis shows an area under the ROC curve of 0. 8\u0026thinsp;\u0026minus;\u0026thinsp;0. 82 (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e), compared to 0. 812 for ferritin/PCT in this report. A systematic review and meta-analysis showed that levels of PCT alone cannot be used to differentiate viral from bacterial causes of pneumonia, with an area under the ROC curve of 0. 73 (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eLimitations of this report and prior assessment of ferritin/PCT to diagnose viral vs bacterial causes of pneumonia (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) include assessment using retrospectively derived databases. Prospective validation that assesses clinical utility in diverse patient populations is needed. Positive features of this report include confirmation of ferritin/PCT performance in a very large clinical database. We point out that this retrospective analysis may underestimate the clinical utility of ferritin/PCT in discriminating between COVID-19 pneumonia and \u003cem\u003eStreptococcus pneumoniae\u003c/em\u003e pneumonia since ferritin and PCT were obtained at the treating clinicians' discretion. Disparate acquisition timing for ferritin and PCT may result in suboptimal biomarker performance. This is a limitation associated with using retrospective data. If ferritin and PCT are obtained simultaneously, the ferritin/PCT ratio may outperform our results as a discriminator between COVID-19 and Pneumococcal pneumonia. Another limitation of our study is the inability to test the accuracy of ferritin/PCT in infections caused by different strains of SARS-CoV-2, since this information is unavailable in the database we used. We summarize by noting some special strengths associated with this report. Since ferritin/PCT is derived from a pathogenesis theory that generally differentiates viral and bacterial infections, ferritin/PCT values may differentiate viral from bacterial infections regardless of specific microbial etiology. It seems reasonable ferritin/PCT can be used to differentiate viral from bacterial infections different from COVID-19 pneumonia and pneumococcal pneumonia. We are pursuing this line of investigation. Prospective studies can also assess the ratio for use in antibiotic stewardship measures in settings with a low likelihood of bacterial infections.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis report supports the concept that sound theoretical understanding of disease can be leveraged to discover novel diagnostic instruments with a high likelihood of clinical utility. We show that a theory-derived biomarker ratio ferritin/PCT can differentiate COVID-19 pneumonia vs pneumococcal pneumonia. This biomarker instrument is inexpensive, rapidly calculated and readily-accessible. It may have special use in low resource settings and during subsequent pandemics where rapid differentiation between viral and bacterial infection is needed in the absence if a definitive diagnostic test.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCOVID\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003e19-Coronavirus Disease 2019\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCKD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003echronic kidney disease\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCRP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eC-reactive protein\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCOPD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003echronic obstructive pulmonary disease\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDM2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003etype 2 diabetes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eICU\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eintensive care unit\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePCT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eprocalcitonin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eROC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ereceiver-operating characteristic\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSARS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eCoV-2-Severe Acute Respiratory Syndrome Coronavirus 2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTNF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003etumor necrosis factor alpha\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eHIV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ehuman immunodeficiency virus infection\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLn\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003enatural logarithm\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eU\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eunits\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eWBC\u0026thinsp;=\u0026thinsp;white blood count\u003c/div\u003e \u003cdiv class=\"Description\"\u003e\u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEthics approval and consent to participate\u003c/em\u003e\u003c/strong\u003e: \u0026nbsp;Human Participants Statement- \u003cem\u003e\u0026ldquo;This retrospective study is exempt from informed consent. \u0026nbsp;The data reviewed is a secondary analysis of existing data, does not involve intervention or interaction with human subjects, and is de-identified per the de-identification standard defined in Section \u0026sect;164. 