Increased perinatal survival and improved ventilation skills over a five-year period: An observational study.

BACKGROUND AND AIM: The Helping Babies Breathe program gave major reductions in perinatal mortality in Tanzania from 2009 to 2012. We aimed to study whether this effect was sustained, and whether resuscitation skills changed with continued frequent training.
METHODS: We analysed prospective data covering all births (n = 19,571) at Haydom Lutheran Hospital in Tanzania from July 2013 -June 2018. Resuscitation training was continued during this period. All deliveries were monitored by an observer recording the timing of events and resuscitation interventions. Heart rate was recorded by dry-electrode ECG and bag-mask-ventilation by sensors attached to the resuscitator device. We analyzed changes over time in outcomes, use of resuscitation interventions and performance of resuscitation using binary regression models with the log-link function to obtain adjusted relative risks.
RESULTS: With introduction of user fees for deliveries since 2014, the number of deliveries decreased by 30% from start to the end of the five-year period. An increase in low heart rate at birth and need for bag-mask-ventilation indicate a gradual selection of more vulnerable newborns delivered in the hospital over time. Despite this selection, newborn deaths <24 hours did not change significantly and was maintained at an average of 8.8/1000 live births. The annual reductions in relative risk for perinatal death adjusted for vulnerability factors was 0.84 (95%CI 0.76-0.94). During the five-year period, longer duration of bag-mask ventilation sequences without interruption was observed. Delivered tidal volumes were increased and mask leak was decreased during ventilation. The time to initiation or total duration of ventilation did not change significantly.
CONCLUSION: The reduction in 24-hour newborn mortality after introduction of Helping Babies Breathe was maintained, and a further decrease over the five-year period was evident when analyses were adjusted for vulnerability of the newborns. Perinatal survival and performance of ventilation were significantly improved.

Introduction

The reduction of neonatal deaths over the last decade has been modest, despite declining mortality of infants between 1 month and five years of age [1]. The estimated global neonatal mortality (≤ 28 days) fell from 31/1000 in 2000 to 18/1000 in 2017 [2]. The annual rate reduction in neonatal mortality was 3.1% compared to 4.7% after the neonatal period. Consequently, the relative contribution of neonatal deaths in the global under-5-mortality has increased from 39% to 45% [1]. The majority of these neonatal deaths are preventable. Safer births are achievable by improvements in a continuum from pregnancy, obstetric to neonatal care, each accounting for 1/3 of preventable deaths [3].

Intrapartum-related events, formerly referred to as birth asphyxia and defined as inadequate onset of spontaneous breathing after birth, are one of the three most prevalent causes of under-5-mortality [1]. The Helping Babies Breathe (HBB) program aimed to reduce death from intrapartum events in low- and middle-income settings, through training of health care workers in early identification and treatment [4].

In Tanzania, the roll-out of HBB from 2010 was associated with a 47% reduction in fresh stillbirths and early neonatal mortality over 24 months in eight large birth facilities [5]. Haydom Lutheran Hospital was one of the eight sites participating in the early roll-out of HBB. HBB training and research related to safer deliveries and newborn resuscitations continued in the Safer Births program [6]. Frequent low-dose training is one of the key elements to improve resuscitation skills [6, 7].

In Tanzania neonatal mortality decreased from 33/1000 in 2000 to 21/1000 in 2017 [2]. However, vital birth statistics are not available [8, 9], and estimates are inaccurate as almost half of the deliveries in Tanzania take place outside birth facilities. In a rural area with a low proportion of deliveries in hospital, this selection may preclude interpretation of temporal trends in mortality rates. It is therefore pertinent to adjust for vulnerability background factors in hospital deliveries.

After introduction of HBB in 2009–2012, efforts to sustain the initial improvements in perinatal mortality continued with onsite low-dose high-frequency HBB training. In the present study, we aimed to document changes in 24-hour neonatal mortality, fresh stillbirths and resuscitation performance during a five-year period from 2013 to 2018.

Methods

Haydom Lutheran Hospital is situated in the rural area of the Manyara region of Northern Tanzania. The hospital has a catchment area of over 2 million inhabitants, and is the only hospital in the region providing comprehensive obstetric care with 250–400 deliveries per month. Haydom is a non-governmental hospital managed by the Lutheran church of Tanzania. All births in the hospital from 1st of July 2013 until 30th of June 2018 were recorded in the study.

This study was approved by the National Institute for Medical Research, Tanzania (NIMR/HQ/R8a/Vol. IX/1434) and the Regional Committee for Medical and Health Research Ethics, Western Norway (2013/110/REK). All health care providers gave informed oral consent. The mothers were informed but formal consent was not obtained for this observational study due to the urgency of mothers arriving in late stages of labor. This consent procedure was approved by the ethical committees.

The initial HBB implementation

The local procedure for newborn resuscitation followed the HBB guidelines as implemented locally from 2009–2012 [5], emphasizing stimulation and early initiation of bag-mask ventilation (BMV) [4]. Newborn resuscitation was mainly the responsibility of midwives. Six labour rooms were each equipped with a resuscitation table. After caesarean section, newborn resuscitation was performed in a room adjacent to the operating theatre. Midwives determined if the newborn remained with the mother or was transferred to a neonatal ward after resuscitation. The neonatal ward provided basic care: respiratory support was limited to provision of supplemental oxygen with nasal prongs.

The Safer Births program

The Safer Births program, starting in 2013, aimed to further facilitate HBB training and improve obstetric and newborn care by low-cost technology [6, 10]. During the study all midwives were encouraged to low-dose, high-frequency skills training as described by Mduma et al. [11]. Briefly, simulation training was performed on manikins, assisted by local-trainers who had undergone facilitator training for the HBB program. Visiting instructors from other places in Tanzania and Norway also participated regularly. The trainings were intended to take place weekly, but with some variations over time.

Trained data collectors observed and recorded the timing of events immediately after birth, and vital data and outcomes from all deliveries were recorded in a database.

Physiological data collected from 1st July 2013 until 30 Jun, 2018 included heart rate data recorded from ECG, and ventilation parameters obtained from sensors connected to the ventilation bag using a Newborn Resuscitation Monitor (Laerdal Global Health, Stavanger, Norway) (Fig 1). The arch-shaped ECG sensor was placed over the thorax or abdomen of the newborn. Two stainless-steel discs arranged on each side of the flexible arch acted as dry electrodes. The monitor used a hot-wire flow sensor connected to the ventilation bag similar to that in the Florian Respiratory Function Monitor (Acutronic Medical Systems, Hirzel, Switzerland). Volume was calculated by flow signal integration. Pressure was measured using an MPX5010 sensor (Freescale Semiconductor, Austin, TX, USA). The HR was displayed on the monitor installed in front of the resuscitation table visible to the provider, ventilation parameters were not displayed [12]. We analyzed data for a five years period with collection of physiological data until 30th of June 2018 (Fig 2, flow chart).

Fig 1

Neonatal resuscitation monitor and its components used for data collection at Haydom Hospital during 1.7.13 to 30.6.18.

Fig 2

Flow chart of participants in a population-based study of births in Haydom hospital during 1.7.2013 to 30.6.2018.

Main and secondary outcomes

The main outcome was early neonatal mortality (ENM) defined as death before 24 hours of life among newborns delivered alive at Haydom Lutheran Hospital. As shown previously, live born babies with severe asphyxia may be misclassified as fresh stillbirths [13, 14]. Therefore, we also used the combined outcome of ENM and fresh stillbirths to analyze perinatal mortality.

The secondary outcomes were clinical performance defined by timing of events during resuscitation: Time from birth to start of BMV, the continuation of BMV by measuring duration of the first ventilation sequence before pause (>5 seconds interruption) and the fraction of time used for BMV once it had been initiated. The total duration of BMV was calculated as the time from the first to the last inflation. To assess actual bag-mask ventilation performance, we studied ventilation frequency, tidal volumes and mask leak the first five minutes of bag-mask ventilation.

Main exposure and covariates

We analyzed by year (1st of July– 30th of June, starting 1st of July 2013) for the five-year period to assess effects of ongoing training in newborn resuscitation over time.

To assess for potential confounding, we obtained information from the prospective database of pregnancy complications, birth weight, prematurity, fetal heart rate abnormalities, the last recorded fetal heart rate and the need for BMV at birth and type of resuscitator used.

