Sepsis is a life-threatening condition that neonatal patients are especially susceptible to. Fortunately, improved bedside monitoring has enabled the collection and use of continuous vital signs data for the purpose of detecting conditions such as sepsis. While current research has found some success in reducing mortality in neonatal intensive care units with linear directly interpretable models, such as logistic regression, accurate detection of sepsis from inherently noisy time-series data still remains a challenge. Furthermore, previous research has generally relied on pre-defined features extracted from rawvital signs data, which may not be optimal for the detection task. Therefore, assessing the overall feasibility of sequential deep learning models, such as recurrent and convolutional models, could improve the results of current research. This task was tackled in three phases. Firstly, baseline scores were established with a logistic regression model. Secondly, three common recurrent classifiers were tested on pre-defined window based features and compared with each other. Thirdly, a convolutional architecture with a recurrent and non-recurrent classifier was tested on raw low frequency (1Hz) signals in order to examine their capability to automatically extract features from the data. The final results from all phases were compared with each other. Results show that recurrent classifiers trained on pre-defined features do outperform automatic feature extraction with the convolutional models. The best model was based on a long-short term memory unit that achieved an area under the characteristic receiver operating unit curve of 0.806, and outperformed the established baseline results. In comparison with previous research, said model performed on par with the examined simple interpretable baseline models. The low results can likely be attributed to a insufficient sample size of patients with sepsis for the examined models and sub-optimal hyperparameter optimization due to the number of possible configurations. Further avenues of research include examination of high frequency data and more complex models for automatic feature extraction. / Sepsis är ett livshotande tillstånd som neonatala patienter är särskilt mottagliga för. Lyckligtvis har förbättrad patientmonitorering möjliggjort kontinuerlig insamling och andvänding av vitalparametrar i syfte att upptäcka tillstånd som sepsis. Medan aktuell forskning har funnit viss framgång i att minska dödligheten på neonatala intensivvårdsavdelningar med hjälp av linjära tolkbara modeller, såsom logistisk regression, är noggrann detektering av sepsis från brusig tidsseriedata fortfarande en utmaning. Dessutom har tidigare forskning i allmänhet förlitat sig på fördefinierade prediktorer extraherade från rå vitalparameterdata, som kanske inte är optimala för detektionsuppgiften. På grund av detta kan en bedömning av den övergripande användbarheten av sekventiella modeller för djupinlärning, såsom RNN- och CNN-modeller, förbättra resultaten av aktuell forskning. Denna uppgift tacklades i tre faser. Först och främst etablerades baslinjeresultat med en logistisk regressionsmodell. För det andra testades tre RNN-baserad klassificerare på data med fördefinierade fönsterbaserade prediktorer och jämfördes med varandra. För det tredje testades en CNN-arkitektur med både en RNN-klassificerare och MLP-klassificerare på råa lågfrekventa (1Hz) signaler för att undersöka deras förmåga att automatiskt extrahera egna prediktorer från datan. Slutresultaten från alla faser jämfördes med varandra. Resultaten visar att RNN-klassificerare som tränats på fördefinierade prediktorer överträffar automatisk extraktion av prediktorer med CNN-modellerna. Den bäst presterande modellen baserades på en långtidsminnesenhet som uppnådde en AUROC på 0.806, och överträffade de etablerade baslinjeresultaten. I jämförelse med tidigare forskning uppnådde ifrågavarande modell lika hög prestation som de väl undersökta enklare tolkbara baslinjemodellerna. De låga resultaten kan sannolikt tillskrivas en otillräcklig provstorlek av patienter med sepsis för de undersökta modellerna och suboptimal hyperparameteroptimering på grund av antalet möjliga konfigurationer. Ytterligare forskningsvägar inkluderar undersökning av högfrekventa data och mer komplexa modeller för automatisk extraktion av prediktorer.

