Reducing admission hypothermia in newborns at a tertiary care NICU of northern India: A quality improvement study

Abstract

BACKGROUND:

Hypothermia at admission to neonatal intensive care units (NICU) is associated with increased morbidity and mortality in newborns. A baseline study at a tertiary care hospital with all out-born babies showed admission hypothermia of 82%.

OBJECTIVE:

To reduce admission hypothermia (moderate) in newborns at least by 50% in next 6 months.

METHODS:

A quality improvement (QI) study was planned using WHO Point of Care Quality Improvement Model (POCQI), [17] using PDSA (Plan-Do-Study-Act) cycle approach from April 2018 to March 2019, and including 427 term and preterm babies. We educated the staff, reinforced the use of caps, cling wraps, warm linen, introduced Ziploc bags and ensured adequate use of transport incubator.

RESULTS:

After 6 months, overall admission hypothermia decreased from 82% to 45%, moderate hypothermia reduced from 46% to <10% (P < 0.001) and severe hypothermia (3%) was completely eliminated. There was also significant reduction in incidence of Intraventricular hemorrhage (13% Vs 4.7%), Late onset neonatal sepsis (38% Vs 19%) and metabolic acidosis (43% Vs 28%). We were able to sustain this improvement for the next 6 months and is ongoing. The strongest predictor of hypothermia was newborns being in the phase before QI initiative was started (OR 2.36, 95% CI 1.47, 3.23).

CONCLUSION:

This study is a cost effective approach in reducing admission hypothermia in NICU in a resource limited setting with all outborn babies, and further decreasing the morbidity associated with it. Hence, emphasizing the importance of maintaining euthermia, not only in delivery rooms, but also during transportation.

Keywords

Admission hypothermia NICU quality improvement PDSA

Abbreviations

VLBW

Very low birth weight

ELBW

Extremely low birth weight

LBW

Low birth weight

WHO

World Health Organization

LONS

Late onset neonatal sepsis

IVH

Intraventricular hemorrhage

NEC

Necrotizing enterocolitis

RDS

Respiratory distress syndrome

NRP

Neonatal rescuscitation program

LSCS

Lowe segment cesarean section

PDSA

Plan–Do-Study-Act

NVD

Normal vaginal delivery

Quality improvement

Admission hypothermia

Transitional hypothermia

Delivery room

Labor room

NICU

Neonatal intensive care unit

MOD

Mode of delivery

Metabolic acidosis

SGA

Small for gestational age

Birth asphyxia

1 Introduction

Hypothermia at admission in NICU (Neonatal Intensive Care Unit) is common but preventable cause of morbidity and mortality in newborn [1, 2]. In developing countries, the prevalence of hypothermia is 11% to 92% in community settings, and 8% to 85% in hospital settings [3]. A recent study from India, revealed a 37% prevalence of moderate hypothermia in newborns admitted to NICU [4].

Neonates cannot regulate body temperature as efficiently as adults and thus hypothermia occurs easily. The smaller and more premature the infant, the higher the risk [5]. Both physical factors (less subcutaneous fat content, lack of non-shivering thermogenesis, poor vasomotor control, larger skin surface area to body volume ratios and immature skin with minimal stratum corneum) [6] and environmental factors (low delivery room temperature) predispose the preterm infant to hypothermia.

There is an association with 28% higher mortality and 11% higher late-onset sepsis for each 1°C lower admission temperature below 36°C but these data are from an association study with high risk for bias [7]. Also studies have shown that moderate and severe hypothermia are associated with increased intraventricular hemorrhage (IVH), worsening respiratory distress [8, 9], hypoglycemia, metabolic acidosis [10], and necrotizing enterocolitis (NEC) [11].

Various methods have been shown to be efficacious in preventing neonatal hypothermia, such as raising ambient temperature, skin-to- skin care, use of plastic wrap/bags immediately after birth, early care under radiant warmer and use of thermal chemical mattress and transport incubator [4 , 13]. The latest NRP guidelines also suggest the use of polyethylene bags or a portable warming pad on maintaining temperature in preterm infants [14]. WHO recommends that the temperature in labor and delivery rooms should be at least 25°C [15].

The WHO recognizes newborn thermal care as an essential component of essential newborn care [12], but still hypothermia continues to remain under-documented, under-recognized and under-managed [16] in a developing country like ours. Our center consists of all outborn babies with >2000 NICU admissions per year, who are transported to the hospital not only from within Jaipur city, but all over from Rajasthan. It was observed that quite a lot of babies had hypothermia at admission to our NICU. Transportation of preterm and sick babies and maintaining euthermia was thus a major challenge we were facing. Hence, to address this major issue, we decided to do a Quality Improvement (QI) study to reduce the rate of admission hypothermia in newborns at least by 50% in next 6 months at our center, and further see its effect on morbidity and mortality in newborns admitted to our hospital.

