Risk estimation for biliary tract cancer: Development and validation of a prognostic score.

BACKGROUND & AIMS: Biliary tract cancer is a rare tumour entity characterized by a poor prognosis. We aimed to identify prognostic factors and create a prognostic score to estimate survival. METHODS: Clinical data of the training set, consisting of 569 patients treated from 2000 to 2010 at Hannover Medical School, were analysed. A prognostic model defining three prognostic risk groups was derived from Cox regression analyses. The score was applied and validated in an independent cohort of 557 patients from four different German centres. RESULTS: Median overall survival (OS) was 14.5 months. If complete resection was performed, the patients had a significantly improved OS (23.9 months; n=242) as compared to patients with non-resectable tumours (9.1 months; n=329, P<.0001). Based on univariable and multivariable analyses of clinical data, a prognostic model was created using variables available before treatment. Those were age, metastasis, C-reactive protein (CRP), international normalized ratio (INR) and bilirubin. The prognostic score distinguished three groups with a median OS of 21.8, 8.6 and 2.6 months respectively. The validation cohort had a median OS of 20.2, 14.0 and 6.5 months respectively. The prognostic impact of the score was independent of the tumour site and of treatment procedures. CONCLUSIONS: Here, we identified prognostic factors and propose a prognostic score to estimate survival, which can be applied to all patients independent of tumour site and before initial treatment. Further validation in prospective trials is required.

A modified Wald interval for the area under the ROC curve (AUC) in diagnostic case-control studies.

BACKGROUND: The area under the receiver operating characteristic (ROC) curve, referred to as the AUC, is an appropriate measure for describing the overall accuracy of a diagnostic test or a biomarker in early phase trials without having to choose a threshold. There are many approaches for estimating the confidence interval for the AUC. However, all are relatively complicated to implement. Furthermore, many approaches perform poorly for large AUC values or small sample sizes. METHODS: The AUC is actually a probability. So we propose a modified Wald interval for a single proportion, which can be calculated on a pocket calculator. We performed a simulation study to compare this modified Wald interval (without and with continuity correction) with other intervals regarding coverage probability and statistical power. RESULTS: The main result is that the proposed modified Wald intervals maintain and exploit the type I error much better than the intervals of Agresti-Coull, Wilson, and Clopper-Pearson. The interval suggested by Bamber, the Mann-Whitney interval without transformation and also the interval of the binormal AUC are very liberal. For small sample sizes the Wald interval with continuity has a comparable coverage probability as the LT interval and higher power. For large sample sizes the results of the LT interval and of the Wald interval without continuity correction are comparable. CONCLUSIONS: If individual patient data is not available, but only the estimated AUC and the total sample size, the modified Wald intervals can be recommended as confidence intervals for the AUC. For small sample sizes the continuity correction should be used.

Diagnostic value of carotid intima-media thickness and clinical risk scores in determining etiology of ischemic stroke.

BACKGROUND: In the general population, carotid intima-media thickness (CIMT) is associated with atherosclerosis as well as atrial fibrillation (AF). However, the extent to which CIMT might be of diagnostic value in clarifying stroke etiology is currently unclear. METHODS: In this retrospective cohort study, we included 800 consecutive patients with acute ischemic stroke. We compared CIMT-values between stroke etiologies. The association between CIMT and cardioembolic stroke was investigated via logistic regression analysis adjusting for vascular risk factors. Receiver operating characteristic analyses were conducted to investigate the diagnostic value of CIMT in comparison to vascular risk factors and clinical AF risk scores (CHA2DS2VASc, HAVOC, and AS5F). RESULTS: CIMT-values were highest in patients with cardioembolic or atherosclerotic stroke origin. CIMT was associated with newly diagnosed AF compared against cryptogenic strokes (crude odds ratio (OR) per 0.1 mm-increase of CIMT: 1.26 (95% confidence interval (CI): 1.13-1.41)). After adjustment for vascular risk factors, the effect of CIMT on AF-diagnosis, however, was weakened (adjusted OR: 1.10 (95% CI: 0.97-1.25)). The diagnostic value of CIMT for detection of AF (AUC: 0.60, 95% CI: 0.54-0.65) was outperformed by AF risk scores. Among the scores investigated, the AS5F-score yielded best accuracy and calibration to predict newly diagnosed AF (AUC: 0.71, 95% CI: 0.65-0.78). CONCLUSIONS: CIMT may help in the diagnosis of stroke etiology. However, compared with vascular risk factors or clinical AF risk scores, CIMT does not provide substantial additional information on the risk of newly detected AF. Thus, stratification of AF risk based on scores, such as the AS5F, is advisable.

Acute gastrointestinal bleeding cases presenting to the emergency department are associated with age, sex and seasonal and circadian factors.

BACKGROUND: Gastrointestinal bleeding (GIB) is one of the most common emergencies in gastroenterology. The aim of this study was to investigate the association between the incidence of GIB and seasonal, circadian and meteorological patterns in the emergency department (ED) of a tertiary hospital. PATIENTS AND METHODS: From January 2007 until December 2012, we retrospectively evaluated patients presenting to the ED with respect to the number and location of GIB, season, time of day and weather. RESULTS: Of 45 458 patients, 578 (1.3%) presented with a GIB. Of these, 62.5% were men compared with 54.7% of all patients in the ED (χ, P=0.0002). Patients with GIB were on average 4.4 years older than those without GIB (95% confidence interval 2.76-5.98, t-test, P<0.001). In addition, 304 (52.6%) patients had upper GIB and 138 (23.9%) had lower GIB. In total, 136 (23.5%) patients required no endoscopy because of initial laboratory and circulatory stability. In univariate analysis, meteorological parameters, including air temperature, cloud cover, relative humidity, vapour pressure, amount of precipitation, sunshine duration and snow height, were each associated with an increased risk of acute GIB (all P-values<0.05). In the 6-year study period, patients with GIB presented to the ED mainly during the winter months. Independent predictors of GIB on multivariate logistic regression were older age, male sex, season and daytime, all P less than 0.005. Emergency admissions during the night were associated with a 54 and 35% higher risk of GIB compared with daytime (8 a.m. to 4 p.m., P=0.0002) and late evening hours (4 p.m. to midnight, P=0.0142), respectively. CONCLUSION: Presentation of patients with acute GIB in the ED is age and sex specific and shows seasonal and circadian differences in distribution, with an increased incidence in winter months and during night-time. This should be considered when determining possible emergency endoscopic interventions and the availability of emergency endoscopy services.