In het algemeen zijn de contact- en zorgmomenten in een poliklinische setting korter en minder intensief in vergelijking met de klinische setting. Een Nederlandse studie toonde eerder aan dat de algemene voorzorgsmaatregelen op de polikliniek voldoende zijn om overdracht van meticilline-resistente Staphylococcus aureus (MRSA) naar medewerkers te voorkomen (Van Rijen, 2014).

Echter, voor andere bijzonder resistente micro-organismen (BRMO) is het onbekend of transmissie plaatsvindt bij polikliniek bezoek van patiënten die deze BRMO bij zich dragen. Als gevolg hiervan is het onduidelijk of er voor deze groep patiënten aanvullende infectiepreventiemaatregelen nodig zijn op de polikliniek. 

De werkgroep heeft de huidige literatuur-search naar de noodzaak van aanvullende infectiepreventiemaatregelen afgebakend tot VRE, CPE/CRE en CP-CRAB. Deze selectie is gebaseerd op het mogelijke risico op verspreiding binnen een instelling en het veroorzaken van infecties die nauwelijks meer met antibiotica te behandelen zijn.

Description of studies

Grabsch (2006) conducted an observational study to assess the risk of environmental and healthcare worker (HCW) contamination with vancomycin-resistant enterococci (VRE) during outpatient procedures performed on fecally continent patients currently colonized with VRE (cVRE) or previously colonized with VRE (pVRE). Patients participated in standardized mock outpatient consultations and radiology procedures and actual hemodialysis sessions. IN total Mock procedures were used to standardize the type and duration of HCW-patient interaction and involved the use of outpatient facilities that had been carefully cleaned and proven to be free of VRE contamination. Actual hemodialysis sessions were already standardized, given the routine duration (4 hours) and procedures involved in hemodialysis, and were also conducted in sites proven to be free of VRE contamination. Environmental samples en samples from HCWs for culture were obtained before and after each mock outpatient consultation, both with saline-moistened swabs (an area of 10-20 cm2 was wiped with each swab) and with contact plates (multiple surface impressions were taken with each plate) covered with bile esculin azide agar, using the Rodac technique. In total, cVRE patients and 13 HCWs to participate in a total of 49 mock outpatient and radiology sessions, and a further 7 cVRE patients and 15 HCWs participated in 26 study hemodialysis sessions. Outcomes were rates of environmental, HCW, and patient contamination with vancomycin-resistant enterococci (VRE).

Heinrichs (2019) conducted an observational study analysing outbreaks of New Delhi metallo-β-lactamase (NDM) producing K. pneumoniae outbreaks in 2 Belgian hospitals (hospital A and hospital B) in 2014–2015. An outbreak case was defined as a patient colonized or infected by carbapenem-nonsusceptible K. pneumoniae between August 2014 and December 2015. Post hoc, outbreak cases were classified as proven (NDM-producing K. pneumoniae of identical pulsed-field gel electrophoresis (PFGE) profile and ST) or possible (carbapenem-nonsusceptible K. pneumoniae but without molecular NDM detection or typing analysis). K. pneumoniae isolates were collected from inpatients of the 2 study hospitals, using cultures of either clinical or rectal screening swab samples. At the time of the outbreak, a screening strategy for gram-negative MDROs was applied systematically at patient admission in 4 high-risk departments in hospital A. In hospital B, there was no systematic active screening surveillance for gram-negative MDROs until the detection of 3 patients positive for NDM-producing K. pneumoniae in the Department of Internal Medicine. A total of 74 NDM-producing K. pneumoniae isolates were analyzed in this study. This included 9 isolates from hospital A and 24 from hospital B. For comparison, 41 NDM-producing K. pneumoniae isolates were included that were collected elsewhere in Belgium and sent to the national reference center in 2014 or 2015. Molecular typing was performed by genomic macrorestriction (XbaI) followed by PFGE. WGS was performed using the Illumina MiSeq platform.

