External Validation and Recalibration of Risk Prediction Models for Acute Traumatic Brain Injury among Critically Ill Adult Patients in the United Kingdom

Harrison, David A; Griggs, Kathryn; Prabhu, Gita; Gomes, Manuel; Lecky, Fiona; Hutchinson, Peter J.; Menon, David; Rowan, Kathy

doi:10.1089/neu.2014.3628

Cited by 18 publications

(6 citation statements)

References 53 publications

(33 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…More extensive model revisions can be considered if the validation dataset is relatively large. Updated by model recalibration and revision, the new three models of 2016E1 showed adequate calibration and small improvements in discrimination, which was similar with previous studies focused on validation and recalibration ( Harrison et al, 2015 ; Wessler et al, 2017 ). Nevertheless, the discrimination ability was merely acceptable, leaving the possibility to develop new models de novo.…”

Section: Discussionsupporting

confidence: 85%

External validation of five predictive models for postoperative cardiopulmonary morbidity in a Chinese population receiving lung resection

Huang

Liu

Wang

et al. 2022

PeerJ

View full text Add to dashboard Cite

Background No postoperative cardiopulmonary morbidity models have been developed or validated in Chinese patients with lung resection. This study aims to externally validate five predictive models, including Eurolung models, the Brunelli model and the Age-adjusted Charlson Comorbidity Index, in a Chinese population. Methods Patients with lung cancer who underwent anatomic lung resection between 2018/09/01 and 2019/08/31 in our center were involved. Model discrimination was assessed by the area under the receiver operating characteristic curve. Model calibration was evaluated by the Hosmer–Lemeshow test. Calibration curves were plotted. Specificity, sensitivity, negative predictive value, positive predictive value and accuracy were calculated. Model updating was achieved by re-estimating the intercept and/or the slope of the linear predictor and re-estimating all coefficients. Results Among 1085 patients, 91 patients had postoperative cardiopulmonary complications defined by the European Society of Thoracic Surgeons. For original models, only parsimonious Eurolung1 had acceptable discrimination (area under the receiver operating characteristic curve = 0.688, 95% confidence interval 0.630–0.745) and calibration (p = 0.23 > 0.05) abilities simultaneously. Its sensitivity, specificity, positive predictive value, negative predictive value and accuracy were 0.700, 0.649, 0.153, 0.960 and 0.653, respectively. In the secondary analysis, increased pleural effusion (n = 94), which was nonchylous and nonpurulent, was labeled as a kind of postoperative complication. The area under the receiver operating characteristic curve of the models increased slightly, but all models were miscalibrated. The original Eurolung1 model had the highest discrimination ability but poor calibration, and thus it was updated by three methods. After model updating, new models showed good calibration and small improvements in discrimination. The discrimination ability was still merely acceptable. Conclusions Overall, none of the models performed well on postoperative cardiopulmonary morbidity prediction in this Chinese population. The original parsimonious Eurolung1 and the updated Eurolung1 were the best-performing models on morbidity prediction, but their discrimination ability only achieved an acceptable level. A multicenter study with more relevant variables and sophisticated statistical methods is warranted to develop new models among Chinese patients in the future.

show abstract

Section: Discussionsupporting

confidence: 85%

External validation of five predictive models for postoperative cardiopulmonary morbidity in a Chinese population receiving lung resection

Huang

Liu

Wang

et al. 2022

PeerJ

View full text Add to dashboard Cite

show abstract

“…Our current study showed that TBI (16.8%, 11.4/100,000 PY) was the second most common cause of ICU admission. TBI has a mortality rate of approximately 40.0% [ 18 ] and a disability rate of 57.4% [ 19 ]. Moreover, our data showed that the fatality rate within the first 7 days of admission was also high in cases of TBI (36.6%).…”

