We tailored the insulating jacket through consulting designers and a textile manufacturer, Poeticgem International Ltd. Furthermore, the quality of fabrics of the insulating jacket had been checked by Intertek Textile Ltd. in Bangladesh. All the laboratory testings of the CWPs and the insulating jackets were carried out in icddr, b and Material Chemistry Research Laboratory under the Department of Chemistry of the University of Dhaka.
Phase 1
Design and development of chemical warming pad (CWP)
In step 1, we chose a phase-change material, Sodium Acetate Trihydrate (SAT)23 for the CWP as the primary source of heat generation due to its high latent heat of fusion (250 kJ/kg) (energy released or a thermodynamic system during a constant-temperature process), extensive cooling range, non-hazardous property, availability and low-cost23,24,25. For thermoregulatory devices, SAT is vastly used for long-term thermal energy storage in a metastable supercooled liquid state24 at low ambient temperature. It releases heat on demand by initiating the crystallisation25 using metal triggers in the aqueous solution. Despite such thermo-physical properties of SAT, there are a few challenges associated, including hard, lumped, big, and sharp edges of SAT crystallites, unintentional nucleation in a metastable supercool liquid state, and fast crystal growth during solidification which causes low retention time for our desired temperature26,27,28. Moreover, when the number of thermal cycles increases, the latent heat capacity of SAT also decreases25. This poor thermal cyclic stability limits its capacity to maintain our desired temperature range.
In step 2, to overcome the limitation of the aquous SAT, we experimented on a few binary composites to land up on the solution with improved thermal cyclic stability. In the experiments, Graphite29, Carboxyl Methyl Cellulose30, Sodium Chloride31, Glycerol31, Ethylene Glycol32, and Paraffin33 were used as additives with the aqueous SAT solution but none could meet the study objective, as uncontrolled nucleation, low heat retention, and attainment of high temperature than desired were observed (Supplementary Table 1).
In step 3, guided by the evidence, glycerol and paraffin compositions had the potentials to achive the desired temperature range and heat retention duration. Therefore, we tested a ternary composite of Glycerol and liquid Paraffin with an aqueous SAT solution through a eutectic system (a homogenous mixture of materials that melts or solidifies at a single temperature)34 approach to obtain the desired latent heat capacity with desired temperature and retention time finally. As the crystallisation temperature and heat retention, both are high for the SAT aqueous solution and binary combination with paraffin, we used varying amounts of glycerol to the binary mixture and tested its heat retention duration in the desired temperature (36–38°C) range. Various mass fractions of glycerol (5%, 15%, 20%, 27%, and 35%) mixed with the binary solution were tried out while crystallisation temperature with retention time was observed.
Design and development of insulating jacket
The selection of the jacket fabrics, and design followed a meticulous process. Expert consultations with apperal designer, safety and quality check officials of the ready-made garments (RMG) industry, public health researchers, and neonatologists were taken place in several rounds, to select the most suitable and comfortable fabrics for the jacket. Moreover, safety, availability, and affordability of the fabrics were also considered during the selection and design of the insultaing jacket.
Through iterative process, we designed a five-layered insulating jacket comprising single jersey, polyester-polar fleece, Taffeta, screemline padding, and water-resistant polyester, from inside to outside and polyester ripstop for the kangaroo pocket to keep the CWP inside the jacket (Supplementary Fig. 1). These fabrics have a high insulating capacity and low-to-moderate air permeability that protect against loss of heat caused by convection, conduction, evaporation and radiation (Supplementary Table 2). Moreover, the fabrics are lightweight, and the outer layer and the fabric of the kangaroo pocket are water-resistant.
The fabric of kangaroo pocket that contains the CWP does not hold any generated heat to itself and therefore, the neonates put inside the jacket will receive the total heat generated by the CWP. The length of the pocket covers the maximum body surface area of the neonate from neck to buttock. Again, the inner body touching layer of the jacket is jersey cotton, followed by fleece, wadding wrapped in taffeta and the outer layer is waterproof polyester. Each fabric serves distinct purpose of the jacket. For example, cotton provides softness, fleece helps to keep the warmth inside by creating an insulating environment. Similarly, the third layer bolsters the first two layers by providing more comfort and restriction of heat transport. Finally, the outermost layer prohibits air and water passing from outside and thus ensures overall protection for the baby. We used soft Velcro on the side and bottom edges of the jacket to make sure the jacket does not get unwrapped during movement. Moreover, the jacket is designed in such a way that it covers a neonate from head to toe given that except from the body cover, it also has a hoodie to support warmth conservation of the head surface area.
Phase 2
Laboratory testing of thermal jacket (CWP and insulating jacket)
The CWP and insulating jacket together were tested on a mannequin in a controlled laboratory environment at icddr, b.
