Table 2 Duration of wound in overall population and by scheduled treatment strategy.

Duration of wound	≤1 month	1–2 months	2–3 months	3–6 months	>6 months
Overall (n=428)	41.8% [36.9%–46.8%]	22.1% [17.9%–26.2%]	20.2% [16.0%–24.5%]	11.7% [8.5%–14.8%]	4.2% [2.2%–6.2%]
Surgical (n=142)	38.0% [28.9%–47.1%]	22.7% [15.4%–30.0%]	20.6% [12.9%–28.2%]	13.4% [7.3%–19.6%]	5.3% [1.3%–9.3%]
Endovascular (n=275)	44.3% [38.3%–50.3%]	21.0% [16.0%–26.0%]	19.9% [14.9%–24.8%]	11.1% [7.2%–14.9%]	3.7% [1.5%–6.0%]
Conservative (n=11)	36.4% [7.9%–64.8%]	27.3% [1.0%–53.6%]	18.2% [0.0%–41.0%]	9.1% [0.0%–26.1%]	9.1% [0.0%–26.1%]

Data are the estimated proportions [95% confidence intervals] obtained using the multiple imputation methods. The original data on the duration of the wound were available in 388 of 428 patients (90.1%): 117 of 142 surgical patients (82.4%), 260 of 275 endovascular patients (94.5%), and 11of 11 conservative patients (100.0%). The duration of the wound were not different among scheduled treatment strategies (P=0.31 between surgical and endovascular, P=0.48 between surgical and conservative, and P=0.38 between endovascular and conservative, by the linear mixed model including log-transformed duration of wound as the dependent variable, treatment strategies as the fixed effects, and the inter-institution variability as the random effects).