Studies aimed at investigating respiratory illness and the microbiology of those illnesses in remote communities have to date been limited by the inability to store and transport clinical specimens requiring freezing/refrigeration to urban laboratories. Our study has demonstrated that freezing of specimens for viral studies at the time of collection is not required,  however the detection of common respiratory bacteria is affected, with lower detection rates in mailed specimens.
The reasons for our findings are not clear, particularly given similar divergence was not identified for viruses . Overall, it is evident that there is at least a moderate loss in respiratory bacteria PCR sensitivity using the mailed samples compared to those that were transported frozen. This is further supported by the fact that the human DNA concentrations in the mailed samples were significantly lower (as indicated by ERV3 PCR) in the mailed samples, suggesting some DNA loss, possibly by degradation, in the posted samples. However, the amount of DNA loss that was observed (ie. to increase ERV3 Ct values by 1.6 cycles) cannot fully explain the lack of concordance between the paired samples, particularly the fact that some individuals had a very high load of a particular bacterial species in one nostril, but were negative for the same species in the other nostril. Focal bacterial colonies may be present in specific individuals but these would be expected to be randomly distributed between the left and right nostrils in the population as a whole. Furthermore for focal colonies to be a plausible explanation for our finding of higher sensitivity in frozen swabs (right nare swabs), this would need to assume that bacteria were more likely to colonise the right rather than the left nare. There are no data to support this assumption and we are currently testing this assumption by evaluating transport methods for swabs from taken from the same nare. Alternatively, given the overlap in confidence intervals for the sensitivities for each organism by transport method (Table 2), our findings may be a factor of inadequate sample size to compare differences in common respiratory bacteria.
We were unable to find similar studies in the literature with which to compare our findings. Our study was based on the assumption that there would be minimal, if any, differences in presence and load of viral and bacterial pathogens in the left and right nostrils of individuals, and appeared valid on the basis of our findings with respect to viruses. However, we now postulate that this is not that case and that higher concordance between transport methods may have been observed for the bacterial pathogens had two swabs from the same nostril been used. As above, a further study is in progress to address this issue.
Our findings need to be considered in light of the feasibility and cost of respiratory pathogen surveillance in remote areas where compromises in sensitivity may be required in order for research to occur . Overall sensitivities for any bacteria detected were acceptable in that context, particularly when the collection of specimens for research purposes does not influence clinical management decisions at the time the specimen was collected. Furthermore, the ability to collect one specimen per participant that can be readily shipped and analysed for both viruses and bacteria is important given the increasing emphasis on virus-bacteria interaction in acute and chronic respiratory disease [18–21]. Ideally, nasal swabs are best stored frozen within a short-time frame at ultra low temperatures and shipping from a remote community using nitrogen shippers or dry ice shippers as the gold standard would be preferred. However, this is not logistically feasible in these settings, particularly when research staff are in the community for extended periods, there are limited facilities to maintain dry ice for these timespans and, air freight in and out of the community only occurs once a week. Similarly, future studies should consider the use of concurrent collection from both nares to maximise sampling efficiency.
Our choice of storage media (Virocult, MW950, Medical Wire & Equipment, England) was also based on logistics, cost and that bacterial culture would not be performed. The medium does not need to be stored refrigerated or frozen prior to use, unlike media commonly used for bacterial culture, and there is no risk of breakage during transport, an important consideration if postal services are used. Furthermore, there is no reason to believe that the media itself would affect PCR detection of bacterial DNA as indicated by our high overall positivity rates that are consistent with other studies in this population. It should also be noted that there have been some reports of false-positive results using the pneumococcal ply PCR used in this study,  and so the results here may represent a slight overestimation of carriage.