Dose length product to effective dose coefficients in adults.

OBJECTIVES: The most accurate method for estimating patient effective dose (a principal metric for tracking patient radiation exposure) from computed tomography (CT) requires time-intensive Monte Carlo simulation. A simpler method multiplies a scalar coefficient by the widely available scanner-reported dose length product (DLP) to estimate effective dose. We developed new adult effective dose coefficients using actual patient scans and assessed their agreement with Monte Carlo simulation. METHODS: A multicenter sample of 216,906 adult CT scans was prospectively assembled in 2015-2020 from the University of California San Francisco International CT Dose Registry and the University of Florida library of computational phantoms. We generated effective dose coefficients for eight body regions, stratified by patient sex, diameter, and scanner manufacturer. We applied the new coefficients to DLPs to calculate effective doses and assess their correlations with Monte Carlo radiation transport-generated effective dose. RESULTS: Effective dose coefficients varied by body region and decreased in magnitude with increasing patient diameter. Coefficients were approximately twofold higher for torso scans in smallest compared with largest diameter categories. For example, abdomen and pelvis coefficients decreased from 0.027 to 0.013 mSv/mGy-cm between the 16-20 cm and 41+ cm categories. There were modest but consistent differences by sex and manufacturer. Diameter-based coefficients used to estimate effective dose produced strong correlations with the reference standard (Pearson correlations 0.77-0.86). The reported conversion coefficients differ from previous studies, particularly in neck CT. CONCLUSIONS: New effective dose coefficients derived from empirical clinical scans can be used to easily estimate effective dose using scanner-reported DLP. CLINICAL RELEVANCE STATEMENT: Scalar coefficients multiplied by DLP offer a simple approximation to effective dose, a key radiation dose metric. New effective dose coefficients from this study strongly correlate with gold standard, Monte Carlo-generated effective dose, and differ somewhat from previous studies. KEY POINTS: • Previous effective dose coefficients were derived from theoretical models rather than real patient data. • The new coefficients (from a large registry/phantom library) differ from previous studies. • The new coefficients offer reasonably reliable values for estimating effective dose.

Associated radiation exposure from medical imaging and excess lifetime risk of developing cancer in pediatric patients with pulmonary hypertension.

Pediatric patients with pulmonary hypertension (PH) receive imaging studies that use ionizing radiation (radiation) such as computed tomography (CT) and cardiac catheterization to guide clinical care. Radiation exposure is associated with increased cancer risk. It is unknown how much radiation pediatric PH patients receive. The objective of this study is to quantify radiation received from imaging and compute associated lifetime cancer risks for pediatric patients with PH. Electronic health records between 2012 and 2022 were reviewed and radiation dose data were extracted. Organ doses were estimated using Monte Carlo modeling. Cancer risks for each patient were calculated from accumulated exposures using National Cancer Institute tools. Two hundred and forty-nine patients with PH comprised the study cohort; 97% of patients had pulmonary arterial hypertension, PH due to left heart disease, or PH due to chronic lung disease. Mean age at the time of the first imaging study was 2.5 years (standard deviation [SD] = 4.9 years). Patients underwent a mean of 12 studies per patient per year, SD = 32. Most (90%) exams were done in children <5 years of age. Radiation from CT and cardiac catheterization accounted for 88% of the total radiation dose received. Cumulative mean effective dose was 19 mSv per patient (SD = 30). Radiation dose exposure resulted in a mean increased estimated lifetime cancer risk of 7.6% (90% uncertainty interval 3.0%-14.2%) in females and 2.8% (1.2%-5.3%) in males. Careful consideration for the need of radiation-based imaging studies is warranted, especially in the youngest of children.

Dose length product to effective dose coefficients in children.

BACKGROUND: The most accurate method for estimating effective dose (the most widely understood metric for tracking patient radiation exposure) from computed tomography (CT) requires time-intensive Monte Carlo simulation. A simpler method multiplies a scalar coefficient by the widely available scanner-reported dose length product (DLP) to estimate effective dose. OBJECTIVE: Develop pediatric effective dose coefficients and assess their agreement with Monte Carlo simulation. MATERIALS AND METHODS: Multicenter, population-based sample of 128,397 pediatric diagnostic CT scans prospectively assembled in 2015-2020 from the University of California San Francisco International CT Dose Registry and the University of Florida library of highly realistic hybrid computational phantoms. We generated effective dose coefficients for seven body regions, stratified by patient age, diameter, and scanner manufacturer. We applied the new coefficients to DLPs to calculate effective doses and assessed their correlations with Monte Carlo radiation transport-generated effective doses. RESULTS: The reported effective dose coefficients, generally higher than previous studies, varied by body region and decreased in magnitude with increasing age. Coefficients were approximately 4 to 13-fold higher (across body regions) for patients <1 year old compared with patients 15-21 years old. For example, head CT (54% of scans) dose coefficients decreased from 0.039 to 0.003 mSv/mGy-cm in patients <1 year old vs. 15-21 years old. There were minimal differences by manufacturer. Using age-based conversion coefficients to estimate effective dose produced moderate to strong correlations with Monte Carlo results (Pearson correlations 0.52-0.80 across body regions). CONCLUSIONS: New pediatric effective dose coefficients update existing literature and can be used to easily estimate effective dose using scanner-reported DLP.

