Abstract
Objectives
Data on antimicrobial resistance (AMR) among children in Nepal are limited. Here we have characterized the causes of bacterial bloodstream infections (BSIs), antimicrobial resistance patterns and the mechanisms of β-lactamase production in Enterobacterales among children attending outpatient and inpatient departments of a secondary care paediatric hospital in Nepal.
Methods
We retrospectively collected demographic and clinical data of culture-proven bacterial BSIs between January 2017 and December 2022 among children <18 years attending a 50-bedded paediatric hospital. Stored isolates were subcultured for antimicrobial susceptibility testing against commonly used antimicrobials. Enterobacterales displaying non-susceptibility to β-lactams were phenotypically and genotypically investigated for ESBLs, plasmid-mediated AmpC (pAmpC) β-lactamases and carbapenemases.
Results
A total of 377 significant bacteria were isolated from 27 366 blood cultures. Among 91 neonates with a BSI, Klebsiella pneumoniae (n = 39, 42.4%), Pseudomonas aeruginosa (n = 15, 16.3%) and Acinetobacter baumannii complex (n = 13, 14.1%) were most common. In the non-neonates, 275/285 (96.5%) infections were community-acquired including Staphylococcus aureus (n = 89, 32.4%), Salmonella Typhi (n = 54, 19.6%) and Streptococcus pneumoniae (n = 32, 11.6%). Among the 98 S. aureus, 29 (29.6%) were methicillin-resistant Staphylococcus aureus. K. pneumoniae and Escherichia coli demonstrated non-susceptibility to extended-spectrum cephalosporins and carbapenems in both community and hospital-acquired cases. For E. coli and K. pneumoniae, blaCTX-M (45/46), blaEBC (7/10) and blaOXA-48 (5/6) were common among their respective groups.
Conclusions
We determined significant levels of AMR among children attending a secondary care paediatric hospital with BSI in Nepal. Nationwide surveillance and implementation of antimicrobial stewardship policies are needed to combat the challenge imposed by AMR.
Introduction
Antimicrobial resistance (AMR) is a leading global cause of mortality. In 2019, an estimated 1.27 million deaths were directly attributed to bacterial AMR infections.1 Death rates per 100 000 population were highest in African regions followed by South Asia and one in every five deaths due to AMR infections occurred in children aged under 5 years.1 The burden of AMR infections is high in Nepal as in many low- and middle-income countries.2 A 23-year retrospective analysis from a tertiary care centre in Kathmandu has shown gradually increasing levels of MDR infections.3 In a further study from the same institution, MDR was reported in 643/719 (89.4%) Escherichia coli, 488/532 (91.7%) Klebsiella spp., 510/520 (98.1%) Enterobacter spp. and 314/382 (82.2%) Acinetobacter spp.4 High levels of non-susceptibility to third-generation cephalosporins, fluoroquinolones and aminoglycosides among Klebsiella spp. and E. coli, multidrug resistance in Acinetobacter spp. and Pseudomonas aeruginosa, and methicillin-resistant Staphylococcus aureus (MRSA) is frequent in tertiary care centres.5–10
Surveillance is a critical component of AMR containment but studies from tertiary care centres may not be representative of the levels of AMR at lower levels of the healthcare system.11 Despite the burden of bacterial infection in children and the high level of antibiotic exposure, especially in those under 5 years,12 there is a lack of focus on this age group. Insufficient laboratory capacity11 and a lack of local epidemiology data encourage clinicians in low- and middle-income countries to prescribe broad-spectrum antimicrobials as empiric therapy potentially amplifying the emergence and spread of AMR infections.
We have studied the bacterial causes and antimicrobial resistance patterns in culture-proven bloodstream infections (BSIs) among children <18 years old attending a secondary care paediatric hospital in the Kathmandu Valley. We characterize hospital and community-acquired infections and explore the mechanisms of β-lactamase production among Enterobacterales.
Materials and methods
Ethics
The study proposal and protocols were reviewed and approved by the ethical review committee of the Nepal Health Research Council (NHRC) (ERB Protocol Registration No. 414/2021 P). All the work was conducted as a part of Siddhi Memorial Hospital’s (SMH) surveillance activities. Anonymous data, with no personal identifiers, were retrieved retrospectively from hospital records. Thus, the need for written consent was waived for the study by the ethical review committee of NHRC.
Study design and setting
This is a retrospective study conducted at SMH. SMH is a 50-bedded paediatric and maternal secondary level hospital located at Bhaktapur, Nepal. The hospital serves the local population of Bhaktapur municipality, in the Kathmandu Valley. SMH microbiology laboratory performs internal quality controls and participates in inter-laboratory external laboratory quality control programmes. We studied all children with positive bacterial growth from blood culture during the 6 years from 1 January 2017 to 31 December 2022.
Sample characteristics and laboratory procedures
Febrile children (<18 years old) attending the inpatient or outpatient facility, or inpatients who become unwell, were investigated with a blood culture at the request of the consulting paediatrician. For inpatients, blood was inoculated into a BD BACTECTM Peds PlusTM bottle (≥2 to 3 mL in infants and children; ≥ 1 mL of blood for neonates or early infants). Commercial brain heart infusion broth bottles (HiMedia, India) supplemented with 0.05% sodium polyanetholsulfonate were used for blood culture for patients attending outpatient department (OPD). Each sample was incubated aerobically at 37°C for up to 5 days. If turbidity was noted, the broth was inoculated onto 5% sheep blood agar and MacConkey agar plates. The plates were then incubated at 37°C for 18 to 24 hours. Bacterial identification was determined by Gram’s staining, colony characteristics and in-house biochemical tests. The identification scheme at SMH is described in Table S1 (available as Supplementary data at JAC-AMR Online). Antimicrobial susceptibility testing was performed by the Kirby–Bauer disc diffusion method. The organisms were stored at the time of isolation in skim-milk tryptone glucose glycerol medium at −40°C.
Isolate recovery and antimicrobial susceptibility testing
Bacteria isolated from the blood culture of patients were recovered from storage in the freezer by inoculation onto 5% sheep blood agar and MacConkey agar plates followed by aerobic incubation at 37°C for 18 to 24 hours. The isolates were subcultured further until pure isolated colonies were obtained. For cases in which the same organism was obtained from the same patient more than once within the same admission, only the first isolate was studied.
