Table 1 Characteristics of all included studies

Abbas 2019 [26]

Multiple countries

English

Large-scale international study to determine the time-trend of surgical site infection (SSI) incidence in SSI surveillance networks. Networks identified through systematic literature review were provided with standardized data template

Cohort

Population

Colorectal surgery (no source)

Intervention

Surveillance over time

1) One-year increase in surveillance time

2) X years of surveillance

Comparison

1) Year One in surveillance time

2) X-1 years of surveillance

Outcome

SSI rate (no source)

STROBE Score

25 in 29

Study size: 320 123

10 networks included

8.5 (IQR 7–11) median years of surveillance

Relative Risk

Intervention 1)

1) 2 years 0.92 (0.89–0.96)

2) 3 years 0.90 (0.87–0.94)

3) 4 years 0.91 (0.87–0.95)

4) 5 years 0.86 (0.83–0.90)

5) 6 years 0.92 (0.87–0.96)

6) 7 years 0.84 (0.79–0.89)

7) 8 years 0.86 (0.80–0.92)

Intervention 2)

1) X = 3 - 0.98 (0.94–1.02)

2) X = 4 − 1.00 (0.96–1.05)

3) X = 5–0.95 (0.91–0.99)

4) X = 6–1.06 (1.01–1.12)

5) X = 7–0.92 (0.86–0.98)

6) X = 8–1.02 (0.04–1.11)

Angel García 2020 [17]

Spain

English

Study data from nine public hospitals in Murcia, Spain, from January 2006 to December 2015. The study developed two risk-adjustment models based on multiple logistic regression, without considering hospital bed size as a candidate variable

Cohort

Population

Colorectal surgery (ICD-9)

Exposure

1) Bed size > 500

2) Bed size 250–500

Comparison

Bed size < 250

Outcome

SSI (ICD-9 codes)

STROBE score

17 in 24

Study size 6 325

423 SSIs (7.32%)

Odds ratio (OR) (overall SSI, univariate analysis)

1) 0.54 (0,41–0,72)

2) 0.95 (0.74–1.23)

Du 2019 [16]

China

English

Multicenter surveillance of radical resection of colon and rectal carcinoma in 26 tertiary hospitals in 14 cities, from January 2015 to June 2016

Surveillance made by infection control professionals until discharge, using a

real-time nosocomial infection surveillance system, and by telephone 30 days postoperatively

Cohort

Population

Radical resection of colon and rectal carcinomas (ICD-9)

Exposure

Beds < 2500

Comparison

Beds ≥ 2500

Outcome

SSI (CDC 1992)

STROBE score

11 in 30

Study size 5 729

3 406 radical resection of colon carcinoma

2 323 radical resection of rectal carcinoma

206 SSIs (3.60%)

87 SSIs after colon resection (2.55%)

• 32 superficial (0.94%)

• 19 deep (0.56%)

• 36 organ space (1.06%)

119 SSIs after rectal resection (5.12%)

• 53 superficial (2.28%)

• 26 deep (1.11%)

• 40 organ space (1.72%)

OR (Multivariable analysis for overall SSI)

0.644 (0.451–0.921) resection of colon carcinoma

0.513 (0.356–0.739) resection of rectal carcinoma

El Aziz 2020 [25]

United States

English

Six-year longitudinal study using the American College of Surgeons—National Surgical Quality Improvement Program (ACS-NSQIP) database, an American quality improvement program gathering abstract patient information through predesigned data extraction sheets manage by trained data abstractors from all participating institutions

SSIs assessed in-hospital before discharge and after discharge until post-operative day 30

ORs adjusted for age > 80 years old, gender, race, body mass index, diabetes mellitus, current smoker within one year, dyspnea, functional health status prior to surgery, history of severe chronic obstructive pulmonary disease, ascites, congestive heart failure in 30 days before surgery, hypertension requiring medication, dialysis, disseminated cancer, open wound infection, steroid use for chronic condition, > 10% loss bodyweight in the last six months, bleeding disorders, transfusion > 1 unit red blood cells 72 h before surgery, pre-operative albumin and hematocrit, diagnosis, extent of resection, operative approach, diversion, operation time, any surgical complication before discharge, any medical complication before discharge, days from operation to discharge

Cohort

Population

Elective surgery for colon or rectal cancer, using Current Procedural Terminology (CPT) codes