Sociodemographic factors associated with tracheostomy and mortality in bronchopulmonary dysplasia.

OBJECTIVES: We sought to investigate how race, ethnicity, and socioeconomic status relate to tracheostomy insertion and post-tracheostomy mortality among infants with bronchopulmonary dysplasia (BPD). METHODS: The Vizient Clinical Database/Resource Manager was queried to identify infants born ≤32 weeks with BPD admitted to US hospitals from January 2012 to December 2020. Markers of socioeconomic status were linked to patient records from the Agency for Healthcare Research and Quality's Social Determinants of Health Database. Regression models were used to assess trends in annual tracheostomy insertion rate and odds of tracheostomy insertion and post-tracheostomy mortality, adjusting for sociodemographic and clinical factors. RESULTS: There were 40,021 ex-premature infants included in the study, 1614 (4.0%) of whom received a tracheostomy. Tracheostomy insertion increased from 2012 to 2017 (3.1%-4.1%), but decreased from 2018 to 2020 (3.3%-1.6%). Non-Hispanic Black infants demonstrated a 25% higher odds (aOR 1.25, 1.09-1.43) and Hispanic infants demonstrated a 20% lower odds (aOR 0.80, 0.65-0.96) of tracheostomy insertion compared with non-Hispanic White infants. Patients receiving public insurance had increased odds of tracheostomy insertion (aOR 1.15, 1.03-1.30), but there was no relation between other metrics of socioeconomic status and tracheostomy insertion within our cohort. In-hospital mortality among the tracheostomy-dependent was 14.1% and was not associated with sociodemographic factors. CONCLUSIONS: Disparities in tracheostomy insertion are not accounted for by differences in socioeconomic status or the presence of additional neonatal morbidities. Post-tracheostomy mortality does not demonstrate the same relationships. Further investigation is needed to explore the source and potential mitigators of the identified disparities.

Association of Ventilator Settings With Mortality in Pediatric Patients Treated With Extracorporeal Life Support for Respiratory Failure.

Extracorporeal life support (ECLS) is a treatment for acute respiratory failure that can provide extracorporeal gas exchange, allowing lung rest. However, while most patients remain mechanically ventilated during ECLS, there is a paucity of evidence to guide the choice of ventilator settings. We studied the associations between ventilator settings 24 hours after ECLS initiation and mortality in pediatric patients using a retrospective analysis of data from the Extracorporeal Life Support Organization Registry. 3497 patients, 29 days to 18 years of age, treated with ECLS for respiratory failure between 2015 and 2021, were included for analysis. 93.3% of patients on ECLS were ventilated with conventional mechanical ventilation. Common settings included positive end-expiratory pressure (PEEP) of 10 cm H 2 O (45.7%), delta pressure (ΔP) of 10 cm H 2 O (28.3%), rate of 10-14 breaths per minute (55.9%), and fraction of inspired oxygen (FiO 2 ) of 0.31-0.4 (30.3%). In a multivariate model, PEEP >10 cm H 2 O ( versus PEEP < 8 cm H 2 O, odds ratio [OR]: 1.53, 95% CI: 1.20-1.96) and FiO 2 ≥0.45 ( versus FiO 2 < 0.4; 0.45 ≤ FiO 2 < 0.6, OR: 1.31, 95% CI: 1.03-1.67 and FiO 2 ≥ 0.6, OR: 2.30; 95% CI: 1.81-2.93) were associated with higher odds of mortality. In a secondary analysis of survivors, PEEP 8-10 cm H 2 O was associated with shorter ECLS run times ( versus PEEP < 8 cm H 2 O, coefficient: -1.64, 95% CI: -3.17 to -0.11), as was ΔP >16 cm H 2 O ( versus ΔP < 10 cm H 2 O, coefficient: -2.72, 95% CI: -4.30 to -1.15). Our results identified several categories of ventilator settings as associated with mortality or ECLS run-time. Further studies are necessary to understand whether these results represent a causal relationship.

Impact of Different Approaches to Preparing Notes for Analysis With Natural Language Processing on the Performance of Prediction Models in Intensive Care.

