External quality assessment of transporting infectious substances in Canada

Clinical Microbiology Proficiency Testing, Department of Pathology and Laboratory Medicine, University of British Columbia and Transport Canada, Pacific Region, Vancouver, British Columbia Correspondence and reprints: Dr MA Noble, Room 328A – 2733 Heather Street, Vancouver, British Columbia V5Z 1M9. Telephone 604-875-4685, fax 604-875-4100, e-mail mnoble@unixg.ubc.ca Received for publication August 11, 1998. Accepted October 5, 1998 SM Tiffin, MA Noble. External quality assessment of transporting infectious substances in Canada. Can J Infect Dis 1999;10(3):246-251.

P atient samples are often collected in locations distant from their final destination.The current trend of downsizing and centralization of laboratory services in Canada has led to an increased volume of patient samples sent from one site to another.The testing site might be down the road, across town, or in an entirely different city or province.Samples are transported for several reasons: initial testing in a central diagnostic laboratory, additional specialized testing of known infectious substances in a reference laboratory, or for research purposes in a research laboratory.Transportation directly influences safety, total cost and the turnaround time of laboratory testing.
Transport Canada is charged with reducing the health risk to the public from unsafe transporting practices.The Transportation of Dangerous Goods (TDG) regulations were significantly amended in 1994.One area of change was in packaging requirements for infectious substances.The regulations describe two packages called type 1A and type 1B.The 1A package is a standardized high containment package, while the 1B package has less stringent design and labelling specifications (1,2).The choice to use 1A or 1B in part depends on the risk group classification of each known infectious substance in the package.Organism risk groups were first introduced by the World Health Organization (WHO) and adopted by Health Canada for their publication, Laboratory Biosafety Guidelines (3).Transport Canada used these same risk group classification criteria and organism examples in the current TDG Regulations.Whether the package travels by air or ground and whether the sample is a culture or specimen may also influence the packaging decision.A third package option is alluded to in the regulations (Part 7.21); this package allows the sender greater choice in packaging but less protection from legal action should an accident occur (1,4,5).
It is important to determine the current understanding and compliance with the Canadian regulatory requirements in transporting samples.Because the majority of samples being transported are for initial diagnostic testing or to confirm the presence of risk group 2 infectious substances, we conducted an anonymous external quality assessment to determine the degree of compliance by clinical laboratories for the packaging and transporting of partially controlled infectious substances.

MATERIALS AND METHODS
Phase 1: With the support and cooperation of the British Columbia Centre of Disease Control (BCCDC, Vancouver, British Columbia), routine sample packages arriving daily were examined.BCCDC is a provincial reference laboratory and receives human, animal, food and water samples from health units, laboratories and environmental centres.A total of 500 packages were observed over eight consecutive days.Each package was described and recorded as per the following parameters: • outer package -intact, firm, size 10 cm or greater, shipping document attached; • external safety marks and labelling present; • inner package -watertight primary container, watertight secondary container and absorbent material present; and • package type -most consistent with type 1A, type 1B or the third packaging option 7.21 (Table 1).
Phase 2: Canadian laboratories participating in the Clinical Microbiology Proficiency Testing (CMPT) clinical bacteriology program were contacted for their approval to participate in this study.Laboratories from all provinces and territories agreed to participate.CMPT laboratories are located in all provinces except Manitoba, Ontario and Quebec.
An information letter was sent to all laboratories, followed by a written request to participate.Participants were randomly selected to submit either a live culture of Escherichia coli or a serum sample for serology testing.Serum for human immunodeficiency virus (HIV), hepatitis B and risk group 3 organisms were specifically excluded.Laboratories that had not responded after two weeks were sent a further request by facsimile to maximize the participation.
Satellite all samples to another laboratory for final packaging were excluded from phase 2 but included in phase 3. Packages from the individual laboratories were examined using identical parameters to the phase 1 group.Phase 3: A questionnaire was mailed out to each CMPT participant.Requested information included the following: • demographics -volume of testing, number of staff, number of send-outs and where the majority of samples are sent; • TDG training -number of sample packers and who trained them; • TDG certification -who certified them, how often re-certification has taken place; and • TDG resources -type of TDG reference material available to the laboratory The authors were unaware of other published studies designed to examine performance in regulatory compliance addressing transport of infectious materials.Therefore, the questionnaire was designed to be consistent with others sent from the CMPT program addressing laboratory characteristics with performance.It was pretested for content, clarity, simplicity and neutrality.All questionnaire data entries were manually checked for entry errors.
Relationships between results of phase 2 performance and phase 3 responses were examined.Results were tabulated in Microsoft Access (Washington).Analyses was performed using Microsoft Excel (Washington), EPI Info (Centers for Disease Control and Prevention, Georgia) and Kwikstat (TexaSoft, Texas) software.
For purpose of statistical analysis, results were aggregated into three categories, overpackaged (1A or 1A-like), acceptably packaged (1B or 1B-like or third option) and inadequately packaged (nonfirm outer package).Statistical examination was performed using c 2 analysis.

