Description of the condition
Definition of enteric fever
Enteric fever is caused by either Salmonella enterica serovar Typhi (S. Typhi; typhoid fever) or Salmonella enterica serovar Paratyphi (S. Paratyphi) A, B, or C (paratyphoid fever), and results in a systemic illness specific to humans.
In the year 2000, typhoid fever was responsible for an estimated 21.7 million global episodes of disease, of which 1% of cases resulted in death, and paratyphoid fever caused a further 5.4 million cases of enteric fever (Crump 2010). The greatest number of cases were caused in children and adolescents in South-Central and South-East Asia (Crump 2004). Both typhoid and paratyphoid fever present as clinically similar, febrile, systemic conditions and the capacity to accurately diagnose these infections is limited by the poor availability of facilities to undertake bone marrow and blood culture in endemic areas, the sensitivity of blood culture as a diagnostic test, and the sensitivity and specificity of currently available serological tests or molecular tests (Baker 2010).
Despite the limitations of the data obtained from enteric fever surveillance, several novel epidemiological features have been observed in recent years. In sub-Saharan Africa, outbreaks of disease with a major risk of intestinal perforation have been described (Muyembe-Tamfum 2009), although infections with non-typhoidal salmonellae are much more common (Gordon 2008; Feasey 2010; Feasey 2012). The high incidence of typhoid fever has been confirmed in Asian and South-East Asian countries, particularly in children, but pan-regional surveillance has demonstrated substantial variation between sites (Ochiai 2008). The incidence of S. Paratyphi A infection appears to be increasing as a cause of enteric fever in Asia, accounting for 50% of bloodstream isolates in enteric fever patients, with implications for the development of a useful vaccine (Vidyalakshmi 2008; Dong 2010). The success of improvements in water and sanitation in preventing enteric fever has been highlighted in Latin America, where the incidence of this disease has declined in response to targeted public health programmes (Crump 2004).
The epidemiology of drug resistance has also changed over the years. Resistance to the older antimicrobials chloramphenicol, ampicillin and co-trimoxazole has been present for many years (Rowe 1997). This led to the use of fluoroquinolones in the management of enteric fever, the subsequent emergence of nalidixic acid resistance (NaR) and decreased susceptibility to fluoroquinolones, and eventually full resistance to fluoroquinolones (Lynch 2009). In fluoroquinolone-resistant isolates a third generation cephalosporin, frequently ceftriaxone, is often the drug of choice (Basnyat 2007). However, sporadic cases of third generation cephalosporin resistance have been reported and there is concern that widespread resistance will leave treatment options severely restricted (Pokharel 2006; Al-Naiemi 2008).
S. Typhi and S. Paratyphi are typically transmitted by the ingestion of contaminated food or water and contamination occurs when bacteria are shed in the faeces of individuals who are acutely unwell, convalescing or are chronic carriers (Bhan 2005). The severity of the infection depends on the initial infective dose, the virulence of the organism and the host's immune response (Adams 1999). The bacteria usually penetrate the intestinal mucosa and proliferate in the underlying lymphoid tissue, from where they disseminate via the lymphatic system or are released into the bloodstream, or both, resulting in spread to other organs. These organs most commonly include the liver, spleen, bone marrow and gall bladder. The bacteria are predominantly intracellular and are able to survive inside monocyte-macrophage cells (Parry 2002; Raffatellu 2008).
Clinical features and diagnosis
The clinical features of enteric fever typically include progressive intermittent fever, headache, abdominal discomfort, anorexia, hepatomegaly, splenomegaly and rose-coloured spots on the torso (Bhan 2005; Walia 2006). It is not possible to distinguish between typhoid and paratyphoid fever on the basis of clinical symptoms.
Complications of enteric fever occur in 10 to 15% of cases and are more frequent in patients whose illness has lasted more than two weeks and can affect multiple organ systems (Parry 2002; Bhan 2005). Intestinal perforation and haemorrhage, shock, pancreatitis, cholecystitis, pneumonia, myocarditis and encephalopathy have all been described (Bhan 2005).