Exposure

1) Discharge to skilled facility

2) Discharge to rehabilitation center

3) Discharge to separate acute care

Comparison

Discharged home

Outcome

SSI (ACS-NSQIP 2016)

STROBE score

21 in 27

Study size 108 617

3476 total SSIs (3.2%)

  1396 superficial SSIs (1.3%)

  349 Deep SSIs (0.3%)

  1915 organ space SSIs (1.8%)

ORs (overall SSIs, adjusted)

1) 1.02 (0.87–1.20)

2) 1.03 (0.81–1.31)

3) 1.25 (0.74–2.09)

Furuya-Kanamori 2017 [20]

Australia

English

New South Wales Admitted Patient Data Collection contains data on all admitted patient services provided by public and private hospitals in the region

Subset population of adult patients who underwent colorectal surgery between January 2002 and December 2013

The annual volume of colorectal surgery in public hospitals was categorized into tertiles, per surgical procedure: low-volume hospitals performed < 45procedures/year, mid-volume performed 45–115 procedures/year and high-volume performed > 115 procedures/year

Outcome includes in-hospital infection only

Cohort

Population

Colorectal surgery (ICD-10, Australian Modification)

Exposure

1) Mid-Hospital Volume

2) High-Hospital Volume

Comparison

Low-Hospital Volume

Outcome

Surgical site infection (ICD-10)

STROBE score

20.5 in 28

Study size 58 096 cases from 59 hospitals

Incidence: 9.64% (9.40–9.88%)

OR (overall SSI, crude analysis)

1) 1.23 (1.08–1.40)

2) 1.50 (1.34–1.69)

When risk-adjusted SSI rates per 1000 admissions were examined, low-volume hospitals performed better for colorectal procedures (91.7 for low, 96.7 for mid and 96.7 for high-volume public hospitals)

Manilich 2013 [23]

United States

English

Single-center study, with exclusion of outpatient surgical cases. 30-day follow-up by letter or phone call

Surgeon volume determined by the number of procedures in each major category that a surgeon performed in 2 years—colorectal surgical procedure as the unit of analysis

Cohort

Population

Major abdominal or transanal colorectal surgery (no source)

Exposure

Surgeon volume < 20 procedures

Comparison

Surgeon volume ≥ 20 procedures

Outcome

SSI (ACS-NSQIP)

STROBE score

23 in 29

Study size 3 552 cases by 15 surgeons

300 incisional SSIs (8.4%)

OR (overall SSI, adjusted OR)

1.38 (1.06–1.79)

Mannien 2008 [28]

Netherlands

English

Data from 1996 to 2006 from the Dutch National Nosocomial Surveillance Network. Hospital participation is voluntary. Hospitals can annually decide which surgical procedures to include, and post-discharge surveillance is strongly recommended

OR adjusted for post-discharge surveillance, American Society of Anesthesiologists (ASA) classification, wound contamination class, duration of surgery, duration of preoperative hospitalization and emergency procedure

Cohort

Population

Colectomy (no source)

Intervention

Surveillance

1) 1-year increase in surveillance time to operation

Comparison

1-less year in surveillance time to operation

Outcome

SSI (CDC 1992)

STROBE score

18.5 in 27

Study size 3 031

370 SSIs (12.2%)

OR (overall SSI, adjusted)

1) 0.92 (0.83–1.02)

Merkow 2013 [19]

United States

English

Multicenter study using centers participating in the ACS-NSQIP, that collects comprehensive data from > 500 hospitals, including 51 National Cancer Institutes

Age, race, ASA class, functional status, preoperative albumin level, hypertension, chronic obstructive pulmonary disease, chemoradiation, disseminated cancer and case complexity were all significantly different at baseline. Adjustment for differences in patient demographics and risk factors, as well as surgical complexity

Cohort

Population

Colorectal cancer surgery (CPT)

Exposure

Oncological Hospital

1) National Cancer Institute

Comparison

Non-oncological hospital

Outcome

SSI (ACS NSQIP)

STROBE score

18.5 in 28

Study size 52,265 from 310 hospitals

Incidence

7.7% superficial SSIs

4.8% deep or organ/space SSIs

OR (adjusted)

Superficial: 1.35 (1.08–1.70)

Deep or organ/space: 1.17 (0.98–1.40)

Schröder 2018 [18]