AST was repeated by Kirby–Bauer disc diffusion method according to CLSI guidelines.13 Clinically important antimicrobials commonly considered in our hospital for the treatment of bacterial infections were tested against each pathogen. Mueller–Hinton Agar and the antimicrobial discs used for AST were purchased from Mast Group Ltd (Liverpool, UK). Vancomycin and teicoplanin susceptibility were determined by the interpretation of MIC by Etest (Liofilchem). E. coli ATCC 25922 and Staphylococcus aureus ATCC 25923 were used for quality control of AST.
The assignment of susceptibility categories (resistant, intermediate and susceptible) was performed as per the CLSI guideline.13 Interpretation of tigecycline susceptibility for E. coli was based on EUCAST clinical breakpoint.14 Staphylococcus aureus non-susceptible to cefoxitin was categorized as MRSA. MDR was defined as non-susceptibility to any of the antimicrobials in ≥3 antimicrobial classes.15 Non-susceptibility was defined as either resistance or intermediate or susceptible dose-dependent phenotype.
Determination of ESBL, AmpC β-lactamase and carbapenemase production
E. coli and K. pneumoniae isolates showing non-susceptibility (intermediate or resistant phenotype) to cefoxitin, any third-generation cephalosporin or carbapenems were investigated by D72C tests (Mast Group Ltd, Liverpool, UK) according to the manufacturer’s instructions to detect ESBL, AmpC β-lactamase and suspected carbapenemase production. Isolates presumptively identified as potential carbapenemase producers by the D72C test were investigated by the modified carbapenem inactivation method (mCIM) for confirmation as per the recommendations of the CLSI guideline.13
For the isolates showing β-lactamase production, crude DNA extracted by the heat lysis method was subjected to PCR screening for genotyping. Isolates were investigated for blaCTX-M,16 blaSHV (not K. pneumoniae) and blaTEM,17 five common carbapenemases (blaOXA-48, blaNDM, blaKPC, blaIMP and blaVIM),18 and six plasmid-mediated AmpC β-lactamase (pAmpC) families (blaMOX, blaCIT, blaDHA, blaACC, blaEBC and blaFOX).19 Carbapenem non-susceptible isolates were investigated for the presence of all three groups of β-lactamases. DNA samples extracted from the clinical isolates previously characterized to harbour the β-lactamases were used as positive controls and the DNA of ATCC E. coli 25922 was used as a negative control. Nuclease-free water was used as a no template control.
Data acquisition and definitions
Demographic and clinical data of the patients were retrieved from the dataset of the microbiology laboratory and medical records at the hospital. BSI cases were considered hospital-acquired if the sample was collected after 48 hours of hospitalization. If the blood was sampled before 48 hours or the child was not admitted to the hospital, the case was categorized as community-acquired. All the cases from the outpatient department and emergency department were considered community-acquired. CoNS (except for cases where CoNS were obtained from two successive blood cultures with identical AST) and Bacillus spp. were considered contaminants. Early-onset neonatal sepsis was defined as onset of clinical signs and symptoms of sepsis within 72 hours from birth. Late-onset neonatal sepsis was defined as sepsis after 72 hours.
Data analysis
The data were entered into an MS Excel sheet (Microsoft) and were imported to SPSS (SPSS v.22; IBM, New York, USA). Variables are presented in proportions and percentages. A Chi-squared test was used to determine the association of β-lactamase production with fluoroquinolone non-susceptibility. A P value of <0.05 was considered for statistical significance. Compliance with the Microbiology Investigation Criteria for Reporting Objectively (MICRO) guideline is in the Supplementary Material (Table S2).20
Results
Demographics and characteristics of infections
A total of 27 366 blood cultures were taken during the 6-year study period. After removing contaminants (coagulase-negative staphylococci and Bacillus spp.) and repeated isolates from the same patients, there were a total of 377 significant isolates from 376 patients (Figure 1).
Figure 1.
Flowchart showing criteria for selecting isolates included in this study.
Of the 376 patients, 91 (24.2%) were neonates and 285 (75.8%) non-neonates. Children aged 1 year to <5 years were the most common age group (138/376, 36.7%) (Table 1). Female to male ratio was 71:117. Of 376 children, 146 (38.8%) were hospitalized at SMH.
Table 1.
Characteristics of 376 children with culture-proven bacterial BSIs
| Characteristic | N (%) |
|---|---|
| Age groups | |
| Neonate (0 to <29 days) | 91 (24.2) |
| Infant (29 days to <1 year) | 60 (16.0) |
| Pre-school (1 to <5 years) | 138 (36.7) |
| Children (5 to <13 years) | 79 (21.0) |
| Adolescent (13 to <18 years) | 8 (2.1) |
| Sex | |
| Male | 234 (62.2) |
| Female | 142 (37.8) |
| Department | |
| Outpatient department | 167 (44.4) |
| Paediatric emergency | 63 (16.8) |
| NICU | 76 (20.2) |
| PICU | 20 (5.3) |
| Paediatric ward | 34 (9.0) |
| Cabin | 13 (3.5) |
| Surgery | 3 (0.8) |
| Site of acquisition | |
| Community-acquired | 303 (80.6) |
| Hospital-acquired | 73 (19.4) |
| Outcomea | |
| Discharged | 109 (74.7) |
| Death | 20 (13.7) |
| Referred | 3 (2.1) |
| Discharges against medical advice | 12 (8.2) |
| Discharged on request | 2 (1.4) |
aFor 146 hospitalized children at SMH.
Bacterial isolates
The bacterial causes of BSI by age group and site of acquisition are in Table 2. Among the 91 neonates with a BSI, 11 were considered early onset, 54 were late onset and for 27 there were insufficient clinical data to be classified. K. pneumoniae (39/92, 42.4%) was the leading cause of BSI among neonates, followed by P. aeruginosa (15/92, 16.3%), A. baumannii complex (13/92, 14.1%) and S. aureus (9/92, 9.8%) (Table 2).