Statistical methods

Descriptive analyses were performed by calculation of medians with interquartile ranges (IQR) or means with standard deviations according to their distribution. Categorical variables were analyzed using percentages and Chi-square test.

We used binary regression with log-link function to estimate relative risks for ENM by time [13, 15]. Risk factors which changed over time and were associated with the outcome (p<0.10) were included in a multivariate regression model if they changed the relative risk estimates by >10%. In the final model we therefore adjusted for birth weight, prematurity, fetal heart rate abnormalities, last recorded fetal heart rate and the need for BMV at birth. Cases with missing variables for any of the included adjustment variables (n = 15) were excluded from the adjusted model.

We also performed sensitivity analyses excluding newborns who died without a history of resuscitation at birth because conditions other than intra-partum hypoxia may cause ENM.

Data were extracted using Matlab 2019a (MathWorks, Natick, MA) and analyzed using Stata SE 15 (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LP.)

Results

During the five-year period, 18,997 infants were born alive, with a reduction over time from 4,662 to 3,197 annual births (Table 1). ENM fluctuated around 8.8/1000 live births, and the combined outcome of perinatal mortality (fresh stillbirths and ENM) around 22.6/1000 with no significant differences over time (Fig 3).

Table 1

Baseline characteristics of live births at Haydom 1.7.2013–30.6.2018 (n = 18,997).

		Period of birth
	1st year	2nd year	3rd year	4th year	5th year	p-value*
	n = 4,662	n = 3,909	n = 3,565	n = 3,664	n = 3,197
Maternal characteristics
No antenatal care, n (%)						0.16
	44 (0.9)	48 (1.2)	36 (1.0)	24 (0.7)	32 (1.0)
Antenatal problem, n (%)						<0.001
	44 (0.9)	29 (0.7)	57 (1.6)	65 (1.8)	66 (2.1)
Preeclampsia/eclampsia, n (%)						0.07
	21 (0.5)	17 (0.4)	25 (0.7)	26 (0.7)	28 (0.9)
Delivery
Mode of delivery, n (%)						<0.001
Vaginal	3,650 (78)	2,961 (76)	2,715 (76)	2,770 (76)	2,409 (75)
Caesarean	924 (20)	943 (24)	824 (23)	858 (23)	749 (23)
Breech	77 (2)	0 (0)	18 (0.5)	19 (0.5)	26 (0.8)
Vacuum/others	11(0.2)	4 (0.1)	8 (0.2)	17 (0.5)	11 (0.3)
Fetal heart rate at admission, n (%)						<0.001
Normal	4,122 (88)	3,510 (90)	3,274 (92)	3,444 (94)	3,006 (94)
Abnormal	42 (1)	33 (1)	31 (1)	38 (1)	43 (1)
Not detectable	2 (0.04)	0 (0)	2 (0.1)	5 (0.1)	0 (0)
Not measured	496 (11)	365 (9)	257 (7)	177 (5)	145 (5)
Fetal heart rate during labour, n (%)						<0.001
Normal	3,917 (84)	3,249 (83)	2,973 (83)	3,157 (86)	2,751 (86)
Abnormal	143 (3)	168 (4)	168 (5)	230 (6)	298 (9)
Not detectable	0 (0)	0 (0)	7 (0.2)	4 (0.1)	1 (0.03)
Not measured	602 (13)	494 (13)	416 (12)	272 (7)	145 (5)
Last fetal heart rate before delivery (bpm), mean (SD)						0.14
	134 (10)	134 (10)	134 (12)	134 (14)	133 (14)
Child characteristics
Birth weight, Mean (SD)	3215 (528)	3253 (520)	3280 (520)	3377 (539)	3342 (546)	<0.001
<1500, n (%)	24 (0.5)	24 (0.5)	15 (0.4)	23 (0.6)	28 (0.9)
1500–2500, n (%)	325 (7)	218 (6)	199 (6)	158 (4)	136 (4)
2500–3499, n (%)	2,895 (62)	2,397 (61)	2,071 (58)	1,821 (50)	1,690 (53)
3500–4499, n (%)	1,370 (29)	1,228 (31)	1,242 (35)	1,594 (44)	1,284 (40)
≥4500, n (%)	48 (1)	42 (1)	38 (1)	68 (2)	57 (2)
Gestational age < 37 weeks, n (%)						0.042
	117 (2.5)	103 (2.6)	96 (2.7)	121 (3.3)	111 (3.5)
Female sex, n (%)						0.049
	2,141 (46)	1,882 (48)	1,672 (47)	1,692 (46)	1,546 (48)

bpm: beats per minute

* p-values calculated with chi-square for categorical and binary regression for continuous variables

1st year: 1.7.2013–30.6.2014

2nd year: 1.7.2014–30.6.2015

3rd year: 1.7.2015–30.6.2016

4th year: 1.7.2016–30.6.2017

5th year: 1.7.2017–30.6.2018

Fig 3

Early neonatal mortality and stillbirths at Haydom Hospital by year during 1.7.13 to 30.6.18.

Vulnerability factors over time

During the observation period, the proportion of deliveries characterized by risk factors for adverse outcome increased. The proportion of pregnancies with antenatal problems including preeclampsia increased significantly from 0.91% to 2.1% (p<0.001). The mean birth weight increased slightly, but an increasing proportion were <1500 g and preterm births increased from 2.5% to 3.5% (Table 1).

Abnormal fetal heart rate during delivery increased gradually from 3% to 9%. The first detected heart rate among resuscitated babies fell from a median of 142 to 108 beats per minute (bpm), and the proportion with heart rate <100 bpm at birth increased from 8% to 22% (p<0.001, Table 2).

Table 2

Postnatal events and heart rate recordings of live births at Haydom 1.7.2013–30.6.2018 (n = 18,997).

		Period of birth
	1st year	2nd year	3rd year	4th year	5th year	p-value*
	n = 4,662	n = 3,909	n = 3,565	n = 3,664	n = 3,197
Apgar <5, n (%)
1 min	37 (0.8)	50 (1.3)	42 (1.2)	45 (1.2)	66 (2.1)	<0.001
5 min	5 (0.1)	7 (0.2)	11 (0.3)	20 (0.6)	23 (0.7)	<0.001
Stimulation/suction*, n (%)						<0.001
No stim./suction	3,950 (85)	3,234 (83)	2,412 (68)	2,684 (73)	2,180 (68)	<0.001
Stimulation only	15 (0.3)	36 (0.9)	100 (2.8)	159 (4.3)	186 (6)
Stim. and suction	697 (15)	641 (16)	1053 (30)	821 (22)	843 (26)
Bag/mask ventilation	361 (7.7)	240 (6.1)	242 (6.8)	272 (7.4)	264 (8.8)	<0.001
Heart rate	n = 389	n = 298	n = 274	n = 262	n = 232
Initial heart rate beats/min median (IQR)	145 (97–179)	142 (94–171)	132 (84–174)	118 (70–162)	102 (67–148)	<0.001
Initial heart rate in categories, n (%)
<60	32 (8)	25 (8)	21 (8)	36 (14)	50 (22)	<0.001
60–100	69 (18)	58 (19)	75 (27)	74 (28)	63 (27)
100–160	139 (36)	98 (33)	85 (31)	80 (31)	75 (32)
>160	149 (38)	117 (39)	93 (34)	72 (27)	44 (19)
Mean heart rate during 0–60 seconds after initial recording, median (IQR)	143 (104–165)	139 (101–165)	139 (105–165)	130 (91–157)	129 (102–152)	<0.001

IQR: Inter-quartile range

* Chi2-test used for categorical analyses, linear regression for continuous analyses. P-values for trend.

Mortality adjusted for vulnerability

In unadjusted analyses, the ENM increased slightly during the five-year period (relative risk per year 1.11, 95%CI 1.00–1.24, Table 3). Vulnerability factors all increased during this period, as shown in Tables 1 and 2. After adjustment for abnormal fetal heart rate during delivery, birth weight, prematurity, initial heart rate and need for BMV at birth, there was no significant reduction in relative risk for ENM (relative risk per year 0.95, 95%CI 0.84–1.07, Table 3). Death the first 24 hours could be due to other reasons than intrapartum events, especially prematurity. Of those who died, 55/169 (33%) had not been ventilated, and thus would likely not be diagnosed as asphyxiated. We performed a sensitivity analysis without these 55: The unadjusted and adjusted analyses were essentially unchanged (Table 3).

