Testing, testing, testing...1,2,3,4

Over the last couple of months, we’ve heard more about testing than in my five decades in testing. The babble has been repetitive enough to remind me of microphone checks at conferences where I was an invited speaker. March was “mask and ventilator” month, April grabbed testing, and May has run strong with testing, drugs and vaccines.

 

As a professional tester, it is flattering to finally be recognized. On the other hand, the ambiguities of diagnostic testing are evidently not widely appreciated. Let’s consider some details with respect to COVID-19. First, there are three quality diagnostic devices all readers should have at home: a blood pressure monitor, a thermometer and a finger pulse oximeter. Together, they cost less than one hour of rent in an ICU. The disease is febrile. Hypertension is a confounder. Reduced blood oxygen saturation is serious. A medical diagnosis almost always uses an algorithm combining observables. These are a very good place to start. Rather than rely on published averages, recommendations, or reference ranges, gather your own “normal” data.

 

Why are we testing?

 

Given that there are no established unique therapies for COVID-19, testing is primarily a population research project to help us understand spread. It is also helpful to protect

 

family, friends and caregivers, especially given the risk of spread days before the onset

 

of symptoms. For those of us who think “I may already have had this,” there is interest in the

 

probability of immunity, although it’s too early to conclude that this solves the problem.

 

How are we testing?

 

The presence of the SARS-CoV-2 virus is explored with rRT-PCR (real-time reverse transcription polymerase chain reaction) utilizing the RNA signature. This amplifies a sequence by replicating many times and now can be done reliably and with tolerable speed and cost, albeit in a lab. An alternative test for the presence of the virus is to use an antibody that recognizes a surface antigen on the virus. This does not permit amplification, and such tests for SARS-CoV-2 are at the exploratory stage as far as I know. The concentration of the virus will matter, but these are yes/no tests. There are opportunities for false positives and false negatives, yet when the symptoms are in full flower, reliability is bound to be better.

 

The tools of CRISPR have now been applied by a team in Boston to track the virus with a development-stage test. This is too new for me to assess, but you can learn more at www.STOPCovid.science. Their team promises minimal sample handling, minimal reagents and a fluorescence readout on a lateral flow (LF) paper strip.

 

A fourth approach is to follow the immune response to the virus, monitoring up to several immunoglobulins. Those tests may be in the format of a LF test strip not requiring

 

instrumentation, or may use sophisticated laboratory instruments for immunoassays that are standard in hospital and commercial laboratories. Questions remain about which immunoglobins are best to measure.

 

What sample?

 

The rRT-PCR test begins with rotating a swab several times within the nasopharyngeal space. Not my idea of fun. Sputum is also a possibility. Saliva is promising for all tests. Feces are getting attention as a virus repository. Blood is the established approach for serology, whether from a finger prick or a venous draw. Detecting this virus in blood is not established. This is a respiratory disease, so it makes sense to search for it there, although increasingly it afflicts the heart, kidney and intestines.

 

When are we testing?

 

At this time, testing has been limited to caregivers, first responders, those known to have been exposed to an infected person, and a few in the general population as a screen. As tests become more available, those in higher-risk categories can request a test. For example, those with a compromised immune system for any reason must be especially on guard. The testing noise has revealed a complex supply chain problem for test components. This is matched by a complex distribution problem not unlike what we are all familiar with from pharmaceuticals.

 

Where are we testing, and who carries out the test?

 

There are regulations for both the laboratory and the personnel who carry out diagnostic tests.

 

Sampling with a nasopharyngeal swab is best done by professionals, who will be at risk for exposure to the virus. Thus, we see drive-up stations in parking lots and near emergency departments with nurses in PPE and patients never leaving their car. Moving unambiguously identified samples to the lab is the next step. Large pharmacy chains have committed to have testing available at their stores. This has yet to be realized. Theranos suggests caution.

 

We do not have an approved over-the-counter test that can be completed at home. The

 

likely suspect for that is a LF test like a home pregnancy test. If COVID-19 hangs around for a while, such a test is likely to be feasible, based on finger-prick blood or perhaps saliva. There is a PCR test using a swab for sampling and another that samples saliva, where the sample is collected at home and sent to a lab for a conclusion. Those are too new to judge fairly.

 

How often should we be testing?

 

This is not clear without reliable test performance metrics. Those closely exposed to patients can be tested every few days. The rest of us should wait until symptoms appear or disappear. If this pandemic lingers, there will be a whole lot of testing going on for little cost. Not yet.

 

Are the tests valid?

 

The performance statistics for the various tests will perhaps settle down soon. This is challenging. Hundreds of tests have been announced, weighing the emergency need above clinical validation. Many announced tests have performed poorly. The American Association for Clinical Chemistry is paying close attention, and the FDA has been holding virtual town hall events on testing for clinical laboratories and test manufacturers to help answer questions. The debate has been vigorous. Sampling supplies, sampling technique, sample preparation, test reagents and instrument performance all can vary from test to test, from lot to lot and from lab to lab. Keep in mind that it is not unusual to take five years to develop and prove a diagnostic test to the satisfaction of regulatory bodies. Feasible tests have been demonstrated for SARS-CoV-2 in weeks, albeit based on mature platforms.

 

What does a test cost?

 

As with every medical procedure, the only answer is, “It depends.” And it depends on many hidden variables.

 

Until there is a choice of over-the-counter tests at CVS or Walgreens, it will be assumed that we don’t need to know. Given the competitive landscape, I suspect a range between $10 and $150. Clinical chemistry has not been on our minds until a couple of months ago. Few have visited a lab. Many don’t track their own test results, although they are available in our EHRs. It’s time we all knew our own baseline data. We paid for it.

---

Peter T. Kissinger (who can be reached at kissinger@ddn-news.com) is a professor emeritus at Purdue University, founder of BASi, chairman of Phlebotics and director of both Prosolia and Tymora.