Table 2.
Bacterial pathogens stratified by age group and site of acquisition
| Age groups (n, column % for each age group) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Organism | Neonate (0 to <29 days) (n = 92) |
Non-neonate (29 days to < 18 years (n = 285) | Infant (29 days to <1 year) (n = 60) | Pre-school (1 to <5 years) (n = 138) | Children (5 to <13 years) (n = 79) |
Adolescent (13 to <18 years) (n = 8) |
||
| Total | Total | CA (n = 52) | HA (n = 8) | CA (n = 136) | HA (n = 2) | CA (n = 79) | CA (n = 8) | |
| Staphylococcus aureus (n = 98) | 9 (9.8) | 89 (31.2) | 15 (28.9) | — | 47 (34.6) | — | 23 (29.1) | 4 (50.0) |
| Salmonella Typhi (n = 54) | — | 54 (19.0) | 1 (1.9) | — | 19 (14.0) | — | 33 (41.8) | 1 (12.5) |
| K. pneumoniae (n = 73) | 39 (42.4) | 34 (11.9) | 8 (15.4) | 6 (75.0) | 18 (13.2) | — | 1 (1.3) | 1 (12.5) |
| S. pneumoniae (n = 34) | 1 (1.1) | 33 (11.6) | 9 (17.3) | — | 18 (13.2) | 1 (50.0) | 5 (6.3) | — |
| E. coli (n = 34) | 6 (6.5) | 28 (9.8) | 6 (11.5) | 1 (12.5) | 13 (9.6) | — | 7 (8.9) | 1 (12.5) |
| A. baumannii complex (n = 40) | 13 (14.1) | 27 (9.5) | 5 (9.6) | 1 (12.5) | 14 (10.3) | 1 (50.0) | 6 (7.6) | — |
| S. pyogenes (n = 6) | — | 6 (2.1) | 2 (3.9) | — | 3 (2.2) | — | 1 (1.3) | — |
| E. cloacae complex (n = 5) | 1 (1.1) | 4 (1.4) | 2 (3.9) | — | 2 (1.5) | — | — | — |
| Salmonella Paratyphi A (n = 4) | — | 4 (1.4) | — | — | 2 (1.5) | — | 2 (2.5) | — |
| Proteus mirabilis (n = 4) | 2 (2.2) | 2 (0.7) | 1 (1.9) | — | — | — | 1 (1.3) | — |
| Citrobacter freundii (n = 2) | — | 2 (0.7) | 2 (3.9) | — | — | — | — | — |
| Pseudomonas aeruginosa (n = 16) | 15 (16.3) | 1 (0.4) | — | — | — | — | — | 1 (12.5) |
| Coagulase-negative Staphylococci (n = 5) | 4 (4.4) | 1 (0.4) | 1 (1.9) | — | — | — | — | — |
| Serratia marscescens (n = 1) | 1 (1.1) | — | — | — | — | — | — | — |
| Burkholderia cepacia complex (n = 1) | 1 (1.1) | — | — | — | — | — | — | — |
‘—’ means no observations.
CA, Community-acquired; HA: Hospital-acquired.
In the non-neonates, 275/285 (96.5%) infections were community-acquired including Staphylococcus aureus (89/275, 32.4%), Salmonella Typhi (54/275, 19.6%), Streptococcus pneumoniae (32/275, 11.6%), Klebsiella pneumoniae (28/275, 10.2%), E. coli (27/275, 9.8%), Acinetobacter baumannii complex (25/275, 9.1%), Streptococcus pyogenes (6/275, 2.2%), Salmonella Paratyphi A (4/275, 1.5%) and Enterobacter cloacae complex (4/275, 1.5%). S. aureus, E. coli and A. baumannii complex were common in all age groups, the typhoidal salmonellae occurred in those aged 1 to 13 years, and S. pneumoniae, K. pneumoniae and E. cloacae complex in those under 5 years. K. pneumoniae was the main bacteria isolated in the hospital-acquired BSI in this group.
Antimicrobial resistance
The antimicrobial non-susceptibility patterns are shown in Table 3(a) (neonates) and Table 3(b) (non-neonates), and the non-susceptibility data stratified by the site of acquisition are presented in Table S3(a) (community-acquired) and Table S3(b) (hospital-acquired). In the neonates, there were high levels of extended-spectrum cephalosporin non-susceptibility among K. pneumoniae (32/39, 82%) and E. coli (3/6, 50%). Non-susceptibility to other antimicrobials was common in K. pneumoniae including meropenem in 4/39 (10%), piperacillin-tazobactam in 12/39 (31%), ciprofloxacin in 26/39 (67%) and gentamicin in 12/39 (31%), but not in E. coli. In the non-neonates, extended-spectrum cephalosporin non-susceptibility was present in 20/34 (59%) K. pneumoniae and 15/28 (54%) E. coli. Non-susceptibility to meropenem was 6/34 (18%) in K. pneumoniae and 3/28 (11%) in E. coli. The proportions of non-susceptibilities ranged from 18% to 26% for gentamicin and piperacillin-tazobactam and 32% to 47% for ciprofloxacin in these two Gram-negative bacteria (GNB). Levels of non-susceptibilities were low for the A. baumannii complex and P. aeruginosa except for trimethoprim-sulfamethoxazole non-susceptibility in A. baumannii complex isolates in 4/13 (31%) neonates and 7/27 (26%) non-neonates. Ciprofloxacin resistance was detected in 21/54 (39%) of the Salmonella Typhi isolates and 2/4 (50%) Salmonella Paratyphi, but there was no resistance to ceftriaxone or azithromycin. Antimicrobial non-susceptibility was infrequent in the small number of other GNBs.
Table 3.
