University of Groningen Early onset sepsis in Suriname Zonneveld, Rens

(1)

University of Groningen

Early onset sepsis in Suriname

Zonneveld, Rens

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2017

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Zonneveld, R. (2017). Early onset sepsis in Suriname: Epidemiology, Pathophysiology and Novel Diagnostic Concepts. Rijksuniversiteit Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

Early Onset Sepsis in Suriname

Epidemiology, Pathophysiology, and Novel Diagnostic Concepts

(3)

ISBN (printed): 978-94-034-0256-7 ISBN (digital): 978-94-034-0257-4

Cover design: Frans Mettes

Lay-out and Printing: Off Page, Amsterdam

Financial support for the research in this thesis is greatly acknowledged. The following institutes and organisations provided funding for completion and printing of this thesis:

Stichting ‘De Drie Lichten’.

Early Onset Sepsis in Suriname

Epidemiology, Pathophysiology, and Novel Diagnostic Concepts

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen

op gezag van de

rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op maandag 11 december 2017 om 16.15 uur

door

Rens Zonneveld

geboren op 8 april 1983 te Breda

(4)

Early Onset Sepsis in Suriname

Epidemiology, Pathophysiology, and Novel Diagnostic Concepts

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen

op gezag van de

rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op maandag 11 december 2017 om 16.15 uur

door

Rens Zonneveld

geboren op 8 april 1983 te Breda

(5)

Promotor Prof. dr. G. Molema Copromotores Dr. F.B. Plötz Dr. M. van Meurs Beoordelingscommissie Prof. dr. J.M. Smit Prof. dr. J.B. van Woensel Prof. dr. J.G. Zijlstra

(6)

1 A COMMON CASE OF SUSPECTED EARLY ONSET SEPSIS IN SURINAME

A day prior to giving birth the mother had taken a boat from her village downstream the Suriname River to the nearest mission post in Debike1_{. She had been pregnant for eight full moons. Her water}

had broken a few days earlier, but the baby had not arrived yet. The friendly datra2_{at the mission}

post phoned somebody in the city of Paramaribo and spoke Bakratongo3_{. People in the village had}

talked about the new at’oso4_{for babies. Many women went there to give birth and they brought}

her there too. After six hours she arrived and spent the night in a room with four other women. She felt like she had korsu5_.

Her daughter was born the next day and although she was crying loudly they still took her to the baby hospital. Doctors and nurses were standing around a glass box that held her daughter. The doctors seemed confused. One of the nurses spoke her tongo6_{and explained that her baby}

was doing fine but could have an infection. They had taken her daughter’s blood to see if it was infected. Depending on her daughter’s condition and the test results they were going to decide whether to continue the antibiotics they had started. The nurse said her daughter could suffer from sepsis, wan takru siki fu brudu7_.

In the next few days she spent many hours next to the glass box in the spacious baby room. To her, her daughter seemed healthy and the same as her four earlier children. After three days, the doctors used a nanai8_{to take her brudu}9_{for the second time. The nurse told her the results}

were fine. However, they were still going to finish her treatment with more antibiotics. Finally, after a total of seven days they started their long journey home.

In this thesis, I focus on newborns admitted to the only neonatal intensive care unit (NICU) in Suriname, which is located in Paramaribo, with a specific focus on dilemmas of Early Onset Sepsis – from epidemiology and prediction towards changes in vascular endothelial integrity, principles of leukocyte-endothelial interaction, and novel diagnostic methodologies for its timely recognition or exclusion.

Rens Zonneveld, M.D. July 2017

1_{Village located along the Suriname River in the district Sipaliwini in the interior of Suriname.} Translated from the Surinamese language (Sranan Tongo)

2_physician; 3_Dutch; 4_hospital; 5_fever; 6_language;

7_{a serious infection of the blood;} 8_needle;

(11)

1 EARLY ONSET SEPSIS

Early onset sepsis (EOS) is defined as onset of sepsis in newborns within 72 hours after birth [1]. When intra-uterine infection is present, the fetus can become infected due to increased permeability of the skin and mucosa for bacterial invasion. EOS is also caused by vertical transmission of pathogens in the vaginal canal from mother to fetus during labor.

EOS is a leading cause of morbidity and mortality amongst newborns [1-6]. In Western (i.e., North American and European) countries incidence of blood culture proven EOS ranges from 0.01 to about 1.2 per 1000 live births. Incidence rates of EOS increase with decreasing gestational age and birth weight, with the highest incidence (i.e., 26 per 1000 live births) and mortality (i.e., 50-60% of blood culture proven cases) amongst infants with a birth weight below 1000 grams (2-4). EOS is associated with colonization of the birth canal (about 30% of mothers in Western countries) with Group B Streptococcus (GBS) [1]. In Western countries over 45% of all cases of culture proven EOS GBS (45%) is the responsible pathogen, followed by Escherichia coli (E.coli) (25%) (5,6). Other bacteria that cause EOS include Listeria Monocytogenes, gram-negative enteric bacilli (i.e., Enterobacter spp., Klebsiella spp.) and Enterococcus spp. [1]. Viruses (predominantly entero and herpes simplex virus) are also identified causes of EOS [1].

After the introduction of intrapartum antibiotics as prophylaxis for GBS, incidence of EOS has decreased about 10-fold over the last 20 years in many Western countries and South Africa [7]. However, recent data indicates that, while incidence of EOS due to GBS is decreasing, incidence of EOS with E.coli increases, probably due to altered resistance patterns of E.coli strains [1,5,8]. Additionally, GBS prevention approaches may have contributed to the rise of multi resistant gram-positive strains, such as Methicillin resistant Staphylococcus aureus, as causes for EOS [1,9-11]. Maternal GBS vaccination to further reduce maternal GBS colonization and incidence of EOS, while preventing antibiotic exposure, is currently under investigation [12].

EARLY ONSET SEPSIS IN THE NON-WESTERN WORLD

Studies of EOS in low resource settings in the non-Western world are severely underrepresented in the literature [13-16]. The vast majority of data on EOS are from upper-middle to high-income countries in North America and Europe. Despite the lack of detailed data on EOS in the non-Western world, there is a strong indication that over 90% of global neonatal deaths due to EOS occurs in these low-to-middle income countries [17,18]. Large meta-analyses revealed incidence of EOS in low-income countries at least similar to Western countries [13,15,16]. However, in these analyses low-income countries represented only 5-10% of the total data leaving the true global impact of EOS underestimated. Additionally, underdiagnosing (i.e., due to lack of resources and logistic or financial constraints) and underreporting of EOS are common issues in low-resource settings further enhancing underestimation of the true global impact of EOS [19,20]. Furthermore, due to limited local availability of proper laboratory facilities, studies from these countries often lack blood culture confirmed results. As a result, the spectrum of bacterial pathogens involved in EOS in the non-Western world remains relatively unclear. More data on incidence, causative organisms, morbidity and mortality from non-Western countries remain critical before proper

(12)

1

prevention strategies and clinical management of suspected EOS can be achieved. Additionally, since there is strong indication that incidence rates of culture proven EOS are substantially higher in the non-Western world versus the Western world, studies from non-Western countries may contribute immensely to our knowledge on basic and pathophysiological principles of EOS.

EARLY ONSET SEPSIS IN SURINAME

Suriname is small developing country on the Northeastern corner of South America with a multiethnic population of about 550,000 people [21]. About half of the population of Suriname lives in its capital, the city of Paramaribo. Medical care is provided by four hospitals in Paramaribo, namely the Academic Hospital Paramaribo, ‘s Lands Hospital, Diakonessen Hospital and St. Vincentius Hospital, and the Streekziekenhuis in Nickerie. Suriname has an annual birth rate of approximately 10,000 births. Over 90% of these births take place at the maternity wards of the hospitals in Paramaribo. In rural parts of Suriname Medical Mission Posts provide primary health care to the inhabitants, including basic obstetric care.