514(a) of the HIPAA Privacy Rule. \u0026nbsp;The process by which the data is de-identified is attested to through a formal determination by a qualified expert as defined in Section \u0026sect;164. 514(b)(1) of the HIPAA Privacy Rule. \u0026nbsp;This formal determination by a qualified expert refreshed on December 2020. \u0026rdquo;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eData Availability Statement\u003c/em\u003e\u003c/strong\u003e: Data analyses available upon reasonable request to corresponding author. \u0026nbsp; All primary data were extracted from the\u0026nbsp;TrinetX database (https://trinetx. com/clinical-trial-design-optimization/#s_0).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eConsent for Publication\u003c/em\u003e\u003c/strong\u003e: not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFunding\u003c/em\u003e\u003c/strong\u003e:\u0026nbsp;\u0026nbsp;Leland Shapiro is supported by a grant from the Emily Foundation for Medical Research.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAuthor contributions\u003c/em\u003e\u003c/strong\u003e: LS, AFH-M. and SS developed ideas expressed in the manuscript. \u0026nbsp; JLS, GR-N, SS, PC, AGB, and DBC performed data extraction and statistical analyses. \u0026nbsp;LS, AFH-M, and CF-P wrote the first draft of the manuscript and LS wrote the final draft. \u0026nbsp;All authors proofread, edited, and \u0026nbsp;approved the final manuscript text.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eDisclosure of interest statement\u003c/em\u003e\u003c/strong\u003e: \u0026nbsp;Dr. \u0026nbsp;Shapiro reports grant support from the Emily Foundation for Medical research. \u0026nbsp;Emily Foundation provided no direct support for this work and had no participation in conduct of this study or writing of the manuscript. \u0026nbsp;All other authors report no actual or potential competing financial interests or personal relationships that could appear to influence this work\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e \u003cli\u003e\u003cspan\u003eGharamti AA, Mei F, Jankousky KC, Huang J, Hyson P, Chastain DB, et al. Diagnostic Utility of a Ferritin-to-Procalcitonin Ratio to Differentiate Patients With COVID-19 From Those With Bacterial Pneumonia: A Multicenter Study. Open Forum Infect Dis. 2021;8(6):ofab124.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKell DB, Pretorius E. Serum ferritin is an important inflammatory disease marker, as it is mainly a leakage product from damaged cells. Metallomics. 2014;6(4):748\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBecze Z, Molnar Z, Fazakas J. The molecular basis of procalcitonin synthesis in different infectious and non-infectious acute conditions. J Hum Virol Retrovirol. 2016;3(2):1\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamade B, Huang DT. Procalcitonin: Where Are We Now? Crit Care Clin. 2020;36(1):23\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKamat IS, Ramachandran V, Eswaran H, Guffey D, Musher DM. Procalcitonin to Distinguish Viral From Bacterial Pneumonia: A Systematic Review and Meta-analysis. Clin Infect Dis. 2020;70(3):538\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCalude CS, Longo G. The deluge of spurious correlations in big data. Found Sci. 2017;22(3):595\u0026ndash;612.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang S. The Tension Between Big Data and Theory in the Omics Era of Biomedical Research. Perspect Biol Med. 2018;61(4):472\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichman JT, Roberts R. Assessing spurious corelations in big search data. Forecasting. 2023;5(1):285\u0026ndash;96.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith G. The AI delusion. First edition. ed. Oxford: Oxford University Press; 2018. p. 249.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith G, Cordes J. The phantom pattern problem: the mirage of big data. First edition. ed. Oxford ; New York, NY: Oxford University Press; 2020. vii, 227 pages.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLingas EC. Empiric Antibiotics in COVID 19: A Narrative Review. Cureus. 2022;14(6):e25596.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeaton NS. Revisiting the concept of a cytopathic viral infection. PLoS Pathog. 2017;13(7):e1006409.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKurian SJ, Mathews SP, Paul A, Viswam SK, Kaniyoor Nagri S, Miraj SS, et al. Association of serum ferritin with severity and clinical outcome in COVID-19 patients: An observational study in a tertiary healthcare facility. Clin Epidemiol Glob Health. 2023;21:101295.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Z, Shi J, Li L, Wang J, Zhao Y, Ma H. Therapy Targets SARS-CoV-2 Infection-Induced Cell Death. Front Immunol. 2022;13:870216.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaushal K, Kaur H, Sarma P, Bhattacharyya A, Sharma DJ, Prajapat M, et al. Serum ferritin as a predictive biomarker in COVID-19. A systematic review, meta-analysis and meta-regression analysis. J Crit Care. 2022;67:172\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChrist-Crain M, Muller B. Procalcitonin in bacterial infections\u0026ndash;hype, hope, more or less? Swiss Med Wkly. 2005;135(31\u0026ndash;32):451\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLinscheid P, Seboek D, Nylen ES, Langer I, Schlatter M, Becker KL, et al. In vitro and in vivo calcitonin I gene expression in parenchymal cells: a novel product of human adipose tissue. Endocrinology. 