Antibiotic and antifungal use in paediatric departments at three academic hospitals in South Africa.

Objectives: South Africa implemented a National Strategic Framework to optimise antimicrobial stewardship in 2014; however, there is limited data on how this has affected prescribing, especially to children treated in academic centres. Methods: We conducted a point prevalence survey using the World Health Organization (WHO) methodology to evaluate antibiotic and antifungal prescribing practices in paediatric departments at three academic hospitals in South Africa. Results: We recorded 1946 antimicrobial prescriptions in 1191 children, with 55.2% and 39.2% of the antibiotics classified as WHO AWaRe Access and Watch drugs, respectively. There were significant differences in prescription of Reserve antibiotics and antifungals between institutions. Receipt of WHO Watch and Reserve antibiotics was independently associated with infancy (<12 months) and adolescents (13-17 years) (adjusted relative risk [aRR]: 2.09-9.95); prolonged hospitalisation (aRR: 3.29-30.08); rapidly or ultimately fatal illness (aRR: 1.94 to 5.52); and blood transfusion (aRR: 3.28-5.70). Antifungal prescribing was associated with treatment of hospital-associated infection (aRR: 2.90), medical prophylaxis (aRR: 3.30), and treatment in intensive care units (aRR: 2.15-2.27). Conclusions: Guidance on optimisation of infection prevention and control practice and strengthening of antimicrobial stewardship would impact positively on the care of sick children in our setting.

Healthcare-associated Infections Drive Antimicrobial Prescribing in Pediatric Departments at Three Academic Hospitals in South Africa.

BACKGROUND: The prevalence of antimicrobial prescriptions for healthcare-associated infections (HAI) in South Africa is largely unknown. This study aimed to estimate the point prevalence of pediatric antibiotic and antifungal usage in 3 South African academic hospitals. METHODS: This cross-sectional study included hospitalized neonates and children (0-15 years). We used the World Health Organization methodology for antimicrobial point prevalence studies, with weekly surveys to achieve a sample size of ~400 at each site. RESULTS: Overall, 1,946 antimicrobials were prescribed to 1,191 patients. At least 1 antimicrobial was prescribed for 22.9% [95% confidence interval (CI): 15.5-32.5%] of patients. The prevalence of antimicrobial prescribing for HAI was 45.6%. In the multivariable analysis, relative to children 6-12 years, neonates [adjusted relative risk (aRR): 1.64; 95% CI: 1.06-2.53], infants (aRR: 1.57; 95% CI: 1.12-2.21) and adolescents (aRR: 2.18; 95% CI: 1.45-3.29) had significantly increased risk of prescriptions for HAI. Being preterm (aRR: 1.33; 95% CI: 1.04-1.70) and underweight (aRR: 1.25; 95% CI: 1.01-1.54) was predictive of antimicrobial usage for HAI. Having an indwelling device, surgery since admission, blood transfusions and classification as rapidly fatal on McCabe score also increased the risk of prescriptions for HAI. CONCLUSIONS: The high prevalence of antimicrobial prescribing for HAI to treat children with recognized risk factors in academic hospitals in South Africa is concerning. Concerted efforts need to be made to strengthen hospital-level infection prevention and control measures, with a critical review of antimicrobial usage through functional antibiotic stewardship programs to preserve the available antimicrobial armamentarium at the hospital level.

Annual Tuberculosis Preventive Therapy for Persons With HIV Infection : A Randomized Trial.

BACKGROUND: Tuberculosis preventive therapy for persons with HIV infection is effective, but its durability is uncertain. OBJECTIVE: To compare treatment completion rates of weekly isoniazid-rifapentine for 3 months versus daily isoniazid for 6 months as well as the effectiveness of the 3-month rifapentine-isoniazid regimen given annually for 2 years versus once. DESIGN: Randomized trial. (ClinicalTrials.gov: NCT02980016). SETTING: South Africa, Ethiopia, and Mozambique. PARTICIPANTS: Persons with HIV infection who were receiving antiretroviral therapy, were aged 2 years or older, and did not have active tuberculosis. INTERVENTION: Participants were randomly assigned to receive weekly rifapentine-isoniazid for 3 months, given either annually for 2 years or once, or daily isoniazid for 6 months. Participants were screened for tuberculosis symptoms at months 0 to 3 and 12 of each study year and at months 12 and 24 using chest radiography and sputum culture. MEASUREMENTS: Treatment completion was assessed using pill counts. Tuberculosis incidence was measured over 24 months. RESULTS: Between November 2016 and November 2017, 4027 participants were enrolled; 4014 were included in the analyses (median age, 41 years; 69.5% women; all using antiretroviral therapy). Treatment completion in the first year for the combined rifapentine-isoniazid groups (n = 3610) was 90.4% versus 50.5% for the isoniazid group (n = 404) (risk ratio, 1.78 [95% CI, 1.61 to 1.95]). Tuberculosis incidence among participants receiving the rifapentine-isoniazid regimen twice (n = 1808) or once (n = 1802) was similar (hazard ratio, 0.96 [CI, 0.61 to 1.50]). LIMITATION: If rifapentine-isoniazid is effective in curing subclinical tuberculosis, then the intensive tuberculosis screening at month 12 may have reduced its effectiveness. CONCLUSION: Treatment completion was higher with rifapentine-isoniazid for 3 months compared with isoniazid for 6 months. In settings with high tuberculosis transmission, a second round of preventive therapy did not provide additional benefit to persons receiving antiretroviral therapy. PRIMARY FUNDING SOURCE: The U.S. Agency for International Development through the CHALLENGE TB grant to the KNCV Tuberculosis Foundation.

