Typhoid Fever in an Inner City Hospital: A 5-Year Retrospective Review

Authors


Corresponding Author: Dimitrios Farmakiotis, MD, Department of Medicine/Division of Infectious Diseases, Baylor College of Medicine, 1709 Dryden Road, BCM 620, 5th Floor, Suite 5.90D, Houston, TX 77030-2348, USA. E-mail: dfarmakiotis@mdanderson.org; farmakio@bcm.edu

Abstract

Background

Typhoid is a leading cause of fever in returning travelers. The prevalence is highest in migrants visiting friends and relatives (VFR travelers) in the Indian subcontinent, where reports of resistance have been of concern. This study is a retrospective analysis of patients with typhoid, seen over a 5-year period, in a tertiary center that serves a large immigrant population.

Methods

Patients with blood cultures positive for Salmonella Typhi were identified between 2006 and 2010. Charts were reviewed for demographic data, travel history, symptoms and signs, basic laboratory results, susceptibility profiles, treatment, and clinical course. Resistance to nalidixic acid was used as a marker of decreased susceptibility to quinolones.

Results

Seventeen patients were identified with S Typhi. The median age was 12 years (range: 2–47 y) and 94% (16 of 17) were hospitalized with a median stay of 7 days; two were admitted to the intensive care unit. Fourteen patients (82%) had a history of recent travel. Twelve were VFR travelers in Bangladesh and Pakistan and two had recently immigrated. In our study, typhoid patients had low eosinophil counts and elevated transaminases. Seventy-six percent (12 of 17) of all isolates were resistant to nalidixic acid, 23.5% (4 of 17) were resistant to ampicillin and co-trimoxazole, and one was resistant to ciprofloxacin. All isolates were susceptible to third-generation cephalosporins.

Conclusions

Younger VFR travelers appear to be at greater risk of acquiring infection and developing complications. Absolute eosinopenia and increased liver function test values could be useful early diagnostic clues in a returning traveler with fever, once malaria has been excluded. There was a high rate of decreased susceptibility to fluoroquinolones, confirming that the use of third-generation cephalosporins or macrolides in patients from the Indian subcontinent is most appropriate. Prevention in VFR travelers to South Asia is critical and efforts should be targeted at better education and pre-travel immunization.

Typhoid fever is endemic in many areas of the world and also a leading cause of fever in the returning traveler.[1] Approximately 21 million people are affected with typhoid each year, which results in 200,000 to 600,000 deaths annually.[2-4] The highest prevalence is among infants, children, and adolescents in South Asia, where poor sanitation and food handling practices continue to make typhoid a persistent public health issue.[1-6]

There is growing concern over the emergence of multidrug-resistant strains of Salmonella Typhi in many parts of Asia and Africa.[7] Since the rapid spread of multidrug-resistant (as defined by resistance to chloramphenicol, amoxicillin, and co-trimoxazole) S Typhi in the 1990s[7] and early 2000s,[8] quinolones have been the mainstay of treatment in adults.[9-11] However, during the last 2 decades, nalidixic acid-resistant strains (NARST) are being isolated with increasing frequency. Despite in vitro sensitivity to ciprofloxacin [minimum inhibitory concentration (MIC) < 1, though usually >0.1], the disease caused by these strains can have a prolonged and sometimes unfavorable course when treated with quinolones.[12]

In the United States, approximately 400 cases of typhoid are reported each year, 70% to 90% of which are associated with recent travel.[7, 13-15] Immigrants and travelers visiting friends and relatives (VFR travelers) are at a higher risk of acquiring typhoid.[8, 9, 16-19] Another 10% to 30% are domestic cases.[14] The vast majority of imported cases come from seven countries: India, Bangladesh, Pakistan, Mexico, the Philippines, El Salvador, and Haiti.[3, 8, 9] The overall risk of acquiring typhoid from travel to the Indian subcontinent is at least 10 to 20[20] and up to 100 times[21] higher than from other geographic areas.

