Indian Journal of Medical Microbiology Medknow Publications on behalf of Indian Association of Medical Microbiology ISSN: 0255-0857 EISSN: 1998-3646 Vol. 25, Num. 3, 2007, pp. 230-235

Indian Journal of Medical Microbiology, Vol. 25, No. 3, July-September, 2007, pp. 230-235

Original Article

One year trends in the gram-negative bacterial antibiotic susceptibility patterns in a medical intensive care unit in South India

Kaul S, Brahmadathan KN, Jagannati M, Sudarsanam TD, Pitchamuthu K, Abraham OC, John G

Department of Medicine, Christian Medical College, Vellore - 632 004, Tamilnadu
Correspondence Address: Department of Medicine, Christian Medical College, Vellore - 632 004, Tamilnadu, thambu@cmcvellore.ac.in

Date of Submission: 11-Dec-2006
Date of Acceptance: 12-Feb-2007

Code Number: mb07065

Abstract

Keywords: Bacteria, antibiotic susceptibility, medical intensive care unit, descriptive study

Antibiotic resistant bacteria are becoming an increasingly difficult problem in intensive care units (ICUs). Infections in the ICU lead to increased mortality and costs. A study from Mumbai reveals ICU mortality of 67.4 vs. 37.1% for patients with and without pneumonia respectively; infected patients stayed an additional 5.8 days in the ICU and the costs of additional stay and antibiotics accounted for 18.6% of the ICU budget. ^[1]

The study of infections in the ICU setting has been well-documented in India, both retrospectively and prospectively. A retrospective review of ICU infections found that Acinetobacter baumannii was the commonest isolate and prevention of multiple drug resistant (MDR) A. baumannii infections was achieved after discontinuation of cefotaxime. ^[2] In the same patients, molecular typing techniques were used to elucidate molecular epidemiology of the Acinetobacter spp. ^[3]

A prospective study of ICU infections from Varanasi evaluated urine, blood, endotracheal secretions and throat found Klebsiella pneumoniae is the most prevalent isolate from respiratory tract infections followed by Proteus spp, Escherichia coli , Staphylococcus spp. and Acinetobacter spp. The gram-negative enteric bacilli were uniformly resistant to β-lactam antibiotics as well as β-lactam-β-lactamase inhibitors. Resistance to ciprofloxacin and ceftrioxone ranged from 50-100% and 25-83.3% respectively. ^[4] A similar study from Mumbai found enteric gram negative organisms as the commonest isolates (61.9%), followed by Staphylococcus aureus (29.8%). ^[5]

A study from Lucknow on the occurrence of hospital-acquired pneumonia (HAP) and ventilator-associated pneumonia (VAP), the intensive care unit found a preponderance of gram negative bacteria, commonest being Pseudomonas spp. along with 16.3% being polymicrobial infections. ^[6] There are data from specialized ICUs-surgical, neonatal, neurological and respiratory. Acinetobacter infections are common in neurosurgical ICUs. ^[7] In these ICUs comatose patients may be at a greater risk of nosocomial pneumonia. ^[8] In a study from Bangalore in a surgical ICU, E. coli was the commonest and the most resistant organism. ^[9] In a respiratory intensive care unit (RICU) in Chandigarh, the most commonly identified organisms were the Acinetobacter spp. (34.8%), Pseudomonas aeruginosa (23.9%) and Escherichia coli (15.2%). ^[10] A surveillance study from a Neonatal ICU in Bangalore found that Klebsiella pneumoniae (27.3%) was the predominant organism followed by E. coli (16.8%), Staphylococcus aureus (11.7%), Staphylococcus epidermidis and Pseudomonas aeruginosa (10.2%), Enterococcus and Proteus (4.7%), Citrobacter freundi (3.5%) and Clostridium tetani (2.4%). These were isolated from the equipment, cradles, other inanimate objects and environmental surfaces. ^[11]