RESULTS
Phase 1: A total of 500 packages transported to BCCDC from hospital laboratories, private laboratories, health units, clinics and environmental testing sites were examined (  Type 1B: An inner package as described above contained within a firm outer package, with appropriate safety marks (TC 125-1B and name or registered symbol of person making the mark).Less rigorous package criteria apply.
Close to 1B: A package resembling 1B, but one component is either missing or incorrect.

7.21:
A watertight primary container and a watertight secondary container all within a firm outer package.
Inconsistent: A package that does not sufficiently meet the requirements of the 1A, 1B or the third option packaging type described in the Transport of Dangerous Goods regulations.
Risk group 1: Organisms unlikely to cause disease in a healthy population (ubiquitous in nature) and that are not regulated.

Risk group 2:
Organisms that can cause disease, but that are unlikely to be a serious hazard to the community (applies to most organisms in a diagnostic laboratory, including those in this study).
Risk group 3: Organisms causing serious disease that are not usually spread by casual contact or for which treatment is available (eg, Mycobacterium tuberculosis).
Risk group 4: Organisms causing serious disease, readily transmitted and often untreatable (there are only viruses in this category and are not dealt with in the routine diagnostic laboratories).mon destinations for samples were the provincial laboratories.Most laboratories (106, 67.5%) designated one to five persons as specimen packers.Most laboratories (106, 55.5%) relied on in-house personnel to train other personnel.Only 15 to 32 (9.2% to 18.2%) laboratories, depending on question response, reported that they did not have certified packers.Only three (1.6%)laboratories reported that they did not have access to information concerning transport regulations.Relationships between laboratory characteristics as reported in phase 3 and performance in phase 2 were sought.No significant relationships were noted between performance and geographic region of the laboratory, whether technologists were certified or whether laboratories had access to a formal copy of regulations.Several relationships were noted that either were statistically significant or almost significant (Tables 4-6).Combined the results presented in these Tables suggest that large laboratories (often private laboratories), measured either by number of technologists or number of samples processed, were more likely to overpackage compared with other laboratories.

