Descriptive statistics for covariates included in the prediction model and outcome variables, excluding individuals with missing data in covariates or the outcomes, by year
| 2015 (n=5929) | 2016 (n=6075) | 2017 (n=5963) | 2018 (n=6407) | ||
| Covariates | |||||
| Age | Median (IQR) | 25.1 (19.0–53.5) | 25.8 (19.0–54.1) | 26.0 (19.3–55.1) | 26.4 (19.5–56.0) |
| Sex | Male, n (%) | 3263 (53.6%) | 3369 (53.7%) | 3442 (53.9%) | 3435 (53.6%) |
| Female, n (%) | 2829 (46.4%) | 2903 (46.3%) | 2938 (46.1%) | 2972 (46.4%) | |
| FEV1% | Median (IQR) | 67.1 (47.6–84.7) | 68.2 (48.0–84.8) | 68.1 (47.8–85.4) | 68.6 (47.7–85.3) |
| FEV1% previous year | Median (IQR) | 69.0 (49.5–85.7) | 69.5 (50.3–85.9) | 70.0 (50.5–86.0) | 70.2 (50.2–86.3) |
| Body mass index | Median (IQR) | 21.7 (19.5–24.2) | 21.8 (19.5–24.3) | 21.8 (19.6–24.5) | 21.8 (19.6–24.4) |
| Pseudomonas aeruginosa | No, n (%) | 2446 (40.2%) | 2817 (44.9%) | 2930 (45.9%) | 2916 (45.5%) |
| Yes, n (%) | 3646 (59.8%) | 3455 (55.1%) | 3450 (54.1%) | 3491 (54.5%) | |
| Staphylococcus aureus | No, n (%) | 3567 (58.6%) | 3976 (63.4%) | 3937 (61.7%) | 3893 (60.8%) |
| Yes, n (%) | 2525 (41.4%) | 2296 (36.6%) | 2443 (38.3%) | 2514 (39.2%) | |
| Burkholderia cepacia | No, n (%) | 5790 (95.0%) | 5956 (95.0%) | 6083 (95.3%) | 6101 (95.2%) |
| Yes, n (%) | 302 (5.0%) | 316 (5.0%) | 297 (4.7%) | 306 (4.8%) | |
| CF-related diabetes | No, n (%) | 3764 (61.8%) | 3877 (61.8%) | 3941 (61.8%) | 4085 (63.8%) |
| Yes, n (%) | 2328 (38.2%) | 2395 (38.2%) | 2439 (38.2%) | 2322 (36.2%) | |
| Genotype | F508del homozygous | 3026 (49.7%) | 3079 (49.1%) | 3102 (48.6%) | 3131 (48.9%) |
| F508+ minimal | 1095 (18.0%) | 1118 (17.8%) | 1112 (17.4%) | 1115 (17.4%) | |
| F508+ residual | 307 (5.0%) | 321 (5.1%) | 337 (5.3%) | 350 (5.5%) | |
| F508+ other/unknown | 523 (8.6%) | 549 (8.8%) | 564 (8.8%) | 559 (8.7%) | |
| Any gating mutation or R117H | 686 (11.3%) | 721 (11.5%) | 765 (12.0%) | 762 (11.9%) | |
| Other/unknown | 455 (7.5%) | 484 (7.7%) | 500 (7.8%) | 490 (7.6%) | |
| Hospital-IV-days, past year* | Zero, n (%) | 3029 (49.7%) | 3830 (61.1%) | 3801 (59.6%) | 3755 (58.6%) |
| Median of non-zeros (IQR) | 25.0 (14.0,42.0) | 15.0 (9.0,32.0) | 14.0 (9.0,31.0) | 15.0 (9.0,32.0) | |
| Home-IV-days, past year* | Zero, n (%) | 3025 (49.7%) | 4400 (70.2%) | 4500 (70.5%) | 4489 (70.1%) |
| Median of non-zeros (IQR) | 26.0 (14.0–43.0) | 18.0 (13.0–33.0) | 20.0 (13.0–34.0) | 19.0 (13.0–34.0) | |
| Outcome variables | |||||
| Hospital-IV-days,*† | Zero, n (%) | 3564 (60.1%) | 3571 (58.8%) | 3407 (57.1%) | – |
| Median of non-zeros (IQR) | 15.0 (9.0–33.0) | 15.0 (9.0–33.0) | 16.0 (10.0–33.0) | – | |
| Home-IV-days,*† | Zero, n (%) | 4079 (68.8%) | 4218 (69.4%) | 4100 (68.8%) | – |
| Median of non-zeros (IQR) | 18.0 (13.0–34.0) | 20.0 (13.0–34.0) | 20.0 (13.0–35.0) | – |
*When the start and end dates for a given episode were the same, the number of days was counted as 1; otherwise, the number of days for that episode was counted as the difference between the start and end dates.
†These are the counts of hospital-IV-days and home-IV-days in the year following the annual review visit in 2015, 2016 and 2017. By contrast, ‘Hospital-IV-days, past year’ and ‘Home-IV-days, past year’ are the counts in the year leading up to the annual review visit.