Table 3

Relative risk for early neonatal death and perinatal death at Haydom 1.7.2013–30.6.2018.

		Period of birth				Overall linear	p-value†
Early neonatal deaths, n/total live births	1st year 30/4,662	2nd year 41/3,909	3rd year 23/3,565	4th year 56/3,720	5th year 37/3,197
Unadjusted
	1 (ref.)	1.63 (1.02–2.61)	1.00 (0.58–1.72)	1.53 (0.94–2.47)	1.63 (1.11–2.90)	1.11 (1.00–1.24)	0.04
Adjusted for abnormal fetal heart rate
	1 (ref.)	1.55 (0.97–2.48)	0.93 (0.54–1.59)	1.34 (0.83–2.18)	1.44 (0.88–2.33)	1.06 (0.95–1.18)	0.31
Adjusted for birth weight and prematurity
	1 (ref.)	1.58 (1.00–2.48)	1.00 (0.59–1.70)	1.53 (0.95–2.44)	1.72 (1.08–2.72)	1.11 (1.00–1.23)	0.05
Adjusted for initial heart rate
1 (ref.)		1.72 (1.09–2.70)	0.96 (0.56–1.63)	1.34 (0.84–2.14)	1.41 (0.88–2.26)	1.04 (0.94–1.16)	0.41
Adjusted for need for BMV						1.09
1 (ref.)		1.88 (1.19–2.98)	1.10 (0.65–1.88)	1.57 (0.98–2.52)	1.72 (1.07–2.74)	1.09 (0.99–1.21)	0.39
Final adjusted model*
	1 (ref.)	1.88 (1.11–3.19)	0.92 (0.50–1.69)	1.25 (0.72–2.18)	0.93 (0.53–1.64)	0.95 (0.84–1.07)	0.39
Sensitivity analysis excluding those never resuscitated (n = 55), adjusted
	1 (ref.)	1.95 (0.99–3.81)	1.35 (0.66–2.73)	1.39 (0.71–2.72)	0.97 (0.49–1.91)	0.96 (0.83–1.11)	0.60
Perinatal deaths n/total births	106/4,738	99/3,967	62/3,604	92/3,720	77/3,235
Unadjusted
	1 (ref.)	1.12 (0.85–1.46)	0.77 (0.56–1.05)	1.11 (0.84–1.46)	1.06 (0.80–1.42)	1.01 (0.95–1.08)	0.77
Final adjusted model*
	1 (ref.)	1.58 (1.01–2.47)	0.88 (0.54–1.43)	0.87 (0.55–1.39)	0.53 (0.32–0.86)	0.84 (0.76–0.94)	0.002

* adjusted for antenatal problem, abnormal fetal heart rate during labour, birth weight, prematurity, initial heart rate and need for BMV after birth.

† p-value for trend

Excluded from adjusted analyses due to missing covariates: n = 15

The perinatal mortality showed no appreciable change during the five-year period in unadjusted analyses. However, after adjustment for vulnerability factors we found a significant annual decrease of 16% in perinatal mortality (relative risk per year 0.84, 95%CI 0.76–0.94, Table 3).

In the fully adjusted model, year of birth, fetal heart rate abnormalities, initial heart rate, birth weight, prematurity and need for BMV predicted perinatal death (Table 4).

Table 4

Predictors (relative risk with 95% confidence intervals) for perinatal death at Haydom Lutheran Hospital during five years (n = 19,264).

	Unadjusted	Adjusted*
Period, per year increase 2013–18
	1.01 (0.95–1.08)	0.84 (0.76–0.94)
Fetal heart rate:
normal	1 (ref.)	1 (ref.)
abnormal	5.7 (4.4–7.3)	3.0 (2.0–4.5)
not detected	74.0 (64.5–84.9)	62.9 (20.5–193)
Birth weight
<1.5 kg	28.7 (21.0–39.2)	23.6 (10.5–53.2)
1.5–2.5 kg	7.7 (5.8–10.1)	2.3 (1.3–4.1)
2.5–3.5 kg	1.4 (1.1–1.8)	1.0 (0.7–1.6)
3.5–4.5 kg	1 (ref.)	1 (ref.)
>4.5 kg	3.1 (1.6–5.9)	1.5 (0.5–5.1)
Prematurity
	12.9 (10.7–15.6)	5.7 (3.3–9.9)
Low initial heart rate after birth
<60 bpm	165 (131–208)	19.6 (11.1–34.9)
60–100 bpm	33.0 (23.3–46.6)	1.8 (1.0–3.2)
100–160 bpm	6.3 (3.5–11.3)	0.3 (0.2–0.8)
>160 bpm	2.4 (0.99–6.0)	0.2 (0.06–0.48)
No need for recording	1 (ref.)	1 (ref.)
Need for BMV after birth
	74.0 (56.1–97.8)	28.7 (16.0–51.2)

*adjustments for fetal heart rate, birth weight category, prematurity, low initial heart rate and need for BMV after birth. In perinatal mortality results the initial heart rate and presumed need for BMV were set to the lowest category.

Changes in resuscitation performance

The proportion who received BMV varied from 6.1–7.7%, but increased to 8.8% the last year (Table 2). Use of stimulation increased substantially (Table 2). Time from birth to start of BMV was lowest the last year (median 99 seconds) but did not change significantly over time (p = 0.76, Table 5). The duration of the first BMV sequence increased over time from a median of 11 to 32 seconds (p<0.001, Table 5). Once initiated, BMV was continued for an increasing part of the first 60 seconds, from a median of 24 seconds the first compared to 48 seconds the last year (p<0.001, Table 5). However, the overall duration of ventilation from the first to the last inflation remained stable around a median of 130 seconds (range 127–142 seconds), explained by a higher number but shorter duration of BMV sequences in the first years (Table 5).

Table 5

Bag-mask ventilation characteristics of live births at Haydom 1.7.2013–30.6.2018 (n = 18,997).

		Period of birth
	1st year	2nd year	3rd year	4th year	5th year	p-value*
	n = 4,662	n = 3,909	n = 3,565	n = 3,664	n = 3,197
Bag/mask-ventilation, time events
	n = 358	n = 234	n = 240	n = 269	n = 262
Time from birth to BMV, seconds median,(IQR)	105 (74–149)	116 (81–164)	119 (88–158)	109 (74–162)	99 (74–153)	0.76
Duration of first BMV sequence in seconds, median (IQR)	11 (4–20)	11 (5–22)	16 (7–34)	23 (9–52)	32 (13–71)	<0.001
Duration of BMV first 60 seconds in seconds, median (IQR)	24 (15–36)	28 (16–41)	38 (22–51)	46 (33–54)	48 (35–59)	<0.001
Duration of BMV sequences the first five minutes in seconds, median (IQR)	12 (5–21)	12 (5–24)	16 (7–33)	23 (10–45)	27 (11–54)	<0.001
Number of BMV sequences the first five minutes, mean (SD)	4.2 (2.7)	4.5 (2.7)	4.1 (2.5)	3.4 (2.2)	3.1 (2.0)	<0.001
Time from the first to the last inflation in seconds, median (IQR)	132 (56–248)	127 (59–340)	127 (66–298)	142 (64–282)	130 (73–301)	0.94
Bag/mask ventilation characteristics†
	n = 293	n = 247	n = 238	n = 173	n = 114
Frequency of ventilations (inflations per minute)	66 (50–84)	52 (39–70)	44 (35–53)	48 (41–54)	47 (41–53)	<0.001
Tidal volume in ml/kg, median (IQR)	5.3 (3.3–8.4)	6.8 (4.3–10.0)	7.5(5.0–11.5)	6.6 (4.5–10.7)	6.2 (3.2–9.8)	<0.001
Mask leak in %, median (IQR)	45 (33–62)	46 (33–61)	42 (30–57)	41 (29–56)	44 (30–57)	0.001
Mean airway pressure mbar, median (IQR)	14 (12–17)	13 (11–16)	15 (12–18)	17 (14–21)	18 (16–21)	<0.001
Maximum airway pressure mbar, median (IQR)	34 (29–38)	33 (28–36)	36 (32–39)	38 (34–43)	40(37–45)	<0.001

IQR: Inter-quartile range.

