Table 5.

Recommendations of the EAPCRI for Aspergillus PCR protocols when testing serum

Result	Recommendation
A positive association between sensitivity and initial sample vol was noted (z value, −2.27; P = 0.023).	The use of larger sample volumes (≥0.5 ml) will improve sensitivity.
When detecting low fungal burdens (<10 genomes/ml), the difference in sensitivity using a starting volume of <0.5 ml serum compared to ≥0.5 ml is 21.8% (95% CI, 5.6 to 37.1; P = 0.0091).	To enhance detection of low fungal burdens (<10 genomes/ml), a minimum of 0.5 ml starting material should be used.
Twenty-eight of 29 protocols evaluated were able to reach the designated threshold of detection on at least one occasion, generating a threshold positivity rate of 91.5%.	Most commercially available nucleic acid extraction systems described in Table 2 can be used to extract Aspergillus DNA using the protocols as described by the manufacturer. However, all kits should be screened for contamination and a limit of detection determined prior to clinical use.
Method 11 (Table 2) was specifically designed to test whole blood and was not able to efficiently process serum samples.	Protocols specifically designed for testing alternative specimen types (e.g., whole blood) should not be used.
A negative association between sensitivity and elution vol was noted (z value, 2.96; P = 0.003).	The use of larger elution volumes (≥100 μl) will reduce sensitivity.
When detecting low fungal burdens (<10 genomes/ml), the difference in Cq for methods eluting in <100 μl (37.8 cycles) compared to ≥100 μl (41.7 cycles) volumes was 3.9 cycles (95% CI, 1.9 to 5.8 cycles; P = 0.0002) and results in a 29.0% reduction in positivity.	For optimal detection of low fungal burdens (<10 genomes/ml), nucleic acid should be eluted in vol of <100 μl. Elution in vol of ≥100 μl delays Cq values to the degree where real-time PCR is inconsistent and positivity will be affected.
A positive association between sensitivity and PCR targeting the ITS region was noted (z value, −2.48; P = 0.013).	The reaction kinetics for each PCR reaction will vary according to the design of specific oligonucleotides and the optimization of each particular PCR assay. Performance may not be directly associated with the target gene but exclusive to the individual assay itself. When designing a new assay, the MIQE guidelines (2a) should be followed.
A negative association between sensitivity and PCR targeting the mitochondrial regions was noted (z value, 2.57; P = 0.010).	The reaction kinetics for each PCR reaction will vary according to the design of specific oligonucleotides and the optimization of each particular PCR assay. Performance may not be directly associated with the target gene but exclusive to the individual assay itself. When designing a new assay, the MIQE guidelines (2a) should be followed.
A positive association between sensitivity and the use of an internal control PCR was noted (z value, −2.18; P = 0.029).	An internal control PCR should be used and incorporated at the start of the extraction procedure, monitoring both PCR inhibition and DNA extraction efficiency. Internal control Cq values should be representative of typical Aspergillus PCR-positive results (≥35 cycles), and human DNA targets should be avoided due to the possibility of variable target amounts within individual specimens generating varied Cq values.
Using a threshold of 43 cycles generates a DOR of 105, and 87.4% of results were correctly classified.	A PCR positivity threshold of 43 cycles provides the greatest degree of diagnostic accuracy.