Background
Traveling to high-prevalence countries (HPC) for multidrug-resistant organisms (MDRO), such as multidrug-resistant
Enterobacteriaceae and
Acinetobacter baumannii (MDRGN) and methicillin-resistant
Staphylococcus aureus (MRSA), as well as contact with the local healthcare-system in HPC have previously been identified as risk factors to for acquiring MDRO [
1‐
9]. UHF is located in the direct vicinity of the Frankfurt international airport which processes almost 60 million passengers per year [
10]. German residents returning after pre-treatment in healthcare-system in HPC, medical travelers seeking healthcare in Germany as well as travelers who experience medical problems during transit represent a relevant number of patients admitted to UHF.
Refugee patients’ countries of origin (COO) have previously been described as HPCs [
11‐
13]. Since 2015, refugee influx to Germany has been high [
12]. Concise hygiene management strategies therefore are required not only for refugees (REF) [
11], but also for patients admitted from health-care systems abroad (ABROAD). Since the first study addressing the prevalence of multidrug-resistant organisms (MDRO) in refugee patients in hospital settings has been published in January 2016 [
11], wide-ranging experiences in management of REF in European hospitals have been achieved. However, apart from a recent Dutch investigation [
13], further data on MDRO in REF are available only to limited degree. This study therefore addresses the epidemiology of MDRO in refugee patients compared to patients after pre-treatment in hospitals abroad admitted to UHF between December 2015 and June 2016. Furthermore, these data were compared to the epidemiology of MDRO of residents, without any history of fleeing or pre-treatment abroad, within the same period. These data are needed to highlight challenges in terms of hospital hygiene and the need for specific demands on infection control and hospital hygiene.
Results
Between December 2015 and June 2016, 117 REF and 84 ABROAD and 495 ICU were screened for MDRGN and MRSA by rectal and nasal swabs, respectively. The median age of REF, ABROAD and ICU was 19, 54 and 66 years, respectively. In the REF, the most frequently reported country of origin was Afghanistan (25.6%). The most frequently reported country ABROAD were admitted from was Turkey (13.1%). REF and ABROAD were most frequently admitted to the department of pediatrics (37.6%) and neurology (15.5%), respectively, as given in Table
1.
Table 1
Patient characteristics in REF, ABROAD and ICU
number of patients | 117 | 84 | 495 |
mean age; standard deviation (years) | 19; 14.7 | 54; 24.0 | 66; 14.4 |
male (n; %; 95%CI) | 85; 72.7; 63.6-80.5 | 59; 70.2; 59.3-79.7 | 341; 68.9; 64.6-72.9 |
COO a / admitted from (n; %) |
Afghanistan | 30; 25.6 | 1; 1.2 | residents without refugee history or pre-treatment abroad |
Syria | 29; 24.8 | - |
Somalia | 8; 6.8 | 3; 3.6 |
Algeria | 7; 5.9 | 1; 1.2 |
Iraq | 5; 4.3 | 1; 1.2 |
Pakistan | 5; 4.3 | - |
Eritrea | 5; 4.3 | 2; 2.4 |
Ethiopia | 3; 2.6 | 3; 3.6 |
Turkey | - | 11; 13.1 |
Greek | - | 4; 4.8 |
Italy | - | 4; 4.8 |
Morocco | - | 4; 4.8 |
Nigeria | - | 4; 4.8 |
India | - | 3; 3.6 |
Iran | - | 3; 3.6 |
Kuwait | - | 3; 3.6 |
Croatia | - | 2; 2.4 |