BMV: Bag-mask ventilation.

* Chi2-test used for categorical analyses, linear regression for continuous analyses. P-values for trend.

† excluding participants in a randomized study using positive end-expiratory pressure in newborn resuscitation, n = 80 and n = 123 in 4th and 5th year, respectively. This study is reported and described elsewhere [34].

Two-step analysis: Mean of ventilation parameters for each newborn during first 5 minutes of active ventilation calculated, and in the second step this mean was used to calculate median/IQR.

Median ventilation frequency decreased from >60 breaths/min the first year to 44–48 breaths/min the last three years of observation (p<0.001). The median weight-adjusted tidal volume, mean and peak airway pressures increased (p<0.001), whereas mask leak decreased slightly but significantly over time (p = 0.001, Table 5).

Discussion

During a five-year period following a successful implementation of frequent HBB training, we found no further significant decline in ENM. A worrying decrease in hospital deliveries was observed, likely due to introduction of user fees since 2014 of 12 and 30 USD for vaginal and caesarean section deliveries, respectively. Consequently, hospitalized newborns included a higher proportion of infants with increased vulnerability for adverse outcomes. In analyses adjusted for vulnerability factors, perinatal deaths decreased significantly during the five years.

Among several strengths in the study is the large size population-based study, performed in a rural setting of a low-income country with high burden of disease. The prospective design ensured high-quality data from the delivery and immediate post-natal period. Collection of physiological data during the first minutes of life provides information which is rarely accessible.

The selection of higher-risk deliveries during the period is evident, and accounted for as far as possible in the adjusted analysis. However, as in any observational study, residual confounding cannot be completely ruled out. A population-based birth registry with inclusion of all pregnancies including home deliveries would capture the complete picture and avoid selection bias. Specific causes of ENM were not possible to identify in our study, though in a sub-study from this cohort half of the deaths were related to intra-partum events [16]. We were, however, able to exclude deaths unlikely to be caused by perinatal hypoxia in a sensitivity analysis. An increased use of BMV ventilation over the 5-year period likely reflects the higher morbidity, and Apgar scores in this setting have previously been shown to have low prognostic value [17]. However, we cannot rule out that a sustained focus on HBB training may increase unnecessary resuscitation procedures.

During the nation-wide HBB roll-out during 2010/2011 in Tanzania, ENM in eight large hospitals including Haydom decreased from 13.4 to 7.1/1000. Fresh stillbirths decreased from 19 to 14.4/1000, and a remarkable fall of perinatal mortality of 34% was reported [5]. Before and after initial implementation of the HBB program ENM at Haydom was 11.1/1000 and 7.2/1000 respectively. The baseline perinatal mortality was 27/1000 and fell to 21.7/1000, and the decrease in ENM thus accounted for 74% of the reduced perinatal mortality [18]. In the present follow-up, our period of observation starts two years after regular HBB implementation and provides information on perinatal mortality, resuscitation performance and physiological data during a five-year period [6]. Our study shows that the initial reduction in perinatal mortality seen after HBB implementation has continued.

HBB has been introduced globally, and overall a reduction in perinatal mortality has been reported [19]. The variation between countries with highly successful implementation [5, 20, 21] to sites reporting no or modest effect on mortality [22, 23] is of note. The relative contribution of intra-partum hypoxia to ENM across countries may explain some differences, because a lower effect would be expected when other causes of ENM dominate. Besides, implementation and uptake of HBB could influence the results, and repeated low-dose onsite training is likely to facilitate successful implementation [11]. Several studies have reported improved resuscitation knowledge and practical skills after HBB training in a simulation setting [4, 24]. Some loss of performance over time is observed [25-Pediatrics. 2017 ">27], but skills retention was improved by frequent re-training and use of self-evaluation checklists [26]. Importantly, changes in clinical management do not necessarily follow simulated performance [28]. The present study documents improved resuscitation skills that are sustained over time, and is likely to improve perinatal mortality. Heart rate feedback during resuscitation may also have improved skills over time, as suggested in a qualitative study from Haydom [7].

In a study from Nepal, a reduction in the percentage of deaths attributable to intra-partum related complications from 51 to 33% was observed, but overall newborn mortality was unchanged [29]. This was also reported in the recent meta-analysis, pointing to the importance of a package of interventions needed to address the main causes of neonatal death [19]. In our study, one third of deaths occurred in newborns that did not receive BMV after birth. Intra-partum events accounted for about half the deaths before 7 days age at Haydom during 2014–17, but preterm birth and sepsis were among other preventable causes of death [16].

Our study is the first to report long-term changes in performance of resuscitation and BMV after HBB implementation. Analyses of actual ventilation performance demonstrate that practical skills improved over time with significantly longer duration of the first ventilation sequence and reduction in ventilation frequency targeting the recommended range. The increase in more complex patients may also change the resuscitation practice over time. The reduction in mask leak over time was desirable though admittedly small. Higher tidal volumes are associated with more rapid HR increase in observational studies from Haydom [30, 31], but should be further studied in controlled trials before we can conclude whether this is recommendable.

In the initial HBB roll-out at Haydom, the use of suction/stimulation increased whereas use of BMV decreased significantly [11]. We found a further increase in suction/stimulation but also of BMV, which could be expected with increased vulnerability over time. In a meta-analysis of the effect of HBB training on resuscitation practices from four studies, the overall use of suctioning, stimulation or BMV was unchanged. However, interventions during the first Golden minute increased 2.5-fold [32]. In our study the increase during the five years of time used to ventilate the first minute after BMV was started is of note. The intensity and frequency of HBB training varied during the 5-year period, and due to shifts in the staff the quality of birth care and post-natal care is likely to be vulnerable. The substantial fluctuation over time in perinatal death rates, was however not reflected in the data on resuscitation performance.

In the present study the quality of immediate newborn care at birth was improved over time, probably due to continuous focus on frequent HBB training. However, coverage and access are likely even more important for the total perinatal mortality than quality of care in the hospital. These factors are interrelated, as increased quality in a delivery unit is likely to encourage more pregnant women to seek care. However, financial constraints may counteract this effect. In Tanzania in 2011/12, 44% had less than 1.25 USD per day for their daily living, and poverty was even more prevalent in rural districts [33]. In this context, 12 to 30 USD in hospital fees for delivery is a considerable burden to the family economy. Free delivery care should be a high priority, as provided in governmental hospitals in Tanzania.

In this study we have shown that neonatal resuscitation skills improved over time at a site with high focus on continued and frequent HBB training. The perinatal mortality adjusted for vulnerability factors decreased, whereas ENM showed a non-significant decline. Further reduction in neonatal mortality in Tanzania will likely require a comprehensive approach to increase facility births and address the major causes of death.

8 May 2020

PONE-D-19-30351

Increased perinatal survival and improved ventilation skills over a five-year period

PLOS ONE

Dear Dr Størdal,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please pay close attention to the reviewers comments and address their concerns, in particular those around overlap with previous mortality data. In addition, PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. We recommend that authors seek independent editorial help before submitting a revision. These services can be found on the web using search terms like “scientific editing service” or “manuscript editing service.”

We would appreciate receiving your revised manuscript by Jun 22 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Kelli K Ryckman

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please amend your current ethics statement to address the following concerns: Please explain why was written consent was not obtained, how you recorded/documented participant consent, and if the ethics committees/IRBs approved this consent procedure.

3. Thank you for stating the following financial disclosure: "No"

At this time, please address the following queries:

a)    Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

b)    State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

c)    If any authors received a salary from any of your funders, please state which authors and which funders.

d)    If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. Thank you for stating the following in your Competing Interests section:  "No"

Please complete your Competing Interests on the online submission form to state any Competing Interests. If you have no competing interests, please state "The authors have declared that no competing interests exist.", as detailed online in our guide for authors at http://journals.plos.org/plosone/s/submit-now

This information should be included in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

5. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

6. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for the opportunity to review this paper entitled, “Increased perinatal survival and improved ventilation skills over a five-year period”.

This prospective analysis evaluates whether the effect of the HBB program was sustained over time.

They showed that the decrease 24h newborn mortality was sustained even improved after adjusting for covariables. They also noted an improvement on ventilation performance. The is a wealth of data collected in this project and showing feasible ways of improving neonatal survival is an important goal.

