Monitoring in-transit temperatures of SARS-CoV-2 PCR test samples mailed through Royal Mail postboxes
Published 27 April 2023
Applies to England
© Crown copyright 2023
This publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. To view this licence, visit nationalarchives.gov.uk/doc/open-government-licence/version/3 or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: psi@nationalarchives.gov.uk.
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
This publication is available at https://www.gov.uk/government/publications/covid-19-monitoring-temperatures-of-sars-cov-2-pcr-samples-via-royal-mail/monitoring-in-transit-temperatures-of-sars-cov-2-pcr-test-samples-mailed-through-royal-mail-postboxes
Introduction
Polymerase chain reaction (PCR) test vials and swab kits and lateral flow devices (LFDs) distributed by the UK Health Security Agency (UKHSA) for testing for the presence or absence of SARS-CoV-2 are highly temperature sensitive and must be stored appropriately. Manufacturer’s guidance stipulates that PCR test vials and swab kits should be stored between 2°C and 25°C, while LFD test kits should be stored between 2°C and 30°C. Stocks of PCR test vials and swab kits and LFDs are stored in temperature-monitored warehouses.
The National Testing Programme validated various PCR combi kits using spike-in virus samples to test how long a test sample could remain in transit. Testing using spike-in virus samples included sample stability validations at 25°C and 35°C to identify any ribonucleic acid (RNA) degradation at higher temperatures to allow for risk based decisions should there be any temperature exceedance in the summer months. Test results were evaluated by the scientific and clinical validation team. The results demonstrated that combi vials that had been spiked with SARS-CoV-2 were stable for up to 240 hours at 35°C with minimal effect on the results obtained in the laboratory.
The high temperatures experienced in the summer of 2022, and the subsequent period of extreme hot weather over a number of days with temperatures exceeding 40°C, resulted in a number of temperature excursions reported by the warehouses holding PCR and LFD stocks. Prolonged exposure to both the high temperatures and the daily cycling of temperatures during the transit of PCR test samples to laboratories is likely to affect the nucleic acid integrity of the samples and therefore the sensitivity of diagnostic PCR testing. This highlighted the need to understand the temperatures and cycle of temperature changes test samples are subjected to when they are returned via postal system by subjects and to understand the impact of raised temperatures on sample integrity.
A rapid literature search at the time the study was conducted did not identify any large peer-reviewed literature on monitoring of temperatures within mailboxes or the impact of temperatures within post and mailboxes on postal tests, barring one conducted by researchers in Phoenix, Arizona, US in 2005 (Robbins and colleagues, 2006). While evaluating the effect of heat on Formoterol capsules (drug used for treatment of asthma, chronic obstructive pulmonary disease (COPD) and other airways-related breathing problems), the researchers identified that temperatures within mailboxes in mid-summer in Arizona exceeded 70°C when the outside temperature was 43°C. When inquired, Royal Mail did not have any information on any postbox temperature monitoring studies that may have been conducted by them.
Given the absence of published reports of temperature excursions, a study was undertaken to monitor the temperatures reached within the standard packaging used for return of PCR test samples during transit when tests are returned via post by subjects. The SARS-CoV-2 samples were returned using a UN3373 specified cardboard box (200mm x 140m x 20mm). Two pre-calibrated data loggers capable of continuously recording temperatures every 5 minutes were packed into standard test return packaging used for returning test samples and deposited into the mailbox to enable the recording of temperatures within mailbox and during transit until the samples reach the labs for testing. The time at which the loggers were posted, and the last time of collection were recorded so the sample journey could be mapped.
Based on the temperatures reached during the summer peak, a bespoke series of studies were undertaken to investigate the effect of high temperatures on PCR sample stability, at the Rosalind Franklin Laboratory (RFL) in Leamington Spa. These studies were designed to include potential additional scenarios of raised temperatures and impact of longer periods of raised temperatures should the samples be posted on a weekend when no collection was available.
Methods
Monitoring temperatures within Royal Mail postboxes
Two pre-calibrated temperature data loggers set up to record temperatures every 5 minutes were packed into standard postal test return boxes, affixed with second class stamps and addressed to UKHSA labs (such as RFL). They were deposited into postboxes on the same day (Friday 12 August 2022) from 2 locations: one each in East Midlands and West Midlands that have average summer temperatures were chosen. One data logger was deposited into a postbox in Rugby at 11:22am and the second data logger was deposited into a postbox in Lutterworth at 10:54am (Figure 1). The test sample boxes containing the data loggers were received at RFL on 16 August 2022 at 11:55pm. Data was extracted from the devices once they were returned to the Supply Chain and Logistics team.
