Tweet
Wkly Epidemiol Rec. WHO external quality assessment for detecting influenza A virus using polymerase chain reaction ? summary, 2010
WHO external quality assessment for detecting influenza A virus using polymerase chain reaction ? summary, 2010 (WHO, Jan 14 2011, edited)
[Source: World Health Organization, full PDF document, (LINK). Extract, edited.]
Weekly epidemiological record
Relev? ?pid?miologique hebdomadaire
14 january 2011, 86th year / 14 janvier 2011, 86e ann?e
No. 3, 2011, 86, 17?24
WHO external quality assessment for detecting influenza A virus using polymerase chain reaction ? summary, 2010
National Influenza Centres (NICs) have been the backbone of the WHO Global Influenza Surveillance Network for more than 50 years. The centres collect specimens, conduct preliminary analyses and send representative virus isolates in a timely manner to the WHO Collaborating Centres for Reference and Research on Influenza to support annual recommendations for the composition of influenza vaccines for the next season. The NICs also play a key part in detecting viruses that have pandemic potential, thus facilitating responses to outbreaks and preparedness for pandemics. This was particularly evident during the influenza A(H1N1) 2009 virus pandemic.
Because polymerase chain reaction (PCR) has been increasingly used as a principal method for both the routine diagnosis and surveillance of influenza viruses, including avian influenza A(H5N1), a WHO external quality assessment project was launched in 2007 to monitor laboratories? performance in rapidly detecting influenza viruses using PCR. The performance of participating laboratories in response to the first 6 panels sent for analysis (2007?2009) has already been summarized in the Weekly Epidemiological Record.(1, 2)
The project, coordinated by the WHO Global Influenza Programme, continued in 2010: the WHO Reference Laboratory for Diagnosis of Influenza A/H5 Infection and the National Influenza Centre at the Centre for Health Protection, China, Hong Kong Special Administrative Region (SAR) sent the panels to participants; additional support was provided by the WHO Collaborating Centres for Reference and Research on Influenza, other H5 reference laboratories and WHO regional offices. The scope of the quality assessment project was expanded to include influenza B in the panels sent for assessment during 2010.
This report summarizes the results of assessments of panels 7 and 8, which were dispatched in 2010 during January?March and June?August, respectively.
Methods
Preparation of panels
Vacuum-dried RNA samples of influenza A(H1N1), influenza A(H3N2), influenza A(H5N1), pandemic influenza A(H1N1) 2009 virus and influenza B viruses were dispatched to participating laboratories. Samples were prepared as described previously.(1, 2)
Composition of panels
Panel 7 and panel 8 each consisted of 10 coded samples containing different concentrations of RNA from different genetic clades of influenza A(H5N1), A(H1N1), A(H3N2), pandemic A(H1N1) 2009 virus and influenza B viruses. Samples that contained no virus were included in both panels. Details of the composition of the panels are shown in Table 1. Participants were instructed to reconstitute each sample with the buffer provided prior to testing. A questionnaire on methods of detection and gene targets used by the laboratories was also included.
Distribution of panels and response of participants
NICs and other national registered influenza laboratories were invited to participate before the panels were dispatched. A total of 157 laboratories were invited to participate in assessing panel 7 and 167 laboratories were invited for panel 8. Panel 7 was dispatched between mid-January 2010 and the beginning of March 2010, and panel 8 was dispatched between mid-June and mid-August 2010. The 2 panels were dispatched at ambient temperature by courier service to participating laboratories in all of the 6 WHO regions. Participating laboratories were requested to notify the centre immediately after receiving the panel, either by fax or e-mail, and to return their results within 1 month. A preliminary report of the correct results was sent to participants shortly after the closing date for receipt of each panel.
Results for panel 7 were reported by 140 laboratories from 111 countries, areas and territories. One laboratory?s results for panel 7 were excluded from analysis due to prolonged shipment time. Thus, the number of laboratories for which results are reported for panel 7 is 139. Results for panel 8 were reported by 158 laboratories from 125 countries, areas and territories. Details of the number of laboratories invited to participate in panels 1?8 and the number that responded from all 6 WHO regions are shown in Figure 1. The majority of participants received their panels within 1 week of dispatch (panel 7, 120/139 [86%]; panel 8, 131/158 [83%]).
For both panel 7 and panel 8, 111/297 (37%) participating laboratories included in the analyses were in the WHO European Region; 59/297 (20%) were in the Region of the Americas; 46/297 (15%) were in the Western Pacific Region; 43/297 (14%) were in the African Region; 24/297 (8%) were in the Eastern Mediterranean Region; and 14/297 (5%) were in the South-East Asia Region.
