Diagnosis of Chronic Pancreatitis Incorporating Endosonographic Features, Demographics, and Behavioral Risk.

OBJECTIVES: Diagnosing chronic pancreatitis remains challenging. Endoscopic ultrasound (EUS) is utilized to evaluate pancreatic disease. Abnormal pancreas function test is considered the "nonhistologic" criterion standard for chronic pancreatitis. We derived a prediction model for abnormal endoscopic pancreatic function test (ePFT) by enriching EUS findings with patient demographic and pancreatitis behavioral risk characteristics. METHODS: Demographics, behavioral risk characteristics, EUS findings, and peak bicarbonate results were collected from patients evaluated for pancreatic disease. Abnormal ePFT was defined as peak bicarbonate of less than 75 mEq/L. We fit a logistic regression model and converted it to a risk score system. The risk score was validated using 1000 bootstrap simulations. RESULTS: A total of 176 patients were included; 61% were female with median age of 48 years (interquartile range, 38-57 years). Abnormal ePFT rate was 39.2% (69/176). Four variables formulated the risk score: alcohol or smoking status, number of parenchymal abnormalities, number of ductal abnormalities, and calcifications. Abnormal ePFT occurred in 10.7% with scores 4 or less versus 92.0% scoring 20 or greater. The model C-statistic was 0.78 (95% confidence interval, 0.71-0.85). CONCLUSIONS: Number of EUS pancreatic duct and parenchymal abnormalities, presence of calcification, and smoking/alcohol status were predictive of abnormal ePFT. This simple model has good discrimination for ePFT results.

Development and validation of a disease-specific risk adjustment system using automated clinical data.

OBJECTIVE: To develop and validate a disease-specific automated inpatient mortality risk adjustment system primarily using computerized numerical laboratory data and supplementing them with administrative data. To assess the values of additional manually abstracted data. METHODS: Using 1,271,663 discharges in 2000-2001, we derived 39 disease-specific automated clinical models with demographics, laboratory findings on admission, ICD-9 principal diagnosis subgroups, and secondary diagnosis-based chronic conditions. We then added manually abstracted clinical data to the automated clinical models (manual clinical models). We compared model discrimination, calibration, and relative contribution of each group of variables. We validated these 39 models using 1,178,561 discharges in 2004-2005. RESULTS: The overall mortality was 4.6 percent (n = 58,300) and 4.0 percent (n = 47,279) for derivation and validation cohorts, respectively. Common mortality predictors included age, albumin, blood urea nitrogen or creatinine, arterial pH, white blood counts, glucose, sodium, hemoglobin, and metastatic cancer. The average c-statistic for the automated clinical models was 0.83. Adding manually abstracted variables increased the average c-statistic to 0.85 with better calibration. Laboratory results displayed the highest relative contribution in predicting mortality. CONCLUSIONS: A small number of numerical laboratory results and administrative data provided excellent risk adjustment for inpatient mortality for a wide range of clinical conditions.

Candidemia on presentation to the hospital: development and validation of a risk score.

INTRODUCTION: Candidemia results in substantial morbidity and mortality, especially if initial antifungal therapy is delayed or is inappropriate; however, candidemia is difficult to diagnose because of its nonspecific presentation. METHODS: To develop a risk score for identifying hospitalized patients with candidemia, we performed a retrospective analysis of a large database of 176 acute-care hospitals in the United States. We studied 64,019 patients with bloodstream infection (BSI) on presentation from 2000 through 2005 (derivation cohort) and 24,685 from 2006 to 2007 (validation cohort). We used recursive partitioning (RPART) to identify the best discriminators for Candida as the cause of BSI. We compared three sets of models (equal-weight, unequal-weight, vs full model with additional variables from logistic regression model) for sensitivity analysis. RESULTS: The RPART identified 6 variables as the best discriminators: age < 65 years, temperature 0.10, indicating predicted and observed candidemia rates did not differ significant across the 7 risk stratus). The full model with 16 risk factors had slightly higher AUROCs (0.74 versus 0.73 for derivation versus validation); however, 7 variables were no longer significant in the recalibrated model for the validation cohort, indicating that the additional items did not materially enhance the model. CONCLUSIONS: A simple equal-weight risk score differentiated patients' risk for candidemia in a graded fashion upon hospital presentation.

Physician reminders to promote surveillance colonoscopy for colorectal adenomas: a randomized controlled trial.

BACKGROUND: Most colorectal cancers develop from adenomatous polyps. National guidelines recommend surveillance colonoscopy within 5 years after such polyps are removed. OBJECTIVE: To determine whether surveillance colonoscopy can be increased among overdue patients by reminders to their primary physicians. DESIGN: Randomized, controlled trial of patient-specific reminders mailed to 141 physicians in 2 Massachusetts primary care networks during April, 2006. PATIENTS: Seven hundred seventeen patients who had colorectal adenomas removed during 1995 through 2000 and no follow-up colonoscopy identified via automated review of electronic records through March, 2006. MEASUREMENTS AND MAIN RESULTS: The use of colonoscopy and detection of new adenomas or cancer were assessed at 6 months by a blinded medical record review in all patients. Among 358 patients whose physicians received reminders, 33 (9.2%) patients underwent colonoscopy within 6 months, compared with 16 (4.5%) of 359 patients whose physicians did not receive reminders (P = 0.009). In prespecified subgroups, this effect did not differ statistically between 2 primary care networks, elderly and nonelderly patients, or women and men (all P > 0.60 by Breslow-Day test). New adenomas or cancer were detected in 14 (3.9%) intervention patients and 6 (1.7%) control patients (P = 0.06), representing 42.4% and 37.5% of patients who underwent colonoscopy in each group, respectively. Despite using advanced electronic health records to identify eligible patients, 22.5% of enrolled patients had a prior follow-up colonoscopy ascertained only by visual record review, and physicians reported 27.9% of intervention patients were no longer active in their practice. CONCLUSIONS: Among patients with prior colorectal adenomas, physician reminders increased the use of surveillance colonoscopy, but better systems are needed to identify eligible patients (ClinicalTrials.gov ID number NCT00397969).