Beryl Thyer

Memorial Africa Trust

Naomi Wright

The evaluation of low-technology radiant heaters for low-income countries

Naomi Wright

Summary

Background: 98% of the 4 million annual neonatal deaths are in developing countries. High-technology radiant heaters contribute significantly to the lower neonatal mortality rates in developed countries, but they are inappropriate for developing countries and hence low-technology radiant heaters are being produced. However, low-technology radiant heaters have not been investigated to ensure they are effective, usable, useful and safe.

Aim: To evaluate the effectiveness, usability, utility and safety of a low-technology radiant heater in the Bansang Hospital, The Gambia.

Methodology: Task 1: All neonates born or admitted to the Bansang Hospital during the 5-week study were placed on the radiant heater and those requiring observation or resuscitation remained on it where their rectal temperature was monitored every 3 minutes. These recordings were utilised to evaluate the effectiveness of the radiant heater at maintaining neonates within the limits of thermoneutrality. Task 2: A survey administered to all staff trained to use the radiant heater to evaluate its usability, utility, and safety.

Findings: 112 neonates were included in the study; 30 required use of the radiant heater; 17 required use of the radiant heat. 15/17 neonates requiring radiant heat were kept within the limits of thermoneutrality or very close giving a success rate of 88% ± 15%. Trained staff stated the radiant heater was easy to use, however some staff had an imperfect knowledge of how to use it. The radiant heater was useful and contributed to saving 30 neonatal lives during the study. Hyperthermia is the main safety risk but it can be eliminated with adequate staff training.

Conclusion: Low-technology radiant heaters are effective at aiding neonatal resuscitation and maintaining neonates within the limits of thermoneutrality and hence have the potential to help significantly decrease neonatal mortality rates in the developing world. However, they will only be effective, usable and safe if implemented with sufficient training.

Introduction

Of the 10 million children under-5 that die in the world every year, 4 million die within the first 4 weeks of life, the neonatal period 1-6. The World Health Organisation (WHO) believe that a failure to address neonatal mortality is what has slowed the decline in childhood mortality in recent years and at the fifty-sixth world assembly (2003) called upon Member States to `make improvements in neonatal health a priority' 2. Public health experts have predicted that unless neonatal mortality is at least halved by 2015, the Millennium Development Goal (MDG) to reduce under-5 mortality by two-thirds will not be met 2.

The WHO has identified immediate post-natal neonatal resuscitation and thermal control in neonates as key interventions to improving neonatal survival 2, 7. In the developed world, high-technology radiant heaters are used in routine neonatal care to aid assessment and resuscitation of neonates immediately after birth while preventing hypothermia using overhead heat 7-22. However, high-technology radiant heaters are inappropriate and unsustainable in developing countries, where 98% of all neonatal deaths occur, and hence low-technology radiant heaters are gradually being introduced. However, as far as I am aware no research has been undertaken on lowtechnology radiant heaters to ensure they are safe, effective and usable like their hightechnology counterparts.

In this study I have evaluated the effectiveness, usability and safety of a lowtechnology radiant heater in the Bansang Hospital, The Gambia. The study consists of a comprehensive literature review of previous research into this area, a methodology, findings, and a discussion of the outcomes and policy recommendations for low-income countries.

Photograph 1: Naomi Wright Standing next to the Low-Technology Radiant Heater in the Labour Ward at the Bansang Hospital, The Gambia

Phototgraph 2: A High-Technology Radiant Heater 64

Photograph 3: The Low-Technology Radiant Heater (without mattress and light-bulbs in this photograph: see

photograph 1)

Background

The Importance of Neonatal Resuscitation

Approximately 5-10% of newborns require some degree of active resuscitation at birth and a further 15-25% require observation until stabilised before being handed to their mother 23. An estimated 19% of the annual 4 million neonatal deaths are attributed to birth asphyxia and the American Heart Association has stated that these can be eliminated using simple resuscitative techniques 23, 24.

The Importance of Preventing Neonatal Hypothermia

Neonatal hypothermia is a worldwide problem across all climates 7, 14, 25-28. Neonates are highly prone to heat loss due to a large surface area relative to weight, an inability to shiver and little insulating fat particularly during resuscitation and observation after birth due to evaporation of amniotic fluid from the skin and the transition from a warm uterus to the cooler, drier delivery room 7, 8, 10, 13, 14, 29-32. Although immediate cold stress may be important to initiate breathing it should be kept to a minimum because prolonged exposure to cool can cause 8, 30, 33.

? Increased oxygen consumption 13, 34, 35

? Hypoxia 13, 14, 36

? Metabolic acidosis 13, 14, 36

? Hypoglycaemia 13, 14, 35, 37

? Coagulative defects 14, 38

? Acute renal failure 14

? Necrotizing enterocolitis 14

? Death 14, 37

There is plentiful evidence to suggest that infants who are not thermally regulated have a significantly higher risk of morbidity and mortality 10, 11, 15, 39-47. Preterm, LBW or distressed neonates are more susceptible to hypothermia than healthy, term neonates because they have less fat for heat production and insulation, decreased glycogen stores, immature skin and poor vascular control and asphyxiated neonates have restricted oxygen supplies 7, 13, 14, 18, 40, 41, 44, 45, 47, 48. Hence, these neonates require warmer environmental conditions than healthy newborns to maintain the same body temperature 10, 39, 41, 49, 50.

