Fit to drive: eyesight, medication and the hours you slept

A car gets a service, an inspection and a date in a calendar. The half nobody schedules is the question you ask in the hallway with the keys already in your hand — am I in a condition to drive this thing right now? The shortest true answer is that two of the three things deciding it have been measured by somebody else and written down, and the third, the one most likely to be wrong tonight, is left to your own judgement, which is precisely the thing the evidence says you judge unreliably.

What a licence measures about your eyes

The eyesight standard is the most concrete part of being fit to drive, and it is written down wherever licences are issued. Britain states it twice over. The Highway Code requires that you be able to read a vehicle number plate, in good daylight, from a distance of 20 metres. The DVLA’s published standard puts the same requirement in clinical terms: a visual acuity of at least decimal 0.5, which is 6/12 on the Snellen scale, measured with glasses or contact lenses if necessary, using both eyes together or the only seeing eye. Its guidance for medical professionals adds the part the number-plate test cannot detect at all — a field of at least 120 degrees on the horizontal, extending at least 50 degrees left and right, with no significant defect in the binocular field encroaching within 20 degrees of fixation. Lorry and bus drivers are held to more on both counts.

Spain puts the acuity figure in the same place and defines the field differently. Anexo IV of the Reglamento General de Conductores requires, for ordinary car and motorcycle licences, a binocular acuity of at least 0.5, with corrective lenses if they are needed; where they are, the fitness report must record the obligation to wear them while driving. The field requirement is written clinically rather than numerically: the binocular field must be normal, each monocular field must show no significant reduction in any of its meridians, and there must be no absolute scotomas.

In the United States there is no national figure, because screening belongs to each state, and two of them show how much room that leaves. New York requires visual acuity of at least 20/40 on the Snellen scale in either or both eyes, with or without corrective lenses. Virginia asks for 20/40 or better acuity and 110 degrees or better of horizontal vision for an unrestricted licence, and will issue a daylight-only licence down to 20/70 and 70 degrees.

Read the three together and something useful falls out. 6/12, decimal 0.5 and 20/40 are the same acuity written three ways, so there is rough agreement about how well a driver must see straight ahead. There is much less agreement about how much of the side the driver must see, and none at all about whether anybody will ever look.

Who checks, and how often

A Spanish renewal application must be accompanied by a fitness report from an authorised driver examination centre; the examination is a condition of the renewal, not something a worried driver arranges privately. New York states that the Commissioner of Motor Vehicles requires each person renewing a licence to submit to a vision examination. Britain, for car drivers, publishes the standard and puts the duty on the person: wear correction every time you drive if you need it to meet the standards of vision, and tell DVLA about any eyesight problem affecting both eyes, or the remaining eye if you have sight in one only.

None of the three is unreasonable. But acuity loss and field loss both arrive slowly, and field loss arrives without symptoms, which is why it is on the standard at all. A duty to report a problem assumes you have noticed one.

The triangle on the box

The second measurable thing is printed on the packaging, and in Spain it is a red equilateral triangle, point upwards, with a black car inside and the legend “Conducción: ver prospecto”. Real Decreto 1345/2007 requires it on the outer carton of any medicine that may reduce the ability to drive or handle dangerous machinery, and the consensus document produced by the medicines agency, the health ministry and the traffic authority is careful about what it means: the pictogram does not prohibit driving, it warns you to read the driving section of the leaflet.

The list of what carries it runs longer than most people expect — medicines for sleep disorders, anxiety, psychosis, depression, epilepsy, Parkinson’s, pain, migraine, dementia, allergy, colds and flu, and eye conditions, with the document saying plainly that other groups not in its table may affect you as well. The effects to watch for are drowsiness, difficulty concentrating, double or blurred vision, vertigo, slowed reactions, loss of coordination and faintness. The moments it singles out are the start of treatment, a change of dose, several medicines at once, alcohol, tiredness and age.

Two things there are easy to miss and do most of the work. The first is that a prescription is not a clearance: some medicines, including ones that need no prescription at all, can reduce your ability to drive safely, and the instruction for the moment you notice it is two clauses long — do not stop taking the medicine, and do not drive. Telling the prescriber that you drive regularly is what lets them look for the option that interferes least. The FDA arrives at the same place, warning that the effects of some medicines can last for several hours and even into the next day, and that antihistamines can slow your reaction time and cause mild confusion even if you don’t feel drowsy.