Microbiological diagnosis is typically confirmed by culture and its sensitivity is dependent on the type of specimen analysed. Blood cultures are positive in 40% to 80% of patients and faecal cultures are positive in 30% to 40% of patients, with the yield from the latter being highest during the second or third week of illness. Culture of urine, rose spots and duodenal contents can also be undertaken (Wain 2008). However, bone marrow culture is the most sensitive test (Farooqui 1991) and cultures can remain positive for up to five days after initiation of the appropriate antibiotic treatment.
Bacterial culture also facilitates antimicrobial susceptibility testing which is helpful in guiding the appropriate antibiotic therapy. Disc diffusion breakpoints incorporating an "intermediate" or decreased ciprofloxacin susceptibility (DCS) category have now been published (CLSI 2012). Alternative options include a test for nalidixic acid susceptibility or confirmation of the ciprofloxacin minimum inhibitory concentration (MIC); NaR organisms have reduced susceptibility to ciprofloxacin (Crump 2003). Susceptibility breakpoints for azithromycin are being developed (Sjolund-Karlsson 2011).
The benefit of serological tests for the diagnosis of enteric fever is limited by the persistence of positive results following on from previous infection. Newer diagnostic tests using ELISA, immunochromatographic platforms and nucleic acid amplification testing are in development, but none have proven to be sensitive and specific enough to be widely adopted in routine clinical diagnostics (Parry 2002).
Mortality due to untreated typhoid is approximately 10% to 15%, and is highest in children below one year of age and the elderly. If the condition is treated with an effective antimicrobial, mortality rates are approximately 1% to 2% (Bhan 2005). Notably, mortality related to complications of the disease depend on the organ system affected and the capacity to provide relevant specialist and supportive care. Mortality rates are greater in resource-poor settings.
Between 1% to 5% of patients develop chronic carriage of salmonellae following infection (defined as over 12 months excretion in faeces or urine) (Ferreccio 1988). Chronic carriage occurs more frequently in women, and in patients with gallstones or other biliary tract abnormalities (Levine 1982). Biliary carriage has been associated with an increased risk of cancer (Caygill 1994).
Antimicrobial monotherapy is usually used to treat enteric fever, but the optimal choice of drug and duration of therapy are uncertain. Combination therapy, such as ceftriaxone/ciprofloxacin, has also been commonly used in the USA (Crump 2008). In the context of multiple-drug resistance (MDR) to first line agents (amoxicillin/ampicillin, cotrimoxazole and chloramphenicol), fluoroquinolones were considered the drugs of choice. However, due to reduced drug susceptibility and complete fluoroquinolone resistance becoming widespread in endemic areas, parenteral therapy with a third generation cephalosporin, such as ceftriaxone, or treatment with azithromycin have become more common (Basnyat 2010). A Cochrane review of the use of fluoroquinolones and azithromycin in the treatment of typhoid has previously been undertaken (Effa 2011), but the use of broad-spectrum beta-lactams such as third and fourth generation cephalosporins has not been systematically investigated to date.
Recommendations for the duration of therapy include five to 10 days for oral treatment with a fluoroquinolone or azithromycin, and seven to 14 days for beta-lactams as there may be a greater risk for relapse with shorter durations of beta-lactam therapy. However, the advice on duration of treatment and the relative benefits of various cephalosporins are not clear from individual studies. Some studies suggest ceftriaxone may have a benefit over cefotaxime, and that oral cefixime has similar treatment efficacy to ceftriaxone.
Description of the intervention
Cephalosporins are beta-lactam compounds in which the beta-lactam ring has been fused to a 6-membered dihydrothiazine ring, forming the cephem nucleus. Historically, they have been divided into four generations depending on their antibacterial spectrum of activity, typically with greater Gram-negative coverage with each successive generation at the expense of Gram-positive coverage (Kalman 1990). More recently 5
Cephalosporins inhibit cell wall synthesis by binding to penicillin binding proteins, and are bactericidal. They have time-dependent killing activity, which requires levels that are continuously above the MIC of the pathogen being treated. Dosing frequency is variable, but some cephalosporins such as ceftriaxone have a distinct advantage in having a half-life sufficiently long to be given once daily (Kalman 1990).
