When you connect a breathalyzer to a smartphone app or a worksite record system, the results shown on screen are stored as numbers and used for searches, statistics, and reports. When a result outside the device's display range comes in, system developers or data managers may be tempted to replace it with the maximum value of that range. For example, when the screen or transmitted value shows an over-range status rather than a number, the upper limit is stored in its place for convenience of analysis.
However, the display range is the interval within which the device is designed to show numbers, and an over-range result may not be any particular value within that interval. So replacing an over-range result with a large number changes the meaning the data expresses from the meaning the test result actually has. The key is not to treat a result without a number as missing but to preserve it as a result status distinct from numeric values.
The display range does not stand in for a reading
DATECH's official product information for the AF-40BT gives the display range as 0.000–0.500 %BAC, 0.00–5.00 ‰, and 0.00–2.50 mg/L. %BAC is a notation unit for blood alcohol concentration, ‰ is a per-mille notation, and mg/L is a notation unit for alcohol concentration in breath. This range is a spec describing the interval the device displays in each unit.
That does not mean that, outside the range, the device always measured some value and that value equaled the display upper limit. How the device actually expresses status in its screen messages, stored records, and communication data must be confirmed separately through official model materials or actual integration documents. That is why you must not infer how over-range results are handled from display range information alone.
Replacing with the upper limit merges two results into one
For instance, as a hypothetical example, suppose the record system received the number 0.500 %BAC. This value could be a quantitative result where the device actually displayed 0.500 %BAC, or it could be a status originally sent as over-range that the system replaced with 0.500. After substitution, the two cases cannot be told apart from the data alone.
What you need to observe is not only the size of the number but whether the device's original output contained a number and whether there was a separate indication meaning over-range. If the original was a quantitative value, you can store its unit and value; if it was an over-range status, you can store it separately from numbers. The completion criterion is a state where, when you query later, you can find "records where exactly the upper limit appeared" and "records with no quantitative value because they were outside the display range" separately.
Keep the original output, status, and reference range separately
It is better to start the storage structure from information categories people can read rather than from complex internal codes. If you keep the expression received from the device as is, and separate the type of result from whether a numeric value exists, you do not have to force the same value to serve both screen display and statistical calculations.
| Information to keep | What to record | Why it is needed |
|---|---|---|
| Original output sent by the device | The value or status message shown on screen | To trace whether the meaning changed during integration |
| Result category | Confirmed statuses such as quantitative value, over-range, or test not completed | To distinguish the reasons a number is missing from one another |
| Number and unit | The value and display unit when a number was actually provided | To keep the basis for unit conversion and numeric comparison |
| Product spec referenced | The display range and product document at the time of checking | To explain the screen design and processing criteria at that time |
Here, "test not completed" must not be lumped together with over-range. Whether the breath sample was insufficient, the power was cut, or the smartphone transmission was missed may be situations different from over-range. However, for the specific status names and the criteria for distinguishing them, use only the expressions found in that model's manual or in the integration materials the manufacturer provides.
Calculate averages and maximums from quantitative results only
If you replace over-range results with the upper limit, averages, maximums, and range distributions look like real quantitative results. In particular, if there are several upper-limit records, the system may calculate as though the same value was measured repeatedly. But since in reality you cannot know whether the value was higher or whether it was a status that was never quantified, the calculation goes beyond what the original data can say.
For operational reports, you can consider presenting the count of over-range results separately from the average or maximum of quantitative values. This way, numeric statistics are limited to records where a number was provided, and results outside the display range can be reviewed as a separate occurrence count. You should also record which records were excluded from which statistics, so you can explain differences in results when the report criteria change.
Show it as a separate result on screen and in search
If an over-range result appears as a large number on the user screen, people looking at the screen are likely to take it as an actual value. Conversely, simply leaving the number blank may lead them to think the result is missing or the device is broken. So it is appropriate to show, based on the confirmed device expression, that it is a status different from a number, such as "over range," and if needed to let users view the device's original output as well.
Search conditions that only use numeric upper limits can also miss over-range records. For example, in a hypothetical case where a manager wants to review high results, a search for "0.500 %BAC or higher" will catch only records stored as the upper limit and may miss over-range records preserved as a status. If you offer numeric conditions and over-range status conditions side by side, reviews of quantitative values and reviews of status records can be carried out without confusing the two.
Do not reverse already-converted past records by guesswork
If past data has already been replaced with the upper limit, it is hard to determine later whether each number was an actual upper-limit value or an over-range status. Reclassifying past values arbitrarily, when the original screen records, device storage records, and raw data from the time of integration are no longer available, can create new errors.
In that case, it is better to separate the records whose originals can be checked from those that cannot, and to separate original output, status, and numeric value starting with the data you collect from now on. A result outside the display range is not "a large number" but information different from a quantitative value. Only by preserving this difference will the device specs, the user screen, and the statistical results not contradict one another.