Kalocheretis (2004) conducted an observational study to determine the epidemiology of VRE colonization within this high-risk population and define the risk factors. Included were swab specimens were obtained from consecutive haemodialysis patients consenting to participate in the study. In total, 334 non replicate swab specimens were included (180 from males, 154 from females, mean age of 63.33 ± 14.12 years, mean time on haemodialysis of 37.80 ± 41.20 months). Swab specimens were obtained from stools or by rectal insertion. Specimens were inoculated on Enterococcosel agar with 6 mg/L vancomycin and incubated aerobically at 37°C for 24 h. Negative cultures were incubated for 48 h. Genus, species identification and susceptibility testing were carried out using the VITEK. A multiplex PCR confirmed species identification of E. faecalis, E. faecium, Enterococcus gallinarum and Enterococcus casseliflavus and detected the major glycopeptide resistance genes (vanA, vanB, vanC₁ and vanC_2,3). DNA fingerprinting was carried out using PFGE. Outcomes were number of positive isolates and corresponding PFGE fingerprint. 

Klein (2021) conducted an observational study describing transmission cluster of blaKPC-2- positive Enterobacter cloacae among patients treated in a highly frequented outpatient department. As part of the routine infection-control measures, patients at risk (including immunocompromised patients) for MDROs were screened on admission on a regular basis. In addition, monthly environmental screening in inpatient wards was performed to identify potential reservoirs for MDROs. Following identification of the first 3 cases in October 2019, infection-control measures were expanded to include screening of all outpatients and monthly environmental screening of outpatient facilities in the hematology department. Patient characteristics and movement data of all patients with Klebsiella pneumoniaecarbapenemase-2 (KPC-2)–producing E. cloacae were extracted from the medical record. Time to first detection is defined as the time span between the first visit in the hospital until the first detection of carbapenem-resistant (CR) E. cloacae in a microbiological sample. Positive clinical and environmental samples were subjected to whole genome sequencing using the Illumina MiSeq platform to analyse genetic relatedness. 

Results

Transmission (outbreaks)

Grabsch (2006) found that VRE contamination of the chair and couch was noted in 36%-48% of outpatient consultation and radiology sessions, and in 58% of all hemodialysis sessions (chair only). Other environmental sites were less frequently colonized. The HCW’s gown was contaminated in 4%-20% of outpatient consultation and radiology sessions and in 30% of hemodialysis sessions. The HCW’s gloved hands (sampled for culture immediately after a rectal swab specimen was obtained), were colonized in 16% of sessions, but after glove removal and handwashing in water and 4% chlorhexidine, HCW’s hands still harboured VRE in 4%- 8% of sessions.

Heinrichs (2009) found that the incidence rate of NDM-producing K. pneumoniae increased from a baseline rate of 1.5 cases per 100 000 patient-days (time period, August 2013 to July 2014) to 9.4 per 100 000 patient-days during the outbreak period (August 2014 to March 2015). The first patient was detected in August 2014, 1 month after being admitted to the ICU. After the index patient transferred to the Department of Pneumology, a second patient was detected with nosocomial acquisition of NDM-producing K. pneumoniae. Later 7 additional infected (n = 4) or colonized (n = 3) patients were identified in 5 wards of hospital A. All 9 isolates showed the same PFGE types and belonged to ST716 indicating a close genetic relationship.

In addition, Heinrichs (2009) found an outbreak was declared at hospital B in January 2015 and persisted until December 2015. The incidence rate of NDM-producing K. pneumoniae increased from 1 case per 100 000 patient-days in 2014 to 102 cases per 100 000 patient-days in 2015. After the identification of the first 3 NDM-positive K. pneumoniae patients, an active surveillance screening program was implemented which included screening for MDROs among (1) contact patients and (2) inpatients staying ≥15 days in hospital B. In parallel, inpatients from all departments of hospital B were screened during 3 point-prevalence surveys. Patients with positive results were placed on contact precautions and were cohorted in a single ward with dedicated nursing staff. In total, 64 additional possible outbreak cases were identified involving the entire hospital. 

Molecular typing using MLST, PFGE, and wgMLST allowed to separate out the 74 NDM-producing K. pneumoniaestrains investigated in this study. Overall, 30 of 74 strains (ie, all 9 strains from hospital A, 20 of 24 from hospital B, and 1 of 41 from the NRC collection) were assigned to ST716, shared the same PGFE type, and clustered together in the wgMLST analysis. These 30 strains were separated by a mean pairwise allelic distance of only 17 alleles (of 4015 alleles analyzed) and by ≤6 SNPs indicating that they were closely related. After construction of a minimum spanning tree, the strain genome from second case patient in hospital A was consistent with a progenitor status relative to hospital B cases. The second case patient in hospital A was also repeatedly seen at the outpatient clinic of hospital B after discharge from hospital A which probably caused the outpatient clinic of hospital B to act as bridging site for the ongoing outbreak between hospitals.