2 Material and methods

2.1 Study design

A Quality improvement prospective study using WHO Point of care Quality Improvement Model (POCQI) using PDSA cycle approach (Plan-Do-Study-Act) was carried out at a Tertiary care private Hospital (Neoclinic), Jaipur, which is one of the largest 75 bedded level III NICU in Rajasthan, from March 2018- February 2019.

Tool: POCQI WHO model [17]: WHO-SEARO in collaboration with the WHO Collaborating Centre for Newborns, All India Institute for Medical Sciences (AIIMS) New Delhi and the United States Agency for International Development Applying Science to Strengthen and Improve Systems (USAID ASSIST), has developed a reliable, validated model to be implemented in health facilities by the teams of healthcare workers to improve the quality of care for mothers and newborns.

Our study was divided into 3 phases: Initial obser- vation phase (2 months), implementation phase (using PDSA cycles: 4 months) and post-implemen- tation phase (6 months).

2.1.1 Initial observation phase

Baseline data was collected from March to April 2018. The temperature at admission to NICU was measured using a digital thermometer placed in the axilla before the neonate was removed from the transport incubator (within 5 minutes of NICU admission). Temperature in Fahrenheit was converted to Celsius for uniformity by the formula: Temperature (°C) = [Temperature (°F) - 32]/1.8 . We measured other parameters like birth weight, gestational age, mode of delivery, birth asphyxia and duration of transport; and other complications associated with hypothermia, such as hypoglycemia (<45 mg/dl), IVH (severe: grade 3,4 as per Volpe’s grading), NEC (Modified Bell’s staging: grade 2,3), metabolic acidosis, LONS: defined as clinical and biochemical evidence (CRP or Blood culture positive) of neonatal sepsis presenting in the neonate after 72 hours of life, and overall mortality.

In our baseline data we observed 63 newborns, 83% babies had admission hypothermia. Out of which, 33 % had mild hypothermia, 47% had moderate hypothermia and 3% had severe hypothermia. Also, 3% babies had hyperthermia. Our data revealed that incidence of hypothermia was increasing with decrease in gestational age and birth weight, which is already a well-known fact. Thus, initially we plan-ned to enroll only preterm and LBW babies, but during our observation it was seen that even term babies admitted to NICU were having hypothermia (30–40%). Hence, we included all the neonates, both preterm and term admitted to our NICU (in first 28 days of life) irrespective of weight and gestation.

Although our admission rate is quite high (2000/year), we could only include 25% of these babies in our study, and rest of them had to be excluded as: many babies were referred to our hospital without any prior information, some babies were brought to NICU directly by parents, and some were brought by accompanying nursing staff of the referring hospital. In all of these cases, actual interventions to be done to reduce hypothermia was before baby reaches our NICU, which was not possible until our team was involved in shifting of the baby or attending its delivery. Also babies who were more than 28 days old, were excluded.

A root cause analysis was done via a Fishbone diagram (Fig. 1) and Pareto principle (Fig. 2) to find out contributing factors of hypothermia. Many causes were identified, such as; lack of awareness and knowledge in the nursing staff, low DR temperature, inadequate resources/materials (Polyethylene wraps, caps, linen), or inadequately applied linen (not covered properly including head and not pre-warmed). Transport incubator was not used mostly, as sometimes it was not working, or the battery was not charged, it was not pre-warmed, and often staff was not aware how to operate it. Often, thermometers were missing or not working. Sometimes, the time taken to shift the baby from DR to NICU was quite long, either due to the long distance or due to the delay by hospital staff due to lack of co-ordination. There was no fixed protocol for shifting of the baby and a proper teamwork was lacking. Then we used the Pareto Principle to prioritize our interventions. As per the Pareto chart 80% of the problem of hypothermia was predominantly due to 20% of the causes such as lack of knowledge and awareness in nursing staff, lack of materials like caps/cling wraps and improper use of linen.

Fig. 1

Fish bone diagram.

Fig. 2

The Pareto chart: X- axis showing various factors identified to be causing hypothermia, and Y- axis shows frequency of factors leading to hypothermia.

2.1.2 Implementation phase

Once we identified our problem, its magnitude, and it’s possible causes, a multidisciplinary thermoregulatory QI team was formed comprising of a Senior Neonatologist, nursing –in-charge, staff, Fellow student, transportation team, and a team leader, who were each assigned a specific duty. Various PDSA cycles were planned to test and adapt possible solutions to the contributing factors of hypothermia as shown in Table 1.