The earliest data on neonatal mortality in Suriname dates back to the detailed documentations by Dr. Paul Christiaan Flu (1884 (Paramaribo, Suriname) - 1945 (Leiden, The Netherlands)) from the early 20th_{century. In his seminal, yet forgotten, work Flu describes the poor socio-economic}

circumstances after over three centuries of slavery and its effect on neonatal and pediatric care and mortality rates [22]. Between 1900 and 1909, 9,259 live births were recorded of whom 474 died within the first 14 days of life, making a high average death rate of 51.2 per 1000 live births for that age category. Over half (N=284) of these deaths were the result of pre- and dysmaturity, yet about one third (N=110) of these deaths were from unknown cause and potentially following neonatal infection.

Currently, neonatal death rate, defined as death within the first month of life, in Suriname has decreased, but remains high with 12.9 per 1000 live births [23]. Early neonatal death (i.e., death within the first 7 days of life) is estimated at 16 per 1000 live births [24]. Preliminary data from the Suriname Perinatal and Infant Mortality Survey estimates contribution of infection to early neonatal mortality at 24% (4 per 1000 live births) of all early neonatal deaths [23]. In contrast, in The Netherlands incidence of EOS alone was 0.19 per 1000 live births in 2014 [25].

These numbers indicate a high burden of neonatal infection in Suriname. About 40 newborns die each year of infection. Despite the overall idea of the impact of infectious disease in Surinamese newborns, detailed information regarding incidence, type of infection (i.e., EOS versus LOS), microbial causes, mortality and morbidity, antibiotic treatment (type and duration), and exact epidemiological determinants are currently unavailable. In Chapter 2 of this thesis we explore the epidemiology and outcomes of newborns admitted to Suriname’s neonatal care facility at the Academic Hospital Paramaribo. This facility was established in 2008 and renewed in 2015 with expansion of intensive care capacity, training of personnel and new equipment. For this chapter we hypothesized that tertiary function and morbidity and mortality rates of treated newborns would improve after the transition to the renewed neonatal care facility. Additionally, the impact of EOS on mortality of Surinamese newborns is explored.

(13)

1 EARLY ONSET SEPSIS: A DIAGNOSTIC AND THERAPEUTIC

_DILEMMA

EOS can present with relatively mild symptoms resulting in late discovery with high risk for mortality and morbidity. Furthermore, clinical symptoms of EOS are extremely diverse and difficult to distinguish from physiologic symptoms of neonatal transition from intra-to-extrauterine life and other non-infectious neonatal disease [3,9,26]. This complicates clinical decision-making on start and duration of antibiotic treatment leading to significant overtreatment. For example, in the European Union almost 8% of newborns are treated with antibiotics for suspected EOS, while incidence rates of bacterial culture proven EOS range from 0.01 to 0.53 per 1000 live births in those countries [3].

Blood culturing is considered the golden standard diagnostic test for EOS and takes several days to become positive. Upon suspicion of EOS, newborns are observed and treated empirically for EOS with antibiotics for at least 48 hours until results of blood culturing are known [1]. However, blood cultures are only positive in 0.01 to 1.2 per 1000 live births in countries in the European Union and North America. Contributing to this low prevalence may be false negativity due to low yield of bacteria in low sample volumes or low-density bacteremia in general. Nonetheless, over 60% of newborns empirically treated with antibiotics for suspected EOS are treated for longer than 72 hours even when blood cultures are negative [27]. Antibiotic stewardship is necessary to reduce this overtreatment [28].

These dilemmas in the management of EOS pose a huge cost and socioeconomic threat, especially in non-Western countries [1,6,16]. Moreover, it is becoming clear that prolonged treatment of newborns with antibiotics also can negatively and severely impact early and long-term neurodevelopment, growth, the developing immune system, and gut microbiome resistance patterns [29-32].

CURRENT APPROACHES IN PREDICTION OF EARLY ONSET SEPSIS

Since clinical presentation and blood culturing have poor specificity for EOS, additional approaches to aid clinical decision-making whether to start and/or continue antibiotic treatment have been developed in the recent decade. Approaches that are commonly used in the clinic include maternal risk factor stratification and serial measurement of C-reactive protein (CRP) levels and leukocyte counts. Each of these has limitations in clinical utility, as will be discussed below.

Maternal Risk Factor Stratification

Maternal risk factors for EOS (i.e., presence and duration of prolonged rupture of the membranes, intrapartum fever or administration of antibiotics, and presence of maternal GBS colonization, as the most common cause of EOS in Western countries, have been used to predict presence of EOS in newborns. In an attempt to overcome the problem of antibiotic overtreatment amongst near and at term newborns with a gestational age equal or above 34 weeks, a risk stratification strategy based on these factors and neonatal clinical findings has been developed in 2010 by Escobar et al., which was revised in 2014 (33). This EOS calculator (available online at https://

(14)

1

neonatalsepsiscalculator.kaiserpermanente.org) provides a quantitative estimation of EOS risk along with a recommendation whether to start antibiotic treatment. Since its inception, two retrospective studies revealed that application of the EOS calculator might help to reduce antibiotic therapy with 50% (34,35). Additionally, the EOS calculator uses local incidence rates of EOS as a variable, which still have to be established in many non-Western countries.

Correlation of results of the EOS calculator with biomarkers of inflammation in the newborn may be helpful in further increasing its clinical utility. Therefore, the study in Chapter 3 explores the relationship of results from the EOS calculator with results of serial measurement of CRP and leukocyte counts in a cohort of Dutch near and at term newborns. For this study we hypothesized that higher EOS calculator result, indicating higher risk for EOS, corresponds with an increase in CRP and low leukocyte counts.

C-reactive Protein

CRP is an endogenous acute phase reactant synthesized by the liver upon infection [36]. Serum CRP in newborns always represents endogenous synthesis since it passes the placenta in extremely low quantities [37]. CRP is constitutively present in serum of newborns at very low concentrations and its levels are dependent on gestational age and birth weight. CRP synthesis starts immediately after an inflammatory stimulus by chemokines, such as interleukin (IL)-1, and IL-6, with serum concentrations rising above the usual laboratory threshold of 5 mg/L after 6 hours and peaking after 48 hours. This delayed synthesis results in poor sensitivity of CRP levels during early EOS. In most practices, in the newborn suspected and treated with antibiotics for EOS, a repeat CRP level below the laboratory threshold measured between 24 to 48 hours after start of antibiotics has negative predictive value of 99% for EOS, yet only in case of a negative blood culture plus a clinically improved newborn [37]. However, in clinical practice, despite this strong negative value, the repeat CRP also leads to even more testing, culturing, and longer treatment duration and hospital stay (38).

Leukocyte Counts

Inflammation and infection causes release of leukocytes from the bone marrow into the circulation. Leukocyte counts (both total and subset, predominantly neutrophil, counts) have been widely used to assess EOS [1,3]. However, both leukocyte and neutrophil counts lack specificity for prediction of EOS [39,40]. Their numbers are dependent on many perinatal factors such as gestational age, birth weight, type of delivery, and post partum age [41]. Neutropenia has shown the most specificity for EOS [42]. However, as discussed above, due to low prevalence of positive blood cultures, clinical decision-making on start and duration of antibiotic treatment is often based on non-specific clinical symptoms and repeated measurement of CRP. Serial measurement of low immature-to-total granulocyte (I/T) ratio has been showen to have a negative predictive value for blood culture positive EOS of 99% [42]. Chapter 4 explores the relevance of a one-point automated I/T ratio determination in prediction of duration of antibiotic therapy in a retrospective cohort of Surinamese newborns with suspected EOS. For this study, we hypothesized that early

(15)

1

establishment of a one-point low I/T ratio is associated with short duration of antibiotic treatment in suspected EOS. This may prevent start of unnecessary antibiotic treatment, which may help to reduce the antibiotic burden in developing countries.

