. 2020 Jul 29;36(1):187–204. doi: 10.1007/s00467-020-04623-2

Table 2.

Effects of gastrostomy tube placement on infection risk

Study [reference]	Patients—number and age (year)	Intervention and comparator groups	Outcome—peritonitis rates	Outcome—PD technique survival/patient survival
Balfe 1990 [30]	N = 20 Age range 0.3 to 12.8	PD patients who were enterally fed – 13 G-tubes (10 were on PD prior to the G-tube insertion) No comparator	G-tube exit site infections in 4, one of these patients developed peritonitis No details on technique of G-tube insertion	Not stated
Coleman 1998 [31]	N = 22 Age range 0.2 to 10.3	n = 14 children with G-tube and PD catheter inserted simultaneously vs. n = 2 with G-button and PD catheter simultaneously vs. n = 2 with G-tube after initiating PD (+ 4 on HD)	2 patients had peritonitis attributed to G-tube (includes 1 fungal peritonitis in a malnourished patient) Open gastrostomy recommended	Not stated
Ramage 1999 [32]	N = 23 Mean age 3.8 ± 3.2	2 groups: 1. G-tube inserted prior to the commencement of PD (n = 9; includes n = 2 who also underwent surgical G-tube replacement while receiving PD) 2. G-tube inserted while receiving PD (n = 14) Type and technique of G-tube insertion—not stated Controls—children on chronic PD without a G-tube (n = 127)	Peritonitis occurred every 18.4 patient-months in controls and 7.8 patient-months in those with a G-tube (p < 0.001) Peritonitis occurred every 6.0 patient-months before and 8.1 patient-months after G-tube insertion in those undergoing G-tube insertion on PD	PD catheter replacement secondary to infection occurred every 109.4 patient-months in controls and 39.9 patient-months in those with a G-tube Peritonitis did not occur before G-tube insertion and occurred every 32.5 patient-months following G-tube insertion
Ledermann 2002 [33]	N = 29 Median age 3.9 (0.5–13.3)	3 groups: 1. G-tube (PEG/open G-tube ± Nissen) before starting PD (n = 15) 2. Open G-tube ± Nissen after PD catheter insertion/start of PD (n = 9) 3. PEG after PD catheter insertion/start of PD (n = 5)	Group 2 (130 months G and PD), 0.2 vs. 1.4 episodes of peritonitis/patient-year before and after G-tube insertion Group 3 0.5 episodes of peritonitis/patient-year before and 4 out of 5 children developed peritonitis soon after PEG placement	In group 3 (PEG after PD) - 2 transferred to hemodialysis, 1 remained on PD after treatment of Candida peritonitis and 1 died
Rahim 2004 [34]	N = 68 Mean 9 (range 0.1 to 19)	Presence of G-tube (n = 68) vs. no G-tube (n = 59) G-tube before PD cath n = 41 G-tube at the same time as a PD catheter n = 13 G-tube after PD catheter n = 14 G-tube insertion technique—not stated	Presence of a G-tube was associated with a higher peritonitis rate (p < 0.001) compared to chronic PD without a G-tube No difference in the peritonitis rate between patients with a G-tube inserted before or after a PD catheter (p = 0.46) No difference in the peritonitis rate between patients with a G-tube inserted before or at the same time as a PD catheter (p = 0.06)	Not stated Note—some patients were switched from PD to HD after G-tube placement for approximately 4–6 weeks while the G-tube tract healed (no clear indications or difference in outcomes with this practice mentioned)
von Schnakenburg 2006 [35]	N = 27 Median 1.3 (range 0.25 to 10.9) years	G-tube insertion after Tenchkoff catheter insertion/patient already on PD 25 PEG and 2 open gastrostomies No comparator group	Early peritonitis < 7 days after PEG in 10/27 (37%); was effectively treated with intraperitoneal antibiotics in 4/10 Fungal peritonitis in 7/27 (26%) patients Note • significant variations in practice across sites. Variations in antibiotic and antifungal prophylaxis used • most patients were malnourished	In those with fungal peritonitis: • 4 cessations of PD and change to hemodialysis • 2 deaths In 18/27 (67%) patients, PD was successfully reinitiated shortly after PEG insertion
Lindley 2012 [42]	N = 33 Mean age 4.9 in OPEN group and 3.7 years in LAP group	2 groups: 1. laparoscopic-assisted PEG and PD catheter insertion (n = 10) – LAP group 2. open gastrostomy and PD catheter (n = 23) – OPEN group	Peritonitis and infection rates* per catheter-year were 0.89 and 0.7 in LAP and 0.59 and 0.5 in OPEN group (not significant) The risk of peritonitis and infection was not related to method of placement (not significant)	PD catheter survival - median 12 months in the LAP group and 17 months in the OPEN group (not significant)
Phan 2013 [36]	N = 7 Median 10	n = 207 on chronic PD n = 7 (3%) with concomitant gastrostomies Timing and technique of G-tube insertion - not stated	Not stated	Hazard ratio for re-operations for infections was 5.01 (95% CI 1.5–16.6) higher in children with gastrostomies compared to those without gastrostomies, p = 0.008 (on fully adjusted multivariable model)
Prestidge 2015 [37]	N = 17 7.2 (range 10 weeks to 17.2 years)	14/17 (82%) open surgical technique, 3 laparoscopic A G-tube was inserted in 15 patients after PD had been established 2 patients had simultaneous G-tube and PD catheter insertion	- 2 cases of early peritonitis with organisms derived from the gastrointestinal tract - no case of fungal peritonitis No statistically significant difference between incident rates of bacterial peritonitis before G-tube placement (0.6 episodes per patient-year; 95% CI 0.26–1.18) and post-G-tube placement (1.21 episodes per patient-year; 95% CI 0.69–1.97)	- No PD technique failure - No deaths
Zaritsky 2018 [38]	N = 156 infants with chronic PD catheter placed prior to their first birthday 53% under 30 days at PD catheter insertion	3 groups: - G-tube placement before or at the same time as PD catheter (n = 47) - G-tube placement after PD catheter insertion/start of PD (n = 49) - No G-tube (n = 61)	G-tube insertion after catheter placement was associated with a nearly 3-fold [OR (95% CI) 2.81 (1.31, 6.01), p < 0.01] increased risk of peritonitis	Not stated