Table 1
Details of PDSA cycles

PDSA Cycle PDSA Cycle Plan Do Study Act

PDSA 1 7-5-2018 to Education of medical and nursing staff –Taking Lectures on hypothermia, re-assessing knowledge after the lectures via questionnaire Mean Temp increased from 35.5°C to 35.7°C Adopt the PDSA cycle and continue with it

31-5-2018 (1 month)

–Teaching material: (pamphlets) distributed to staff and posters kept in each NICU/transport ambulance Moderate Hypothermia decreased from 47% baseline to 33%.

–Digital Thermometer made available, motivated staff to use it always

PDSA 2 1-6-2018 to Using Caps, Polyethylene wraps, linen –Caps (of all sizes) and cling wraps made available Mean Temp increased from 35.7°C to 36.3°C Adopt the PDSA cycle and continue with it.

30-6-2018

(1 month) –Staff taught how to apply cling wrap, warm linen. Moderate Hypothermia decreased from 33% to 25%.

PDSA 3 1-7-2018 to Introducing Ziploc bags –Ziploc bags (10×16 inches) were made available. Mean Temp slightly decreased from 36.3°C to 36.15°C Adopt the PDSA cycle and continue with it.

31-7-2018

(1 month) –Staff was taught how to put baby in bags after delivery before drying. Moderate Hypothermia decreased from 25% to 12.7%.

PDSA 4 1-8-2018 to –Use of Transport Incubator –Transport Incubator was repaired Mean Temp increased from 36.15°C to 36.3°C We could not raise the DR temperature as desired, due to lack of adequate team work/co-operation. We were still facing problem using TI due to some technical issues: So we made some changes and adapted the PDSA cycle.

31-8-2018

(1month) –Increasing the ambient temperatute of Delivery room –Use of transport incubator was demonstrated to staff

–Co-ordinated team work –Team was informed before shifting, Radiant warmer was put on manual mode 100% heater output. Moderate Hypothermia decreased from 12.7% to 9.6%.

temp increased from 36.3°C to 36.4°C

–Nurses of delivery room were encouraged to keep radiant warmer on, to switch off the AC, try to maintain temp to 25°C. In sustainability period mean

PDSA Cycle	PDSA Cycle	Plan	Do	Study	Act
PDSA 1	7-5-2018 to	Education of medical and nursing staff	–Taking Lectures on hypothermia, re-assessing knowledge after the lectures via questionnaire	Mean Temp increased from 35.5°C to 35.7°C	Adopt the PDSA cycle and continue with it
	31-5-2018 (1 month)
			–Teaching material: (pamphlets) distributed to staff and posters kept in each NICU/transport ambulance	Moderate Hypothermia decreased from 47% baseline to 33%.
			–Digital Thermometer made available, motivated staff to use it always
PDSA 2	1-6-2018 to	Using Caps, Polyethylene wraps, linen	–Caps (of all sizes) and cling wraps made available	Mean Temp increased from 35.7°C to 36.3°C	Adopt the PDSA cycle and continue with it.
	30-6-2018
	(1 month)		–Staff taught how to apply cling wrap, warm linen.	Moderate Hypothermia decreased from 33% to 25%.
PDSA 3	1-7-2018 to	Introducing Ziploc bags	–Ziploc bags (10×16 inches) were made available.	Mean Temp slightly decreased from 36.3°C to 36.15°C	Adopt the PDSA cycle and continue with it.
	31-7-2018
	(1 month)		–Staff was taught how to put baby in bags after delivery before drying.	Moderate Hypothermia decreased from 25% to 12.7%.
PDSA 4	1-8-2018 to	–Use of Transport Incubator	–Transport Incubator was repaired	Mean Temp increased from 36.15°C to 36.3°C	We could not raise the DR temperature as desired, due to lack of adequate team work/co-operation. We were still facing problem using TI due to some technical issues: So we made some changes and adapted the PDSA cycle.
	31-8-2018
	(1month)	–Increasing the ambient temperatute of Delivery room	–Use of transport incubator was demonstrated to staff
		–Co-ordinated team work	–Team was informed before shifting, Radiant warmer was put on manual mode 100% heater output.	Moderate Hypothermia decreased from 12.7% to 9.6%.
				temp increased from 36.3°C to 36.4°C
			–Nurses of delivery room were encouraged to keep radiant warmer on, to switch off the AC, try to maintain temp to 25°C.	In sustainability period mean

1st PDSA cycle: The nursing staff was educated and awareness about hypothermia reduction in newborns was increased. Digital thermometers were placed in each NICU as well as in transport bag and proper method of temperature measurement was demonstrated.

