2.3 Statistical analyses
The annual incidence and prevalence of CRSwNP were assessed during 1 January 2012 – 31 December 2019 by dividing the number of incident and prevalent cases by the total Finnish population. Patients who fulfilled the criteria for CRSwNP during 1 January 2000 – 31 December 2011 were included in the prevalence assessment. Incident patients were defined as patients with a first record fulfilling the criteria for CRSwNP during 1 January 2012 – 31 December 2019. The presence of a monotonic time trend in prevalence and incidence was assessed using the Mann-Kendall test.
Patient demographics, clinical characteristics and disease burden were tabulated at baseline, overall and by subgroups, and described using summary statistics (mean with standard deviation, median with interquartile range, or frequency as the total number of patients and proportion (%) of all patients). The Charlson comorbidity index was calculated using ICD-10 codes as implemented in the comorbidity R package 21,22. Chi-squared test was used for assessing differences in systemic corticosteroid (SCS) use between the asthma subgroups during baseline.
Time to first and repeated ESS were assessed using Kaplan-Meier fits, where time from index to surgery was defined as an event and the end of follow-up or death as a censoring event. In the Kaplan-Meier fit of repeated ESS, only patients with at least one ESS were included. The Kaplan-Meier fits were visualized including the 95% Confidence Intervals (CIs), and median survival was reported, if reached. A mean cumulative count of ESS (with 95% CIs) was calculated over time overall and per subgroup using a Mean Cumulative Function (MCF)23.
All statistical analyses were run using R version 4.0.3 on RStudio Server version 1.4.1103 24, under Microsoft Windows Server 2016 Standard.