SKRIFTLIG RAPPORT Läkarprogrammet, självständigt arbete (30 hp)
Factors associated with an increased risk of death in sepsis at a neonatal intensive
care unit in Kathmandu, Nepal
Student: David Sjöstedt
Supervisor: Mats Målqvist
Uppsala University
Department of Women’s and Children’s Health International Maternal and Child Health (IMCH) Degree Project, 30c
Date: January 10, 2018
Factors associated with an increased risk of death in sepsis at a neonatal intensive care unit in Nepal.
David Sjöstedt
Populärvetenskaplig sammanfattning ... 4
Abstract ... 5
Background ... 6
Millennium Development Goal 4 ... 6
Neonatal death world wide ... 6
Infections and antibiotics... 7
Risk factors of neonatal sepsis... 7
Sustainable Development Goals ... 7
Nepal and Nepal’s progress towards Millennium Development Goal 4 ... 8
Study objective ... 8
Hypotheses ... 9
Methods ...10
Data and study design ... 10
Variables ... 10
Statistical methods ... 11
Ethical considerations ... 11
Results ...12
Neonatal death and death in sepsis ... 12
Background data describing the population at the hospital ... 12
Results of logistic regressions... 16
Discussion ...18
Findings ... 18
External validity ... 18
Internal validity... 19
Conclusion:... 20
References ...21
Appendixes ...24
Appendix a - Variables and their definitions ... 24
Appendix b - Information about the deliveries at the hospital: ... 25
4
Populärvetenskaplig sammanfattning
Det fjärde milleniemålet var att minska antalet barn som dör innan sin femte födelsedag med två tredjedelar mellan 1990 och 2015. Underfem-dödligheten minskade dock endast med ca 50%. Det som i efterhand identifierats som orsaken till att målet inte kunde nås är att barnadödligheten under första månaden i livet inte minskade i samma takt som den resterande underfem-dödligheten.
De tre vanligaste dödsorsakerna under den första månaden i livet är för tidig födsel, förlossningskomplikationer och blodförgiftning. I resurssvaga länder är blodförgiftning
svårdiagnostiserad hos nyfödda. För det första så skiljer sig inte symptomen mycket från andra allvarliga sjukdomar och dessutom är tillgängligheten på labprover och bakterieodlingar, vilka ofta behövs för att säkerställa diagnosen, i många fall låg. För att underlätta diagnostiseringen av blodförgiftning så har forskare under senare år börjat fokusera på riskfaktorer för död i blodförgiftning för att på så sätt kunna identifiera riskpatienter.
Den här studien gjordes på data insamlad under ett års tid från ett förlossningssjukhus i Nepal.
Fördelningen mellan de olika dödsorsakerna under första levnadsmånaden kartlades. Dessutom jämfördes de barn som dog i blodförgiftning under första månaden med en referensgrupp för att identifiera riskfaktorer för död i blodförgiftning.
I studien var de tre vanligaste dödsorsakerna förlossningskomplikationer (40%), för tidig födsel (27%) och blodförgiftning (18%). Utebliven mödravård, kejsarsnitt, låg födelsevikt och för tidig födsel kunde kopplas till en riskökning för död i blodförgiftning. Att dessa fyra faktorer skulle vara kopplade till en ökad risk finns det stöd för i tidigare studier och därför ses dessa resultat som tillförlitliga. Studien kan inte förklara varför fördelningen av dödsorsaker ser ut som den gör och att undersöka överrepresentationen av förlossningskomplikationer skulle vara ett intressant område för fortsatt forskning.
5
Abstract
The neonatal mortality is a growing proportion of the total under-five mortality. The objective of this study was to map the causes of neonatal death and to identify risk factors for neonatal death in sepsis, which is one of the most common causes of neonatal death at a Maternity hospital in Nepal.
Data was collected between July 2012 and September 2013 and resulted in a case-referent cohort study. The distribution of the causes of death was mapped. Cases of neonatal death in sepsis were compared to a referent group to identify risk factors. Fifteen variables were analysed separately and all variables that showed a statistical significant association with sepsis death were analysed in one multivariate logistic regression model.
The most common causes of death were intrapartum complications (40.1%), pre-term birth (27.4%) and Sepsis/pneumonia (17.7%). Low birthweight (OR:7.9 (3.3-19)), prematurity (OR:5.4 (1.3-21)), caesarean section (OR:3.0 (1.4-6.4)) and no antenatal care (OR:2.4 (1.0-5.2)) were identified as risk factors for sepsis death.