The optimum temperature for healthy, term neonates has been defined as between 36·5-37·5ºC with a skin temperature of 0·5-1·0ºC lower 5, 14, 39. The optimum temperature for preterm infants however has yet to be defined due to an ongoing debate about the most accurate method of temperature measurement 14, 51, 52. The WHO classifications for hypothermia are shown in Table 1 7. Table 1: WHO's Classifications for Hypothermia 7

Category of hypothermia Temperature of neonate (ºC)

Mild 36-36.4

Moderate 32-35.9

Severe <32

Interventions to Prevent Hypothermia during Observation and

Resuscitation of Neonates

Initial standard care for all neonates includes: a warm delivery room at 25ºC or more 7, drying immediately after birth 29, 53, wrapping in a pre-warmed blanket, prewarming any contact surfaces and eliminating draughts 14, 39, 54. Following this, evidence suggests that full-term, healthy neonates should be placed in skin- to-skin

contact between the mothers breasts in a method called Kangaroo Mother Care (KMC) 8, 10, 14, 55-63. Neonates requiring observation or resuscitation however must be easily accessible in the minutes following birth and hence KMC is impractical 16. The radiant heater provides such access and has been proven to effectively maintain neonates within the limits of thermoneutrality and hence is used as part of routine neonatal care in all parts of the world that can afford one 7-9, 12-22, 46. Photograph 2 shows the design of a typical high- technology radiant heater 64. There are three reported problems with radiant heaters: increased evaporation and hence insensible water loss, continued risk of hypothermia and the additional risk of hyperthermia. Insensible water loss (IWL) is increased under a radiant heater, but is only clinically relevant for longer-term care 10, 22, 32, 40, 65-72. Evaporative heat loss is also increased but Marks et al has shown that energy expenditure in the form of evaporation is balanced by an increase in heat gain from the heater and hence does not significantly increase total energy expenditure or risk of mortality 14, 45, 65, 73. Hyperthermia is a potential problem since neonates are prone to overheating with a large surface area, limited insulation and limited sweating ability 30, 41, 45, 50, 74. Overheating increases fluid loss and frequency of apnoea and can raise metabolic rate to potentially fatal levels, resulting in an increased risk of mortality 30, 45, 75. WHO's definition for hyperthermia is above 37·5ºC 7, 75.

Appropriateness of Low-Technology rather than High-Technology

Radiant Heaters for Low-Income Countries

High-technology equipment from developed countries is inappropriate for lowincome countries not only because it is expensive but also because it requires sophisticated management, organisation, trained technicians and supplies to operate, monitor and repair it, which are scarcely available 76-82. Radiant heaters in the UK are currently on the market from £8,120 - £11,680 each and even if donated are often outdated and money is not spent on proper packaging and shipping which limits their lifespan 83-86. In Sub-Saharan Africa up to 70% of equipment lies idle for these reasons 87. In response, many doctors have started to develop much cheaper and simpler low-technology equipment which is far more usable and sustainable in developing countries 74, 78, 88-94 . A 5-year project on a neonatal ward in China showed that Western equipment is inappropriate and equipment that can be maintained and repaired using local resources is most effective 95, 96. Daga has shown how the introduction of low-technology equipment in a neonatal unit in Bombay has led to a significant decline in neonatal mortality 89, 97.

A Gap in the Research

Although many studies have been performed on high-technology radiant heaters I have found no research on their low-technology counterpart. One paper explains how to make a low-technology `Kira Kira' radiant heater but no trials have been performed on it 94. This is a concern because potentially these low-technology radiant heaters could be unbeneficial, and hence a waste of time and efforts, or even harmful to neonates. Alternatively, they could be highly effective in preventing hypothermia and aiding resuscitation whilst being affordable, usable and sustainable and hence should be produced and promoted worldwide. It is for these reasons that I propose to perform

an evaluation of a low-technology radiant heater in the Bansang Hospital, The Gambia.

The Situation at Bansang

Before this study, if a neonate was distressed or not breathing at birth at the Bansang hospital it would be `wobbled' or sometimes suctioned if available; no staff were trained in neonatal resuscitation, no resuscitation equipment was available and no interventions were used to prevent hypothermia. In May 2005, McCormick introduced a low-technology radiant heater containing resuscitation equipment to the labour ward (Appendix 1: list of equipment). It was implemented with on-the-job training for the current nurses and midwives and a neonatal course at the State Enrolled Nursing School for future staff. The radiant heater consists of 6 x 100 Watt light bulbs, each with an individual switch, situated 60cm above the mattress (Photographs 1 and 3: the low-technology radiant heater, Appendix 2: Design details).

Statement of Problem

Neonates, even in tropical countries, are prone to hypothermia, particularly during resuscitation and observation after birth. Hypothermia leads to an increased risk of neonatal morbidity and mortality. High-technology radiant heaters used to prevent neonatal hypothermia during resuscitation and observation in developed countries are inappropriate and unsustainable in developing countries. Low-technology radiant heaters being produced for the developing world have not been investigated to ensure they are effective, usable, useful and safe.

Aim

To evaluate whether the low-technology radiant heater at the Bansang Hospital is effective at maintaining neonates wit hin the limits of thermoneutrality (36·5-37·5ºC) and whether it is usable, useful and safe.

Objectives

? To record measurements of the rectal temperatures of all the neonates placed on the radiant heater within the 5-week study period in order to analyse whether they are kept within the limits of thermoneutrality.

? To evaluate the usability, utility and safety of the radiant heater through surveying all the staff that have been trained to use it.

Assumptions

Using evidence from previous studies, it was assumed that neonates would become cold after birth in Bansang, The Gambia 7, 14, 25-28.

Methodology

Literature Review

Initially, I undertook a systematic and thorough literature search on neonatal thermoregulation, resuscitation, low-technology equipment and radiant heaters using grey literature, peer reviewed literature, journal articles and the World Wide Web.

Research Design

I used a combination of quantitative and qualitative research methods to ensure a comprehensive evaluation of the radia nt heater. This consisted of two tasks: Task 1: An Evaluation of the Effectiveness of the Low-Technology Radiant Heater at Maintaining Neonates within the Limits of Thermoneutrality. All neonates born by normal delivery, by caesarean section or diagnosed with hypothermia at the Bansang Hospital within the 5-week study period were included in the study. Neonates born in the labour ward were dried and wrapped after birth and placed directly on the radiant heater. Stable neonates were returned to their mother after 2-3 minutes. Neonates requiring observation or resuscitation remained on the radiant heater until stable or transferred to the paediatric ward for incubator care. Neonates diagnosed with hypothermia (primarily due to poor thermal maintenance following a home delivery or washing the neonate in cold water) were kept on the radiant heater until their temperature was stabilised or until the mother was available and willing to provide KMC. Neonates born by caesarean section had to be resuscitated in theatre and carried over to the radiant heater in the labour ward when stable enough.

