From roadside tests to diagnosing Covid-19, Clare Sansom looks at how breathalysers have developed

Until very recently, most of us hardly gave a second thought to breathing. As essential as it for life and survival, we took it for granted, an act as natural as … well breathing. Since the Covid-19 pandemic swept the world, however, we have all become more aware of our own breath and especially the breath of those with whom we come into close contact. But there is a lot more that can be learnt from the composition and chemistry of the air we exhale.

An image showing a driver using a breathalizer

Source: © Getty Images

Roadside breath tests for alcohol levels have been around for decades

Smelling a person’s breath was one of the first diagnostic tests known, even as far back as Roman times. The sweetish smell of acetone on a person’s breath was a strong indication that they were suffering from diabetes mellitus. Breath analysis technology has advanced dramatically since then, of course, but the technology that people are most familiar with is a forensic one: the roadside breath test for alcohol.

Many people will have seen the occasional driver being asked to blow into a hand-held device to test their sobriety. Similar tests for driving under the influence of drugs are beginning to be developed and comparable techniques are now used in diagnosing lung conditions including Covid-19. The hope is that, after the pandemic passes, techniques developed for diagnosing this viral infection will find other applications in respiratory medicine and beyond.

Roadside revolution

The presence of alcohol in the exhaled breath of drinkers is obvious to anyone who has spent time in crowded pubs and bars. There is a strong, direct correlation between the concentration of alcohol in a sample of blood and in the exhaled breath, which permits the analysis of breath as a surrogate or proxy for laboratory analysis of blood samples. Furthermore, it is relatively easy to detect and measure alcohol (the simple ethanol molecule, C2H5OH) in the breath because it is volatile: that is, it evaporates readily at body temperature and is emitted and easily measurable in exhaled breath.

Today’s devices used by police forces throughout the world are all based on electrochemical oxidation

In the 19th century, people curious about the effects of alcoholic drinks on behaviour wondered whether ethanol should be treated as a food, a medicine or a poison. This coincided with rapid developments in analytical chemistry and applying these techniques to analyse biological specimens. The discovery that a small proportion of the alcohol a person had consumed was measurable in samples of blood, urine and exhaled breath was a prerequisite for the later development of tests for intoxication.

The first practically useful instrument for analysis of breath was christened the Breathalyser – the name given to the first instrument produced in the mid-1950s, but a word now used for any such device. ‘The first instruments for analysis of breath were very bulky and awkward to carry around’, says Wayne Jones, a forensic toxicologist at the University of Linköping in Sweden. ‘But even then, as with the hand-held electronic devices we use today, the instruments were calibrated to translate the readings into the equivalent blood-alcohol concentrations.’ Today most nations have enacted statutory alcohol limits for driving in terms of the equivalent concentrations measured in blood, urine or breath. For example, in England and Wales, ‘over the limit’ means at least 80mg alcohol in 100ml of blood, 107mg in 100ml of urine or 35µg in 100ml of breath.

Over the years, three basic analytical technologies have been incorporated in instruments used by the police for breath-alcohol analysis – photometry, spectroscopy and fuel cell technology – and these differ in accuracy, precision, specificity and ease of use. In the original breathalyser, exhaled breath was passed through a mixture of potassium dichromate (K2Cr2O7) and sulfuric acid in a glass vial. The potassium dichromate crystals oxidise any ethanol present, and the associated colour change in the oxidising agent from orange–yellow to blue–green can be measured and used to calculate the concentration of alcohol in the original breath sample. More precise quantitative analytical methods, particularly infra-red spectroscopy, came to the fore in the 1970s and 80s, and the predominant method incorporated in the hand-held roadside screening testing in use today is electrochemical oxidation. Ethanol serves as a fuel and, when oxidized, generates a measurable electrical signal. ‘Today’s small, handheld electronic devices used by police forces throughout the world are all based on electrochemical oxidation’, says Jones.

These small devices, however, are not accurate or precise enough to be used as evidence in court. If the roadside breath test is positive, the suspect is taken to the nearest police station for a more sophisticated breath test – most often using infra-red spectroscopy – under controlled environmental conditions. This is necessary if a prosecution is to succeed, but it is not ideal: blood alcohol concentrations fall as ethanol is being metabolised in the liver, so a driver who, according to the roadside test, is just over the legal limit is likely to be under the limit when the second evidential test is carried out. This can, of course, be allowed for, but it is difficult to account for varying metabolic rates of ethanol in the body, which might vary from 10 to 25mg per 100ml of blood per hour. It can be a particular problem in remote areas where the nearest police station may be many miles away.