EARLY ONSET SEPSIS: A NEED FOR NOVEL DIAGNOSTIC

STRATEGIES

The approaches described above have been used for over 20 years and have remained virtually unchanged. A recent international survey established that in practice only 31% of clinicians use CRP levels and leukocyte counts as arguments for the decision to start antibiotics [43]. Many other biomarkers have been investigated, but have not made it into the clinic for various reasons such as poor specificity, short half lives of biomarkers, lack of reproducibility, or technical issues [44]. At this point, serial measurement of procalcitonin, an acute phase reactant similar to CRP, is showing promise in negative prediction of EOS and reduction of antibiotic treatment in Western countries [45]. However, novel and practical approaches for early and prompt confirmation or exclusion of EOS remain necessary to reduce antibiotic overtreatment, while improving outcomes. Elements of the vascular pathophysiology may be relevant for development of these novel approaches, which will be discussed below.

THE VASCULAR PATHOPHYSIOLOGY OF EARLY ONSET SEPSIS

The diagnostic and therapeutic dilemmas of EOS occur, at least in part, because its pathophysiology remains poorly understood. Endothelial inflammatory activation and leukocyte-endothelial interactions are key processes in sepsis pathophysiology. Part 3 is aimed to provide more insight into these processes in newborns to unravel aspects of EOS pathophysiology and provide novel concepts for its timely diagnosis and management.

LEUKOCYTE-ENDOTHELIAL INTERACTIONS: SHEDDING OF

ADHESION MOLECULES IN EARLY ONSET SEPSIS

Leukocyte-endothelial interactions are involved in any infectious pathophysiology [46]. A body of evidence is indicating that aberrant leukocyte, mostly neutrophil, activation and recruitment towards the endothelium plays a pivotal role in breakdown of the vascular endothelium, which, in turn, is associated with organ failure and death [47,48]. Bacterial derived lipopolysaccharide (LPS) drives release of cytokines, such as tumor necrosis-α and interleukins, known as the ‘cytokine storm’. Additionally, the endothelium becomes activated and increased presence of LPS in the vasculature is associated with increased expression of endothelial cell adhesion molecules (CAM) P-selectin, E-selectin, vascular cell adhesion molecule (VCAM-1), intercellular adhesion molecule (ICAM-1) and platelet and endothelial cell adhesion molecule-1 (PECAM-1) [49]. These adhesion molecules orchestrate tethering, rolling and firm adhesion of leukocytes on and transmigration across the endothelium [50]. During sepsis, soluble isoforms of adhesion molecules (sCAMs) accumulate in the bloodstream due to shedding [51]. Shedding represents removal of CAMs from cell surfaces

(16)

1

by enzymes called sheddases, in particular matrix metalloproteinase-9 (MMP-9) and neutrophil elastase (NE), released from tertiary granules in neutrophils [51]. Both MMP-9 and NE prepare the extracellular matrix underlying the endothelium to allow transmigration of leukocytes into inflammatory sites. The activity of MMP-9 is tightly regulated by sheddase antagonist tissue-inhibitor of metalloproteinases-1 (TIMP-1) to reduce damage to host-tissues and an increased TIMP-1/MMP-9 ratio was associated with severity and outcome of sepsis in adults [52,53].

Chapter 5 reviews mechanisms for changes in levels of circulating adhesion molecules and their

sheddases during sepsis and age-dependency of their levels in newborns, children and adults. For Chapter 6 we applied the concept of simultaneous measurement of circulating adhesion molecules and their sheddases in a cohort of healthy newborns and newborns with suspected EOS. We hypothesized that higher circulating levels of adhesion molecules sP-selectin, sE-selectin sVCAM-1, sICAM-1 and sPECAM-1, coincide with higher levels of sheddases MMP-9 and NE, and sheddase antagonist TIMP-1 in newborns with culture proven EOS versus healthy controls.

ENDOTHELIAL INTEGRITY DURING EARLY ONSET SEPSIS:

THE ANGIOPOIETINS

Endothelial integrity is maintained by the Angiopoietin/Tie2 Receptor Tyrosine Kinase - system, which consists of the endothelial restricted receptor Tie-2 and its ligands Angiopoietin (Ang)-1 and Ang-2 [54]. In health, Ang-1 is present in human serum at higher levels than Ang-2 and promotes endothelial stability through continuous endothelial Tie-2 receptor phosphorylation [55]. Inflammation leads to higher circulating levels of Ang-2 that is being release from endothelial cells. Circulating Ang-2 dose-dependently inhibits Tie-2 signaling and acts as an antagonist of Ang/ Tie-2, driving vascular permeability. Emerging clinical evidence indicates a positive correlation of high Ang-2 levels, and subsequent high Ang-2/Ang-1 ratio with presence, severity, and outcome of pediatric and adult sepsis [56,57]. It was recently suggested that the Angiopoietins may be relevant as biomarkers of EOS [58]. Additionally, investigating the dynamics of Ang-1 and Ang-2 in healthy and infected newborns may unravel changes in their levels during EOS. In Chapter 7, these changes are explored in a large cohort of healthy newborns and newborns with suspected and culture proven EOS. For this study, we hypothesized that low Ang-1 and high Ang-2 levels are associated with presence of bacterial culture positive EOS.

NOVEL ASPECTS OF NEUTROPHILS IN SEPSIS

Manual microscopic analysis of neutrophils and their counts have been part of the clinical assessment of bacterial infection for over a century [59]. However, manual analysis of counts and morphology is time consuming, requires experienced laboratory technicians, and lacks reproducibility. Novel methods allow for measurement of several aspects of neutrophils, in particular morphology, mechanics and motility. Flow-based automated hematology analysers (AHAs) are able to determine leukocyte subsets and different granulocyte fractions [42]. Additionally, recent developments in the performance of these AHAs have enabled measurement of neutrophil size and scatter properties and determination of neutrophil cell surface markers with

(17)

1

immunofluorescence, each with their own sensitivity for presence of sepsis in patients. Chapter 8 and Chapter 9 discuss basic and clinical aspects of neutrophil morphology, mechanics and motility during sepsis, along with current evidence and future possibilities for the use of these parameters into the management of sepsis.

(18)

1 REFERENCES

1. Simonson KA, Anderson-Berry AL, Delair SF, Davies HD. Early-onset neonatal sepsis. Clin Microbiol Rev 2014, 27(1):21-47.

2. Schrag SJ, Farley MM, Petit S, Reingold A, Weston EJ, Pondo T, Hudson Jain J, Lynfield. Epidemiology of Invasive Early-Onset Neonatal Sepsis, 2005 to 2014. Pediatrics 2016, 138(6), pii: e20162013. 3. van Herk W, Stocker M, van Rossum AM.

Recognising early onset neonatal sepsis: an essential step in appropriate antimicrobial use. J Infect 2016, 72:S77-82.

4. Cohen-Wolkowiez M, Moran C, Benjamin DK, Cotten CM, Clark RH, Benjamin DK, Jr, Smith PB. 2009. Early and late onset sepsis in late preterm infants. Pediatr. Infect. Dis. J. 2009, 28:1052–1056. 5. Stoll BJ, Hansen NI, Sanchez PJ, Faix RG,

Poindexter BB, Van Meurs, KP, Bizzarro MJ, Goldberg RN, Frantz ID, III, Hale EC, Shankaran S, Kennedy K, Carlo WA, Watterberg KL, Bell EF, Walsh MC, Schibler K, Laptook AR, Shane AL, Schrag SJ, Das A, Higgins RD, Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. Early onset neonatal sepsis: the burden of group B streptococcal and E. coli disease continues. Pediatrics 2011, 127:817–826. 6. Weston EJ, Pondo T, Lewis MM, Martell-Cleary

P, Morin C, Jewell B, Daily P, Apostol M, Petit S, Farley M, Lynfield R, Reingold A, Hansen NI, Stoll BJ, Shane AJ, Zell E, Schrag SJ.The burden of invasive early-onset neonatal sepsis in the United States, 2005-2008. Pediatr Infect Dis J 2011, 30(11):937-41.

7. Schrag SJ, Verani JR. Intrapartum antibiotic prophylaxis for the prevention of perinatal group B streptococcal disease: experience in the United States and implications for a potential group B streptococcal vaccine. Vaccine 2013, 31(4):D20-6.

