In what was one of the standout sessions at this year's Summit, seven diagnostic product developers were asked to present on what level of accuracy will be required to bring a commercially viable product to market, and also how large the clinical studies will need to be to certify sufficient accuracy in their tests.

Introduction into Diagnostic Measures
CEO of BlinkLab, Henk-Jan Boele stressed that what is much more important than sensitivity and specificity is NPV (Negative Predictive Value) and PPV (Positive Predictive Value). "What is the chance that I really have the disease," is what the patients (or carers) want to know. "That is the most relevant question here."

With a data set, the sensitivity (proportion of positive cases picked up) can be increased, but this compromises the specificity (i.e. more false positives). Conversely the specificity can be increased but with a reduced sensitivity.

But importantly, the prevalence of the disease needs to be considered, which will directly impact the NPV or PPV (which effectively takes into account sensitivity, specificity and disease prevalence). For instance with a low prevalence, say 15%, a lower sensitivity would make sense to increase the Positive Predictive Value argued Boele.

Sensitivity = (true positives) / (true positive + false negatives)

Positive predictive value (PPV) is the probability that a person who gets a positive test result actually has the disease or condition.

Negative predictive value (NPV) is the chance that a person with a negative test result truly does not have the disease.

BlinkLab (Autism & ADHD)
BlinkLab (BB1: $0.76) was founded five years ago as a Princeton University start-up focused on diagnosing autism and ADHD via a smartphone. With autism, the prevalence is increasing, which CEO Henk-Jan Boele said was a global phenomenon. The issues at the moment are the long wait lists to diagnosis and treatment.

"Keep in mind there is no brain scan, there is no blood test, there is no biomarker for autism." Autism continues to be measured subjectively. The only tools available are questionnaires. But early diagnosis leads to better clinical outcomes believes Boele as well as significantly reducing medical expenses.

BlinkLab is looking at blinking reflexes, and other facial movements, that similar to a knee stimulus response, can not be controlled. The company's data is published in peer-reviewed journals, where a sensitivity and specificity of 84% and 85% respectively has been achieved most recently.

The company is currently conducting a pivotal study in the US for 510k clearance for use as a diagnostic aid for specialists – not for self-diagnosis – where the test can be conducted in the home setting, with the data going to the prescribing doctor.

One of the challenges for BlinkLab is that the current diagnosis is subjective and clinicians can differ with their opinions in 70%-80% of cases. This means that a 100% 'accuracy' with a new technology can never be achieved.

As such the performance window for BlinkLab is 85%-100%. "We will never reach 100%" under the current reference system.

One of the key learnings for BlinkLab has been to agree with the FDA on your thresholds before you start your pivotal studies, which in autism is 65% for sensitivity and specificity. Another learning has been to install a second independent review, that requires a third reviewer if the first two can not reach consensus. Also important has been to run pilot studies that mirror the pivotal study as well as understanding where the new diagnostic will fit into the current clinical workflow.

A benefit that BlinkLab has is that it can sell its product through the specialist network without reimbursement because the pricing of the test is so low. The key initially is to drive early adoption. After that will come a wide use screening device, where the parameters of the test will need to be adjusted.

In ADHD, BlinkLab recently generated highly positive results in a study in 208 subjects, achieving a sensitivity and specificity of 82% and 83% respectively. The company will now look into adults as well. It has also been working with a pharmaceutical client, showing the impact of the drug methylphenidate in subjects with ADHD. Using the BlinkLab technology, the hyper-responsiveness was shown to fall away within two hours following treatment, and then to return as the drug's effect dissipated over the day. Monitoring drug responsiveness is a major commercial opportunity for BlinkLab.

Boele's five tips are:
1. Start with the gap in the market, not the technology
2. The real asset is the area under the curve
3. Intended use determines the threshold
4. One asset can have more than one use, such as a screening and diagnostic, depending on the threshold. Many diagnostics fail because of the wrong threshold
5. Adoption is driven by the NPV and PPV

Disclaimer:
Information contained in this newsletter is not a complete analysis of every material fact respecting any company, industry or security. The opinions and estimates herein expressed represent the current judgement of the publisher and are subject to change. Blake Industry and Market Analysis Pty Ltd (BIMA) and any of their associates, officers or staff may have interests in securities referred to herein (Corporations Law s.849). Details contained herein have been prepared for general circulation and do not have regard to any person’s or company’s investment objectives, financial situation and particular needs. Accordingly, no recipients should rely on any recommendation (whether express or implied) contained in this document without consulting their investment adviser (Corporations Law s.851). The persons involved in or responsible for the preparation and publication of this report believe the information herein is accurate but no warranty of accuracy is given and persons seeking to rely on information provided herein should make their own independent enquiries. Details contained herein have been issued on the basis they are only for the particular person or company to whom they have been provided by Blake Industry and Market Analysis Pty Ltd. The Directors and/or associates declare interests in the following ASX Healthcare and Biotechnology sector securities: Analyst MP: 1AD, ACR, AVR, CGS, CUV, CYC, DXB, IMM, LBT, MX1, OPT, NEU, PAB, PXS,RNO,SOM. These interests can change at any time and are not additional recommendations. Holdings in stocks valued at less than  $100 are not disclosed.