Antimicrobial non-susceptibility patterns of bacterial isolates obtained from blood cultures of the three hundred and 76 paediatric patients
| (a) Neonates (n = 92) | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| K. pneumoniae (n = 39) | E. coli (n = 6) | A. baumannii complex (n = 13) | Pseudomonas aeruginosa (n = 15) | E. cloacae complex (n = 1) | Proteus mirabilis (n = 2) | Burkholderia cepacia complex (n = 1) | Serratia marcescens (n = 1) | Staphylococcus aureus (n = 9) | MRSA (n = 5) | MSSA (n = 4) | S. pneumoniae (n = 1) | CoNS (n = 4) | |||
| Antimicrobial non-susceptibility, No. of non-susceptible isolates/No. of total isolates tested (%) | AP | — | 5/6 (83) | — | — | — | — | — | — | PG | 7/9 (78) | 5/5 (100) | 2/4 (50) | 0/1 (0) | 0/4 (0) |
| AUG | 22/39 (56) | 2/6 (33) | — | — | — | — | — | — | FOX | 5/9 (56) | 5/5 (100) | 0/4 (0) | — | — | |
| PTZ | 12/39 (31) | 0/6 (0) | 1/13 (8) | 0/15 (0) | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | CPD | 5/9 (56) | 5/5 (100) | 0/4 (0) | — | — | |
| SAM | — | — | 1/13 (8) | — | — | — | — | — | E | 7/9 (78) | 4/5 (80) | 3/4 (75) | 0/1 (0) | 2/4 (50) | |
| CTX | 35/39 (90) | 3/6 (50) | — | — | 0/1 (0) | 0/2 (0) | — | — | CD | 1/9 (11) | 1/5 (20) | 0/4 (0) | 0/1 (0) | 1/4 (25) | |
| CAZ | 32/39 (82) | 3/6 (50) | 1/13 (8) | 2/15 (13) | 0/1 (0) | 0/2 (0) | 1/1 (100) | 0/1 (0) | CIP | 6/9 (67) | 4/5 (80) | 2/4 (50) | 0/1 (0) | 0/4 (0) | |
| CPM | 19/39 (49) | 0/6 (0) | 1/13 (8) | 2/15 (13) | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | VAN | 0/9 (0) | 0/5 (0) | 0/4 (0) | 0/1 (0) | 0/4 (0) | |
| IMI | — | — | 1/13 (8) | 0/15 (0) | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | TEI | 0/5 (0) | 0/3 (0) | 0/2 (0) | — | 0/4 (0) | |
| MEM | 4/39 (10) | 0/6 (0) | 1/13 (8) | 0/15 (0) | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | GM | 1/9 (11) | 1/5 (20) | 0/4 (0) | 0/1 (0) | 0/4 (0) | |
| CIP | 26/39 (67) | 1/6 (17) | 1/13 (8) | 6/15 (40) | 0/1 (0) | 0/2 (0) | 1/1 (100) | 0/1 (0) | DXT | 0/5 (0) | 0/3 (0) | 0/2 (0) | — | 0/4 (0) | |
| AK | 8/39 (21) | 0/6 (0) | 0/13 (0) | 4/15 (27) | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | C | 0/5 (0) | 0/3 (0) | 0/2 (0) | 0/1 (0) | 0/4 (0) | |
| GM | 12/39 (31) | 0/6 (0) | 0/13 (0) | — | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | LZD | 0/5 (0) | 0/3 (0) | 0/2 (0) | — | 0/4 (0) | |
| TGC | — | 0/6 (0) | — | — | — | — | — | — | TS | 2/9 (22) | 1/5 (20) | 1/4 (25) | 1/1 (100) | 1/4 (25) | |
| TS | 18/39 (46) | 4/6 (67) | 4/13 (31) | — | 0/1 (0) | 0/2 (0) | 0/1 (0) | 0/1 (0) | LEV | — | — | — | — | — | |
| (b) Non-neonates (n = 285) | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| K. pneumoniae (n = 34) | E. coli (n = 28) | A. baumannii complex (n = 27) | Pseudomonas aeruginosa (n = 1) | E. cloacae complex (n = 4) | Proteus mirabilis (n = 2) | Citrobacter freundii (n = 2) | Salmonella Typhi (n = 54) | Salmonella Paratyphi A (n = 4) | Staphylococcus aureus (n = 89) | MRSA (n = 24) | MSSA (n = 65) | S. pneumoniae (n = 33) | S. pyogenes (n = 6) | CoNS (n = 1) | ||||
| Antimicrobial non-susceptibility, No. of non-susceptible isolates/No. of total isolates tested (%) | AP | — | 15/28 (54) | — | — | — | — | — | AP | 2/54 (4) | 0/4 (0) | PG | 53/89 (60) | 24/24 (100) | 29/65 (45) | 1/33 (3) | 0/6 (0) | 0/1 (0) |
| AUG | 15/34 (44) | 8/28 (29) | — | — | — | — | — | C | 0/54 (0) | 0/4 (0) | FOX | 24/89 (27) | 24/24 (100) | 0/65 (0) | — | — | — | |
| PTZ | 9/34 (26) | 5/28 (18) | 1/27 (4) | 0/1 (0) | 0/4 (0) | 0/2 (0) | 0/2 (0) | TS | 0/54 (0) | 0/4 (0) | CPD | 24/89 (27) | 24/24 (100) | 0/65 (0) | — | — | — | |
| SAM | — | — | 2/27 (7) | — | — | — | — | PEF | 11/14 (79) | 1/1 (100) | E | 56/89 (63) | 17/24 (71) | 39/65 (60) | 4/33 (12) | 0/6 (0) | 0/1 (0) | |
| CTX | 20/34 (59) | 15/28 (54) | — | — | 1/4 (25) | 0/2 (0) | — | CIP | 21/54 (39) | 2/4 (50) | CD | 18/89 (20) | 6/24 (25) | 12/65 (18) | 0/33 (0) | 0/6 (0) | 0/1 (0) | |
| CAZ | 20/34 (59) | 14/28 (50) | 2/27 (7) | 0/1 (0) | 1/4 (25) | 0/2 (0) | 0/2 (0) | CRO | 0/54 (0) | 0/4 (0) | CIP | 29/89 (33) | 16/24 (67) | 13/65 (20) | 1/33 (3) | 0/6 (0) | 0/1 (0) | |
| CPM | 13/34 (38) | 10/28 (36) | 1/27 (4) | 0/1 (0) | 0/4 (0) | 0/2 (0) | 0/2 (0) | ATH | 0/54 (0) | 0/4 (0) | VAN | 0/89 (0) | 0/24 (0) | 0/65 (0) | 0/33 (0) | 0/6 (0) | 0/1 (0) | |
