Dynamic heterogeneity in COVID-19: Insights from a mathematical model.

Critical illness, such as severe COVID-19, is heterogenous in presentation and treatment response. However, it remains possible that clinical course may be influenced by dynamic and/or random events such that similar patients subject to similar injuries may yet follow different trajectories. We deployed a mechanistic mathematical model of COVID-19 to determine the range of possible clinical courses after SARS-CoV-2 infection, which may follow from specific changes in viral properties, immune properties, treatment modality and random external factors such as initial viral load. We find that treatment efficacy and baseline patient or viral features are not the sole determinant of outcome. We found patients with enhanced innate or adaptive immune responses can experience poor viral control, resolution of infection or non-infectious inflammatory injury depending on treatment efficacy and initial viral load. Hypoxemia may result from poor viral control or ongoing inflammation despite effective viral control. Adaptive immune responses may be inhibited by very early effective therapy, resulting in viral load rebound after cessation of therapy. Our model suggests individual disease course may be influenced by the interaction between external and patient-intrinsic factors. These data have implications for the reproducibility of clinical trial cohorts and timing of optimal treatment.

In silico clinical studies for optimal COVID-19 vaccination schedules in patients with cancer.

This study introduces a tailored COVID-19 model for patients with cancer, incorporating viral variants and immune-response dynamics. The model aims to optimize vaccination strategies, contributing to personalized healthcare for vulnerable groups.

How Treatment Effect Heterogeneity Works.

How Treatment Effect Heterogeneity WorksThis Stats, STAT! animated video explores the concept of treatment effect heterogeneity. Differences in the effectiveness of treatments across participants in a clinical trial is important to understand when deciding how to apply clinical trial results to clinical practice.

Mechanistic model for booster doses effectiveness in healthy, cancer, and immunosuppressed patients infected with SARS-CoV-2.

SARS-CoV-2 vaccines are effective at limiting disease severity, but effectiveness is lower among patients with cancer or immunosuppression. Effectiveness wanes with time and varies by vaccine type. Moreover, previously prescribed vaccines were based on the ancestral SARS-CoV-2 spike-protein that emerging variants may evade. Here, we describe a mechanistic mathematical model for vaccination-induced immunity. We validate it with available clinical data and use it to simulate the effectiveness of vaccines against viral variants with lower antigenicity, increased virulence, or enhanced cell binding for various vaccine platforms. The analysis includes the omicron variant as well as hypothetical future variants with even greater immune evasion of vaccine-induced antibodies and addresses the potential benefits of the new bivalent vaccines. We further account for concurrent cancer or underlying immunosuppression. The model confirms enhanced immunogenicity following booster vaccination in immunosuppressed patients but predicts ongoing booster requirements for these individuals to maintain protection. We further studied the impact of variants on immunosuppressed individuals as a function of the interval between multiple booster doses. Our model suggests possible strategies for future vaccinations and suggests tailored strategies for high-risk groups.

Prolonged Prone Position Ventilation Is Associated With Reduced Mortality in Intubated COVID-19 Patients.

BACKGROUND: Prone position ventilation (PPV) is resource-intensive, yet the optimal strategy for PPV in intubated patients with COVID-19 is unclear. RESEARCH QUESTION: Does a prolonged (24 or more h) PPV strategy improve mortality in intubated COVID-19 patients compared with intermittent (∼16 h with daily supination) PPV? STUDY DESIGN AND METHODS: Multicenter, retrospective cohort study of consecutively admitted intubated COVID-19 patients treated with PPV between March 11 and May 31, 2020. The primary outcome was 30-day all-cause mortality. Secondary outcomes included 90-day all-cause mortality and prone-related complications. Inverse probability treatment weights (IPTW) were used to control for potential treatment selection bias. RESULTS: Of the COVID-19 patients who received PPV, 157 underwent prolonged and 110 underwent intermittent PPV. Patients undergoing prolonged PPV had reduced 30-day (adjusted hazard ratio [aHR], 0.475; 95% CI, 0.336-0.670; P < .001) and 90-day (aHR, 0.638; 95% CI, 0.461-0.883; P = .006) mortality compared with intermittent PPV. In patients with Pao2/Fio2 ≤ 150 at the time of pronation, prolonged PPV was associated with reduced 30-day (aHR, 0.357; 95% CI, 0.213-0.597; P < .001) and 90-day mortality (aHR, 0.562; 95% CI, 0.357-0.884; P = .008). Patients treated with prolonged PPV underwent fewer pronation and supination events (median, 1; 95% CI, 1-2 vs 3; 95% CI, 1-4; P < .001). PPV strategy was not associated with overall PPV-related complications, although patients receiving prolonged PPV had increased rates of facial edema and lower rates of peri-proning hypotension. INTERPRETATION: Among intubated COVID-19 patients who received PPV, prolonged PPV was associated with reduced mortality. Prolonged PPV was associated with fewer pronation and supination events and a small increase in rates of facial edema. These findings suggest that prolonged PPV is a safe, effective strategy for mortality reduction in intubated COVID-19 patients.

Good Intentions to Treat.

Good Intentions to Treat This Stats, STAT! animated video explores common approaches to analyzing data from randomized controlled trials, including intention-to-treat, per-protocol, and as-treated analyses.

How Statistical Power Works.

How Statistical Power WorksThis Stats, STAT! animated video explores the concept of statistical power and explains how clinical investigators determine how many participants to enroll in a randomized trial.

Bayesian Way.

Bayesian WayThis animated video explores two possible approaches to analyzing data in a randomized controlled trial: "Frequentist" versus "Bayesian."

Large Language Models.

Large Language ModelsIn the latest edition of Stats, STAT!, Fralick and colleagues explain the statistics behind large language models - used in chat bots like ChatGPT and Bard. While these new tools may seem remarkably intelligent, at their core they just assemble sentences based on statistics from large amounts of text.

How Censoring Works.

How Censoring WorksA common challenge in clinical research is determining the time to occurrence of a given event. This animated video explores the concept of censoring in survival analysis and how investigators deal with ambiguity in the time of an event's occurrence.