Background: New onset fever is a common symptom among hospitalized patients and it may be a manifestation of fatal illnesses such as infection. However, its epidemiology and predictors for mortality have not been fully determined in a Japanese teaching hospital.
Methods: We investigated adult patients with new onset elevated temperature of 37.5 degrees Celsius or greater the 3^rd day after admission during a 4-month study period. Only the first, single episode per patient was analyzed. We determined the causes of fever among these patients with new onset fever. We also analyzed predictors for in-hospital mortality among these patients. These predictors were based on multivariable adjusted logistic regression using demographics, vital signs at the time of fever onset, baseline diseases, and basic laboratory data.
Results: From a total of 2,271 admitted patients, 126 patients (5.6%) developed fever. Among these febrile patients, 98 (78%) had infectious diseases with a prevalence of 4.3% in all admitted patients. The most common cause of infection among those patients was respiratory tract infection, followed by urinary tract infection. Causes for non-infectious fever included neoplastic diseases, inflammatory diseases, and drug fever. In-hospital mortality was associated with lower mean blood pressure <60 mmHg with odds ratio (OR) of 12.7 (95% CI, 1.3–121), tachycardia >90/min with OR 4.1 (95% CI, 1.2–13.5), tachypnea >20/min with OR 10.0 (95% CI, 2.8–35.2), and neoplastic disease with OR 4.1 (95% CI, 1.3–13.1). Infection as a cause of fever was not associated with mortality.
Conclusion: The majority of inpatients with new onset fever had infectious diseases, however fever was also caused by neoplastic diseases, inflammatory diseases and drug fever. Abnormality of vital signs and neoplastic disease were related to in-hospital mortality.

Purpose: To investigate the clinical significance of adaptive radiotherapy (ART) in locally advanced nasopharyngeal carcinoma treated with intensity-modulated radiotherapy (IMRT). Materials and Methods: Eligible patients were treated with concurrent chemoradiotherapy using IMRT. Planning computed tomography in ART was performed during radiotherapy, and replanning was performed. Since ART was started in May 2011 (ART group), patients who were treated without ART up to April 2011 (non-ART group) were used as the historical control. The Kaplan-Meier method was used to calculate overall survival (OS), locoregional recurrence-free survival (LRFS), progression-free survival (PFS), and distant metastasis-free survival (DMFS). LRFS for the primary tumor (LRFS_P) and regional lymph node (LRFS_LN) were also studied for more detailed analysis. Statistical significance was evaluated using the log-rank test for survival. Results: The ART group tended to have higher radiation doses. The median follow-up period was 127 months (range, 10 to 211 months) in the non-ART group and 61.5 months (range, 5 to 129 months) in the ART group. Compared to the non-ART group, the ART group showed significantly higher 5-year PFS (53.8% vs. 81.3%, p = 0.015) and LRFS (61.2% vs. 85.3%, p = 0.024), but not OS (80.7% vs. 80.8%, p = 0.941) and DMFS (84.6% vs. 92.7%, p = 0.255). Five-year LRFS_P was higher in the ART group (61.3% vs. 90.6%, p = 0.005), but LRFS_LN did not show a significant difference (91.9% vs. 96.2%, p = 0.541). Conclusion Although there were differences in the patient backgrounds between the two groups, this study suggests the potential effectiveness of ART in improving locoregional control, especially in the primary tumor.