DORA Health Service:

Comparison of Random Forest and Stepwise Regression for Variable Selection Using Low Prevalence Predictors: A case Study in Paediatric Sepsis

2026-08-11T01:00:01Z

Title: Comparison of Random Forest and Stepwise Regression for Variable Selection Using Low Prevalence Predictors: A case Study in Paediatric Sepsis Authors: Gilholm, Patricia; Lister, Paula; Irwin, Adam; Harley, Amanda; Raman, Sainath; Schlapbach, Luregn J; Gibbons, Kristen S Abstract: Variable selection is a common technique to identify the most predictive variables from a pool of candidate predictors. Low prevalence predictors (LPPs) are frequently found in clinical data, yet few studies have explored their impact on model performance during variable selection. This study compared the Random Forest (RF) algorithm and stepwise regression (SWR) for variable selection using data from a paediatric sepsis screening tool, where 18 out of 32 predictors had a prevalence < 10%. Variable selection using RF was compared to forward and backward SWR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), and the variables retained. Additionally, a simulation study assessed how increasing the prevalence of the predictors impacted the variable selection results. The best fitting RF and SWR models retained were 22, and 17 predictors, respectively, with 14 and 10 predictors having a prevalence < 10%. Both the RF and SWR models had similar predictive performance (RF: AUC [95% Confidence Interval] 0.79 [0.77, 0.81], LR: 0.80 [0.78, 0.82]). The simulation study revealed differences for both RF and SWR models in variable importance rankings and predictor selection with increasing prevalence thresholds, particularly for moderately and strongly associated predictors. The RF algorithm retained a number of very low prevalence predictors compared to SWR. However, the predictive performance of both models were comparable, demonstrating that when applied correctly and the number of candidate predictors is small, both methods are suitable for variable selection when using low prevalence predictors.

"You're listening but you're not hearing": qualitative exploration of parents' lived experience of paediatric sepsis

2026-08-10T05:00:01Z

Title: "You're listening but you're not hearing": qualitative exploration of parents' lived experience of paediatric sepsis Authors: O'Keefe, Meagan; Crowe, Liz; Lister, Paula; Schlapbach, Luregn J; Simons, Megan Abstract: Sepsis is a life-threatening condition of significant mortality and morbidity burden in children. There is limited evidence exploring the lived experiences of parents whose children have survived or died from sepsis, nor their support needs, despite growing awareness of poor sepsis outcomes. This study aims to understand the lived experiences of parents of children affected by sepsis, both surviving and bereaved, and identify gaps in support services to inform future interventions. A phenomenological-hermeneutic design was used, with data collected through a focus group and individual interviews with 11 parents of children treated for paediatric sepsis in Queensland, Australia. Reflexive thematic analysis was used to analyse the data. Four themes were generated that present an in-depth understanding of parents' lived experiences of having a child with sepsis and what their support needs are: 1. Rupture of Life as We Knew It-parents experienced a distinct rupture between life before and after their child's sepsis diagnosis, often accompanied by trauma and grief; 2. Navigating in the Dark-parents described feeling confused and isolated, driven by a lack of awareness about sepsis, inadequate communication from healthcare providers, and limited access to specialised support; 3. The Weight of 'What If'-guilt, regret and disillusionment with the healthcare system were common among parents, particularly regarding missed early warning signs and opportunities for intervention; and 4. Call for Change-parents advocated for improved public and professional education about paediatric sepsis, trauma-informed communication from healthcare professionals, and the development of structured, specialised support networks. Parents of children affected by sepsis experience profound distress and isolation, compounded by a lack of awareness, inadequate communication, and limited specialised support services. They identified that urgent action is needed to develop paediatric sepsis-specific family support networks, enhance public and clinical education, and implement trauma-informed care to reduce the burden of sepsis on all families.

Executive Summary: Surviving Sepsis Campaign International Guidelines for the Management of Septic Shock and Sepsis-Associated Organ Dysfunction in Children

2026-08-03T01:00:00Z

Title: Executive Summary: Surviving Sepsis Campaign International Guidelines for the Management of Septic Shock and Sepsis-Associated Organ Dysfunction in Children Authors: Weiss, Scott L; Peters, Mark J; Alhazzani, Waleed; Agus, Michael S D; Flori, Heidi R; Inwald, David P; Nadel, Simon; Schlapbach, Luregn J; Tasker, Robert C; Argent, Andrew C; Brierley, Joe; Carcillo, Joseph; Carrol, Enitan D; Carroll, Christopher L; Cheifetz, Ira M; Choong, Karen; Cies, Jeffry J; Cruz, Andrea T; De Luca, Daniele; Deep, Akash; Faust, Saul N; De Oliveira, Claudio Flauzino; Hall, Mark W; Ishimine, Paul; Javouhey, Etienne; Joosten, Koen F M; Joshi, Poonam; Karam, Oliver; Kneyber, Martin C J; Lemson, Joris; MacLaren, Graeme; Mehta, Nilesh M; Møller, Morten Hylander; Newth, Christopher J L; Nguyen, Trung C; Nishisaki, Akira; Nunnally, Mark E; Parker, Margaret M; Paul, Raina M; Randolph, Adrienne G; Ranjit, Suchitra; Romer, Lewis H; Scott, Halden F; Tume, Lyvonne N; Verger, Judy T; Williams, Eric A; Wolf, Joshua; Wong, Hector R; Zimmerman, Jerry J; Kissoon, Niranjan; Tissieres, Pierre

Identification of distinct clinical profiles of sepsis risk in paediatric emergency department patients using Bayesian profile regression

2025-06-23T07:25:34Z

Title: Identification of distinct clinical profiles of sepsis risk in paediatric emergency department patients using Bayesian profile regression Authors: Patricia Gilholm; Sainath Raman; Adam Irwin; Paula Lister; Amanda Harley; Luregn J Schlapbach; Kristen S Gibbons