Keeping a track of ICU infections is also important for outbreak surveillance-these are common in ICUs and early detection is vital. In an outbreak of Acinetobacter spp. infection in the neonatal unit at Lok Nayak Hospital, during six-month period, 68 strains of Acinetobacter baumannii were isolated from blood and CSF of 47 neonates admitted to the intensive care unit. Acinetobacter spp. isolates with similar antibiograms were recovered from intravenous catheter and washbasin. Control of the outbreak was possible only after strict infection control practices in the unit. ^[12]

Data on ICU infections collected prospectively also give us an idea of success or failure of infection control programmes. A study from Mumbai evaluated the absolute numbers of microbial isolates as an indirect reflection of infection rate in the intensive care unit (ICU) for six months in 1992, 1994 and 1996. They found that in spite of the total admission to the ICU increasing, the absolute numbers of infections reduced showing the impact of infection control protocols. ^[13] Some of the measures that have been tried include trials in whether pharmacologically increasing gastric pH affects development of pneumonia in intubated critically-ill patients. ^[14]

As is evident from these studies, it is essential to have an awareness of the prevailing infections in the ICU, the antibiotics likely to work against these infections, a good awareness of any newly emerging pathogen and the early detection of outbreaks. Central to this effort is a database created specifically with this in mind. Unfortunately the microorganisms that cause infections in one part of the world may not be the same in other parts. Thus country specific data is required to help formulate antibiotic policies, test the effectiveness of antibiotic stewardship programmes and to nip any emerging outbreak early before, it leads to serious consequences.

One of the areas of interest in infections in the ICU is the trend in the antibiotic susceptibility patterns (AST) of common pathogens in the ICU. These are important as antibiotic guidelines are commonly made based on them and changes in the AST patterns need to be reflected in regularly updated antibiotic guidelines.

We undertook the present study to prospectively evaluate the changes in the AST patterns over one year of the common gram negative pathogens seen in our medical ICU (MICU). This data we hope will be of use to health planners in India who plan antibiotic guidelines, as well as antimicrobial stewardship programmes such as antibiotic cycling.

Materials and Methods

The Christian Medical College (CMC) is a 2100 bed, urban, tertiary care, teaching hospital, located in Vellore, South India. The study unit was an 11-bed MICU. The MICU is a semi-closed unit with a multi-disciplinary care team, which includes two full-time critical care consultants, 18 internal medicine residents who rotate every two months, four full-time respiratory therapists and one ICU trained nurse for every three patients.

In the year August 2004 to July 2005 there were 491 admissions and the mean length of stay was 7.5 days. Data was prospectively collected on all patients admitted to the MICU between August 2004 and May 2005. Data collected included patient profile, medical history, hospital and ICU admission dates. In addition, process of care information, (mechanical ventilation, central venous catheter use, enteral nutrition data), ICU treatment and events including ICU outcomes and organ system failure (OSF) scores ^[14] were recorded.

Data pertaining to all microbial cultures with special reference to gram-negative bacterial infections, including antibiotic susceptibility data, were prospectively collected and entered into microsoft excel database. The data was analyzed separately for the four predominant gram negative isolates, E. coli, K. pneumoniae, P. aeroginosa and other non-fermenting gram negative bacteria (NFGNB).

Identification of all causative microorganisms was performed by standard microbiologic ethods. ^[15] Susceptibility testing was performed using disk-diffusion method. The breakpoints were those defined by the Clinical and Laboratory Standards Institute (CLSI). ^[16] We do not routinely test for the presence of extended spectrum β-lactamase (ESBL) or carbapenamases. However, resistance to the third generation cephalosporins (cefotaxime and ceftazidime) is highly suggestive of the presence of ESBL in E. coli and K. pneumoniae and resistance to carbapenams in P. aeruginosa is usually mediated by metalo-β-lactamases.

Prior to September 2004, the empirical antibiotic of choice for presumed infections by gram-negative bacteria was intravenous cefotaxime. From September 16, 2004 as part of a planned antibiotic change, cefepime was used instead of cefotaxime for all clinically suspected gram negative bacterial infections.

