Table 2.

Summary of key findings and limitations of the studies included in the systematic review to assess how the use of vaccines outside the cold chain or in controlled temperature chain contributes to improving immunization coverage in low- and middle-income countries (LMICs)

Author(s) Year	Relevant key findings	Study limitations.
Zipursky et al., 2011 [15]	The use of a CTC-based strategy allowed for vaccination to happen in areas with unreliable cold chain proffers logistical and operational advantages including allowing health workers to conduct home vaccination anytime. Health Workers were comfortable, but the parents were not. Ambient temperature exposure for the vials reached a maximum of 47.1°C. A total of 6 vials reached the VVM discard point at end of day 2.	Vaccines used in the study were from the same batch and manufacturer, and the study did not allow for a detailed correlation to be drawn between the length of time and temperature exposure and vaccine potency reached and VVMs.
Lee et al., 2012 [16]	Making vaccines thermostable had positive effects by reducing supply chain bottlenecks and increase the availability of all EPI vaccines and decreased cold chain space utilization. Thermostable pentavalent had the highest positive effect with its availability increasing from 87% to 97%, and the availability of other non-thermostable EPI vaccines increased to over 93%.	A study is a Model that cannot capture every detail of reality. Assumptions that the physical characteristics of the vaccine remain the same when made thermostable.
Shrivastava, et al., 2012 [17]	The 10 test vials reached the discard point in 43 hours at 37°C storage, 24 hours at 41°C storage, 16 hours at 45°C, and 9 hours at 49.5°C storage temperatures. Findings suggest VVMs are not reliable indicators of vaccine potency at high environmental temperatures.	The study did not conduct a laboratory test for vaccine potency.
Juan-Giner et al., 2014 [18]	Following vaccination, overall seroprotection was the same in both groups; 99.34% in the CTC and 99.45% in the SCC groups. Few adverse events were noted. Thus, the study demonstrated immunogenicity and safety of TT vaccines in CTC at <40°C for <30 days. Maximum recorded ambient temperature of 43.1°C with no damage to vaccines from heat exposure.	Only vaccines from one manufacturer were used for the study – Serum Institute India.
Lydon et al., 2014 [19]	The study demonstrated large cost savings that could be obtained when vaccines are kept at CTC more especially from district-level storage down to service delivery points.	Used a single scenario in the model, could not estimate additional costs of CTC, and data not generalizable.
Steffen, et al., 2014 [20]	Incidence rates of AEFIs in the CTC group were the same or less than those in the non-CTC group (No hospitalizations record). Ambient temperatures ranged from 19°C to 46°C during the vaccination period.	Non-randomization and a non-representative population sample.
Zipursky, et al., 2014 [21]	An overall vaccination coverage of 105.7% was achieved (155 596 people vaccinated) with vaccines stored under three different CTC scenarios. All vaccinators and 98% of supervisors' preference to use CTC for the next campaign. No VVMs reached the discard point, and no temperature reading up to 40°C. Minimal challenge with CTC.	The sample was not representative of the population.
Kolwaite et al., 2016 [22]	A 27% median increase (interquartile range [IQR] 58%, P < 0.0001) in HepB-BD coverage in the intervention districts, compared with a 0% median change (IQR 25%, P = 0.03) in comparison districts. No adverse reactions were reported. Median temperature exposure for the vaccine was 27°C in the intervention group and 4.6°C in the control group.	Several health facilities were not enrolled due to access issues and some selected villages only had children from one age group, which made a comparison impossible.
Breakwell, et al., 2017 [23]	Timely HepB-BD vaccination coverage increased from 30% (n = 38/125) to 68% (n = 104/152) (P = 0.0005) and from 4% (n = 2/46) to 24% (n = 9/38) among facility and home births respectively. Additionally, BCG 24-h coverage increased from 15% (n = 19/125) to 28%. By 42 d post-birth, HepB-BD coverage had reached 80% (n = 121/152). Rarely temperatures exceeded 37°C, but vaccine wastage was high and shortages common. Where home births are common, an outside cold chain policy could improve birth dose coverage.	Some health facilities had very low births and sample not representative of the total population.
Landoh et al., 2017 [24]	No statistical differences in vaccination coverage between CTC and non-CTC areas (AOR = 0.09; 95%CI = -0.27, 0.45). The overall vaccination rate was 98% of the surveyed population. Mild to moderate AEFI in 2.3% following vaccine administration.	A non-representative sample of the regions was taken.
Lee et al., 2017 [25]	Replacing a particular vaccine with a thermostable version yielded cost savings in many cases even when charging a price premium. For instance, replacing the current pentavalent vaccine with a thermostable version with or without increasing the vaccine price was cost-saving (US$366 to US$10 945 per 100 members of the vaccine's target population). Cost savings observed even when vaccine prices were doubled or tripled.	The study is a model and may not capture all real-life scenarios. Model assumed vaccine will maintain the same character when it becomes thermostable.
Mvundura et al., 2017 [26]	The cost of logistics per dose administered was not statistically different between CTC and standard storage. There is a possibility of increased cost per dose if the facilities without refrigerators had not used a CTC. The analysis showed that the strongest case for CTC use is for remote health centers without cold chain equipment.	Possibly underestimated the cold chain costs because the costs at the regional level were not included in the analysis. Transport costs may have been under-budgeted.
Coldiron et al., 2018 [27]	Children who received the RotaSIIL vaccine had similar safety outcomes compared to placebo. Only one case of intussusception reported.107 deaths split near half among the two groups. SAEs occurred in 814 participants, 395 (19.3%) RotaSIIL and 419 (20.5%) control. A total of 7385 child-years of follow-up.	The complexity of making an accurate diagnosis of intussusception in remote settings.

SAE – serious adverse effects, CTC – controlled temperature chain, AEFI – adverse events following immunization, HepB-BD – hepatitis B birth dose, BCG – bacillus Calmette Guerin, CI – confidence interval, AOR – adjusted odds ratio, EPI – expanded programme on immunization, VVM – vaccine vial monitor, SCC – standard cold chain storage, TT – tetanus toxoid, IQR – interquartile range