Prevalence of Device-associated Nosocomial Infections Caused By Gram-negative Bacteria in a Trauma Intensive Care Unit in Libya

Abstract

Objectives

Device-associated nosocomial infections (DANIs) have a major impact on patient morbidity and mortality. Our study aimed to determine the distribution rate of DANIs and causative agents and patterns of antibiotic resistance in the trauma-surgical intensive care unit (ICU). 

Methods

Our study was conducted at Abusalim Trauma Hospital in Tripoli, Libya. All devices associated with nosocomial infections, including central venous catheters (CVC), endotracheal tubes (ETT), Foley’s urinary catheters, chest tubes, nasogastric tubes (NGT), and tracheostomy tubes, were removed aseptically and examined for Gram-negative bacteria (GNB). 

Results

During a one-year study period, 363 patients were hospitalized; the overall mortality rate was 29%. A total of 79 DANIs were identified, the most common site of infection was ETT (39.2%), followed by urinary catheters (19%), NGTs (18%), tracheostomy tubes (11%), CVCs (10%), and chest tubes (3%). The most frequently isolated organisms were Klebsiella pneumonia, Acinetobacter baumannii, and Pseudomonas aeruginosa (30%, 20%, and 14%, respectively). Extremely high resistance rates were observed among GNB to ampicillin (99%), cefuroxime (95%), amoxicillin-clavulante (92%), and nitrofurantoin (91%). Lower levels of resistance were exhibited to amikacin (38%), imipenem (38%), and colistin (29%). About 39% of the isolates were defined as multi-drug resistant (MDR). Overall, extended spectrum β-lactmase producers were expressed in 39% of isolates mainly among K. pneumonia (88%). A. baumannii isolates exhibited extremely high levels of resistance to all antibiotics except colistin (100% sensitive). In addition, 56.3% of A. baumannii isolates were found to be MDR. P. aeruginosa isolates showed 46%–55% effectiveness to anti-pseudomonas antibiotics. 

Conclusion

High rates of DANI’s and the emergence of MDR organisms poses a serious threat to patients. There is a need to strengthen infection control within the ICU environment, introduce surveillance systems, and implement evidence-based preventive strategies.

Introduction

Infections acquired in intensive care units (ICUs) are a major health concern worldwide, particularly in developing countries.¹ Patients admitted to ICUs are at an increased risk for acquiring device-associated nosocomial infections (DANIs) because of their debilitated immune systems and exposure to invasive devices.² Device use is recognized as creating a high risk of catheter-associated urinary tract infections (CAUTIs), central line-associated bloodstream infections (CLABSIs), and ventilator-associated pneumonia (VAP) in ICU patients.^3,4 They are considered a principal threat to patient safety and are among the main causes of morbidity, mortality, and increased costs.^5-7

Data using standardized international case definitions and methodology on DANIs rates from different ICUs worldwide are published through the International Nosocomial Infection Control Consortium (INICC) network. Analysis of the data showed that DANI rates were increased three- to five-fold in ICUs in developing countries compared to the United States.^2,8

There is limited data available for countries in the Eastern Mediterranean region using standardized international case definitions.⁹ The Centers for Disease Control (CDC) has implemented the National Nosocomial Infection Surveillance System (NNISS) to standardize hospital associated infection (HAI) surveillance.¹⁰

DANIs are a serious cause of patient attributable mortality in developing countries.¹¹ Gram-negative bacteria (GNB) are a common cause of sepsis, pneumonia, and urinary tract infections in ICU patients.^11-13 Antibiotic resistance among GNB are increasing continuously^14,15 and surveillance studies across the world have demonstrated an increase in resistance especially in critically ill patients.¹⁶

The aim of our study was to determine the distribution of DANIs and the microbiological and antibiotic resistance profiles of the infecting pathogens. To the best of our knowledge, this is the first study of its kind in Libya.

Methods

Between January and December 2014, we evaluated patients who developed DANIs in the trauma/surgical intensive care unit (SICU) at Abusalim Trauma Hospital, Tripoli, Libya, which provides care for both ventilated and non-ventilated patients. The hospital is a 500-bed tertiary referral and teaching hospital and the ICU comprises of two rooms with a total of 10 beds and no barriers between patients. There are only two single isolation rooms.

