PubMed Health⌕ Search

Biomedical subjects

James R Carpenter

Publications and source records attributed to James R Carpenter.

5 recordsLinked to original sources
PubMed2006

Using SAS to conduct nonparametric residual bootstrap multilevel modeling with a small number of groups.

In multilevel modeling, researchers often encounter data with a relatively small number of units at the higher levels. As a result, of this and/or non-normality of the residuals, model parameter estimates, particularly the variance components and standard errors of parameter estimates at the group level, may be biased, thus the corresponding statistical inferences may not be trustworthy. This problem can be addressed by using bootstrap methods to estimate the standard errors of the parameter estimates for significance testing. This study illustrates how to use statistical analysis system (SAS) to conduct nonparametric residual bootstrap multilevel modeling. Specific SAS programs for such modeling are provided.

Models, Statistical↗
PubMed2006

Recalibration of risk prediction models in a large multicenter cohort of admissions to adult, general critical care units in the United Kingdom.

OBJECTIVE: To assess the performance of published risk prediction models in common use in adult critical care in the United Kingdom and to recalibrate these models in a large representative database of critical care admissions. DESIGN: Prospective cohort study. SETTING: A total of 163 adult general critical care units in England, Wales, and Northern Ireland, during the period of December 1995 to August 2003. PATIENTS: A total of 231,930 admissions, of which 141,106 met inclusion criteria and had sufficient data recorded for all risk prediction models. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The published versions of the Acute Physiology and Chronic Health Evaluation (APACHE) II, APACHE II UK, APACHE III, Simplified Acute Physiology Score (SAPS) II, and Mortality Probability Models (MPM) II were evaluated for discrimination and calibration by means of a combination of appropriate statistical measures recommended by an expert steering committee. All models showed good discrimination (the c index varied from 0.803 to 0.832) but imperfect calibration. Recalibration of the models, which was performed by both the Cox method and re-estimating coefficients, led to improved discrimination and calibration, although all models still showed significant departures from perfect calibration. CONCLUSIONS: Risk prediction models developed in another country require validation and recalibration before being used to provide risk-adjusted outcomes within a new country setting. Periodic reassessment is beneficial to ensure calibration is maintained.

APACHE↗
PubMed2005

Survival following the development of ascites and/or peripheral oedema in primary biliary cirrhosis: a staged prognostic model.

OBJECTIVE: Current prognostic models in primary biliary cirrhosis (PBC) have low precision, partly due to the restricted inclusion criteria of some cohorts used for modelling but also because of the prolonged natural course of the disease. It is hypothesized that better precision could be achieved with a staged model, using ascites or peripheral oedema as a new starting-point for prediction. MATERIAL AND METHODS: The study was based on an established database of 289 consecutive patients, followed between 1977 and 1998. Stepwise Cox regression was used to construct a staged model based on 143 patients who first developed ascites (n=111) or peripheral oedema (n=32) at entry or during subsequent follow-up. The model was compared with published models using graphical methods and receiver operating characteristics (ROCs). RESULTS: Mean time from clinical diagnosis of ascites or peripheral oedema to death was 3.1 years. The following variables had independent prognostic significance: log10(bilirubin) (p<0.001), albumin (p<0.001), age (p<0.001) and history of encephalopathy (p<0.001). Goodness of fit showed that the survival probabilities predicted by the Ascites Stage Model fitted well with the observed data. The Ascites Stage Model (ROC 0.8324 (SE 0.0348)) was a better predictor of survival than the Mayo long-term model (ROC 0.7833 (SE 0.0397)), the Mayo Repeated Patient Visits Model (ROC 0.7779 (SE 0.0399)) and the Royal Free PBC Prognostic Model (ROC 0.7785 (SE 0.0396)). CONCLUSIONS: The Ascites Stage Model gives a better survival estimate for PBC patients once they have developed ascites or peripheral oedema compared with the current models, and demonstrates an advantage of staged models in diseases with a prolonged natural history.

Ascites↗
PubMed2004

Winter excess mortality in intensive care in the UK: an analysis of outcome adjusted for patient case mix and unit workload.

OBJECTIVE: To investigate whether mortality in UK intensive care units is higher in winter than in non-winter and to explore the importance of variations in case mix and increased pressure on ICUs. DESIGN AND SETTING: Cohort study in 115 adult, general ICUs in England, Wales and Northern Ireland. PATIENTS AND PARTICIPANTS: 113,389 admissions from 1995 to 2000. MEASUREMENTS AND RESULTS: Hospital mortality following admission to ICU was compared between winter (December-February) and non-winter (March-November). The causes of any observed differences were explored by adjusting for the case mix of admissions and the workload of the ICUs. Crude hospital mortality was higher in winter. After adjusting for case mix using the APACHE II mortality probability this effect was reduced but still significant. When additional factors reflecting case mix and workload were introduced into the model, the overall effect of winter admission was no longer significant. Factors reflecting both the case mix of the individual patient and of the patients in surrounding beds were found to be significantly associated with outcome. After adjustment for other factors, the occupancy of the unit (proportion of beds occupied) was not significantly associated with mortality. CONCLUSIONS: The excess winter mortality observed in UK ICUs can be explained by variation in the case mix of admissions. Unit occupancy was not associated with mortality.

APACHE↗