Ghana | - | 2; 2.4 |
Hungary | - | 2; 2.4 |
Saudi-Arabia | - | 2; 2.4 |
Spain | - | 2; 2.4 |
Sri Lanka | - | 2; 2.4 |
Thailand | - | 2; 2.4 |
Egypt | 2; 1.7 | 4; 4.8 |
Other | 8; 6.8 b
| 19; 22.6 c
|
Unknown | 15; 12.8 | - |
department patients admitted to (n; %; 95%CI) |
intensive care unit | - | - | 495; 100 |
Surgery | 13; 11.1; 0.6-18.3 | 12; 14.3; 7.6-23.6 | 495; 100; 99.3-100 |
general / vascular | 6; 5.1 1.9-10.8 | 4; 4.7; 1.3-11.7 | 153; 30.9; 26.8-35.2 |
thoracic | - | 2; 2.3; 0.2-8.3 | 273; 55.2; 50.7-59.6 |
traumatology | 7; 5.9; 2.4-11.9 | 6; 7.1; 2.7-14.9 | 69; 13.9; 11.0-17.3 |
Internal medicine | 20; 17.1; 10.8-25.2 | 39; 46.4; 35.5-57.6 | |
Infectious diseases | 10; 8.5; 4.1-15.2 | 9; 10.7; 5.0-19.4 | - |
Gastroenterology | 3; 2.6; 0.5-7.3 | 9; 10.7; 5.0-19.4 |
Pneumology | 2; 1.7; 0.2-6.0 | 5; 5.9; 1.9-13.3 |
Hematology / Oncology | - | 9; 10.7; 5.0-19.4 |
Cardiology | 3; 2.6; 0.5-7.3 | 7; 8.3; 3.4-16.4 |
Angiology | 1; 0.9; 0.0-4.7 | - |
Rheumatology | 1; 0.9; 0.0-4.7 | - |
Gynecology / Obstetrics | 10; 8.5; 4.1-15.2 | 3; 3.6; 0.7-10.1 | - |
Urology | 6; 5.1; 1.9-10.8 | 1; 1.2; 0.0-6.5 |
Pediatrics | 44; 37.6; 28.8-47.0 | 8; 9.5; 4.2-17.9 |
Neurology | 4; 3.4; 0.9-8.5 | 13; 15.5; 8.5-25.0 |
Psychiatrics | 7; 5.9; 2.4-11.9 | 1; 1.2; 0.0-6.5 |
otherd
| 12; 10.3; 5.4-17.2 | 7; 8.3; 3.4-16.4 |
The prevalence of at least one MDRGN in REF, ABROAD and ICU amounted to 52.1% (95% CI =42.7-61.5), 41.6% (31.0-52.9) and at a significantly lower prevalence 7.9% (5.6-10.6), respectively. Thereof, 0.9% (0.0-4.7) of REF, 15.5% (8.5-25.0) of ABROAD and 0% (0.0-0.7) of ICU, were positive for at least one MDRGN with additional carbapenem resistance (CR).
Six ABROAD were found positive for more than one MDRGN species with CR, resulting in a total number of 19 MDRGN with CR in ABROAD. Thereof, the most frequently detected MDRGN with CR in ABROAD was A. baumannii with 42.1% (20.3-66.5).
Interestingly, no significant difference was found in the overall prevalence of MDRGN between REF and ABROAD, however, the prevalence of MDRGN with CR in ABROAD (15.5%; 8.5-25.0) significantly exceeded the prevalence of MDRGN with CR in REF (0.9%; 0.0-4.7) by 17-fold (Table
2). The prevalence of at least one MDRGN with CR in both REF and ABROAD significantly exceeded the prevalence of MDRGN with CR in ICU (0%; 0.0-0.7). All carbapenemases detected are given in Table
2. Of the 117 REF tested, 61 were tested positive for at least one MDRGN, three individuals were found positive for two different MDRGN each and one individual was found positive for three different MDRGN, resulting in a total number of 66 MDRGN in REF. Of the 84 ABROAD thested, 35 patients were tested positive for at least one MDRGN, three individuals were positive for two different MDRGN each, two individuals were positive for three MDRGN each, and two individuals were positive for four different MDRGN each, resulting in a total number of 48 MDRGN in ABROAD. Of note, two individuals were tested positive for three different MDRGN with CR each:
n = 1 with
A. baumannii, K. pneumoniae and
E. coli, and
n = 1 with
A. baumannii, K. pneumoniae and
Proteus mirabilis. In contrast, all of 495 ICU were tested single-positive for MDRGN, resulting in 39 MDRGN in this group.