A nosocomial infection was defined as an infection that developed 48 hours after admission and up to 10 days after discharge from hospital (in accordance with the CDC criteria).^2,10,17 Data were obtained from ICU daily reports, microbiology laboratory reports, and patient medical charts.

All devices associated with nosocomial infections, including central venous catheters (CVCs), endotracheal tubes (ETTs), Foley’s urinary catheters, chest tubes, nasogastric tubes (NGT), and tracheostomy tubes were removed aseptically and the distal 5cm of the devices was removed and cultured using a standardized method.¹⁸ Isolated organisms were identified to the species level and tested for their susceptibility to a variety of antimicrobial agents using the BD Phoenix Automated Microbiology System (BD Diagnostic Systems, Sparks, Md, USA) and employed a Gram-negative panel, which was inoculated according to the manufacturer’s instructions. Reading and interpretation of panels was also performed according to the manufacturer’s instructions. Standardized methods were used to detect extended-spectrum ß-lactamase (ESBL) producers.¹⁹ All specimens were collected under approved ethical standards. This study was approved by the administration. Quality control was performed by testing Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, and Klebsiella pneumoniae ATCC 700603.

Only consecutive, non-duplicate, and GNB were included in the study. Multidrug resistant bacteria (MDR) were defined as showing resistance to three different classes of antibiotics such as fluoroquinolones, aminoglycosides, and cephalosporins.²⁰ Infections occurring at more than one site due to different organisms in the same patient were reported as separate infection events.

The clinical outcome was assessed until hospital discharge or death. Patients were considered to have appropriate antimicrobial therapy if the isolated organism was susceptible to a previous antimicrobial regimen within the first 24 hours of clinical infection.

Results

During the one-year study period, 363 patients were hospitalized in the ICU. The mortality rate was 29%, and 30% of discharged patients (serious cases) were transferred abroad for advanced medical treatment. A total of 79 DANIs were identified during the study period; their distribution is shown in Table 1.

Table 1. Distribution of DANIs and related microorganisms.

Microorganisms	n (%)	Devices n (%)
		ETT 31 (39)	CAUTI 15 (19)	NGT 14 (18)	Tracheostomy 9 (11)	CVC 8 (10)	Chest tube 2 (3)
K. pneumonia	24 (30)	8	5	5	3	2	1
A. baumanii	16 (20)	6	6	3	0	1	0
P. aeroginosa	1 (14)	5	1	1	4	0	0
Pro. mirabilis	6 (8)	1	0	3	0	2	0
E. cloacae	4 (5)	4	0	0	0	0	0
Ser. marcescens	4 (5)	3	0	0	0	0	1
K. ozaenae	4 (5)	0	1	0	2	1	0
Other	10 (13)	4	2	2	0	2	0

ETT: endotracheal tubes; CAUTI: catheter-associated urinary tract infection; NGT: nasogastric tubes; CVC: central venous catheters.

VAP, CLABSI, and CAUTI were studied. The leading DANI was ETT (39%), followed by Foley’s urinary catheters (19%), NGT (18%) tracheostomy tubes (11%), CVCs (10%), and chest tubes (3%). The most frequently isolated organisms were K. pneumonia, A. baumannii, and P. aeroginosa (30%, 20%, and 14%, respectively). Other isolates included: Serattia marcescens, E. coli, Providencia sturatii, and Enterobacter aerogenes. All E. cloacae isolates were associated in ETT. Similarly Proteus mirabilis were mainly isolated from NGTs.