Figure 1a. Packaging and mailing of dataloggers to Rosalind Franklin Laboratory. Data logger 1 (serial number 65039260) deposited into a postbox in Rugby
Figure 1b. Packaging and mailing of dataloggers to Rosalind Franklin Laboratory. Data logger 2 (serial number 65039603) deposited into a postbox in Lutterworth
Effect of prolongued exposure to raised temperature on PCR sample stability
One hundred and four PCR sample collection tubes were spiked with ribonuclease P (RNAseP) material from anonymised volunteer buccal swabs and a 2000 copies/mL of SARS-CoV-2 (SCV2AQP01-S01, SARS-CoV-2 Analytical Q Panel 01, Qnostics) to mimic sample collection from individuals. These were then incubated at 40°C, 45°C, 50°C, and 55°C using 4 ThermoMixers™ while control was set to standard laboratory temperature (19°C). PCR samples (n=8) for each temperature point was retrieved at 24 hours, 48 hours and 72 hours and processed by quantitative polymerase chain reaction (qPCR) using a LightCycler® 480 System along with n=8 t=0h control. Any sample returning a negative SARS-CoV-2 test result was classed as a sample dropout due to lack of assay detection.
Results
Temperatures within Royal Mail postboxes
The transit temperature data from the 2 data loggers were extracted and this is shown in Figures 2 and 3. Given the data loggers were posted on a Friday (12 August), they were not returned to the lab until the following Tuesday (16 August) following collection on Saturday. The loggers remained within the postal transit system until their delivery. The time of the last collection on Saturday (red vertical line) is indicated, along with periods where the temperatures were above the excursion temperature of 25°C (amber) and temperatures beyond 35°C (red).
Figure 2. Plot showing the transit temperatures recorded by data logger posted in Rugby until it was delivered to the lab
The data logger posted from Rugby shows that it was subject to 2 rounds of temperature excursions above 25°C (amber). Temperature exceeded 25°C within 5 minutes of the data logger being deposited in the postbox and a maximum temperature of 38°C was recorded around midday on 12 August 2022 as shown in Figure 2 (highlighted in orange). The second round of excursion likely occurred post scheduled collection at 7am on Saturday, 13 August 2022 (red line), and the temperatures dropped to below 25°C around 9:15am on 13 August 2022 and remained below 25°C until the samples were delivered to RFL. The temperatures recorded in Rugby show that temperatures reached 34°C at midday and the highest temperature of 36°C was recorded at 3pm local time.
The data logger posted from Lutterworth was subjected to temperature above 25°C within 10 minutes of being deposited into the postbox (amber), temperatures reaching a maximum of 32°C around midday on 12 August 2022 as shown in Figure 3. Temperatures remained above 25°C until 9:15am on 13 August 2022 after which they dropped below 25°C. The temperature remained under 25°C post scheduled collection (11am on Saturday, 13 August 2022 marked by red line) until the samples are delivered to RFL. The temperatures recorded in Lutterworth show that temperatures reached 30°C at midday and the highest temperature of 36°C was recorded from 2pm to 5pm local time.
Figure 3. Plot showing the transit temperatures recorded by data logger posted in Lutterworth until it was delivered to the lab
Differential temperature was calculated for the period covering the time within the postboxes. The differential calculation suggests considerable difference in the temperatures observed within the postboxes (Figure 4). While the temperature within the postbox remains warmer compared to the ambient temperature during the day, it declines slowly through the evening, resulting in a higher average differential.
Figure 4. Plot showing calculated differential temperatures and the average temperature differential observed within the postbox compared to the outside ambient temperature.
Stability of PCR samples upon prolonged exposure to raised temperatures
All samples incubated at raised temperatures were processed along with control samples. For control samples, an average value of 32.94 cycle threshold (Ct) was observed for the SARS-CoV-2 target, and an average value of 32.94 Ct was observed for the RNaseP target. The average Ct values for SARS-CoV-2 and RNaseP samples observed at elevated temperatures are presented relative to that observed for the control sample.
Samples from T=40°C and 45°C did not have any sample dropout across all 3 time points for the SARS-CoV-2 target and had lower Ct values at all time points compared to the control. (Table 1). Samples from T=50°C and 55°C had both higher Ct values and sample dropouts across all 3 time points for the SARS-CoV-2 target except for the 24-hour time point for T=50°C, which had slightly lower Ct compared to the control. Samples at 55°C had high dropout rates, suggesting increasing sample degradation at this temperature.
