Predicting mortality for paediatric inpatients where malaria is uncommon

Predicting Mortality for Adolescent and Adult Patients with Fever in Resource-Limited Settings

Febrile illnesses are a major cause of mortality in sub-Saharan Africa. Early identification of patients at increased risk of death may avert adverse outcomes. We aimed to independently evaluate the performance of the Modified Early Warning Score, quick Sequential Organ Failure Assessment (qSOFA) score, and Integrated Management of Adolescent and Adult Illness (IMAI) emergency signs and severity criteria to predict in-hospital mortality among a prospective cohort of febrile patients in Tanzania. We evaluated 419 patients aged ≥ 10 years in the period 2007-2008. Of the 44 patients who died, 31 (70.5%) were human immunodeficiency virus (HIV) infected. On univariate analysis, in-hospital mortality was associated with HIV infection, oxygen saturation < 90%, respiratory distress, Glasgow Coma Scale < 15, neck stiffness, unconsciousness, convulsions, hemoglobin < 9 g/dL, absence of a systemic syndrome, and neurologic syndrome. A qSOFA score ≥ 2, the presence of at least one, two, or three IMAI emergency signs, and IMAI severe respiratory distress syndrome without shock were significantly associated with in-hospital mortality. The criterion "presence of at least one IMAI emergency sign" showed a good diagnostic accuracy, as highlighted by the high sensitivity, low negative likelihood ratio, and wide area under the receiver operating characteristics curve. The remaining scores showed a poor performance in predicting fatal outcomes in our study population. Further studies are needed to validate our findings and to derive early warning scores that have good clinical performance in settings throughout sub-Saharan Africa.

Variation in and risk factors for paediatric inpatient all-cause mortality in a low income setting: data from an emerging clinical information network

BACKGROUND

Hospital mortality data can inform planning for health interventions and may help optimize resource allocation if they are reliable and appropriately interpreted. However such data are often not available in low income countries including Kenya.

METHODS

Data from the Clinical Information Network covering 12 county hospitals' paediatric admissions aged 2-59 months for the periods September 2013 to March 2015 were used to describe mortality across differing contexts and to explore whether simple clinical characteristics used to classify severity of illness in common treatment guidelines are consistently associated with inpatient mortality. Regression models accounting for hospital identity and malaria prevalence (low or high) were used. Multiple imputation for missing data was based on a missing at random assumption with sensitivity analyses based on pattern mixture missing not at random assumptions.

RESULTS

The overall cluster adjusted crude mortality rate across hospitals was 6 · 2% with an almost 5 fold variation across sites (95% CI 4 · 9 to 7 · 8; range 2 · 1% - 11 · 0%). Hospital identity was significantly associated with mortality. Clinical features included in guidelines for common diseases to assess severity of illness were consistently associated with mortality in multivariable analyses (AROC =0 · 86).

CONCLUSION

All-cause mortality is highly variable across hospitals and associated with clinical risk factors identified in disease specific guidelines. A panel of these clinical features may provide a basic common data framework as part of improved health information systems to support evaluations of quality and outcomes of care at scale and inform health system strengthening efforts.

Teaching Pediatric Life Support in Limited-Resource Settings: Contextualized Management Guidelines

Of the estimated 6.3 million global annual deaths in children younger than the age of 5 years, nearly all (99%) occur in low- to middle-income countries (LMIC). Existing management guidelines for children with emergency conditions as taught in a variety of current pediatric life support courses are mostly applicable to high-income countries with a different disease range and full resources compared with LMIC. A revised curriculum with evidence-based application to limited-resource settings would expand their potential for reducing pediatric mortality worldwide. This review provides a supplemental curriculum of standards for selected pediatric emergency conditions with attention to the context of disease range and level-specific resources in LMIC. During training sessions, contextualized management guidelines create the framework for realistic and fruitful case simulations.

Application of Sepsis Definitions to Pediatric Patients Admitted With Suspected Infections in Uganda

OBJECTIVES

Acute infectious diseases are the most common cause of under-5 mortality. However, the hospital burden of nonneonatal pediatric sepsis has not previously been described in the resource poor setting. The objective of this study was to determine the prevalence of sepsis among children 6 months to 5 years old admitted with proven or suspected infection and to evaluate the presence of sepsis as a predictive tool for mortality during admission.

