Analysis of seasonal variation of antibiotic prescribing for respiratory tract diagnoses in primary care practices.

Objective: To determine antibiotic prescribing appropriateness for respiratory tract diagnoses (RTD) by season. Design: Retrospective cohort study. Setting: Primary care practices in a university health system. Patients: Patients who were seen at an office visit with diagnostic code for RTD. Methods: Office visits for the entire cohort were categorized based on ICD-10 codes by the likelihood that an antibiotic was indicated (tier 1: always indicated; tier 2: sometimes indicated; tier 3: rarely indicated). Medical records were reviewed for 1,200 randomly selected office visits to determine appropriateness. Based on this reference standard, metrics and prescriber characteristics associated with inappropriate antibiotic prescribing were determined. Characteristics of antibiotic prescribing were compared between winter and summer months. Results: A significantly greater proportion of RTD visits had an antibiotic prescribed in winter [20,558/51,090 (40.2%)] compared to summer months [11,728/38,537 (30.4%)][standardized difference (SD) = 0.21]. A significantly greater proportion of winter compared to summer visits was associated with tier 2 RTDs (29.4% vs 23.4%, SD = 0.14), but less tier 3 RTDs (68.4% vs 74.4%, SD = 0.13). A greater proportion of visits in winter compared to summer months had an antibiotic prescribed for tier 2 RTDs (80.2% vs 74.2%, SD = 0.14) and tier 3 RTDs (22.9% vs 16.2%, SD = 0.17). The proportion of inappropriate antibiotic prescribing was higher in winter compared to summer months (72.4% vs 62.0%, P < .01). Conclusions: Increases in antibiotic prescribing for RTD visits from summer to winter were likely driven by shifts in diagnoses as well as increases in prescribing for certain diagnoses. At least some of this increased prescribing was inappropriate.

Impact of Interobserver Variability in Manual Segmentation of Non-Small Cell Lung Cancer (NSCLC) Applying Low-Rank Radiomic Representation on Computed Tomography.

This study tackles interobserver variability with respect to specialty training in manual segmentation of non-small cell lung cancer (NSCLC). Four readers included for segmentation are: a data scientist (BY), a medical student (LS), a radiology trainee (MH), and a specialty-trained radiologist (SK) for a total of 293 patients from two publicly available databases. Sørensen-Dice (SD) coefficients and low rank Pearson correlation coefficients (CC) of 429 radiomics were calculated to assess interobserver variability. Cox proportional hazard (CPH) models and Kaplan-Meier (KM) curves of overall survival (OS) prediction for each dataset were also generated. SD and CC for segmentations demonstrated high similarities, yielding, SD: 0.79 and CC: 0.92 (BY-SK), SD: 0.81 and CC: 0.83 (LS-SK), and SD: 0.84 and CC: 0.91 (MH-SK) in average for both databases, respectively. OS through the maximal CPH model for the two datasets yielded c-statistics of 0.7 (95% CI) and 0.69 (95% CI), while adding radiomic and clinical variables (sex, stage/morphological status, and histology) together. KM curves also showed significant discrimination between high- and low-risk patients (p-value < 0.005). This supports that readers' level of training and clinical experience may not significantly influence the ability to extract accurate radiomic features for NSCLC on CT. This potentially allows flexibility in the training required to produce robust prognostic imaging biomarkers for potential clinical translation.