In The Lancet Infectious Diseases, Karthikeyan Kumarasamy and colleagues1 reported that blaNDM-1-containing bacteria was isolated from a few patients in India, Pakistan, and Bangladesh. Subsequently, Timothy Walsh and colleagues2 identified blaNDM-1-carrying bacterial species from the Enterobacteriaceae, Aeromonadaceae, Vibrionaceae, and Pseudomonadaceae, and some other non-fermenters, in New Delhi water samples. The emergence and rapid environmental dissemination of NDM-1 in a densely populated country like India is a very worrying scenario.

We believe that poor management of the hospital environment and medical waste could be one of many routes of environmental dissemination of pathogens, especially in south Asian countries. For example, in Bangladesh and India, medical facilities rarely have provision for internal waste treatment, and solid hospital waste is often disposed of without proper care on hospital premises3 and liquid waste from hospitals flows with sewage into open ditches and urban river systems.

Wild animals, and birds in particular, can carry and spread multidrug-resistant bacteria and CTX-M-carrying bacteria.4 However, the prevalence of infection in birds living close to people is probably related to the extent of antibiotic resistance in the environment.4, 5

During a research trip to Bangladesh in 2010, we analysed 150 faecal samples from gulls along the coastline and 238 faecal samples from crows in the grounds of Rajshahi Medical College Hospital and Chittagong Medical College Hospital to detect blaNDM-1 expression. We enriched every sample with brain—heart infusion broth supplemented with vancomycin (16 mg/L) for 18 h at 37°C, followed by inoculation and overnight culture at 37°C on chromID extended-spectrum β-lactamase (ESBL) plates. We isolated bacterial colonies and verified the identity of species by biochemical testing and with API 20E biochemical strips. We used the cefpodoxime and cefpodoxime plus clavulanic acid double disc test to measure ESBL production. We isolated ESBL-producing Escherichia coli, Enterobacter cloacae, Klebsiella pneumoniae, Citrobacter freundii, and Raoultella terrigena from the bird faeces. More than 59% of crow faeces and 27% of gull faeces were ESBL-positive phenotypically. To detect blaNDM-1 in these samples, we developed a quantitative PCR with the K pneumoniae plasmid pKpANDM-1 sequence6 and Primer3 software.7 We designed a forward primer (NDM1_123F: 5′-CGG CGT AGT GCT CAG TGT C-3′) and reverse primer (NDM1_281R: 5′-GCC CGC TCA AGG TAT TTT AC-3′) and used iQ SYBR Green Supermix. All isolates were negative for blaNDM-1.

We believe that birds that live close to people might be transmitting ESBL-producing bacteria from hospital to community. Bangladesh shares borders with India, and many people from both countries cross them for trade, medical treatment, and other reasons. The presence of NDM-1 in the environment is a sign of wide dissemination in India, and Bangladesh will probably be the next country to be affected by NDM-1-carrying bacteria. Although we did not detect NDM-1 in Bangladeshi bird faeces, the presence of NDM-1 in Bangladesh cannot be ruled out. Environmental reservoirs of NDM-1-carrying bacterial strains in Bangladesh now need to be identified to stop further spread of these dangerous pathogens.

We declare that we have no competing interests. We thank C G Giske for providing the control strain of NDM-1-carrying Klebsiella pneumonia 05-506; Ahasanul Alam of Rajshahi Medical College; Laboni Akter of Chittagong Medical College, Bangladesh; Uppsala University, Sweden; and International Centre for Diarrhoeal Disease Research, Bangladesh, for logistic support during sample collection. Our study was financially supported by Uppsala University, the Swedish Research Council Formas, and Karin Korsner’s foundation.