The Cape Town Declaration on Vaccines 2012: Unlocking the full potential of vaccines in Africa.

Delegates at the first International African Vaccinology Conference noted, with dismay, that many African children have limited access to existing and new vaccines as a consequence of weak immunisation programmes, lack of political will, and high vaccine prices. This inequality is a denial of the African child her basic right to a healthy life, and jeopardises long term economic growth on the continent. In addition, there is insufficient emphasis in Africa on adolescent and adult immunisation. The delegates documented various concerns and made various commitments; contained in this Cape Town Declaration on Vaccines, adopted on 11 November 2012. Finally, delegates confirmed their agreement with the goals and strategic objectives of the Global Vaccine Action Plan, and committed to hold African leaders accountable for its implementation during the Decade of Vaccines. The full list of registered conference delegates is provided as supplementary data to this manuscript.

Advocating for efforts to protect African children, families, and communities from the threat of infectious diseases: report of the First International African Vaccinology Conference.

One means of improving healthcare workers' knowledge of and attitudes to vaccines is through running vaccine conferences which are accessible, affordable, and relevant to their everyday work. Various vaccinology conferences are held each year worldwide. These meetings focus heavily on basic science with much discussion about new developments in vaccines, and relatively little coverage of policy, advocacy, and communication issues. A negligible proportion of delegates at these conferences come from Africa, home to almost 40% of the global burden of vaccine-preventable diseases. To the best of our knowledge, no major vaccinology conference has ever been held on the African continent apart from World Health Organization (WHO) meetings. The content of the first International African Vaccinology Conference was planned to be different; to focus on the science, with a major part of discussions being on clinical, programmatic, policy, and advocacy issues. The conference was held in Cape Town, South Africa, from 8 to 11 November 2012. The theme of the conference was "Advocating for efforts to protect African children, families, and communities from the threat of infectious diseases". There were more than 550 registered participants from 55 countries (including 37 African countries). There were nine pre-conference workshops, ten plenary sessions, and 150 oral and poster presentations. The conference discussed the challenges to universal immunisation in Africa as well as the promotion of dialogue and communication on immunisation among all stakeholders. There was general acknowledgment that giant strides have been made in Africa since the global launch of the Expanded Programme on Immunisation in 1974. For example, there has been significant progress in introducing new and under-utilised vaccines; including hepatitis B, Haemophilus influenza type b, pneumococcal conjugate, rotavirus, meningococcal A conjugate, and human papillomavirus vaccines. In May 2012, African countries endorsed the Global Vaccine Action Plan at the World Health Assembly. However, more than six million children remain incompletely vaccinated in Africa leading to more than one million vaccine-preventable deaths annually. In addition, there are persistent problems with leadership and planning, vaccine stock management, supply chain capacity and quality, provider-parent communication, and financial sustainability. The conference delegates agreed to move from talking to taking concrete actions around children's health, and to ensure that African governments commit to saving children's lives. They would advocate for lower costs of immunisation programmes in Africa, perhaps through bulk buying and improved administration of vaccine rollout through the New Partnership for Africa's Development.

Randomized trial of type 1 and type 3 oral monovalent poliovirus vaccines in newborns in Africa.