2nd PDSA cycle: Caps, polyethylene cling wrap and warm linen’s availability was ensured in the transport bag, and its appliance was ensured during transport. Its proper application was demonstrated to the staff.

3rd PDSA cycle: VLBW babies were placed in Ziploc bags, below the neck, immediately after delivery without drying. This was followed by drying of the head and covering with caps. The bags were kept on during transport and were removed in the NICU only when the core temperature was more than 36.5°C in a stable thermal environment.

4th PDSA cycle: Transport Incubator was repaired and a nursing staff was assigned for its care and charging. The staff was instructed to pre-warm it, before use. We tried to maintain ambient temperature of DR (>25°C) by introducing Digital thermometers in DR, requesting the Obstetric team to switch off the AC, and make radiant warmers available in DR. This required a well-coordinated team work between our transportation team and the shifting team.

2.1.3 Post implementation/sustainability phase

Data was collected from September 2018-February 2019. Sustaining an improvement is a must and for this the QI team met weekly to discuss the various PDSA cycles and review the data. A regular monitoring of temperatures, a checklist to ensure that all supplies and equipment are available, and a regular audit by feedback system in monthly statistical meet was held.

2.2 Measures

Our primary outcome measures was the percentage of admission hypothermia to NICU (mainly moderate) and the mean admission temperature which was calculated on monthly basis during each PDSA cycle, during observation, implementation and post-implementation phase. Other outcomes measures were to compare the incidence of IVH, LONS, hypoglycemia, metabolic acidosis and overall mortality during each PDSA cycles.

2.3 Analysis

Continuous analysis was done on statistical process control charts (p-chart) and control limits were calculated to evaluate the trend of hypothermia reduction. Preliminary analysis was expressed as Mean (SD), Median (range), percentage and frequencies. For parametric data, Chi square/Fisher exact test and for continuous variables t-test were used. Comparison between observation and implementation phase was done in terms of baseline characteristics, various interventions and complications of hypothermia. The p value of less than 0.05 was taken as significant. Data was further analyzed using linear regression analysis. Analysis was done using the Statistical Package for Social Sciences (SPSS) Version 20.0 for Windows (IBM Corporation, New York, USA).

2.4 Ethical consideration

This study was done as a part of quality im- provement initiative in our hospital and informal consent was taken from all parents/guardians before participation.

3 Results

Our study included a total of 427 newborns, 63 in the observation phase, 211 during the implementation phase (PDSA 1, 2, 3 and 4), and 153 in post-implementation phase.The baseline characteristics of our study population in terms of gender, birth weight, gestational age, mode of delivery, duration of transport, any support requirement in form of intravenous fluids/oxygen requirement, birth Asp-hyxia (BA), SGA (Small for gestational age), was found to be similar in different phases of the study as shown in Table 2.verall hypothermia decreased from 83% in observation phase to 47% by PDSA 4, moderate hypothermia had fallen from 47% to less than 10% and severe hypothermia was completely eliminated (P < 0.001). Interestingly, mild hypothermia increased from baseline of 33% to 38.7%. Also baseline hyperthermia of 3% increased to 9% in PDSA 1, but then after it was also controlled (Table 3).

Table 2
Comparison of baseline characteristics of newborns in observation and implementation phase

Baseline characteristics Observation phase Implementation phase P value

(n = 63) (PDSA cycles) (n = 211)

Male 46 (73.02%) 160 (75.8%) 0.65

Females 17 (26.9%) 51 (24.17%)

Gestational age (Mean) 34±4.45 34.8±4 0.4

<28 weeks 6 (9.5%) 15 (7.1%)

28–31 6/7 weeks 20 (31.75%) 47 (22.27) 0.33

32–36 6/7 weeks 14 (22.22%) 64 (30.3%)

>37 weeks 23 (36.5%) 85 (40.28%)

Birth weight (Mean) 2017±947 2120±771 0.85

<1000 gm 8 (12.7%) 18 (8.5%)

1000–1499 gm 14 (22.2%) 32 (15.17%) 0.37

1500–2499 gm 20 (31.7%) 78 (36.97%)

>2500 gm 21 (33.3% ( 83 (39.34%)

Mode of delivery

LSCS 42 (66.67%) 123 (58.29%) 0.23

NVD 21 (33.33%) 88 (41.75%)

Perinatal Asphyxia 4 (6.3%) 21 (9.9%) 0.38

SGA 4 (6.3%) 12 (5.7%) 0.83

IV fluid 29 (46%) 124 (58.7%) 0.074

O2 support 49 (77.7%) 142 (67.3%) 0.112

Duration of transport

<1 hour 47 (74.6%) 182 (86.26%) <0.01

>1 hour 16 (25.4%) 29 (13.74%)