The second is that sleep and medication are not separate subjects. Anexo IV refuses narcolepsy and non-respiratory daytime hypersomnias outright, and where the finding is insomnia it directs particular attention to the risks attached to the medicines that might be taken for it.

The hours you actually slept

The third thing is the one with arithmetic behind it, and nobody measures it but you.

The AAA Foundation’s case-control study of a national crash database — 7,234 drivers in 4,571 crashes — compared how long each had slept in the 24 hours before the crash against whether investigators found they had contributed to it. Set against drivers who had slept 7 hours or more, those who had slept 6 to 7 hours had an estimated 1.3 times the crash rate, 5 to 6 hours 1.9 times, 4 to 5 hours 4.3 times, and under 4 hours 11.5 times. The study is unusually frank about that last figure: only 80 drivers in the sample had slept so little, six of them controls, and the confidence interval runs from under 3 to over 45. The ordering is the finding.

Its own comparison is the part worth carrying around. The estimated rate ratio for driving on 4 to 5 hours of sleep is similar to the same government’s estimates of the crash risk of driving at the blood-alcohol concentration every American state has set as its per se limit, and the figure for driving on less than 4 hours is a good deal worse. A separate strand of the analysis found that drivers who had slept an hour or more less than their own usual amount had elevated crash rates independent of how much they usually slept — so the number that matters is how much you are short by, and not simply how much sleep you had.

The hedges belong with the figures. Sleep was self-reported, alcohol and drug use could not be controlled for, and the underlying database excluded crashes between midnight and 6 AM altogether, so the study is more likely to understate the risk than overstate it.

Why the self-check is the weak part of it

The same foundation later measured drowsiness instead of asking about it, running the PERCLOS eye-closure measure over in-vehicle video from the minutes before real crashes. It found drowsiness in 8.8 to 9.5 percent of all crashes and 10.6 to 10.8 percent of police-reportable ones, against the 1.4 percent of police-reported crashes in the official national statistics it cites. Its explanation for the gap is the sentence this piece turns on: a driver who was not actually asleep but was operating at a reduced level of alertness may not even recognise that he or she was drowsy, nor that drowsiness contributed at all.

So the one input you are asked to supply from introspection is the input introspection handles worst. That is an argument for making the check arithmetic instead of a feeling, not for abandoning it.

When somebody else sets the schedule

A company that writes the roster is a party to the fatigue, and the American federal rule says so without naming a single hour. Under 49 CFR 392.3 no driver shall operate a commercial motor vehicle, and a motor carrier shall not require or permit a driver to operate one, while the driver’s ability or alertness is so impaired, or so likely to become impaired, through fatigue, illness or any other cause as to make it unsafe to begin or continue. Two phrases carry the weight: “or permit”, which reaches the dispatcher who knew, and “so likely to become impaired”, which reaches the schedule before the shift has started.

The Spanish consensus document points the same way from the clinical side, asking prescribers to take particular care with professional drivers and with anyone who drives more than 40 minutes at a stretch daily. A company that never asks has left the information with the only person who cannot change the roster.

What we cannot tell you

We cannot tell you whether you meet the eyesight standard attached to your own licence, because that is a measurement rather than an opinion. We cannot tell you how a particular medicine affects you, which is a question for the prescriber or the pharmacist and for the leaflet in the box. We cannot give you a number of hours below which you must not drive, because the study that comes closest reports population rate ratios built on self-reported sleep, and the band where the effect is largest is the band where its estimate is least precise. And we cannot tell you how an impaired driver’s claim would be treated afterwards — that turns on fault and on the wording of the policy, and it is a different piece.

Before a long drive

Go outside and read a number plate from about the length of five parked cars, in daylight, wearing whatever correction you normally wear; if that is a struggle, the answer is an appointment rather than an argument. Look at the cartons of everything you are taking, find the driving pictogram, and open the leaflet at the driving and machinery section rather than remembering what it said — treating a treatment or a dose that changed this week as the most relevant fact about the drive, and never taking a medicine for the first time on a day you have to drive. Then do the part that is arithmetic: count the hours you actually slept in the last 24, and how many short of your own usual that leaves you. Plan the stops before setting off rather than when you need one.

The eyesight and the medication are checks you can pass or fail on the spot. The sleep is the one you already decided, hours earlier, while you were doing something else.