Invasive salmonellosis among children admitted to a Rural Tanzanian Hospital and a comparison with previous studies

Background: The importance of invasive salmonellosis in African children is well recognized but there is inadequate information on these infections. We conducted a fever surveillance study in a Tanzanian rural hospital to estimate the case fraction of invasive salmonellosis among pediatric admissions, examine associations with common co-morbidities and describe its clinical features. We compared our main findings with those from previous studies among children in sub-Saharan Africa. Methodology/Principal Findings: From 1 March 2008 to 28 Feb 2009, 1,502 children were enrolled into the study. We collected clinical information and blood for point of care tests, culture, and diagnosis of malaria and HIV. We analyzed the clinical features on admission and outcome by laboratory-confirmed diagnosis. Pathogenic bacteria were isolated from the blood of 156 (10%) children, of which 14 (9%) were S. typhi, 45 (29%) were NTS and 97 (62%) were other pathogenic bacteria. Invasive salmonellosis accounted for 59/156 (38%) bacteremic children. Children with typhoid fever were significantly older and presented with a longer duration of fever. NTS infections were significantly associated with prior antimalarial treatment, malarial complications and with a high risk for death. Conclusions/Significance: Invasive salmonellosis, particularly NTS infection, is an important cause of febrile disease among hospitalized children in our rural Tanzanian setting. Previous studies showed considerable variation in the case fraction of S. typhi and NTS infections. Certain suggestive clinical features (such as older age and long duration of fever for typhoid whereas concomitant malaria, anemia, jaundice and hypoglycemia for NTS infection) may be used to distinguish invasive salmonellosis from other severe febrile illness. © 2010 Mtove et al. The study was conducted at Teule Hospital, which is the designated district hospital of Muheza in north-eastern Tanzania. The hospital serves a catchment population of about 277,000 of whom 17% are aged less than five years. Child mortality in the area is 165/1000 [5]. The majority of inhabitants reside in rural settings, mainly practicing subsistence farming and informal trade. The area is highly endemic for Plasmodium falciparum malaria with perennial transmission and two seasonal peaks coinciding with the short and long rains [6]. HIV sero-prevalence among antenatal clinic attendees was about 7% in 2007 [7]. The Tanzanian Expanded Programme of Immunization includes the following: Bacille Calmette-Guérin, live oral polio, diphtheria-whole cell pertussis-tetanus-hepatitis B and monovalent measles vaccines for children, as well as supplemental tetanus toxoid vaccine for women of child-bearing age. Tanzania had just started immunization against Haemophilus influenzae type b in March 2009 and hopes to introduce pneumococcal vaccine in 2010 [8]. Typhoid vaccine is not routinely administered in the country. Prior to the start and during the course of the study, emergency triage and hospital care guidelines were implemented in the ward [9]. On admission, children aged 2 months to 14 years were screened for eligibility during the study hours from 7am to 7pm, Monday-Sunday. Children with fever of 3 or more days prior to admission or fever of less than 3 days but with at least one severity criteria (respiratory distress, deep breathing, severe pallor with respiratory distress, prostration, capillary refill ≥3 seconds, temperature gradient, systolic blood pressure <70 mm Hg, coma, severe jaundice, history of 2 or more convulsions in last 24 hours, hypoglycemia, neck stiffness, bulging fontanel or desaturation) were recruited into the study. All clinical information was recorded in a standard case record form. Treatment was provided as per national guidelines. Outcome was recorded at discharge or death in a discharge form. Surveillance procedures were supervised by experienced study physicians (GM, IH, JD). We collected 1 to 10 millilitres of blood (depending on body weight) from each eligible child. Immediate bedside testing included those for hemoglobin concentration (Hemocue™, Anglholm, Sweden) and blood glucose level (Accu-check™, Roche Diagnostics GmbH, Germany). We performed two types of rapid diagnostic test (RDT) for P. falciparum malaria: HRP-2 based (Paracheck™, Orchid Biomedical, Mumbai, India or Parahit™, Span Diagnostics, Surat, India) and LDH based (OptiMAL-IT, DiaMed AG, Switzerland). From each child, thin and thick blood films were prepared, Giemsa-stained and read by experienced laboratory technicians. At least 100 high power microscopic fields of the thin film were examined to exclude the diagnosis of malaria. Blood for culture was inoculated