Table 4.
Preoperative Predictors of Treatment Failure After Global Endometrial Ablation
| Unadjusted Analyses | Adjusted Analyses (Final Model)a,b | |||||
|---|---|---|---|---|---|---|
| Predictors | Univariable Hazard Ratio |
95% CI | P Value | Multivariable Hazard Ratio |
95% CI | P Value |
| Thermal balloon ablation vs radiofrequency ablation | 1.5 | 0.8–2.9 | .26c | 1.5 | 0.7–2.9 | .27 |
| Age <45 y | 2.4 | 1.2–4.7 | .01 | 2.6 | 1.3–5.1 | .008 |
| Parity ≥5 | 4.8 | 2.0–11.4 | <.001 | 6.0 | 2.5–14.8 | <.001 |
| Body mass index ≥30 kg/m2 | 0.6 | 0.3–1.3 | .20 | |||
| Previous cesarean delivery | 0.7 | 0.3–1.6 | .40c | |||
| Tubal ligation | 2.5 | 1.4–4.5 | .002c | 2.2 | 1.2–4.0 | .01 |
| Preoperative bleeding ≥7 d | 1.1 | 0.5–2.4 | .78 | |||
| Preoperative menstrual accidents or clots | 1.6 | 0.8–3.1 | .16 | |||
| Irregular uterine bleeding (metrorrhagia) | 1.5 | 0.8–2.7 | .18 | |||
| Preoperative dysmenorrhea | 3.9 | 1.7–8.7 | .001c | 3.7 | 1.6–8.5 | .003 |
| Uterine length ≥9 cm | 1.0 | 0.5–1.8 | .94c | |||
| Retroverted uterus | 1.0 | 0.3–3.3 | .97c | |||
| Hemoglobin ≥12 g/dL | 1.8 | 0.9–3.6 | .08 | |||
| Ultrasonogram suggestive of adenomyosis | 1.5 | 0.5–4.9 | .51 | |||
| Endometrial thickness ≥4 mm | 0.8 | 0.3–2.4 | .73c | |||
| Intracavitary lesions (hysteroscopy) | 0.8 | 0.4–1.7 | .58 | |||
| Uterine polyp | 0.6 | 0.3–1.4 | .22c | |||
| Submucous fibroid | 1.0 | 0.3–3.1 | .94c | |||
| Nonsubmucous fibroid | 1.1 | 0.6–2.3 | .75 | |||
Abbreviation: CI, confidence interval.
a
The c index of this model was 0.755.
b
Variables with P values less than .20 in the unadjusted (univariable) analysis were considered in the multivariable Cox proportional hazard regression model.
c
This variable previously was shown in the literature to be associated with treatment failure.