Section: Discussionmentioning

confidence: 99%

Acute spontaneous intracerebral hemorrhage and traumatic brain injury are the most common causes of critical illness in the ICU and have high early mortality

et al. 2018

View full text Add to dashboard Cite

BackgroundCritical care covers multiple disciplines. However, the causes of critical illness in the ICU, particularly the most common causes, remain unclear. We aimed to investigate the incidence and the most common causes of critical illness and the corresponding early mortality rates in ICU patients.MethodsA retrospective cohort study was performed to examine critically ill patients (aged over 15 years) in the general ICU in Shuyang County in northern China (1/2014–12/2015). The incidences and causes of critical illnesses and their corresponding early mortality rates in the ICU were determined by an expert panel.ResultsDuring the 2-year study period, 1,211,138 person-years (PY) and 1645 critically ill patients (mean age, 61.8 years) were documented. The median Glasgow Coma Scale (GCS) score was 6 (range, 3–15). The mean acute physiology and chronic health evaluation II (APACHE II) score was 21.2 ± 6.8. The median length of the ICU stay was 4 days (range, 1–29 days). The most common causes of critical illness in the ICU were spontaneous intracerebral hemorrhage (SICH) (26%, 17.6/100,000 PY) and traumatic brain injury (TBI) (16.8%, 11.4/100,000 PY). During the first 7 days in the ICU, SICH was the most common cause of death (42.2%, 7.4/10,000 PY), followed by TBI (36.6%, 4.2/100,000 PY). Based on a logistic analysis, older patients had a significantly higher risk of death from TBI (risk ratio [RR], 1.7; 95% CI, 1.034–2.635), heart failure/cardiovascular crisis (RR, 0.2; 95% CI, 0.083–0.484), cerebral infarction (RR, 0.15; 95% CI, 0.050–0.486), or respiratory failure (RR, 0.35; 95% CI, 0.185–0.784) than younger patients. However, the risk of death from SICH in the two groups was similar.ConclusionsThe most common causes of critical illness in the ICU were SICH and TBI, and both critical illnesses showed a higher risk of death during the first 7 days in the ICU.

show abstract

“…Like the CRASH prediction methodology, the IMPACT model is available online, thus allowing access for risk prediction in individual subjects (http://www.tbi-impact.org/). The IMPACT model has been validated in multiple settings, and shows good discrimination of favourable and unfavourable outcomes (Maas et al, 2015), but sometimes poor calibration in new patient populations (Harrison et al, 2013), which can be substantially improved by recalibration in these contexts (Harrison et al, 2015).…”

Section: Outcome Predictionmentioning

confidence: 99%

Critical care management of traumatic brain injury

Menon

Ercole

2017

Critical Care Neurology Part I

View full text Add to dashboard Cite

Traumatic brain injury (TBI) is a growing global problem, which is responsible for a substantial burden of disability and death, and which generates substantial healthcare costs. High-quality intensive care can save lives and improve the quality of outcome. TBI is extremely heterogeneous in terms of clinical presentation, pathophysiology, and outcome. Current approaches to the critical care management of TBI are not underpinned by high-quality evidence, and many of the current therapies in use have not shown benefit in randomized control trials. However, observational studies have informed the development of authoritative international guidelines, and the use of multimodality monitoring may facilitate rational approaches to optimizing acute physiology, allowing clinicians to optimize the balance between benefit and risk from these interventions in individual patients. Such approaches, along with the emerging impact of advanced neuroimaging, genomics, and protein biomarkers, could lead to the development of precision medicine approaches to the intensive care management of TBI.

show abstract

External Validation and Recalibration of Risk Prediction Models for Acute Traumatic Brain Injury among Critically Ill Adult Patients in the United Kingdom

Cited by 18 publications

References 53 publications

External validation of five predictive models for postoperative cardiopulmonary morbidity in a Chinese population receiving lung resection

External validation of five predictive models for postoperative cardiopulmonary morbidity in a Chinese population receiving lung resection

Acute spontaneous intracerebral hemorrhage and traumatic brain injury are the most common causes of critical illness in the ICU and have high early mortality

Critical care management of traumatic brain injury

Contact Info

Product

Resources

About