# Table 3 Distributional cost composition analysis

Deciles HCE 4-class combination (main analysis) 3- & 4-class combination (sensitivity analysis)
Full population (n = 1840)
1 55 − 303 [− 659; 53] −220 [− 732; 292]
2 1837 −90 [− 473; 293] 107 [− 397; 611]
3 2898 162 [− 345; 669] 379 [− 229; 988]
4 4386 510 [−164; 1184] 693 [− 106; 1493]
5 6779 995 [33; 1958] 982 [− 98; 2062]
6 10,062 1730 [268; 3192] 1304 [− 194; 2801]
7 15,135 2504 [357; 4652] 1475 [− 674; 3624]
8 23,369 3516 [182; 6850] 2291 [− 1341; 5923]
9 42,046 4984 [− 1360; 11,328] 6987 [− 823; 14,798]
Not pre-exposed (n = 542)
1 629 − 340 [− 1074; 395] − 1082 [− 5818; 3654]
2 1400 − 332 [− 836; 172] − 479 [− 1779; 821]
3 1979 −315 [− 933; 303] −72 [− 1250; 1105]
4 2544 − 132 [− 1007; 743] 328 [− 1258; 1914]
5 3455 202 [− 1124; 1529] 857 [− 1388; 3101]
6 4879 587 [− 1548; 2722] 1277 [− 2013; 4567]
7 7949 1200 [− 1968; 4367] 1716 [− 3359; 6792]
8 12,058 2087 [− 2436; 6610] 3881 [− 4434; 12,196]
9 21,928 4317 [− 2969; 11,603] 4233 [− 21,459; 29,924]
1. This table illustrates results from the counterfactual distribution analysis (total costs and costs attributed to compliance). Numbers in [square brackets] represent bootstrap-based 95% confidence intervals. HCE amounts represent Swiss francs (CHF). The decomposition analysis took the following potential confounders into account age, sex, living in a French-speaking or Italian-speaking canton, degree of urbanity of place of living, having a high deductible, participating in a managed care model, having at least one supplementary insurance, having had high medication expenditures of at least CHF 5′000 within 360 days before the index date, having had an inpatient hospital stay within 360 days before the index date, number of pharmaceutical cost groups (which are drug-prescription based indicators for co-morbidities). Deciles represent the 9 points in the HCE distribution that split the full sample into 10 equally large parts
2. Positive values indicate lower health care expenditures (HCE) in compliers, and vice versa