My major concern is that the is significant overlap in the mortality data with a previous published paper with only 1.5 years of additional data. I think that focusing more on the ventilation portion versus mortality would add more to the literature given that the mortality data on the majority of the mortality data of this data-set has already been published depicting the fact that the effect of the program can be sustained with frequent training.

Abstract: Predictors of survival can be added to the aims as this was a big focus. In the methods, some information regarding training schedule and data analysis can be added to the methods. I think the first 2 sentences in the results regarding the change in population can be condensed. I would remove the last sentence of the conclusion.

Introduction: I think adding the already known data regarding frequent training and the effect of HBB from reference 6 would be helpful in the intro.

Methods:

1. Though the frequency of the skills training is referenced to another paper, I think this paper would benefit from a brief description of the “low dose high frequency skills training” that is a major intervention thought to be responsible for the sustained outcomes. I think the same can be said for the monitoring set up – it is frustrating for me as the reader to have to reference several articles to determine the major determinants of the methods – though these can be referenced some details are required or the paper does not stand alone.

2. I would like to know from reading this paper alone the basics of the training and the set-up of the ventilation data collection. In addition I feel that the description of neonatal resuscitation performance was sparse and needed more detail.

3. I think that the adjustment for vulnerability factors should go under the statistical analysis section.

4. I recommend that the statistical section be reviewed by the statistician involved in the project as this section feels incomplete. For example, the categorical variable comparisons in the table was noted to be chi squared but this is not mentioned in the methods. In addition, your mortality risk ratios I would assume are based off of multi variable logistic regression given that mortality is a binary variable. Please also include the statistical software used.

Results:

1. The first sentence of the second paragraph read awkwardly.

2. In the second sentence please give exact numbers.

3. Did you do/can you add analysis to determine if there was a significant differences between specific time points versus just an overall difference as presented? This may be helpful.

4. I think graphic depiction of the mortality data versus the many tables would add to the paper.

5. It would be beneficial to break up table 3 as there is so much information. For the linear regression, was the outcome modeled per year? If so the r-square value would be useful in determining the strength of the association. As noted above, the methods in general need more clarification.

6. Was there a qualitative portion of the respiratory data – ie was BMV indicated when administered and appropriately stopped, were TV given appropriate for the size of the infant, etc

Discussion:

1. Generally the for me it makes sense for strengths and limitations to come right before the conclusion.

2. I don’t think the subheadings are needed

3. I do not think the 3rd paragraph under comparison with other studies adds much to the discussion, and could be removed.

4. Can you comment on the fact that continued exposure to higher acuity resuscitations may alone have some impact on these skills.

5. Can you say that this is the first study that shows the long-term sustained effects if your reference 6 by the same groups shows this as well (acknowledged that yours is an additional 18mo follow-up).

6. I think it would be helpful in the discussion to add more on the simulation/training aspect with the ventilation outcome with respect to other studies.

Reviewer #2: Nicely done paper. I've made comments on the attached file. In brief I had a few questions:

- Can you describe who provided the "low-dose, high-frequency skills training"? Was this done by local personnel, or did it involve visiting instructors from Europe, the USA, etc.?

- Can you provide data on emergency vs scheduled/repeat C-Section rates over time? This would be helpful in understanding the changing risk profile of your OB population.

- Can you comment on why year 2 was so much worse than other years (Table 4)?

- Do you know if the increased use of BVM was all due to appropriate and indicated use? For example, it appears that incidence of BVM was 4-fold the incidence of 1-min Apgar < 5, & 10-fold the incidence of 5-min Apgar < 5 (Table 3).

- Can you comment on any potential downsides of HBB, for instance, the incidence of Pneumothoraces?

Reviewer #3: Stordal and colleagues present an observational study reporting on perinatal survival and ventilation skills over a five year period of ongoing training following an initial intervention associated with the Helping Babies Breathe program.

Strengths of this study include the prospective nature, the large and relatively complete data obtained, and the inclusion of physiological data to assist in explaining the mortality outcomes. Key limitations of the study include confounding associated with a change in health funding that altered the complexity of neonatal admissions. This confounding required results to undergo statistical adjustment for increasing vulnerability of the infant case mix, to demonstrate an annual reduction in relative risk for perinatal death.

An annotated manuscript is uploaded to assist authors with their revisions. This manuscript makes some suggestions for improved text – but I would strongly recommend the authors to seek professional English editing: overall I think the text length could be reduced by at least 10 % with more efficient word use, and clearer sentence structure.

MAJOR COMMENTS:

The authors should consult the STROBE guidelines for reporting an observational study.

It would be preferable to make deidentified data available for transparency. De-identified data should not compromise any data privacy restrictions.

Specific statements regarding Financial Disclosure, Competing Interests and Ethics statements should be entered in line with suggested response types in the online submission form rather than just stating “no” or “as stated in the manuscript”.

TITLE: the study design should be indicated in the title and abstract

ABSTRACT

Background/aim: - the years of the initial program should be studied.

Methods should summarise statistical methodology including method used for adjustment of outcomes. rather than refering loosely to analysis of changes over time… (e.g., prospective observational cross-sectional cohort study).

Results – please see suggestions for adjusted wording of the last few sentences.

Conclusions – please see suggestions for adjusted wording. The final conclusion is not justified without data relating to population level neonatal mortality/morbidity over the same interval (i.e., was not being born in a hospital associated with increased neonatal mortality.

INTRODUCTION

Overall – the rationale for the study is not well presented. If the HBB decreased mortality, the authors need to justify why a follow-up study was necessary and what it aimed to achieve.

P9, Sentence 1 and 2: It would be helpful to present the fall in global mortality in both 1-5 years and neonatal population to support the statement in the first sentence that reduction in neonatal deaths has been slower over the last 2 decades

P9 final paragraph. This paragraph is unclear. The abstract suggests 1 follow-up study 2013-2018 and previous paragraphs in the introduction suggest the HBB program commenced in 2010 with the initial study extending for 24 months This paragraph suggests a study 2011-2016, and another study with ongoing onsite-low-dose HF HBB. The crossover in dates is confusing to the reader and it is not clear if this is a hybrid of the initial study or a separate follow-up study. I suggest the authors revise the wording of this paragraph to make it explicity clear when the five year study was completed and the timing of the initial study. This paragraph should provide a hypothesis for the study.

METHODS

Overall, the methodology section appeared to include elements of the previous study (2010-2011) and the current 2013-2016 study which was confusing to read. As this paper is about the 2013-2018 study, the methodology should exclusively relate to what was done in this study, with a citation used to reference back to the previous study.

Key elements of the study design and ethical approvals should be provided early in the methodology section.

The setting should simply state the hospital/location and relevant dates (including day not just month and year) for recruitment and data collection. Detail about fees for delivery and the effect on hospital delivery numbers should be part of a discussion.

Participants should be described including eligibility and exclusion criteria

Outcomes should be described separately to statistical methodology/treatment, and any explanation/justification of the methodology should go into the discussion.

Statistical elements that are missing include management of missing data, methods to examine subgroups/interactions, how confounding variables were managed, the approach to regression modelling and treatment of collinear variables (e.g., birthweight and gestation).

RESULTS

The exclusion of premature infants from the analysis is not well justified – death in the first 24 hours from prematurity is most often related to respiratory distress and may stem from insufficient resuscitation.

Some results are presented without accompanying statistics

It is not clear why infants involved in the PEEP trial were excluded from the ventilation analysis but not from the mortality outcomes, given PEEP may alter effectiveness of resuscitation and hence the resulting mortality.

Ventilation frequency should be presented as inflations/min or breaths/min not as 60/min

Tidal volume was corrected for weight and this should be clear in the results.

DISCUSSION

Paragraph 1: suggest that a full stop is provided after “user fees” and that the rest of the sentence becomes a new sentence as “Consequently, hospitalised newborns included a higher proportion of infants with increased vulnerability for adverse outcomes”.

Reference to the previous study would be better phrased as “before and after initial implementation of the HBB program was 11.1/1000 and 7.2/1000 respectively. The following sentence could be revised to indicate that reduction in early neonatal mortality accounted for the majority of the overall decrease in baseline perinatal mortality from 27/1000 to 21.7/1000 over the same period.

