Incidence of Device Associated-Healthcare Associated Infections from a Neurosurgical Intensive Care Unit of a Tertiary Care Center: A Retrospective Analysis.

BACKGROUND: Deviceassociated infections (DAIs) increase the morbidity and mortality in the intensive care unit (ICU). Studies from the neurosurgical ICU in developing countries are sparse. AIMS: The aim of this study was to assess the incidence of device-associated healthcare associated infections, pathogens isolated, antibiotic resistance, and mortality in neurosurgical ICU. SETTINGS AND
DESIGN: A retrospective study was conducted in the neurosurgical ICU of a tertiary care center.
MATERIALS AND METHODS: This study was done by analyzing data of patients admitted in a neurosurgical ICU with one or more devices during the period from January 2011 to July 2017. STATISTICAL ANALYSIS: Quantitative variables were expressed as mean and standard deviation; qualitative variables were expressed as frequency and percentage.
RESULTS: During this period, 6788 patients with devices were admitted in the ICU, and 316 patients developed DAI. Two hundred and forty-eight patients had catheter-associated urinary tract infection (CAUTI), 78 had ventilator-associated pneumonia (VAP), and 53 had central line-associated bloodstream infection (CLABSI). The incidence rate for CAUTI was 17.83, VAP - 16.83, and CLABSI - 4.39 per 1000 device days. The device utilization ratio was highest for urinary catheter - 0.76, followed by central line - 0.66 and ventilator - 0.25. Predominant pathogens were Klebsiella - 90, Escherichia coli - 77, Pseudomonas - 40, Candida - 39, Acinetobacter - 30, and Enterobacter - 21. Carbapenem resistance was found in Acinetobacter (73.4%), Pseudomonas (45%), and Enterobacter (38%). S. aureus isolated in six cases; four being MRSA (66.7%). Multidrug resistance was found in Acinetobacter (80%), Pseudomonas (60%), Enterobacter (52.3%), Klebsiella (42.3%), and E. coli (33.7%). No colistin resistant Gram negative bacilli or vancomycin resistant enterococci were isolated. During this period 124 patients with DAI died, of which 52 patients had sepsis. The crude mortality rate was 1.83%.
CONCLUSION: The DAI with the highest incidence was CAUTI, followed by VAP and CLABSI. With the implementation of insertion bundles and adherence to aseptic precautions, the DAI rate had come down. Copyright:

INTRODUCTION

Nosocomial infection is defined as a localized or systemic reaction caused by an infectious agent or toxin that was not present or incubating at the time of hospital admission.[1]

Device-associated infections (DAI) such as ventilator-associated pneumonia (VAP), central line-associated bloodstream infections (CLABSI), and catheter-associated urinary tract infections (CAUTI) are a major threat to patient survival in the intensive care unit (ICU). The Centre for Disease Controls (CDC) National Nosocomial Infection Surveillance (NNIS) has laid down definitions for surveillance of device-associated healthcare associated infections (DA-HAI).[1]

In general, most cases of infection in critically ill patients are associated with medical devices. In the study by Richard et al.,[2] 95% of cases of urinary tract infection were catheter related, 87% of cases of bloodstream infection originated from an indwelling vascular catheter, and 86% of cases of pneumonia were associated with mechanical ventilation.

Majority of catheter-related urinary tract infection involve only the lower urinary tract and have a low attributable mortality rate (<5%), whereas 5%–25% of patients may not survive an episode of vascular catheter-related bloodstream infection.[3]

Nosocomial infections are a major threat to patient safety. DAI increases the morbidity and mortality, delays discharge from the hospital, and increases the financial burden of the patient. Most of the studies on DA-HAI are from developed countries. Studies on DAI from neurosurgical ICU in developing countries are sparse to date. Hence, this retrospective study was conducted from a neurosurgical ICU of a tertiary care center.

The aim of this study was to assess the incidence of device-associated health care associated infections (DA-HAI), patterns of pathogens isolated, antibiotic resistance, and mortality due to DAI in a neurosurgical ICU of a tertiary care center.