While on the radiant heater neonates had their rectal temperature measured every 3 minutes using a flexible neonatal thermometer inserted 2cm into the rectum 7, 98, 99 (Appendix 3: details of the thermometer). The first temperature reading was measured 10 minutes after birth because it was established that it took this amount of time to fall from the intra-uterine temperature of 38·1-38·5ºC to within the normal range. A staff member controlled the neonate's temperature by turning light bulbs situated above the neonate on and off according to the rectal temperature readings (Appendix 4: details of how to use the radiant heater). The staff member was initially a doctor, but after 2- 3 weeks the midwives and nurses were trained in resuscitation of the newborn and use of the radiant heater and took over the role to ensure sustainability after the study.

Other details recorded included: mode of delivery, gestational age, weight, whether the neonate was dried and wrapped adequately, whether the neonate required

resuscitation, and whether the electricity went off at any point during use of the radiant heater (Appendix 5: the results sheet filled in for each neonate).

Initially, a Thermospot was applied to the upper right quadrant of the neonate's abdomen to additionally monitor the skin temperature. However, subsequently this was discontinued due to inaccurate readings (Appendix 6). If the electricity went off while the neonate was on the radiant heater, the neonate would be returned to the mother for KMC if she was available and the neonate stable, or the neonate would remain on the radiant heater and the generator sent for if unstable, which took approximately 10 minutes.

Constants maintained throughout the study

? The time of birth (time zero) was marked by the complete delivery of the neonate, when all body parts were free and visible.

? Temperature readings were taken to an accuracy of 0·1ºC

? The two digital rectal thermometers used were regularly calibrated against two mercury thermometers.

? Each neonate was exposed to the same external conditions of drying, being placed on a warm towel and remaining naked while on the radiant heater 100.

? The radiant heater was placed out of direct sunlight to avoid false recordings from an additional source of radiant heat 7.

Data Analysis

Each neonate that received radiant heat to maintain their temperature was categorised into a success or a failure. A success was initially defined as, `the neonate's temperature remaining within the limits of thermoneutrality (36·5-37·5ºC) once it had initially dropped within these limits from the higher intra- uterine temperature'. However, during the study the definition was adapted to allow for neonates born by caesarean section, at home and those with hypothermia. For these neonates success was defined as, `a neonate warmed to within the limits of thermoneutrality within 3- hours of being placed on the radiant heater and remaining within these limits whilst on the radiant heater'.

95% confidence intervals were calculated using the following formula:

P ± 1.96v(P (1-P)/N)

P = proportion of successes or failures

N = sample size

Excel was used to collate the rectal temperatures of neonates born by normal delivery, caesarean section, home delivery and those with hypothermia into graphs.

Task 2: An Evaluation of the Usability, Utility and Safety of the Low-Technology

Radiant Heater through a Survey of the Staff Trained to use it.

During the first 2-4 weeks of the study, all the nurses and midwives who worked on the maternity unit and in theatre at the Bansang Hospital were trained how to use the radiant heater and the resuscitation equipment inside it. Subsequently, they were

asked to answer a verbal survey on their knowledge of how to use the radiant heater and their opinion of its usability, utility and safety. The survey was administered verbally rather than in the written form in order to make the process quicker and easier for the staff and to help improve compliance rates.

Initially, I piloted the survey on two nurses to ensure the questions were comprehensive, useful, understandable and suitable (Appendix 7). In response to this I adapted and added questions and surveyed the remaining staff with the modified version (Appendix 8). All the staff were asked to read an information sheet about the study (Appendix 9), were given the opportunity to ask questions about the study and were asked to sign a consent form before the survey was initiated (Appendix 10).

Data analysis consisted of adding the number of ticks in each box and summarising longer-answer questions.

Ethical Consideration

With regards to task 1, data on neonatal temperatures was collected from an intervention already in place because rectal temperatures were routinely recorded to control the radiant heater, and hence no additional harm or distress was caused to the neonates. With regards to task 2, I gained fully informed consent from every staff member before completing the survey. All data collected and used throughout the study will remain anonymous. The study design was ethically approved by the ethics board at the Nuffield Centre for International Health and Development, Leeds.

Ethical Approval was not required by an ethics board in The Gambia.

Methodological Study Limitations

? Though Randomised Controlled Trials are the gold standard for research it was not possible in this study because it would not be ethical to expose some neonates to hypothermia.

? The short or long-term clinical outcomes of the neonates on the radiant heater were not assessed. However, arguably there is plentiful evidence already available to prove the benefits of keeping a neonate within the limits of thermoneutrality. This study sought to evaluate whether the radiant heater is effective at keeping the neonates within this range.

? Air flow rate, humidity, and ambient air temperature were not measured in this study but they do affect the optimum environment for a neonate 10, 30, 39, 41, 44, 49, 65, 101. However, it has been shown that if temperature is maintained at a constant level then these other factors do not affect the survival of premature neonates 45. Also, the study represents a true evaluation of the radiant heater because the above factors would not be controlled in clinical practice.

? IWL was not monitored during this study, however previous research suggests it is not relevant for short-term care 10, 22, 32, 65-72.

? The accuracy of the rectal temperature recordings can be affected by the depth of insertion of the thermometer and whether the neonate had passed stool 102.

? The results are dependant on the weather conditions at Bansang during the study and may not represent an accurate evaluation of the radiant heater for other climates.

Findings

Task 1: The Effectiveness of the Low-Technology Radiant Heater at

Maintaining Neonates within the Limits of Thermoneutrality

112 neonates were included in the study; 87 (78%) were born by normal delivery in the labour ward, 20 (18%) were born by caesarean section, 3 (3%) were delivered at home and admitted later and 2 (2%) were treated for hypothermia. 64/112 (57%) were normal at birth and were given directly to the mother, 18/112 (16%) were stillborn and 30/112 (27%) required resuscitation and/or observation on the radiant heater (15/30 resuscitation and 15/30 observation).

Of the 30 neonates requiring care on the radiant heater, 15 (50%) were delivered normally (Figure 1), 10 (33%) were delivered by caesarean section (Figure 2), 3

(10%) were delivered at home and admitted later (Figure 3) and 2 (7%) were neonates with hypothermia (Figure 4).