Results
The criteria used to evaluate results of the analyses of panels 7 and 8 were the same as those used for panels 1?6.1, 2 In addition, the following criteria were used to assess results for samples containing influenza B virus:
Laboratories? performance
Panel 7
Of the 139 laboratories included in the analysis, 118 (85%) returned correct results for all 10 samples.
An additional 7 (5%) participants returned 9 correct results, and 14 (10%) returned 5?8 correct results (Table 2).
Two laboratories mistakenly reported positive results for a negative sample (sample 2010-09) (Table 1); the false-positive rate was 1%.
For the paired samples containing similar concentrations of genetic clade 2.2 H5 virus (samples 2010-02 and 2010-06), 131 (94%) laboratories correctly reported the results. For the other paired samples containing similar concentrations of clade 2.3.4 H5 virus (2010-04, 2010-08), 132 (95%) laboratories reported correct results for sample 2010-04 and 133 (96%) reported correct results for sample 2010-08.
For the paired pandemic samples containing similar concentrations of (H1N1) 2009 virus (2010-05, 2010-10), 137 (99%) laboratories reported correct results for sample 2010-05 and 136 (98%) reported correct results for sample 2010-10.
The sample containing H1 (2010-01) was reported correctly by 129 (93%) laboratories; the sample containing H3 (2010-03) was reported correctly by 136 (98%) laboratories.
The sample containing influenza B virus (2010?07) was reported correctly by all laboratories that assessed panel 7.
Panel 8
Of 158 laboratories included in the analysis, 136 (86%) returned correct results for all 10 samples. An additional 8 (5%) laboratories returned 9 correct results; 10 (6%) returned 5?8 correct results; and 4 (3%) returned <5 correct results (Table 2).
The false-positive rate was 1%: 2 participating laboratories reported a positive result for a negative sample (2010-16) (Table 1).
For the paired samples containing similar concentrations of clade 2.3.2 H5 virus (2010-13, 2010-19), 146 (92%) laboratories reported correct results. For the other paired samples containing different concentration of clade 1 H5 virus (2010-15, 2010-18), 149 (94%) laboratories reported correct results for sample 2010-15 and 147 (93%) reported correctly on sample 2010-18.
For the paired samples containing different concentrations of pandemic (H1N1) 2009 virus (2010-12, 2010-20), 154 (98%) laboratories reported correct results for both samples.
Samples containing H1 virus (2010-17) and H3 virus (2010-14) were both correctly reported by 154 (98%) laboratories.
The sample containing influenza B virus (2010-11) was reported correctly by 152 (96%) laboratories.
Methods of detection
There was considerable variation in the testing strategies and PCR protocols used by laboratories to screen for influenza type A and type B viruses and subtypes H1, H3, H5 and pandemic (H1N1) 2009 viruses. To screen for influenza A virus, 136 (99%) laboratories participating in panel 7 and 156 (99%) laboratories participating in panel 8 performed M gene detection. However, a majority of laboratories did not specify the target gene used to screen for influenza B virus (panel 7, 91 [66%] did not specify target gene; panel 8, 115 [73%] did not specify). The protocol from the United States Centers for Disease Control and Prevention (CDC) for detecting H5 virus had been adopted by 68 laboratories participating in panel 7 and 80 laboratories in panel 8. The CDC protocol for pandemic (H1N1) 2009 virus had been adopted by 96 laboratories in panel 7 and 109 laboratories in panel 8. There was little difference in test results despite the use of different PCR protocols by different laboratories. Details of the target genes, detection methods and source of primers or probes, and enzymes used were included in the summary report that was distributed to all participants along with the results.
Comparison of laboratories? performance for all panels
From 2007 to 2010, the number of laboratories participating in external quality assessments and reporting their results in time for analysis increased, from 64 for panel 1 to 139 for panel 7 and 158 for panel 8. The percentage of laboratories that correctly assessed all panels has also increased, from 43/64 (67%) in panel 1 to 118/139 (85%) in panel 7 and 136/158 (86%) in panel 8. The percentage of laboratories correctly detecting H5 virus increased from 49/64 (77%) for panel 1 to 125/139 (90%) for panel 7 and 144/158 (91%) for panel 8. The laboratories continued to perform well in detecting pandemic (H1N1) 2009 virus, with 132/139 (95%) returning correct results for panel 7 and 152/158 (96%) for panel 8.
Factors affecting performance
An analysis of the results of panel 7 showed that laboratories using real-time PCR in ≥1 test for the H5 gene performed significantly better than those using only conventional PCR (P<0.05). The use of commercial kits did not have a significant effect on a laboratory?s performance (P=0.507).