Sepsis is a clinical syndrome characterized by the systemic activation of inflammatory and coagulation pathways in response to microbial infection of normally sterile parts of the body. Despite considerable advances in our understanding of sepsis pathophysiology, sepsis remains the leading cause of death in non-coronary intensive care units (ICU) with a global disease burden between 15 and 19 million cases per year (Dellinger et al., 2008). Severe sepsis, defined as sepsis associated with organ dysfunction is associated with mortality rates of 33% to 45%. The incidence of severe sepsis continues to increase by 1.5% per annum due to the aging population, a rise in the prevalence of comorbidities, and the wider use of immunosuppressive agents and invasive procedures (Angus et al., 2001). Over the past several decades, many potential treatments for sepsis have shown early promise, yet have failed to improve survival in over 100 Phase II and Phase III clinical trials (Marshall, 2014) suggesting that some fundamental knowledge is lacking in our understanding of sepsis pathophysiology. Emerging studies on cell-free DNA (cfDNA), DNA released extracellularly into the circulation, demonstrate that cfDNA is a crucial link between inflammation and coagulation . In various conditions characterized by excessive inflammatory responses or aberrant prothrombotic responses, cfDNA has been implicated in exacerbating disease pathology (Atamaniuk, Kopecky, Skoupy, Säemann, & Weichhart, 2012; Fuchs, Brill, & Wagner, 2012; Swystun, Mukherjee, & Liaw, 2011). In clinical sepsis, levels of cfDNA upon admission into the ICU have strong prognostic value in predicting mortality (Dwivedi et al., 2012; Saukkonen et al., 2008). However, it is unclear whether these increases in cfDNA are an epiphenomenon during sepsis progression, or whether cfDNA actively plays a role in sepsis pathophysiology. In this work, in vivo studies were conducted to characterize the role of cfDNA in sepsis, the effects of DNase administration, and the potential mechanism by which cfDNA is released during experimental sepsis. In addition, mortality studies were conducted to identify surrogate markers of death to promote the design of humane and ethical animal studies in conducting sepsis research. Polymicrobial sepsis was induced via a surgical procedure whereby the cecum is exteriorized, ligated and punctured twice to introduce a continuous source of microorganisms, a model termed cecal ligation and puncture (CLP). In our CLP sepsis model, levels of cfDNA increased in a time-dependent manner. These increases accompanied an early pro-inflammatory response marked by increased pro-inflammatory IL-6, a transient increase in anti-inflammatory IL-10, and elevated lung myeloperoxidase (MPO) activity. Septic mice with elevated cfDNA levels also had high bacterial loads in the lungs, blood, and peritoneal cavity fluid. Organ damage was also observed in mice following CLP surgery versus mice subjected to the non-septic sham control surgery marked by increased levels of creatinine and alanine aminotransferase (ALT) indicative of kidney and liver injury, respectively. Histological analyses further confirmed lung and kidney damage following CLP surgery. Changes in coagulation were also observed in septic mice as mice subjected to CLP had sustained increases in thrombin-antithrombin (TAT) complexes. In addition, plasma from CLP-operated mice had increased thrombin generation (i.e. increased endogenous thromin potential, increased peak thrombin, decreased time to peak, and decreased lag time) mediated by FXIIa and enhanced by platelets. Following CLP-induced sepsis, elevations in cfDNA levels accompanied pro-inflammatory and pro-coagulant responses. The effects of in vivo DNase treatment in septic mice were time-dependent. Early DNase treatment when cfDNA levels were low resulted in an exaggerated pro-inflammatory response marked by increased plasma IL-6 levels and increased lung damage. In contrast, delayed DNase treatment at time-points when cfDNA levels were elevated suppressed inflammation characterized by an increase in anti-inflammatory IL-10 and reductions in cfDNA, IL-6, lung MPO, and ALT activity. Furthermore, delayed DNase administration resulted in decreased bacterial load in the lungs, blood, and peritoneal cavity fluid. Delayed DNase treatment also resulted in blunted pro-coagulant responses as levels of TAT complexes