Figure 1: The Rectal Temperatures of the Neonates born by Normal Delivery whilst on the Low-Technology Radiant Heater

32.5

33.5

34.5

35.5

36.5

37.5

38.5

39.5

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Time After Birth (Minutes)

Rectal Temperature (Celcius)

Neonate 1

Neonate 2

Neonate 3

Neonate 4

Neonate 5

Neonate 6

Neonate 7

Neonate 8

Neonate 9

Neonate 10

Neonate 11

Neonate 12

Neonate 13

Neonate 14

Neonate 15

17/30 of the neonates on the radiant heater required radiant heat to maintain their temperature; these consisted of 4/15 born by normal delivery, 8/10 born by caesarean section, 3/3 admitted after a home delivery and 2/2 with hypothermia. Of the 17/30 requiring radiant heat for temperature maintenance, for 12/17 (71%) it was successful and for 5/17 (29%) it was a failure (Figure 5). However, the rectal temperature in 2 of the failure cases only fell to 36·4ºC temporarily and was soon corrected (Neonates 23 and 25, Figure 5), and in 1 case the temperature dropped to 36·2ºC before correction within 6 minutes (neonate 3, figure 5). In the other 2 failure cases the radiant heater

had not been used correctly, the rectal temperatures were not measured and the light bulbs not used (Neonates 14 and 15, Figure 5). When a member from the study arrived and measured the rectal temperatures the decision was made to warm the neonates by KMC.

Figure 2: The Rectal Temperatures of the Neonates born by Caesarean Section whilst on the Low-Technology Radiant Heater

32.5

33.5

34.5

35.5

36.5

37.5

38.5

39.5

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180

Time After Birth (Minutes)

Rectal Temperature (Celcius)

Neonate

Figure 3: The Rectal Temperatures of Neonates born at Home and admitted later whilst on the Low-Technology Radiant Heater

32.5

33.5

34.5

35.5

36.5

37.5

38.5

39.5

0 10 20 30 40 50 60 70

Time After Admittance to Hospital (Minutes)

Rectal Temperature (Celcius)

Neonate

In 2/30 cases the electricity failed whilst the neonate was on the radiant heater. This was potentially life-threatening for one neonate who required suction at birth but had to wait 10 minutes for the generator to be turned on. 9/30 neonates were not dried and wrapped adequately at birth, of which 8/9 (89%) became hypothermic and required radiant heat. Of the 21/30 who were dried adequately, only 6/21 (29%) became hypothermic.

42/112 (38%) of the neonates in the study were LBW (= 2·5kg). 11/30 (37%) of neonates requiring care on the radiant heater were LBW. 8/11 (73%) of the LBW neonates requiring resuscitation or observation on the radiant heater required use of the radiant heat, compared to 9/19 (47%) of term neonates.

Figure 4: The Rectal Temperatures of Neonates requiring Treatment for Hypothermia whilst on the Low-Technology Radiant Heater

32.5

33.5

34.5

35.5

36.5

37.5

38.5

39.5

0 10 20 30 40 50 60 70

Time after Diagnosis of Hypothermia (Minutes)

Rectal Temperature (Celcius)

Neonate

Figure 5: The Rectal Temperatures of the Neonates whose Temperatures Failed to remain within the Limits of Thermoneutrality

32.5

33.5

34.5

35.5

36.5

37.5

38.5

39.5

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Time after Birth or Admittance (Minutes)

Rectal Temperature (Celcius)

Neonate

Task 2: A Survey of the Usability, Utility and Safety of the Low

Technology Radiant Heater

10 midwives and nurses at the Bansang Hospital were surveyed in total. All staff had been trained in the use of the radiant heater, but only 5 staff members had used it in clinical practice. The survey findings are shown in Table 2.

Table 2: The Results of the Verbal Survey administered to 10 Staff Trained in the use of the Radiant Heater at the Bansang Hospital.

Subject Survey Question Answers

Usability How easy is the radiant heater to use?

3 OK

1 Easy

6 Very Easy

Could the radiant heater be improved to make it easier to use? Yes/ No

If Yes, how?

10 No

Is using the radiant heater too timeconsuming? Yes/ No

If Yes, why?

4 Yes

6 No

Key comments: Staff shortages is a problem because at times there could be only one staff member working in maternity and their time has to be prioritised between the neonate and the mother.

Knowledge of What is the minimum temperature 7 36.5ºC

how to use the Radiant Heater.

the neonate should be? 1 35.5ºC

1 34.5ºC

1 could not remember

What would you do if the neonate's temperature fell below this limit whilst on the radiant heater?

8 turn the light bulbs on

2 nothing

What is the maximum temperature the neonate should be whilst on the radiant heater?

7 37.5ºC

1 37ºC

1 36ºC

1 could not remember

What would you do if the neonate's temperature rose above this limit whilst on the radiant heater?

9 turn the light bulbs off

1 call the doctor because the neonate may have a fever

What would you do if the electricity went off while the neonate was receiving temperature maintenance on the radiant heater?

5 wrap the neonate and send for the generator to be turned on

3 just wrap the neonate

1 put the neonate in KMC and send for the generator if need be.

1 wrap the neonate and place a hot water bottle beside it

How long should a healthy neonate remain on the radiant heater?

10 a few minutes

Where should it go after that?

10 with the mother

Where should it go if the mother is not available?

7 cot beside the bed

2 returned to the radiant heater

1 with a family member

How long should a neonate requiring observation or resuscitation remain on the radiant

heater?

3 until the neonate is stable

3 30 minutes

1 15 minutes

1 10 minutes

2 5 minutes

Where should it go once stabilised?

10 to the mother

When should the first temperature reading be taken whilst on the radiant heater?

4 after 10 minutes

3 immediately

1 after 3 minutes

1 after 15 minutes

1 not sure

How frequently should temperature readings be taken whilst the neonate is on the radiant heater?

6 3 minutes

2 5 minutes

1 10 minutes

1 not sure

Usefulness Does the radiant heater provide a useful working area? Yes/ No

If Yes, why?

10 Yes

Key comments: “all the equipment is at hand”, “the baby can be kept warm while resuscitating”, “it provides easy access to the baby”.