The findings on risk factors for sepsis align with previous studies and the findings are interpreted as credible. The overrepresentation of intrapartum complications could not be explained within the frames of this study and calls for further research.
6
Background
Millennium Development Goal 4
Millennium Development Goal 4, MDG 4, was to reduce the under-five mortality by two-thirds between 1990 and 2015. The target was not reached but substantial progress was made. The global under-five mortality declined from 90 to 43 deaths per 1000 live births, a reduction by more than 50%. (1) (2)
The neonatal mortality, death during the first 28 days of life, during the same period of time has been reduced from 33 to 19 per 1000 live births. (1) This makes the neonatal mortality a growing part of the total under-five mortality and the main reasons that the MDG4 was not reached. If the neonatal mortality would have had the same reduction rate as the mortality of children 1-59 months the goal would have been reached ahead of time. (3)
Neonatal death world wide
The neonatal deaths are not equally distributed in the world. It is shown that 99% of neonatal deaths occur in middle- and low-income countries while the richest countries only account for 1%. (4) This shows that the large majority of neonatal deaths can be avoided with the right resources used in the right way. Not only is the number of deaths unevenly spread over the world, the causes of death also vary. Globally 36% of neonatal deaths are caused by complications due to preterm birth, 23%
due to intrapartum-related conditions, 23% due to infections and 10% due to congenital
malformations (Figure 1). (5)(6) Intrapartum-related conditions and infections normally account for a higher proportion of deaths in high mortality settings than in low-mortality settings. (6)
Figure 1: Cause of neonatal death (6)
7
Infections and antibiotics
Infections are one of the three most common causes of neonatal death. The symptoms of neonatal sepsis and other infections are not specific and the clinical presentation does not differ much from other common diseases during the neonatal period. The problems with a lack of high specificity and sensitivity clinical diagnosis combined with limited access to laboratory tests in many developing countries results in that correctly diagnosing neonatal sepsis remains a challenge. (7) (8) (9) A consequence of the difficulties of correctly diagnosing infections is that antibiotics is the most frequently used drug in neonatal intensive care units. (10) In fact some studies show that 89% of all neonates at neonatal intensive care units receive antibiotics some time during their stay. (11)
Risk factors of neonatal sepsis
Instead of only focusing on the clinical presentation of sepsis, research is starting to focus on identifying neonates of greater risk of sepsis and sepsis death to provide care with higher sensitivity and specificity to the neonates. Studies have identified low maternal age, education less than 9 years, primigravida or parity more than 6, multiple pregnancy and low attendance to antenatal care as maternal factors associated with an increased risk of neonatal death in general. Factors during delivery which are associated with an increased risk of neonatal death are non-cephalic
presentation, caesarean section, prolonged labour and prelabour rupture of membranes, PROM.
Characteristics of the child that further increases the risk of neonatal death are prematurity, low birthweight and male sex. (4) (12) (13) Factors that are found to be associated with an increased risk of sepsis and death in sepsis are primigravida, PROM, meconium stained amniotic fluid, caesarean section, low birthweight, prematurity and low Apgar-scores .(14)(15)(16)
Sustainable Development Goals
The era of the Millennial Development Goals (MDG), 1990 to 2015, has come to an end and the Sustainable Development Goals (SDG) have taken their place. The SDG consists of 17 different goals with a total of 169 targets and sets the agenda of Sustainable Development from 2015 to 2030. The SDGs are not only built on the MDGs but they aim to finish what the MDGs were not able to do. The SDG3, “Ensure healthy lives and promote well-being for all at all ages”, includes an aim of reducing under-five mortality and neonatal mortality to at least 25 and 12 deaths per 1000 live births respectively. This part of SDG3 can be seen as an extension and a sharpening of the MDG4, which the world was not able to reach. (17) (18) Projections of under-five mortality shows that in 2030, given that the annual reduction rate of under-five mortality in every country continues as between 2000 and 2015, will be 26,2 deaths per 1000 live birth. (19) Meaning that even if the reduction rates stay the same the world will still fall short of the SDG3.