Table 1. Summary table for SARS-CoV-2 target for each time point and temperatures
| Temperature (°C) | 19 | 40 | 40 | 40 | 45 | 45 | 45 | 50 | 50 | 50 | 55 | 55 | 55 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time (hours) | 0 | 24 | 48 | 72 | 24 | 48 | 72 | 24 | 48 | 72 | 24 | 48 | 72 |
| Mean Ct | 32.94 | 30.81 | 31.06 | 31.45 | 31.09 | 32.47 | 31.90 | 32.67 | 33.57 | 34.04 | 34.13 | 34.93 | 34.72 |
| Standard deviation (Ct) | 1.04 | 0.71 | 0.84 | 1.08 | 0.85 | 0.83 | 0.76 | 1.10 | 1.77 | 0.95 | 0.91 | 1.62 | 0.85 |
| Dropout (%) | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 2 | 4.08 | 6.67 | 14 | 57.14 | 66 |
Table 2. Summary table for RNaseP target for each time point and temperatures
| Temperature (°C) | 19 | 40 | 40 | 40 | 45 | 45 | 45 | 50 | 50 | 50 | 55 | 55 | 55 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time (hours) | 0 | 24 | 48 | 72 | 24 | 48 | 72 | 24 | 48 | 72 | 24 | 48 | 72 |
| Mean Ct | 30.84 | 30.62 | 31.46 | 31.36 | 31.09 | 29.94 | 28.68 | 32.05 | 30.66 | 30.00 | 31.40 | 31.13 | 31.08 |
| Standard deviation (Ct) | 0.83 | 2.39 | 2.86 | 2.74 | 1.65 | 1.92 | 1.81 | 2.31 | 2.60 | 1.21 | 1.92 | 2.66 | 1.37 |
| Dropout (%) | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Figure 5. Box plot distribution of mean Ct values for each time point and temperatures for the SARS-CoV-2 target
No dropouts were observed for RNaseP targets as all samples returned a positive test result. Samples from T=40°C returned a lower average Ct value compared to control at the 24 hour time point, while the average Ct was higher at the 48 hours and 72 hours time points (Table 2). Samples at T=45°C, 50°C had higher average Ct values at 24 hour compared to control, but had lower averages at 48 hours and 72 hours. Samples for T=55°C had higher Ct values compared to control at all time points. Interestingly, for T=45°C, 50°C and 55°C, longer exposure to higher temperature resulted in a lower RNaseP CT value, average observed Ct value at 72 hours was lower than that at 48 hours, which in turn was lower than that at 24 hours.
Figure 6. Boxplot distribution of mean Ct values for each time point and temperatures for the RNaseP target
Boxplot showing the distribution of the average Ct values for SARS-CoV-2 and RNaseP targets for T= 40°C, 45°C, 50°C and 55°C at t= 24 hours, 48 hours and 72 hours are shown in Figure 5 and Figure 6 respectively. Each box displays the first and third quartile range (Q1 and Q3) while the line through the box displays the median value; the entire box is defined as the interquartile range (IQR = Q3 to Q1). The dotted line across the plot indicates the Ct value observed for the control sample at 19°C. The whiskers extending out of the box are the furthest data points within 1.5 times the interquartile range.
Conclusion
The data collected from the 2 data loggers indicates slight variations in the relationship between external temperature measurements and the temperature measurements taken inside the postboxes. The temperatures inside the postboxes peaked after midday and were found to be higher than the local atmospheric temperatures. Based on differential temperature calculations, the temperature inside the postboxes was estimated to be 8°C to 10°C warmer than the ambient temperature through the day. In the evening, while external temperatures dropped rapidly below 20°C, the temperature inside the postbox decreased slowly and remained well above this range.
The temperature excursion observed in the data logger posted in Rugby occurred after scheduled last collection time, suggesting that this happened during initial transit. No such excursion was observed for the data logger posted from Lutterworth. The temperatures remained below 20°C until the packages were delivered to RFL.
While additional data is required to draw more robust conclusions, the evidence suggests that the relationship between external and internal temperature measurements varies slightly within the postboxes. During hot periods (greater than 35°C), the temperature inside the postboxes exceeds the external temperature, and the temperature decline inside the postboxes is slower than outside. It is likely that when the outside ambient temperature is around 25°C, the temperature within the postboxes would remain around or under this observed temperature. The temperature post-collection is influenced by the mode of transportation and is likely to remain below 25°C.
Evaluation of sample stability in the labs suggests that SARS-CoV-2 positive can degrade when exposed to temperatures above 45°C. The degree of degradation depends on the duration and the temperature that the samples are exposed to. While the study did not evaluate how cycling of the temperatures would affect sample stability, continuous exposure to temperatures above 45°C results in increased Ct values with time and higher dropout rates, particularly at temperatures over 55°C, suggesting increased sample (viral RNA) degradation (or SARS-CoV-2 targets detected by PCR). On the contrary, no dropouts were observed for RNAseP across any of the temperatures over the 72-hour period, suggesting that exposure to these raised temperatures does not affect the integrity of the target binding points for RNaseP.
The laboratory studies clearly suggest that there is an increased likelihood of false negative PCR test results, due to degradation of SARS-CoV-2 RNA. SARS-CoV-2 positive samples with a low viral load (high Ct value) are likely to be impacted should they be exposed to elevated temperatures for longer periods.
Further studies are required to assess the impact of temperature cycling and exposure to raised temperatures on PCR test samples. The findings from this study should be considered while delivering future testing for detection or surveillance of any pathogen, particularly when samples are intended to be returned via postal system. Changing global weather conditions warrant a more robust evaluation of temperatures reached within postboxes and the impact of such temperature excursions on various postal testing regimes.
Limitations
The major limitation of the study is that there are only 2 data sets monitoring the temperature within postboxes during warm weather. A wider assessment of temperatures within postboxes during extreme weather is required given the different types of postboxes that exist in the UK, their locations, or exposure to direct sunlight.