History of travel to the above regions,[15-17] in conjunction with clinical and laboratory features unique to typhoid, may be helpful in the initial diagnosis, prior to blood culture results being available.[22] Clinically, typhoid is typically characterized by a syndrome of prolonged high fever, relative bradycardia, splenomegaly, and abdominal symptoms.[1-3] Laboratory abnormalities often consist of pancytopenia with zero or near-zero eosinophils[10, 23-25] and mild transaminitis.[1-3, 9, 10, 15]

This study is a retrospective analysis of the epidemiologic, clinical, and basic hematologic features of patients diagnosed with typhoid, as well as an analysis of the sensitivity profiles of S Typhi isolates collected over a 5-year period at Jacobi Medical Center, a municipal tertiary center that serves a large immigrant population.

Methods

We queried all positive S Typhi isolates over a 5-year period, from January 2006 to December 2010. Charts were reviewed for demographic data, travel history, symptoms and signs, basic laboratory results [complete blood count (CBC) with automated differential, basic metabolic panel, and liver function tests (LFTs)], treatment, and clinical course. Susceptibility profiles of all isolates were reviewed, and resistance to nalidixic acid was used as a marker of decreased susceptibility to quinolones.

Results

During the study period, 17 individuals were identified with S Typhi. Fourteen patients (82%) had a history of recent travel and 11 were children and adolescents <18 years. Twelve patients (nine < 18 y) were VFR travelers in Bangladesh and Pakistan and two children had recently immigrated. All 11 children were traveling with adult family members, none of whom developed typhoid fever. Two adolescents were family members of imported cases (one from Bangladesh and one from Pakistan) but had no travel history themselves. For one patient, the mode of transmission remained unknown.

None of the travelers had been vaccinated or formally educated about preventive measures regarding safe food and water, prior to their trip. Salmonella Typhi was thought to have been domestically acquired in one patient with typhoid fever and no history of recent travel, through contact with her grandmother, who had recently visited from Bangladesh. That patient reported vaccination more than 1 year ago, prior to a trip to Bangladesh.

The median age of our patients was 12 years (range: 2–47 y). Ninety-four percent of positive typhoid cases (16 of 17) were hospitalized (median stay of 7 d), and two children were admitted to the intensive care unit (both of them with hypotension and respiratory distress, one with a pleural effusion). Eighty-eight percent (15 of 17) of patients had been previously evaluated and discharged, either from the emergency department or by their primary care physician. One 7-year-old patient developed osteomyelitis, despite 8 days of appropriate intravenous antibiotics (ceftriaxone). Patients with typhoid had a history of prolonged and high fevers, elevated LFT values, and low eosinophil counts (Tables 1-3). In specific, 58.8% (10 of 17) of our patients with typhoid had an absolute eosinophil count of 0 (range: 0–50,000/mcL) by automated differential (Table 2).

Table 1. Demographic and clinical parameters of patients with typhoid fever (median, 25th–75th percentiles)
Age (y)12 (4–19.5)
Gender (M/F)10/7
Duration of fever (d)8 (5–12)
T max (°F)103.4 (103–104)
Systolic blood pressure (mm Hg)116 (101–123)
Diastolic blood pressure (mm Hg)68 (64–80.5)
Heart rate124 (105.5–138.5)
Abdominal pain (%)7/17 (41.2)
Diarrhea (%)10/17 (90.9)
Hepatosplenomegaly (%)2/17 (11.8)
Rose spots (%)1/17 (5.9)
Admitted to the hospital (%)16/17 (94)
Length of stay (d)7 (6–8)
Complications (%)3/17 (17.6)
Table 2. Hematologic parameters of patients with typhoid fever (median, 25th–75th percentiles)
White blood cell count (103/μL)7.6 (4.9–9.05)
Hematocrit (%)38.5 (34.15–40.15)
Hb (g/dL)12.9 (11.43–13.65)
Neutrophils (%)68.8 (60.05–76.7)
Neutrophils (103/μL)5.56 (3.25–7.7)
Lymphocytes (%)21.1 (14.5–26)
Lymphocytes (103/μL)1.48 (1.24–2.3)
Monocytes (%)6.6 (3.1–9.5)
Monocytes (103/μL)0.31 (0.14–0.662)
Eosinophils (%)0 (0–0.15)
Eosinophils (103/μL)0 (0–0.015)
Basophils (%)0.1 (0–0.4)
Basophils (103/μL)0.01 (0–0.025)
Platelet count (103/μL)203 (150–309)
Table 3. Biochemical parameters of patients with typhoid fever (median, 25th–75th percentiles)
  1. BUN, blood urea nitrogen; AST, aspartate aminotransferase; ALT, alanine aminotransferase.