There was a variable degree of resistance against commonly used antibiotics. Extremely high resistance rates were observed to ampicillin (99%), cephalothin (95%), cefuroxime (95%), amoxicillin-clavulante (92%), and nitrofurantoin (91%). High levels of resistance were detected in third generation cephalosporines ceftriaxone (85%) and ceftazidime (73%), cefepime (70%), and fluoroquinolones ciprofloxacin (73%) and levofloxacin (73%). High rates of resistance were also demonstrated to cefoxitin (84%), ertapenem (79%), azteronam (77%), gentamicin (72%), pipracillin-tazobactam (66%), trimethoprim-sulfamethoxazole (56%), and meropenem (49%). On the other hand, lower levels of resistance were exhibited to amikacin (38%), imipenem (38%), and colistin (29%). Overall MDR were detected in 39% of isolates. ESBL producers were expressed in 39% of isolates mainly K. pneumonia (88%) and K. ozaenae (100%). The pattern of resistance of the three most commonly isolated GNB, K. pneumonia, A. baumannii, and P. aeruginosa, can be seen in Table 2.

Table 2. Percentage of resistance rates of major isolates associated with DANIs.

Antibiotic	Organisms (percentage resistance)
	K. pn. n=24	K. ox. n=4	A. ba. n=16	P. ae. n=11	E. cl. n=4	Pr. mi. n=6	Se. ma. n=4	Others n=10
Amikacin	25	0	81	36	0	17	0	50
Gentamicin	79	100	100	64	0	50	75	50
Ertapenem	92	100	94	100	25	17	25	60
Imipenem	17	25	94	36	0	33	0	40
Meropenem	50	75	94	46	0	17	0	30
Cephalothin	92	0	100	100	100	83	100	100
Cefuroxim	92	100	100	100	100	67	100	90
Cefoxitin	70	75	100	100	100	33	100	50
Ceftazidim	92	100	100	55	0	17	25	80
Ceftriaxon	92	100	100	100	0	67	75	70
Cefepime	83	100	100	46	25	33	25	50
Azitronam	92	100	100	64	0	17	75	60
Ampicillin	100	100	100	100	100	83	100	100
Amoxicillin/clavulanic	92	100	100	100	100	50	100	90
Piperacillin/tazobactam	88	100	100	46	0	0	0	40
Colistin	25	0	0	0	25	100	100	50
Trimethoprim-sulfamethoxazole	59	0	75	64	0	67	0	60
Nitrofuranton	83	100	100	100	50	83	100	60
Ciprofloxacin	83	100	100	55	25	67	25	70
Levofloxacin	79	100	94	55	25	67	0	70

K. pn.: K. pneumonia; K. oz.: K. ozaenae; A. ba.: A. baumannii; P. ae.: P. aeruginosa; E. cl.: E. cloacae; P. mi.: Prot. mirabilis; S. ma.: Ser. Marcescens.

A. baumannii isolates exhibited extremely high level of resistance to all tested antibiotics (75–100%), except colistin (100% sensitive). Out of 16 isolates, nine (56%) were found to be MDR. Resistance of K. pneumonia isolates to imipenem (17%) and amikacin (25%) antibiotics were the least compared to the majority of other tested antibiotics (79%–100%). ESBL producers were detected in 21 out of 24 (88%) isolates of K. pneumonia. The anti-pseudomonas antibiotics, which include cefepime, pipracillin-tazobactam, and ceftazidime showed effectiveness to P. aeruginosa isolates (46%, 46%, and 55%, respectively). Only four of 11 (36%) P. aeruginosa isolates were considered MDR.

Discussion

Infection control measures that include intensive surveillance can reduce the incidence of nosocomial infections. Therefore, standards of institutional HAI surveillance and infection control have been adopted in developed countries.^16,21

The overall mortality was 29% due to the fact that the majority of patients suffered gunshot wounds, blasts and blunts. The mortality rate of our trauma/SICU patients was higher than Thailand (20%).²² In Ethiopia, the mortality rate among patients (aged 21–30 years old) admitted to SICU was 9%.²³ A meta-analysis reported lower trauma mortality rate (13%) than that found in our study.²⁴ The incidence and mortality rates of VAP vary due to several factors, such as the study population, time of onset, causative organisms, and appropriate antibiotic therapy.²⁵ Our patients were ventilated for the majority of their time in the ICU. However, we were unable to determine a relationship between nosocomial infections of ventilated and non-ventilated patients. Our ventilated trauma patients had higher rates of ETT (39%) compared to other device infections. This finding was in keeping with the higher incidence rate in trauma patients reported by the CDC.¹⁶ Trauma patients had a higher occurrence of VAP (63%) than non-trauma patients (38%). The most frequently detected DANIs for ETT was K. pneumonia (26%) followed by A. baumannii (19%) and P. aeruginosa (16%), these isolates are commonly isolated from VAP infection. Similar results were found in previous studies.^12,26-29 A recent study showed that the VAP incidence of mixed medical-SICUs in Thailand, GNB were the most common organisms particularly A. baumannii, P. aeruginosa, and K. pneumoniae, which were comparable with our results.³⁰