It is not clear why the authors say that the period of observation started 2 years after regular HBB implementation and extended for a further 1.5 years, when the actual study went for 5 years. The numbers don’t add up?

When other studies are quoted, the discussion would have more impact if the information was synthesised and presented as a comparison to the current data to indicate what the literature is telling us overall about the issue at hand. Reference to the study from Nepal does not indicate what the intervention was that led to a reduction of deaths from intra-partum complications but didn’t change overall neonatal mortality, and hence the reader is left confused about the relevance of this information.

It was not clear that the increase in (weight corrected) tidal volumes was due to the HBB training or due to the change in patient mix. It is odd that the increase in tidal volume during resuscitation was not matched by an increase in heart rate. How do the authors explain this?

MINOR COMMENTS

Numbers < 10 should be written in full

Overall, there are too many unnecessary abbreviations in the manuscript that make it difficult to read and digest.

Table 1 – data are usually presented as n (%) rather than % (n). As it is not clear what the numbers are for the child characteristics weight ranges it would be best to indicate in the table legend at the bottom of the table that data are presented as % (n) unless indicated otherwise. Any abbreviations (e.g., bpm) need to be explained in the table legend (which goes for other tables also – e.g., MSB in Table 2).

Table 5 needs to indicate what the numbers are – presume relative risk and 95 % CI?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-19-30351_reviewer.pdf

Click here for additional data file.

13 Aug 2020

We would like to thank the editors and reviewers for their work and valuable comments. The editors’ remarks have been accommodated. In the resubmitted version, we have incorporated changes according to the comments below, and in the following respond point-by-point to the reviewer’s comments.

Reviewer #1: Thank you for the opportunity to review this paper entitled, “Increased perinatal survival and improved ventilation skills over a five-year period”.

This prospective analysis evaluates whether the effect of the HBB program was sustained over time.

They showed that the decrease 24h newborn mortality was sustained even improved after adjusting for covariables. They also noted an improvement on ventilation performance. The is a wealth of data collected in this project and showing feasible ways of improving neonatal survival is an important goal.

My major concern is that the is significant overlap in the mortality data with a previous published paper with only 1.5 years of additional data. I think that focusing more on the ventilation portion versus mortality would add more to the literature given that the mortality data on the majority of the mortality data of this data-set has already been published depicting the fact that the effect of the program can be sustained with frequent training.

The current manuscript adds to the previous paper by ventilation data over time, which have not been published before. In the revised manuscript. We have aimed to further highlight the changes in resuscitation performance stating clearly that this is one of the aims (introduction).

Abstract: Predictors of survival can be added to the aims as this was a big focus. In the methods, some information regarding training schedule and data analysis can be added to the methods. I think the first 2 sentences in the results regarding the change in population can be condensed. I would remove the last sentence of the conclusion.

We have included more information in the aim and the methods, and condensed the results to stay within word limits. As the study has several aims, we intend to narrow down and not include predictors of survival in the aim (end of introduction). We have removed the last sentence of the conclusion.

Introduction: I think adding the already known data regarding frequent training and the effect of HBB from reference 6 would be helpful in the intro.

We have added one sentence and an additional reference (third paragraph, page 2):

“Frequent low-dose training is one of the key elements to improve resuscitation skills.6,7»

Methods:

1. Though the frequency of the skills training is referenced to another paper, I think this paper would benefit from a brief description of the “low dose high frequency skills training” that is a major intervention thought to be responsible for the sustained outcomes. I think the same can be said for the monitoring set up – it is frustrating for me as the reader to have to reference several articles to determine the major determinants of the methods – though these can be referenced some details are required or the paper does not stand alone.

Thank you for this suggestion. In the revised manuscript we have included more information regarding the training and monitor set-up as suggested. A figure of the monitor has been added (Fig 2).

(page 4): Briefly, weekly simulation training was performed on manikins, assisted by local-trainers who had undergone facilitator training for the HBB program. Visiting instructors from other places in Tanzania and Norway also participated regularly.

(page 5): Physiological data collected from 1st July 2013 until 30 Jun, 2018 included heart rate data recorded from ECG, and ventilation parameters obtained from sensors connected to the ventilation bag using a Newborn Resuscitation Monitor (Laerdal Global Health, Stavanger, Norway) (Fig 1). The arch-shaped ECG sensor was placed over the thorax or abdomen of the newborn. Two stainless-steel discs arranged on each side of the flexible arch acted as dry electrodes. The monitor used a hot-wire flow sensor connected to the ventilation bag similar to that in the Florian Respiratory Function Monitor (Acutronic Medical Systems, Hirzel, Switzerland). Volume was calculated by flow signal integration. Pressure was measured using an MPX5010 sensor (Freescale Semiconductor, Austin, TX, USA). The HR was displayed on the monitor installed in front of the resuscitation table visible to the provider, ventilation parameters were not displayed.

2. I would like to know from reading this paper alone the basics of the training and the set-up of the ventilation data collection. In addition I feel that the description of neonatal resuscitation performance was sparse and needed more detail.

Please refer to above corrections for training and monitor setup. Resuscitation performance has been detailed more (page 6, second paragraph).

“The secondary outcomes were clinical performance defined by timing of events during resuscitation: Time from birth to start of BMV, the continuation of BMV by measuring duration of the first ventilation sequence before pause (>5 seconds interruption) and the fraction of time used for BMV once it had been initiated. To assess actual bag-mask ventilation performance, we studied ventilation frequency, tidal volumes and mask leak the first five minutes of bag-mask ventilation.”

3. I think that the adjustment for vulnerability factors should go under the statistical analysis section.

We have moved the description of adjustment variables including “vulnerability factors” to the statistics section as suggested.

4. I recommend that the statistical section be reviewed by the statistician involved in the project as this section feels incomplete. For example, the categorical variable comparisons in the table was noted to be chi squared but this is not mentioned in the methods. In addition, your mortality risk ratios I would assume are based off of multi variable logistic regression given that mortality is a binary variable. Please also include the statistical software used.

The method for analysis of categorical date has been added (Page 6, last paragraph).

The regression method is not a logistic regression model with logit-link (which provides odds ratios) but as the study is a cohort design we have chosen to use a binary regression model with log-link (which provides relative risks). This information has been added to the text. Description of selection of adjustment variables has been moved and expanded in the revised manuscript, please refer to the manuscript for details.

We have also included the statistical software:

“Data were extracted using Matlab 2019a (MathWorks, Natick, MA) and analyzed using Stata SE 15 (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LP.)»

Results:

1. The first sentence of the second paragraph read awkwardly.

We have simplified the sentence to clarify: “During the observation period, the proportion of deliveries characterized by risk factors for adverse outcome increased.”

2. In the second sentence please give exact numbers.

This has been corrected.

3. Did you do/can you add analysis to determine if there was a significant differences between specific time points versus just an overall difference as presented? This may be helpful.

We believe that adding more details would risk to overload the presentation. However, we kindly refer to the paper by Mduma et al (ref. 6) for more details in this respect.

4. I think graphic depiction of the mortality data versus the many tables would add to the paper.

Thank you for this suggestion. In the revised manuscript the mortality data from Table 2 are replaced by a line diagram (Figure 3).

5. It would be beneficial to break up table 3 as there is so much information. For the linear regression, was the outcome modeled per year? If so the r-square value would be useful in determining the strength of the association. As noted above, the methods in general need more clarification.

We agree that Table 3 (now Table 2 and 3) was large and that breaking up to display bag/mask-ventilation in a separate table provides better readability. For the new Table 2, only two of the analyses are linear, so we chose not to include the R square.

6. Was there a qualitative portion of the respiratory data – ie was BMV indicated when administered and appropriately stopped, were TV given appropriate for the size of the infant, etc

This information would certainly be of interest. However, other studies from the research group have analyzed these data (video analysis of appropriate BMV: Haug et al, Clinical Simulation in Nursing, for Tidal volumes: J.E. Linde et al. Resuscitation 117 (2017) 80–86).

Discussion:

1. Generally the for me it makes sense for strengths and limitations to come right before the conclusion.

This could be done but preferences differ among author, reviewers and editors. We feel the logical flow in the discussion is maintained without this change.

2. I don’t think the subheadings are needed

We have removed the subheadings as suggested.

3. I do not think the 3rd paragraph under comparison with other studies adds much to the discussion, and could be removed.