8
Nepal and Nepal’s progress towards Millennium Development Goal 4
When summarising Nepal’s work with the MDG4 it has been shown that the under-five mortality has been reduced from 141 to 36 per 1000 live births between 1990 and 2015. This is a reduction of just over two-thirds making Nepal successful in reaching their MDG4 target. As in the rest of the world the neonatal proportion of the under-five mortality is growing in Nepal and in 2015 the neonatal mortality, 22 per 1000 live births, accounted for 61% of the total under-five mortality. (2)
Nepal is located between India and China. More than 90% of the population in Nepal works with agriculture and with 82% of the population living in rural areas Nepal is one of the least urbanized countries in the world. (20) In April and May 2015 two devastating earthquakes occurred in Nepal, killing over 9000 people and destroyed more than half a million private houses. The rebuilding process has been started but there is still a long way to go. (21) In Nepal 15% of all women are illiterate, 18% have one to five years of education and 67% has an education level of 6 years or more.(24) The people in Nepal are divided in to different ethnic groups by their caste and large inequalities exists between these groups. A simple way to categorise the groups in hierarchic order is “Upper caste” made up by the Brahman and Chhetri groups (31.7%), the Janajatis made up by the non-Hindu groups (35.7%), the non Daliti groups (14.8%), the Dalitis made up by the lower ranked Hindu-groups (12.4%) and the Muslim population (5%). (22) (23)
There are large differences in under-five mortality between socioeconomic groups and geographic areas in Nepal. In 2011 the under-five mortality in the wealthiest quintile of the population was less than half of the under-five mortality in the least wealthy quintile, 36 compared to 75 deaths per 1000 live births. The under-five mortality in the mountains was 87 compared to 58 deaths per 1000 live birth in the hills. (25) This intra-national inequality and the growing proportion of neonatal mortality of the under-five mortality are pointed out as areas in need of improvement if Nepal is going to further improve its under-five mortality and reach the SDG3. (21)
Study objective
The objective of this study is to map the causes of death in a neonatal intensive care unit in Nepal and to investigate associations between independent variables, such as gestational age, birth weight and APGAR-score, and neonatal death in sepsis.
9
Hypotheses
Maternal factors thought to be associated with an increased risk of death in sepsis are maternal age under 20, education level less than 9 years, belonging to a lower ethnic group, not receiving antenatal care, primigravida and multiple pregnancy.
Factors during delivery that are thought to be associated with a higher risk of death in sepsis are prolonged labour, PROM or PPROM, fetal distress, preterm labour and caesarean section.
Factors with the child that are thought to be associated with a greater risk of sepsis are prematurity, low birthweight, low Apgar-scores at one and five minutes and male sex.
10
Methods
Data and study design
The data used is from a previous dataset from the implementation of the “Helping Babies Breath”
(HBB) protocol in Nepal. The data was collected between July 2012 and September 2013 at
Paropakar Maternity and Women´s Hospital, Katmandu, and resulted in a case-referent prospective cohort study. During 2011 and 2012 the hospital had an early neonatal mortality rate at 9 death per 1000 live births. (26) In the study 20% of all deliveries at the hospital were randomly selected as referents and all stillbirths and neonatal deaths were included as cases. The data was collected by surveillance officers from patient records, clinical observations and interviews. The referent population was followed from admission and the cases of stillbirth and neonatal death were traced back using patient records. Referents who turned out as either stillbirths or neonatal deaths were later moved to the case-group. Observations were made both during a baseline period and during a period of implementation of the HBB guidelines. The dataset contained 394 variables and 5155 observations. More details about the data collection can be found in Wrammerts thesis “Surviving birth : Studies of a simplified neonatal resuscitation protocol in a low-income context using a mixed-methods approach”. (27)
This study was divided into two parts, one descriptive and one analytical. Observations during the baseline and the implementation period were included and all stillbirths were excluded.
The descriptive part of the study used all cases of neonatal deaths and the original referent group as study populations, aiming to map the causes of death and to present background data on the patients at the hospital.
The analytical part of the study was a case-referent study and used all cases of neonatal death in sepsis as cases. Cases of sepsis death in the referent group were moved to the case group. Resulting in a case population of n=53 and referent population of n=4472.
Variables
The classification of cause of death in the data was made by neonatologists at the neonatal intensive care unit and was based on clinical presentation and basic lab tests. Parity was divided in two categories, primipara or multipara, since primipara is a known risk factor for neonatal death and death in sepsis in other studies. (4) (14) Maternal age was divided in to two categories, 20 years or older or under 20 years old, because maternal age<20 or <18 is a known risk factor of neonatal death and death in sepsis in other studies. (4) (12) Previous studies have shown that maternal
11 education level less than 9 years is a risk factor, which is why education was divided in to two categories, education more or equal to 9 years or education less than 9 years. (12) Complete list of variables and how they are defined in the study can be found in Appendix A.