To evaluate whether different approaches in note text preparation (known as preprocessing) can impact machine learning model performance in the case of mortality prediction ICU. DESIGN: Clinical note text was used to build machine learning models for adults admitted to the ICU. Preprocessing strategies studied were none (raw text), cleaning text, stemming, term frequency-inverse document frequency vectorization, and creation of n-grams. Model performance was assessed by the area under the receiver operating characteristic curve. Models were trained and internally validated on University of California San Francisco data using 10-fold cross validation. These models were then externally validated on Beth Israel Deaconess Medical Center data. SETTING: ICUs at University of California San Francisco and Beth Israel Deaconess Medical Center. SUBJECTS: Ten thousand patients in the University of California San Francisco training and internal testing dataset and 27,058 patients in the external validation dataset, Beth Israel Deaconess Medical Center. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Mortality rate at Beth Israel Deaconess Medical Center and University of California San Francisco was 10.9% and 7.4%, respectively. Data are presented as area under the receiver operating characteristic curve (95% CI) for models validated at University of California San Francisco and area under the receiver operating characteristic curve for models validated at Beth Israel Deaconess Medical Center. Models built and trained on University of California San Francisco data for the prediction of inhospital mortality improved from the raw note text model (AUROC, 0.84; CI, 0.80-0.89) to the term frequency-inverse document frequency model (AUROC, 0.89; CI, 0.85-0.94). When applying the models developed at University of California San Francisco to Beth Israel Deaconess Medical Center data, there was a similar increase in model performance from raw note text (area under the receiver operating characteristic curve at Beth Israel Deaconess Medical Center: 0.72) to the term frequency-inverse document frequency model (area under the receiver operating characteristic curve at Beth Israel Deaconess Medical Center: 0.83). CONCLUSIONS: Differences in preprocessing strategies for note text impacted model discrimination. Completing a preprocessing pathway including cleaning, stemming, and term frequency-inverse document frequency vectorization resulted in the preprocessing strategy with the greatest improvement in model performance. Further study is needed, with particular emphasis on how to manage author implicit bias present in note text, before natural language processing algorithms are implemented in the clinical setting.

Variation in Arterial and Central Venous Catheter Use in Pediatric Intensive Care Units.

OBJECTIVE: Describe patient and hospital characteristics associated with Arterial Catheter (AC) or Central Venous Catheter (CVC) use among pediatric intensive care units (ICUs). DESIGN: Hierarchical mixed effects analyses were used to identify patient and hospital characteristics associated with AC or CVC placement. The ICU adjusted median odds ratios (ICU-AMOR) for the admission ICU, marginal R2, and conditional intraclass correlation coefficient were reported. SETTING: 166 PICUs in the Virtual PICU Systems (VPS, LLC) Database. PATIENTS: 682,791 patients with unscheduled admissions to the PICU. INTERVENTION: None. MEASURES AND MAIN RESULTS: ACs were placed in (median, [interquartile range]) 8.2% [4.9%-11.3%] of admissions, and CVCs were placed in 14.9% [10.4%-19.3%] of admissions across cohort ICUs. Measured patient characteristics explained about 25% of the variability in AC and CVC placement. Higher Pediatric Index of Mortality 2 (PIM2) illness severity scores were associated with increased odds of placement (Odds Ratio (95th% Confidence Interval)) AC: 1.88 (1.87-1.89) and CVC: 1.82 (1.81-1.83) per 1 unit increase in PIM2 score. Primary diagnoses of cardiovascular, gastrointestinal, hematology/oncology, infectious, renal/genitourinary, rheumatology, and transplant were associated with increased odds of AC or CVC placement compared to a primary respiratory diagnosis. Presence of in-house attendings 24/7 was associated with increased odds of AC placement 1.32 (1.11-1.57). Admission ICU explained 4.9% and 3.5% of the variability in AC or CVC placement, respectively. The ICU-AMOR showed a patient would have a median increase in odds of 55% and 43% for AC or CVC placement, respectively, if the same patient moved from an ICU with lower odds of placement to an ICU with higher odds of placement. CONCLUSIONS: Variation in AC or CVC use exists among PICUs. The admission ICU was more strongly associated with AC than with CVC placement. Further study is needed to understand unexplained variation in AC and CVC use.

Predicting NICU admissions in near-term and term infants with low illness acuity.

OBJECTIVE: Describe NICU admission rate variation among hospitals in infants with birthweight ≥2500 g and low illness acuity, and describe factors that predict NICU admission. STUDY DESIGN: Retrospective study from the Vizient Clinical Data Base/Resource Manager®. Support vector machine methodology was used to develop statistical models using (1) patient characteristics (2) only the indicator for the inborn hospital and (3) patient characteristics plus indicator for the inborn hospital. RESULTS: NICU admission rates of 427,449 infants from 154 hospitals ranged from 0 to 28.6%. C-statistics for the patient characteristics model: 0.64 (Confidence Interval (CI) 0.62-0.65), hospital only model: 0.81 (CI, 0.81-0.82), and patient characteristic plus hospital variable model: 0.84 (CI, 0.83-0.84). CONCLUSION/RELEVANCE: There is wide variation in NICU admission rates in infants with low acuity diagnoses. In all cohorts, birth hospital better predicted NICU admission than patient characteristics alone.