were suppressed and thrombin generation from septic mouse plasma was normalized. Moreover, DNase treatment when cfDNA levels were elevated increased survival in CLP-operated mice by 80% and reduced lung and liver damage. These findings suggest that administration of DNase when cfDNA levels are elevated may reduce pro-inflammatory and pro-coagulant responses and that delayed DNase treatment may infer protection in the CLP model of sepsis. One mechanism by which cfDNA is released is via the formation of neutrophil extracellular traps (NETs). Upon inflammatory stimulation, some neutrophils release chromatin material and antimicrobial proteins (i.e. neutrophil elastase, MPO, and histones) in an active process termed NETosis. Although NETs ensnare bacteria and exert antimicrobial properties, NETs may also exert harmful effects on the host by activating inflammation and coagulation. While some in vitro evidence suggest that neutrophils are the main source of cfDNA released following inflammatory stimulation, others have reported that neutrophils are not the main source of circulating cfDNA following septic challenge. To determine whether NETs contribute to cfDNA released during CLP sepsis, genetically modified mice that are incapable of forming NETs, PAD4-/- mice, were used. Levels of cfDNA in PAD-/- mice were significantly lower than cfDNA levels in C57Bl/6 mice following CLP surgery, suggesting that NETs were a source of cfDNA in our model. Levels of IL-6, MPO, and bacterial load in the lungs, blood, and peritoneal cavity were significantly reduced, indicating that NETs exert pro-inflammatory effects in CLP sepsis. Thrombin generation was also suppressed in PAD4-/- mice which suggests that NETs contribute to thrombin generation following CLP sepsis. NETs contribute to increases in circulating cfDNA and may exacerbate pathology by driving pro-inflammatory and pro-coagulant responses in CLP-induced sepsis. Appreciating the implications of conducting research using animals, it is pertinent that researchers ensure the highest ethical standards and design animal studies in the most humane, yet scientifically rigorous manner. Using mortality studies, we validated the utility of physiological and phenotypic markers to assess disease severity and predict death in murine sepsis. Temperature via a rectal probe monitor and sepsis scoring systems which assess components such as orbital tightening, level of consciousness, and activity were effective surrogate markers of death. These tools offer a non-invasive assessment of disease progression which do not artificially exacerbate sepsis pathology and immediate information regarding any changes in the health status. Surrogate markers of death also provide reliable monitoring to meet increasing standards of ethical, humane animal research and a feasible and cost-efficient means to obtain vital signs in small rodents. We have proposed a scoring system which can be used for assessing disease severity, endpoint monitoring, and predicting death to obviate inhumane methods of using death as an endpoint in sepsis studies. In summary, cfDNA levels are elevated in CLP-induced sepsis and these elevations accompany pro-inflammatory and pro-coagulant responses. NETosis may be a mechanism by which cfDNA is released and NETs may drive inflammation and coagulation in CLP sepsis. Delayed DNase administration may suppress inflammation and coagulation and may be protective in polymicrobial sepsis. In future animal sepsis studies, surrogate markers of death and a sepsis scoring system can be used in place of death as an endpoint to raise the standards in conducting ethical, humane sepsis research. / Thesis / Doctor of Philosophy (PhD)

Sepsis characterized by severe bloodstream bacterial infection that results in high mortality rates and medical expenses has become a global healthcare challenge. The current clinical sepsis diagnosis is based on blood culture and requires a long waiting time, delaying effective treatments. The emerging sepsis diagnosis shortens the turn-around of time but requires an efficient bacteria separation from the blood beforehand. In this master’s degree project, a new filter-based method of separating bacteria from the blood is investigated to meet the clinical need for future sepsis diagnosis. The separation device is controlled within the size of a centrifuge falcon tube and integrated with a slanted filter