A 23-year retrospective investigation of Salmonella Typhi and Salmonella Paratyphi isolated in a tertiary Kathmandu hospital

Abstract

Background: Salmonella serovars Typhi (S. Typhi) and Paratyphi A (S. Paratyphi A), the causative agents of enteric fever, have been routinely isolated organisms from the blood of febrile patients in the Kathmandu Valley since the early 1990s. Susceptibility against commonly used antimicrobials for treating enteric fever has gradually changed throughout South Asia since this time, posing serious treatment challenges. Here, we aimed to longitudinally describe trends in the isolation of Salmonella enterica and assess changes in their antimicrobial susceptibility in Kathmandu over a 23-year period.

Methods: We conducted a retrospective analysis of standardised microbiological data from April 1992 to December 2014 at a single healthcare facility in Kathmandu, examining time trends of Salmonella-associated bacteraemia and the corresponding antimicrobial susceptibility profiles of the isolated organisms.

Results: Over 23 years there were 30,353 positive blood cultures. Salmonella enterica accounted for 65.4% (19,857/30,353) of all the bacteria positive blood cultures. S. Typhi and S. Paratyphi A were the dominant serovars, constituting 68.5% (13,592/19,857) and 30.5% (6,057/19,857) of all isolated Salmonellae. We observed (i) a peak in the number of Salmonella-positive cultures in 2002, a year of heavy rainfall and flooding in the Kathmandu Valley, followed by a decline toward pre-flood baseline by 2014, (ii) an increase in the proportion of S. Paratyphi in all Salmonella-positive cultures between 1992 and 2014, (iii) a decrease in the prevalence of MDR for both S. Typhi and S. Paratyphi, and (iv) a recent increase in fluoroquinolone non-susceptibility in both S. Typhi and S. Paratyphi isolates.

Conclusions: Our work describes significant changes in the epidemiology of Salmonella enterica in the Kathmandu Valley during the last quarter of a century. We highlight the need to examine current treatment protocols for enteric fever and suggest a change from fluoroquinolone monotherapy to combination therapies of macrolides or cephalosporins along with older first-line antimicrobials that have regained their efficacy.

Fig 1

The annual distribution of bacteria positive blood cultures from 1992 to 2014. Bar graph showing the annual counts (A) and percentages (B) of Salmonella-positive cultures in all positive blood samples collected from 1992 to 2014. The annual counts and percentages of S. Typhi, S. Paratyphi, and Salmonella spp. in all Salmonella-positive cultures collected from 1992 to 2014 are shown in (C) and (D), respectively.

Fig 2

The annual distribution of ward from which Salmonella positive blood cultures were taken from 1992 to 2014. Annual counts (A) and percentages (B) of the ward of origin for S. Typhi positive cultures collected from 1992 to 2014. The annual counts and percentages of the ward of origin for the S. Paratyphi positive cultures collected from 1992 to 2014 are shown in (C) and (D), respectively.

Fig 3

Time series for Salmonella-positive cultures from 1992 to 2014. Plots showing the monthly time series of all Salmonella positive blood cultures (absolute numbers in (A) and trends in (B)) between April 1992 and December 2014. Time series (absolute numbers and trends) for S. Typhi are shown in (C) and (D), and for S. Paratyphi in (E) and (F).

Fig 4

The antimicrobial susceptibility profiles of isolated Salmonella from 1992 to 2014. Plots summarising antimicrobial susceptibility data for all Salmonella (A), S. Typhi only (B), and S. Paratyphi only (C) from 1992 to 2014. The colour range on the heat map tiles indicate the proportion of non-susceptible isolates (blue, low; red; high), the number of isolates tested is indicated on the respective tile. Tiles with fewer than 10 isolates tested for the respective antimicrobial are not shown.

Fig 5

The development of non-susceptibility against selected antimicrobials and multidrug resistance. Plots showing changes in non-susceptibility against i) nalidixic acid for S. Typhi (A) and S. Paratyphi A (B), ii) against ciprofloxacin for S. Typhi (C) and S. Paratyphi (D), iii) against cefotaxime for S. Typhi (E) and S. Paratyphi A (F), and iv) multidrug resistance for S. Typhi (G) and S. Paratyphi (H) between 1992 and 2014. The linear trend-lines (blue) and 95% CI for the regression lines (shaded) are overlaid on the data points. The slope of the regression line (with 95% CI) is indicated on each plot; two regression lines are present when a breakpoint in the slope was detected.