Postoperative three-dimensional endoanal ultrasound findings and relation to anal fistula plug failure

Figures & data

Table 1. Characteristics of participants in the study at inclusion. Values are given as median (range) or n (%).

Characteristic	Total (N = 95)
Age	44 (10–80)
Gender
Female	34 (36)
Male	61 (64)
Etiology
Cryptoglandular	65 (68)
Crohn’s disease	30 (32)
Earlier procedure
None	66 (70)
Fistulectomy	12 (13)
Advancement flap	3 (3)
Other	13 (14)
Unknown	1 (1)
Internal opening
Anterior	30 (32)
Lateral	15 (16)
Posterior	49 (50)
Unknown	1 (1)

Figure 1. Flow-chart summarising the treatment and follow-up process of the study.

Figure 2. The visible fistula was noted as a hypoechogenic area (white arrow) at 3D EAUS in axial plane after the AFP procedure. IAS: internal anal sphincter.

Figure 3. Inflammation was noted as a slightly hyperechogenic area (red arrow) at 3D EAUS in multiplanar reconstruction after the AFP procedure. Strongly hyperechogenic dots (white arrow) represent gas bubbles in the fistula tract. IAS: internal anal sphincter; PR: puborectalis muscle.

Table 2. The relevant 3D EAUS findings related to the results of the AFP procedure as identified in the long-term follow-up. Values are given as n(positive parameter)/n(plugs in group) and (%).

EAUS parameter	Time	Total	Failure	Success	Not possible to analyse	EAUS missing	p
Gas	2 weeks	65/121 (54)	49/94 (52)	16/27 (59)	20/141 (14)	5/146 (3)	p = 0.66
	3 months	55/98 (56)	47/71 (66)	8/27 (30)	14/112 (13)	7/119 (6)	p = 0.0014
	Late control	35/64 (55)	32/43 (74)	3/21 (14)	13/70 (19)	7/77 (9)	p < 0.001
Inflammation	2 weeks	71/125 (57)	57/97 (59)	14/28 (50)	16/141 (11)	5/146 (3)	p = 0.52
	3 months	51/98 (52)	43/71 (61)	8/27 (30)	14/112 (13)	7/119 (6)	p = 0.0072
Visible fistula	3 months	89/98 (91)	69/73 (95)	20/25 (80)	14/112 (13)	7/119 (6)	p = 0.045
	Late control	52/64 (81)	43/45 (96)	9/19 (47)	6/70 (9)	7/77 (9)	p < 0.001

Table 3. Diagnostic probabilities for AFP failure based on 3D EAUS findings. PPV: positive predictive value; NPV: negative predictive value.

EAUS parameter	Time	Sensitivity	Specificity	PPV	NPV
Gas	3 months	66%	70%	85%	44%
	Late control	74%	86%	91%	62%
Inflammation	3 months	61%	70%	84%	40%
Visible fistula	3 months	95%	20%	76%	56%
	Late control	96%	53%	83%	83%

Table 4. Odds for AFP failure based on logistic regression model analysis of the relevant 3D EAUS findings. 95% CI = 95% Confidence Interval.

EAUS parameter	P	Odds	95% CI
Gas 3 months	0.0017	4.65	1.78–12.16
Gas late control	<0.001	17.46	4.30–70.86
Inflammation 3 months	0.0078	3.65	1.41–9.46
Visible fistula 3 months	0.0042	4.31	1.06–17.59
Visible fistula late control	<0.001	23.89	4.45–128.13

Figure 4. Result algorithm for the relationship between the clinical symptom of fluid discharge through the external fistula opening combined with the 3D EAUS parameter of gas in fistula 3 months after the AFP procedure analysed in relation to AFP failure for “the double positive” and “double negative” plugs (Fischer’s exact test, p < 0.001).

Data availability statement

Data is available on request due to privacy/ethical restrictions.