Review 2: "Virus Testing Optimisation Using Hadamard Pooling"

RR\ID Evidence Scale rating by reviewer:

• Reliable. The main study claims are generally justified by its methods and data. The results and conclusions are likely to be similar to the hypothetical ideal study. There are some minor caveats or limitations, but they would/do not change the major claims of the study. The study provides sufficient strength of evidence on its own that its main claims should be considered actionable, with some room for future revision.

***************************************

Review: The development of additional methods to reduce waste and increase efficiency in detection of viral infections within populations is necessary for responding to the next pandemic and current public health needs.  The authors present Hadamard matrices as a simple sample pooling design method to address this need and allow deconvolution of positive samples from the pooled testing without the need for complex computation.  Hadamard matrices are an interesting method to address this need and are compatible with other pooling and sample handling methods.  The argument for the use of these matrices at 1% positive rate, or lower, is well laid out.  A minor comment is that the statement at the end of the Figure 1 legend “In the case of pool A samples 1, 2, 3 and 5 would be grouped” appears to be wrong or misleading, and the grouping should be samples 3, 5, 6 and 7.

However, the need for pooling increases dramatically with increased rates of sampling present with increased infection rates and the background positivity rate in surveillance samples can also vary wildly even in non-symptomatic populations.  It would greatly strengthen the understanding provided to the reader if the authors would explore more levels of positivity to determine at what rates Hadamard matrices retain advantages.  For example, is this method still advantageous at 2, 5, 7, or 10 % positivity? This increased understanding would allow users flexibility to know when to use the method and when to pivot based on tracking of the current testing population positivity rates.  Giving the reader this information would also greatly increase the potential impact of the research.

Finally, the authors should take into account the loss in signal and sensitivity of the various assays they propose to combine with the Hadamard matrices.  There are accepted pooling numbers for viral detection while maintaining signal detection levels necessary to avoid missing a positive sample.  By increasing the linkage to the known sensitivities, this would increase the impact of the research and provide guidance for the use of the Hadamard matrices in real world situations.

While there are several things that could significantly increase the impact of the work, the authors do show that Hadamard matrices have potential to increase testing throughput while reducing waste generation meeting a critical need for responding to public health and pandemic events.