Discussion
The number of laboratories participating in external quality assessments has increased steadily since the first panel was sent out. As shown in Figure 1, the number of participating laboratories has increased steadily in all 6 WHO regions. In terms of overall performance, the percentage of laboratories reporting all results correctly also increased steadily (Figure 2). Improvement was also observed in the detection of H5 virus. H5 viruses from different clades (1, 2.2, 2.3.4) were included in panels 7 and 8, and the percentage of laboratories reporting all results correctly for samples containing H5 virus increased from 49/64 (77%) for panel 1 to 144/158 (91%) for panel 8 (Figure 2). Of the 17 laboratories that reported incorrect results for H5 samples in panels 7 and 8, 9 (53%) had previously reported incorrect results for H5 samples. It is hoped that identifying these gaps in proficiency will help laboratories to improve their performance.
Because the 2 influenza type-B lineages (B/Victoria/2/87 and B/Yamagata/16/88) continue to circulate worldwide, samples containing influenza B virus were included in the external quality assessment in 2010 for the first time.
All participating laboratories were capable of correctly detecting influenza B viruses in panel 7. In the assessment of panel 8, 11 (7%) laboratories did not test for influenza B virus. Among the 147 laboratories that reported detecting influenza B virus, 141 (96%) reported correctly.
The results of the external quality assessment of panels 7 and 8 provided a means of following up and monitoring laboratories? proficiency in diagnosing influenza A viruses and influenza B viruses, and showed that an increasing number of laboratories are capable of correctly identifying all samples.
NICs and designated national influenza laboratories in countries without NICs are encouraged to continue to participate in the external quality assessment project to benefit from the monitoring of their performance, optimizing protocols and exchanging information. It is expected that this project will help to raise the quality and standard of the global capacity to diagnose influenza.
Editorial note.
The full report for each panel contains more information than is presented in this summary. All data from the project will be used by the Global Influenza Surveillance Network as a foundation for assessment of the needs of the laboratories and develop plans for action.
WHO thanks all the NICs and other influenza laboratories that participated in the project, for the time they spent completing the questionnaires and for their willingness to share information for this analysis.
For more information, please contact the WHO Global Influenza Programme, Geneva, Switzerland (email: GISN@who.int).
(1) See No. 45, 2008, pp. 401?412.
(2) See No. 48, 2009, pp. 493?504.
-
------
[Source: World Health Organization, full PDF document, (LINK). Extract, edited.]
Weekly epidemiological record
Relev? ?pid?miologique hebdomadaire
14 january 2011, 86th year / 14 janvier 2011, 86e ann?e
No. 3, 2011, 86, 17?24
WHO external quality assessment for detecting influenza A virus using polymerase chain reaction ? summary, 2010
National Influenza Centres (NICs) have been the backbone of the WHO Global Influenza Surveillance Network for more than 50 years. The centres collect specimens, conduct preliminary analyses and send representative virus isolates in a timely manner to the WHO Collaborating Centres for Reference and Research on Influenza to support annual recommendations for the composition of influenza vaccines for the next season. The NICs also play a key part in detecting viruses that have pandemic potential, thus facilitating responses to outbreaks and preparedness for pandemics. This was particularly evident during the influenza A(H1N1) 2009 virus pandemic.
Because polymerase chain reaction (PCR) has been increasingly used as a principal method for both the routine diagnosis and surveillance of influenza viruses, including avian influenza A(H5N1), a WHO external quality assessment project was launched in 2007 to monitor laboratories? performance in rapidly detecting influenza viruses using PCR. The performance of participating laboratories in response to the first 6 panels sent for analysis (2007?2009) has already been summarized in the Weekly Epidemiological Record.(1, 2)
The project, coordinated by the WHO Global Influenza Programme, continued in 2010: the WHO Reference Laboratory for Diagnosis of Influenza A/H5 Infection and the National Influenza Centre at the Centre for Health Protection, China, Hong Kong Special Administrative Region (SAR) sent the panels to participants; additional support was provided by the WHO Collaborating Centres for Reference and Research on Influenza, other H5 reference laboratories and WHO regional offices. The scope of the quality assessment project was expanded to include influenza B in the panels sent for assessment during 2010.
This report summarizes the results of assessments of panels 7 and 8, which were dispatched in 2010 during January?March and June?August, respectively.