MATERIALS AND METHODS

Setting

The study was conducted by analyzing the medical records of a total of 6788 patients admitted to the neurosurgical ICU of a tertiary care referral center after getting the approval from the Hospital Ethics Committee (IEC-AIMS-2018-ANES115). There are 14 beds in the neurosurgical ICU, of which 10 beds are dedicated to neurosurgery patients and 4 beds to orthopedic patients.

Study design and data collection

A retrospective analysis of 6788 patients admitted in the neurosurgical ICU with one or more devices (central line, ventilator, and urinary catheter) during the period from January 2011 to July 2017 was carried out. CDC NNIS definition was used for surveillance of DA-HAI. VAP, CLABSI, and CAUTI were included as DA-HAI.

The incidence of DA-HAI is calculated as the rate of DAIs per 1000 device days (incidence/1000 device days = number of DAI due to a device/total number of days with that device × 1000)

Device utilization ratio is calculated by dividing the total device days by total patient days multiplied by 100 (DU ratio: total number of device days/total patient days × 100).

Neurosurgical patients admitted in the ICU during the period of January 2011 to July 2017, having one or more of the above-mentioned devices, and coming under the definition of CDS NNIS for DA-HAI were noted. The type of surgeries patients underwent, pathogens isolated, antibiotic susceptibility, and mortality were taken from the medical records of the patient.

Sampling

Samples for cultures from the respective devises were taken from patients who had symptoms and/or signs of infection.

Culture techniques

For VAP: Lower respiratory tract secretions were collected either by tracheal aspiration and/or by bronchoalveolar lavage. For CLABSIs: Central venous catheters were removed aseptically, and the distal 5 cm of the catheter was cut and cultured using a standardized semi-quantitative method. Concomitant blood cultures were drawn percutaneously in all cases. For CAUTI: Urine sample was aseptically aspirated from the sampling port of the urinary catheter and was cultured semiquantitatively. In all cases, standard laboratory methods were used to identify microorganisms and standardized antibiotic susceptibility testing was performed.

There is a dedicated infection control department in the institute. The infection control nurses daily record the number of patients in the ICU with devices. If any patient shows clinical features of infection, the sample from the respective devices was taken for culture and sensitivity as mentioned earlier. From the records, the number of infections acquired with each device, device days, and total patient days were calculated [Tables 1-3]. DAI rates per 1000 device days and device utilization ratio were derived from this. The microorganisms isolated, susceptibility to antibiotics, and mortality details were collected from the patient's medical records.

Table 1

Device associated infection rate and device utilization ratio for catheter-associated urinary tract infection

Year	Pt Days	U.Cath Days	UTI Rate	No of UTI	U.Cath DU Ratio
2011	3120	2190	13.70	30	0.71
2012	2607	1984	35.28	70	0.76
2013	2852	2144	27.05	58	0.75
2014	2461	1942	11.84	23	0.79
2015	2686	2137	18.25	39	0.8
2016	2783	2111	9.00	19	0.76
2017	1775	1402	6.42	9	0.79
Total	18254	13910	17.83	248	0.76

Table 3

Device associated infection rate and device utilization ratio for central line-associated blood stream infection

Year	Pt Days	CL Days	BSI Rate	No of BSI	CL DU Ratio
2011	3120	2353	2.12	5	0.75
2012	2607	2274	6.59	15	0.87
2013	2852	2475	5.65	14	0.87
2014	2461	1845	1.62	3	0.74
2015	2686	1132	4.41	5	0.42
2016	2783	1242	7.25	9	0.45
2017	1775	745	2.68	2	0.42
Total	18254	12066	4.39	53	0.66

Sample size

Based on the proportion of CAUTI (19%) observed in an earlier publication[4] and with 95% confidence and 10% allowable error, the minimum sample size calculated was 1612.