Na+ (mEq/L)134 (133–138)
K+ (mEq/L)4 (3.6–4.5)
Cl (mEq/L)102 (96.5–103.5)
tCO2 (mEq/L)23 (19.5–25)
BUN (mg/dL)11 (9.5–13)
Cr (mg/dL)0.6 (0.35–0.85)
Glc (mg/dL)95 (76.5–114)
Ca (mg/dL)8.8 (8.2–9.4)
AST (IU/L)57 (47–84.5)
ALT (IU/L)53 (32.5–93)

With respect to S Typhi cases, 76% (12 of 17) of all isolates were resistant to nalidixic acid, 23.5% (4 of 17) were resistant to ampicillin and co-trimoxazole, and one strain was resistant to ciprofloxacin. All isolates were susceptible to third-generation cephalosporins. The isolates were not tested for susceptibility to the newer macrolides.

Discussion

New York City residents, representing 3% of the US population, account for 12% of US overseas travelers. Moreover, the immigrant population of New York City is approximately 3.5 times that of the national average.[13] Therefore, it is not surprising that, from 1999 to 2006, New York State had the second highest incidence of imported typhoid (15% of total cases reported to the Centers for Disease Control and Prevention), after California.[14] VFR travelers returning to the United States,[14, 20] as well as Europe[26] and Canada,[27] seem to be at high risk of contracting typhoid, compared to those visiting typhoid-endemic areas for business or tourism. In addition, travel to the Indian subcontinent is associated with a 10 to 100 times greater risk of infection than travel to other geographic areas.[20, 21, 27]

In agreement with the above, 12 of 17 (70%) patients diagnosed with typhoid at our institution from 2006 to 2010 were VFR travelers in the Indian subcontinent. Most of them were children and young adolescents, whose adult companions did not develop the disease. This could be due to immunity acquired earlier in life or better adherence to safe food and water precautions.[28] Younger VFR travelers seem to be at greater risk of acquiring infection and developing complications and are, therefore, most likely to benefit from travel consultation and vaccination.[5, 6]

High fever in VFR travelers returning from the Indian subcontinent should prompt a strong clinical suspicion for typhoid. However, the majority (88%) of our patients had had previous health care visits and were discharged with the diagnosis of a viral infection. Three of them had a complicated course, leading to prolonged hospitalization. Therefore, given the mostly nonspecific symptoms and signs of typhoid, it would be useful to identify features from the clinical presentation and initial laboratory results (CBC and metabolic profile) that could help differentiate typhoid from other causes of fever in returning travelers, early in the course of the disease.

In a prospective surveillance study of 82 cases in an endemic area,[22] duration of fever >7 days, chills, and absence of cough were found to be of diagnostic value. However, the authors could not formulate a specific prediction rule that could be reproducible in clinical decision making. In our case series of returning travelers, we confirmed that the magnitude and duration of measured or reported fever could be useful diagnostic clues (Table 1). Two of the classic features of typhoid in the literature, constipation and bradycardia, were not observed frequently in our group of patients with S Typhi. On the contrary, our patients with typhoid reported more frequently loose bowel movements, possibly because most of them were diagnosed later in the course of the disease (Table 1).