inside. Through a centrifuge, blood cells and bacteria are maneuvered toward the filter and separated by it. Using the optimal design, at concentrations as low as 100 CFU/mL, the device can recover around 30% of bacteria of E. coli, K. pneumoniae, and S. aureus from 1 mL of whole blood within 60 minutes, and reject 99.4% RBCs from going through the filter. Bacteria recovery at 10 CFU/mL concentration demonstrated all-positive results, making this new separation method a promising candidate for future clinic needs in sepsis diagnosis. / Sepsis, som kännetecknas av svår bakterieinfektion i blodomloppet och leder till hög dödlighet och medicinska kostnader, har blivit en global utmaning inom hälso- och sjukvården. Den nuvarande kliniska diagnosen av sepsis baseras på blododling och kräver lång väntetid, vilket försenar effektiva behandlingar. Den framväxande diagnosmetoden för sepsis förkortar väntetiden men kräver effektiv separation av bakterier från blodet. I detta masterprojekt undersöks en ny filterbaserad metod för att separera bakterier från blodet för att möta den kliniska efterfrågan inom framtida sepsisdiagnostik. Separationsenheten kontrolleras inom storleken av en falcon-rör och är integrerad med en lutande filterinsats. Genom en centrifug manövreras blodceller och bakterier mot filtret och separeras av det. Genom att använda den optimala designen kan enheten återhämta cirka 30% av bakterierna E. coli, K. pneumoniae och S. aureus från 1 ml helblod med koncentrationer så låga som 100 CFU/ml inom 60 minuter, och avvisa 99,4% av de röda blodkropparna från att passera genom filtret. Återhämtningen av bakterier vid en koncentration av 10 CFU/ml visade helt positiva resultat, vilket gör denna nya separationsmetod till en lovande kandidat för framtida kliniska behov inom sepsisdiagnostik.

Neonatal sepsis is a potentially fatal medical conditiondue to an infection and is attributed to about 200 000annual deaths globally. With healthcare systems that are facingconstant challenges, there exists a potential for introducingmachine learning models as a diagnostic tool that can beautomatized within existing workflows and would not entail morework for healthcare personnel. The Herlenius Research Teamat Karolinska Institutet has collected neonatal sepsis data thathas been used for the development of many machine learningmodels across several papers. However, none have tried to studydecision tree ensemble methods. In this paper, random forestand XGBoost models are developed and evaluated in order toassess their feasibility for clinical practice. The data contained24 features of vital parameters that are easily collected througha patient monitoring system. The validation and evaluationprocedure needed special consideration due to the data beinggrouped based on patient level and being imbalanced. Theproposed methods developed in this paper have the potentialto be generalized to other similar applications. Finally, usingthe measure receiver-operating-characteristic area-under-curve(ROC AUC), both models achieved around ROC AUC= 0.84.Such results suggest that the random forest and XGBoost modelsare potentially feasible for clinical practice. Another gainedinsight was that both models seemed to perform better withsimpler models, suggesting that future work could create a moreexplainable model. / Nenatal sepsis är ett potentiellt dödligt‌‌‌ medicinskt tillstånd till följd av en infektion och uppges globalt orsaka 200 000 dödsfall årligen. Med sjukvårdssystem som konstant utsätts för utmaningar existerar det en potential för maskininlärningsmodeller som diagnostiska verktyg automatiserade inom existerande arbetsflöden utan att innebära mer arbete för sjukvårdsanställda. Herelenius forskarteam på Karolinska Institet har samlat ihop neonatal sepsis data som har använts för att utveckla många maskininlärningsmodeller över flera studier. Emellertid har ingen prövat att undersöka beslutsträds ensemble metoder. Syftet med denna studie är att utveckla och utvärdera random forest och XGBoost modeller för att bedöma deras möjligheter i klinisk praxis. Datan innehör 24 attribut av vitalparameterar som enkelt samlas in genom patientövervakningssystem. Förfarandet för validering och utvärdering krävde särskild hänsyn med tanke på att datan var grupperad på patientnivå och var obalanserad. Den föreslagna metoden har potential att generaliseras till andra liknande tillämpningar. Slutligen, genom att använda receiveroperating-characteristic area-under-curve (ROC AUC) måttet kunde vi uppvisa att båda modellerna presterade med ett resultat på ROC AUC= 0.84. Sådana resultat föreslår att både random forest och XGBoost modellerna kan potentiellt användas i klinisk praxis. En annan insikt var att båda modellerna verkade prestera bättre med enklare modeller vilket föreslår att ete skulle kunna vara att skapa en mer förklarlig skininlärningsmodell. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm

Der Skelettmuskel ist ein dynamisches Gewebe, das seine Funktionalität durch Anpassung des Gleichgewichts zwischen Proteinabbau und Proteinsynthese aufrechterhält. Kritisch kranke septische Patienten in der Intensivstation erleiden häufig eine dort erworbene tiefgreifende Muskelschwäche und –atrophie (intensive care unit acquired weakness, ICUAW). Es gibt Hinweise darauf, dass die Regenerationsfähigkeit des Muskels bei kritisch kranken Patienten beeinträchtigt ist. Die Pathogenese der ICUAW ist nur unzureichend verstanden, jedoch werden Sepsis und Entzündungen als führende Risikofaktoren angesehen. In früheren RNA-Sequenzierungsanalysen aus Muskeln septischer Mäuse wurde eine Herunterregulierung des Transforming Growth Factor beta (TGF-β)-Signals und eine erhöhte Genexpression von SPRY domain and SOCS-box containing protein 1 (SPSB1) gefunden. Ich stellte also die Hypothese auf, dass SPSB1 die Proteinhomöostase im Muskel beeinträchtigt, indem es den TGF-β/TβRII-Signalweg beeinflusst und eine entzündungsinduzierte Muskelatrophie verursacht. Die quantitative Echtzeit-Polymerase-Kettenreaktion verifizierte eine erhöhte Spsb1/SPSB1-Expression im Skelettmuskel von septischen Mäusen und ICUAW-Patienten. Die inflammatorischen Zytokine IL-6, IL-1β und Tumor-Nekrose-Faktor induzierten Spsb1 in C2C12-Myozyten in vitro. Die Überexpression von SPSB1 hemmte den TGF-β-Akt-Signalweg durch Destabilisierung von TβRII, was zu einer reduzierten Proteinsynthese in Myozyten führte. Als Konsequenz beeinträchtigte die SPSB1-Überexpression die Myogenese von C2C12-Myoblasten, gemessen an reduzierten Differenzierungs- und Fusionsindizes, sowie einer verminderten Protein- und mRNA-Expression der Differenzierungsfaktoren Mymk, Mymx, Myog, Myh1, 3 und 7. Zusammengenommen hemmt SPSB1 den TβRII-Signalweg im entzündeten Skelettmuskel, was die Myogenese beeinträchtigt. Daher könnte die Hemmung von SPSB1 hilfreich sein, um entzündungsinduziertes Muskelversagen zu verhindern. / Skeletal muscle is a dynamic tissue which maintains its functionality by adapting the balance between protein degradation and protein synthesis. Critically ill septic patients often develop intensive care unit acquired weakness (ICUAW), characterized by profound muscle weakness and atrophy. Emerging evidence suggests that the regenerative ability is impaired in patients with ICUAW. However, the pathogenesis of this disease is poorly understood. Sepsis and inflammation are considered as leading risk factors for ICUAW. In previous RNA sequencing analyses from muscle of septic mice, downregulation of transforming growth factor beta (TGF-β)-signaling and an increased gene expression of SPRY domain and SOCS-box containing protein 1 (SPSB1) have been found. If SPSB1 and TGF-β signaling play a role in inflammation-induced muscle atrophy was unknown. I hypothesized that SPSB1 impairs protein homeostasis in muscle by affecting TGF-β/TβRII signaling and causes inflammation-induced muscle atrophy. Quantitative real-time polymerase chain reaction verified increased Spsb1/SPSB1 expression in skeletal muscle of septic mice and ICUAW patients. The inflammatory cytokines IL-6, IL-1β and tumor necrosis factor induced Spsb1 in C2C12 myocytes in vitro. Overexpression of SPSB1 inhibited the TGF-β-Akt signaling pathway by destabilization of TβRII, leading to reduced protein synthesis in myocytes. These effects on TβRII signaling were mediated by the SPRY- and SOCS-box domains of SPSB1. As a consequence, SPSB1 overexpression impaired myogenesis of C2C12 myoblasts as measured by reduced differentiation and fusion indices, decreased protein and mRNA expression of the differentiation factors Mymk, Mymx, Myog, Myh1, 3 and 7. Taken together, SPSB1 binds and inhibits TβRII signaling in the inflammatory skeletal muscle resulting in impaired myogenesis. Therefore, inhibition of SPSB1 could be useful to prevent inflammation-induced muscle failure.