Statistics

The statistical analysis was done using the IBM SPSS version 20.0 software (IBM SPSS Inc, Chicago, USA). Quantitative variables were expressed as mean and standard deviation. Qualitative variables were expressed as frequency and percentage. For classification of the different causative pathogens associated with DAIs, all the microorganisms isolated from culture were recorded and their relative frequency of isolation was determined as percentage. Incidence and resistance were expressed in terms of frequency and percentage.

RESULTS

All neurosurgical cases admitted with the three devices (central venous line, ventilator, and urinary catheter) were included in this retrospective study.

Neurosurgical ICU admits both preoperative and postoperative patients. Most of the preoperative cases were trauma patients with either head or spine injury. Craniotomy cases for tumor surgeries accounted for the highest number of patients with DAI (80), followed by decompressive craniectomy (44) and aneurysm clipping (38).

During the study period, 6788 patients were admitted in the neurosurgical ICU with one or more of the above devices, of which 316 patients developed DAI. There were 193 (61.1%) male and 123 (38.9%) female patients. The age of the patients varied between 9 months and 89 years. The median age was 53 years (interquartile range 30.75). The total patient days with devices were 18,254. The number of device days was highest for urinary catheter (13,910), followed by central line (12,066), and ventilator (4634). During this period, the total number of CAUTI was 248, VAP was 78, and CLABSI was 53. The incidence rate calculated for CAUTI/VAP/CLABSI per 1000 device days was 17.83/16.83/4.39, respectively. Device utilization ratio was highest for urinary catheter (0.76), followed by central line (0.66) and ventilator (0.25) [Tables 1-3].

Predominant organisms isolated from CAUTI were Escherichia coli (27.75%), Klebsiella species (18.06%), Candida species (17.18%), Enterococcus species (11.89%), and Pseudomonas species (7.9%) [Table 4]. Organisms such as Morganella morganii, Citrobacter freundii, Burkholderia cepacia, Sphingomonas paucimobilis, and Proteus vulgaris were also isolated from few patients. Predominant organisms isolated from VAP were Klebsiella pneumoniae (36.17%), Pseudomonas aeruginosa (19.14%), Acinetobacter species (17.02%), and E. coli (8.51%) [Table 5]. There were two VAP cases with Candida species isolated. Predominant organisms isolated from CLABSI were K. pneumoniae (31.2%), Acinetobacter species (18.7%), and E. coli (12.5%) [Table 6].

Table 4

Predominant organisms isolated from catheterassociated urinary tract infection

Organism	Frequency	Percentage (%)
Escherichia coli	63	27.75
Klebsiella species	41	18.06
Candida species	39	17.18
Enterococcus species	27	11.89
Pseudomonas species	18	7.9
Enterobacter species	16	7.04
Acinetobacter species	5	2.2
Miscellaneous organisms	18	7.9
Total	227

Table 5

Predominant organisms isolated from ventilatorassociated pneumonia

Organism	Frequency	Percentage (%)
Klebsiella pneumoniae	34	36.17
Pseudomonas aeruginosa	18	19.14
Acinetobacter species	16	17.02
Escherichia coli	8	8.51
Enterobacter species	4	4.25
Staphylococcus aureus	4	4.25
Miscellaneous organisms	10	10.63
Total	94

Table 6

Predominant organisms isolated from central line-associated blood stream infection

Organism	Frequency	Percentage (%)
Klebsiella pneumoniae	15	31.2
Acinetobacter species	9	18.7
Escherichia coli	6	12.5
Miscellaneous organisms	18	37.5
Total	48

In this study, the common pathogens isolated from DA-HAI patients were Klebsiella species (90), E. coli (77), Pseudomonas species (40), Candida species (39), Acinetobacter species (30), and Enterobacter species (21) [Table 7].