BACKGROUND: The Global Polio Eradication Initiative aims to eradicate wild poliovirus by the end of 2012. Therefore, more-immunogenic polio vaccines, including monovalent oral poliovirus vaccines (mOPVs), are needed for supplemental immunization activities. This trial assessed the immunogenicity of monovalent types 1 and 3, compared with that of trivalent oral poliovirus vaccine (tOPV), in South Africa. METHODS: We conducted a blinded, randomized, 4-arm controlled trial comparing the immunogenicity of a single dose of mOPV1 (from 2 manufacturers) and mOPV3 (from 1 manufacturer), given at birth, with the immunogenicity of tOPV. RESULTS: Eight hundred newborns were enrolled; 762 (95%) were included in the analysis. At 30 days after vaccine administration, seroconversion to poliovirus type 1 was 73.4% and 76.4% in the 2 mOPV1 arms, compared with 39.1% in the tOPV arm (P < .0000001), and seroconversion to poliovirus type 3 was 58.0% in the mOPV3 arm, compared with 21.2% in the tOPV arm (P < .0000001). The vaccines were well tolerated, and no adverse events were attributed to trial interventions. CONCLUSION: A dose of mOPV1 or mOPV3 at birth was superior to that of tOPV in inducing type-specific seroconversion in this sub-Saharan African population. Our results support continued use of mOPVs in supplemental immunization activities in countries where poliovirus types 1 or 3 circulate. Clinical Trials Registration. ISRCTN18107202.

'Communicate to vaccinate' (COMMVAC). building evidence for improving communication about childhood vaccinations in low- and middle-income countries: protocol for a programme of research.

BACKGROUND: Effective provider-parent communication can improve childhood vaccination uptake and strengthen immunisation services in low- and middle-income countries (LMICs). Building capacity to improve communication strategies has been neglected. Rigorous research exists but is not readily found or applicable to LMICs, making it difficult for policy makers to use it to inform vaccination policies and practice.The aim of this project is to build research knowledge and capacity to use evidence-based strategies for improving communication about childhood vaccinations with parents and communities in LMICs. METHODS AND DESIGN: This project is a mixed methods study with six sub-studies. In sub-study one, we will develop a systematic map of provider-parent communication interventions for childhood vaccinations by screening and extracting data from relevant literature. This map will inform sub-study two, in which we will develop a taxonomy of interventions to improve provider-parent communication around childhood vaccination. In sub-study three, the taxonomy will be populated with trial citations to create an evidence map, which will also identify how evidence is linked to communication barriers regarding vaccination. In the project's fourth sub-study, we will present the interventions map, taxonomy, and evidence map to international stakeholders to identify high-priority topics for systematic reviews of interventions to improve parent-provider communication for childhood vaccination. We will produce systematic reviews of the effects of high-priority interventions in the fifth sub-study. In the sixth and final sub-study of the project, evidence from the systematic reviews will be translated into accessible formats and messages for dissemination to LMICs. DISCUSSION: This project combines evidence mapping, conceptual and taxonomy development, priority setting, systematic reviews, and knowledge transfer. It will build and share concepts, terms, evidence, and resources to aid the development of communication strategies for effective vaccination programmes in LMICs.

Developing vaccinology expertise for Africa: six years and counting.

There exists high quality evidence showing that interactive educational meetings and workshops can improve healthcare worker performance. This evidence formed the basis for establishing the annual African Vaccinology Course in 2005 at the University of Cape Town in South Africa. The course, which is designed to develop vaccinology expertise for Africa, covers relevant basic sciences pertaining to vaccine-preventable diseases such as epidemiology, immunology and microbiology; discusses specific vaccine-preventable diseases; provides information on vaccine safety, vaccination strategies and evaluation of vaccines; discusses new vaccines in the pipeline; and promotes vaccine advocacy. We hope that course alumni would become strong advocates for childhood immunisation in their respective countries. Such dedicated advocacy should contribute to reducing the time gap between the development of new vaccines and the formulation of policies enabling their introduction in African countries, as well as contributing to more equitable increase in immunisation coverage in our continent.

Review of rotavirus studies in Africa: 1976-2006.

BACKGROUND: Rotavirus is a common cause of severe diarrheal disease in children worldwide. Ninety percent of the associated deaths occur in sub-Saharan Africa and South East Asia. Our aim was to review the prevalence of rotavirus infection in Africa over the past 30 years. METHODS: Eligible studies were identified from 3 Medline searches. Only studies including children <5 years of age that included >50 children and had an observational period >3 months were included. The data were analyzed during 3 periods (1976-1985, 1986-1995, and 1996-2006), as a summary (1976-2006), and by different study settings (ie, hospital, outpatient department, and combined). RESULTS: The initial search identified 206 studies from 27 countries during 1976-2006. The refined search yielded 101 studies, of which 58 (57%) were hospital based, 25 (25%) were outpatient studies, and 18 (18%) were combined. Rotavirus was detected in 25% (interquartile range, 16%-32%) of stool samples. Rotavirus was the most common agent identified in 73% of studies in which multiple diarrheal agents were determined. CONCLUSION: Rotavirus is an important cause of severe diarrheal disease in children <5 years of age in Africa. Clinical trials in South Africa and Malawi have shown that severe rotavirus disease is a vaccine-preventable entity in Africa.