Do you feel confident using the resuscitation equipment in the radiant heater? Yes/ No

10 Yes

Has the radiant heater improved the care for neonates at the Bansang Hospital? Yes/ No

10 Yes

Will the radiant heater save lives?

Yes/ No

10 Yes

Safety How often should the mattress cover be cleaned?

10 between every use

How often should the resuscitation equipment be cleaned?

10 between every use

How often should inside the drawers and cupboard be cleaned?

6 once a day

3 every few days

1 between every use

Have you been involved in cleaning the radiant heater before? Yes/ No

5 Yes (the staff that have used it in practice)

5 No

If Yes, how easy is the mattress cover to clean?

2 OK

3 very easy

How easy is it to clean the inside of the drawers and cupboard?

1 OK

1 easy

3 very easy

Are there any safety concerns for the neonate whilst on the radiant heater?

Yes/ No

6 Yes

4 No

If yes, what are they?

3 hyperthermia

2 neonate rolling off the mattress

2 infection control

Improvements How were you trained to use the radiant heater?

9 on-the-job training

1 neonatal course

Could you improve the training you received? Yes/ No

8 Yes

2 No

If Yes, how?

4 a few more weeks assistance

3 refresher course after 6 months

1 who was trained through the neonatal course would have liked on-the-job training too.

Comments: “all the staff are regularly rotated in the hospital so continued training is vital to the sustainability

of its use”.

Could the design of the radiant heater be improved? Yes/ No

5 Yes

5 No

If Yes, how?

3 sides to prevent the neonate rolling off

2 the light bulbs should be a locally manufactured brand

1 battery back-up

1 inbuilt suction and oxygen machines

Discussion

Is there a need for Radiant Heaters in The Gambia?

During the study 30/112 (27%) neonates required observation and/or resuscitation on the radiant heater, identifying a need for such apparatus in that setting. 17/30 neonates required use of the radiant heat to maintain their temperature, despite the ambient air temperature in the labour ward remaining between 29-39ºC throughout the study, which supports previous evidence that neonatal thermoregulation is required even in tropical climates 77, 11, 25-28. It is important to note that this study was undertaken during the warmest season in The Gambia and hence the radiant heat will be required at the same level or more frequently during the other seasons when the ambient air temperature can fall as low as 7ºC 103.

Is the Low-Technology Radiant Heater Effective?

This was evaluated by analysing the rectal temperatures of the neonates placed on the radiant heater to determine whether they remained within the limits of thermoneutrality (36·5-37·5ºC). Of the 17/30 neonates requiring radiant heat, 12 remained constantly within the boundaries of thermoneutrality and a further three remained very close (Neonates 3, 23 and 25, Figure 5), which if counted collectively give an overall success rate of 88% ± 15%. Hence, 15/17 neonates were given a better standard of care and a lowered risk of morbidity and mortality through the use of the radiant heater. This is demonstrated by the 2 neonates that did not have their temperatures measured or radiant heat provided whilst on the radiant heater and subsequently became hypothermic (Neonates 14 and 15, Figure 5). It must be noted that these two failure cases were caused by a failure in the staff training, which prevented proper use of the radiant heater, rather than the radiant heater being ineffective. Hence, this radiant heater has the potential to lower the incidence of neonatal hypothermia and therefore neonatal morbidity and mortality at the Bansang Hospital and other developing countries if used correctly. However, the sample size in this study was relatively small and a larger study would be useful to confirm these findings.

The radiant heater may be less effective in countries with cooler climates or in Bansang during the cooler seasons, which warrants further investigation. However, the 600-Watt output of this radiant heater is comparable to high-technology radiant heaters in developed countries with cooler climates, which range from 400 - 750 Watts, and hence it should be powerful enough 7, 13, 29, 46, 64, 104. If required, the radiant heater could be adapted to include more light bulbs for countries with cooler climates.

The effectiveness of the radiant heater could also be affected by the method of thermometry used to measure neonatal core temperature. After reviewing the literature, rectal thermometry was concluded to be the most accurate and appropriate method to use. This is supported by many paediatricians and a recent systematic review 102, 105-122.

Is the Low-Technology Radiant Heater Usable?

This was evaluated using the staff's opinion of how easy the radiant heater is to use, their knowledge of how to use it and any problems they encountered.

Staff Opinion

All the staff stated that the radiant heater was very easy, easy or OK to use and no one could suggest any improvements to make it easier to use.

Staff Knowledge

Not all the staff answered the questions correctly on how to use the radiant heater. 2/10 staff thought the minimum rectal temperature for the neonate should be 35ºC and 34·5ºC and stated if the temperature were to fall below this level they would do nothing. This would result in neonates becoming hypothermic despite the availability of the radiant heater. Similarly 3/10 staff did not know the maximum rectal temperature for neonates on the radiant heater, however none of the answers were above 37·5ºC suggesting the neonates would not be at risk of hyperthermia. 9/10 staff stated they would turn the light bulbs off if the temperature rose above 37.5ºC to prevent hyperthermia. However, one stated they would call a doctor in case the neonate had a fever. Although this is not the correct answer, it is a sensible response because some neonatal temperatures were higher than normal due to neonatal or maternal infection and in these cases the doctor should be informed. It is unclear whether this staff member knew whether to turn the light bulbs off if the temperature rose above 37·5ºC, which is a potential danger. It is important to note that the boundaries of thermoneutrality are clearly displayed on the radiant heater to inform the staff at the point of use, as are the instructions of how to use it. There was some confusion regarding when the temperature readings should be taken. 8/10 stated the first temperature reading should be taken at or before 10 minutes, which would ensure good temperature maintenance, however, 1/10 staff stated after 15 minutes, which could potentially allow the neonate to become hypothermic before receiving radiant heat. 6/10 staff knew to take the rectal temperature every 3 minutes, however, 2/10 stated every 5 minutes and 1/10 every 10 minutes, which are acceptable but could result in larger temperature fluctuations. There were mixed views on how long the neonate should remain on the radiant heater. This could have been because the y did not know or it could have been a confusing question because although only 3/10 correctly stated `until the neonate is stable', the answers given by the other 7/10 staff were between 5-30 minutes which could be correct from their experience; it varies for every neonate. It is unlikely that a staff member would return a neonate to the mother before they had finished caring for it

and none of the times given would keep the neonate separated from the mother for an unacceptable length of time.