Kalocheretis (2004) found that the frequency of VRE colonization was 3.9% (13 out of 334 patients) separated over 4 dialysis units (unit I to IV). DNA fingerprinting allocated the 13 VRE isolates into seven distinct PFGE types designated A to G. Type B was the predominant type being represented by six isolates (46.2%), type A was represented by two isolates (15.4%), while there was one isolate each of types C to G (7.7% each). Two isolates of distinct types were obtained from Unit I (one type E, one type F), Unit II (one type B, one type D) and Unit III (one type B, one type G), while seven isolates representing three types originated from Unit IV (two type A, four type B and one type C). The predominance of one clone and its isolation from several units strongly indicate interfacility transmission of strains.

Klein (2021) identified 6 patients (patient [P] 1 to P6) with blaKPC-2-positive E. cloacae. Initial MDRO screening was negative for all 6 patients. Patient 1, P2, and P3 were only rectally colonized without evidence of infections. Patient 4 had a urinary tract infection (UTI) and was rectally colonized later on, P5 was rectally colonized and developed a UTI and pneumonia with a fatal outcome. Patient 6 was initially only rectally colonized but developed bloodstream infection, a UTI, and skin and soft tissue infection afterwards. Time to first detection ranged between 20 and 83 days after initial negative screening. Molecular typing of patient isolates by WGS indicated close genetic relatedness between all 6 isolates on multilocus sequence typing (MLST) level (ST96). Four isolates (P1, P3–P5) were very closely related as indicated by single nucleotide polymorphisms (SNPs) less than 10 over a core genome of 4 223 058 nucleotides (4318 genes), whereas isolates from P2 and P6 were genetically less related to the isolates of P1 and P3–P5 (18 and 92 SNPs respectively). According to patient-movement data, the 4 patients with the closely related clones (P1 and P3–P5) were never located on the same ward at the same time or in the same room at any given time during inpatient treatment. Klein (2021) observed temporal overlap in patient visits in the outpatient clinics. Patients 1 and 3 had only 1 overlap, whereas P3–P5 had overlapping clinic visits on multiple occasions. Patient 3 visited the outpatient department more than 50 times and P4 more than 60 times in 1 year. Patients 2 and 6 had no spatiotemporal overlap with any of the other patients. Overall, patient-movement data and epidemiological overlap were consistent with the molecular typing data.

A systematic review of the literature was performed to answer the following question: Are there outbreaks seen in outpatient clinics caused by VRE, CPE/CRE or CP-CRAB if no additional infection prevention measures are taken in addition to the standard infection prevention measures?

P:         Patients/clients visiting an outpatient clinic who are infected with or carrier of VRE, CPE/CRE, or CP-CRAB.

I:          Additional infection control measures used to control the outbreak.

C:         Standard infection prevention measures (no isolation).

O:        Transmission (outbreaks) involving patients, clients or healthcare workers.

Relevant outcome measures

The guideline development group considered transmission (outbreaks) as a critical outcome measure for decision making.

A priori, the working group did not define the outcome measures listed above but used the definitions used in the studies. 

Search and select (methods)

The databases Medline (via OVID), Embase (via embase.com), Web of Science, and Cinahl, were searched with relevant search terms from 1 January 2016 until 13 July 2022. The detailed search strategy is available upon reasonable request via info@sri-richtlijnen.nl. The systematic literature search resulted in 425 hits. Studies were selected based on the following criteria: systematic reviews, randomized controlled trials, or observational studies describing transmission or outbreaks in an outpatient clinic caused by VRE, CRE or Acinetobacter. Sixteen studies were initially selected based on title and abstract screening. After reading the full text, 12 studies were excluded (see the table with reasons for exclusion under the tab Methods), and 4 studies were included (Grabsch, 2006; Heinrichs, 2019; Kalocheretis, 2004; Klein, 2012).

Results

Four studies were included in the analysis of the literature. The summary of literature and results are included in Samenvatting literatuur.