Stata mp version 16

All statistical analyses were tabulated and analyzed by Stata/MP version 16. The Chi-square test was used to evaluate the association between categorical variables and the Mann–Whitney U test was applied to compare the continuous variables. The Kaplan–Meier method with log-rank test and Cox proportional hazard models were utilized to compare progression–free survival duration among the patient groups. The logistic regression models were used to compare PSA and radiological response among the patient groups. All the above models were based on the patients with available information on outcomes, such as progression-free survival duration (N = 192), PSA response (N = 146), and radiological response (N = 131). The factors with statistical significance (p < 0.1 in univariable analysis) were selected for the multivariable model. All tests were two-sided with p < 0.05 considered statistically significant. The missing value would not be included in the Cox proportional analysis.

Statistical analysis was performed from September 1, 2022, through April 10, 2023. Self-reported confirmed and convincing FA prevalence estimates were calculated using complex survey weighted proportions.^{1 (link),2 (link)} Pearson χ² statistics were calculated to test the independence of key study variables. Covariate-adjusted, complex survey-weighted logistic regression models compared relative prevalence and other convincing FA outcomes by participant characteristics, including interaction terms to assess moderation effects by demographic information. Two-sided hypothesis tests were used, and conventional thresholds of P < .05 denoted statistical significance. Stata MP, version 16 (StataCorp LLC), was used for all analyses.

Microbiological characteristics were assessed both on an isolate-level and a collapsed wastewater sample-level. Respective differences in distributions were calculated using Chi-squared, Fisher’s exact, and Wilcoxon rank sum/Kruskal-Wallis tests—as appropriate. Missing data is indicated throughout. Correlation analyses of sample, socio-economic and meteorological parameters were performed using the Spearman’s rank correlation coefficients (rho) ranging from −1 to 1. Meteorological data were standardized across time and merged on a day-level for Basel overall. Socio-economic/population characteristics data were collapsed on a district-year level (medians) by weighting, where appropriate, for population size. Correlation strengths (rho) were defined as no correlation (0), very weak (0.01–0.19), weak (0.20–0.39), moderate (0.40–0.59), strong (0.60–0.79), and very strong (0.80–1.00). To visualize the strength of the relationships between the different variables, the matrix information were plotted using heatmaps (Stata packages “heatplot” (Jann, 2019 ), “palettes” and “colrspace”). Intracluster correlation was marginal overall and was therefore not considered in the explorative hypothesis tests. All analyses were performed on a multicore system with Stata/MP version 16 (Stata Corp., College Station, Texas, United States). All reported p-values are two-sided.

Since some patients might have undergone surgery with a previously unrecognized cancer, we did a supplementary analysis, by restriction of the study population to patients free from cancer at baseline. In this sub-cohort, we applied a time-dependent stratification on patients who did and did not develop cancer during follow-up. Furthermore, to clarify the short-term VTE risk after surgery, the VTE risk was also investigated within 7 days after surgery.
The following additional analyses were conducted: (1) VTE risk stratifying patients in age categories as suggested in the Caprini score [22 ] (< 41 years, 41–59 years, 60–74 years, > 74 years); and (2) VTE risks after surgery according to varying anatomic areas (surgeries related to mouth/throat, nose/sinuses, ear, and endoscopies, respectively).
Analyses were conducted using Stata/MP version 16 (StataCorp LP). This study was conducted in compliance with the General Data Protection Regulation and is part of North Denmark Region’s record of processing activities (j.no. 2017–68). Other approvals were not necessary according to Danish legislation.

The final developed model equation was applied to the validation dataset, and calibration and discrimination were examined as above.⁴⁴ (link)^,45 Calibration of 5-year risks was examined by plotting agreement between estimated risk from the model and observed outcome risks. Predicted and observed risks were divided into 10 equally sized groups. Additionally, pseudo-observations were used to construct smooth calibration curves across all individuals via a running non-parametric smoother. Separate calibration plots were provided for each imputation. Age-group, inflammatory disease type, and whether the patient was commenced on thiopurine after the year 2010, formed the basis of sub-group analyses. Stata-MP version 16 was used for all statistical analyses and data visualisation.
This study was reported in line with the transparent reporting of a multivariate prediction model for individual prediction or diagnosis (TRIPOD) guidelines.⁴⁶ (link)