Germany

English

Data from surgical site infection module of the German national nosocomial infection surveillance system, the component for surgical site infections, which is patient based and voluntary. SSI following laparoscopic colon resection from 145 hospitals (44 public, 65 non-profit and 36 private) and following open colon resection in 159 (45, 67 and 37, respectively)

Cohort

Population

Colon procedures (national codes)

Exposure

Public ownership

Non-profit ownership

Bed size < 400

Comparison

Private ownership

Bed size < 400

Outcome

SSI

STROBE score

19 in 29

Study size 28,291

• 19 453 open colon procedures

• 8838 laparoscopic colon procedures

ORs (overall SSIs, multivariate analysis)

1.12 (0.86–1.47) for public ownership

0.85 (0.66–1.09) for non-profit ownership

0.81 (0.51–1.29) for bed size < 400

Serra-Aracil 2011 [15]

Spain

English

Multicenter study of 19 public hospitals in Catalonia, Spain, between June 2007 and March 2008

Colon operation defined as any resection above the peritoneal reflection. Rectal operation defined as any resection below the same point. Inclusion criteria were the application of all preventive measures and rectal cancer operations with oncologic resections. Outpatient visit after 30 days

Cohort

Population

Elective operations for colon or rectal cancer

Exposure

1) > 500 beds

2) 250–500 beds

Comparison

< 250 beds

Outcome

SSI (CDC 1992)

STROBE score

18 in 29

Study Size: 611

383 colon cancer operation

89 total SSIs (23.2%)

• 49 superficial SSIs (12.8%)

• 8 deep SSIs (2.1%)

• 32 organ space SSIs (8.4%)

228 rectal cancer operation

63 SSIs (27.6%)

• 31 superficial SSIs (13.6%)

• 13 deep SSIs (5.7%)

• 19 organ space SSIs (8.3%)

OR (univariate analysis)

Colon cancer operations

Overall SSI

1) 0.48 (0.25–0.95)

2) 0.41 (0.17–0.95)

Incisional SSI

1) 0.36 (0.18–0.76)

2) 0.26 (0.09–0.68)

Organ/space SSI

1) 1.25 (0.41–5.68)

2) 1.52 (0.39–7.80)

Rectal cancer operations

Overall SSI

1) 0.69 (0.30–1.67)

2) 0.68 (0.24–1.94)

Incisional SSI

1) 0.89 (0.35–2.63)

2) 0.90 (0.27–3.13)

Organ/space SSI

1) 0.51 (0.16-2.00)

2) 0.49 (0.08–2.54)

Spolverato 2019 [21]

Italy

English

Data from National Italian Hospital Discharge Dataset, from January 2002 to November 2014

Adult patients only

Hospital volume calculated as the average annual number of rectal cancer procedures performed at each hospital, during study period, divided into tertiles (respectively 1–12, 13–31, > 31 surgeries/year)

Main outcome is failure to rescue, defined as the mortality rate among patients with complications, which is why there is no adjusted analysis specifically for SSI; however, low-volume hospitals, in multivariable analysis, are associated with higher rate of failure to rescue and any complication, when compared with high volume hospitals

Cohort

Population

Major surgical procedure for primary rectal cancer (ICD-9)

Exposure

1) High-volume hospital

2) Intermediate-volume hospital

Comparison

Low-volume hospital

Outcome

SSI (ICD-9)

STROBE score

20.5 in 30

Study size 75 280

3.9% SSI incidence

OR (overall SSI, crude analysis)

1) 0.99 (0.90–1.08)

2) 0.95 (0.87–1.04)

Staszewicz 2014 [29]

Switzerland

English

Data collected from 1998 to 2010 from the Swissnoso consortium, a voluntary participation surveillance network of Swiss public hospitals

OR adjusted for age, sex, ASA Score ≥ 3, delay from admission to operation > 2 days, emergency, antibiotic prophylaxis, contamination class ≥ 3, multiple procedures, laparoscopy, duration > T, re-intervention for non-infectious complications

Cohort

Population

Colectomy (no source)

Intervention

Surveillance

1) time to operation

Outcome

SSI (CDC 1992)

STROBE score

18 in 28

Study size 7411

1349 SSIs (18.2%)

555 superficial SSIs (7.5%)

308 deep SSIs (4.2%)

486 organ/space SSIs (6.6%)

OR (overall SSI, adjusted)

1.05 (1.03–1.07)