We have removed this sentence as suggested.

4. Can you comment on the fact that continued exposure to higher acuity resuscitations may alone have some impact on these skills.

Thank you for this comment: In a qualitative study from Haydom hospital, the midwives have indicated that feedback from the monitor improves their skills. We have included a new sentence:

“Heart rate feedback during resuscitation may also have improved skills over time, as suggested in a qualitative study from Haydom.7»

5. Can you say that this is the first study that shows the long-term sustained effects if your reference 6 by the same groups shows this as well (acknowledged that yours is an additional 18mo follow-up).

We have modified this sentence, as the current study adds more data and time but the previous study also has long-term data. Instead, we include that the study is the first to monitor changes in ventilation performance with physiological data over time.

“…and provides information on perinatal mortality, resuscitation performance and physiological data during a five-year period”

6. I think it would be helpful in the discussion to add more on the simulation/training aspect with the ventilation outcome with respect to other studies.

We have added new text to several of the paragraphs to highlight the improved resuscitation performance over time, likely due to a high focus on this at the site.

Reviewer #2: Nicely done paper. I've made comments on the attached file. In brief I had a few questions:

- Can you describe who provided the "low-dose, high-frequency skills training"? Was this done by local personnel, or did it involve visiting instructors from Europe, the USA, etc.?

We have included more information regarding the low-dose, high-frequency skills training in the revised version: Local personnel were trained as HBB-trainers, and some were also facilitated by visiting instructors.

“Briefly, simulation training was performed on manikins, assisted by local-trainers who had undergone facilitator training for the HBB program. Visiting instructors from other places in Tanzania and Norway also participated regularly. The trainings were intended to take place weekly, but with some variations over time.»

- Can you provide data on emergency vs scheduled/repeat C-Section rates over time? This would be helpful in understanding the changing risk profile of your OB population.

We have some information regarding planned vs emergency causes of C-Section over time. However, almost all C-Sections are emergency C-Sections, and the data regarding causes of emergency C-Sections do not provide sufficient quality to answer this question.

- Can you comment on why year 2 was so much worse than other years (Table 4)?

We have also seen this fluctuations over time, with year 2 as an outlier. The resuscitation performance (Table 2 and 3) was relatively similar in year 2. The HBB training varied during the 5-year period, and with shifts in the staff the quality of birth care and post-natal care is likely to have varied though not directly reflected in the data on resuscitation performance. We have added comments on this in the discussion, underlying factors not recorded in our data are likely needed to explain this change.

“The intensity and frequency of HBB training varied during the 5-year period, and due to shifts in the staff the quality of birth care and post-natal care is likely to be vulnerable. The substantial fluctuation over time in perinatal death rates, was however not reflected in the data on resuscitation performance.”

- Do you know if the increased use of BVM was all due to appropriate and indicated use? For example, it appears that incidence of BVM was 4-fold the incidence of 1-min Apgar < 5, & 10-fold the incidence of 5-min Apgar < 5 (Table 3).

In this setting, research from our group has documented that the use of Apgar is highly inaccurate (Ersdal et al, Pediatrics 2012; 129: e1238-1243): 50% of those who died secondary to asphyxia had received a 5-minute Apgar score of ≥7. Thus, the Apgar scores are highly inaccurate and likely to explain this discrepancy. We have included a comment and reference to this finding.

“An increased use of BMV ventilation over the 5-year period reflects the higher morbidity, and Apgar scores in this setting have previously been shown to have low prognostic value.”

- Can you comment on any potential downsides of HBB, for instance, the incidence of Pneumothoraces?

We unfortunately do not have access to portable chest x-rays at the hospital, so pneumothoraces are not possible to confirm accurately. Potentially, the focus on stimulation and suctioning from the start of the HBB program may have resulted in unnecessary suctioning that is now discouraged in resuscitation algorithms. We have added the following:

“However, we cannot rule out that a sustained focus on HBB training may increase unnecessary resuscitation procedures.”

Reviewer #3: Stordal and colleagues present an observational study reporting on perinatal survival and ventilation skills over a five year period of ongoing training following an initial intervention associated with the Helping Babies Breathe program.

Strengths of this study include the prospective nature, the large and relatively complete data obtained, and the inclusion of physiological data to assist in explaining the mortality outcomes. Key limitations of the study include confounding associated with a change in health funding that altered the complexity of neonatal admissions. This confounding required results to undergo statistical adjustment for increasing vulnerability of the infant case mix, to demonstrate an annual reduction in relative risk for perinatal death.

An annotated manuscript is uploaded to assist authors with their revisions. This manuscript makes some suggestions for improved text – but I would strongly recommend the authors to seek professional English editing: overall I think the text length could be reduced by at least 10 % with more efficient word use, and clearer sentence structure.

Thank you for this comment, we have revised the text according to the annotated manuscript and aimed to make the revision more to-the-point. We have also revised the language according to the comments.

MAJOR COMMENTS:

The authors should consult the STROBE guidelines for reporting an observational study.

It would be preferable to make deidentified data available for transparency. De-identified data should not compromise any data privacy restrictions.

Data are available upon specific request. However, we are not allowed to make these openly available due to regulations from the National Institute of Medical Research in Tanzania.

De-identified individual participant data will be made available to researchers whose methodologically sound proposal has been approved by the Scientific Steering Comitee for Safer Births Study Group. Proposals may be submitted up to 36 months following article publication to hege.ersdal@safer.net.

Specific statements regarding Financial Disclosure, Competing Interests and Ethics statements should be entered in line with suggested response types in the online submission form rather than just stating “no” or “as stated in the manuscript”.

This has been corrected.

TITLE: the study design should be indicated in the title and abstract

In the title we have now included: An observational study

ABSTRACT

Background/aim: - the years of the initial program should be studied.

The years of the initial program have been included.

Methods should summarise statistical methodology including method used for adjustment of outcomes. rather than refering loosely to analysis of changes over time… (e.g., prospective observational cross-sectional cohort study).

In the revision we have added “…We analyzed changes over time in outcomes, use of resuscitation interventions and performance of resuscitation using regression models to obtain adjusted relative risks.

Results – please see suggestions for adjusted wording of the last few sentences.

In the revision we have changed the wording of the last few sentences to: “During the five-year period, longer duration of bag-mask ventilation sequences without interruption was observed. Delivered tidal volumes were increased and mask leak was decreased during ventilation.”

Conclusions – please see suggestions for adjusted wording. The final conclusion is not justified without data relating to population level neonatal mortality/morbidity over the same interval (i.e., was not being born in a hospital associated with increased neonatal mortality.

We have modified the wording of the conclusion, and removed the last sentence as we do not have population-based data of perinatal mortality for the region.

“The reduction in 24-hour newborn mortality after introduction of Helping Babies Breathe was maintained, and a further decrease over the five-year period was evident when analyses were adjusted for vulnerability of the newborns.”

INTRODUCTION

Overall – the rationale for the study is not well presented. If the HBB decreased mortality, the authors need to justify why a follow-up study was necessary and what it aimed to achieve.

P9, Sentence 1 and 2: It would be helpful to present the fall in global mortality in both 1-5 years and neonatal population to support the statement in the first sentence that reduction in neonatal deaths has been slower over the last 2 decades

The phrasing has been changed according to suggestions. We have added accurate numbers for comparison between neonatal and post-neonatal mortality rates, please refer to the revised manuscript.

P9 final paragraph. This paragraph is unclear. The abstract suggests 1 follow-up study 2013-2018 and previous paragraphs in the introduction suggest the HBB program commenced in 2010 with the initial study extending for 24 months This paragraph suggests a study 2011-2016, and another study with ongoing onsite-low-dose HF HBB. The crossover in dates is confusing to the reader and it is not clear if this is a hybrid of the initial study or a separate follow-up study. I suggest the authors revise the wording of this paragraph to make it explicity clear when the five year study was completed and the timing of the initial study. This paragraph should provide a hypothesis for the study.

We have changed the wording to clarify the periods of time included in our study:

“After introduction of HBB in 2009-2012, efforts to sustain the initial improvements in perinatal mortality continued with onsite low-dose high-frequency HBB training. In the present study, we aimed to document changes in 24-hour neonatal mortality, fresh stillbirths and resuscitation performance during a five-year period from 2013 to 2018.”

