30 Chapter 2 Table 1. Estimated HR for ongoing pregnancy of oil versus water HSG at different time points after HSG Follow-up time HR (95%CI) for ongoing pregnancy using log time Immediately after HSG 1.71 (1.27-2.31) 1 month 1.57 (1.24-1.99) 3 months 1.36 (1.17-1.59) 6 months 1.25 (1.09-1.44) 9 months 1.19 (1.03-1.38) 1 year 1.15 (0.98-1.35) 2 years 1.06 (0.86-1.3) 3 years 1.00 (0.79-1.28) CI, confidence interval; HR, hazard ratio; HSG, hysterosalpingography The numbers of pregnancies for the first six cycles are shown in Table 2. The relative risks ranged from 1.16 to 1.83. The effect seemed to last during these six cycles but due to a small number of pregnancies per cycle, the estimates for this approach were uncertain, making it difficult to ascertain a clear trend. Table 2. Number of pregnancies per group, per cycle Cycle after HSG Ongoing pregnancies oil group Ongoing pregnancies water group Relative Risk (95% CI) 1 47/550 31/557 1.54 (0.99-2.38) 2 39/502 30/526 1.36 (0.86-2.16) 3 41/463 24/496 1.83 (1.12-2.98) 4 24/422 23/471 1.16 (0.67-2.03) 5 22/398 19/447 1.30 (0.71-2.37) 6 23/374 19/428 1.39 (0.77-2.50) CI, confidence interval; HSG, hysterosalpingography Sensitivity analyses When repeating the primary analysis only in women who were asked to judge their pain on a visual analogue scale (n = 401) and scored at least 6 points (n = 152: 73 oil and 79 water), the estimated hazard ratio over 3 years was 1.47 (95% CI 1.03–2.12). There was no evidence of a change in effect of oil versus water over time as the scaled Schoenfeld residuals plot showed a slight increase rather than a decrease, and the tests for regular and log-transformed time were not significant (Supplementary Figures 3 and 4, P = 0.88 and P = 0.71, respectively).
RkJQdWJsaXNoZXIy MTk4NDMw