Training

It is clear that adequate staff training is vital for the effective use of this lowtechnology radiant heater, since although none of the staff's answers exposed any serious dangers to the neonates, their knowledge of how to use it was imperfect. The staff member who had not received on-the-job training knew least about how to use the radiant heater, proving its value. 4/10 stressed they would have liked a few more weeks of on-the-job assistance, 3/10 suggested the need for a refresher course and it was also stressed that the nurses at the hospital are regularly rotated and hence continued training would be required. One solution for this could be for two permanent, senior members of staff to be highly trained and examined on the use of the radiant heater and resuscitation equipment, and own literature detailing these for future use, and for these staff members to be in charge of continued training and refresher courses. The Neonatal Resuscitation Program implemented across India utilised this technique and led to a significant decline in asphyxiated deaths 123. Dunn performed a one-day resuscitation programme with 190 nurses and showed it significantly increased their knowledge and skill performance, however this was not maintained at 6 months, confirming the need for refresher training at Bansang 124. It is important to note the value of this survey in improving the knowledge of the staff; it allowed all problem areas to be identified and for incorrect answers to be explained. Hence, staff should be examined at the end of training and refresher courses. Time Constraints 4/10 staff thought the radiant heater was too time-consuming, and at times, when only one staff member is working, the radiant heater would not be used. There are two ways to reduce the time-consuming process of measuring rectal temperatures: setting the radiant heater at one temperature for all neonates or servo-control. It has been well established that neonates of different weights and gestational ages require different levels of radiant heat and to set the radiant heater at one temperature could overheat some neonates and continue to expose others, particularly LBW neonates, to hypothermia 10, 13, 32, 35, 39, 41, 45, 47, 49, 50. This is supported by the 73% (8/11) of LBW neonates requiring use of the radiant heat on the radiant heater, compared to 47% (9/19) of term neonates. Servo-control is not possible because it requires hightechnology equipment 64, 125, 126. Hence, the manual method used is most appropriate for low-income countries. Furthermore, this radiant heater was not designed to provide long-term care; its purpose is to keep neonates warm during resuscitation and observation immediately after birth, a time when it is vital for the staff member to remain with the neonate anyway.

Is the Low-Technology Radiant Heater Useful?

All of the staff stated that the radiant heater provides an ideal working area for resuscitating newborns; the surface provides access to the neonate from three sides, the resuscitation equipment is readily assessible and the neonate can be maintained at the optimum temperature for survival. All of the staff stated they were confident in using the resuscitation equipment, and in combination with the decreased risk of

morbidity and mortality provided by the radiant heat, the low-technology radiant heater has the potential to improve the care and save many neonatal lives at the Bansang Hospital, as stated by all 10 staff. This is demonstrated by the 15/112 neonates that were successfully resuscitated and the further 15/112 neonates observed on the radiant heater during the study, who previously would have received no neonatal care or thermoregulation. Using the hospital records it can be estimated that if used correctly, approximately 389 (27%) of the 1442 neonates born at the Bansang Hospital each year will benefit from a lower risk of morbidity and mortality through these interventions 127.

Is the Low-Technology Radiant Heater Safe?

Hyperthermia

As mentioned above, lack of knowledge and awareness about the hazards of using the radiant heater could be dangerous to the neonate if allowed to become hyperthermic, and hence an examination at the end of the training course should be mandatory to ensure this knowledge. Only 3/10 staff stated that hyperthermia was a risk factor for the neonate, however this could have been a misleading question because it did not directly ask about the risk of hyperthermia and they could have just not mentioned it. Le Blanc describes cases where neonates have died of heat stroke on radiant heaters but these involved neonates receiving long-term care on servo-controlled radiant heaters where the alarm failed 128. This radiant heater has no alarm and is designed only for short-term attended use and hence even if temperature measurements were not taken correctly the staff member would see if the neonate became distressed. All staff members were taught that if for any reason they must leave the neonate unattended on the radiant heater they must turn the lights off because the risks of hyperthermia are as great as the risks of hypothermia 7.

Re-warming Hypothermic Neonates

Rapid re-warming of hypothermic neonates has been associated with convulsions, apnoea and worsening of metabolic acidosis 11, 73, 129. However, Tafari and Gentz and Sarman et al have shown that rapid re-warming over hours is associated with a lower neonatal death rate than slow re-warming over days and hence the WHO advise rapid re-warming 7, 130, 131. The two neonates in this study who were admitted with hypothermia at 35·6ºC and 36·3ºC were warmed to within the normal temperature range within 52 and 38.5 minutes respectively and showed no signs of apnoea or seizures (Figure 4).

Infection

Spread of infection is a risk with all hospital equipment. 10/10 staff stated they would clean the mattress cover of the radiant heater and resuscitation equipment between every use and 10/10 stated the inside of the draws and cupboard should be washed every few days or more frequently. All these answers provide a sufficient level of cleanliness; however even in developed countries with well-developed infection control systems cases of cross- infection from radiant heaters have been reported 132, 133. Added concerns for the Bansang Hospital include frequent lack of running water and poor funding for disinfectants. However, placing the neonates on the radiant

heater surface provides no greater risk of infection than the previous practice of placing 3-4 neonates in one cot, which was rarely cleaned 134. The radiant heater is sufficiently easy to clean, with answers varying from very easy to OK. A greater concern is the rectal thermometer as a route of infection, which must be sterilised between each use.

The Neonate Rolling off the Radiant Heater

This is not a safety issue since neonates cannot roll until 9-10 weeks of age 135.

Is the Low-Technology Radiant Heater Appropriate?

This low-technology radiant heater cost less than £100 to produce, does not require technical expertise to operate it and is easy to maintain because the only maintenance problem could be a light-bulb or fuse needing changing. Hence, it is highly appropriate and sustainable in low-income countries.