Table 2
Microbiological findings in REF, ABROAD and ICU
number of patients | 117 | 84 | 495 |
patients positive for at least one MDRGN (n;%; 95% CI) | 61; 52.1; 42.7-61.5 | 35; 41.6; 31.0-52.9 | 39; 7.9; 5.6-10.6 |
subgroup surgery | |
general / vascular | 4; 66.7; 22.2-96.6 | 1; 25; 0.6-80.6 | 22; 14.4; 9.2-20.9 |
thoracic | - | 1; 50; 12.6-98.7 | 14; 5.1; 2.8-8.5 |
traumatology | 3; 42.9; 9.9-81.6 | 2; 33.3; 4.3-77.7 | 3; 4.3; 0.9-12.2 |
patients positive for at least one MDRGN with CR (n;%; 95%CI) | 1; 0.9; 0.0-4.7 | 13; 15.5; 8.5-25.0 | 0; 0; 0.0-0.7 |
total number of MDRGN | 66 | 48 | 40 |
total number of MDRGN with CR | 1 | 19 | - |
Escherichia coli
| |
ESBL (n;%; 95CI) | 24; 20.5; 13.6-28.9 | 5; 5.9; 1.9-13.3 | 14; 2.8; 1.5-4.7 |
ESBL/FQ (n;%;95CI) | 34; 29.1; 21.0-38.2 | 17; 20.2; 12.3-30.4 | 19; 3.8; 2.3-5.9 |
ESBL/FQ + CR (n;%;95CI) | 1; 0.9; 0.0-4.7 | 2; 2.4; 0.3-8.3 | - |
carbapenemases (n) | none detected (1) | OXA-181 + NDM-5 (1) none detected (1) | - |
Klebsiella pneumoniae
| |
ESBL (n;%;95CI) | 3; 2.6; 0.5-7.3 | - | 1; 0.2; 0.0-1.1 |
ESBL/FQ (n;%;95CI) | 4; 3.4; 0.9-8.5 | 5; 5.9; 1.9-13.3 | 2; 0.4; 0.0-1.5 |
ESBL/FQ + CR (n;%;95CI) | - | 7; 8.3; 3.4-16.4 | - |
carbapenemases (n) | - | OXA-48 (2) OXA-181 (1) KPC-3 (1) KPC-9 (1) NDM-1 (1) none detected (1) | - |
other Enterobacteriaceae
| |
Ceph/FQ (n;%;95CI) | - | 2 a; 2.4; 0.3-8.3 | 4 b; 0.8; 0.2-2.1 |
Ceph/FQ + CR (n;%;95CI) | - | 1 c; 1.2; 0.0-6.5 | - |
carbapenemase (n) | - | none detected (1) | - |
Acinetobacter baumannii
| |
Ceph/FQ (n;%;95CI) Ceph/FQ + CR (n;%;95CI) | - | - | - |
- | 8; 9.5; 4.2-17.9 | - |
carbapenemase (n) d
| - | OXA-23 (4) OXA-24 (1) OXA-58 (1) NDM-1 + OXA-23 (1) OXA-23 + OXA-24 (1) | - |
Pseudomonas aeruginosa
| |
Pip/Ceph/FQ + CR (n;%;95CI) | - | 1; 1.2; 0.0-6.5 | - |
carbapenemase (n) | - | none detected (1) | - |
patients positive for MRSA (n;%; 95% CI) | 12; 10.3; 5.4-17.2 | 4; 4.8; 1.3-11.7 | 7; 1.4; 0.6-2.9 |
spa types (n) | t304 (2) t4892 (2) | | |
t044 (1) t131 (1) t311 (1) t325 (1) t386 (1) | t003 (1) t217 (1) t325 (1) t608 (1) | t020 (2) t003 (1) t223 (1) t309 (1) t668 (1) t3758 (1) |
t688 (1) t790 (1) t3422 (1) | | |
The most common MDRGN in REF, ABROAD and ICU was
E. coli expressing resistance due to ESBL and additional resistance to fluoroquinolones (Table
2), with 34/117 (29.1%; 21.0-38.2) in REF, 17/84 (20.2%; 12.3-30.4) in ABROAD, and, significantly lower, 19/495 (3.8%; 2.3-5.9) in ICU (Table
2). Furthermore, 10.3% of samples (5.4-17.2) in REF and 4.8% of samples (1.3-11.7) in ABROAD, respectively, were positive for MRSA. The prevalence of MRSA in ICU (1.4%; 0.6-2.9) was significantly lower compared to REF.
Spa types detected in REF, ABROAD and ICU are given in Table
2.
Furthermore, we found a low prevalence of MDRGN in the REF, ABROAD and ICU surgical group (Table
2). In particular, in the traumatology ICU group, the prevalence of MDRGN was lowest with 4.3% (0.9-12.2).