Most cephalosporins distribute well into the extracellular fluid of most tissues, and some of the later-generation cephalosporins also sufficiently penetrate into cerebrospinal fluid to be used in the treatment of central nervous system infections. However, penetration of cephalosporins into the intracellular compartment is poor. Elimination is mostly through the renal system, although significant biliary excretion is a feature of some cephalosporins, such as ceftriaxone and cefoperazone. They are generally well-tolerated, although some patients may display hypersensitivity; hepatic dysfunction and interstitial nephritis can also be features (Kalman 1990).
Third and fourth generation cephalosporins typically have activity against salmonellae and can be used in the management of enteric fever. Plasmid-mediated extended-spectrum cephalosporin resistance has been widely reported in non-typhoidal salmonellae (Paterson 2006), with sporadic cases of third-generation cephalosporin resistance in S. Typhi isolates from Pakistan, Bangladesh and the Phillipines (Saha 1999; Al-Naiemi 2008; Abdullah 2012). The potential horizontal transfer of extended-spectrum beta-lactamase genes to Typhi and/or Paratyphi isolates is a real and major clinical concern.
Why it is important to do this review
Although broad-spectrum cephalosporins are widely used in the management of enteric fever, clear guidance on the most appropriate drug and treatment regimen amongst different patient groups is lacking. Cephalosporins are considered safe in children, whereas sufficient concern exists in relation to the safety profile of fluoroquinolones in paediatric medicine for them to remain unapproved by the US Food and Drug Administration (FDA). Developing a more comprehensive understanding of how these drugs can be best used in the treatment of enteric fever will have an impact on the cost of treatment, the improvement of clinical outcomes and the provision of baseline data for the development of future clinical trials.
To evaluate third and fourth generation cephalosporins in the treatment of enteric fever in children and adults compared with other antibiotics. The primary objectives are:
- to identify whether the drugs under scrutiny are clinically effective
- which is the optimal agent and at what dose
- what is the most appropriate route of administration
- what is the ideal treatment duration
Therefore, we will perform within-class comparisons and an assessment of third and fourth generation cephalosporin treatment duration.
Criteria for considering studies for this review
Types of studies
Randomized controlled trials
Types of participants
Adults and children diagnosed with typhoid or paratyphoid fever on the basis of blood culture, bone marrow culture or molecular tests.
Types of interventions
Treatment with a third or fourth generation cephalosporin antimicrobial with respect to drug type, treatment dose and treatment duration.
Cephalaoporins considered in this review are shown in Table 1.
Other antimicrobials, including:
- Different third or fourth generation cephalosporin antimicrobial
- Different treatment duration of the same third or fourth generation cephalosporin antimicrobial
Types of outcome measures
- Clinical failure - defined as the presence of symptoms or development of complications that necessitate a change in antibiotic therapy or prolongation of existing therapy at the time period specified by trial authors; or, death related to the disease as opposed to an adverse event arising from any therapy administered.
- Microbiological failure - defined as a positive culture from blood, bone marrow or any sterile anatomical site during the period specified by the trial authors
- Relapse - defined as the recurrence of symptoms with a positive culture from blood, bone marrow or any sterile anatomical site to the point of follow-up defined by the trial author
- Fever clearance time – defined as the time in hours or days taken to defervesce from the start of the intervention or control drug with the definition of fever clearance as defined by the trial authors
- Length of hospital stay – defined as the time in days from study entry until discharge from hospital
- Convalescent faecal carriage – defined as a positive faecal culture detected at any time after the end of treatment up to one year of follow-up
Adverse events for treatment regimens will be determined under the following categories where possible:
- Serious adverse events (SAEs) - these are defined as any untoward events thought to be related to the administration of the treatment under consideration, including those which:
- result in death
- are life-threatening
- require patient hospitalization or prolongation of existing hospitalization
- result in persistent or significant disability/incapacity or require intervention to prevent permanent impairment or damage
- Events requiring the discontinuation of treatment
- All other adverse events discussed by authors of the studies reviewed
Search methods for identification of studies
We will attempt to identify all potential studies regardless of language or publication status (published, unpublished, in press, and in progress).