DESIGN

In this prospective cohort study, we used the pediatric International Consensus Conference definition of sepsis to determine the prevalence of sepsis among children admitted to the pediatric ward with a proven or suspected infection. The diagnosis of sepsis, as well as each individual component of the sepsis definition, was evaluated for capturing in-hospital mortality.

SETTING

The pediatric ward of two hospitals in Mbarara, Uganda.

PATIENTS

Admitted children between 6 months and 5 years with a confirmed or suspected infection.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

One thousand three hundred seven (1,307) subjects with a confirmed or suspected infection were enrolled, and 65 children died (5.0%) during their admission. One thousand one hundred twenty-one (85.9%) met the systemic inflammatory response syndrome criteria, and therefore, they were defined as having sepsis. The sepsis criteria captured 61 deaths, demonstrating a sensitivity and a specificity of 95% (95% CI, 90-100%) and 15% (95% CI, 13-17%), respectively. The most discriminatory individual component of the systemic inflammatory response syndrome criteria was the leukocyte count, which alone had a sensitivity of 72% and a specificity of 56% for the identification of mortality in hospital.

CONCLUSIONS

This study is among the first to quantify the burden of nonneonatal pediatric sepsis in children with suspected infection, using the international consensus sepsis definition, in a typical resource-constrained setting in Africa. This definition was found to be highly sensitive in identifying those who died but had very low specificity as most children who were admitted with infections had sepsis. The systemic inflammatory response syndrome-based sepsis definition offers little value in identification of children at high risk of in-hospital mortality in this setting.

Pediatric in-Hospital Death from Infectious Disease in Uganda: Derivation of Clinical Prediction Models

Background

Pediatric hospital mortality from infectious diseases in resource constrained countries remains unacceptably high. Improved methods of risk-stratification can assist in referral decision making and resource allocation. The purpose of this study was to create prediction models for in-hospital mortality among children admitted with suspected infectious diseases.

Methods

This two-site prospective observational study enrolled children between 6 months and 5 years admitted with a proven or suspected infection. Baseline clinical and laboratory variables were collected on enrolled children. The primary outcome was death during admission. Stepwise logistic regression minimizing Akaike’s information criterion was used to identify the most promising multivariate models. The final model was chosen based on parsimony.

Results

1307 children were enrolled consecutively, and 65 (5%) of whom died during their admission. Malaria, pneumonia and gastroenteritis were diagnosed in 50%, 31% and 8% of children, respectively. The primary model included an abnormal Blantyre coma scale, HIV and weight-for-age z-score. This model had an area under the curve (AUC) of 0.85 (95% CI, 0.80–0.89) with a sensitivity and specificity of 83% and 76%, respectively. The positive and negative predictive values were 15% and 99%, respectively. Two alternate models with similar performance characteristics were developed withholding HIV and weight-for-age z-score, for use when these variables are not available.

Conclusions

Risk stratification of children admitted with infectious diseases can be calculated based on several easily measured variables. Risk stratification at admission can be used for allocation of scarce human and physical resources and to guide referral among children admitted to lower level health facilities.

Etiologies of illness among patients meeting integrated management of adolescent and adult illness district clinician manual criteria for severe infections in northern Tanzania: implications for empiric antimicrobial therapy

We describe the laboratory-confirmed etiologies of illness among participants in a hospital-based febrile illness cohort study in northern Tanzania who retrospectively met Integrated Management of Adolescent and Adult Illness District Clinician Manual (IMAI) criteria for septic shock, severe respiratory distress without shock, and severe pneumonia, and compare these etiologies against commonly used antimicrobials, including IMAI recommendations for emergency antibacterials (ceftriaxone or ampicillin plus gentamicin) and IMAI first-line recommendations for severe pneumonia (ceftriaxone and a macrolide). Among 423 participants hospitalized with febrile illness, there were 25 septic shock, 37 severe respiratory distress without shock, and 109 severe pneumonia cases. Ceftriaxone had the highest potential utility of all antimicrobials assessed, with responsive etiologies in 12 (48%) septic shock, 5 (14%) severe respiratory distress without shock, and 19 (17%) severe pneumonia illnesses. For each syndrome 17-27% of participants had etiologic diagnoses that would be non-responsive to ceftriaxone, but responsive to other available antimicrobial regimens including amphotericin for cryptococcosis and histoplasmosis; anti-tuberculosis therapy for bacteremic disseminated tuberculosis; or tetracycline therapy for rickettsioses and Q fever. We conclude that although empiric ceftriaxone is appropriate in our setting, etiologies not explicitly addressed in IMAI guidance for these syndromes, such as cryptococcosis, histoplasmosis, and tetracycline-responsive bacterial infections, were common.