Other Considerations

Neonates born by Caesarean Section

20/112 (28%) of neonates in the study were born by caesarean section. 10/20 (50%) of these required resuscitation in theatre and/or observation on the radiant heater when returned to the labour ward. This is a much larger proportion than the 15/87 (17%) neonates born by normal delivery requiring observation and/or resuscitation, which is not surprising because many of the caesarean sections were undertaken due to dystocia and foetal distress and neonates required radiant heat to maintain their temperature until the mother returned from theatre. Hypothermia was a significantly greater problem amongst neonates born by caesarean section than those born by normal delivery, with 8/10 (80%) of the neonates born by caesarean section requiring use of the radiant heat on the radiant heater compared to 4/15 (27%) of neonates born by normal delivery. Reasons for this increased heat loss include: air conditioning in theatre, delay in putting the neonate under the radiant heat, transport of the neonate from theatre to the labour ward, and delay in putting the neonate with the mother. This additional heat loss in caesarean section neonates is illustrated by the number of neonates below 36·5ºC on using the radiant heater in figure 2, compared to the majority of neonates born by normal delivery remaining above 36·5ºC in figure 1. 4/20 (20%) of the neonates born by caesarean section were already hypothermic on arrival to the labour ward, a figure that could be even higher during the colder seasons. This is significant because low admission temperature has been independently associated with an increased risk of morbidity and mortality in preterm neonates, and hence interventions are required to reduce such heat loss 21, 40, 136.

For the first caesarean section delivery, the radiant heater was pushed over to theatre, the neonate received radiant heat and remained within the limits of thermoneutrality (Neonate 16, Figure 2). However, transporting the radiant heater with an uneven ground surface proved too difficult and hence subsequently it remained in the labour ward. Therefore, a second low-technology radiant heater is required for theatre, not only to keep neonates warm during resuscitation and observation but also to provide a better working area for resuscitation. Currently, neonates are resuscitated on a cold,

moving trolley with resuscitation equipment borrowed from the radiant heater and with suction and oxygen shared with the surgeon. It is advisable that the radiant heater remains outside the main operating room however to minimise risk of explosion due to ignition of combustible anaesthetic gases by the hot surface of the radiant heater 137.

To reduce heat loss during transportation from theatre to the labour ward an occlusive plastic wrapping could be considered, since Baum and Scopes, Hobbs and Besch et al have shown they can significantly reduce heat loss 46, 138, 139. However, these studies were evaluated with the wrapping administered directly from birth and hence further studies are required to evaluate whether they are effective just during transportation.

In this situation, it would not be necessary to apply the plastic wrap immediately after birth, as some studies suggest, because this study has shown that neonatal temperature can be maintained on the radiant heater without additional interventions and wrapping could delay the onset of resuscitation and obscure examination of the neonate even if transparent. Furthermore, Björklund noted that a low Apgar score was more common in infants enclosed in plastic bags immediately after birth, possibly due to the removal of an initial cold stimulus and melting of the plastic wrap onto the neonates skin is a further risk 18. If a wrapping is used during transportation it should be removed on arrival to the labour ward because it is unnecessary and could lead to hyperthermia if kept on neonates not requiring additional heat 110.

Transport using an incubator is inappropriate at Bansang because there is only one incubator available and the ground surface is too uneven to allow safe pushing 136. However, the WHO has suggested a locally produced insulated box with holes in for ventilation could be used for transport and pre-warmed using a hot-water bottle if required 7.

Obstruction to the Clinical Value of Neonatal Temperature

Messaritakis et al has shown that septicaemic neonates show a thermoregulatory reaction, reflected by a widening of the rectal-sole temperature difference of more than 3·5ºC, and Yashiro et al has stated that one major disadvantage of radiant heaters is that they deprive the doctor of this clinical value 40, 41, 140. However, conversely, use of the radiant heater meant that temperature readings were measured routinely, unlike previous practice, and if a neonate's temperature remained abnormally high after birth, the staff were prompted to monitor its temperature at regular intervals and administer antibiotics if need be. This should actually improve diagnosis and treatment of neonatal sepsis at the hospital. Febrile neonates would not receive radiant heat because the lights are only turned on if the rectal temperature is going to drop below 36·5ºC and hence the clinical finding would not be masked. Furthermore, neonates allowed to become chilled following birth are virtually unable to become febrile in response to infection and hence hypothermia is more likely to mask sepsis in neonates than a neonate kept thermally stable on the radiant heater 17.

The Effect of Socio-Cultural Practices on Hypothermia in Neonates at the Bansang Hospital

Warmth during resuscitation and observation is only one of the 10 `warm chain' stages identified by WHO for keeping neonates warm during the first hours of life.

Table 3 illustrates the other 9 7.

Table 3: The WHO Warm Chain 7

Findings from this study support evidence that the drying and wrapping of neonates at birth decreases heat loss, since 8/9 (89%) of the neonates not dried and wrapped properly became hypothermic compared to 6/21 (29%) of the neonates who were adequately dried and wrapped 11. Prior to this study, drying and wrapping was not practiced in Bansang and it required education and persistence to implement it.

Availability of towels was a problem because even if donated they were often stolen by staff or patients, resulting in unsustainability. The 3 neonates delivered at home and admitted later in this study were not dried and wrapped properly and became hypothermic (Neonates 26-28, Figure 3), a practice that can only be changed with parental education possibly through trained birth attendants.

KMC was also not practiced prior to this study and was difficult to implement because close proximity between mother and child was not the cultural norm. In one case a hypothermic neonate was given to the mother to re-warm with KMC, but an hour later when checked had been placed in the cot and had a core temperature of 36ºC; a common scenario. Conversely, in another case a hypothermic neonate was given to the mother for re-warming with KMC, which was administered, and resulted in the neonate's temperature returning to 36·7ºC within 40 minutes. Evidence suggests that with education of staff and persistence with the patients socio-cultural practices can be changed 7, 141. This was demonstrated by the complete transition from before this study when neonates were separated from their mothers for hours after birth, sometimes in different wards, to after the study when all neonates were kept with their mothers and breast- feeding initiated within half an hour of life if possible. This change was reflected in the survey results, with all staff members stating that healthy neonates should be returned to their mothers within minutes of birth and neonates requiring observation and/or resuscitation as soon as they were stable.