DISCUSSION
The observations of this paper demonstrate that most laboratories are aware of transporting requirements and are capable of providing safe packages but are less likely to package in a manner fully compliant with regulations.This was seen to a greater extent in phase 1, where senders were not informed that their packaging would be examined versus phase 2 where senders were notified.Several factors may account for this difference.In phase 2, each package was sent from a separate laboratory, whereas in phase 1 there was no such restriction, Thus, part of the difference observed may be that some laboratories in phase 1 sent more than one package.In addition, the group of senders was different in the two phases.In phase 1, 233 (46.5%) of the senders were hospital laboratories, with the balance of packages coming from clinics and private laboratories.In phase 2, hospital laboratories constituted 109 (88.6%) of the group.In phase 1, laboratories were more likely to send samples in envelope packaging.Unfortunately, facilities sending samples in envelopes were not contacted by the study group.Thus, no additional information was available to ascertain other characteristics in common.
The three factors associated with overpackaging (number of samples, number of technologists and being a private laboratory) are probably related.It is difficult to ascertain to what extent these laboratories use 1A packages regularly or how much of the use was influenced by 'faking good' bias.In the process of wanting to be seen to be correct, the laboratories may have overcompensated, resulting in overpackaging.While this may have played a role, had this been a major factor, it would have been more likely to have been a more widespread effect.Other studies on the impact of blind versus open proficiency challenges have demonstrated improved but not significant differences in performance (6).Routine overpackaging can increase the overall cost of sample transportation.The packages are costly themselves and as such are often returned for reuse, resulting in transporting charges in both directions.
When preparing a sample for transport, the sender (consignor) must first assess the risk of the package contents.A sample for initial diagnostic testing has a low risk of containing a fully controlled infectious substance and, as such, does not require the standardized high containment package.If the package is known to contain a pathogen, the risk level must be considered when choosing an appropriate package (1,4,5).Phase 2 of this study focused on the majority of samples transported, which consist of diagnostic specimens and risk group 2 organisms.Thus, the expected package was a type 1B package, though the regulations do allow for variation under the 7.21 guideline.We did not accept a nonfirm outer package as meeting the intent of the 7.21 guideline.
In phase 1, packages containing water samples were examined.Whether these samples require 1B packaging is debatable.While water samples may indeed carry bacteria, the concentration is exceedingly low.On the other hand, the consequence of a leaked sample raises the perception of danger by package handlers regardless of whether microorganisms are present.Because a complete 1B package with waterproofing, and absorbent and firm packaging would reduce the potential for leakage, we conditionally included them in the analysis.
Of the 638 packages examined, it is important to appreciate that 122 (19.1%) would have been considered not only inconsistent with current regulation, but also potentially unsafe because they were transported in packages that were not firm.Of the 122 packages, one (0.8%) actually leaked outside the pack-250 Can J Infect Dis Vol 10 No 3 May/June 1999    When we compared the reported transport training and certification (questionnaire answers) with the actual performance, no trends or associations were noted.This may have been due to the size of the groups being examined or may reflect the quality of training and certification.Although Transport Canada provides guidelines on training, there is no standardization for either training or the requirements for certification, other than that the employers needs to be satisfied that the employee is knowledgeable.Four of the six laboratories that sent improperly packaged samples stated that they had certified packers.
There are few published studies looking at transportation of packages containing infectious substances.In a four-year study, Garner and Masterton (7) demonstrated that unsafe packages are rare but that compliance with regulation increased with awareness.
Samples that are transported in Canada fall within Transport Canada's guidelines.In 1994, changes in regulation resulted in a code that would allow for simple, safe transport of samples for domestic transport in Canada.As opposed to international transport, most samples and isolates can be trans-ported in inexpensive, firm boxes without a shipper's declaration or biohazard label.Requirements for the outside of the package are that it be marked with the type of packaging (TC-125-1B) and the shipper's identification.Only for special organisms of a higher risk nature (special groups within risk group 2, most risk group 3 and all risk group 4) require full packaging and labelling (1)(2)(3)(4)(5).This is not the case for international transport, making domestic transport simpler and less expensive.

CONCLUSIONS
Despite trained packers being available, most diagnostic samples and infectious substances were transported in a manner that is safe but not compliant with current TDG regulations.Packages are often both overpackaged and underpackaged.More effective education specific to transportation of infectious substances is needed for all sites shipping samples.

Table 2
).All 248 Can J Infect Dis Vol 10 No 3 May/June 1999 Tiffin and Noble

TABLE 1 Glossary of terms used in study of packaging used for transport of biological samples Inner package
: A watertight primary container, a watertight secondary container and enough absorbent to contain the contents of the primary container should a leak occur.

TABLE 2 Comparison between phase 1 and phase 2 packages
laboratories that did not participate, 33 were satellite laboratories that refer to a single central testing site.Thus, the true participation rate by laboratories that routinely transport materials was 141 of 152 (92.8%).Three packages were excluded because the outer package was removed before descriptions could be made (Table2).All outer packaging was intact.In this group, only six (4.3%) used envelope packaging as op-Phase 3: Of the 185 questionnaires requested from CMPT participants, 171 (92.4%) were returned and tabulated.Some respondents chose not to answer all questions (Table3).Of the laboratories participating in the study, the annual number of samples received ranged broadly as did the number of samples requiring packaging for transport.The most com-Can J Infect Dis Vol 10 No 3 May/June 1999 249EQA of transporting practices