External Validation of Prognostic Models in Critical Care: A Cautionary Tale From COVID-19 Pneumonitis.

OBJECTIVES BACKGROUND: To externally validate clinical prediction models that aim to predict progression to invasive ventilation or death on the ICU in patients admitted with confirmed COVID-19 pneumonitis. DESIGN: Single-center retrospective external validation study. DATA SOURCES: Routinely collected healthcare data in the ICU electronic patient record. Curated data recorded for each ICU admission for the purposes of the U.K. Intensive Care National Audit and Research Centre (ICNARC). SETTING: The ICU at Manchester Royal Infirmary, Manchester, United Kingdom. PATIENTS: Three hundred forty-nine patients admitted to ICU with confirmed COVID-19 Pneumonitis, older than 18 years, from March 1, 2020, to February 28, 2022. Three hundred two met the inclusion criteria for at least one model. Fifty-five of the 349 patients were admitted before the widespread adoption of dexamethasone for the treatment of severe COVID-19 (pre-dexamethasone patients). OUTCOMES: Ability to be externally validated, discriminate, and calibrate. METHODS: Articles meeting the inclusion criteria were identified, and those that gave sufficient details on predictors used and methods to generate predictions were tested in our cohort of patients, which matched the original publications' inclusion/exclusion criteria and endpoint. RESULTS: Thirteen clinical prediction articles were identified. There was insufficient information available to validate models in five of the articles; a further three contained predictors that were not routinely measured in our ICU cohort and were not validated; three had performance that was substantially lower than previously published (range C-statistic = 0.483-0.605 in pre-dexamethasone patients and C = 0.494-0.564 among all patients). One model retained its discriminative ability in our cohort compared with previously published results (C = 0.672 and 0.686), and one retained performance among pre-dexamethasone patients but was poor in all patients (C = 0.793 and 0.596). One model could be calibrated but with poor performance. CONCLUSIONS: Our findings, albeit from a single center, suggest that the published performance of COVID-19 prediction models may not be replicated when translated to other institutions. In light of this, we would encourage bedside intensivists to reflect on the role of clinical prediction models in their own clinical decision-making.

A Tale of Two Waves: Changes in the Use of Noninvasive Ventilation and Prone Positioning in Critical Care Management of Coronavirus Disease 2019.

New treatments and increased experience are changing the management of hospitalized coronavirus disease 2019 patients but the impact on ICU management is unclear. OBJECTIVES: To examine characteristics, ventilatory management, and outcomes of critically ill patients in two distinct waves of the pandemic. DESIGN SETTING AND PARTICIPANTS: Observational cohort study in an ICU in a single-center university-affiliated U.K. hospital. Two-hundred ten adults with coronavirus disease 2019 admitted to ICU between March 17, 2020, to May 31, 2020, and September 1, 2020, to December 10, 2020, with hourly data and 100% follow-up to ICU discharge. MAIN OUTCOMES AND MEASURES: Data were extracted from the electronic medical record for patient characteristics and clinical data. Patients were classified into distinct waves of the pandemic and assessed for differences between the two waves. RESULTS: The duration of noninvasive ventilation/nasal high flow increased in wave 2 versus wave 1, both in self-ventilating patients (107 vs 72 hr; p = 0.02), and in those ultimately requiring invasive mechanical ventilation (34 vs 10 hr; p = 0.02). The proportion of survivors treated without invasive mechanical ventilation increased in wave 2 (59% vs 39%; p = 0.01). In both waves, longer duration of noninvasive ventilation/nasal high flow prior to intubation was associated with higher ICU mortality (survivors 10 hr [4-21 hr] vs nonsurvivors 50 hr [23-124 hr]; p < 0.01). Proned invasive mechanical ventilation was common (54.7%) and prolonged. In wave 2, invasive mechanical ventilation patients were generally more hypoxic with proning initiated at lower Pao2/Fio2 ratios (81 vs 116 mm Hg; p = 0.02) and yielding smaller improvements in Fio2 requirements. Continued proning episodes despite poor responses were commonplace and typically futile. Length of stay for patients requiring tracheostomy increased markedly in wave 2 (51.3 vs 33.7 d; p = 0.03). Overall survival remained similar in wave 2 (68.0% vs 60.9%; p = 0.31). CONCLUSIONS AND RELEVANCE: Our data suggest that management of critically ill coronavirus disease 2019 patients is changing with more survivors avoiding invasive mechanical ventilation. Duration of noninvasive ventilation/nasal high flow use is increasing, which may be associated with worsening outcomes for individuals who require invasive mechanical ventilation. Among invasively ventilated patients, changes in the use of and response to prone positioning and increased length of stay following tracheostomy may imply that the care of these patients is becoming more challenging.

Maternal smoking during pregnancy and fetal organ growth: a magnetic resonance imaging study.

OBJECTIVE: To study whether maternal cigarette smoking during pregnancy is associated with alterations in the growth of fetal lungs, kidneys, liver, brain, and placenta. DESIGN: A case-control study, with operators performing the image analysis blinded. SETTING: Study performed on a research-dedicated magnetic resonance imaging (MRI) scanner (1.5 T) with participants recruited from a large teaching hospital in the United Kingdom. PARTICIPANTS: A total of 26 pregnant women (13 current smokers, 13 non smokers) were recruited; 18 women (10 current smokers, 8 nonsmokers) returned for the second scan later in their pregnancy. METHODS: Each fetus was scanned with MRI at 22-27 weeks and 33-38 weeks gestational age (GA). MAIN OUTCOME MEASURES: Images obtained with MRI were used to measure volumes of the fetal brain, kidneys, lungs, liver and overall fetal size, as well as placental volumes. RESULTS: Exposed fetuses showed lower brain volumes, kidney volumes, and total fetal volumes, with this effect being greater at visit 2 than at visit 1 for brain and kidney volumes, and greater at visit 1 than at visit 2 for total fetal volume. Exposed fetuses also demonstrated lower lung volume and placental volume, and this effect was similar at both visits. No difference was found between the exposed and nonexposed fetuses with regards to liver volume. CONCLUSION: Magnetic resonance imaging has been used to show that maternal smoking is associated with reduced growth of fetal brain, lung and kidney; this effect persists even when the volumes are corrected for maternal education, gestational age, and fetal sex. As expected, the fetuses exposed to maternal smoking are smaller in size. Similarly, placental volumes are smaller in smoking versus nonsmoking pregnant women.