METHODS

Overall, the methodology section appeared to include elements of the previous study (2010-2011) and the current 2013-2016 study which was confusing to read. As this paper is about the 2013-2018 study, the methodology should exclusively relate to what was done in this study, with a citation used to reference back to the previous study.

Key elements of the study design and ethical approvals should be provided early in the methodology section.

The setting should simply state the hospital/location and relevant dates (including day not just month and year) for recruitment and data collection. Detail about fees for delivery and the effect on hospital delivery numbers should be part of a discussion.

We have moved the section of ethical approvals to the end of the first paragraph.

Further, we have divided with inclusion of separate headlines the text describing the initial HBB study and the Safer Births study with exact time period and interventions. The text has been revised according to suggestions from the reviewer, please refer to the “track changes” document. The reduction in births and user fees is moved to the discussion.

Participants should be described including eligibility and exclusion criteria

All births in the hospital including stillbirths are recorded in the dataset. This is stated in the revised manuscript: “All births in the hospital from 1st of July 2013 until 30th of June 2018 were recorded in the study.”

Outcomes should be described separately to statistical methodology/treatment, and any explanation/justification of the methodology should go into the discussion.

Statistical elements that are missing include management of missing data, methods to examine subgroups/interactions, how confounding variables were managed, the approach to regression modelling and treatment of collinear variables (e.g., birthweight and gestation).

The outcomes are described under a separate headline, now also including the secondary outcomes of resuscitation performance.

Exposure follows this section, as shown in the revised manuscript (“track changes”). We have added adjustment for type of bag used during the study period as explained.

In the statistics section we have added information regarding handling of missing data, adjustment for confounding and regression models. Subgroup or stratified analyses could be done, but would add further complexity to the presentation. Therefore we prefer to adjust for subgroup characteristics (for example term/preterm birth). The term “gestational age” appears once in the results section, but the correct term should be “prematurity”: Gestational week was not used (this has not been recorded with sufficient precision), only as a dichotomous variable for preterm/term birth. With a Spearman rho of 0.2 between prematurity and birth weight, collinearity does not preclude the analyses.

RESULTS

The exclusion of premature infants from the analysis is not well justified – death in the first 24 hours from prematurity is most often related to respiratory distress and may stem from insufficient resuscitation.

We include all – also preterms - in the main analyses. The analysis excluding those that had not been ventilated at all is performed as a sensitivity analysis only, and this is clarified in the revised manuscript. Whether any of these should have been ventilated is possible, however other causes of death including sepsis or severe prematurity are certainly likely in this category. We have modified the wording to clarify, and added the results from this sensitivity analysis to Table 3.

Some results are presented without accompanying statistics

The numbers from statistical analyses have been added to the text where appropriate, in addition to the original version displaying this in Tables.

It is not clear why infants involved in the PEEP trial were excluded from the ventilation analysis but not from the mortality outcomes, given PEEP may alter effectiveness of resuscitation and hence the resulting mortality.

Thank you for this important comment. The type of bag used did not predict mortality (p=0.25) and therefore did not satisfy the criteria for inclusion in the adjusted model. This is now explained in the Methods/statistics section.

Ventilation frequency should be presented as inflations/min or breaths/min not as 60/min

Tidal volume was corrected for weight and this should be clear in the results.

We have changed this in the revised manuscript.

DISCUSSION

Paragraph 1: suggest that a full stop is provided after “user fees” and that the rest of the sentence becomes a new sentence as “Consequently, hospitalised newborns included a higher proportion of infants with increased vulnerability for adverse outcomes”.

We have revised this paragraph according to suggestions.

Reference to the previous study would be better phrased as “before and after initial implementation of the HBB program was 11.1/1000 and 7.2/1000 respectively. The following sentence could be revised to indicate that reduction in early neonatal mortality accounted for the majority of the overall decrease in baseline perinatal mortality from 27/1000 to 21.7/1000 over the same period.

Thank you for this suggestion, we have rephrased and added the number showing that decrease in ENM accounted for most of the reduction in perinatal mortality:

“The baseline perinatal mortality was 27/1000 and fell to 21.7/1000, and the decrease in ENM thus accounted for 74% of the reduced perinatal mortality”

It is not clear why the authors say that the period of observation started 2 years after regular HBB implementation and extended for a further 1.5 years, when the actual study went for 5 years. The numbers don’t add up?

We have clarified in the revised manuscript, stating time periods clearly in the Methods section and highlighting what is new in the present study.

“In the present follow-up, our period of observation starts two years after regular HBB implementation and provides information on perinatal mortality and resuscitation performance during a five-year period”

When other studies are quoted, the discussion would have more impact if the information was synthesised and presented as a comparison to the current data to indicate what the literature is telling us overall about the issue at hand. Reference to the study from Nepal does not indicate what the intervention was that led to a reduction of deaths from intra-partum complications but didn’t change overall neonatal mortality, and hence the reader is left confused about the relevance of this information.

We have added to the discussion points that we believe are relevant to explain why HBB implementation have different impact across countries (page 12 and 13).

It was not clear that the increase in (weight corrected) tidal volumes was due to the HBB training or due to the change in patient mix. It is odd that the increase in tidal volume during resuscitation was not matched by an increase in heart rate. How do the authors explain this?

The data shown for HR in Table 2 are absolute numbers of the first recorded heart rate and during the first 60 seconds thereafter in order to characterize vulnerability of the newborns over time. Thus, the HR shown is not after one minute of BMV, which is now stated more clearly in the revised table 2.

“Mean heart rate during 0-60 seconds after initial recording, median (IQR)»

MINOR COMMENTS

Numbers < 10 should be written in full

This has been corrected.

Overall, there are too many unnecessary abbreviations in the manuscript that make it difficult to read and digest.

We have replaced FSB and MSB with the phrases “fresh stillbirth” and “macerated stillbirths” throughout.

Table 1 – data are usually presented as n (%) rather than % (n). As it is not clear what the numbers are for the child characteristics weight ranges it would be best to indicate in the table legend at the bottom of the table that data are presented as % (n) unless indicated otherwise. Any abbreviations (e.g., bpm) need to be explained in the table legend (which goes for other tables also – e.g., MSB in Table 2).

We have changed the tables according to this comment to indicate % (n). The legend includes any abbreviation, and MSB is not used now.

Table 5 needs to indicate what the numbers are – presume relative risk and 95 % CI?

Correct, Table 5 (now Table 4) is relative risk and 95% confidence interval as now included in the headline.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

27 Aug 2020

PONE-D-19-30351R1

Increased perinatal survival and improved ventilation skills over a five-year period: An observational study

PLOS ONE

Dear Dr. Størdal,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

Please make the minor changes suggested by reviewer #1. Once this is done the paper should be acceptable for publication.

==============================

Please submit your revised manuscript by Oct 11 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Kelli K Ryckman

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for the opportunity to rereview this paper entitled, “Increased perinatal survival and improved ventilation skills over a five-year period”.

I have reviewed the responses and changes made, and am satisfied with the revised manuscript.

Reviewer #2: Thanks for responding to our prior editorial comments. I still have a few questions as indicated on the attached PDF.

I think you should clarify your description of the amount of BMV the subjects received. In methods, you indicate your analyzed this over the first 5 minutes, but you only present data on the first 60 seconds.

Please clarify how "duration of the first BMV sequence increased" & "BMV was continued for an increasing part of the first 60 seconds", & yet the duration of BMV did not increase?

In Table 5, you refer to a "PEEP-study" but you've eliminated any other mention of this study - you need to clarify that.

There is a truly dramatic increase in the amount of "stimulation" used for resuscitation (Table 2); please comment on that. Was this in part due to the training?

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-19-30351_R1_reviewer.pdf

Click here for additional data file.

15 Sep 2020

Please find attached rebuttal letter in a separate file.

Submitted filename: Rebuttal letter v2 PONE.docx

Click here for additional data file.

29 Sep 2020

Increased perinatal survival and improved ventilation skills over a five-year period: An observational study

PONE-D-19-30351R2

Dear Dr. Størdal,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Kelli K Ryckman

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

1 Oct 2020

PONE-D-19-30351R2

Increased perinatal survival and improved ventilation skills over a five-year period: An observational study

Dear Dr. Størdal:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Kelli K Ryckman

Academic Editor

PLOS ONE