Baseline characteristics	Observation phase	Implementation phase	P value
Male	46 (73.02%)	160 (75.8%)	0.65
Females	17 (26.9%)	51 (24.17%)
Gestational age (Mean)	34±4.45	34.8±4	0.4
<28 weeks	6 (9.5%)	15 (7.1%)
28–31 6/7 weeks	20 (31.75%)	47 (22.27)	0.33
32–36 6/7 weeks	14 (22.22%)	64 (30.3%)
>37 weeks	23 (36.5%)	85 (40.28%)
Birth weight (Mean)	2017±947	2120±771	0.85
<1000 gm	8 (12.7%)	18 (8.5%)
1000–1499 gm	14 (22.2%)	32 (15.17%)	0.37
1500–2499 gm	20 (31.7%)	78 (36.97%)
>2500 gm	21 (33.3% (	83 (39.34%)
Mode of delivery
LSCS	42 (66.67%)	123 (58.29%)	0.23
NVD	21 (33.33%)	88 (41.75%)
Perinatal Asphyxia	4 (6.3%)	21 (9.9%)	0.38
SGA	4 (6.3%)	12 (5.7%)	0.83
IV fluid	29 (46%)	124 (58.7%)	0.074
O2 support	49 (77.7%)	142 (67.3%)	0.112
Duration of transport
<1 hour	47 (74.6%)	182 (86.26%)	<0.01
>1 hour	16 (25.4%)	29 (13.74%)

Data expressed as %, Mean (±SD) wherever applicable. P < 0.05 Significant, <0.01 highly significant.

Table 3

Percentage of normothermia, hypothermia and hyperthermia in observation phase and different PDSA cycles

	Temperature: No. of newborns (%)
	Normal	Mild Hypothermia	Moderate Hypothermia	Hyperthermia
	(>36.5–37.5°C)	(36–36.5°C)	(32–36°C)	(>37.5°C)
Observation phase	9(14.29%)	21(33.33%)	31 (47%)	2 (3.17%)
PDSA 1	18(42.86%)	6(14.29%)	14(33.33%)	4(9.52%)
PDSA 2	30(57.69%)	9(17.31%)	13(25.00%)	0
PDSA 3	31(56.36%)	17(30.91%)	7(12.73%)	0
PDSA 4	32(51.61%)	24(38.71%)	6(9.68%)	0
Sustainability phase	75(49%)	59(38.7%)	19(12.3%)	0

Neonatal caps and cling wraps were used only in 23.8% cases which increased in subsequent PDSA cycles, with maximum leap after PDSA 2 cycle, 78% and 46% respectively (P < 0.001). Similarly, usage of linen increased from baseline of 62% to 86% after PDSA 2 (P < 0.001). The use of Ziploc bags was null, until PDSA 3, where Ziploc bags were introduced, after which its usage increased to 71% (P < 0.05). The use of TI increased from 20% to 54% only after PDSA 4 cycle and also it was pre-warmed more often than earlier. The DR temperature could be recorded in only 1/3 of deliveries, among which it was seen that 88.8% had low temperature (<25°C) and after PDSA 4 it decreased to 56%.

We compared the various complications of hypothermia in between the Observation and Implementation phase, and it was seen that there was no significant difference in rates of hypoglycemia, respiratory distress, NEC and mortality with reduction in admission hypothermia. However, complications like IVH (8% Vs 4.7%, P < 0.01), LONS (38% Vs 19.4%, P < 0.01) and metabolic acidosis (42.8% Vs 28.4%, P < 0.05) showed significant reduction in Implementation phase as compared to Observation phase. Also, severe IVH (grade 3 and 4) decreased from 4.7% to 1.9% in the implementation phase. (Table 4)

Table 4

Comparison of complications of hypothermia between Observation phase and Implementation phase

Complications	Observation phase	Implementation phase	P-value
Hypoglycemia	7 (11%)	31 (14.7%)	0.47
Respiratory distress	46 (73%)	164 (77.7%)	0.44
IVH	8 (12.7%)	10 (4.74%)	<0.01^**
Severe IVH (3,4)	3 (4.7%)	4 (1.9%)
NEC	5 (7.9%)	10 (4.7%)	0.63
LONS	24 (38%)	14 (19.4%)	<0.002^**
Metabolic Acidosis	27 (42.86%)	60 (28.4%)	<0.03^*
Outcome
Discharge	80.95%	89.5%
LAMA	14.3%	6.2%
Death	4.7%	4.26%	0.53

^*P value < 0.05 significant, ^**P value < 0.01 highly significant.