Methods
Preparation of panels
Vacuum-dried RNA samples of influenza A(H1N1), influenza A(H3N2), influenza A(H5N1), pandemic influenza A(H1N1) 2009 virus and influenza B viruses were dispatched to participating laboratories. Samples were prepared as described previously.(1, 2)
Composition of panels
Panel 7 and panel 8 each consisted of 10 coded samples containing different concentrations of RNA from different genetic clades of influenza A(H5N1), A(H1N1), A(H3N2), pandemic A(H1N1) 2009 virus and influenza B viruses. Samples that contained no virus were included in both panels. Details of the composition of the panels are shown in Table 1. Participants were instructed to reconstitute each sample with the buffer provided prior to testing. A questionnaire on methods of detection and gene targets used by the laboratories was also included.
Distribution of panels and response of participants
NICs and other national registered influenza laboratories were invited to participate before the panels were dispatched. A total of 157 laboratories were invited to participate in assessing panel 7 and 167 laboratories were invited for panel 8. Panel 7 was dispatched between mid-January 2010 and the beginning of March 2010, and panel 8 was dispatched between mid-June and mid-August 2010. The 2 panels were dispatched at ambient temperature by courier service to participating laboratories in all of the 6 WHO regions. Participating laboratories were requested to notify the centre immediately after receiving the panel, either by fax or e-mail, and to return their results within 1 month. A preliminary report of the correct results was sent to participants shortly after the closing date for receipt of each panel.
Results for panel 7 were reported by 140 laboratories from 111 countries, areas and territories. One laboratory?s results for panel 7 were excluded from analysis due to prolonged shipment time. Thus, the number of laboratories for which results are reported for panel 7 is 139. Results for panel 8 were reported by 158 laboratories from 125 countries, areas and territories. Details of the number of laboratories invited to participate in panels 1?8 and the number that responded from all 6 WHO regions are shown in Figure 1. The majority of participants received their panels within 1 week of dispatch (panel 7, 120/139 [86%]; panel 8, 131/158 [83%]).
For both panel 7 and panel 8, 111/297 (37%) participating laboratories included in the analyses were in the WHO European Region; 59/297 (20%) were in the Region of the Americas; 46/297 (15%) were in the Western Pacific Region; 43/297 (14%) were in the African Region; 24/297 (8%) were in the Eastern Mediterranean Region; and 14/297 (5%) were in the South-East Asia Region.
Results
The criteria used to evaluate results of the analyses of panels 7 and 8 were the same as those used for panels 1?6.1, 2 In addition, the following criteria were used to assess results for samples containing influenza B virus:
- (i) failing to detect influenza B virus or reporting the result as influenza type A, or both, was recorded as an incorrect response;
- (ii) failing to report a negative test result for influenza type A if influenza B typing was not performed was recorded as an incorrect response.
Laboratories? performance
Panel 7
Of the 139 laboratories included in the analysis, 118 (85%) returned correct results for all 10 samples.
An additional 7 (5%) participants returned 9 correct results, and 14 (10%) returned 5?8 correct results (Table 2).
Two laboratories mistakenly reported positive results for a negative sample (sample 2010-09) (Table 1); the false-positive rate was 1%.
For the paired samples containing similar concentrations of genetic clade 2.2 H5 virus (samples 2010-02 and 2010-06), 131 (94%) laboratories correctly reported the results. For the other paired samples containing similar concentrations of clade 2.3.4 H5 virus (2010-04, 2010-08), 132 (95%) laboratories reported correct results for sample 2010-04 and 133 (96%) reported correct results for sample 2010-08.
For the paired pandemic samples containing similar concentrations of (H1N1) 2009 virus (2010-05, 2010-10), 137 (99%) laboratories reported correct results for sample 2010-05 and 136 (98%) reported correct results for sample 2010-10.
The sample containing H1 (2010-01) was reported correctly by 129 (93%) laboratories; the sample containing H3 (2010-03) was reported correctly by 136 (98%) laboratories.
The sample containing influenza B virus (2010?07) was reported correctly by all laboratories that assessed panel 7.
Panel 8
Of 158 laboratories included in the analysis, 136 (86%) returned correct results for all 10 samples. An additional 8 (5%) laboratories returned 9 correct results; 10 (6%) returned 5?8 correct results; and 4 (3%) returned <5 correct results (Table 2).
The false-positive rate was 1%: 2 participating laboratories reported a positive result for a negative sample (2010-16) (Table 1).
For the paired samples containing similar concentrations of clade 2.3.2 H5 virus (2010-13, 2010-19), 146 (92%) laboratories reported correct results. For the other paired samples containing different concentration of clade 1 H5 virus (2010-15, 2010-18), 149 (94%) laboratories reported correct results for sample 2010-15 and 147 (93%) reported correctly on sample 2010-18.
For the paired samples containing different concentrations of pandemic (H1N1) 2009 virus (2010-12, 2010-20), 154 (98%) laboratories reported correct results for both samples.