The outcome variable in the analytical part of the study was sepsis death.
Statistical methods
The first part of the study used descriptive statistics to map the characteristics of the patients at the hospital.
To find statistically significant associations between the independent variables and the odds ratio of death in sepsis in the analytical part of the study logistic regressions were performed. First each independent variable was analysed separately in bivariate Logistic Regressions to find significant associations with death in sepsis. Variables with less than n=8 in one of the groups (referents or cases) where analysed with Fisher’s exact test instead. Second all the variables with statistically significant associations to death in sepsis were analysed in one multivariate logistic regression.
Statistical significance was set to p<0,2 in the bivariate analyses and to p<0,05 in the multivariate analysis.
All calculations, graphs and statistics were made in RStudio 1.0.153.
Ethical considerations
Since this study used an old dataset with no way to identify the patients in the study no new ethics approval was necessary. Ethics approval for the original study and data collection was received from Uppsala University’s Ethical Review Board, DNR 2012/267, and from Nepal Health Research Council, Reg. No. 37/2012. (27)
12
Results
Neonatal death and death in sepsis
During the period of data collection there were 25108 deliveries at Paropakar Maternity and Women´s Hospital. The number of stillbirths was 443, resulting in 24665 live births. The total number of neonatal deaths was 299 making the neonatal mortality rate 12.1 deaths per 1000 live births. The causes of neonatal death are shown in figure 2. Intrapartum complications, pre-term birth and Sepsis/Pneumonia were the three big causes of death followed by others and congenital anomalies while there were no deaths caused by neonatal tetanus or diarrhoea. The neonatal mortality rate caused by sepsis was 2.1 deaths per 1000 live births.
Figure 2 Causes of neonatal death
Background data describing the population at the hospital
Data was collected from 20% of all deliveries at Paropakar Maternity and Women´s Hospital during the study. Information from these deliveries is presented in the following pages. More details about this population can be found in Appendix B.
The distribution of APGAR-score at five minutes after birth is shown in figure 3. The mean APGAR-score was 7,72 ± 0.92.
13
Figure 3 Distribution of APGAR-score at five minutes after birth
The distribution of birthweight is shown in figure 4. The mean birthweight was 2919 ± 493.59grams.
Figure 4 Distribution of birthweight.
The distribution of gestational age is shown in figure 5. The mean gestational age was 39.05 ± 2.06 weeks.
14
Figure 5 Distribution of gestational age.
The proportions of the ethnic groups are shown in figure 6. The groups in order of largest proportion were “Upper caste”, “Janajatie groups without advantages”, “Janajatie groups with advantages”, “Non-Daliti groups” and “Muslims”.
Figure 6 Proportions of Ethnic groups
The maternal education levels are shown in figure 7. The most common length of education was 9- 10 years followed by 10+2 years or more, 6-8 years, illiterate, 1-5 years and lastly the group that had School leaving certificates, meaning that they completed their tenth year of study but had no further education.
15
Figure 7 Maternal education level
The distribution of maternal age is shown in figure 8. The mean maternal age was 23,66 ± 4,3778 years.
Figure 8 Maternal age
16
Results of logistic regressions
Figure 9 shows the study flow chart for the analytic part of the study.
Figure 9 Study flow chart for the analytic part of the study
When the independent variables were analysed in separate bivariate logistic regressions thirteen of them showed a statistical significant association with death in sepsis, p<0.2. These were “Multiple pregnancy”, “Mothers first birth”, Maternal education less than 9 years” “Ethnic group- non-daliti groups” “No antenatal care”, “Preterm labour”, “Prom or PPROM”, “Caesarean section”, “Apgar 1 min”, “Apgar 5 min”, “Low birthweight”, “Prematurity” and “implementation observation”. All variables and their respective p-values are shown in table 1. Since the variable that showed if the observation was made during the baseline or the implementation period was statistically significant it was added in the multivariate analyse to adjust for when the observation was made.
