, Volume 40, Issue 4, pp 425–429

Frequency of extended-spectrum β-lactamase (ESBL)-producing Gram-negative bacilli in a 200-bed multi-specialty hospital in Vellore district, Tamil Nadu, India


    • Division of Biomedical ResearchSri Narayani Hospital and Research Center
  • H. Narayanan
    • Division of Biomedical ResearchSri Narayani Hospital and Research Center
  • S. Kuppanan
    • Division of Biomedical ResearchSri Narayani Hospital and Research Center
  • B. Nandagopal
    • Division of Biomedical ResearchSri Narayani Hospital and Research Center
Clinical and Epidemiological Study

DOI: 10.1007/s15010-012-0261-6

Cite this article as:
Sankar, S., Narayanan, H., Kuppanan, S. et al. Infection (2012) 40: 425. doi:10.1007/s15010-012-0261-6



Extended-spectrum β-lactamase (ESBL) producers were reported only from hospital settings previously, but, nowadays, its common presence in community settings is evident from reports. Our primary aim was to assess the frequency of ESBL-producing Gram-negative bacteria (GNB) and their antibiogram pattern among the clinical isolates received during an 8-month time period.


The clinical isolates which belonged to the family Enterobacteriaceae from the clinical specimens were included in the study. These clinical isolates were tested for ESBL production using the double-disk synergy test.


In total, 301 patients were included in this study, of which 146 (48.50 %) were found to harbor strains of ESBL producers. The acquisition of ESBL in relation to age, sex, inhabitancy, inpatients, and outpatients was also analyzed. In our study, 50.29 % of inpatients and 45.86 % of outpatients were found to harbor ESBL producers. The difference between the two groups was not statistically significant. We found five meropenem-resistant ESBL-producing strains among the 146 ESBL producers. Rural inhabitants were found to contain more ESBLs when compared to peri-urban inhabitants.


The study showed a high frequency of ESBLs in both community-acquired and hospital-acquired infections. The frequency of ESBLs was higher among isolates from patients who were from rural populations than those from peri-urban populations. The data on ESBL frequency suggests the need for a rational antibiotic use which would reduce the spread of ESBL-producing members of the family Enterobacteriaceae.


ESBLsEnterobacteriaceaeAntibiotic resistanceRuralPeri-urban


The frequency of infections caused by extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae is increasing worldwide [1]. It is essential to assess the frequency of ESBLs and the antibiogram from time to time in both hospital and community care settings. This will allow us to devise treatment and preventive measures such as a rational antibiotic policy. This will help stop the dissemination of ESBL-producing bacteria. Furthermore, limiting drugs to those that act against such organisms will help conserve their efficaciousness. In this context, we measured the frequency of ESBL producers among Gram-negative bacteria (GNB) and the antibiogram pattern of isolates recovered from patients attending a 200-bed multi-specialty hospital in Vellore district of Tamil Nadu, India. Numerous studies have been carried out on the frequency of ESBLs from different parts of India. However, few studies exist on the frequency of ESBLs in rural and peri-urban communities in India. The study was carried out among clinical isolates received for tests in the microbiology department of our hospital. We analyzed the data in terms of sex, age, inhabitancy (rural and peri-urban), and hospitalization status (inpatient and outpatient) in relation to harboring ESBL producers.

Materials and methods

The study was carried out during a period of 8 months (from November 2009 to July 2010). This study has been approved by the Institutional Review Board (Scientific and Medical Ethics Scrutiny). The strains of bacteria belonging to the family Enterobacteriaceae from all the samples collected during the study were included for the study. The hospital is engaged in providing medical care to the rural (3 km or more outside the perimeter of municipal corporation zonal limits), urban (municipal corporation limits), and peri-urban (3 km or less outside the perimeter of municipal corporation limits) populations living in and around Vellore district of Tamil Nadu, India. The patients studied were categorized into rural and peri-urban populations and also as patients within 48 h of hospitalization (OP) and those after 48 h of hospitalization (IP). We considered the IP category to represent hospital-acquired infections (HAIs). This is in accordance with the criteria described by Jacobs and Chin-Hong [2], i.e., infection that becomes clinically evident after 48 h of hospitalization was considered as being hospital acquired.

Culture identification

A total of 1,218 clinical isolates from samples such as urine, exudates, blood, tracheal aspirate, sputum, tip culture, and bile were received during the study period. Of these, 915 (75.1 %) samples either produced growth other than Enterobacteriaceae or normal flora. Such samples were excluded from our study. The remaining 301 (24.71 %) clinical isolates from as many patients which belonged to the family Enterobacteriaceae were included in our study and were screened for the presence of ESBLs. Strains were identified according to the standard microbiology procedures [3].