We decided to further explore the potential diagnostic utility of a CBC and comprehensive metabolic panel, which are part of the routine work-up for the returning travelers with fever at most Western institutions. The most striking feature of the hematologic profile seems to be the well-described feature from decades ago: “aneosinophilia.”[23, 24] Specifically, more than half (10 of 17;58.8%) of our patients with typhoid had an absolute eosinophil count of 0 by automated differential. According to a recent task force on the diagnosis of enteric fever in children,[25] eosinopenia does not necessarily differentiate typhoid from other acute bacterial or viral infections, but a normal eosinophil count makes enteric fever less likely. The exact mechanism by which eosinopenia develops is unclear, but our findings suggest that it can be a useful diagnostic clue.[25] LFT values were significantly increased in patients with S Typhi, although not high enough to qualify as “typhoid hepatitis,” which has been previously described.[29, 30] It should be noted, though, that in cases of markedly elevated LFT values, the clinician should also look for water-borne co-transmission of hepatitis viruses, namely hepatitis A and E.[30]

In the present case series, we report a high rate of nalidixic acid resistance (76%). In 2006, the overall rate of NARST was 54% and it was 65% for India for the period 1999 to 2006.[14] On the basis of these results, third-generation cephalosporins should now be considered the antibiotics of choice for the initial empiric treatment of typhoid that requires parenteral therapy, especially when there is a history of travel to India, Pakistan, or Bangladesh.[7-10] The recommended duration of treatment is 10 to 14 days,[1, 7] and one of our patients who had been treated with ceftriaxone for 8 days developed Salmonella osteomyelitis.

In our study, S Typhi isolates were not tested for susceptibility to the newer macrolides. The use of macrolides in endemic areas is limited, because of their high cost and low availability. It should be noted, though, that azithromycin is a promising option for oral treatment of typhoid in returning travelers, as no resistance has been reported yet and the cure rate is >90%.[9, 15, 31] A very recent randomized study showed that combination therapy of ceftriaxone with azithromycin, compared to ceftriaxone alone, significantly decreased the time to defervescence and the length of hospital stay, in a group of Israeli travelers to Nepal who had acquired Salmonella Paratyphi.[32]

None of our VFR travelers had been vaccinated or formally educated about preventive measures prior to travel. Safe food and water practices are of utmost importance; however, the evidence on pre-travel vaccination is quite controversial.[33-35] In the study by Lynch and colleagues,[14] only 5% of all US travelers found to have typhoid fever, over a 10-year period, had actually received the vaccine. On the contrary, in a large nation-wide study, 62% of the Israeli travelers who acquired typhoid fever had received vaccination within 3 years.[21] However, the same study[21] showed that the incidence of enteric fever in Israeli travelers to Nepal declined, compared to the prevaccination era. A single case-control study of travelers to India estimated the efficacy of the Ty21a vaccine to be only 23%.[34] Nevertheless, in that study, only three doses of the oral vaccine were used, which may, in part, explain its low efficacy. It is interesting that our only patient who had received the vaccine developed typhoid after 1 year, when visited by her grandmother, most likely a chronic carrier. All travelers should be included in a discussion about vaccine indications and participate in the decision process,[35] but we believe that VFR travelers to South Asia could benefit from vaccination, despite its limited efficacy in previous reports.

In conclusion, younger VFR travelers in areas that are endemic for typhoid fever seem to be at greater risk of acquiring infection and developing complications. Absolute eosinopenia and increased LFT values could be useful early diagnostic clues in a returning traveler with fever. In our study, there was a high rate of decreased susceptibility to fluoroquinolones, confirming that the use of third-generation cephalosporins and macrolides in patients from the Indian subcontinent is most appropriate for the treatment of typhoid. Prevention in this group of travelers is critical and efforts should be targeted at better education and pre-travel immunization.

Declaration of Interests

The authors state that they have no conflicts of interest.

Box.

image

This is a restaurant entrance in Senegal. The name of the restaurant is “birds' net”. There is a warning sign in the background that indicates “Kitchen Hygiene = health”, which leads customers to presume that standard hygiene precautions are followed. Food hygiene is an important preventive measure to avoid typhoid fever in travelers (see the article by Farmakiotis et al, pages 17–21) Photo Credit: Eric Caumes (Setting: Senegal)

Ancillary