8. Bauserman MS, Laughon MM, Hornik CP, Smith PB, Benjamin DK Jr, Clark RH, Engmann C, Cohen-Wolkowiez M. Group B Streptococcus and Escherichia coli infections in the intensive care nursery in the era of intrapartum antibiotic prophylaxis. Pediatr Infect Dis J 2013, 32(3):208-12.

9. Stoll BJ. Early-onset neonatal sepsis: a continuing problem in need of novel prevention strategies. Pediatrics 2016, 138(6): e20163038.

10. Ohlsson A, Shah VS. Intrapartum antibiotics for known maternal Group B streptococcal colonization. Cochrane Database Syst Rev 2014, (6):CD007467. 11. Li S, Huang J, Chen Z, Guo D, Yao Z, Ye X.

Antibiotic Prevention for Maternal Group B Streptococcal Colonization on Neonatal GBS-Related Adverse Outcomes: A Meta-Analysis. Front Microbiol 2017, 8:374.

12. Dangor Z, Cutland CL, Izu A, Kwatra G, Trenor S, Lala SG, Madhi SA. Temporal Changes in Invasive Group B Streptococcus Serotypes: Implications for Vaccine Development. PLoS One 2016, 11(12):e0169101.

13. Chan GJ, Lee AC, Baqui AH, Tan J, Black RE. Prevalence of early-onset neonatal infection among newborns of mothers with bacterial infection or colonization: a systematic review and meta-analysis. BMC Infect Dis 2015, 15:118. 14. Le Doare K, Heath PT. An overview of global

GBS epidemiology. Vaccine 2013, 28:31(4):7-12. 15. Edmond KM, Kortsalioudaki C, Scott S, Schrag

SJ, Zaidi AK, Cousens S, Heath PT. Group B streptococcal disease in infants aged younger than 3 months: systematic review and meta-analysis. Lancet 2012, 379(9815):547-56. 16. Sinha A, Russell LB, Tomczyk S, Verani JR, Schrag

SJ, Berkley JA, Mohammed M, Sigauque B, Kim SY; GBS Vaccine Cost-Effectiveness Analysis in Sub-Saharan Africa Working Group. Disease Burden of Group B Streptococcus Among Infants in Sub-Saharan Africa: A Systematic Literature Review and Meta-analysis. Pediatr Infect Dis J 2016, 35(9):933-42.

17. Lawn JE, Cousens S, Zupan J; Lancet Neonatal Survival Steering Team. 4 million neonatal deaths: when? Where? Why? Lancet 2005, 365(9462):891-900.

18. Chow S, Chow R, Popovic M, et al. A Selected Review of the Mortality Rates of Neonatal Intensive Care Units. Front Public Health 2015, 7;3:225 19. Quan V, Verani JR, Cohen C, von Gottberg

(19)

1

SA. Invasive Group B Streptococcal Disease in South Africa: Importance of Surveillance Methodology. PLoS One 2016, 11(4):e0152524. 20. Dagnew AF, Cunnington MC, Dube Q, Edwards

MS, French N, Heyderman RS, Madhi SA, Slobod K, Clemens SA. Variation in reported neonatal group B streptococcal disease incidence in developing countries. Clin Infect Dis 2012, 55(1):91-102.

21. Census Statistics Suriname, 2012. http://www. statistics-suriname.org

22. Mazoembaks N. Verzwegen werk van P.C. Flu: kindersterfte in Suriname, een erfenis uit de slavernij. Uitgeverij De Woordenwinkel, Zierikzee 2014, pp 48 and 80.

23. Zijlmans W, Hindori-Mohangoo A. Determinants of neonatal mortality in Suriname: preliminary findings from a perinatal and infant mortality survey. Ann of Glob Health 2015, 81(1): 121-122.

24. WHO data, 2015. http://www.who.int

25. Bekker V, Bijlsma MW, van de Beek D, Kuijpers TW, van der Ende A. Incidence of invasive group B streptococcal disease and pathogen genotype distribution in newborn babies in the Netherlands over 25 years: a nationwide surveillance study. Lancet Infect Dis 2014, 14(11):1083-9.

26. van den Anker JN. How to optimize the evaluation and use of antibiotics in neonates. Early Hum Dev 2014, 90(1):S10-2. 27. Canty JB, Wozniak PS, Sanches PJ. Prospective

surveillance of antibiotic use in the neonatal intensive care unit: results from the SCOUT study. Pediatr Infect Dis J 2015;34:267-72. 28. Cantey JB, Wozniak PS, Pruszynski JE, Sánchez

PJ. Reducing unnecessary antibiotic use in the neonatal intensive care unit (SCOUT): a prospective interrupted time-series study. Lancet Infect Dis 2016, 16(10):1178-84.

29. Cotten CM. Adverse consequences of neonatal antibiotic exposure. Curr Opin Pediatr 2016, 28(2):141-9.

30. Langdon A, Crook N, Dantas G. The effects of antibiotics on the microbiome throughout development and alternative approaches for therapeutic modulation. Genome Med, 2016, 8(1):39.

31. Cotten CM, Taylor S, Stoll B, Goldberg RN, Hansen NI, Sánchez PJ, Ambalavanan N, Benjamin DK; NICHD Neonatal Research Network. Prolonged duration of initial empirical antibiotic treatment is associated with increased rates of necrotizing enterocolitis and death for extremely low birth weight infants. Pediatrics 2009, 123(1):58–66. 32. Kuppala VS, Meinzen-Derr J, Morrow AL,

Schibler KR. Prolonged initial empirical antibiotic treatment is associated with adverse outcomes in premature infants. J Pediatr 2011, 159(5):720–725.

33. Escobar GJ, Puopolo KM, Wi S, et al. Stratification of risk of early-onset sepsis in newborns ≥ 34 weeks’ gestation. Pediatrics. 2014;133:30–36. 34. Shakib J, Buchi K, Smith E, et al. Management

of newborns born to mothers with chorioamnionitis: Is it time for a kinder, gentler approach? Acad. Pediatr, 2015;15:340–344. 35. Kerste M, Corver J, Sonnevelt MC, van Brakel

M, van der Linden PD, M Braams-Lisman BA, Plötz FB. Application of sepsis calculator in newborns with suspected infection. J Matern Neonatal Med 2016;7058:1–6.

36. Pepys MB, Hirschfield GM. C-reactive protein: a critical update. J Clin Invest 200, ;111:1805–1812. 37. Hofer N, Zacharias E, Müller W, Resch B. An

update on the use of C-reactive protein in early-Onset neonatal sepsis: Current insights and new tasks. Neonatology 2012, 102:25–36. 38. Mukherjee A, Davidson L, Anguvaa L, Duffy

DA, Kennea N. NICE neonatal early onset sepsis guidance: greater consistency, but more investigations, and greater length of stay. Arch Dis Child Fetal Neonatal Ed 2015, 100(3):F248-9. 39. Newman TB, Puopolo KM, Wi S, Draper D,

Escobar GJ. Interpreting complete blood counts soon after birth in newborns at risk for sepsis. Pediatrics 2010, 126:903–90.

40. MacQueen BC, Christensen RD, Yoder BA, Henry E, Baer VL, Bennett ST, Yaish HM. Comparing automated vs manual leukocyte differential counts for quantifying the ‘left shift’ in the blood of neonates. J Perinatol 2016, 36(10):843-8. 41. Wiland EL, Sandhaus LM, Georgievskaya Z, Hoyen

(20)

1

automated immature granulocyte norms are inappropriate for evaluating early-onset sepsis in newborns. Acta Paediatr 2014;103(5):494-7. 42. Mikhael M, Brown LS, Rosenfeld CR. Serial

neutrophil values facilitate predicting the absence of neonatal early-onset sepsis. J Pediatr 2014;164(3):522-8.

