Development and clinical impact assessment of a machine-learning model for early prediction of late-onset sepsis.

BACKGROUND AND AIM: Preterm infants are prone to neonatal infections such as late-onset sepsis (LOS). The consequences of LOS can be severe and potentially life-threatening. Unfortunately, LOS often presents with unspecific symptoms, and early screening laboratory tests have limited diagnostic value and are often late. This study aimed to build a predictive algorithm to aid doctors in the early detection of LOS in very preterm infants. METHODS: In a retrospective cohort study, all consecutively admitted preterm infants (GA ≤ 32 weeks) from 2008 until 2019 were included. They were classified as LOS or control according to blood culture results, currently the gold standard. To generate features, routine and continuously measured oxygen saturation and heart rate data with a minute-by-minute sampling rate were extracted from electronic medical records. Care was taken not to include variables indicative of existing LOS suspicion. The timing of a positive blood culture served as a proxy for LOS-onset. An equivalent timestamp was generated in gestational-age-matched control patients without a positive blood culture. Three machine learning (ML) techniques (generalized additive models, logistic regression, and XGBoost) were used to build a classification algorithm. To simulate the performance of the algorithm in clinical practice, a simulation using multiple alarm thresholds was performed on hourly predictions for the total hospitalization period. RESULTS: 292 infants with LOS were matched to 1497 controls. The median gestational age before matching was 28.1 and 30.3 weeks, respectively. Evaluation of the overall discriminative power of the LR algorithm yielded an AUC of 0.73 (p < 0.05) at the moment of clinical suspicion (t = 0). In the longitudinal simulation, our algorithm detects LOS in at least 47% of the patients before clinical suspicion without exceeding the alarm fatigue threshold of 3 alarms per day. Furthermore, medical experts evaluated the algorithm as clinically relevant regarding the feature contributions in the model explanations. CONCLUSIONS: An ML algorithm was trained for the early detection of LOS. Performance was evaluated on both prediction horizons and in a clinical impact simulation. To the best of our knowledge, our assessment of clinical impact with a retrospective simulation on longitudinal data is the most extensive in the literature on LOS prediction to date. The clinically relevant algorithm, based on routinely collected data, can potentially accelerate clinical decisions in the early detection of LOS, even with limited inputs.

Strategies to prevent and manage running-related knee injuries: a systematic review of randomised controlled trials.

OBJECTIVE: To evaluate the effectiveness of interventions to prevent and manage knee injuries in runners. DESIGN: Systematic review and meta-analysis. DATA SOURCES: MEDLINE, EMBASE, CINAHL, Web of Science and SPORTDiscus up to May 2022. ELIGIBILITY CRITERIA FOR SELECTING STUDIES: Randomised controlled trials (RCTs) with a primary aim of evaluating the effectiveness of intervention(s) to prevent or manage running-related knee injury. RESULTS: Thirty RCTs (18 prevention, 12 management) analysed multiple interventions in novice and recreational running populations. Low-certainty evidence (one trial, 320 participants) indicated that running technique retraining (to land softer) reduced the risk of knee injury compared with control treadmill running (risk ratio (RR) 0.32, 95% CI 0.16 to 0.63). Very low-certainty to low-certainty evidence from 17 other prevention trials (participant range: 24 -3287) indicated that various footwear options, multicomponent exercise therapy, graduated running programmes and online and in person injury prevention education programmes did not influence knee injury risk (RR range: 0.55-1.06). In runners with patellofemoral pain, very low-certainty to low-certainty evidence indicated that running technique retraining strategies, medial-wedged foot orthoses, multicomponent exercise therapy and osteopathic manipulation can reduce knee pain in the short-term (standardised mean difference range: -4.96 to -0.90). CONCLUSION: There is low-certainty evidence that running technique retraining to land softer may reduce knee injury risk by two-thirds. Very low-certainty to low-certainty evidence suggests that running-related patellofemoral pain may be effectively managed through a variety of active (eg, running technique retraining, multicomponent exercise therapy) and passive interventions (eg, foot orthoses, osteopathic manipulation). PROSPERO REGISTRATION NUMBER: CRD42020150630.

Population Trends of the Sugarcane Borer (Lepidoptera: Crambidae) in Louisiana Sugarcane.

The sugarcane borer, Diatraea saccharalis (F.) (Lepidoptera: Crambidae), is the primary pest of sugarcane, Saccharum spp., in Louisiana. Spring populations are not considered economically damaging, but quantifying infestations can provide an indication of the spatial and temporal character of the damaging summer populations. Statewide surveys quantified the density of sugarcane tillers killed by D. saccharalis (deadhearts) from sugarcane fields across the state in spring from 2003 to 2020. Deadheart density varied greatly among years with a high of 1,318/ha in 2003 to a low of 0/ha in 2018. Linear regressions of the 3-yr rolling average showed declines in spring D. saccharalis populations and the percentage of acreage treated with insecticides over 17 yr. Weather factors including minimum winter temperatures and average spring temperatures were poor predictors of D. saccharalis populations. Only total precipitation in the month of April was positively correlated with numbers of deadhearts per hectare. Results suggest overwintering mortality is not a key factor influencing populations of the first generation of D. saccharalis in Louisiana. Total precipitation in the month of July was positively associated with percentage of treated acreage. Spring deadheart density was directly related to percentage of acreage treated with insecticides during the summer. Quantifying first-generation D. saccharalis populations by recording deadheart density can aid in predicting pest pressure later in the growing season.

West Indian Canefly (Hemiptera: Delphacidae): An Emerging Pest of Louisiana Sugarcane.

The West Indian canefly, Saccharosydne saccharivora (Westwood) (Hemiptera: Delphacidae), is a sporadic pest of sugarcane in Louisiana which has recently emerged as a more consistent threat with outbreaks occurring in 2012, 2016, 2017, and 2019. Surveys of commercial fields in 2016 revealed that S. saccharivora infestations were present throughout Louisiana sugarcane and populations peaked in mid-June before declining. High minimum winter temperatures are generally associated with S. saccharivora outbreaks. Six insecticide evaluations demonstrated effective control with several insecticides including λ-cyhalothrin, flupyradifurone, acetamiprid, and imidacloprid. In five of the six insecticide trials, S. saccharivora infestations had substantially declined by 21 d after treatment. Effects of insecticidal control of S. saccharivora on sugar yields were detected in one of four small plot trials in which yield data were collected. Linear regression revealed S. saccharivora cumulative insect days in a grid sampling study were inversely associated with sugar yields. Results from these collective experiments suggest impacts on sugar yields are influenced by pest density and infestation duration. Differences were detected in numbers of S. saccharivora nymphs and adults as well as sooty mold coverage among commercial sugarcane cultivars with more than twofold increases in the most susceptible compared to resistant cultivars. The research presented herein documents the impact of S. saccharivora to Louisiana sugarcane and provides important ground work for developing effective pest management strategies. Future research efforts should aim to identify ecological factors influencing population dynamics, varietal preferences, and economic thresholds.