The data was classified in five time periods such as: 15-08-04 to 15-09-04; 16-09-04 to 15-11-04, 16-11-04 to 15-01-05; 16-01-05 to15-03-05 and16-03-05 to 15-05-05.

Results

During the study period, 491 patients were admitted to the MICU. The details are seen in the tables and figures as mentioned below. [Table - 1] reveals the number of patients seen in the different time periods of the study.

[Table - 2], [Table - 3], [Table - 4], [Table - 5] and [Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4], [Figure - 5] show the comparison of the different time periods and the resistance patterns of the different gram-negative bacteria. Being multiple antibiotics and different organisms we did not calculate the statistical value of the trends. The visual inspection of the graphs reveals a better idea of this data. We see cefotaxime resistance in non-fermenting gram negative bacteria while there is increased cefotaxime resistance in E. coli and Klebsiella spp. However, ceftazidime resistance decreased in all organisms except in E. coli . Resistance to both cefotaxime and ceftazidime is shown separately as is resistance to imipenum or meropenum.

Discussion

Reduction in antimicrobial resistance in the ICUs has been a goal for all intensive care units as it improves outcome and cost to the patients in terms of the expenses for costly antibiotics as well as duration of ICU stay. Mortality associated with multi-drug resistant organisms is also a concern.

We see a difference between E. coli and Klebsiella with regards to their antibiotic susceptibility patterns (ASP). ceftazidime resistance reduced in Klebsiella while cefotaxime resistance increased. In E. coli however, ceftazidime and cefotaxime resistance increased. The reason for this difference is not clear. However, this data is important, given the increase in the incidence of extended spectrum β-lactamase (ESBL) producing gram-negative bacteria that are mostly resistant to ceftazidime and cefotaxime in the ICUs. Any reduction in the magnitude of ESBL infections is welcome. However, as we mentioned in our methods section, ESBL production has to be confirmed by specific tests as other mechanisms of cephalosporin resistance other than ESBL such as amp C β-lactamases, efflux pumps and porin channel alterations also play a role in this resistance.

Why did cefotaxime resistance not decrease? Possibly in our situation cefotaxime resistance was increasing prior to the change of antibiotic from cefotaxime to cefepime and hence we have not yet seen a change in its antibiotic susceptibility pattern. Also 50% of admissions to the MICU come from the emergency department and the rest from wards. The use of cefotaxime in the community is widespread in the treatment of suspected gram-negative infections-this might have also affected their ASPs. However, we did not differentiate between ICU acquired and previously acquired infections in this phase of the trial. However, we have now currently begun attempting to document this data as well.

In pseudomonas and the NFGNB ceftazidime resistance decreased. Pseudomonas being among the commonest organisms associated with ICU infections, this is a useful change given the fact that ceftazidime is the first line antibiotic for Pseudomonas and NFGNB. Cephalosporin resistance seems to be reducing in the NFGNB which is a useful change.

However, carbapenum resistance seems to be increasing, possibly due to their increased use. A previous study from Vellore, evaluating respiratory isolates found that 12.2% of the NFGNB were resistant to imipenum and meropenem. ^[17] It is likely that this carbapenum resistance is mediated by metalo-β-lactamases- however this needs confirmatory tests that were not done on these samples. ^[18]

As expected, cefepime resistance gradually increased. This needs to be constantly monitored as we change to the next antibiotic. In a study comparing mixing versus cycling of antibiotics in MICUs, Martinez and colleagues demonstrated that cycling of anti-pseudomonal drugs including Cefepime lead to reduced Cefepime resistance. ^[19] However, they cycled antibiotics every month.

We present this data from our ICU as an observation of the trends in antimicrobial susceptibility patters over a year of the most common pathogens seen in our ICU, possible reasons for these and the implications of these changing trends. We hope that this data will be of use to health care professionals in areas such as deciding antibiotic cycling policies, analyzing the emergence of new pathogens and the effects of various strategies such as hand washing. We believe that this continuous surveillance is vital to attaining a goal of low levels of ICU infections.

References

Copyright 2007 - Indian Journal of Medical Microbiology

The following images related to this document are available:

Photo images