2003;144(12):5578\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAzijli K, Minderhoud TC, de Gans CJ, Lieveld AWE, Nanayakkara PWB. Optimal use of procalcitonin to rule out bacteremia in patients with possible viral infections. J Am Coll Emerg Physicians Open. 2022;3(3):e12621.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGendrel D, Raymond J, Coste J, Moulin F, Lorrot M, Guerin S, et al. Comparison of procalcitonin with C-reactive protein, interleukin 6 and interferon-alpha for differentiation of bacterial vs. viral infections. Pediatr Infect Dis J. 1999;18(10):875\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLubell Y, Blacksell SD, Dunachie S, Tanganuchitcharnchai A, Althaus T, Watthanaworawit W, et al. Performance of C-reactive protein and procalcitonin to distinguish viral from bacterial and malarial causes of fever in Southeast Asia. BMC Infect Dis. 2015;15:511.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorner E, Lord JM, Hazeldine J. The immune suppressive properties of damage associated molecular patterns in the setting of sterile traumatic injury. Front Immunol. 2023;14:1239683.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan Z, Kernan KF, Qin Y, Canna S, Berg RA, Wessel D, et al. Hyperferritinemic sepsis, macrophage activation syndrome, and mortality in a pediatric research network: a causal inference analysis. Crit Care. 2023;27(1):347.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou KH, O'Malley AJ, Mauri L. Receiver-operating characteristic analysis for evaluating diagnostic tests and predictive models. Circulation. 2007;115(5):654\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHwang H, Hwang BY, Bueno J. Biomarkers in Infectious Diseases. Dis Markers. 2018;2018:8509127.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGupta RK, Kunst H, Lipman M, Noursadeghi M, Jackson C, Southern J, et al. Evaluation of QuantiFERON-TB Gold Plus for Predicting Incident Tuberculosis among Recent Contacts: A Prospective Cohort Study. Ann Am Thorac Soc. 2020;17(5):646\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Pneumonia, Infection pathogenesis, COVID-19, Ferritin, Procalcitonin, Ferritin-Procalcitonin ratio, Biomarker, Theory","lastPublishedDoi":"10.21203/rs.3.rs-5581463/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5581463/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground: \u003c/strong\u003eRapid and inexpensive biomarker-based clinical instruments that can diagnose infectious diseases are desired, but developing clinical instruments has proved challenging. \u0026nbsp;Proliferation of large clinical databases and expansive computational capability risks uncovering spurious associations that cannot be reproduced.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eObjectives\u003c/strong\u003e: We present an approach to biomarker instrument creation that may enhance clinical applicability. \u0026nbsp;We prospectively derived a biomarker instrument from a theoretical model of infection pathogenesis. \u0026nbsp;Our theory-derived ferritin/procalcitonin (ferritin/PCT) ratio was designed to differentiate Coronavirus Disease 2019 (COVID-19) pneumonia from bacterial pneumonias.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMaterials and Methods\u003c/strong\u003e: We assessed this ratio in over 30,000 patients in the TrinetX global database containing over 200 million patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e: Ferritin/PCT was significantly increased in COVID-19 pneumonia patients compared to bacterial pneumonia pateints. \u0026nbsp;\u0026nbsp;Ferritin/PCT accuracy for separating pneumonia due to COVID-19 vs Pneumococcus was assessed by calculating area under Receiver Operating Characteristic curve, which revealed a value of 0. 812.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e: The ferritin/PCT ratio may have clinical use for differentiating COVID-19 pneumonia vs Pneumococcal pneumonia. \u0026nbsp;Calculating the ferritin/PCT ratio is easy, rapid, and inexpensive. \u0026nbsp;Clinical utility in resource-poor locations is an especially attractive application. \u0026nbsp;Moreover, the conceptual model of infection pathogenesis that underlies this ratio may have broad applicability to differentiate other viral from bacterial infections.\u003c/p\u003e","manuscriptTitle":"A Theory-Based Ferritin-Procalcitonin Ratio Differentiates COVID-19 Pneumonia vs Bacterial Pneumonia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-17 05:58:50","doi":"10.21203/rs.3.rs-5581463/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b63bde4f-ed5b-4bbb-a331-ced13d6721bf","owner":[],"postedDate":"December 17th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-12-17T05:58:50+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-17 05:58:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5581463","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5581463","identity":"rs-5581463","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}