Study [reference]

Patients—number and age (year)

Intervention and comparator groups

Outcome—peritonitis rates

Outcome—PD technique survival/patient survival

Balfe 1990

[30]

N = 20

Age range 0.3 to 12.8

PD patients who were enterally fed – 13 G-tubes (10 were on PD prior to the G-tube insertion)

No comparator

G-tube exit site infections in 4, one of these patients developed peritonitis

No details on technique of G-tube insertion

Not stated

Coleman 1998

[31]

N = 22

Age range 0.2 to 10.3

n = 14 children with G-tube and PD catheter inserted simultaneously

vs. n = 2 with G-button and PD catheter simultaneously

vs. n = 2 with G-tube after initiating PD

(+ 4 on HD)

2 patients had peritonitis attributed to G-tube (includes 1 fungal peritonitis in a malnourished patient)

Open gastrostomy recommended

Not stated

Ramage 1999

[32]

N = 23

Mean age 3.8 ± 3.2

2 groups:

1. G-tube inserted prior to the commencement of PD (n = 9; includes n = 2 who also underwent surgical G-tube replacement while receiving PD)

2. G-tube inserted while receiving PD (n = 14)

Type and technique of G-tube insertion—not stated

Controls—children on chronic PD without a G-tube (n = 127)

Peritonitis occurred every 18.4 patient-months in controls and 7.8 patient-months in those with a G-tube (p < 0.001)

Peritonitis occurred every 6.0 patient-months before and 8.1 patient-months after G-tube insertion in those undergoing G-tube insertion on PD

PD catheter replacement secondary

to infection occurred every 109.4 patient-months in controls and 39.9 patient-months in those with a G-tube

Peritonitis did not occur before G-tube insertion and occurred every 32.5 patient-months following G-tube insertion

Ledermann 2002

[33]

N = 29

Median age 3.9 (0.5–13.3)

3 groups:

1. G-tube (PEG/open G-tube ± Nissen) before starting PD (n = 15)

2. Open G-tube ± Nissen after PD catheter insertion/start of PD (n = 9)

3. PEG after PD catheter insertion/start of PD (n = 5)

Group 2 (130 months G and PD), 0.2 vs. 1.4 episodes of peritonitis/patient-year before and after G-tube insertion