Table 1 P-values for each independent variable when analysed by univariate logistic regressions
Variable Prevalence: P-value:
Case n=53 (%)
Referent n=4472 (%)
Multiple pregnancy 2 (3.8) 38 (0,85) 0.08
Mothers first delivery 23 (43) 2417 (54) 0.13
17
Maternal education less than 9 years 24 (45) 2147 (48) 0.05
Mothers age less than 20 11 (21) 724 (16) 0.37
Ethnic group Janajatie groups with advantages
11 (21) 812 (18) 0.33
Janajatie groups without
advantages
13 (25) 1292 (29) 0.53
Non-Daliti groups
8 (15) 368 (8.2) 0.05
Dalitis 2 (3.8) 235 (5,3) 0.91
Muslims 1 (1.9) 34 (0.76) 0.27
No antenatal care 23 (43) 572 (13) <0.01
Preterm labour 22 (42) 342 (7.6) <0.01
Prom or PProm 8 (15) 178 (4.0) <0.01
Fetal distress 7 (13) 391 (8.7) 0.23
Caesarean section 27 (51) 1011 (23) <0.01
APGAR 1min - - <0.01
APGAR 5min - - <0.01
Low birthweight 34 (64) 511 (11) <0.01
Prematurity 26 (49) 362 (8.1) <0.01
Sex (female=0 male=1) 27 (51) 2371 (53) 0.76
implementation observation 37 (70) 2591 (58) 0.08
The multivariate logistic regression of all the statistically significant variables showed that only four of the thirteen variables had a significant association with death in sepsis. Sorted by largest odds ratio these were “low birthweight”, “prematurity”, “Caesarean section” and “No antenatal care”.
The analysed variables, their odds ratios with 95% confidence intervals and p-values are shown in table 2.
Table 2 Independents variables effect on the risk of sepsis death Variable: Adjusted odds ratio (95%
Confidence interval)
P-value
Multiple pregnancy 1.0 (0.14 - 4.7) 1.00
Mothers first delivery 0.89 (0.42 - 1.9) 0.77
Education less than 9 years 1.5 (0.7-3.5) 0.29
Ethnic group Non-dalitis groups
2.5 (0.78-6.9) 0.089
No antenatal care 2.4 (1.0-5.2) 0.039
Preterm labour 1.1 (0.3- 4.7) 0.86
Prom or PProm 1.67 (0.5- 4.4) 0.34
Caesarean Section 3.0 (1.4-6.4) <0.01
APGAR 1min 0.80 (0.45-1.4) 0.45
APGAR 5min 1.2 (0.7-20) 0.56
Low birthweight 7.9 (3.3-19) <0.01
Prematurity 5.4 (1.3-21) 0.019
Implementation observation 5.9e+7 (8.6e-18 - 1.9e+117) 0.99
18
Discussion
Findings
The aim with this study was to map the causes of death at a neonatal intensive care unit in Nepal and to identify factors that had an association with increased risk of sepsis death. The causes of death were intrapartum complications (40.1%), pre-term birth (27.4%), Sepsis/pneumonia (17.7%), Others (9.7%) and congenital anomalies (5%). Out of fifteen tested independent variables four showed a significant association with the risk of sepsis death in the multivariate logistic regression.
The four significant variables were low birthweight 7.9 (3.3-19), prematurity 5.4 (1.3-21),
caesarean section 3.0 (1.4-6.4) and no antenatal care 2.4 (1.0-5.2). Notable is that the lower bound of the odds ratios confidence interval for no antenatal care was 1.0 and that this result therefore must be interpreted with caution.
External validity
When comparing the proportion of causes of death in the world with the causes of death in this study some differences are clear. Firstly intrapartum complications is over represented in this study.
Secondly preterm birth is less common in this study. Infections, such as sepsis/pneumonia, diarrhoea and tetanus, and congenital anomalies also account for smaller proportions of the total number of neonatal death. (5)(6) An explanation for this could be that the proportion of intra- partum complications in high mortality settings often is higher than in low mortality settings.
However previous studies also show that infections often account for a larger proportion of deaths in high mortality settings than in low mortality settings and that preterm birth in both high and low mortality settings is the most common cause of death, ratios that cannot be seen in this study. (6)
The neonatal mortality rate found in this study 12.1 deaths per 1000 live births is substantially lower than 22-23 deaths per 1000 live births reported in previous studies and reports of Nepal’s neonatal mortality rate. (2)(21) The big difference between the study data and the national data could partly be explained by the large intra-national differences in Nepal between the urban and rural parts of the population and between different ecological zones. (25) Another factor that can contribute to the low number of neonatal deaths in the study is that the study only registered neonatal deaths at the hospital and that deaths at home therefore were missed.
