Rapid Identification and Antibiotic Susceptibility Testing of Salmonella enterica Serovar Typhi Isolated from Blood: Implications for Therapy

Abstract

The turnaround time (TAT) for Salmonella enterica serovar Typhi identification and reporting of the antibiotic susceptibility profile was determined for 391 cases of typhoid fever, using the lysis direct plating or lysis centrifugation method of blood culture along with rapid antimicrobial susceptibility testing. The TAT was more rapid (TAT for 90% of the patients [TAT⁹⁰] = 30 h; TAT¹⁰⁰ ≤ 67 h) than was possible with conventional methodologies and was equivalent to that reported previously using more advanced, costly technologies that are largely unavailable in developing countries. Antibiotic susceptibility profiles, determined by the rapid antimicrobial susceptibility testing method, of randomly selected 60 S. enterica serovar Typhi isolates were identical to those determined by overnight conventional testing. Preliminary assessment of the impact of the reduced TAT on physician practices revealed that initial empirical therapy was prescribed at the time of presentation in most cases (87 of 108 [81%]) despite awareness that the final report would be available on the following day. Patients treated empirically with first-line antibiotics and shown subsequently to be infected with a multidrug-resistant strain benefited most (8 cases), since therapy was changed appropriately on the following day. In an additional 21 cases, therapy with an appropriate antibiotic was initiated after culture results were available. Thus, in approximately one-fourth (29 of 108 [27%]) of the cases, a change in management to an agent active for treatment of the isolate was made after receipt of the test results. However, in no case was therapy changed from a second-line to a first-line agent, even if the isolate was reported on the day after presentation to be sensitive to first-line therapy (33 cases). Ways in which to utilize rapid-TAT result reporting in order to positively influence physicians' prescribing in Bangladesh are the subject of ongoing research.

Despite the increased availability of innovative molecular technologies for detecting and reporting microbial pathogens, most clinicians still regard the antimicrobial susceptibility report as the most important test result generated by the clinical microbiology laboratory (3). Several studies have shown that antimicrobial susceptibility results, however, generally have not significantly influenced physicians' choices of appropriate antimicrobial therapies (5). The long turnaround time (TAT) required to generate the report using conventional methods and poor communication between the laboratory and physicians (3) likely contribute to the lack of impact of the antibiotic susceptibility report on clinical practice. Rapid identification of the etiologic agent of infection and its antimicrobial susceptibility profile by using advanced laboratory technology and instant dissemination of the information through hospital- and laboratory-based computer network systems could rationalize antimicrobial therapy. These methods, however, are largely unavailable in the developing world. Thus, the development of low-technology, low-cost methods for reducing TAT and improving flow of information to physicians is needed to effectively treat infections while minimizing unwarranted use of broad-spectrum antibiotics in the developing world.

The goals of this study were (i) to evaluate the TAT of the lysis direct plating-lysis centrifugation (LDP-LC) method of blood culture, along with rapid antimicrobial susceptibility testing (RAST), for processing isolates of Salmonella enterica serovar Typhi and (ii) to perform a preliminary assessment of the impact of the rapid TAT of the LDP-LC–RAST method on the therapeutic management of typhoid fever by clinicians.

MATERIALS AND METHODS

Patients.

Patients were referred by community practitioners, clinics, and hospitals throughout Dhaka City, Bangladesh, to the Popular Diagnostic Centre, a private laboratory and clinical diagnostic facility, for evaluation of possible serious bacterial infection from August 1998 to November 1999 (9). Isolates of S. enterica serovar Typhi from these patients were chosen for further study.

Laboratory procedure.

Blood cultures were performed by the LDP-LC method (10) with 2 ml of blood from patients up to 12 years of age and 5 ml of blood from patients in older age groups. Specimens were cultured on sheep blood and MacConkey agar plates, and the number of bacteria per milliliter of blood was determined by plate counting. Suspected colonies of S. enterica serovar Typhi were immediately identified by agglutination with specific antisera (Murex, Dartford, United Kingdom), and RAST was performed as described by Granato (3). In brief, instead of conventional overnight incubation of susceptibility test plates, the readings for antibiotic sensitivity results were obtained after 8 h of incubation (Table 1). If agglutination testing was negative or doubtful, identification was performed by standard biochemical tests and repeat agglutination testing 24 h later (9). In these cases, however, the growth was reported generically as gram-negative bacilli, and RAST results were provided with a note that precise identification of the isolate would be confirmed the following day. Antibiograms of the strains were interpreted according to National Committee for Clinical Laboratory Standards recommendations (7). Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used for quality control. The antibiograms of randomly selected multidrug-resistant (MDR) (n = 40) and multidrug-susceptible (n = 20) strains of S. enterica serovar Typhi were obtained by both RAST and conventional antibiotic sensitivity testing methods to determine their concordance (7).