A statistical process control chart (p-chart) was plotted throughout the study, (Fig. 3) which showed that there was gradual reduction in admission Hyp-othermia (moderate) from 47% in April 2018 to <20% in September 2018 and then maintained till February 2019.Also the monthly mean temperature of all the babies increased from 35.5°C (April) to 36.4°C (June) which was further sustained throughout till Feb 2019. It was encouraging to know that we were able to sustain this low rate of hypothermia even in next 6 months.

Fig. 3

Process control chart of the moderately hypothermic babies from April 2018 to February 2019with control limits. The different arrows represents the time of implementation of each PDSA cycle.

In the present study, admission hypothermia was seen in almost 100% of extremely preterm and ELBW babies. Whereas in other gestational age or weight groups 80–90% babies were hypothermic. With our QI initiative, hypothermia decreased significantly in all groups (up to 30–50%, P < 0.01), except in <28 weeks and ELBW babies where incidence of hypothermia was still high (93%, P = 0.64).

A multivariate regression analysis showed that duration of transport (<1 hr Vs >1 hr: OR 1.63, 95% CI 1.2, 2.06), and Birth Asphyxia (Apgar <7: OR 0.83, 95% CI 0.49, 1.18) were strong risk factors for hypothermia and the strongest predictor of hypothermia was babies being in study before QI initiatives were started (OR 2.36, 95% CI 1.47, 3.23) as shown in Table 5.

Table 5

Multivariable model for hypothermia (<36.5°C)

	OR	P value	95% Confidence	Interval
			Lower bound	Upper bound
Before Vs After^**	2.36	0.001^*	1.47	3.23
Gestational age	0.49	0.35	–0.55	1.54
Mode of delivery	0.09	0.71	–0.39	0.57
(LSCS Vs NVD)
Birth Asphyxia	0.83	0.01^*	0.49	1.18
Duration of transport	1.63	0.01^*	1.2	2.06
(<1 hr Vs >1 hr)
SGA	0.22	0.70	–0.91	0.135

^*P < 0.05: significant. OR Odds Ratio. ^**Before Vs After implementation of QI study.

4 Discussion

Overall, the QI approach was quite effective, as we were able to reduce admission hypothermia significantly, where moderate hypothermia was reduced to less than 10%, and severe hypothermia was completely eliminated. There was also slight increase in mild hypothermia from 33% to 38%, which could be due to shifting of moderate hypothermia cases to the mild category. During the study, there was also increase in hyperthermia from baseline of 3% to 9% in 1st PDSA cycle. Previous studies have shown that using a combination of intervention such as cling wrap/cap may lead to hyperthermia [18]. Hence, we avoided unwanted combination of interventions, to eventually control hyperthermia in subsequent cycle.

Our study showed a higher incidence of hypothermia as compared to other studies from developing countries [3, 4]. A study from Northern India quoted admission hypothermia of 70%, [19] whereas a recent study from AIIMS showed moderate hypothermia of about 50% [20]. This variation could due to the different cut-offs taken to define hypothermia. We have taken a cut-off of <36.5°C for hypothermia and <36°C for moderate hypothermia as per WHO, whereas studies which have taken a lower cut-off of 36°C, will obviously have a lower incidence. Also, all our babies are outborn, transported from all over Rajasthan, which itself is a risk factor for developing hypothermia.

There was a reasonably good compliance with the various interventions we introduced during the study. There was drastic increase in the use of neonatal caps, cling wraps, linen and Ziploc bags after their respective PDSA cycles. In fact their use increased immediately after the 1st PDSA cycle itself, which was all about training and educating the staff. Therefore, we regularly had these training sessions during the study period so that staff remained motivated. The Cochrane review in 2004 [13] and previous studies [21] have also shown that using caps, polyethylene wraps and bags have been quite effective in preventing hypothermia in especially preterm babies. One area in which we could not have significant improvement was in maintaining DR temperature and the mean temperature increased from 22.7°C to 23.7°C, but that was still below the recommended temperature as per WHO [12].

We did face a lot of challenges during the study. The most difficult part was in trying to get the staff and doctors to change their behavior, without actually blaming or offending anyone. We could overcome this by actually involving the defaulter person himself into the QI team so that he would feel and act more responsibly. We realized poor compliance of the staff, especially if a staff was new or the workload had increased which was overcome by continuous motivation, education and awareness programs, and holding weekly team meetings to address staff’s concerns. The other problematic area was in using Transport Incubator. Despite, training and demonstration, often TI was not used, sometimes it was not charged, and despite of repair, sometimes it would not work properly. Datta et al. [19] from Delhi, also identified TI dysfunction as one of the key barriers during transportation. The most challenging of all was to maintain the temperature of the delivery rooms. Many studies have shown that maintaining DR temp >25°C leads to increase in admission temperature of premature babies [22]. We were able to document DR temperature only in about 1/3 of deliveries, due to lack of facility. Most of the deliveries were LSCS, conducted in AC rooms, and it was difficult to convince the operating team to switch off the AC.