Table 7

Distribution of predominant Gram-negative bacilli isolated in device-associated healthcare-associated infections and pattern of antibiotic resistance

Pathogens identified	Number of cases	Carbapenem Resistant(%)
Klebsiella spp.	90	23 (25.6)
E. coli	77	7 (9)
Pseudomonas spp.	40	18 (45)
Acinetobacter spp.	30	22 (73.4)
Enterobacter spp.	21	8 (38.1)
Pathogens identified	Number of cases	Multidrug Resistant (%)
Klebsiella spp.	90	38 (42.3)
E. coli	77	26 (33.7)
Pseudomonas spp.	40	24 (60)
Acinetobacter spp.	30	24 (80)
Enterobacter spp.	21	11 (52.3)

Carbapenem resistance was found in 73.4% of Acinetobacter spp., 45% of Pseudomonas spp., and 38% of Enterobacter spp. [Table 7 and Figure 1]. Multidrug resistance (MDR) was seen in 80% of Acinetobacter spp., 60% of Pseudomonas spp., 52.3% of Enterobacter spp., 42.3% of Klebsiella spp., and 33.7% of E. coli [Table 7 and Figure 2]. None of the isolates were colistin resistant. Of the six Staphylococcus aureus isolated, four were methicillin-resistant S. aureus (MRSA) (66.7%). There were no vancomycin-resistant enterococci isolated [Table 8].

Figure 1

Distribution of predominant Gram-negative bacilli isolated from device-associated healthcare-associated infections and pattern of carbapenem resistance

Figure 2

Distribution of predominant Gram-negative bacilli and pattern of multidrug resistance

Table 8

Distribution of Gram-positive Cocci isolated in device-associated healthcare-associated infections and pattern of antibiotic resistance

Pathogens identified	Number of cases	Resistance pattern	Resistant (%)
Enterococcus spp.	29	VRE	0
Staphylococcus aureus	6	MRSA	4 (66.7)

During the study period, 139 patients died in the neurosurgical ICU, of which 124 had DAI. Fifty-two patients had evidence of sepsis of which 32 cases had CAUTI, 27 cases had VAP, and 16 cases had CLABSI [Table 9]. Among patients with sepsis, 23 were infected with multidrug-resistant organisms and 4 were infected with MRSA. The crude mortality rate of patients with devices was 1.8% and that related to sepsis was 0.76%. Deaths due to infection were highest in 2012 (16), after which the rates declined, and in 2017 till July, there was only one death related to sepsis. It was also noted that many a times, the same patient had more than one DAI during the same time period.

Table 9

Distribution of deaths among device-associated healthcare-associated infections due to sepsis

Year	Death	CA UTI	CLABSI	VAP
2011	9	6	4	4
2012	16	7	5	11
2013	6	5	3	6
2014	9	6	1	3
2015	6	4	1	1
2016	5	3	1	1
2017	1	1	1	1
Total	52	32	16	27

DISCUSSION

Diagnosis of infection is a challenge in the neuro-ICU population as fever is highly prevalent with the incidence in the 1st week of hospitalization being as high as 87%.[5] In a prospective study of fever in neuro-ICU patients, nearly half were noninfectious in etiology.[6] Acute brain injury leads to an inflammatory response and is influenced by patient age, gender, genetics, mechanism and degree of injury, and therapeutic interventions. Tumor patients have compromised immune response and many of these patients may be on steroids. Neurosurgical patients are hence highly susceptible to DAI. Various methods including infection prevention bundles and standardization of care have been initiated to prevent DAI.

As per CDC-NNIS System reports, in the US, pooled mean rates of DAI for CAUTI were 3.9 per 1000 urinary catheter days, VAP – 5.4 per 1000 mechanical ventilated days, and CLABSI – 4.0 per 1000 CVC days.[7] Compared to the above study, we had high rates of UTI (17.83/3.9) and VAP (16.83/5.4) but slightly comparable rate of CLABSI (4.9/4). With the implementation of appropriate insertion and maintenance bundles, our infection rates have come down. The DAI rates are still very high in developing countries. Studies from Colombia[8] showed very high DA-HAI rates per 1000 device days – The CAUTI/VAP/CLABSI rates were 20.3/32.3/47.4.