The findings that low birthweight, prematurity and caesarean section all are associated with an increased risk of death in sepsis can all be strengthened by previous studies.(14)(15)(16) Lack of
19 antenatal care also showed an association with increased risk of sepsis death. Its relationship with sepsis death is not well studied but has previously shown an association with increased risk of neonatal death witch strengthens the finding. (12) (13) If there is a true association or if the
relationship is established through confounding factors, for example socioeconomic factors, cannot be shown in this study.
When comparing the background data in the study to population data from Nepal, shown in brackets, differences can be seen. Upper caste 38.7% (31.7%), Janajatie groups 47.0% (35.7%), Non-daliti groups 8.2% (14.8%), Dalitis 5.3% (12.4%) and Muslims 0.8% (5%). The ethnic groups of higher hierarchy were over represented in the study. Since no association could be shown
between different ethnic groups and the risk of death in sepsis one could argue that this over
representation should not be a problem. However, reports show that socioeconomic inequalities still exist between castes and that one factor that affects antenatal care is wealth. This results in that the overrepresentation of the ethnic groups could have influenced the neonatal mortality rate and the number of sepsis deaths in the study.(23) Education levels in the background data matches the population data, in brackets, well. Illiterate 16.7% (15%), 1-5 years of education 16.6% (18%) and more than 5 years 66.7% (67%). (24)
Internal validity
The study design, a case-referent study, does not enable causal relationships to be found and limits the study to findings of associations between variables. The strengths with the design is however that even a rare event as sepsis death, 2.1 sepsis deaths per 1000 live births in this setting, can be studied. Even though the study population for the analytic part contained 4525 observations there were still no mothers diagnosed with amnionitis and no deliveries with prolonged labour in the case group why these variables could not be analysed.
The fact that the causes of death were classified by neonatologists at the hospital without advanced lab tests and blood cultures accounts for a risk of misclassification, especially since the symptoms of many severe illnesses in neonates have similar clinical presentation.(7)(8)(9) Another limitation is that, as previously mentioned, only deaths at the hospital were noted as neonatal deaths in the study, making it possible that neonatal deaths that occurred after discharge were missed. Since sepsis death is most common in the late neonatal period the importance of this limitation increases.
(6)
The fact that the variable for caesarean section included both elective and acute caesarean sections resulted in that some information was lost. To add variables describing the use of antibiotics both during and after labour would further the study and the knowledge about risk factors for death in
20 sepsis.
A final limitation is that the data collected form patient records was dependent on that the variables and interventions measured were mentioned by the physicians in their documentation, resulting in that some data could have been lost due to lack of notations in patient records.
Conclusion:
Low birthweight, prematurity, caesarean section and lack of antenatal care were found as factors associated with an increased risk of sepsis death in neonates. Since these findings accord with studies in similar settings the findings are interpreted as credible. The causes of neonatal death were intrapartum complications (40.1%), pre-term birth (27.4%), Sepsis/pneumonia (17.7%), Others (9.7%) and congenital anomalies (5%). The ratio of death in intrapartum complications is higher than found in previous studies. Within the frames of this study the overrepresentation of intrapartum complications cannot be explained but it calls for further research. The reason for the large ratio of death by intrapartum complications must be found and interventions to improve the care of the affected neonates must be taken if Nepal is going to further reduce their neonatal mortality rate and reach the Sustainable Development Goal 3.
21
References
1. United Nations. The Millennium Development Goals Report 2015 [Internet]. New York: United Nations; 2015. Available from:
http://www.undp.org/content/undp/en/home/librarypage/mdg/the-millennium-development- goals-report-2015.html
2. United Nations Children’s Fund. Levels & Trends in Child Mortality [Internet]. New York:
United Nations Children’s Fund; 2015 Sep. Available from:
http://www.who.int/maternal_child_adolescent/documents/levels_trends_child_mortality_2015/
en/
3. Liu L, Oza S, Hogan D, Chu Y, Perin J, Zhu J, et al. Global, regional, and national causes of under-5 mortality in 2000–15: an updated systematic analysis with implications for the Sustainable Development Goals. The Lancet. 2016 Dec 17;388(10063):3027–35.
4. Lawn JE, Cousens S, Zupan J. 4 million neonatal deaths: When? Where? Why? The Lancet.
2005 Mar 5;365(9462):891–900.
5. Oza S, Lawn JE, Hogan DR, Mathers C, Cousens SN. Neonatal cause-of-death estimates for the early and late neonatal periods for 194 countries: 2000-2013. World Health Organ Bull World Health Organ Geneva. 2015 Jan;93(1):19–28.