THE WARM CHAIN

1 Warm delivery room

2 Immediate drying

3 Skin -to-skin contact

4 Breast-feeding

5 Bathing and weighing

postponed

6 Appropriate clothing/bedding

7 Mother and baby together

8 Warm transportation

9 Warm resuscitation

10 Training and awareness

raising

Before this study, bathing the neonate in a cold bucket of water after birth was common practice (Neonate 29 and 30, Figure 4). This was difficult to change because it is part of the Muslim culture. However, it was advised that bathing should be delayed for the first 12 hours of life or at least the adverse effects minimised by washing the neonate away from draughts, in warm water, drying it immediately afterwards and placing it in KMC 10, 39, 142.

Electricity

In 2/30 cases the electricity failed while a neonate was receiving care on the radiant heater, one of the times being potentially fatal. It must become routine practice for the generator to be turned on in advance before every delivery to power the radiant heater, suction and oxygen machines. This routine can be achieved because it is already practiced for operations at the hospital. A separate generator for the labour ward would be at risk of running out of fuel and not functioning anyway. The staff gave mixed responses when asked what they would do if the generator went off while using the radiant heater, however all answers given were safe and acceptable except putting a hot water bottle next to the neonate which can cause dangerous fluctuations in heat 7. The correct answers were explained.

Could the Design of the Low-Technology Radiant Heater be Improved?

Sides on the Mattress Surface

Having sides on the radiant heater has been proven to decrease air velocity around the neonate and cooling by convection and hence could be useful 40, 143. The sides should be no higher than 4 inches because they could obstruct access to the neonate.

Light Bulbs

Ideally the light bulbs used on the low-technology radiant heater should be a locally manufactured brand to ensure they are sustainable. In this case the light bulbs are sustainable through Anita Smith running the Bansang Hospital Appeal, but this might not be the case for other low- income facilities.

Back-up Source of Power

A low-technology anaesthetic machine has been produced which can still function if the electricity and oxygen fails and a similar design could be considered for a radiant heater 78. However, the machine itself is expensive and not appropriate for this lowtechnology radiant heater.

In-built Suction and Oxygen Machines

This would make the radiant heater too high-technology and liable to break down. It worked well to have the suction machine and oxygen concentrator next to the radiant heater and if any of the two dysfunction they can be replaced separately.

Further Limitations identified during the Study

? The sample size was limited to 30 neonates on the radiant heater and 17 requiring radiant heat, which has resulted in relatively wide confidence intervals. However, the results are still useful as an approximate of the success of the radiant heater and will hopefully guide future research.

? Initially a doctor controlled the radiant heater and then was present to assist the other staff using it, which could have improved their skills in maintaining the neonates within the limits of thermoneutrality, and hence the results might not be truly representative of the staffs' skills in everyday practice.

? Three of the questions in the survey were worded in a way that in use did not give a clear answer. However, I did not want to prompt the staff too much for fear of getting biased results.

Conclusion

Policy Recommendations

The low-technology radiant heater is effective at maintaining neonates within the limits of thermoneutrality and is appropriate for low- income countries. Hence it should be promoted and produced for use in the developing world to help reduce neonatal morbidity and mortality. However, sufficient training is essential to ensure that it is used effectively and safely. All staff members should be trained and examined on how to use the radiant heater and neonatal resuscitation equipment and this should be refreshed every 6months-1year. To ensure the sustainability of this it is advised that two permanent, senior members of staff at the health facility are highly trained and examined in the use of the radiant heater and resuscitation equipment and have literature to consult and hence can ensure continual training and refresher courses for the other staff members.

It is recommended that one low-technology radiant heater is situated in the labour ward and another is situated outside the theatre fo r neonates born by caesarean section. An intervention is required to maintain the neonates' temperature between the theatre and the labour ward and evidence suggests an occlusive plastic wrapping or a locally produced, insulated box could be appropriate and effective for this purpose 46, 138, 139. If replicated the design of the radiant heater could be altered to include small sides around the mattress area to shield the neonate and light bulbs manufactured in the country where the radiant heater will be in use. It is vital that the adverse sociocultural practices such as leaving neonates in a wet cloth, separating the mother and baby, and washing the neonate after birth are not re-administered after this study. The Bansang Hospital would greatly benefit from the introduction of the WHO's `warm chain' policy 7.

Implications for Future Research

The sample size in this study was relatively small and hence a larger study would be useful to substantiate these findings. Further research is required to determine whether this radiant heater would be as effective at maintaining neonates within the limits of

thermoneutrality in colder climates. A follow-up study would be useful to investigate whether the staff are still using and know how to use the radiant heater effectively and safely after 6 months or a year and to check the sustainability of the machine in terms of functionality after this time period.

Acknowledgements

Thanks to Dr Peter McCormick and Mrs Anita Smith from the Bansang Hospital Appeal, Revd Jane Shaw, Dr Nick Emmel and staff at the Nuffield Centre for International Health and Development in Leeds and all staff who participated in the study at the Bansang Hospital in The Gambia.

References

1. United Nations Children's Fund. Neonatal Deaths: 5million each year. DHS and government reports, Nov 2004: http://www.unicef.org/pon97/p31a.htm (Accessed Oct 10, 2004).

2. World Health Organisation (2003). Health of the Newborn, Regional Committee Fifty-sixth Session 10-13 September 2003. World Health Assembly, 2003: http://www.who.int/child-adolescenthealth/New_publications/NEWS/doc56.pdf (Accessed Nov 7, 2004)

3. Bryce J, Victora C. Child survival: countdown to 2015. Lancet 2005; 365:2153-2154.

4. Lawn J E, Cousens S, Zupan J. 4 million neonatal deaths: When? Where? Why? Lancet 2005; 365: 891-900.

5. Darmstadt G L, Bhutta Z A, Cousens S, Adam T, Walker N et al. Evidencebased, cost-effective interventions: how many newborn babies can we save? Lancet 2005; 365: 977-988.

6. Knippenberg R, Lawn J E, Darmstadt G L. Systematic scaling up of neonatal care in countries. Lancet 2005; 365: 977-988.

7. World Health Organisation (1997). Thermal protection of the new born: a practical guide out of print. Geneva. WHO, 1997: http://www.who.int/reproductionhealth/publication/MSM_97_2_Thermal_protection_of_the

_newborn/MSM_97_2_chapter4-I.en.h