TABLE 1.

TATs obtained by LDC-LC

Day and exam	Time	Incubation period (h)	Final report time with antibiogram	TAT (h)
1 (inoculation)	10 a.m. to 10 p.m.
2 (plate examination)
First exam	10 a.m.	12	6 p.m.	≤30
Second exam	6 p.m.	20–32	9 a.m. (third day)	≤47
3 (plate examination)
First exam	10 a.m.	36–48	6 p.m.	≤54
Second exam	8 p.m.	46–58	9 a.m. (fourth day)	≤67

Physicians' practice.

Several seminars and meetings were organized (i) to inform referring physicians about the rapid reporting system and the availability of isolate identification and antibiogram results within 24 h of obtaining blood cultures and (ii) to encourage them to refrain, when feasible, from prescribing empirical antibiotic therapy until after receiving the antibiogram; to utilize first-line therapy when prescribing antibiotics empirically; and to make appropriate and prompt adjustments in therapy once the antibiogram was available. Impact of rapid reporting on management of the S. enterica serovar Typhi-infected patients of three of the physicians with the busiest practices was determined retrospectively by either reviewing the physician's prescriptions or contacting them personally to determine prescribing patterns.

RESULTS

A total of 4,650 blood cultures were received in the 15-month study period. Of these cultures, 538 (11.6%) produced significant growth, and 72.7% (391 of the 538) of this group of cultures were determined to be S. enterica serovar Typhi, as reported previously (9). More than 90% (342 of 391) of the typhoid cases were identified by the LDP-LC method within 12 to 24 h of inoculation (by 10 a.m. on the second day [Table 1]), and RAST results were available within the subsequent 8 h. The TAT¹⁰⁰ value (i.e., the TAT for all [100%] of the patients) was ≤67 h (Fig. 1). The susceptibilities of 60 S. enterica serovar Typhi strains to ampicillin, cotrimoxazole, chloramphenicol, ciprofloxacin, and ceftriaxone were found to be identical by the RAST method to those obtained by overnight conventional testing (data not shown).

FIG. 1.

TAT for culture-positive cases of typhoid

Follow-up of treatment by physicians revealed that empirical therapy was started in 81% of the cases evaluated (87 of 108) before the culture sensitivity report was available (Table 2). In approximately half (47 of 87 [54%]) of these cases, the empirical choice was an appropriate first-line antibiotic (amoxicillin, cotrimoxazole, or chloramphenicol). In only 17% (8 of 47) of these cases was it necessary to change from a first-line to a second-line agent (ceftriaxone or ciprofloxacin), either because the isolate was MDR (n = 3) or because it displayed isolated resistance to the empirically prescribed first-line agent (n = 5). Moreover, in 21 cases, therapy with an appropriate antibiotic was initiated after culture results were available. Thus, in a total of 27% (29 of 108) of the cases, a change in management to an agent active for treatment of the isolate was made after the receipt of test results. On the other hand, empirical therapy was initiated in 40 cases with a second-line antibiotic (ciprofloxacin, n = 17; ceftriaxone, n = 23); 33 (83%) of these isolates, however, were sensitive to the first-line antibiotic. In none of these cases was the prescribed antibiotic changed to a first-line antibiotic after the susceptibility report was received. However, there was no difference in final outcome since all patients were cured without any complications.

TABLE 2.

