Use of Filter Paper as a Transport Medium for Laboratory Diagnosis of Cholera under Field Conditions

TEXT

Cholera outbreak confirmation and isolate characterization are critical to set up an appropriate response. However, cholera outbreaks frequently occur in remote areas with limited access to adequate microbiology laboratories. Samples need to be sent to reference laboratories which are sometimes located outside the country, with shipment times reaching a few days. The only recommended transport medium for Vibrio cholerae according to the World Health Organization (WHO) is Cary-Blair transport medium (3, 4). Stuart's (9) and Amie's (1) transport media are inferior to Cary-Blair medium for transport of V. cholerae, and alkaline peptone water (APW) is recommended for use only when Cary-Blair medium is not available and subculture can be done within 6 h of collection (4).

The use of strips of blotting paper in sealed plastic envelopes for stool transportation for cholera testing was first developed in the late 1960s, following the satisfactory use of this alternative mode of transportation for the isolation of Salmonella and Shigella spp. (2, 8). Despite demonstrations of its efficiency, this mode of transportation has not been added to the list of WHO-recommended transport media for the isolation of V. cholerae, possibly due to the difficulties involved in packing the samples in plastic envelopes and ensuring that they remain sufficiently moist during shipment. In the early 1990s, the Institut Pasteur and the nongovernmental medical organization Médecins Sans Frontières (MSF) improved this alternative mode of transportation by standardizing the packaging (7) and modifying the procedure used. This simple method has now been used by MSF for over 20 years and has made it possible to recover and to characterize samples from various locations around the world.

We compared this improved filter paper transport method with the method of using Cary-Blair medium for the recovery of V. cholerae from stool samples. The findings reported here relate to a substudy in the evaluation of a rapid test for the diagnosis of cholera during an outbreak in Lubumbashi, Democratic Republic of Congo (DRC), which will be published elsewhere. The study took place in two Ministry of Health cholera treatment centers (CTC) supported by MSF in Lubumbashi. Ethical approval was obtained from the Ethics Committee of the Ecole de Santé Publique, Kinshasa, DRC, and the Comité de Protection des Personnes, Ile de France XI, France.

Patients presenting to the CTC were included in the study if they were over 5 years of age, if they had acute watery diarrhea with or without vomiting, and if they or their guardian signed written informed consent. Exclusion criteria were ingestion of antibiotics in the previous 7 days and absence of stool during the observation period. The sample size required for McNemar statistical comparison was initially estimated as 148 specimens based on the following hypotheses and parameters: 10% difference in culture results, 20% discordant pairs, an alpha risk of 5%, and study power of 80%.

Samples were prepared in Cary-Blair medium according to the manufacturer's recommendations (Copan Diagnostics, Italy), i.e., the swab was dipped into the nonchlorinated liquid stool specimen and immersed in the transport medium. For transportation on filter paper, a blotting paper disc (commercially available nonimpregnated paper discs [6.35 mm in diameter, not sterilized]; Bio-Rad, Marnes-la-Coquette, France) was dipped into the nonchlorinated liquid stool and placed in a screw-cap microtube. A few drops (2 or 3, around 200 μl) of normal saline solution (0.9% NaCl) were added to stop the sample from drying out, as our experience in practice and published data (2, 8) showed that the maintenance of moisture is essential for the recovery of strains (Fig. 1). The tube was hermetically closed with Parafilm. Both transport media were stored at room temperature and sent weekly to the Institut Pasteur, Paris, France, following International Air Transport Association (IATA) regulations for infectious substances. The liquid supernatant from the filter paper tube or the swab from the Cary-Blair medium was plated directly on appropriate medium. The paper disc or the swab was transferred directly to hyperalkaline peptone water for enrichment. Isolation of choleragenic vibrios (Vibrio cholerae serogroup O1 or O139) was performed by culture following enrichment steps according to standard methods (5).

• Open in new tab
• Download powerpoint

Fig. 1.

Device/method used by MSF for the transport to the laboratory of stool specimens for research of Vibrio cholerae. Filter paper was standardized by using commercialized nonimpregnated discs (Bio-Rad, Marnes-la-Coquette, France). (Copyright Institut Pasteur.)

The study started on 2 March 2008 and ended on 2 May 2008. In total, 296 patients were included in the general study (i.e., the evaluation of a rapid diagnostic test), 234 of whom were included in the substudy comparing the filter paper and Cary-Blair medium methods for stool transport. The patients included in this substudy had a median age of 29 years (interquartile range [IQR], 19 to 40 years), and the ratio of males to females was 1.3; these values were similar to those for the patients in the general study. Dehydration was assessed on the basis of clinical signs at admission according to WHO guidelines (10) and was considered severe in 33 (14%) of the patients included, moderate in 54 patients (23%), and absent in 147 patients (63%). Among the first 34 samples sent both on filter paper and in Cary-Blair medium, 20 tested positive for V. cholerae O1 when cultured from filter paper, while only 10 tested positive when cultured from Cary-Blair medium. This difference was unexpected, and we thought that it might be an anomaly. We therefore decided to exclude these 34 samples from the study and to use a new, single batch of Cary-Blair medium for subsequent comparisons. This resulted in a final sample size of 200 for the comparison of filter paper and Cary-Blair medium (Table 1). The median delay between sample collection and culture in Paris was 13 days (range, 7 to 17 days) for both Cary-Blair medium and filter paper. There was a good correlation, with a kappa value of 0.76 (95% confidence interval [CI], 0.62 to 0.89), and no statistically significant difference (McNemar's exact χ² test P value = 0.10) between the two transport media. The proportion of positive samples was still high even when culture was performed more than 2 weeks after collection, with no statistically significant differences in proportions of positive samples between transport media (Table 2). These results show that the filter paper method is as effective as use of the recommended Cary-Blair medium for specimen transport for culture in field conditions, confirming the results obtained during earlier studies (6, 8).

We added a small volume of saline to the specimen. This was not done in previous studies and was found to be sufficient to prevent dehydration problems, even for shipment periods of up to 17 days. Moreover, the filter paper method has several advantages over use of Cary-Blair medium. Filter paper and normal saline are easy to find and inexpensive and can be stored for long periods at room temperature, in contrast to Cary-Blair medium, which requires storage at a temperature less than 25°C or even refrigeration (2 to 8°C), according to the manufacturer, and has a relatively short shelf life. In addition, as observed in this study, there appears to be performance variation between batches of Cary-Blair medium, which may lead to false-negative culture results. Although the high batch-to-batch consistency of filter paper has yet to be proven, the simplicity of the method and medium might make it less susceptible to variations and better adapted to field conditions than the more sophisticated Cary-Blair medium.