43. van Herk W, el Helou S, Janota J, Hagmann C, Klingenberg C, Staub E, Giannoni E, Tissieres P, Schlapbach LJ, van Rossum AM, Pilgrim SB, Stocker M. Variation in Current Management of Term and Late-preterm Neonates at Risk for Early-onset Sepsis: An International Survey and Review of Guidelines. Pediatr Infect Dis J 2016, 35(5):494-500. 44. Chauhan N, Tiwari S, Jain U. Potential

biomarkers for effective screening of neonatal sepsis infections: An overview. Microb Pathog 2017, 107:234-242.

45. Stocker M, van Herk W, El Helou S, Dutta S, Fontana MS, Schuerman FABA, van den Tooren-de Groot RK, Wieringa JW, Janota J, van der Meer-Kappelle LH, Moonen R, Sie SD, de Vries E, Donker AE, Zimmerman U, Schlapbach LJ, de Mol AC, Hoffman-Haringsma A, Roy M, Tomaske M, Kornelisse RF, van Gijsel J, Visser EG, Willemsen SP, van Rossum AMC; NeoPInS Study Group. Procalcitonin-guided decision making for duration of antibiotic therapy in neonates with suspected early-onset sepsis: a multicentre, randomised controlled trial (NeoPIns). Lancet 2017, pii. 6736(17)31444-7. 46. Aird WC. The role of the endothelium in

severe sepsis and multiple organ dysfunction syndrome. Blood 2003, 101(10):3765-77. 47. Brown KA, Brain SD, Pearson JD, Edgeworth

JD, Lewis SM, Treacher DF. Neutrophils in development of multiple organ failure in sepsis. Lancet 2006, 368:157-69.

48. Shapiro NI, Schuetz P, Yano K, Sorasaki M, Parikh SM, Jones AE, Trzeciak S, Ngo L, Aird WC. The association of endothelial cell signalling, severity of illness, and organ dysfunction in sepsis. Crit Care 2010, 14(5):R182.

49. Kümpers P, van Meurs M, David S, Molema G, Bijzet J, Lukasz A, Biertz F, Haller H, Zijlstra JG. Time course of angiopoietin-2 release during

experimental human endotoxemia and sepsis. Crit Care 2009, 13(3):R64.

50. Ley K, Laudanna C, Cybulsky MI, Nourshargh S. Getting to the site of inflammation: the leukocyte adhesion cascade updated. Nat Rev Immunol 2007, 7(9):678-89.

51. Garton KJ, Gough PJ, Raines EW. Emerging roles for ectodomain shedding in the regulation of inflammatory responses. J Leukoc Biol 2006, 79(6):1105-16.

52. Lorente L, Martín MM, Solé-Violán J, Blanquer J, Labarta L, Díaz C, Borreguero-León JM, Orbe J, Rodríguez JA, Jiménez A, Páramo JA. Association of sepsis-related mortality with early increase of TIMP-1/MMP-9 ratio. PLoS One 2014, 9(4):e94318.

53. Wang M, Zhang Q, Zhao X, Dong G, Li C. Diagnostic and prognostic value of neutrophil gelatinase-associated lipocalin, matrix metalloproteinase-9, and tissue inhibitor of matrix metalloproteinases-1 for sepsis in the Emergency Department: an observational study. Crit Care 2014, 18(6):634.

54. van Meurs M, Kümpers P, Ligtenberg JJ, Meertens JH, Molema G, Zijlstra JG. Bench-to-bedside review: Angiopoietin signalling in critical illness - a future target? Crit Care 2009, 13(2):207.

55. Parikh SM. Dysregulation of

the angiopoietin-Tie-2 axis in sepsis and ARDS. Virulence 2013, 4(6):517-24.

56. Fang Y, Li C, Shao R, Yu H, Zhang Q, Zhao L. Prognostic significance of the angiopoietin-2/ angiopoietin-1 and angiopoietin-1/Tie-2 ratios for early sepsis in an emergency department. Crit Care 2015, 14(19):367.

57. Mussap M, Cibecchini F, Noto A, Fanos V. In search of biomarkers for diagnosing and managing neonatal sepsis: the role of angiopoietins. J Matern Fetal Neonatal Med 2013, 26(2):24-6. 58. Giuliano JS Jr, Tran K, Li FY, Shabanova V,

Tala JA, Bhandari V. The temporal kinetics of circulating angiopoietin levels in children with sepsis. Pediatr Crit Care Med 2014, 15(1):e1-8. 59. Cornbleet PJ. Clinical utility of the band count.

(21)

(22)

(23)

(24)

I

Epidemiology of

Early Onset Sepsis in Suriname

(25)

(26)

(27)

(28)

2

Improved Referral and Survival of

Newborns after Scaling Up of

Intensive Care in Suriname

Rens Zonneveld, Natanael Holband, Anna Bertolini, Francesca Bardi, Neirude Lissone, Peter Dijk, Frans B. Plötz, Amadu Juliana

(29)

N EO N A TA L I N TE N SIV E C A RE IN S U RIN A M E 28

2 ABSTRACT

Background

Scaling up neonatal care facilities in developing countries can improve survival of newborns. Recently, the only tertiary neonatal care facility in Suriname transitioned to a modern environment in which interventions to improve intensive care were performed. This study evaluates impact of this transition on referral pattern and outcomes of newborns.

Methods

A retrospective chart study amongst newborns admitted to the facility was performed and outcomes of newborns between two 9-month periods before and after the transition in March 2015 were compared.

Results

After the transition more intensive care was delivered (RR 1.23; 95% CI 1.07-1.42) and more outborn newborns were treated (RR 2.02; 95% CI 1.39-2.95) with similar birth weight in both periods (P=0.16). Mortality of inborn and outborn newborns was reduced (RR 0.62; 95% CI 0.41-0.94), along with mortality of sepsis (RR 0.37; 95% CI 0.17-0.81) and asphyxia (RR 0.21; 95% CI 0.51-0.87). Mortality of newborns with a birth weight <1000 grams (34.8%; RR 0.90; 95% CI 0.43-1.90) and incidence of sepsis (38.8%, 95% CI 33.3-44.6) and necrotizing enterocolitis (NEC) (12.5%, 95% CI 6.2-23.6) remained high after the transition.

Conclusions

After scaling up intensive care at our neonatal care facility more outborn newborns were admitted and survival improved for both in- and outborn newborns. Challenges ahead are sustainability, further improvement of tertiary function, and prevention of NEC and sepsis.

(30)

2 BACKGROUND

Neonatal mortality in developing countries continues to be a chief global health challenge [1,2]. A recent global report indicates that over 40% reduction of neonatal mortality can be achieved by implementation of institutional care in lower resource countries [3]. In particular, local or regional neonatal care facilities with integrated availability of perinatal and neonatal intensive care can reduce mortality [4]. For example, newborns born in a rural hospital featuring a neonatal intensive care unit (NICU) in Uganda were almost twice as likely to survive than those born outside [5]. Moreover, introduction of a neonatal care facility in a low-income district in India reduced neonatal mortality rate (NMR) by 21% after the first two years [6]. Improving interventions within existing neonatal care facilities (e.g., training of personnel, refurbishment, infection prevention) can improve mortality and enhance tertiary function for newborns in need of intensive care [6-9].

In Suriname NMR in 2009 was 16.0 per 1000 live births. However, detailed data on demographics and outcomes of newborns are lacking. In 2008 the neonatal care facility at the Academic Hospital Paramaribo (AHP), which also incorporated the first and only NICU in Suriname, opened its doors. The ability to treat premature and critically ill newborns was an important step towards reducing mortality. At the end of March 2015 the facility moved to a new and modern environment. This transition solidified availability of neonatal intensive care in Suriname with reinforcement and training of personnel, new equipment, continuous availability of supplies, and protocol-based care. Since this facility is the only referral center for newborns requiring intensive care in Suriname, morbidity and mortality of newborns treated here reflect their outcomes at the national level.

Therefore, as a benchmark for future investigations, we developed a registry to describe demographics and outcomes of newborns admitted to the neonatal care facility. Additionally, to evaluate the impact of improvements we compare referral pattern, mortality and morbidity of newborns treated in periods before and after the transition. Ultimately, this could lead to better prospective registry and care for critically ill newborns in Suriname.