Improving communication for immunisation in Africa: contribution of the Vaccines for Africa website.

ABOUT THE AUTHORS: C. Wiysonge is a medical epidemiologist and Vaccinology Programme Manager at the Institute of Infectious Disease and Molecular Medicine, University of Cape Town, South Africa. He previously worked for the Expanded Programme on Immunisation in Cameroon and has been a consultant on vaccines and immunisation for WHO and the GAVI Alliance. Z. Waggie is a Paediatrician and Senior Clinical Research Officer at the Institute of Infectious Disease and Molecular Medicine, University of Cape Town, South Africa. L. Rhoda is a communications specialist and Communications Manager at the South African Tuberculosis Vaccine Initiative, University of Cape Town, South Africa. G. Hussey is a Paediatric Infectious Diseases Clinical Specialist and Professor of Child and Adolescent Health, Director of the Institute of Infectious Disease and Molecular Medicine, and Director of the South African Tuberculosis Vaccine Initiative, University of Cape Town, South Africa. He has been a WHO part-time consultant on vaccines and immunisation for the past 10 years.

Relationship between fluid management, changes in serum sodium and outcome in hypernatraemia associated with gastroenteritis.

OBJECTIVE: To describe the relationship between fluid management, serum sodium and outcome in critically ill children with hypernatraemic gastroenteritis. METHODS: A retrospective study of 57 children with hypernatraemic gastroenteritis admitted to a paediatric intensive care unit in Cape Town, South Africa. Data were collected on fluid management, serum electrolytes and adverse outcome (seizures, new neurological deficit and mortality) and analysed using univariate and multivariate statistics. RESULTS: Median admission sodium was 165 mmol/L (145-199). Median volume of intravenous rehydration fluid was 6 mL/kg/h (144 mL/kg/day), with sodium concentration of 61 mmol/L (0-154 mmol/L), resulting in a median fall in sodium of 0.6 mmol/L/h (14.4 mmol/L/day). Fourteen children (25%) had seizures during rehydration, four children (7%) died and five children (9%) developed neurological deficit. Median admission sodium in children with adverse outcome was 172 mmol/L, with rate of fall of 0.63 mmol/L/h, compared with median admission sodium of 163 mmol/L and rate of fall of 0.48 mmol/L/h, in children with good outcome (P=0.068 and P=0.08, respectively). Median sodium content of intravenous solution was 61 mmol/L in both groups (P=0.68). Multivariate analysis demonstrated that neither sodium content of intravenous solution (P=0.59), nor rate of fall of sodium (P=0.31), was independently associated with adverse outcome. CONCLUSIONS: Rehydration in hypernatraemic gastroenteritis using intravenous solutions containing 61 mmol/L sodium would be expected to correct serum sodium at a rate of approximately 0.6 mmol/L/h. Neither sodium content of the intravenous solution, nor rate of correction of sodium, was independently associated with adverse outcome.

The lactate:pyruvate ratio following open cardiac surgery in children.

OBJECTIVE: To explore the relationship between lactate:pyruvate ratio, hyperlactataemia, metabolic acidosis, and morbidity. DESIGN AND SETTING: Prospective observational study in the paediatric intensive care unit (PICU) of a university hospital. PATIENTS: Ninety-seven children after open cardiac surgery. Most children (94%) fell into low-moderate operative risk categories; observed PICU mortality was 1%. INTERVENTIONS: Blood was sampled on admission for acid-base analysis, lactate, and pyruvate. Metabolic acidosis was defined as standard bicarbonate lower than 22 mmol/l, raised lactate as higher than 2 mmol/l, and raised lactate:pyruvate ratio as higher than 20. MEASUREMENTS AND RESULTS: Median cardiopulmonary bypass and aortic cross-clamp times were 80 and 46 min. Metabolic acidosis occurred in 74%, hyperlactataemia in 42%, and raised lactate:pyruvate ratio in 45% of children. In multivariate analysis lactate:pyruvate ratio increased by 6.4 in children receiving epinephrine infusion and by 0.4 per 10 min of aortic cross-clamp. Duration of inotropic support increased by 0.29 days, ventilatory support by 0.27 days, and PICU stay by 0.42 days, for each 1 mmol/l increase in lactate. Neither standard bicarbonate nor lactate:pyruvate ratio were independently associated with prolongation of PICU support. CONCLUSIONS: Elevated lactate:pyruvate ratio was common in children with mild metabolic acidosis and low PICU mortality. Hyperlactataemia, but not elevated lactate:pyruvate ratio or metabolic acidosis, was associated with prolongation of PICU support. Routine measurement of lactate:pyruvate ratio is not warranted for children in low-moderate operative risk categories.

