Retrospective analysis of factors influencing the self-healing of patients with enterocutaneous fistulas receiving conservative treatment

Table 1 Characteristics of enterocutaneous fistula patients compared between uncured and cured groups, n (%)

Category	Uncured group (n = 36)	Cured group (n = 55)	χ2	P value
Age (years)	54.64 ± 17.90	56.55 ± 14.56	0.557	0.579
Gender			2.552	0.110
Male	28 (77.78)	34 (61.82)
Female	8 (22.22)	21 (38.18)
BMI (kg/m²)	20.57 ± 3.76	21.86 ± 3.72	1.610	0.111
Albumin level (g/L)	33.95 ± 5.68	35.12 ± 5.59	0.967	0.336
Primary disease				0.044
Abdominal wall mass	1 (2.78)	2 (3.64)
Abdominal external hernia	1 (2.78)	0 (0.00)
Abdominal trauma	2 (5.56)	7 (12.73)
Gastroduodenal disease	3 (8.33)	8 (14.55)
Intestinal disease	9 (25.00)	4 (7.27)
Appendix disease	0 (0.00)	1 (1.82)
Colorectum disease	10 (27.78)	25 (45.45)
Hepatobiliary/pancreatic disease	4 (11.11)	6 (10.91)
Acute abdominal disease	2 (5.56)	0 (0.00)
Vascular disease	1 (2.78)	0 (0.00)
Urological disease	1 (2.78)	2 (3.64)
Gynecological disease	2 (5.56)	0 (0.00)
Fistula reason				0.055
Surgery	33 (91.67)	52 (94.55)
Trauma	0 (0.00)	3 (5.45)
Inflammatory bowel disease	2 (5.56)	0 (0.00)
Radio-/chemotherapy	1 (2.78)	0 (0.00)
Fistula location				0.006
Esophagus/stomach	0 (0.00)	1 (1.82)
Duodenum	2 (5.56)	11 (20.00)
Bile bowel/pancreatic bowel	1 (2.78)	1 (1.82)
Anastomosis fistula
Intestinal fistula	20 (55.56)	14 (25.45)
Colorectum fistula	13 (36.11)	21 (38.18)
Complex fistula	0 (0.00)	7 (12.73)
Fistula number			7.354	0.007
Single fistula	20 (55.56)	45 (81.82)
Multiple fistula	16 (44.44)	10 (18.18)
Fistula nature				0.277
Tubular fistula	33 (91.67)	48 (87.27)
Labral fistula	2 (5.56)	7 (12.73)
Internal fistula	1 (2.78)	0 (0.00)
Infection status			0.617	0.432
None	14 (28.89)	17 (30.91)
Yes	22 (61.11)	38 (69.09)
Diagnostic methods				0.249
Drainage fluid observation	16 (44.44)	31 (56.36)
Abdominal wound observation	12 (33.33)	13 (23.64)
DFO + AWO	6 (16.67)	11 (20.00)
Imaging diagnosis	2 (5.56)	0 (0.000
Nutrition feeding			1.043	0.307
None	1 (2.78)	6 (10.91)
Yes	35 (97.22)	49 (89.09)
Somatostatin			0.009	0.925
None	20 (55.56)	30 (54.55)
Yes	16 (44.44)	25 (45.45)
Growth hormone			3.376	0.066
None	25 (69.44)	47 (85.45)
Yes	11 (30.56)	8 (14.55)
Fibrin glue			12.194	< 0.001
None	29 (80.56)	24 (43.64)
Yes	7 (19.44)	31 (56.36)

BMI: Body mass index; DFO: Drainage fluid observation; AWO: Abdominal wound observation.

Table 2 Univariate regression analysis of factors associated with fistula cure

	b	Sb	χ2	OR (95%CI)	P value
Constant	0.199	0.319	0.389		0.533
Fistula number (multiple)	-1.309	0.526	6.179	0.270 (0.096, 0.758)	0.013
Fibrin glue application	1.697	0.523	10.528	5.459 (1.958, 15.219)	0.001

OR: Odds ratio; Sb: Standard bias.

Table 3 Interaction analysis between fistula number and fibrin glue application

Category	b	Sb	χ2	OR (95%CI)	P value
Constant	0.223	0.335	0.443	1.250	0.506
Fistula number (multiple)	-1.402	0.663	4.472	0.246 (0.067, 0.903)	0.034
Fibrin glue application	1.609	0.634	6.436	5.000 (1.142, 17.338)	0.011
Fistula number + fibrin glue application	0.262	1.109	0.056	1.300 (0.148, 11.422)	0.813

OR: Odds ratio; Sb: Standard bias.

Table 4 Comparison of predictive effects among different models

Constant	Fistula number		Fibrin glue application
	z	P value	z	P value
Fistula number + fibrin glue application	-2.242	0.025	-2.517	0.012

Table 5 Evaluation of efficacy combined with fistula number, fibrin glue and therapeutic prediction

Category	Therapeutic prediction of fistula number and fibrin glue application
	Uncured (n = 53)	Cured (n = 38)
Outcome of conservative treatment
Uncured (n = 36)	29	7
Cured (n = 55)	24	31
Sensitivity	56.36
Specificity	80.56
Positive predictive value	81.58
Negative predictive value	54.72
Youden index	36.92
AUC	0.747 (0.644, 0.850)

AUC: Area under the curve.