| IMI | — | — | 0/27 (0) | 0/1 (0) | 0/4 (0) | 0/2 (0) | 0/2 (0) | MEM | 0/54 (0) | 0/4 (0) | TEI | 0/33 (0) | 0/13 (0) | 0/20 (0) | — | — | 0/1 (0) | |
| MEM | 6/34 (18) | 3/28 (11) | 0/27 (0) | 0/1 (0) | 0/4 (0) | 0/2 (0) | 0/2 (0) | GM | 0/54 (0) | 0/4 (0) | GM | 7/89 (8) | 3/24 (13) | 4/65 (6) | 2/33 (6) | 0/6 (0) | 0/1 (0) | |
| CIP | 16/34 (47) | 9/28 (32) | 1/27 (4) | 0/1 (0) | 0/4 (0) | 1/2 (50) | 0/2 (0) | DXT | 0/33 (0) | 0/13 (0) | 0/20 (0) | 0/5 (0) | — | 0/1 (0) | ||||
| AK | 6/34 (18) | 5/28 (18) | 0/27 (0) | 0/1 (0) | 0/4 (0) | 0/2 (0) | 0/2 (0) | C | 2/36 (6) | 2/14 (14) | 0/22 (0) | 0/33 (0) | 0/6 (0) | 0/1 (0) | ||||
| GM | 8/34 (24) | 7/28 (25) | 0/27 (0) | — | 0/4 (0) | 0/2 (0) | 0/2 (0) | LZD | 0/33 (0) | 0/13 (0) | 0/20 (0) | — | — | 0/1 (0) | ||||
| TGC | — | 0/28 (0) | — | — | — | — | — | TS | 25/89 (28) | 8/24 (33) | 17/65 (26) | 9/33 (27) | 2/6 (33) | 0/1 (0) | ||||
| TS | 11/34 (32) | 6/28 (21) | 7/27 (26) | — | 1/4 (25) | 0/2 (0) | 0/2 (0) | LEV | — | — | — | 0/33 (0) | 0/6 (0) | — | ||||
‘—’ means not-tested.
AP, ampicillin; AUG, amoxicillin-clavulanic acid; PTZ, piperacillin-tazobactam; SAM, ampicillin-sulbactam; CTX, cefotaxime; CAZ, ceftazidime; CPM, cefepime; IMI, imipenem; MEM, meropenem; CIP, ciprofloxacin; AK, amikacin; GM, gentamicin; TGC, tigecycline; TS, trimethoprim-sulfamethoxazole; C, chloramphenicol; PEF, pefloxacin; CRO, ceftriaxone; ATH, azithromycin; PG, penicillin; FOX, cefoxitin; CPD, cefpodoxime; E, erythromycin; CD, clindamycin; VAN, vancomycin; TEI, teicoplannin; DXT, doxycycline; LZD, linezolid; LEV, levofloxacin; MSSA, methicillin-susceptible Staphylococcus aureus.
In non-neonates, 24/89 (27.0%) of the S. aureus isolates were MRSA, and many of these isolates were also resistant to erythromycin, ciprofloxacin and trimethoprim-sulfamethoxazole. Notably, 5/9 (55.6%) of S. aureus isolates in the neonates were MRSA. Of the 29 MRSA BSI, 26 were community-acquired and most were in the pre-school age group (13/29, 44.8%). No S. aureus was non-susceptible to vancomycin. Additional testing on a sub-set of 38 isolates confirmed susceptibility to vancomycin and teicoplanin [median MIC 0.25 mg/L (IQR 0.19–0.5); teicoplanin, (median MIC 0.5 mg/L (IQR 0.38–1)]. Levels of non-susceptibility to relevant antimicrobials for S. pneumoniae and S. pyogenes were low.
Overall, 105/377 (27.9%) isolates were MDR. Three bacteria, K. pneumoniae (51/105), S. aureus (33/105) and E. coli (17/105), accounted for 96.2% of all MDR isolates. The proportion of MDR was highest among K. pneumoniae (51/73, 69.9%), neonates (41/92, 44.6%), females (45/143, 31.5%) and hospital-acquired infections (32/74, 43.2%).
Characteristics of β-lactamase producing isolates
Of 18 E. coli and 55 K. pneumoniae non-susceptible to third-generation cephalosporins and/or cefoxitin, 15 E. coli and 41 K. pneumoniae were available for further analysis. Among these isolates, 13 E. coli and 40 K. pneumoniae were found to display at least one phenotypic mechanism of β-lactamase production. ESBL was the most common mechanism of β-lactamase production (n = 46) followed by pAmpC β-lactamase (n = 10) and carbapenemase (n = 6) (Table 4).
Table 4.
Distribution of ESBL, pAmpC β-lactamase and carbapenemase genotypes among 13 E. coli and 40 K. pneumoniae isolates
| Pathogen | β-lactamase type | n (%) | Neonates n (% of 28) |
Non-neonates n (% of 25) |
||
|---|---|---|---|---|---|---|
| ESBL | pAmpC | Carbapenemase | ||||
| E. coli | bla CTX-M | 8 (15.1) | 2 (7.1) | 6 (24.0) | ||
| bla EBC | 1 (1.9) | — | 1 (4.0) | |||
| bla TEM | bla DHA | 1 (1.9) | — | 1 (4.0) | ||
| bla CTX-M bla SHV | bla EBC | 1 (1.9) | — | 1 (4.0) | ||
| bla CTX-M | bla NDM bla OXA-48 | 2 (3.8) | — | 2 (8.0) | ||
| K. pneumoniae | bla CTX-M | 24 (45.3) | 14 (50.0) | 10 (40.0) | ||
| bla CTX-M bla TEM | 5 (9.4) | 4 (14.3) | 1 (4.0) | |||
| bla NDM | 1 (1.9) | — | 1 (4.0) | |||
| bla OXA-48 | 1 (1.9) | — | 1 (4.0) | |||
| bla DHA blaEBC | 1 (1.9) | 1 (3.6) | — | |||
| bla EBC | 3 (5.7) | 3 (10.7) | — | |||
| bla CTX-M | bla DHA | 1 (1.9) | 1 (3.6) | — | ||
| bla CTX-M bla TEM | bla DHA | 1 (1.9) | 1 (3.6) | — | ||
| bla CTX-M bla TEM | bla EBC | 1 (1.9) | — | 1 (4.0) | ||
| bla CTX-M | bla OXA-48 | 1 (1.9) | 1 (3.6) | — | ||
| bla CTX-M | bla NDM bla OXA-48 | 1 (1.9) | 1 (3.6) | — | ||
‘—’ means no observation.