METHODS

Study Design

We performed a retrospective (pre-and post transition) study in the neonatal care facility of the AHP during the periods July 1st_{2014 to March 29}th _{2015 (Period 1) and March 31}st_{to December}

31st_{2015 (Period 2). The impact of the transition was described by analyzing demographics and}

outcomes of all inborn and outborn newborns admitted within these two periods. Excluded were newborns whom were treated in both periods and of whom insufficient information (i.e., no or incomplete paper charts) was available to confirm outcomes. We received a waiver from our institutional ethical board.

Setting and Interventions

Suriname is a small middle-income country with a multiethnic society and has an annual birth rate of about 10,000 births. Over 90% of births take place at delivery rooms of one of four hospitals situated in Suriname’s capital Paramaribo (inhabited by more than half of Suriname’s population).

(31)

2

About 30% of births take place at the delivery room of the AHP. The neonatal care facility at the AHP serves as the only referral hospital for critically ill newborns. Since the opening in 2008, between 350-400 newborns are treated each year in one room with 12 beds, with NICU capacity operating at Level III [9]. Newborns are generally only actively treated with a birth weight (BW) ≥ 750 grams and/or gestational age (GA) ≥ 27 weeks.

On March 30th_{2015 the facility moved to a completely new, modern and spacious environment}

with central climate control and new equipment (i.e., ventilators, incubators, air-humidifiers, ultrasound machines and multi-parameter monitors). Capacity for mechanical ventilation and continuous positive airway pressure (CPAP) was doubled. The NICU (6 beds), high care (HC) (6 beds), and medium care (MC) (4 beds) capacity in the new facility remained the same until February 2016 (when a separate space for the MC was opened and the NICU capacity increased to 10 beds).

Total expense for the new building and equipment was 2.6 million US dollars. Funds were collected from kind donations from governmental and private organizations and from Surinamese companies. Since there were no architects or contractors available within Suriname with experience in designing a NICU level neonatal care facility, we relied on guidelines from developed countries and local creativity and practical experience to realize the project within budget, without the need for expensive consultants. For example, one of the savings came from using venturi mechanism based suction devices powered by compressed air, avoiding the need for a separate central vacuum system.

Admission criteria remained the same. Obstetric nurses were trained in neonatal life support and the number of residents in the obstetric and pediatric department was increased. For both day and evening shifts a separate resident was assigned to the NICU exclusively. Shortly before the transition, nurses were trained in intensive neonatal care and their number was expanded to 1 per 3 or 4 beds. New charts for vital signs, ventilation settings, and fluid management were implemented. A breast-feeding and nutrition program was started to help reduce cases of necrotizing enterocolitis (NEC) and mothers were allowed at the bedside twice as long as before. Systematic infection prevention (i.e., stringent guidelines and more facilities for hand washing, providing of patient specific (disposable) materials, Extended Spectrum Beta-Lactamase (ESBL) outbreak control) was enforced.

Data Collection and Analysis

Data were collected from paper medical records on maternal, obstetric and perinatal history, birth location, reason for admission, hospital course, and outcomes. A single major cause of death was determined. For each included newborn we determined the highest level of care during their stay by assigning criteria for NICU, HC or MC retrospectively according to local protocol (Supplemental Table 1). Primary outcome was mortality: NMR at the AHP and at the neonatal care facility divided in early (i.e., in-hospital death before 7 days of life) and late (i.e., in-hospital death of at term newborns after 7 days of life), GA-specific mortality, BW-specific mortality, and cause-specific mortality. Secondary outcomes were highest level of care, respiratory treatments (CPAP, mechanical ventilation, surfactant), use of antibiotics, development of respiratory complications,

(32)

2

i.e., pneumothorax, bronchopulmonary dysplasia (BPD; i.e., oxygen dependence > 28 days of age), ventilator-associated pneumonia (VAP; i.e., positive tracheal aspirate culture after ventilation), development of NEC and sepsis (i.e., early (<72 hours after birth) and late (>72 hours after birth) onset clinical (i.e., clinical suspicion, treated with antibiotics for 7 days, raised c-reactive protein levels)) and blood culture positive sepsis, blood and ESBL culture results, and duration of stay.

Statistical Analysis

Incidence rates and epidemiological determinants were calculated for the inclusion period. Categorical variables are presented as numbers and percentages with 95% confidence intervals (CI) and continuous variables as means with standard deviations (SD) or, if not normally distributed, as medians with ranges. Continuous variables were compared with a student t-test and categorical variables were compared with Chi-Square. Relative risk (RR) and 95% CI were calculated. P-values < 0.05 were considered statistically significant.

RESULTS

Demographics and Referral

A total of 626 newborns were treated at the neonatal care facility of whom 601 (320 before and 281 after the transition) were included (Table 1). Overall demographics were comparable between both periods, with similar percentages of missing data, showing high prevalence of (antenatal) risk factors for mortality and morbidity (Table 1). In period 2 significantly more outborn newborns (RR 2.02; 95% CI 1.39-2.95; P<0.001) were treated with similar mean birthweight (2183 ± 845 grams vs. 1915 ± 990 grams; P=0.16). Prematurity was the main reason for admission for all inborn (48.3%; 95% CI 44.0-52.7) and outborn (66.0%; 95% CI 56.3-74.5) newborns, followed by respiratory distress and suspected infection (Table 1).

Mortality

NMR of inborn newborns born at the AHP was lower in period 2 (P=0.02) (Table 2). After the transition, reduction in mortality was greatest in newborns treated at NICU level care (P<0.01), with a GA above 28 weeks (RR 0.42; 95% CI 0.25-0.72; P=0.002), and outborn newborns (P=0.02). A trend in decrease in mortality was observed in late mortality (P=0.06), inborn newborns (P=0.07), and in newborns with a birth weight (BW) above 1500 grams (P=0.07). A significant reduction in mortality was observed in cases of sepsis (P=0.01) and perinatal asphyxia (P=0.03). Sepsis was the main cause of death in period 1 (34.5%; 95% CI 23.4-47.7), and second in period 2 (26.7%; 95% CI 14.2-44.4). For newborns with a BW<1000 grams late-onset sepsis was the main cause of death in both periods (44.8%; 95% CI 28.4-62.5).

Treatments and Morbidity

Based on our criteria (Supplemental Table 1) significantly more NICU level care was given in period 2 (P<0.01) (Table 3a). More mechanical ventilation and surfactant were applied after the transition. No difference in prevalence of VAP or pneumothorax was observed and there was

(33)

2

Table 1. Demographics of newborns admitted to the neonatal care facility before and after the transition Period 1

(July 2014-March 2015)

Period 2

(April 2015-December 2015)

N % (95% CI) N % (95% CI)

Live births Total at AHP 2353 1972

Admissions to facility Total Included Inborn Outborn2 331 320 284 36 96.7 88.7 (84.8-91.8) 11.3 (8.2-15.2) 295 281 217 64 95.3 77.2 (72.0-81.7) 22.8 (18.3-28.0) Maternal age (Years) <20 20-34 ≥35 Missing 54 168 46 52 16.9 (13.2-21.4) 52.5 (47.0-57.9) 14.4 (11.0-18.6) 16.3 36 140 24 81 12.8 (9.4-17.2) 49.8 (44.0-55.6) 8.5 (5.8-12.4) 28.8 Pregnancy HIV Diabetes PIH / Preeclampsia Antenatal steroids3 Infection risk4 6 18 60 47 47 1.9 (0.9-4.0) 5.6 (3.6-8.7) 18.8 (14.9-23.4) 14.7 (11.2-19.0) 14.7 (11.2-19.0) 2 20 62 55 38 0.7 (0.2-2.6) 7.1 (4.7-10.7) 22.1 (17.6-27.3) 19.6 (15.4-24.6) 13.5 (10.0-18.0) Mode of delivery Vaginal