Sample size Estimation for laboratory testing of the thermal jacket
Our objectives were to determine whether the thermal jacket attains and maintains a temperature ranging between 36 and 38°C for 120 minutes and the effect of ambient temperature on the thermal jacket’s performance. Moreover, we were also interested in determining the effect of the reusability of a CWP on the insulating jackets’ performance. For this, we calculated the required number of events for capturing a 95% success rate for the CWPs as well as the insulating jackets using the below formula:
$$\:n=\:\frac{{Z}^{2}P(1-P)}{{d}^{2}}\:$$
Considering,
\(\:Z\) is the Z-score corresponding to 95% confidence level = 1.96, \(\:\:d\) is the margin of error = 0.05, \(\:P\) is the rate of success = 95%, the calculated sample size was 73 events. The definition of an event was to put the mannequin inside the thermal jacket and observe the temperature of the CWP and insulating jacket simultaneously for 120 minutes. Since, the events were correlated within the CWP, multiplying the sample size by a design effect of 1.08 considering the intra-cluster correlation coefficient 0.01, the minimum required events became 73 × 1.08 \(\:\approx\:\) 79. Since each CWP was planned to reuse nine times due to its functional performance limit to maintain the temperature range 36–38⁰C for 120 minutes, we needed nine CWPs to observe the required number of events. Hence, the CWPs and insulating jacket were tested 81 times with a mannequin to assess their performance to attain and maintain the desired temperature range (36–38⁰C) for 2 hours.
Sampling technique for the laboratory testing of the thermal jacket
For assessing the effect of ambient temperature on the performance of the CWP and the insulating jacket, we artificially created room temperature ranging from 18–34⁰C. Three temperature bands were created within the range, including 18–24⁰C, 25–29⁰C, and 30–34 ⁰C. Furthermore, we tested 81 sample events in each temperature band more than once. The detailed description is provided in the supplementary Table 3.
Temperature measurement of the CWP and insulating jacket
To test the thermal jacket in the laboratory, we have devised LabQuest to measure the real-time temperature of different surfaces of the CWP and insulating jacket (Fig. 2). The first sensor was attached to the surface of the CWP, the second one was on the surface of kangaroo pocket of the insulating jacket, and the third one was kept outside of the jacket to measure ambient temperature. One additional sensor was also used to measure the humidity of the room, separately. Once all the sensors were placed correctly, we then flexed the metal trigger inside of the CWP and observed the initiation of crystalisation. Once the liquid starts to solidify, we then put the CWP in the kangaroo pocket of the insulating jacket instantly. A mannequin was then placed on the kangaroo pocket immediately and wrapped inside the the jacket.
Temperature measurement process of the CWP and insulating jacket.
Once the jacket is prepared, we started recording the time to measure the temperature and humidity. The data points were collected on an interval of every 6 seconds. Therefore, for every one minute, we got 10 data points. Though, our objective was to observe if the thermal jacket can attain and maintain the temperature range between 36⁰C and 38⁰C for 120 minutes, but we recorded the performance for 130 minutes, so that we could observe if the jacket has potential to perform beyond 120 minutes and give caregivers enough time to put the neonates off the jacket in a real-life setting. We performed all the events in the same manner.
Data analysis
We used the data analysis software ‘R’, version 4.4.1 (https://www.r-project.org/)35 to compute the result. The temperature was measured in degree celsius and the time was in minutes. The simple mean and standard deviation had been estimated both for the CWP and insulating jacket, seperately. The Welch’s t-test had also been performed to see any potential differences in the performances between the CWPs and the insulating jacket. Moreover, these performances had been assessed by ambient temperature, humidity, number of times CWP were reused, and number of CWPs were used with simple mean and standard deviation. Analysis of Variance (ANOVA) had also been performed to test whether there is any statistical differences between the mentioned factors.
An event had been labelled as successful if it attained a temperature of 36°C within three minutes (maximum) from the baseline point and then maintained a temperature range between 36°C and 38°C for 120 minutes. If the temperature exceeded or dropped from the above range at any time point and returned to the range with a rate of 0.5°C/hour or more, we considered it as a successful event. The percentage of events that successfully attained and maintained the temperature range 36–38⁰C for 120 minutes was then calculated and the success rate was disaggregated by ambient temperature, number of times CWP were reused, and number of CWPs were used to see whether there was any significant association between the success status of an event and these factors.
To estimate the adjusted effect of the above factors on the success status of an event, an extended version of the generalized estimating equation (GEE) known as ‘geefirth’ model was performed36. Since each of the CWPs was reused several times, the panel data-based method ‘geefirth’ model was then used to adjust correlation within the CWPs while estimating the effect of ambient temperature bands, average humidity, and number of times CWP were reused on the success of an event. This model also reduces bias due to the small sample size, as well as resolves the separation issue (that occurs when covariates perfectly [or near perfectly] predict the outcome) arising from the skewed distribution of the success rate of the events36.