Device utilization varies in the ICUs. Our analysis showed a lower utilization of mechanical ventilation (0.25 vs. 0.45) and almost similar utilization of central line (0.66 vs. 0.59) and urinary catheter (0.76 vs. 0.76) compared to the National Healthcare Safety Network (NHSN) of US. The longer the device is in place the greater the chance of getting infections. The device utilization is high in head injury, polytrauma, and brain surgeries but comparatively less in elective spine cases.

The risk of CAUTI is high in the neuro-ICU patients due to the need for catheter placement and maintenance in patients who sustained brain or spinal cord injuries and/or paraplegics and in patients with cerebrovascular disease.[9] Puri et al.[10] reported the prevalence of CAUTI in neurosurgical and neurology patients to be around 8%–10%. After the implementation of the urinary catheter insertion and maintenance bundle [Figure 3], CAUTI rate in our institute had steadily dropped from 35.28 in 2012 to 9.00 in 2016 [Table 1].

Figure 3

Urinary catheter insertion bundle checklist and follow-up

Incidence of VAP varies between 9% and 27% of all mechanically ventilated patients, with the highest risk being early in the course of hospitalization. VAP is classified as early onset (within 96 h) or late onset (after 96 h of ventilation); the latter may be due to multidrug-resistant pathogens. The risk factors for VAP in neurosurgical ICU are prolonged sedation/unconsciousness, delayed enteral feeding, and raised intracranial hypertension. Head injury patients have a high risk of developing VAP.[11] Early tracheostomy and early extubation reduces the duration of mechanical ventilation and the risk of VAP.[121314] Following the implementation of the VAP bundle [Figure 4], the incidence rate of VAP in our ICU had steadily declined from 30.01/1000 device days in 2012 to 3.85/1000 device days in 2016 [Table 2].

Figure 4

Intubation bundle checklist and follow-up

Table 2

Device associated infection rate and device utilization ratio for ventilator-associated pneumonia

Year	Pt Days	Vent Days	VAP Rate	No Of VAP	Vent DU Ratio
2011	3120	798	16.29	13	0.26
2012	2607	933	30.01	28	0.36
2013	2852	972	18.52	18	0.34
2014	2461	511	15.66	8	0.21
2015	2686	543	12.89	7	0.2
2016	2783	520	3.85	2	0.19
2017	1775	357	5.60	2	0.2
Total	18254	4634	16.83	78	0.25

Airway intubation is associated with colonization of the respiratory tract by gram-negative bacteria. Studies showed a higher incidence of P. aeruginosa and Acinetobacter spp. among patients who have previously received antibiotics, compared with patients who have not received antibiotics.[15] K. pneumonia, Acinetobacter species, and Pseudomonas are responsible for most of the cases of VAP. In our ICU also, the predominant organisms isolated from VAP patients were K. pneumoniae (43.59%) and P. aeruginosa (23.07%).

The risk of CLABSI varies with the site of catheterization; femoral central venous catheters are associated with the highest risk followed by the internal jugular and subclavian catheters. CLABSI can occur either due to the migration of the organisms harboring near the exit site of the catheter to the intravascular space or due to hematogenous spread of bacteria from a contaminated infusate. The International Nosocomial Infection Control Consortium surveillance data reported a CLABSI rate of 4.1 per 1000 central line days. The risk of CLABSI varies with the insertion site, catheter type, conditions of insertion, and catheter care. Host factors such as extremes of age, immunosuppressed states, and chronic illness also increase the incidence of CLABSI. Our study shows a decline in the incidence of CLABSI with the introduction of insertion and maintenance bundle [Figure 5] from 6.29 (15 patients) in 2012 to 2.68 (2 patients) in 2017 (up to July) [Table 3].

Figure 5

Central line insertion bundle checklist and follow-up

The data of NHSN[16] from January 2006 to October 2007 showed that Gram-positive organisms were the most common pathogen in CLABSI, followed by Gram-negative organisms and Candida species. The pathogens isolated from CLABSI in our ICU were predominantly Gram-negative-K. pneumoniae (31.2%), Acinetobacter (18.7%), and E. coli (12.5%).