Group 3 0.5 episodes of peritonitis/patient-year before and

4 out of 5 children developed peritonitis soon after PEG placement

In group 3 (PEG after PD) - 2 transferred to hemodialysis,

1 remained on PD after treatment of Candida peritonitis and 1 died

Rahim 2004

[34]

N = 68

Mean 9 (range 0.1 to 19)

Presence of G-tube (n = 68) vs. no G-tube (n = 59)

G-tube before PD cath n = 41

G-tube at the same time as a PD catheter n = 13

G-tube after PD catheter n = 14

G-tube insertion technique—not stated

Presence of a G-tube was associated with a higher peritonitis rate (p < 0.001) compared to chronic PD without a G-tube

No difference in the peritonitis rate between patients with a G-tube inserted before or after a PD catheter (p = 0.46)

No difference in the peritonitis rate between patients with a G-tube inserted before or at the same time as a PD catheter (p = 0.06)

Not stated

Note—some patients were switched from PD to HD after G-tube placement for approximately 4–6 weeks while the G-tube tract healed (no clear indications or difference in outcomes with this practice mentioned)

von Schnakenburg 2006

[35]

N = 27

Median 1.3 (range 0.25 to 10.9) years

G-tube insertion after Tenchkoff catheter insertion/patient already on PD

25 PEG and 2 open gastrostomies

No comparator group

Early peritonitis < 7 days after PEG in 10/27 (37%); was effectively treated with intraperitoneal antibiotics in 4/10

Fungal peritonitis in 7/27 (26%) patients

Note

• significant variations in practice across sites. Variations in antibiotic and antifungal prophylaxis used

• most patients were malnourished

In those with fungal peritonitis:

• 4 cessations of PD and change to hemodialysis

• 2 deaths

In 18/27 (67%) patients, PD was successfully reinitiated shortly after PEG insertion

Lindley 2012

[42]

N = 33

Mean age 4.9 in OPEN group and 3.7 years in LAP group

2 groups:

1. laparoscopic-assisted PEG and PD catheter insertion (n = 10) – LAP group

2. open gastrostomy and PD catheter (n = 23) – OPEN group

Peritonitis and infection rates* per catheter-year were 0.89 and 0.7 in LAP and 0.59 and 0.5 in OPEN group (not significant)

The risk of peritonitis and infection was not related to method of placement (not significant)

PD catheter survival - median 12 months in the LAP group and 17 months in the OPEN group (not significant)

Phan 2013

[36]

N = 7

Median 10

n = 207 on chronic PD

n = 7 (3%) with concomitant gastrostomies

Timing and technique of G-tube insertion - not stated

Not stated

Hazard ratio for re-operations for infections was 5.01 (95% CI 1.5–16.6) higher in children with gastrostomies compared to those without gastrostomies, p = 0.008 (on fully adjusted multivariable model)

Prestidge

2015

[37]

N = 17

7.2 (range 10 weeks to

17.2 years)

14/17 (82%) open surgical technique, 3 laparoscopic

A G-tube was inserted in 15 patients after PD had been established

2 patients had simultaneous G-tube and PD catheter insertion

- 2 cases of early peritonitis with organisms derived from the gastrointestinal tract

- no case of fungal peritonitis

No statistically significant difference between incident rates of bacterial peritonitis before G-tube placement (0.6 episodes per patient-year; 95% CI 0.26–1.18) and post-G-tube placement (1.21 episodes per patient-year; 95% CI 0.69–1.97)

- No PD technique failure

- No deaths

Zaritsky 2018

[38]

N = 156 infants with chronic PD catheter placed prior to their first birthday

53% under 30 days at PD catheter insertion

3 groups:

- G-tube placement before or at the same time as PD catheter (n = 47)

- G-tube placement after PD catheter insertion/start of PD (n = 49)

- No G-tube (n = 61)

G-tube insertion after catheter placement was associated with a nearly 3-fold [OR (95% CI) 2.81 (1.31, 6.01), p < 0.01] increased risk of peritonitis

Not stated

*Peritonitis was defined as the presence of a white blood cell count > 100/mm3 with at least 50% being polymorphonuclear leukocytes, and infection was defined as the presence of positive peritoneal cultures with peritonitis

CI, confidence interval; G, gastrostomy; HD, hemodialysis; PD, peritoneal dialysis; PEG, percutaneous endoscopic gastrostomy; OR, odds ratio