Phenotypic confirmatory test for ESBL production

All the strains after culture identification were tested for ESBLs using the double-disk synergy (DDS) test according to Clinical and Laboratory Standards Institute (CLSI, 2010). Antibiotic disks of ceftazidime and a combination of ceftazidime with clavulanic acid (30/10 μg), cefotaxime and the combination of cefotaxime with clavulanic acid (30/10 μg), and cefpodoxime and the combination of cefpodoxime with clavulanic acid (30/10 μg) were used for testing. An increase in the zone diameter by ≥5 mm for antibiotics with clavulanic acid versus the antibiotic tested alone was considered to be indicative of ESBL production [4]. The strains were also tested against other antimicrobial agents for susceptibility and zone sizes were interpreted as per the CLSI guidelines [4].

Escherichia coli (ATCC 25922) strain was used as a negative reference and Klebsiella pneumoniae (ATCC 700603) strain was used as a positive reference for ESBL production. The two strains were obtained through the kind courtesy of Dr. G. Thangamathi (Chennai Port Trust Hospital, Chennai, Tamil Nadu, India).

Statistical analysis

Statistical software Epi Info 6.04d for MS-DOS was used for the data computation purposes. The Chi-squared test was used to analyze the association of patients’ age, sex, inhabitancy, and category (inpatient or outpatient) with the carriage of ESBLs. A p-value ≤0.05 was considered to be statistically significant. This software was chosen for its convenience and as per previously published experience [5, 6].


Our study subjects consisted of males and females with mean ages of 53.95 (±2.06) and 49.97 (±1.95) years, respectively. Among the 301 bacterial isolates, E. coli (83.1 %) was the most common organism, followed by Klebsiella spp. (12.6 %) and Proteus mirabilis (4.3 %). The DDS test for ESBL was positive for 146 (48.50 %) of 301 bacterial isolates. Of these, 131 (89.7 %) were E. coli, 11 (7.5 %) were Klebsiella spp., and four (2.7 %) were Proteus mirabilis. The distribution of isolates in different clinical specimens of subjects of this study is presented in Table 1.
Table 1

Distribution of isolates from different types of clinical specimens

Type of specimen

No. of clinical isolates (%)

No. of ESBL producers (%)


223 (74.08)

113 (77.4)


47 (15.51)

20 (13.7)

Tracheal aspirate

5 (1.66)

4 (2.7)


5 (1.66)

1 (0.7)

Urinary Foley’s catheter tip

5 (1.66)

2 (1.4)


2 (0.66)

1 (0.7)


14 (4.65)

5 (3.4)

A majority of ESBL-positive strains showed resistance to co-trimoxazole, ciprofloxacin, and gentamicin. The resistance pattern of these ESBL strains to an array of antimicrobial agents is shown in Fig. 1.
Fig. 1

Percentage of extended-spectrum β-lactamase (ESBL)-producing Gram-negative bacteria (GNB) showing resistance to an array of antibiotics tested

Of the 301 patients analyzed, 167 were inpatients with more than 48 h of hospitalization, which included both general and private wards (single/double occupancy rooms), and 133 patients were from the outpatients department. Information on one patient was not available. Among the 167 inpatients, 84 (50.29 %) patients were found to harbor ESBL producers [hospital acquired infections (HAIs)] and of the 133 outpatients, 61 (45.86 %) were found to harbor ESBL producers [community-acquired infections (CAIs)]. The difference in the proportions of HAIs and CAIs was, however, not statistically significant (χ2 = 0.58; p = 0.44). The association of age, inhabitancy, and sex in harboring ESBLs was also analyzed. In the four different age groups analyzed, the age group 41–60 years harbored more ESBLs (40.9 %) than any other group and this was statistically significant (p ≤ 0.001). When females are compared with males, they exhibited a slightly increased carriage of ESBL producers than males, but the difference was not statistically significant (p = 0.81). The rural inhabitants were found to contain more ESBLs when compared to peri-urban inhabitants (p ≤ 0.001) (Table 2).
Table 2

ESBL-producing Enterobacteriaceae stratified by age, sex, and inhabitancy of patients


Total number of isolates (%)

ESBL producers, no. (%)

χ2 and p-value



148 (49.16)

72 (49.32)

χ2 = 0.05; p = 0.81


153 (50.83)

74 (50.68)

Age (years)


28 (9.30)

9 (6.16)

χ2 = 72; p ≤ 0.0001


39 (12.95)

21 (14.39)


123 (40.86)

63 (43.15)

 61 and above

111 (36.87)

53 (36.30)



220 (73.08)

113 (77.39)

χ2 = 87.7; p ≤ 0.0001


81 (26.91)

33 (22.60)

 Rural males

111 (36.87)

57 (50.44)

χ2 = 0.02; p = 0.89

 Rural females

109 (36.21)

56 (49.56)

 Peri-urban males

37 (12.29)

15 (45.45)

χ2 = 0.55; p = 0.46

 Peri-urban females

44 (14.61)

18 (54.54)



167 (55.48)

84 (57.53)

χ2 = 6.63; p = 0.010


133 (44.51)

62 (42.47)

p < 0.05 denotes statistically significant


In our study, among the 301 isolates of the family Enterobacteriaceae from as many patients tested, 146 were found to be positive for ESBLs. Patients from the age group 41–60 years harbored a higher number of ESBLs.