6. Lawn JE, Blencowe H, Oza S, You D, Lee AC, Waiswa P, et al. Every Newborn: progress, priorities, and potential beyond survival. The Lancet. 2014 Jul 12;384(9938):189–205.
7. Clinical signs that predict severe illness in children under age 2 months: a multicentre study.
The Lancet. 2008 Jan 12;371(9607):135–42.
8. Opiyo N, English M. What clinical signs best identify severe illness in young infants aged 0–59 days in developing countries? A systematic review. Arch Dis Child. 2011 Jan 10;96(11):1052–
9.
9. Zea-Vera A, Ochoa TJ. Challenges in the diagnosis and management of neonatal sepsis. J Trop Pediatr. 2015 Feb;61(1):1–13.
10. Clark RH, Bloom BT, Spitzer AR, Gerstmann DR. Empiric Use of Ampicillin and Cefotaxime, Compared With Ampicillin and Gentamicin, for Neonates at Risk for Sepsis Is Associated With an Increased Risk of Neonatal Death. Pediatrics. 2006 Jan 1;117(1):67–74.
11. Hauge C, Stålsby Lundborg C, Mandaliya J, Marrone G, Sharma M. Up to 89% of neonates received antibiotics in cross-sectional Indian study including those with no infections and unclear diagnoses. Acta Paediatr. 2017 Oct 1;106(10):1674–83.
12. Vogel J, Souza J, Mori R, Morisaki N, Lumbiganon P, Laopaiboon M, et al. Maternal
complications and perinatal mortality: findings of the World Health Organization Multicountry Survey on Maternal and Newborn Health. BJOG Int J Obstet Gynaecol. 2014 Mar;121:76–88.
13. Dey SK, Afroze S, Islam T, Jahan I, Shabuj MKH, Begum S, et al. Death audit in the neonatal ICU of a tertiary care hospital in Bangladesh: a retrospective chart review. J Matern Fetal Neonatal Med. 2017 Oct 26;0(0):1–5.
22 14. Schrag SJ, Cutland CL, Zell ER, Kuwanda L, Buchmann EJ, Velaphi SC, et al. Risk Factors for
Neonatal Sepsis and Perinatal Death Among Infants Enrolled in the Prevention of Perinatal Sepsis Trial, Soweto, South Africa: Pediatr Infect Dis J. 2012 Aug;31(8):821–6.
15. Shah GS, Budhathoki S, Das BK, Mandal RN. Risk factors in early neonatal sepsis.
http://www.kumj.com.np [Internet]. 2006 Apr 8 [cited 2017 Oct 21]; Available from:
http://imsear.hellis.org/handle/123456789/46611
16. View of Neonatal Sepsis as a Major Cause of Morbidity in a Tertiary Center in Kathmandu [Internet]. [cited 2017 Nov 7]. Available from:
http://jnma.com.np/jnma/index.php/jnma/article/view/2424/1662
17. World Health Organization. Health in 2015: from MDGs, Millennium Development Goals to SDGs, Sustainable Development Goals [Internet]. Geneva Schweitz: World Health
Organization; 2015 Dec. Available from: http://www.who.int/gho/publications/mdgs-sdgs/en/
18. United Nations. Transforming our World: The 2030 Agenda for Sustainable Development [Internet]. New York: United Nations; 2015. Available from:
https://sustainabledevelopment.un.org/post2015/transformingourworld/publication
19. You D, Hug L, Ejdemyr S, Idele P, Hogan D, Mathers C, et al. Global, regional, and national levels and trends in under-5 mortality between 1990 and 2015, with scenario-based projections to 2030: a systematic analysis by the UN Inter-agency Group for Child Mortality Estimation.
The Lancet. 2015 Dec 5;386(10010):2275–86.
20. Nepal - Uppslagsverk - NE [Internet]. [cited 2017 Sep 15]. Available from:
http://www.ne.se.ezproxy.its.uu.se/uppslagsverk/encyklopedi/l%C3%A5ng/nepal
21. National Planning Commission, 2016. Nepal and the Millennium Development Goals: Final Status Report 2000-2015 [Internet]. Kathmandu, Nepal: Government of Nepal, National Planning Commission; 2016 Dec [cited 2017 Sep 13]. Available from:
http://www.np.undp.org/content/nepal/en/home/library/mdg/mdg-status-report-2016.html 22. Bennet, Lynn, Dilli Ram Dhal, Pav Govindasamy. Caste, Ethnic and Regional Identity in
Nepal: Further Analysis of the 2006 Nepal Demographic and Health Survey [Internet].