Studies from a trauma care ICU by Sharma et al.[17] showed that Candida (29%) was the predominant pathogen followed by Acinetobacter spp. (23%), Pseudomonas spp. (21%), Klebsiella spp. (10%), and E. coli (4%). In our ICU, we found a higher rate of Klebsiella species (28.48%) and E. coli (24.37%), and a lower rate of Pseudomonas species (12.65%) and Candida species (12.34%).

About 66.7% of the S. aureus isolated in our ICU were methicillin resistant. In a similar study by Sharma et al., it was found that 58% of S. aureus were methicillin resistant.[17] In the study by Khan et al.,[18] all the Acinetobacter baumannii, more than 80% K. pneumoniae and E. coli, and more than 70% of P. aeruginosa isolated were susceptible only to colistin and tigecycline.

Our data also shows the increased rates of Carbapenem resistant and MDR organisms in neurosurgical ICU. There were no vancomycin-resistant enterococci isolated in this study, which is a notable observation. The only ray of hope is the fact that none of the isolates were colistin resistant. Drug resistance is a key area to be considered by every critical care physician as the pipeline for new antibiotics is relatively dry. Multiple factors such as inappropriate antimicrobial use, hygiene policy, and virulence and adaptation characteristics of pathogens are important factors in the emergence of drug resistance. However, the most reliable factor to curb drug resistance is to limit and avoid the inappropriate use of antimicrobials.[17] Strategies suggested by various healthcare authorities both national and international should be tailored to ensure feasibility across centers with varying practices and patterns of antimicrobial resistance.[19]

Mortality in a neurosurgical ICU is tied to various factors, of which an infective process may contribute to the rapid decline of patients. The crude mortality in the ICU was attributable to geriatric patients, patients with severe neurological insults, and patients with comorbidities, resulting in decompensation and deterioration. Transfer-in to the ICU after unexpected patient events in lower dependency settings also contributed to the mortality of the ICU.[18] The crude ICU mortality rate is generally higher in patients with DAI and varies with different devices. Iordanou et al. found a crude mortality rate of 40% in patients who acquired HA-DAI and 17.9% in patients who did not have infection.[20] In their study, the crude excess mortality rate for VAP (21.9%) was higher than CLABSI (16.7%) and CAUTI (16.7%).

Considering the scope of the problem of DAIs and its impact on morbidity and mortality in these patients, developing interventions that further decrease the incidence of DAI are mandated. Continuing audit of the quality of implementation of care bundles is also vital to ensure that the improvements seen are sustained to cover all the subsequent ICU admissions. Further research into the morbidity that DAI imposes on neurosurgical patients is also warranted.

Limitations of the study

Our ICU being a common ICU for all patients admitted under the neurosurgical department, we included patients who were admitted following elective surgery and those who required admission due to neurotrauma including significant spine injury. Our study did not differentiate between the incidences of DAI in patients who had spinal or cranial pathology nor did we differentiate between emergency and elective procedures, as these groups may have shown significant variation in incidence. We also did not include other devices present in our patients such as ventricular drainage catheters and lumbar drains.

Future study

A prospective study to understand the incidence and complications of DAI will better serve to understand its true association with morbidity and mortality in the population.

CONCLUSION

In this retrospective study of DAI in a tertiary care center neurosurgical ICU, during the period from January 2011 to July 2017, we had 6788 patients admitted in the ICU with devices including urinary catheter, ventilator, and central line. DA-HAI was diagnosed in 316 patients with one or more of the three devices. There were 139 deaths in the ICU during this period, of which 124 (1.82%) patients had DAI. Fifty-two (0.79%) patients died of sepsis and related complications, and in other cases, the cause of death was related to neurosurgical issues.

The emergence of MDR organisms and high rate of DAI pose threat to survival in ICU. Implementation of evidence-based strategies such as insertion bundles and training personnel in ensuring asepsis during maintenance of devices have reduced the incidence rate of DAI and related complications in our neurosurgical ICU.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.