Caesarean section Missing 187 105 28 58.4 (53.0-63.7) 32.8 (27.9-38.1) 8.8 167 94 20 59.4 (53.6-65.0) 33.5 (28.2-39.2) 7.1 Sex Male Female 162 158 50.6 (45.2-56.1) 49.4 (43.9-54.8) 155 126 55.2 (49.3-60.9) 44.8 (39.1-50.7) Gestational age (Weeks) <28 28-32 33-36 ≥37 Missing 16 48 114 132 10 5.0 (3.1-8.0) 15.0 (11.5-19.3) 35.6 (30.6-41.0) 41.3 (36.0-46.7) 3.1 13 47 100 110 11 4.6 (2.7-7.8) 16.7 (12.8-21.5) 35.6 (30.2-41.3) 39.1 (33.6-45.0) 3.9 Birth weight (Grams) <1000 ≥1000-1499 ≥1500 Missing 26 48 242 4 8.1 (5.6-11.6) 15.0 (11.5-19.3) 75.6 (70.6-80.0) 1.3 23 33 221 4 8.2 (5.5-12.0) 11.7 (8.5-16.0) 78.6 (73.5-83.0) 1.4 Apgar Score at 5’ <5 Missing 24 45 7.5 (5.1-10.9) 14.1 7 47 2.5 (1.2-5.1) 16.7 (12.8-21.5)

(34)

2

a trend in increases incidence of BPD (P=0.07) (Table 3b). Grade 2 or higher NEC was present at high incidence in newborns with a BW<1500 grams in both periods (5.4% and 12.5%, respectively). Sepsis (either early or late-onset) was prevalent in over 30% of patients in both periods, of which half was LOS. During both periods, outbreaks with ESBL bacteria led to a significant prevalence of ESBL positive cultures.

Table 1. (continued) Period 1 (July 2014-March 2015) Period 2 (April 2015-December 2015) N % (95% CI) N % (95% CI) Ethnicity Maroon Creole Hindo-Surinamese Javanese Amerindian Chinese Other5 Missing 87 85 59 15 10 2 31 31 27.2 (22.6-32.3) 26.2 (22.0-31.7) 18.4 (14.6-23.1) 4.7 (2.9-7.6) 3.1 (1.7-5.7) 0.6 (0.2-2.2) 9.7 (6.9-13.4) 9.7 72 72 55 21 7 2 32 20 25.6 (20.9-31.0) 25.6 (20.9-31.0) 19.6 (15.4-24.6) 7.5 (4.9-11.2) 2.5 (1.2-5.1) 0.7 (0.2-2.6) 11.4 (8.2-15.6) 7.1

Initial reason for admission1 Prematurity Respiratory distress6 Suspected infection7 Perinatal asphyxia8 Congenital malformations9 Other10 152 119 91 39 42 71 47.5 (42.1-53.0) 37.2 (32.1-42.6) 28.4 (23.8-33.6) 12.2 (9.0-16.2) 13.1 (9.9-17.3) 22.2 (18.0-27.1) 148 122 97 30 35 49 52.7 (46.8-58.4) 43.4 (37.7-49.3) 34.5 (29.2-40.3) 10.7 (7.6-14.8) 12.5 (9.1-16.8) 17.4 (13.4-22.3)

AHP = Academic Hospital Paramaribo; NICU = neonatal intensive care unit; HC = high care; MC = medium care; PIH = pregnancy-induced hypertension; RDS = respiratory distress syndrome.

1 _{Newborns could have more than one reason for admission.}

2 _{Includes: delivery rooms of four other hospitals in Paramaribo and one other hospital in Nickerie, birth clinics in rural and}

interior parts of Suriname, and home births.

3_{Administered in two doses of dexamethasone in the case of suspected premature birth before GA of 34 weeks.}

4 _{Includes: premature rupture of membranes (PROM), intrapartum fever and/or antibiotics, positive maternal Group-B}

streptococcus culture.

5 _{Includes: Caucasian, Brazilian, or mixed,}

6 _{Includes: neonatal respiratory distress syndrome, congenital pneumonia, pulmonary hemorrhage, pneumothorax,}

meconium aspiration syndrome, and transient neonatal tachypnea.

7 _{Includes: newborns defined with clinical symptoms of infection by admitting physician.}

8_{Includes: asphyxia defined by admitting physician (e.g., in the case of either need for resuscitation or Apgar <5 beyond}

5 minutes; lactate acidosis with base excess <16; coma or seizures after birth; findings with cerebral ultrasound such as edema).

9 _{Includes: diaphragmatic hernia, congenital heart defects, gastro-intestinal anomalies and neurological malformations.} 10_{Includes: hypoglycemia, dysmaturity, jaundice, and social indications.}

(35)

2

Ta b le 2 . M o rt al it y o f n ew b o rn s t re ate d a t t he fa ci lit y b ef o re a nd a fte r t he tr an si ti o n Peri o d 1 ( N =3 20 ) (J ul y 20 14 -M ar ch 2 0 15 ) Pe ri o d 2 ( N= 28 1) (A p ri l 2 0 15 -D ec em b er 2 0 15 ) R el at iv e R is k (9 5% C I) P-va lu e N % N % O ve ra ll m o rt al it y To ta l a t A HP ( p er 10 0 0 li ve bi rt hs ) 1 To ta l a t fac ili ty To ta l e ar ly n eo na ta l m o rt al it y To ta l l at e ne on at al m o rt al it y In b o rn O ut b o rn N ew b o rn s w it h N IC U le ve l c ar e 23 .4 55 /3 20 29 /3 20 26 /3 20 42 /2 84 13 /3 6 52 /1 59 17 .2 9. 1 8. 1 14 .8 36 .1 32 .7 13 .2 30 /2 81 18 /2 81 12 /2 81 20 /2 17 10 /6 4 29 /1 72 10 .7 6. 4 4. 3 9. 2 15 .6 16 .9 0 .5 6 (0 .36 -0 .90 ) 0 .6 2 (0 .4 1-0 .9 4) 0 .7 0 ( 0 .4 0 -1 .2 4) 0 .5 3 (0 .27 -1 .0 2) 0 .6 2 (0 .3 8-1. 0 3) 0 .4 3 (0 .2 1-0 .8 9) 0 .5 2 (0 .35 -0 .7 7) 0 .0 2 0 .0 2 0 .2 3 0 .0 6 0 .0 7 0 .0 2 <0 .0 1 G es ta tio na l a ge -s p ec ifi c m o rt al it y <2 8 w eek s 28 -3 2 w eek s 33 -3 6 w ee ks ≥3 7 w eek s M is si ng 6/ 16 12 /4 8 14 /1 14 20 /1 32 3 37 .5 25 .0 12 .3 15 .2 8/ 13 5/ 47 4/ 10 0 8/ 11 0 5 61 .5 10 .6 4. 0 7. 3 1. 64 ( 0 .7 6-3. 53 ) 0 .4 3 (0 .16 -1 .11 ) 0 .3 3 (0 .11 -0 .9 6) 0 .4 8 (0 .2 2-1. 0 5) 0 .2 0 0 .0 8 0 .0 4 0 .0 7 Bi rt h w ei gh t-sp ec ifi c m o rt al it y <1 0 0 0 g ≥1 0 0 0 -14 99 g ≥1 50 0 g _Missi ng 10 /2 6 13 /4 8 30 /2 42 2 38 .5 27 .1 12 .4 8/ 23 6/ 33 16 /2 21 0 34 .8 18 .2 7. 2 0 .9 0 ( 0 .4 3-1.9 0 ) 0 .6 7 (0 .2 8-1. 59 ) 0 .5 8 (0 .3 3-1. 0 4) 0 .7 9 0 .3 6 0 .0 7

(36)

