Skip to main content

Table 2 Hazard ratios and time to next relapse by number of prior relapses (primary proxy and secondary proxy 1)

From: Predicting time to relapse in patients with schizophrenia according to patients’ relapse history: a historical cohort study using real-world data in Sweden

Number of prior relapses

Primary proxy

Secondary proxy 1

Patients, na

Median time to next relapse, years

HR (95% CI)b

Patients, na

Median time to next relapse, years

HR (95% CI)b

0

2994

1.52

-

2993

-

-

1

2024

1.23

1.84 (1.71–1.99)

779

6.47

2.76 (2.39–3.18)

2

1278

0.89

2.77 (2.53–3.03)

278

2.26

6.10 (5.00–7.43)

3

829

0.85

3.55 (3.19–3.94)

128

1.80

5.69 (4.32–7.48)

4

530

0.59

5.06 (4.48–5.73)

60

0.72

10.11 (7.03–14.54)

5

362

0.46

6.38 (5.53–7.37)

32

1.23

10.82 (6.77–17.31)

6

248

0.44

6.98 (5.89–8.27)

19

1.08

7.55 (3.72–15.32)

7

164

0.43

9.55 (7.79–11.70)

8

0.84

16.92 (6.95–41.22)

8

109

0.51

9.23 (7.21–11.81)

5

1.31

13.72 (4.37–43.11)

9

71

0.39

11.59 (8.69–15.47)

3

0.07

35.98 (11.34–114.20)

10

49

0.22

18.65 (15.42–22.56)

3

0.04

10.02 (2.48–40.48)

  1. aAfter each relapse there was a censoring time of seven days after each event. Since not all patients could be followed after the time-window (due to the end of the follow-up period), the number of patients at risk of the n'th relapse was slightly smaller than the number of patients with the n'th-1 prior relapses
  2. bPatients with no history of relapse (following the index hospitalisation) were used as the reference group in the Cox model. Primary proxy: identified a relapse episode based on a psychiatric hospitalisation ≥7 days. Secondary proxy 1: identified a relapse episode as a psychiatric hospital contact with >1 overnight stay, followed by a switch in AP treatment. Time to next relapse reflects the time by which 50% of patients would be estimated to have experienced their next relapse; this information was extracted from the Aalen-Johansen plot. Statistical significance was estimated by the 95% CI. If the 95% CI included 1, then the time to relapse was not statistically significantly different from the reference group. If the 95% CI excluded 1, then time to relapse was significantly different from the reference group. AP: antipsychotic; CI: confidence interval; HR: hazard ratio