Bakgrund: 49 miljoner människor runt om i världen drabbas årligen av sepsis samt orsakar 11 miljoner dödsfall. Det är ett allvarligt och livshotande tillstånd där varje timme räknas för att minska dödligheten och därmed krävs en tidig identifiering för en snabb och adekvat behandling samt en säker vård. Syfte: Syftet var att beskriva sjuksköterskors kliniska erfarenheter av att identifiera sepsis hos patienter inom hälso- och sjukvården. Metod: En litteraturöversikt som inkluderade åtta vetenskapliga artiklar publicerade mellan 2003 och 2022. Data inhämtades från databaserna CINAHL och MEDLINE samt genom sekundär sökning. Innehållsanalys utfördes enligt Fribergs fem steg. Resultat: Två huvudkategorier och sex underkategorier identifierades. Sjuksköterskors kliniska erfarenheter av att identifiera sepsis (erfarenhet påverkar tidig identifiering, symtom och varningstecken, klinisk erfarenhet av diffusa symtom). Faktorer som påverkar sjuksköterskornas identifiering av sepsis (kunskap och arbetsstöd, samverkan i team och kommunikation, arbetsbelastningens betydelse).   Slutsats: Sepsis är ett komplext tillstånd som kan vara svårt att identifiera när diffusa symtom och tecken uppvisas. Sjuksköterskornas förmågor att identifiera sepsis påverkas av olika faktorer såsom kunskap, klinisk erfarenhet och arbetsmiljö. / Background: 49 million people around the world suffer from sepsis annually and cause 11 million deaths. It is a serious and life-threatening condition where every hour counts to reduce mortality and thus early identification is required for prompt and adequate treatment for a safe care. Aim: The purpose was to describe nurses' experiences of identifying sepsis in patients within the health care. Method: A literature review that included eight scientific articles published between 2003 and 2022. Data were obtained from the databases CINAHL, MEDLINE and by secondary search. Content analysis was performed according to Friberg's five steps. Results: Two main categories and six subcategories were identified. Nurses' clinical experience in identifying sepsis (experience affects early identification, symptoms and warning signs, clinical experience of diffuse symptoms). Factors influencing the nurses' identification of sepsis (knowledge and work support, teamwork and communication, workload importance). Conclusion: Sepsis is a complex condition that can be difficult to identify when diffuse symptoms and signs are present. The nurses' abilities to identify sepsis are affected by various factors such as knowledge, clinical experience and work environment.

No description available.