Increasing accuracy

There is, therefore, a need for new tests that are both compact enough to be used at the roadside as well as accurate and precise enough for the results to be used in court. The Parliamentary Advisory Council on Traffic Safety (Pacts), a charity that works with, and provides the secretariat for, the All-Party Parliamentary Group for Transport Safety, is running a competition for companies interested in developing reliable roadside breath testing technology that meets the high standards required. ‘The roadside tests that we currently use aren’t getting it wrong, but we still need improved accuracy, precision and reliability, and that need has sparked the competition,’ says Evan Webster, policy and research officer at Pacts.

An image showing an AlcoSense breathalizer

Source: © AlcoSense

Commercially available breathalysers – such as this one made by AlcoSense – allow the public to check their own alcohol levels before driving

Hunter Abbott, managing director of AlcoSense, based in Maidenhead, UK, is one of those who this competition aims to reach. He set up the company, which designs and makes commercial breathalysers, after a close call in 2005. ‘I was matching my drinks with a friend at another friend’s wedding; we both drove home at noon the next day, he was stopped and found to be over the limit, which led to him losing his licence and later his job,’ he explains. Now, AlcoSense’s products are principally used by transport companies and by individuals who need to avoid this unintentional ‘morning after effect’. They are relatively cheap and easy to use, with improved versions of the 1960s potassium dichromate test at the lower end of the range and fuel cells at the upper end. Their top-end gadgets are close to the accuracy and precision required for evidential use.

AlcoSense manufactures breathalysers for the European market, and so must be able to adapt them to test for the different – and stricter – limits that apply there. Most of the continent uses a blood alcohol limit of 50mg per 100ml, with Poland and a few Scandinavian countries opting for an even stricter 20mg per 100ml. Scotland adopted the European 50mg per 100ml standard in 2014. The difference, as Abbott explains, is stark. ‘At the Scottish limit, you are five times as likely to be involved in a fatal crash as you would sober; at the English limit, you are 13 times as likely and even with only 10mg/ml you are still 37% more likely,’ he explains. ‘Most UK drivers wouldn’t dream of drink driving, but it’s surprisingly easy for residual alcohol to be left in your system the morning after the night before. This risk curve is why it’s so important for drivers to know when they’re clear of alcohol the next day.’

But even the tiniest amounts of alcohol can affect your ability to drive. Theoretically, the best limit to set would be zero, but this is technologically problematical as illness, medication or even in some cases normal metabolism might lead to false positives. Diabetic ketoacidosis is a key example. People with diabetes who are dependent on insulin may, if their insulin levels drop too far, develop this condition in which the body burns fat instead of sugar for energy releasing volatile ketones, and these could trigger false positive readings in some breathalysers if the limit is set too low. ‘The safest and most practicable limit is probably 20µg/ml’, says Abbott, and Webster agrees.

Checking for drugs

As drug abuse has become prevalent, so, too, has the crime of driving while under the influence of drugs. Today, a police officer who observes someone driving erratically but finds no alcohol in their breath must decide if the putative offender is ill, tired or distressed (or perhaps just a poor driver) or whether they are under the influence of any other drug. Breath tests for drugs are now being developed, but this is a much more complex analytical exercise than testing for alcohol could ever be. Few drugs are as volatile as alcohol is, and there are many different compounds to test for. A negative test for ethanol is enough to prove that a driver is not drunk, but a negative drug test might mean that the drug taken was not one of those tested for. And almost all drugs – prescription drugs, which can seriously affect a patient’s ability to drive, as well as drugs of abuse – are chemically much more complex than ethanol.

It is much harder to cheat in a breath test

Testing drivers for drug use is still a complex analytical business involving blood or urine assays in a hospital clinical lab – but perhaps for not much longer. ‘A bit over 10 years ago we were challenged to develop a non-invasive breath test for drugs that could be used at the roadside,’ says Olof Beck, a toxicologist at the Karolinska Institute in Stockholm, Sweden. ‘Many people thought this would be extremely difficult, but we have finally succeeded.’ And when such a piece of kit is eventually rolled out, roadside tests will not be its only use. Drug tests on prisoners have generally involved urine samples, but prisoners can find ways to cheat. ‘It is much harder to cheat in a breath test, as these can be observed easily,’ adds Beck.