Resistance pattern and impact of RAST on antibiotic treatment in 108 typhoid fever cases

Resistance patterna	Total no.	Initial antibiotic	No.	Antibiotic used after receipt of culture and sensitivity report
Ampr	1	Ceftriaxone	1	Unchanged
Ampr Cotr	3	Ceftriaxone	1	Unchanged
		Ciprofloxacin	2	Unchanged
Cotr	7	Amoxicillin	1	Unchanged
		Cotrimoxazole	2	Changed to second-line antibiotic
		Ceftriaxone	1	Unchanged
		Ciprofloxacin	1	Unchanged
		None	2	Treated with amoxicillin
Cotr Chlr	2	Amoxicillin	1	Unchanged
		Ceftriaxone	1	Unchanged
MDR (Ampr Cotr Chlr)	15	Amoxicillin	3	Changed to second-line antibiotic
		Cotrimoxazole	3	Changed to second-line antibiotic
		Ceftriaxone	4	Unchanged
		Ciprofloxacin	3	Unchanged
		None	2	Treated with second-line antibiotic
Sensitive	80	Amoxicillin	16	Unchanged
		Cotrimoxazole	21	Unchanged
		Ceftriaxone	15	Unchanged
		Ciprofloxacin	11	Unchanged
		None	17	Treated with first-line antibiotic

^aAmp^r, ampicillin resistant; Cot^r, cotrimazole resistant; Chl^r, chloramphenicol resistant. 

DISCUSSION

Typhoid fever is endemic in Bangladesh, where there is a high incidence in children (9). The emergence of MDR S. enterica serovar Typhi isolates in the early 1990s, particularly from the Indian subcontinent, prompted the suggestion that ciprofloxacin or ceftriaxone should be the drug of choice for empirical treatment of typhoid fever (4, 8, 11). Initially, reduced use of amoxicillin, cotrimoxazole, or chloramphenicol was associated with a decreased prevalence of MDR strains (12), but more recently, continued dependence on ciprofloxacin or ceftriaxone for the empirical treatment of typhoid fever in Bangladesh and elsewhere has led to the emergence of resistance of S. enterica serovar Typhi to these drugs (6, 13).

In this context of changing the dynamics of resistance to antibiotics, it is imperative for optimal patient care that accurate information on S. enterica serovar Typhi isolation and its antibiotic susceptibility pattern be available to the clinician as rapidly as possible. A recent study demonstrated that faster reporting with advanced technology (e.g., the Vitek system for isolate identification and antibiotic susceptibility testing or automated, computerized laboratory information systems), along with changes in the workflow, could save hospital costs and improve patient care (1). The advanced equipment needed for such rapid reporting is expensive, however, and incurs a high recurring cost. Consequently, these methods are not yet available in most clinical laboratories in developing countries. Even in the developed part of the world, the use of these advanced technologies is feasible only for relatively large laboratories and hospitals.

We showed here that the TAT for the identification and anitimicrobial sensitivity testing of S. enterica serovar Typhi can be substantially reduced to the level of modern technologies by using rapid but relatively low-cost, low-technology blood culture and susceptibility testing methods. A TAT⁹⁰ of ≤30 h achieved using these methods was at least as rapid as that of advanced “state-of-the-art” methodologies (1, 14). On the other hand, the minimum TAT achievable by conventional broth culture would be 96 h (10, 15). The rapid method described here did not require any additional equipment or laboratory personal and produced antibiogram results identical to those obtained by conventional methods.

In Bangladesh, empirical therapy is the rule rather than the exception (2). In the present study, >80% of cases evaluated were treated empirically, even though a report was expected within the next 24 h and the referring physicians had received instructions about the methodology and the benefits of waiting for test results before initiating therapy. In more than one-fourth of these cases, however, including patients infected with MDR strains, a change from ineffective to effective therapy was made after the receipt of test results and the ultimate outcome was good since the patients recovered without complications. On the other hand, the availability of results did not result in a change if therapy was initiated with a relatively expensive second-line empirical antibiotic, even though the isolate was susceptible to less-expensive first-line agents.

These data are preliminary and do not include an evaluation of prescribing patterns for either culture-negative patients or for typhoid patients before introduction of the rapid-TAT system, both of which would provide additional important insights into the influence of rapid TAT on prescribing patterns. Finding ways in which to utilize reporting of rapid-TAT results to positively influence physician prescribing in Bangladesh is the subject of ongoing research.

In summary, a reduced TAT for isolate identification and antibiogram results by using the rapid LDP-LC–RAST method was achieved with equal precision but without additional costs compared to conventional methodologies. This approach may improve the quality and cost of patient care in the developing world, where typhoid fever is endemic, MDR strains are relatively common, and the availability of advanced diagnostic laboratory methods is limited.