Bakgrund: Sepsis är ett allvarligt tillstånd med hög risk för mortalitet och drabbar miljontals människor världen över. Att som sjuksköterska tidigt kunna identifiera tillståndet innebär större chans till adekvat behandling och mindre lidande för patienten. Sjuksköterskan ansvarar för säker vård och god omvårdnad, där kunskap och klinisk intuition utgör en viktig del i observationen av patienten. Syfte: Syftet var att undersöka vilka faktorer som kan påverka sjuksköterskans identifiering av sepsis.MetodEn allmän litteraturöversikt valdes som metod. Relevanta artiklar hämtades ur databaserna PubMed och Cinahl Complete. Datainsamlingen gjordes på tolv originalartiklar, där majoriteten av studierna var av kvantitativ design. Samtliga artiklar genomgick kvalitetsgranskning och analyserades enligt relevant kurslitteratur. Resultat: Två huvudteman identifierades: arbetsmiljöfaktorer och kunskap hos sjuksköterskan. Två underkategorier till arbetsmiljöfaktorer framkom: kommunikation inom vårdteamet och vikten av mätverktyg och observationer. Hög arbetsbelastning, bristande kunskap och kommunikation samt otillräcklig användning av mätverktyg vid observationer, var faktorer som hindrade identifieringen av sepsis. Arbetslivserfarenhet, god kunskap och självförtroende samt adekvat användning av mätverktyg visade sig vara främjande faktorer. Slutsats: Mätverktyg, kommunikation, kunskap och arbetsmiljö var faktorer som påverkade sjuksköterskans identifiering av sepsis. Sjuksköterskan bör ges rätt förutsättningar för att tidigt kunna identifiera sepsis. En god arbetsmiljö och kommunikation i vårdteamet skulle kunna minska sjuksköterskans etiska stress, höja arbetsmoralen och leda till en mer säker vård. Kontinuerlig kompetensutveckling på arbetsplatsen kan också öka sjuksköterskans självförtroende. En god arbetsmiljö, kunskap, kommunikation och adekvat användning av mätverktyg är således av största vikt för sjuksköterskans identifiering av sepsis. / Background: Sepsis is a common condition with a high risk of mortality, affecting millions of people worldwide. As a nurse, being able to identify the condition early means a higher chance of adequate treatment and less suffering for the patient. The nurse is responsible for safe care and good nursing, where knowledge and clinical intuition are an important part of the observation of the patient. Aim: The aim was to investigate the factors that influence nurse´s identification of sepsis.MethodA general literature review was chosen as the method. Data collection was based on the PubMed and Cinahl Complete databases. Twelve original articles were selected, where most of the studies were of quantitative design. All articles underwent quality review and were analyzed according to relevant course literature. Results: Two main themes were identified: work environment factors and nurse knowledge. Two subcategories to work environment factors emerged: communication within the care team and the importance of measurement tools and observations. High workload, lack of knowledge and communication, and insufficient use of measurement tools during observations were factors that hindered the identification of sepsis. Work experience, good knowledge and confidence, and adequate use of measurement tools were found to be facilitating factors. Conclusion: Measurement tools, communication, knowledge, and work environment were factors that influenced nurses' identification of sepsis. Nurses should be given the right conditions to identify sepsis at an early stage. A good working environment and communication within the healthcare team could reduce nurses' ethical stress, increase morale and lead to safer care. Continuous professional development in the workplace could also increase nurses' confidence. A good work environment, knowledge, communication, and adequate use of measurement tools is shown to be of great importance for nurses’ identification of sepsis.

Monocytic human leukocyte antigen-D related (mHLA-DR) expression determined by flow cytometry has been suggested as a biomarker of sepsisinduced immunosuppression. In order to facilitate use of HLA-DR in clinical practice, a quantitative real-time PCR technique measuring HLA-DR at the transcription level was developed and evalutated. Levels of HLA-DR mRNA correlated to mHLADR expression and were robustly measured, with high reproducibility, during the course of infection. Dynamics of mHLA-DR expression was studied during the first weeks of bloodstream infection (BSI) and was found to be dependent on the bacterial etiology of BSI. Moreover, mHLA-DR was shown to be inversely related to markers of inflammation. In patients with unfavourable outcome, sustained high C-reactive protein level and high neutrophil count were demonstrated along with low mHLA-DR expression and low lymphocyte count. This supports the theory of sustained inflammation in sepsis-induced immunosuppression. The association between mHLA-DR and bacterial etiology may be linked to the clinical trajectory via differences in ability to cause intractable infection. Staphylococcus aureus was the dominating etiology among cases with unfavourable outcome. With focus on patients with S. aureus BSI, those with complicated S. aureus BSI were found to have lower HLA-DR mRNA expression during the first week than those with uncomplicated S. aureus BSI. If these results can be confirmed in a larger cohort, HLA-DR measurement could possibly become an additional tool for early identification of patients who require further investigation to clear infectious foci and achieve source control. In conclusion, PCR-based measurement of HLA-DR is a promising method for measurements of the immune state in BSI, but needs further evaluation in the intensive care unit setting to define the predictive and prognostic value for deleterious immunosuppression. The etiology of infection should be taken into consideration in future studies of translational immunology in sepsis.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.