2

Ta b le 2 . ( con ti ue d ) Pe ri o d 1 ( N =3 20 ) (J ul y 20 14 -M ar ch 2 0 15 ) Pe ri o d 2 ( N= 28 1) (A p ri l 2 0 15 -D ec em b er 2 0 15 ) R el at iv e R is k (9 5% C I) P-va lu e C au se -s p ec ifi c m o rt al it y Se ps is 2 Ea rl y-o ns et s ep si s La te -o ns et s ep si s Pe ri na ta l a sp hy xi a Pr em at ur it y co m pl ic at io n 3 C o ng en it al m al fo rm at io ns 4 O th er 5 19 /9 6 10 /4 4 9/ 52 12 /3 8 7/ 15 7 12 /4 2 5 19 .8 22 .7 17 .3 31 .6 4. 5 28 .6 8/ 10 9 3/ 59 5/ 50 2/ 30 5/ 14 8 9/ 35 6 7. 3 _5.1 ₁₀.0 _6.7 3. 4 ₂₅.7 0 .3 7 (0 .17 -0 .8 1) 0 .2 2 (0 .0 7-0 .7 7) 0 .5 8 (0 .2 1-1. 61 ) 0 .2 1 ( 0 .5 1-0 .8 7) 0 .7 6 (0 .2 5-2. 34 ) 0 .9 0 ( 0 .4 3-1. 88 ) 0 .0 1 0 .0 2 0 .2 9 0 .0 3 0 .6 3 0 .7 8 A H P = A ca d em ic H o sp it al P ar am ar ib o ; N IC U = n eo na ta l i nt en si ve c ar e un it ; 1 In cl ud in g de at hs a t th e d el iv er y ro o m ( 13 b efo re a nd 6 a ft er t he t ra ns it io n) . 2 In cl ud es: n ew b o rn s w it h cl in ic al s us pi ci o n, t re at ed w it h an ti bi o ti cs fo r 7 d ay s, r ai se d c -r ea ct iv e pr o te in le ve ls , a nd p o si ti ve b lo o d c ul tu re . 3 In cl ud es: r es pi ra to ry in su ffic ie nc y o r pn eu m o th o ra x w it h RD S an d e xt re m e pr em at ur it y, n ec ro ti zi ng e nt er o co lit is ; i nt ra ve nt ri cu la r he m o rr ha ge . 4In cl ud es: d ia ph ra gm at ic h er ni a, c o ng en it al h ea rt d ef ec ts , g as tr o -i nt es ti na l a no m al ie s an d n eu ro lo gi ca l m al fo rm at io n. 5 In cl ud es: p er si st en t pu lm o na ry h yp er te ns io n o f t he n eo nat e (P PH N ), p ne um o th o ra x, c ar d ia c ta m p o na d e, a nd k er ni ct er us .

(37)

2

Table 3a. Trends in treatments at the facility in two time periods Period 1 (N=320) (July 2014-March 2015) Period 2 (N=281) (April 2015-December 2015) Relative Risk (95% CI) P-value N % N % Highest level of care1 NICU HC MC 159 75 86 49.7 23.4 26.9 172 60 49 61.2 21.4 17.4 1.23 (1.07-1.42) 0.91 (0.68-1.23) 0.65 (0.47-0.87) <0.01 0.54 <0.01 Respiratory treatment CPAP Mechanical ventilation Surfactant 100 38 15 31.3 11.9 4.7 106 55 21 37.7 19.6 7.5 1.21 (0.97-1.51) 1.65 (1.13-2.41) 1.59 (0.84-3.03) 0.10 0.01 0.16 Antibiotics received Total 173 54.1 170 60.5 1.12 (0.97-1.29) 0.11

NICU = neonatal intensive care unit; HC = high Care; MC = medium Care; CPAP = continuous positive airway pressure.

1 _{Determined with local criteria given in Supplemental Table 1.}

Table 3b. Morbidity of newborns treated at the facility in two time periods Period 1 (N=320) (July 2014- March 2015) Period 2 (N=281) (April 2015-December 2015) Relative Risk (95% CI) P-value N % N % Respiratory morbidity BPD VAP Pneumothorax 4 9 4 1.3 2.8 1.3 10 5 7 3.6 1.8 2.5 2.85 (0.90-8.98) 0.63 (0.21-1.87) 1.99 (0.59-6.74) 0.07 0.41 0.27 NEC1 _Total ≥Stage 2 10 4 13.5 5.4 12 7 21.4 12.5 1.59 (0.74-3.40) 2.31 (0.71-7.51) 0.24 0.16 Sepsis2 _Total Positive blood culture 96 38 30.0 11.9 109 25 38.8 8.9 1.29 (1.03-1.62) 0.75 (0.46-1.20) 0.02 0.24 Positive ESBL culture3 Total 34 10.6 39 13.9 1.31 (0.85-2.01) 0.22 Duration of stay (days) Mean 13 SD 16 Mean 14 SD 18 0.44

BPD = bronchopulmonary dysplasia; VAP = ventilator-associated pneumonia; NEC = necrotizing enterocolitis; ESBL = extended spectrum beta-lactamase.

1 _{Calculated for newborns with a birthweight below 1500 grams (N=74 and N=56 in period 1 and period 2, respectively).} 2 _{Includes: early and late-onset clinical (i.e., high clinical suspicion, treated with antibiotics for 7 days; raised C-reactive}

protein levels) and blood culture positive sepsis.

3 _{Includes: blood and urine cultures and cultures on (tracheal aspirate, skin and anal) swabs, central lines or}

(38)

2 DISCUSSION

Improvements at the neonatal care facility led to an increase of newborns that received intensive care with a significant reduction in their mortality. Furthermore, newborns with a GA above 28 weeks and/or BW≥1500 grams showed a significantly reduced mortality rate. A striking reduction in mortality was seen in cases of perinatal asphyxia and sepsis. In addition, after the transition a two-fold increase in admission of outborn newborns, with similar demographics and increased survival rates, was observed. These findings indicate enhanced tertiary function and centralization of neonatal intensive care in Suriname, which may play a significant role in reducing neonatal mortality in Suriname.

Other studies performed in developing countries have shown similar patterns in improvement of mortality after scaling up of neonatal care facilities. Creation of a level II sick newborn care unit (SNCU) (i.e., with introduction of bed warmers and central oxygen) in a district hospital in India led to a significant reduction of regional NMR of mostly newborns with a BW<1500 grams [6]. Another pre-and-post intervention study in India showed that basic interventions (i.e., promotion of enteral nutrition, asepsis regulations and training of nurses) led to an immediate and stable reduction of NMR and birth-weight specific survival of newborns with a BW<1500 grams, but not with a BW<1000 grams, primarily after reduced incidence and mortality of sepsis [7]. Introduction of nasal CPAP at a NICU in Nicaragua reduced mortality amongst total newborns receiving ventilation assistance (i.e. either mechanical ventilation or CPAP) [8]. Improvement (i.e., new equipment, refurbishment and training of personnel) of a newborn unit to a Level III NICU at a teaching hospital in Ghana led to significant reduction of mortality amongst newborns with a BW<2500 grams, mostly secondary to significantly reduced incidence of perinatal asphyxia [9].

In these studies, training and expansion of personnel was a universal denominator for improvement of care, which was also part of our intervention. Systematic training of midwives in neonatal resuscitation has been a challenge in low resource countries and so far has yielded positive results only in low risk settings, and takes time with need for strong re-enforcement and repetition before an effect on neonatal mortality is observed [10-12]. However, increasing the number of nurses per infant at the NICU may have a beneficial effect on neonatal outcome [13,14]. Further improvement of survival may then be accomplished with increased capacity for neonatal intensive care (e.g., increased capacity for (modernized) ventilation). We observed a significant increase of use of neonatal intensive care commodities in the post-transition period. Indeed, both higher level and volume of neonatal intensive care have been associated with better survival of newborns with a BW<1500 grams [15,16]. While this seems an intuitive and logical effect, it is important to realize that positive effects of higher capacity can only be sustained with continuous and balanced availability of trained personnel, which can be challenging in the lower resource setting [17,18]. Illustratively, in our population the reduction of admission rates in the post transition period coinciding with increased number of nurses per bed may have been beneficial for survival. However, the amount of nurses per infant at our facility is still less than recommended for the intended level of care (i.e., one nurse per one or two beds), which may partially explain our finding that the mortality rate in the most vulnerable small preterm infants (i.e., with a BW<1000

University of Groningen Early onset sepsis in Suriname Zonneveld, Rens