Among the β-lactamases investigated, blaCTX-M (n = 45), blaEBC (n = 7) and blaOXA-48 (n = 5) were the most frequent ESBL, pAmpC β-lactamase, and carbapenemase among E. coli and K. pneumoniae, respectively (Table 4). Four and five isolates had the simultaneous presence of ESBL-carbapenemase and ESBL-pAmpC β-lactamase combination, respectively.
Among 45 neonatal BSIs caused either by E. coli or K. pneumoniae, 20 were ESBL, four were pAmpC β-lactamase, two were ESBL-pAmpC and two were ESBL-carbapenemase producers (Table 4). One carbapenem-resistant A. baumannii complex was confirmed to be positive for blaNDM.
Overall, ciprofloxacin non-susceptibility in E. coli (10/34, 29.4%) and K. pneumoniae (42/73, 57.5%) were significantly associated with β-lactamase production (χ2 = 34.8, P = 0.000). Consequently, non-susceptibility to ciprofloxacin was common among isolates with non-susceptibility to extended-spectrum cephalosporins (48/73; 65.8%) (Figure 2a). Analysis of β-lactamase producing isolates revealed additional higher proportions of non-susceptibility to gentamicin (22/53; 41.5%) and amikacin (15/53; 28.3%) (Figure 2b).
Figure 2.
Antibiogram of (a) 73 K. pneumoniae and 34 E. coli and (b) 53 β-lactamase-producing E. coli and K. pneumoniae isolates.
Characteristics of fatal BSIs
There were 20 deaths among the 146 patients with positive blood cultures who were hospitalized at SMH, of whom 17 were neonates, two were infants and one was a 12-year-old child. One each of the neonatal deaths was associated with an A. baumannii complex isolate harbouring blaNDM (newborn had hypoxic-ischaemic encephalopathy), CoNS (newborn had congenital heart disease, CHD) and E. coli with no acquired resistance (newborn had moderate to late prematurity). Of nine fatal K. pneumoniae BSIs (all ceftazidime resistant), four were ESBL positive, one was pAmpC β-lactamase positive and three had both ESBL and pAmpC β-lactamase. Three of these neonates were premature and had low birth weight (two very pre-term and one moderate to late pre-term). Another six neonatal deaths were associated with P. aeruginosa; one had birth asphyxia and five were premature (one moderate to late pre-term and four very pre-term) and of low birth weight. The remaining two deaths were associated with Serratia marscescens (neonate with CHD) and MRSA (children aged 12 years).
Discussion
We report the bacterial causes of BSIs in children attending a secondary healthcare level in the Kathmandu Valley, Nepal. In neonates, GNB, which were probably hospital-associated, were the main cause of BSI. In the non-neonates, most BSIs were community-acquired. Six of the WHO GLASS AMR pathogens (Acinetobacter spp., E. coli, K. pneumoniae, Salmonella spp., S. aureus and S. pneumoniae)2 comprised 337/377 (89.4%) of the BSIs. We found a high level of MDR in K. pneumoniae (51/73; 69.9%), E. coli (17/34; 50.0%) and S. aureus (33/98; 33.7%), in agreement with previous findings from Nepal.3–10
There were high levels of extended-spectrum cephalosporin resistance among BSIs in this population, particularly among E. coli and K. pneumoniae, and most of these infections were from children <5 years. These isolates were often also non-susceptible to fluoroquinolones and aminoglycosides.4,21 Carbapenem non-susceptibility levels of 3/34 (9%) in E. coli and 10/73 (14%) in K. pneumoniae were higher than in other settings such as in Cambodia,22 Malawi,23 but lower than in India.24 BlaCTX-M was the most common ESBL gene in E. coli and K. pneumoniae, consistent with findings from surveillances in Asia-Pacific,25 Thailand,21 Malawi23 and India.26 Surprisingly, blaEBC (7/10) was the most common pAmpC and we did not detect blaCIT (such as CMY-2). In another study of clinical isolates from Nepal, the positivity for blaCIT was 25/72 (34.7%) and 9/25 (36.0%) for E. coli and K. pneumoniae, respectively.27 BlaDHA was present in 25/72 (34.7%) and 10/25 (40.0%) among E. coli and K. pneumoniae, respectively. We found a small number of carbapenemase-harbouring isolates with blaNDM and blaOXA-48 which are common carbapenemases in Asia,21,25 including Nepal.4 The rare or low occurrence of blaKPC, blaIMP and blaVIM in Nepal is consistent with our observations.4 We also demonstrate the co-existence of ESBL-carbapenemase and ESBL-pAmpC β-lactamase combinations in individual isolates. A. baumannii complex and P. aeruginosa were mostly susceptible to antimicrobials tested with only two A. baumannii complex (5.0%) and one P. aeruginosa (6.3%) displaying MDR. This is in contrast to high levels of resistance reported in Acinetobacter spp. and P. aeruginosa in other studies in Nepal.4,8,9
S. aureus and S. pneumoniae were the first and third most common community-acquired BSIs in non-neonates. The proportion of MRSA in this study (29/98; 29.6%) was consistent with a pooled estimate for Nepal, 38.2% (95% CI: 31.4%–45.2%),6 and WHO’S GLASS reports’ median MRSA BSIs of 34.7%.2 Most MRSA BSI [26/29 (89.7%%)] were thought to be community-acquired, mostly among children aged 1 to <13 years, although it is possible previous hospital contact was not recorded. In view of the nationwide vaccine coverage of pneumococcal conjugate vaccine, it would be instructive to understand the pneumococcal conjugate vaccine status among these children and the serotype of the invasive isolate. Salmonella Typhi and Paratyphi A accounted for 58 (19.1%) of community-acquired infections. Ciprofloxacin non-susceptibility (23/58; 39.7%) was lower than reported in other South Asian countries.11 The recent introduction of the Vi capsular polysaccharide conjugate vaccine into the routine immunization programme should reduce the future burden of enteric fever among children in Nepal.