Samples containing H1 virus (2010-17) and H3 virus (2010-14) were both correctly reported by 154 (98%) laboratories.
The sample containing influenza B virus (2010-11) was reported correctly by 152 (96%) laboratories.
Methods of detection
There was considerable variation in the testing strategies and PCR protocols used by laboratories to screen for influenza type A and type B viruses and subtypes H1, H3, H5 and pandemic (H1N1) 2009 viruses. To screen for influenza A virus, 136 (99%) laboratories participating in panel 7 and 156 (99%) laboratories participating in panel 8 performed M gene detection. However, a majority of laboratories did not specify the target gene used to screen for influenza B virus (panel 7, 91 [66%] did not specify target gene; panel 8, 115 [73%] did not specify). The protocol from the United States Centers for Disease Control and Prevention (CDC) for detecting H5 virus had been adopted by 68 laboratories participating in panel 7 and 80 laboratories in panel 8. The CDC protocol for pandemic (H1N1) 2009 virus had been adopted by 96 laboratories in panel 7 and 109 laboratories in panel 8. There was little difference in test results despite the use of different PCR protocols by different laboratories. Details of the target genes, detection methods and source of primers or probes, and enzymes used were included in the summary report that was distributed to all participants along with the results.
Comparison of laboratories? performance for all panels
From 2007 to 2010, the number of laboratories participating in external quality assessments and reporting their results in time for analysis increased, from 64 for panel 1 to 139 for panel 7 and 158 for panel 8. The percentage of laboratories that correctly assessed all panels has also increased, from 43/64 (67%) in panel 1 to 118/139 (85%) in panel 7 and 136/158 (86%) in panel 8. The percentage of laboratories correctly detecting H5 virus increased from 49/64 (77%) for panel 1 to 125/139 (90%) for panel 7 and 144/158 (91%) for panel 8. The laboratories continued to perform well in detecting pandemic (H1N1) 2009 virus, with 132/139 (95%) returning correct results for panel 7 and 152/158 (96%) for panel 8.
Factors affecting performance
An analysis of the results of panel 7 showed that laboratories using real-time PCR in ≥1 test for the H5 gene performed significantly better than those using only conventional PCR (P<0.05). The use of commercial kits did not have a significant effect on a laboratory?s performance (P=0.507).
Discussion
The number of laboratories participating in external quality assessments has increased steadily since the first panel was sent out. As shown in Figure 1, the number of participating laboratories has increased steadily in all 6 WHO regions. In terms of overall performance, the percentage of laboratories reporting all results correctly also increased steadily (Figure 2). Improvement was also observed in the detection of H5 virus. H5 viruses from different clades (1, 2.2, 2.3.4) were included in panels 7 and 8, and the percentage of laboratories reporting all results correctly for samples containing H5 virus increased from 49/64 (77%) for panel 1 to 144/158 (91%) for panel 8 (Figure 2). Of the 17 laboratories that reported incorrect results for H5 samples in panels 7 and 8, 9 (53%) had previously reported incorrect results for H5 samples. It is hoped that identifying these gaps in proficiency will help laboratories to improve their performance.
Because the 2 influenza type-B lineages (B/Victoria/2/87 and B/Yamagata/16/88) continue to circulate worldwide, samples containing influenza B virus were included in the external quality assessment in 2010 for the first time.
All participating laboratories were capable of correctly detecting influenza B viruses in panel 7. In the assessment of panel 8, 11 (7%) laboratories did not test for influenza B virus. Among the 147 laboratories that reported detecting influenza B virus, 141 (96%) reported correctly.
The results of the external quality assessment of panels 7 and 8 provided a means of following up and monitoring laboratories? proficiency in diagnosing influenza A viruses and influenza B viruses, and showed that an increasing number of laboratories are capable of correctly identifying all samples.
NICs and designated national influenza laboratories in countries without NICs are encouraged to continue to participate in the external quality assessment project to benefit from the monitoring of their performance, optimizing protocols and exchanging information. It is expected that this project will help to raise the quality and standard of the global capacity to diagnose influenza.
Editorial note.
The full report for each panel contains more information than is presented in this summary. All data from the project will be used by the Global Influenza Surveillance Network as a foundation for assessment of the needs of the laboratories and develop plans for action.
WHO thanks all the NICs and other influenza laboratories that participated in the project, for the time they spent completing the questionnaires and for their willingness to share information for this analysis.
For more information, please contact the WHO Global Influenza Programme, Geneva, Switzerland (email: GISN@who.int).
(1) See No. 45, 2008, pp. 401?412.
(2) See No. 48, 2009, pp. 493?504.
-
------