Tserenpuntsag 2014 [11]

United States

English

Multicenter study of 174 New York State hospitals, with mandatory surveillance of SSIs following colon procedures through the NHSN, including post-discharge detection of SSI. Authors used 2009–2010 data of an audit of the surveillance program, performed by trained program staff certified in infection control

If a small bowel procedure, kidney transplant, liver transplant, or bile duct, liver pancreatic or rectal procedure was performed through the same incision, and it was not clear which procedure was associated with infection, the SSI was attributed to 1 of the above listed procedures

Case-control

Population

Colon procedures, using ICD-9 codes

Exposures

Teaching hospitals

Bed size > 500

Comparison

Nonteaching hospitals

Bed size ≤ 500

Outcome:

SSI (CDC 1992)

STROBE Score:

18.5 in 28

Study Size: 2 656 cases from 175 hospitals

698 SSIs identified

• 355 superficial incisional

• 343 deep incisional and organ space

ORs (overall SSI, bivariable analysis):

Teaching hospitals:

1.88 (1.55–2.95) for superficial

1.86 (1.45–2.39) for deep incisional and organ space

Bed size:

2.32 (1.82–2.95) for superficial incisional and 2.08 (1.54–2.80) for deep and organ space

Vicentini 2019 [27]

Italy

English

32 Piedmont hospitals (primary, secondary and tertiary) participating in the voluntary Italian surveillance system for SSIs, using data from January 2009 to December 2015

Surveillance must be performed at least 6 months/year or a minimum of 50 procedures must be monitored

Surveillance time is equivalent to the number of years of participation in a surveillance program

Cohort

Population

Colon surgery (ICD-9)

Intervention

Surveillance

1) Participating in surveillance program for over 5 years

2) 1-unit increase in the number of monitored procedures

3) 1-year increase in surveillance time

Comparison

1) No participation in surveillance network

Outcome

SSI (ECDC)

STROBE score

16 in 27

Study size 6 060

595 SSIs (9.83%), 172 post-discharge

370 incisional SSIs (6.13%)

96 deep SSIs (1.59%)

97 organ/space SSI (1.60%)

RR (overall SSI)

1) 0.64 (0.46–0.90)

2) 0.99 (0.98-1.00)

3) 0.93 (0.89–0.97)

Wang 2021 [22]

United States

English

Safety-net hospitals are mandated to treat patients regardless of their ability to pay, and consequently carry a high safety-net burden (SNB), defined as the proportion of patients who are either uninsured or Medicaid-insured. Hospitals were stratified into tertiles of low, medium and high SNB

Data from State Inpatient Databases (2009–2014)

Hospital volume stratified into quartiles by colectomy case volume

Adult patients only

Cohort

Population

Colectomy (ICD-9)

Exposure

1. High safety-net burden

2. Medium safety-net burden

3. Hospital volume—4th quartile

4. Hospital volume—3rd quartile

5. Hospital volume—2nd quartile

Comparison

1. Low safety-net burden (1–2)

2. Hospital volume—1st quartile (3–5)

Outcome

SSI (ICD-9)

STROBE score

21.5 in 30

Study size 459 568

29 117 SSIs (6.3%)

OR (overall SSI, crude analysis)

1. 1.35 (1.31–1.40)

2. 0.97 (0.94-1.00)

3. 0.51 (0.50–0.53)

4. 0.64 (0.62–0.66)

5. 0.70 (0.68–0.73)

Yi 2018 [24]

United States

English

Administrative databases used for colorectal surgical patients discharged in years 2013 and 2014, in a hybrid tertiary referral central with 8 campuses and 2247 beds. Over 80% of study patients were from campuses with high colorectal surgery volume

Volume of surgery determined by the actual number of patients operated on per surgeon. High volume surgeon with case volume of more than 34 cases in the last 2 years, medium volume with case volume between 14 and 34, and low volume with fewer than 14 cases

Cohort study

Population

Colorectal procedures (ICD-9)

Exposure

Medium-volume Surgeon

Low-volume Surgeon

Comparison

High-volume Surgeon

Outcome

SSI (no source provided)

STROBE Score

16,5 in 27

Study size 1 190 cases by 44 surgeons

ORs (overall SSI, adjusted)

0.23 (0.08–0.65) for medium-volume surgeons

0.39 (0.09–1.71) for low-volume surgeons