Discussion
The rapid global spread of MDRO is a serious global health risk and must direct attention to the development of effective strategies to prevent the spread of antibiotic resistances and life-threatening infections caused by MDRO. Diligent prevention strategies of MDRO transmission in hospital settings should therefore be the focus of preventive efforts. Traveling to HPC, medical tourism, contact to local health care systems as well as history of refugee status have previously been identified as risk factors for carrying MDRO [
1‐
11,
19,
20].
The objective of our study was to give a firm insight into the epidemiology of MDRO in refugee patients (REF), patients admitted from abroad (ABROAD) and German resident patients admitted to a intensive care unit at UHF (ICU). The major strength of our investigation is the direct comparison of these three groups as this issue for our knowledge has not been addressed in scientific literature so far.
No significant difference in the overall prevalence of MDRGN between REF and ABROAD was found, the prevalence of MDRGN with CR in ABROAD (15.5%), however, significantly exceeded the prevalence in REF (0.9%) by 17-fold (Table
2). This phenomenon might reflect an inherent risk to acquire highly resistant MDRGN in hospitals in HPC, from which the majority of ABROAD are admitted from to UHF, such as e.g. Turkey, Egypt, Ethiopia or Somalia (Table
1). For repatriates, this aspect has previously also been mentioned by
Josseaume et al. [
8]. Considering that conditions under which refugees make their way to Germany might have been poor, we hypothesize that refugees in good health condition are more likely to make the journey than refugees suffering from severe disease or disability and with a strong history of hospital treatment in their country of origin. We however cannot quantify the level of contact REF have had with their local health care system, which therefore might be a source of bias in our setting. In a future setting, it therefore might be interesting to evaluate the MDRO prevalence in refugees without any contact to their local health care system compared to refugees without.
Our data thus indicate, that ABROAD are more likely to represent the prevalence of highly resistant bacteria in hospitals abroad, whereas REF are more likely to portray the general prevalence of highly resistant bacteria in their COO. Furthermore, this finding is additionally underlined by previous findings showing a high prevalence of
A. baumannii with CR in hospitals in Syria and Iran [
21‐
24].
However, due to language barriers, data regarding e.g. exact previous medical treatment or length of stay in the hospitals abroad was available only in a minority of REF patients as well as a low number of ABROAD patients. This aspect could be interesting to evaluate the timeframe that MDRO might be acquired within in hospitals abroad. Moreover, exact phylogeny analysis of isolated resistant strains (e.g., via NGS techniques) was not performed which might have demonstrated a potential clonal spread of resistant pathogens in some particular countries of patients from ABROAD.
In the traumatology ICU group we found a MDRGN prevalence of 4.3% (0.9-12.2), which is in the range of general German population’s MDRGN prevalence [
25].
Concerning prevalence of MRSA, no significant difference between REF (10.3%; 5.4-17.2) and ABROAD (4.8%; 1.3-11.7) was observed, even though MRSA prevalence in ICU (1.4%; 0.6-2.9) was significantly lower than in REF (Table
2). This finding might reflect the limited space conditions in refugee accommodations: in parallel with e.g. an outbreak of measles in a refugee settlement in Calais, France, [
26], or clustering of shigellosis in refugees in Austria [
27]. The conditions refugees live in might also be conducive to the spread of MRSA within the community. Investigating the
spa types in the three cohorts, we found ICU patients harbouring
spa types known to be common in Germany, e.g. t003, t020, or t223 [
28,
29]. In contrast, many of the
spa types detected in REF (t131, t304, t325, t790, t4892; Table
2) have frequently been recorded from e.g. Iran, Jordan, United Arab Emirates or Lebanon [
28]. However, the number of MRSA detected in REF in this study (
n = 12) seems to be too low to assess any epidemiological trends.
These mentioned countries are in immediate neighbourhood to the most frequently recorded COO of REF in this study (e.g. Afghanistan and Syria, Table
1), which might indicate that REF bring new
spa types to Germany and Europe. For future settings, we suggest to perform genotyping on the core chromosomes, e.g. via multilocus sequence typing (MLST), whole genome sequencing or SCCmec typing for MRSA, and plasmid typing for more exact phylogenetic and epidemiological investigations.
Acknowledgements
Authors would like to declare that this report is exclusively based on epidemiological findings and is not influenced by any political opinion. Authors would furthermore like to emphasize that all patients are treated equally and in conditions of best medical care at UHF, regardless of their origin.