There were 20 deaths: 17 in neonates and two were infants. BSIs may have been a contributor rather than a direct cause of death since most children had underlying health conditions, mostly prematurity and low birth weight. Deaths associated with culture-proven BSI as a proportion of the total neonatal intensive care unit (NICU) admitted cases in our institution were low [17 of 961 (1.8%) NICU and 2 of 624 (0.3%) paediatric intensive care unit (PICU) admissions within the study period]. The high proportions of BSI in neonates non-susceptible to ampicillin and gentamicin and to extended-spectrum cephalosporins challenge empiric antimicrobial therapy for neonatal sepsis in this setting.28 Carbapenem-sparing options would be welcomed. Piperacillin-tazobactam, colistin and tigecycline could be options, but clinical data in neonates and young children are needed.
This study has limitations. The proportion of positive blood cultures was low, possibly due to the use of in-house blood culture methods for the outpatients. Also, we did not record the volume of blood inoculated in each instance and low blood volumes may have been contributory. We categorized hospital-acquired from community-acquired infections as per the CDC’s recommendations for children hospitalized at SMH, as advised in the MICRO guideline.20 In this retrospective study, we were unable to subdivide the community-onset infections in children with known co-morbidities and previous healthcare contact (healthcare-associated) and those in otherwise healthy children (community-acquired). We excluded most Coagulase-negative staphylococci BSIs even though a small proportion of these cases in neonates may have been true pathogens. We did not determine other mechanisms of resistance in extended-spectrum cephalosporin-non-susceptible isolates, such as efflux pump over-expression, porin alterations and the presence of other resistance genes not covered by the assay.
MDR BSI is a significant challenge in the paediatric population in Nepal. K. pneumoniae, E. coli and S. aureus were among the leading BSI-causing pathogens and also demonstrated high levels of MDR. The former two were found to be mostly ESBL producers and our data suggest the need for carbapenem-sparing options for the treatment of ESBL BSIs particularly in neonatal sepsis. Nationwide AMR surveillance, including in secondary and primary care settings, and antimicrobial stewardship initiatives are necessary to combat AMR in Nepal.
Supplementary Material
Acknowledgements
We would like to thank Mr Anil Rajbhandari for supporting this study. We also extend our thanks to laboratory members of the microbiology department, NICU and PICU staff who offered valuable help in data collection.
Contributor Information
Raj Kumar Shrestha, Department of Research, Siddhi Memorial Hospital, Bhaktapur, Nepal.
Dhruba Shrestha, Department of Research, Siddhi Memorial Hospital, Bhaktapur, Nepal.
Ashok Kumar Sah, Department of Research, Siddhi Memorial Hospital, Bhaktapur, Nepal.
Ashmita Thapa, Department of Research, Siddhi Memorial Hospital, Bhaktapur, Nepal.
Nipun Shrestha, Department of Research, Siddhi Memorial Hospital, Bhaktapur, Nepal.
Ganendra Bhakta Raya, Department of Research, Siddhi Memorial Hospital, Bhaktapur, Nepal.
Kenshi Furushima, School of Tropical Medicine and Global Health (TMGH), Nagasaki University, Nagasaki, Japan.
Bhim Gopal Dhoubhadel, School of Tropical Medicine and Global Health (TMGH), Nagasaki University, Nagasaki, Japan; Department of Respiratory Infections, Institute of Tropical Medicine, Nagasaki University, Nagasaki, Japan.
Christopher M Parry, School of Tropical Medicine and Global Health (TMGH), Nagasaki University, Nagasaki, Japan; Clinical Sciences and Education, Liverpool School of Tropical Medicine, Liverpool, UK.
Funding
This study was supported by internal funding.
Transparency declarations
None to declare.
Data availability
All data generated or analysed during this study are included in this published article.
Supplementary data
Tables S1, S2, and S3 are available as Supplementary data at JAC-AMR Online.
