Matteo Sesia

This paper develops a conformal method to compute prediction intervals for non-parametric regression that can automatically adapt to skewed data. Leveraging black-box machine learning algorithms to estimate the conditional distribution of the outcome using histograms, it translates their output into the shortest prediction intervals with approximate conditional coverage. The resulting prediction intervals provably have marginal coverage in finite samples, while asymptotically achieving conditional coverage and optimal length if the black-box model is consistent. Numerical experiments with simulated and real data demonstrate improved performance compared to state-of-the-art alternatives, including conformalized quantile regression and other distributional conformal prediction approaches.

Supratherapeutic oxygen levels consistently cause oxygen toxicity in the lungs and other organs. The prevalence and severity of hyperoxemia among pediatric intensive care unit (PICU) patients remain unknown. This was the first study to examine the prevalence and duration of hyperoxemia in PICU patients receiving oxygen therapy. This is a retrospective chart review. This was performed in a setting of 36-bed PICU in a quaternary-care children's hospital. All the patients were children aged <18 years, admitted to the PICU for ≥24 hours, receiving oxygen therapy for ≥12 hours who had at least one arterial blood gas during this time.
There was no intervention. Of 5,251 patients admitted to the PICU, 614 were included in the study. On average, these patients received oxygen therapy for 91% of their time in the PICU and remained hyperoxemic, as measured by pulse oximetry, for 65% of their time on oxygen therapy. Patients on oxygen therapy remained hyperoxemic for a median of 38 hours per patient and only 1.1% of patients did not experience any hyperoxemia. Most of the time (87.5%) patients received oxygen therapy through a fraction of inspired oxygen (FiO₂)-adjustable device. Mean FiO₂ on noninvasive support was 0.56 and on invasive support was 0.37. Mean partial pressure of oxygen (PaO₂) on oxygen therapy was 108.7 torr and 3,037 (42.1%) of PaO₂ measurements were >100 torr. Despite relatively low FiO₂, PICU patients receiving oxygen therapy are commonly exposed to prolonged hyperoxemia, which may contribute to ongoing organ injury.

Robotic-assisted esophagectomy results in similar pain with significantly less analgesic
consumption compared to open and hybrid surgery. Despite longer operation time, RAMIE
patients require shorter post-operative ICU stays compared to open and hybrid surgery
patients, and experience lower blood loss compared to open surgery patients.

In this article we develop a method based on model-X knockoffs to find conditional associations that are consistent across environments, while controlling the false discovery rate. The motivation for this problem is that large datasets may contain numerous associations that are statistically significant and yet misleading, as they are induced by confounders or sampling imperfections. However, associations replicated under different conditions may be more interesting. In fact, sometimes consistency provably leads to valid causal inferences even if conditional associations do not. Although the proposed method is widely applicable, in this paper we highlight its relevance to genome-wide association studies, in which robustness across populations with diverse ancestries mitigates confounding due to unmeasured variants. The effectiveness of this approach is demonstrated by simulations and applications to UK Biobank data.

Genome-wide association studies compare a phenotype to thousands of genetic variants, searching for associations of potential biological interest. Standard analyses rely on linear models of the phenotype given one variable at a time. However, their assumptions are difficult to verify and their univariate approaches make it hard to recognize interesting associations from spurious ones. Our work takes a different path: We analyze all variants simultaneously, modelling the randomness in the genotypes, which is better understood, instead of the phenotype. Our solution accounts for linkage disequilibrium and population structure, controls the false discovery rate, and leverages powerful machine-learning tools. Applications to the UK Biobank data indicate increased power compared to state-of-the-art alternatives and high replicability.