Invasive Salmonella infections in areas of high and low malaria transmission intensity in Tanzania

BACKGROUND

 The epidemiology of Salmonella Typhi and invasive nontyphoidal Salmonella (NTS) differs, and prevalence of these pathogens among children in sub-Saharan Africa may vary in relation to malaria transmission intensity.

METHODS

 We compared the prevalence of bacteremia among febrile pediatric inpatients aged 2 months to 13 years recruited at sites of high and low malaria endemicity in Tanzania. Enrollment at Teule Hospital, the high malaria transmission site, was from June 2006 through May 2007, and at Kilimanjaro Christian Medical Centre (KCMC), the low malaria transmission site, from September 2007 through August 2008. Automated blood culture, malaria microscopy with Giemsa-stained blood films, and human immunodeficiency virus testing were performed.

RESULTS

 At Teule, 3639 children were enrolled compared to 467 at KCMC. Smear-positive malaria was detected in 2195 of 3639 (60.3%) children at Teule and 11 of 460 (2.4%) at KCMC (P < .001). Bacteremia was present in 336 of 3639 (9.2%) children at Teule and 20 of 463 (4.3%) at KCMC (P < .001). NTS was isolated in 162 of 3639 (4.5%) children at Teule and 1 of 463 (0.2%) at KCMC (P < .001). Salmonella Typhi was isolated from 11 (0.3%) children at Teule and 6 (1.3%) at KCMC (P = .008). With NTS excluded, the prevalence of bacteremia at Teule was 5.0% and at KCMC 4.1% (P = .391).

CONCLUSIONS

 Where malaria transmission was intense, invasive NTS was common and Salmonella Typhi was uncommon, whereas the inverse was observed at a low malaria transmission site. The relationship between these pathogens, the environment, and the host is a compelling area for further research.

Etiology of severe non-malaria febrile illness in Northern Tanzania: a prospective cohort study

Introduction

The syndrome of fever is a commonly presenting complaint among persons seeking healthcare in low-resource areas, yet the public health community has not approached fever in a comprehensive manner. In many areas, malaria is over-diagnosed, and patients without malaria have poor outcomes.

Methods and Findings

We prospectively studied a cohort of 870 pediatric and adult febrile admissions to two hospitals in northern Tanzania over the period of one year using conventional standard diagnostic tests to establish fever etiology. Malaria was the clinical diagnosis for 528 (60.7%), but was the actual cause of fever in only 14 (1.6%). By contrast, bacterial, mycobacterial, and fungal bloodstream infections accounted for 85 (9.8%), 14 (1.6%), and 25 (2.9%) febrile admissions, respectively. Acute bacterial zoonoses were identified among 118 (26.2%) of febrile admissions; 16 (13.6%) had brucellosis, 40 (33.9%) leptospirosis, 24 (20.3%) had Q fever, 36 (30.5%) had spotted fever group rickettsioses, and 2 (1.8%) had typhus group rickettsioses. In addition, 55 (7.9%) participants had a confirmed acute arbovirus infection, all due to chikungunya. No patient had a bacterial zoonosis or an arbovirus infection included in the admission differential diagnosis.

Conclusions

Malaria was uncommon and over-diagnosed, whereas invasive infections were underappreciated. Bacterial zoonoses and arbovirus infections were highly prevalent yet overlooked. An integrated approach to the syndrome of fever in resource-limited areas is needed to improve patient outcomes and to rationally target disease control efforts.

The syndrome of fever is caused by a large number of infectious diseases. Malaria is thought to have been declining in the tropics since 2004. Increasing use of malaria diagnostic tests reveal a growing proportion of patients with fever who do not have malaria. While malaria diagnostic tests may be available, healthcare workers have few tools to diagnose causes of fever other than malaria. In order to identify major causes of fever other than malaria in northern Tanzania, we studied 870 patients with fever who were sufficiently ill to require admission to hospital. Malaria was uncommon and over-diagnosed, whereas invasive infections, including bloodstream infections, were underappreciated. Infections associated with animals such as brucellosis, leptospirosis, Q fever, and spotted fever group rickettsioses as well as viral infections transmitted by mosquitoes were common yet overlooked. We recommend that research on the syndrome of fever in resource-limited areas should focus on a wide range of potential causes. Animal-associated infections should be prioritized in patient management and disease control.