We also evaluated many complications related to hypothermia, and it was quite reassuring to see that with decrease in hypothermia there was significant decrease in complications like IVH, LONS and Metabolic acidosis. There was also mild decrease in NEC rates from 8% to 4.7%, but it was not statistically significant. Other complications like hypoglyemia, respiratory distress and overall outcome in form of mortality was almost similar throughout the study. Laptook et al. [7] also showed significant decrease in LONS with decrease in hypothermia, however there was no effect on incidence of IVH and NEC. Some studies have shown significant reduction in mortality as well with decrease in hypothermia [19, 23], however our study was not able to demonstrate so. This could be due to a shorter study period leading to insufficient power to detect a small difference in outcome and may be after sustaining our QI efforts for a longer period we might be able to demonstrate a decrease in overall mortality in the long run.

One of the major strengths of our study was to have a highly motivated QI team who could not only effectively implement PDSA cycles, but also sustain it, which is a major success for any QI study. The study was done without any extra major cost, additional staff or new equipment. Hence, just by taking a few preventive measures we could easily reduce admission hypothermia at NICU and this could be easily applicable to any other resource limited settings. Also our study highlights the importance of maintaining euthermia during transportation of sick babies, often a neglected aspect, which would be extremely useful to hospitals which cater outborn babies like ours. Our study includes both term and preterm babies whereas in most of the previous studies only VLBW babies have been included being the high risk group. However even term babies with hypothermia may have increased morbidity and mortality. As the major chunk of babies transported to our center were term (36%) and almost 1/3^rd of our hypothermic babies were term, we included them in the study. This reflected huge lapse and ignorance in maintaining temperature of the babies being transported to our hospital which needed to be rectified, thus highlighting the importance of maintaining euthermia not only in preterm but also in term babies.

Our study did have some limitations though. Our study population could have a lot of variation as all the babies were transported to our center from different parts of the state. However, the baseline characteristics of the babies in observation and implementation phase, were found to be almost similar, except the duration of transport. Here, babies with shorter duration of transport (<1 hour) were significantly more in the implementation phase (86%) as compared to the observation phase (75%, P < 0.01). This could be one of the confounding factors, leading to decreased rate of hypothermia in Implementation phase. However, we compared hypothermia rates in Observation phase with respect to duration of transport (<1 hr: 75% Vs >1 hr: 25%, P = 0.45) and similarly in the Implementation phase (<1 hr: 73% Vs >1 hr: 27%, P = 0.58). This suggests that there was no significant difference in hypothermia due to difference in duration of transport. Also the babies shifted to our NICU, being a tertiary center, were relatively more sick babies, thus we could have a higher morbidity and mortality which cannot be just attributed to hypothermia. Also this study period was relatively for a shorter duration, leading to a smaller sample size which could limit the power of the study, however this being a QI study, and seeing the significant improvement we are motivated enough and continuing our sustainability phase, so as to come up with more long term outcome data in future. Also, In our study, hypothermia rates were as high as 94% in ELBW babies, even after all the QI measures, although statistically not significant, as the number of babies in this group were quite small. However, we still need to work more on this group, as may be the set of interventions used to prevent hypothermia are not sufficient.So, we introduced thermal mattress while transporting these babies in addition to all other interventions.

We have conducted our study mostly in summer and rainy season, and sustainability phase was continued up to winter season. It has been well known that seasonal variation effects the incidence of hypothermia. A study in northern India has reported a 70% incidence of hypothermia among newborns during January to March, 20% during April to June, 32% in July to September and 55% in October to December [24]. Thus, we were expecting higher admission hypothermia rates during winter season, so we were more vigilant, careful and used an extra thermal mattress to sustain euthermia.

5 Conclusion

The study successfully demonstrates that Quality improvement method using the WHO-POCQI model, is a very cost effective approach in reducing admission hypothermia at NICU (especially at a center catering all outborn babies) in both pre-term and term babies in a resource-limited setting, and further decreasing the morbidity associated with it, thus highlighting the importance of maintaining euthermia not only in delivery rooms, but also during transportation. Hence, QI studies should be encouraged in other hospitals of similar settings, where hypothermia is still under recognized and under-managed. We have also been able to sustain our QI efforts, and seeing the positive results, are further motivated to continue the sustainability phase.

Conflicts of interest

None

Funding

None

Contribution of authors

JP, JM and NK conceptualized the study; JP, VS, MV and MR were involved in data collection and analysis. JP and VS have drafted the manuscript and JM has critically reviewed it.