Transport of critically ill children in a resource-limited setting.

OBJECTIVE: To audit paediatric intensive care unit (PICU) transfer activity and transfer-related adverse events in a resource-limited setting. DESIGN AND SETTING: Twenty-two bed regional PICU of a university children's hospital in Cape Town, South Africa. Prospective 1-year audit of all children transferred directly from other hospitals. Data were collected for patient demographics, diagnostic category, referring hospital, transferring personnel, mode of transport, and technical, clinical, and critical adverse events. Data are median (interquartile range) or percentages. The transfers of 202 children, median age 2.8 months (1.1-14), median weight 3.5 kg (2.5-8.1) were analysed. RESULTS: Most transfers were performed by paramedic personnel (82%) and via road ambulance (76%). One or more technical adverse events occurred in 36%, clinical adverse events in 27%, and critical adverse events in 9% of children. Retrievals by intensive care staff (10%), all from rural hospitals, had a lower incidence of technical adverse events (0%). Children transferred from non-academic hospitals within the metropolitan area had the highest incidence of technical (44%), clinical (39%), and critical (17%) adverse events. Crude mortality was 17% ( n=34). Technical adverse events were not associated with mortality. Non-survivors were more likely to develop shock (32%) or hypoxia (26%) during transfer than survivors (10% and 11%, respectively). CONCLUSIONS: There is a high incidence of transfer-related adverse events, most commonly in transfers from non-academic metropolitan hospitals. Further studies are needed to assess the impact of regional paediatric life support training or a specialised retrieval team on clinical adverse events and mortality.

Mortality and the nature of metabolic acidosis in children with shock.

HYPOTHESIS: Mortality in children with shock is more closely related to the nature, rather than the magnitude (base deficit/excess), of a metabolic acidosis. OBJECTIVE: To examine the relationship between base excess (BE), hyperlactataemia, hyperchloraemia, 'unmeasured' strong anions, and mortality. DESIGN: Prospective observational study set in a multi-disciplinary Paediatric Intensive Care Unit (PICU). PATIENTS: Forty-six children, median age 6 months (1.5-14.4), median weight 5 kg (3.2-8.8), admitted to PICU with shock. INTERVENTIONS: Predicted mortality was calculated from the paediatric index of mortality (PIM) score. The pH, base excess, serum lactate, corrected chloride, and 'unmeasured' strong anions (Strong Ion Gap) were measured or calculated at admission and 24 h. MEASUREMENTS AND RESULTS: Observed mortality ( n=16) was 35%, with a standardised mortality ratio (SMR) of 1.03 (95% CI 0.71-1.35). There was no significant difference in admission pH or BE between survivors and nonsurvivors. There was no association between elevation of 'unmeasured' anions and mortality, although there was a trend towards hyperchloraemia in survivors ( P=0.08). Admission lactate was higher in nonsurvivors (median 11.6 vs 3.3 mmol/l; P=0.0003). Area under the mortality receiver operating characteristic curve for lactate was 0.83 (955 CI 0.70-0.95), compared to 0.71 (95% CI 0.53-0.88) for the PIM score. Admission lactate level >5 mmol/l had maximum diagnostic efficiency for mortality, with a likelihood ratio of 2.0. CONCLUSION: There is no association between the magnitude of metabolic acidosis, quantified by the base excess, and mortality in children with shock. Hyperlactataemia, but not elevation of 'unmeasured' anions, is predictive of a poor outcome.

Retropharyngeal abscess complicated by carotid artery rupture.

OBJECTIVE: To report the case of a child with a retropharyngeal abscess complicated by carotid artery rupture. DESIGN: Descriptive case report. SETTING: Pediatric intensive care unit of a children's hospital. Patient: A 2-month old girl. INTERVENTION: Surgical ligation of the right common carotid artery was performed. MEASUREMENTS AND MAIN RESULTS: Carotid artery involvement is a rare but life-threatening complication of a retropharyngeal abscess. The child developed carotid arteritis, with pseudoaneurysm formation and acute rupture of the right carotid artery. Surgical ligation was performed, with no accompanying neurologic sequelae. A high index of suspicion is needed for the possibility of carotid space involvement if there is a delay in the resolution of symptoms, despite adequate treatment. CONCLUSION: If a retropharyngeal abscess extends into the carotid space or there is a slow resolution of symptoms, carotid artery involvement must be excluded.