into a BactALERT™ Pediatric-fan bottle (bioMérieux, Marcy l'Etoile, France) and incubated in the BacT/ALERT 3D automated microbial detection system. Blood cultures were processed according to standard methods. Colonies with biochemical reactions on API20E suggestive of Salmonellae were confirmed serologically by slide and tube agglutination testing using specific O and H antisera (Becton Dickinson, NJ, USA). Sera were tested for the presence of HIV-1 and HIV-2 antibodies according to the National HIV rapid testing algorithm [10] using Capillus HIV-1, HIV-2 test (Trinity Biotech, Bray, Ireland) or SD Bioline (Standard Diagnostics, Kyonggi-do, Korea) followed by Determine HIV-1/2 test (Abbott Laboratories, IL, USA) if the first test was positive. Discordant results were resolved by a third antibody test, Unigold (Trinity Biotech, Bray, Ireland), which if positive rendered the sample as positive and if negative, the sample was considered as negative. Children aged less than 18 months with positive results were not tested by polymerase chain reaction for viral antigen and for that reason were excluded in the final analysis. Data were double-entered into custom-made data entry programs using MS-Access (Microsoft Corp, VA, USA). Data management programs included error, range and consistency check programs. Fever was defined as history of a rise in body temperature as recalled by a care-giver or presence of axillary temperature ≥37.5°C on presentation. Bacteremia was defined as fever with isolation of pathogenic bacteria from blood culture, further classified as those caused by S. typhi (typhoid fever), NTS, and other (non-Salmonellae) pathogenic bacteria. Malaria was defined as fever with a positive RDT or blood film. HIV infection was defined as a positive Capillus test or SD bioline, confirmed either by a positive Determine HIV-1/2 or a positive Unigold test. Low maternal education was considered as schooling to less than the National Curriculum standard 7. Diarrhea was defined as loose or watery stools ≥3 times per day. A seizure was regarded as abnormal movements with altered consciousness. Desaturation was defined as oxygen saturation less than 90% measured by pulse oximetry. Acute severe malnutrition was defined as the presence of bilateral pedal edema or severe wasting. We also assessed the mid-upper arm circumference (MUAC) of children between 12 to 59 months of age. Signs of shock were temperature gradient in the lower extremities and delayed capillary refill of ≥3 seconds or systolic blood pressure <70 mm Hg. Signs of dehydration were delayed skin pinch >2 seconds or sunken eyes. Prostration was defined as inability to sit unsupported (for children over 9 months of age) or to drink/breastfeed. Coma was defined as Blantyre coma score ≤2 for children less than 2 years of age or a Glasgow coma score ≤10 for older children. Hypoglycemia was defined as blood glucose level of <2.5 mmol/litre. Anemia was defined as hemoglobin of <8 g/dl and severe anemia <5 g/dl. To assess potentially important distinguishing factors, we classified the cases into 5 non-mutually-exclusive groups: typhoid fever, invasive NTS infection, other pathogenic bacteremia, malaria, and those without bacteremia and malaria. Comparisons of categorical data were made using the Chi square or Fishers' Exact test, as appropriate. Comparisons of continuous data were made using student's t-test for data with equal variance or Welch's t-test for those with unequal variance. All analyses were performed using Stata™ v 10.0 (Stata Corp., Tx, USA). We conducted a literature review to compare our main findings with those from previous studies. Potential articles for inclusion were identified by direct searches of the MEDLINE database through PubMed. We included facility-based studies of children ≤16 years in sub-Saharan Africa that reported case fractions of S. typhi and NTS infection from sterile-site specimen (blood, CSF, lung or joint aspirate) bacterial cultures. The searches were restricted to publications from 1987 to date. For study sites with several publications generated from the same study population, only one citation was included unless the time period varied. We also conducted supplementary searches of the references in retrieved articles. Abstracts were reviewed and if relevant, the article was included. The fever surveillance was conducted following the principles governing biomedical research involving human subjects. Prior written informed consent was obtained from the parent or guardian of each eligible child. Pre-test counseling was provided before HIV testing in accordance with local guidelines. The study was approved by the National Institute for Medical Research, Tanzania (NIMR/HQ/R.8a/Vol.IX/666) and the International Vaccine Institute - Institutional Review Boards, South Korea (IRB# 2007-017).