We will search the following electronic databases using the search terms and strategy described in Appendix 1: Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library; MEDLINE; EMBASE; and LILACS. We will also search the metaRegister of Controlled Trials (mRCT) and the WHO International Clinical Trials Registry Platform (ICTRP) using "typhoid fever" and "cephalosporin*" as search terms.
Searching other resources
We will handsearch abstracts from the following annual meetings: International Symposium on Typhoid Fever and Other Salmonelloses; the Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC); the Infectious Diseases Society of America (IDSA); the Western Pacific Congress on Chemotherapy and Infectious Diseases; the European Congress of Clinical Microbiology and Infectious Diseases; and the American Society of Tropical Medicine and Hygiene.
We will contact researchers in the field to identify additional studies that may be eligible for inclusion.
We will check the reference lists of all studies identified by the above methods.
Data collection and analysis
Selection of studies
Nicole Stoesser (NS) and David Eyre (DE) will screen the title, abstract and keywords of each record identified using the search strategy. They will retrieve the full-text article for all potentially relevant studies and all studies where the relevance is unclear from screening.
NS and DE will independently apply the inclusion criteria to each of the studies which pass the screening process in order to determine their eligibility for inclusion.
Any disagreements will be resolved through discussion with Christopher Parry (CP) or Buddha Basnyat (BB). Also, attempts will be made to contact the study authors if there are further doubts. We will tabulate the excluded studies and reasons for exclusion in a “Characteristics of excluded studies” section. We will enter data from each eligible study only once.
Data extraction and management
NS and DE will independently extract data using a standardized, pre-tested data extraction form incorporating information about the study population, intervention used (type of drug, means of administration, duration of treatment) and outcomes (side effects, success of treatment).
For dichotomous outcomes, we will extract data such as clinical failure, the total number of participants randomized, the number of participants analysed and the total number of participants that experienced that event.
For continuous outcomes, we will extract data such as fever clearance time, the total number of participants, arithmetic means and standard deviations. If the standard deviation is not reported, we will attempt to use the confidence interval (CI) or P-value to calculate it.
We will attempt to contact authors for additional data if it is not available or not in the format required to undertake the analyses.
We will compare extracted data between review authors to identify errors. We will resolve any conflicts by discussion with CP or BB.
In addition, data will be entered into RevMan 5.0 by NS; correct data entry will be verified by DE.
Assessment of risk of bias in included studies
NS and DE will assess the methodological quality of each included study using the Cochrane Collaboration's tool for assessing risk of bias with reference to the following: sequence generation; allocation concealment; blinding of participants, personnel and outcome assessors (blinding to be defined as double (trial uses a placebo or a double dummy technique such that neither the participant or care provider/assessor know which treatment is given), single (participant or care provider/assessor is aware of the treatment given), or open (all parties are aware of treatment)); incomplete outcome data (follow-up will be considered adequate if 90% or more of the randomized culture-positive participants are in the final analysis and inadequate if this figure was less than 90%); selective outcome reporting (his will be considered present if any of the study’s pre-specified outcomes are not reported or not reported in the pre-specified way; other potential threats to validity (such as funding from pharmaceutical bodies or the stopping of the trial by an independent process); study and protocol registered; and protocol available on-line.
We will define the risk of bias as "low", "high" or "uncertain" in accordance with the assessment tool. We will resolve any disagreements through discussion with CP or BB.
Measures of treatment effect
We will calculate risk ratios (RR) for dichotomous data and mean differences (MD) for continuous data. We will present each result with a 95% CI.
Unit of analysis issues
We anticipate the following issues related to the units of analysis:
1. Studies with more than two treatment groups
We will analyse data using pair-wise comparisons. We will compare the group under investigation (third or fourth generation cephalosporin) with each alternative antibiotic and divide them into subgroups by specific third or fourth generation cephalosporin.