Non-volatile compounds are exhaled in aerosol particles: tiny droplets of liquid that are formed during normal breathing. If a subject has recently consumed a non-volatile drug substance, that substance can be detected in a sample condensed from those aerosol particles. Beck and his co-workers are developing a device that will collect a useful sample if a subject breathes normally into it for a few minutes. This sample can be taken to a forensic lab and tested for a variety of drug compounds using tandem liquid chromatography mass spectrometry (LC–MS), a technique that is sensitive and reliable enough for its readings to be used in court. The compounds tested for can include all main drugs of abuse, some newer psychoactive substances and therapeutic drugs that may be misused, including opioid painkillers.

No LC–MS machine has yet been designed that is fast, simple and portable enough for use at the roadside, even for screening, but it is not impossible to imagine that such a device will be used one day. And moving just the sample collection from the police station to the roadside would have many advantages. Cannabinoids, for example, are metabolised rapidly, so transporting a suspected cannabis user to the police station in time to obtain a sample that will test positive is a challenge.

One further problem that forensic toxicology needs to address is a growing tendency to combine alcohol and drugs. A user may even think that half a pint and a single spliff will be OK, as both these are relatively harmless separately. However, the combination is more than additive, even when the quantities are small enough for both tests to separately come out negative. ‘Drivers under the influence of both drink and drugs have been involved in many fatal accidents’, says Beck.

A promising diagnosis

And breath analysis probably has more diagnostic applications than it has forensic ones. Exhalation Technology is a small start-up company based in Cambridge, UK, specialising in testing exhaled breath condensate (EBC). It has an interesting history, as Stig Brejl, one of its directors, explains. ‘Racehorses often suffer from an asthma-like condition that affects their form, and it is important to diagnose this early enough to pull an affected horse from a race. Working with David Marlin at the Animal Health Trust in Newmarket, UK, we identified hydrogen peroxide (H2O2) in horse breath as an inflammatory biomarker for this condition and developed a diagnostic device that can be used without veterinary training.’

An image showing breathing into CoronaCheck CTTO

Source: © Jon Klasbu

Researchers are working on a breath test for Covid-19 

Understanding that humans have similar respiratory physiology to horses, Brejl and his colleagues investigated the possibility of using a similar test for clinical diagnostics. The outcome of this study was Inflammacheck, a hand-held device for collecting breath condensate and measuring hydrogen peroxide concentrations. In humans, this is a biomarker for oxidative stress caused by lung inflammation that is found in patients with chronic obstructive pulmonary disease (COPD) and severe asthma. Interestingly, the dominant inflammatory biomarker in patients with mild asthma is a different one – fractional exhaled nitric oxide – and this, plus the spirometry (peak flow) tests that patients with severe disease find difficult to use, is often used for diagnosis. With Inflammacheck, patients breathe normally into the device for a minute or two, generating about 20µl of condensate that flows into a capillary chamber attached to a sensor with a coating based on horseradish peroxidase. When the sensor is exposed to the sample, an enzymatic process takes place generating a tiny but detectable signal on the sensor; the signal strength indicates the hydrogen peroxide concentration. ‘This assay is fast, and it has two further advantages: it is precise enough to be used with a tiny amount of sample, and it can be embedded in a hand-held device’, explains Brejl.

Inflammacheck was first marketed in September 2019, a mere six months before respiratory medicine – and much else – was upended by the Covid-19 pandemic and most COPD clinics forced online or made to close. Brejl and his team, however, rapidly pivoted to developing a diagnostic test for Sars-Cov-2 infection. ‘Inflammatory markers are not specific enough to diagnose a particular viral infection, so we had to use a different type of test,’ he explains.

As many of us have learned to our cost, active virus particles are carried on an infected individual’s breath. If that individual breathes into Exhalation Technology’s new CoronaCheck device, the condensate flows to a sensor where virus particles are bound. These can then be detected by a Sars-Cov-2 specific biosensor, generating an electrochemical readout in only 2–5 minutes, faster than the approved lateral flow tests. This non-invasive device is currently in clinical trials at Portsmouth Hospitals NHS Trust. ‘The kit includes a multi-use reader and a set of single-use consumables for sample collection, so it is completely safe,’ explains Brejl. ‘We are now investigating ways to reduce the amount of plastic in the single-use set from 85g to about 6g.’ Furthermore, if the test is approved, it should be quite easy to adapt it for detecting other respiratory viruses.

Vaccines are already promising us one way out of this pandemic. Once it has passed, will a new ‘roaring twenties’ unleash a further wave of reckless drivers? Or do further and, potentially, more dangerous pandemics lie ahead? Whatever the future holds, both forensic and diagnostic breath testing will undoubtedly help us to stay safe and well.

Clare Sansom is a science writer based in Cambridge, UK

Updated 16 April 2021 to clarify the legal blood alcohol levels in different countries