It is estimated that 80,000 infections related to CVC occur in ICU patients each year and these infections are associated with a mortality rate as high as 25%.³¹ In a study carried out in Brazil, 6% of admitted patients suffered from CVC infection.³² K. pneumoniae, Acinetobacter, and P. aeruginosa were the three most common organisms associated with CVC infection in SICU.^33,34 In this study, 10% of DANIs were related to CLABSI and K. pneumonia accounted for 25%.

Patients on NGT might be predisposed to colonization by Proteus and Pseudomonas spp. within 48–72 hours.³⁵ Similarly, our study noted that Prot. mirabilis was the main encountered organism along with Klebsiella, Acienetobacter and Pseudomonas. The main isolates associated with CAUTI were Acinetobacter (40%) and Klebsiella (33%). Pseudomonas was the predominant organism in a SICU in Turkey.²⁸

Extremely high resistance rates to ampicillin, cephalothin, cefuroxime, amoxicillin-clavulante, and nitrofurantoin (91.1%–99.0%) were observed. Lower levels of resistance were exhibited to amikacin, imipenem, and colistin (29%–38%), but amikacin was broadly active; 39% of the isolates were defined as MDR. Overall ESBL producers were expressed in 39% of isolates mainly among K. pneumonia (88%) and K. ozaenae (100%) isolates.

Our results show that the high level of resistance was quite problematic in the ICU. The high resistance rates might be associated with antibiotic abuse and prolonged ICU stays.³⁶ A. baumannii isolates exhibited extremely high levels of resistance to all antibiotics (75–100%) with the exception of colistin. In addition, 56% of A. baumannii isolates were considered MDR. Our data support previous results that indicated an alarming pattern of antibiotic resistance to A. baumannii, particularly in the ICU setting, and is frequently associated with outbreaks of VAP.³⁷ The high rate of resistance to ESBL producers in K. pneumonia (88%) was similar to that in other developing countries reported by the INICC, where the rate of ceftazidime resistance was 92%.^2,38 Both of those rates were much higher than those reported by the National Healthcare Safety Network (NHSN) in the US (6%).³⁹ Resistance of P. aeruginosa to imipenem was 36% similar to that reported by the INICC (37%) and lower than Egypt (56%).^2,38 The anti-pseudomonas antibiotics (cefepime, pipracillin-tazobactam, and ceftazidime) showed effectiveness to P. aeruginosa isolates (46%, 46%, and 55%, respectively). Only four of 11 (36%) P. aeruginosa isolates were considered as MDR. Similar results were demonstrated in other ICU settings.^28,40 The high rates of antimicrobial resistance identified in the present study might be attributed to the lack of antibiotic use policies and guidelines in the majority of hospitals in Libya. As in most developing countries, administrative and financial support is lacking, which results in limited funds and resource availability to deal with infection control. In this ICU setting, guidelines on specific infection control practices are not adhered to adequately, infection control surveillance is not conducted, and hospital accreditation is not mandatory.

Our study had several limitations including that the data was not enough to reflect the whole country and different types of ICUs. Additionally, DANI rates per 1,000 device-days and DANI’s mortality rates were not demonstrated.

Conclusion

High rates of DANI’s and the emergence of MDR organisms poses a threat to physicians and patients alike. With limited therapeutic options, there is a need to strengthening infection control measures within the ICU, introduce surveillance systems, and implement evidence-based preventive strategies against nosocomial infections.

Disclosure

The authors declared no conflicts of interest. No funding was received for this work.