2. Multiple outcome events
We anticipate this to be of relevance to adverse events analysis and will be handled as described in the "Measures of treatment effect" section above.
We will undertake a count of adverse and serious adverse events by patient associated with particular treatment alternatives. We do not expect issues relating to study designs that feature cross-over, repeated measurements, cluster randomization or body-part randomization to be an problem within this review.
Dealing with missing data
In the event that data is missing, we will attempt to contact the original investigators to request missing data.
If more than 90% of the relevant data is missing, we will use this as a reason for exclusion.
In order to confirm the robustness of this methodological approach, we will perform a sensitivity analysis to identify the sensitivity of results to reasonable changes in these assumptions. We will examine the potential impact of missing data in the discussion section.
Intention-to-treat (ITT) analyses
We will apply ITT principles to cases where follow-up data are available. However, in situations where this is not the case, we will undertake the analyses on an available-case basis.
We will analyse data using Review Manager (RevMan), with measures of treatment effect recorded as per the "Measures of treatment effect" section. We will combine trials of different cephalosporins in the evaluation of different treatment durations.
We will present a stratified analysis by resistance type for all trials where these data are available (MDR; NaR or DCS, or both). MDR is defined as resistance to all three first-line antibiotics (chloramphenicol, co-trimoxazole and ampicillin or amoxicillin). Where this information is not available, we will stratify results according to the presence or absence of drug-resistant strains. If stratification is necessary, we will stratify trials into those that report the presence of resistance, the absence of resistance and those that do not report these distinctions. For the latter, we will attempt to contact the authors to identify if this data is available.
Given that antibiotic resistance will affect the performance of first-line antibiotics, we will undertake stratification of results by the presence of MDR strains for trials comparing:
- Third generation cephalosporins and first-line antibiotics
- Fourth generation cephalosporins and first-line antibiotics
Given that the presence of NaR/DCS strains will affect the performance of older fluoroquinolones and possibly the newer fluoroquinolones, we will perform stratification of results by the presence of NaR/DCS strains for trials comparing:
- Third generation cephalosporins and fluoroquinolones
- Fourth generation cephalosporins and fluoroquinolones
We will not undertake stratification for studies making comparisons within groups of cephalosporins.
Subgroup analysis and investigation of heterogeneity
We will undertake separate analyses for paediatric populations (0 to 16 years) and adults (> 16 years). We will endeavour to make contact with the trial author if these distinctions are unclear from the data. We also plan to perform subgroup analyses for hospitalization (hospitalized or not), presence of MDR/NaR/DCS and duration of treatment (see also "Stratification" section).
We will make an assessment of heterogeneity by visually inspecting the forest plots and by comparing the heterogeneity statistic, Q, with the Chi
Where there is no statistical heterogeneity, we will apply the fixed-effect model. If we detect heterogeneity but we still decide to pool the data, we will apply a random-effects model.
We will undertake sensitivity analyses for each of the risk of bias assessment factors. We will assess the presence of publication bias using a funnel plot if there are five or more studies available for analysis of each primary outcome.
We will undertake a sensitivity analysis of the data excluding any antimicrobials included in the initial analysis, but which have subsequently been removed from the market for clinical reasons, such as adverse events.
The editorial base of the Cochrane Infectious Diseases Group is funded by UKaid from the UK Government for the benefit of developing countries.
Appendix 1. Search Strategies
Contributions of authors
NS and DE drafted the protocol, with input from CP and BB.
Declarations of interest
NS and DE: None known
CP: Co-author of "A comparative study of ofloxacin and cefixime for treatment of typhoid fever in children" (Pediatric Infectious Disease Journal 1999; 18(3):245-8), which is likely to be assessed as part of the review process.
BB: Co-author of "An open randomized comparison of gatifloxacin versus cefixime for the treatment of uncomplicated enteric fever" (PLoS One 2007 2(6):e542), which is likely to be assessed as part of the review process.
Sources of support
- No sources of support supplied
- National Institute for Health Research, UK.DE is supported by a doctoral training fellowship
- Wellcome Trust, UK.NS was supported by a clinical research training fellowship