Calverton, Maryland, USA: Macro International Inc.; 2008. Available from:
https://dhsprogram.com/pubs/pdf/FA58/FA58.pdf
23. Ministry of Health, Nepal; New ERA; and ICF. Nepal Demographic and Health Survey 2016:
Key Indicators. Kathmandu, Nepal: Ministry of Health, Nepal; 2017.
24. Central Bureau of Statistics, 2015. Nepal Multiple Indicator Cluster Survey 2014, Key fidings.
Kathmandu, Nepal: Central Bureau of Statistics and UNICEF Nepal;
25. Ministry of Health and Population (MOHP) [Nepal], New ERA, and ICF International Inc.
Nepal Demographic and Health Survey 2011. Kathmandu, Nepal: Ministry of Health and Population (MOHP) [Nepal], New ERA, and ICF International Inc.; 2012.
26. Ministry of Health and Population GoN. Annual Report “Smarika” 2068/69. In. Kathmandu, Nepal: Paropakar Maternity and Women’s Hospital; 2011 2012.
27. Wrammert J. Surviving birth : Studies of a simplified neonatal resuscitation protocol in a low- income context using a mixed-methods approach. 2017 [cited 2017 Oct 23]; Available from:
http://uu.diva-portal.org/smash/record.jsf?pid=diva2:1079015
23 28. Kc A. Neonatal Resuscitation : Understanding challenges and identifying a strategy for
implementation in Nepal. 2016 [cited 2017 Nov 2]; Available from: http://uu.diva- portal.org/smash/record.jsf?pid=diva2:874833
24
Appendixes
Appendix a - Variables and their definitions
Neonatal mortality: Death of an infant between 0-27 day of birth.
Parity: Number of times a woman has given birth after the age of viability, i.e. 22 weeks, including both live and still births. This variable was used to create the variable mothers first delivery.
Ethnic group: The group within the social hierarchical system of Nepal to which the woman’s family belongs. Divided in to six categories in hierarchic order where upper cast were used as the referent group.
Maternal age less than 20: maternal age divided in to two categories, more or equal to 20 or less than 20 years old.
Maternal education less than 9 years: reported education of mother. Divided in to two categories, more or equal to 9 years or less than 9 years of education.
No Antenatal care: no attendance to any antenatal care visits, during which the mother received clinical examination, counselling and medication (if needed) from a health worker.
Multiple pregnancy: Woman pregnant with more than one fetus.
Prolonged labour: When the cervix is not dilated beyond 4cm after eight hours of regular contractions or if cervical dilatation was to the right of the alert line on the partogram.
Gestational age: Gestational age measurement based on the mother’s last menstrual period.
Birth weight: Weight of the baby measured within one hour of delivery using a weighing scale.
Prematurity: Babies born before 37 completed weeks of gestation, estimated by the date of the mother’s last menstrual period.
Low Birth Weight: Babies who weigh less than 2500 grams at the time of birth.(27) (28)
Preterm labour: Regular uterine contractions accompanied by progressive cervical dilation and/or effacement at less than 37 weeks.
Prom or PProm: Prelabour rupture of membranes or preterm prelabour rupture of membranes.
Fetal distress: Abnormal fetal heart rate, less than 100 or more than 180 beats per minute, or thick meconium-stained amniotic fluid.
Caesarean section: baby delivered by caesarean section.
APGAR 1 min and 5 min: the documented Apgar-score at one and five minutes after birth.
Sex: two categories were female act as referent group.
Implementation observation: Observation made during the period of implementation of the Helping babies breath protocol.
25
Appendix b - Information about the deliveries at the hospital:
Table 1 Statistics describing the deliveries during the data collection
Variable Number of deliveries (percent)
Number of children in current pregnancy
One 4438 (99)
Two 37 (0.83)
Three 1 (0.022)
Number of previous pregnancies
Zero 2418 (54)
One 1372 (30)
Two 497 (11)
Three or more 189 (4.0)
No antenatal care 572 (13)
Preterm labour 343 (7.8)
Prom or PProm 178 (4.0)
Prolonged labour 28 (0.6)
Amnionit 0 (0)
Fetal distress 393 (8.8)
Caesarean section 1012 (23)
Sex Female 2103 (47)
Male 2373 (53)