References
- 1. Murray CJL, Ikuta KS, Sharara F et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet 2022; 399: 629–55. 10.1016/S0140-6736(21)02724-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. WHO . Global Antimicrobial Resistance and use Surveillance System (GLASS) Report 2022. World Health Organization, 2022. Licence: CC BY-NC-SA 3.0 IGO. https://www.who.int/publications/i/item/9789240062702. [Google Scholar]
- 3. Zellweger RM, Basnyat B, Shrestha P et al. Changing antimicrobial resistance trends in Kathmandu, Nepal: a 23-year retrospective analysis of bacteraemia. Front Med (Lausanne) 2018; 5: 262. 10.3389/fmed.2018.00262 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Manandhar S, Zellweger RM, Maharjan N et al. A high prevalence of multi-drug resistant Gram-negative bacilli in a Nepali tertiary care hospital and associated widespread distribution of extended-spectrum beta-lactamase (ESBL) and carbapenemase-encoding genes. Ann Clin Microbiol Antimicrob 2020; 19: 48. 10.1186/s12941-020-00390-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kayastha K, Dhungel B, Karki S et al. Extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella species in pediatric patients visiting international friendship children’s hospital, Kathmandu, Nepal. Infect Dis (Auckl) 2020; 13: 1178633720909798. 10.1177/1178633720909798 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Khanal A, CS G, Gaire A et al. Methicillin-resistant Staphylococcus aureus in Nepal: a systematic review and meta-analysis. Int J Infect Dis 2021; 103: 48–55. 10.1016/j.ijid.2020.11.152 [DOI] [PubMed] [Google Scholar]
- 7. Manandhar S, Amatya P, Ansari I et al. Risk factors for the development of neonatal sepsis in a neonatal intensive care unit of a tertiary care hospital of Nepal. BMC Infect Dis 2021; 21: 546. 10.1186/s12879-021-06261-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Pandey R, Mishra SK, Shrestha A. Characterisation of ESKAPE pathogens with special reference to multidrug resistance and biofilm production in a Nepalese Hospital. Infect Drug Resist 2021; 14: 2201–12. 10.2147/IDR.S306688 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Parajuli NP, Acharya SP, Mishra SK et al. High burden of antimicrobial resistance among Gram negative bacteria causing healthcare associated infections in a critical care unit of Nepal. Antimicrob Resist Infect Control 2017; 6: 67. 10.1186/s13756-017-0222-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Yadav NS, Sharma S, Chaudhary DK et al. Bacteriological profile of neonatal sepsis and antibiotic susceptibility pattern of isolates admitted at Kanti Children’s Hospital, Kathmandu, Nepal. BMC Res Notes 2018; 11: 301. 10.1186/s13104-018-3394-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Gandra S, Alvarez-Uria G, Turner P et al. Antimicrobial resistance surveillance in low- and middle-income countries: progress and challenges in eight South Asian and Southeast Asian countries. Clin Microbiol Rev 2020; 33: e00048-19. 10.1128/CMR.00048-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Fink G, D’Acremont V, Leslie HH et al. Antibiotic exposure among children younger than 5 years in low-income and middle-income countries: a cross-sectional study of nationally representative facility-based and household-based surveys. Lancet Infect Dis 2020; 20: 179–87. 10.1016/S1473-3099(19)30572-9 [DOI] [PubMed] [Google Scholar]
- 13. Clinical and Laboratory Standards Institute . Performance Standards for Antimicrobial Susceptibility Testing—Thirty-Second Informational Supplement: M100-S30. 2022. [Google Scholar]
- 14. EUCAST . Clinical Breakpoints. http://www.eucast.org/clinical_breakpoints.
- 15. Magiorakos AP, Srinivasan A, Carey RB et al. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance. Clin Microbiol Infect 2012; 18: 268–81. 10.1111/j.1469-0691.2011.03570.x [DOI] [PubMed] [Google Scholar]
- 16. Lewis JS II, Herrera M, Wickes B et al. First report of the emergence of CTX-M-type extended-spectrum beta-lactamases (ESBLs) as the predominant ESBL isolated in a U.S. health care system. Antimicrob Agents Chemother 2007; 51: 4015–21. 10.1128/AAC.00576-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Dallenne C, Da Costa A, Decré D et al. Development of a set of multiplex PCR assays for the detection of genes encoding important beta-lactamases in Enterobacteriaceae. J Antimicrob Chemother 2010; 65: 490–5. 10.1093/jac/dkp498 [DOI] [PubMed] [Google Scholar]
- 18. Poirel L, Walsh TR, Cuvillier V et al. Multiplex PCR for detection of acquired carbapenemase genes. Diagn Microbiol Infect Dis 2011; 70: 119–23. 10.1016/j.diagmicrobio.2010.12.002 [DOI] [PubMed] [Google Scholar]
- 19. Pérez-Pérez FJ, Hanson ND. Detection of plasmid-mediated AmpC beta-lactamase genes in clinical isolates by using multiplex PCR. J Clin Microbiol 2002; 40: 2153–62. 10.1128/JCM.40.6.2153-2162.2002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Turner P, Fox-Lewis A, Shrestha P et al. Microbiology investigation criteria for reporting objectively (MICRO): a framework for the reporting and interpretation of clinical microbiology data. BMC Med 2019; 17: 70. 10.1186/s12916-019-1301-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Yungyuen T, Chatsuwan T, Plongla R et al. Nationwide surveillance and molecular characterization of critically drug-resistant gram-negative bacteria: results of the research university network Thailand study. Antimicrob Agents Chemother 2021; 65: e0067521. 10.1128/AAC.00675-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Fox-Lewis A, Takata J, Miliya T et al. Antimicrobial resistance in invasive bacterial infections in hospitalized children, Cambodia, 2007–2016. Emerg Infect Dis 2018; 24: 841–51. 10.3201/eid2405.171830 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Lester R, Musicha P, Kawaza K et al. Effect of resistance to third-generation cephalosporins on morbidity and mortality from bloodstream infections in Blantyre, Malawi: a prospective cohort study. Lancet Microbe 2022; 3: e922–30. 10.1016/S2666-5247(22)00282-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Mathur P, Malpiedi P, Walia K et al. Health-care-associated bloodstream and urinary tract infections in a network of hospitals in India: a multicentre, hospital-based, prospective surveillance study. Lancet Glob Health 2022; 10: e1317–25. 10.1016/S2214-109X(22)00274-1 [DOI] [PubMed] [Google Scholar]
- 25. Jean SS, Hsueh PR, SMART Asia-Pacific Group . Distribution of ESBLs, AmpC β-lactamases and carbapenemases among Enterobacteriaceae isolates causing intra-abdominal and urinary tract infections in the Asia-Pacific region during 2008–14: results from the study for monitoring antimicrobial resistance trends (SMART). J Antimicrob Chemother 2017; 72: 166–71. 10.1093/jac/dkw398 [DOI] [PubMed] [Google Scholar]
- 26. Patil S, Wagh K, Lopes B et al. ESBL encoding third-generation cephalosporin resistance observed in bloodstream infection in India. J Hosp Infect 2023; 133: 98–9. 10.1016/j.jhin.2022.12.019 [DOI] [PubMed] [Google Scholar]
- 27. Aryal SC, Upreti MK, Sah AK et al. Plasmid-mediated AmpC β-lactamase CITM and DHAM genes among Gram-negative clinical isolates. Infect Drug Resist 2020; 13: 4249–61. 10.2147/IDR.S284751 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Russell NJ, Stöhr W, Plakkal N et al. Patterns of antibiotic use, pathogens, and prediction of mortality in hospitalized neonates and young infants with sepsis: a global Neonatal Sepsis Observational Cohort Study (NeoOBS). PLoS Med 2023; 20: e1004179. doi: 10.1371/journal.pmed.1004179 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data generated or analysed during this study are included in this published article.