Conflicting evidences are available concerning the role of the sex of the individual as being one of the other risk factors. Colodner et al. [7] reported that male sex was a risk factor for the acquisition of ESBLs; in contrast, Peña et al. [8] reported that female sex was a major risk factor. In this study, we did not find any significant association between patients’ sex, age, inhabitancy, and acquisition of ESBL producers. In this study, association of the older age groups with the acquisition of ESBL producers was significant (p < 0.05). A multivariate analysis was not carried out as only the above three risk factors were considered. Harris et al. [9], however, showed that patients over 60 years of age harbored higher ESBL counts. Subsequently, Mumtaz et al. [10] reported 27.9 % of ESBLs among patients between 61 and 70 years of age. In contrast to these studies, we found more ESBLs (43.15 %) in the 41–60 years age group.

Among the 146 ESBL-positive patients, women with complaints of urinary tract infections (UTIs) constituted 43.15 %, with a preponderance of rural women (79.36 %). It has been reported that adult women have a higher frequency of UTI than men, principally owing to anatomic and physical factors [11, 12]. In this study, E. coli was predominantly isolated from rural women (77.7 %) and peri-urban women (20.63 %) with UTIs. Chatterjee et al. [13] reported that chances exist of uropathogenic E. coli spreading through the fecal contamination of ground water in rural India where sanitation facilities are poor. The reason behind the higher frequency of ESBLs among rural women in our study may be due to the lack of sanitary facilities in the villages and the disposal of human waste in open areas that pave the way for the propagation of pathogens to uninfected new hosts through different routes. Earlier, the presence of NDM-1 β-lactamase-producing bacteria has been demonstrated in environmental samples obtained in New Delhi. The authors indicate the importance of this finding in terms of the risk of clinically significant infections produced by such organisms [14].

A recent nationwide study [15] reported an alarming incidence of ESBL-producing Enterobacteriaceae in both CAIs and HAIs across India. Recent studies have shown that, in southern parts of India, varying numbers of ESBL-producing Enterobacteriaceae were isolated from clinical specimens, 69 % in Pondicherry [16], 64.2 % in Mangalore [17], 60 % in Chennai [18], 32.1 % in Bijapur, Karnataka [19], and 40 % of E. coli and 41 % of Klebsiella pneumoniae in Coimbatore [20]. Another study from Vellore district [21] reported a high incidence (73.28 %) of ESBLs. The higher frequency was probably due to the study setting (a large tertiary care referral center). In our study, we report that 48.50 % of isolates showed ESBL production (57.2 % of HAIs and 42.5 % of CAIs). This could be due to the study center being more accessible to rural, peri-urban, and contiguous tribal populations of Vellore district compared to the urban population. Our data indicates that the occurrence of ESBLs would vary between different geographical regions within the country. Livermore [22] reported that the frequency of ESBLs among clinical isolates vary greatly worldwide and in geographical areas, changing rapidly overtime. In the antibiogram analysis of 146 ESBL-positive strains, five ESBL-positive strains unusually showed resistance to meropenem. This shows that these strains may belong to the Bush–Jacoby group (2009) 3a subclass B, which has a characteristic feature of hydrolyzing carbapenems [23]. However, further confirmatory studies are needed in order to verify the enzyme type involved.

A previous study in healthy asymptomatic pregnant women in this region (rural and urban areas of Vellore) reported that 42 % of commensal E. coli were resistant to at least one antibiotic and 8 % were resistant to more than one antibiotic [24]. Further, strains causing infections were more likely to be resistant than the commensal strains. Our study, however, has certain limitations, such as the unavailability of patient information about previous antibiotic use, as the majority of individuals are not sufficiently literate and may self-medicate with antimicrobials, often illegally using over-the-counter purchases. Hawkey [25] reviewed the frequency of ESBLs in Asia (including India) and suggested that the spread of Enterobacteriaceae through poor-quality drinking water and poor sewage disposal probably accounts for the fecal carriage of ESBL producers.

As reported by Global Antibiotic Resistance Partnership [26], the β-lactam antibacterials and penicillins are commonly sold over the counter in India. Necessary steps in terms of the rational use of antibiotics and strict adherence to the concept of “reserve drug” to minimize the unnecessary use of available antimicrobials should be followed. Health practitioners and physicians should give priority to infection control practices, as this is the best way to prevent the spread of bacterial resistance, because only limited drugs are available as alternatives.

Conflict of interest


Copyright information

© Springer-Verlag 2012