Footnotes

Acknowledgments

NICU nursing staff, transportation team.

References

de Mota Silveira

, Goncalves de Mello

, de Arruda Vidal

, et al. Hypothermia on admission: a risk factor for death in newborns referred to the Pernambuco Institute of Mother and Child Health. J Trop Pediatr. 2003;49:115–20.

Perlman

, Wyllie

, Kattwinkel

, et al. Part neonatal resuscitation: international consensus on cardiopulmonary resuscitation and emergency cardiovascular care science with treatment recommendations. Circulation. 2015;132:204–41.

Lunze

, Bloom

, Jamison

, et al. The global burden of neonatal hypothermia: systematic review of a major challenge for newborn survival. BMC Med. 2013;11:24.

Suman

, Udani

, Nanavati

. Kangaroo mother care for low birth weight infants: a randomized controlled trial. Indian Pediatr. 2008;45:17–23.

Knobel

, Holditch-Davis

, Schwartz

, Wimmer

Jr. . Extremely low birth weight preterm infants lack vasomotor response in relationship to cold body temperatures at birth. J Perinatol. 2009;29(12):814–21.

Cramer

, Wiebe

, Hartling

, Crumley

, Vohra

. Heat loss prevention: a systematic review of occlusive skin wrap for premature neonates. J Perinatol. 2005;25(12):763–9.

Laptook

, Slab

, Bhaskar

. Neonatal Research Network Admission temperature of low birth weight infants: predictors and associated morbidities. Pediatrics. 2007;119:643–9.

Miller

, Lee

, Gould

. Hypothermia in very low birth weight infants: distribution, risk factors and outcomes. J Perinatol. 2011;31:49–56.

Herting

, Speer

, Harms

, et al. Factors influencing morbidity and mortality in infants with severe respiratory distress syndrome treated with single or multiple doses of a natural porcine surfactant. Biol Neonate. 1992;61:26–30.

10.

Bhatt

, White

, Martin

, et al. Transitional hypothermia in preterm newborns. J Perinatol. 2007;27:45–7.

11.

, Joseph

, Bajuk

, Orgill

, Astbury

. Perinatal risk factors for necrotizing enterocolitis. Arch Dis Child. 1984;59(5):430–4.

12.

World Health Organization, Maternal and Newborn Health/Safe Motherhood. Thermal protection of the newborn: a practical guide. Geneva; 1997; pp. 64.

13.

McCall

, Alderdice

, Halliday

, et al. Interventions to prevent hypothermia at birth in preterm and/or low birth weight infants. Cochrane Database Syst Rev. 2008;1:CD004210.

14.

AAP. Textbook of Neonatal Resuscitation. 6th ed. American Academy of Pediatrics. 2011.

15.

American Society of Heating. 2003 ASHRAE handbook: HVAC applications. Atlanta, GA: American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc (ASHRAE); 2003. Chapter 7: pp 7.5-7.7.

16.

Lawn

, Cousens

, Zupan

. 4 million neonatal deaths: When? Where? Why? Lancet. 2005;365(9462):891–900.

17.

www.newbornwhocc.org/POCQI-Learner-Facilitator.pdf

18.

McCarthy

, Molloy

E J

, Twomey

, Murphy

, O’Donnell

. A randomized trial of exothermic mattresses for preterm newborns in polyethylene bags. Pediatrics. 2013;132(1):135–41.

19.

Datta

, Saili

, Goel

, et al. Reducing hypothermia in newborns admitted to a neonatal care unit in a large academic hospital in New Delhi, India. BMJ Open Qual. 2017;6:183.

20.

Sivanandan

, Sethi

, Joshi

, et al. Gains from quality improvement initiatives–experience from a tertiary-care institute in India. Indian Pediatr. 2018;55:809–17.

21.

Trevisanuto

, Doglioni

, Cavallin

, et al. Heat loss prevention in very preterm infants in delivery rooms: a prospective, randomized, controlled trial of polyethylene caps. J Pediatr. 2010;156(6):914–7.

22.

Kent

, William

. Increasing ambient operating theatre temperature and wrapping in polyethylene improves admission temperature in premature infants. J Pediatr Child Health. 2008;44:325–31.

23.

Leng

, Wang

, Lin

, Cheng

, Wang

. Reducing transitional hypothermia in outborn very low birth weight infants. Neonatology. 2016;109:31–6.

24.

Darmstadt

, Kumar

, Yadav

, Singh

, Mohanty

, et al. Introduction of community based skin-to-skin care in rural Uttar Pradesh, India. J Prenatal. 2006;26(10):597–604.