Route stop 1
The Detail: Give Every Line of Small Print a Job
Small print is not one category. A unit tells how a value is expressed. An observation window defines the period summarized. A completeness note describes how much expected data was available. A method note identifies how a laboratory or device produced a measure. A legend defines symbols. A limitation says where interpretation may fail. A version line tells when the report was generated or amended. Sorting notes by job prevents a disclaimer about one field from being applied to the entire page.
This route uses three independent educational documents: a one-row A1C laboratory slip, a fourteen-day fictional sensor summary, and a portal trend card. They were designed for this guide and do not represent people, products, laboratories, or promised outcomes. Each uses a different visual hierarchy. The lab slip emphasizes a row and footnote; the device summary emphasizes a time window and coverage; the portal card emphasizes a trend assembled from selected records.
Before interpreting anything, make a margin list with seven headings: unit, window, coverage, calculation, legend, limitation, and version. Copy each note under the heading that best describes its function. If a note fits two roles, record both.
Route stop 2
Anchor Every Number to a Unit
Illustrative data, not patient results.A number without a unit is incomplete. A1C is commonly reported as a percentage in the United States. Glucose values commonly use milligrams per deciliter, written mg/dL. Device summaries may also display percentages for time in a defined range, coefficients of variation, or data captured. Those percentages answer different questions. Never compare two percent signs until you have read the labels attached to them.
In the fictional lab slip, 6.4% is an illustrative A1C formatting example. In the fictional sensor summary, 82% means the proportion of expected readings present, not an A1C value and not time in range. In the portal card, 3 entries means the number of selected laboratory documents, not the number of days. Illustrative data, not patient results. The exercise is intentionally designed so similar-looking numerals have different meanings.
Check whether the axis, table header, tooltip, or legend supplies the unit. On a phone, scroll only within a table container and verify that the header still belongs to the visible column. When comparing sources, do not convert or round the original. Preserve the reported unit and ask for qualified help if a conversion is actually needed.
| Document | Illustrative display | Meaning in this example |
|---|---|---|
| Lab slip | 6.4% | A1C row formatting |
| Device summary | 82% captured | Expected readings available |
| Portal trend card | 3 entries | Documents selected for display |
Route stop 3
Find the Observation Window and the Clock Rules
A summary needs a start and end. Fourteen days, thirty days, and ninety days can produce different views even when they end on the same date. A calendar label may use local time, device time, or the portal's account setting. Travel, daylight-saving transitions, disconnected devices, and delayed synchronization can complicate timestamps. Read the report's stated convention rather than assuming every midnight means the same boundary.
The fictional sensor summary covers June 1 through June 14, 2026, and says days begin at 12:00 a.m. in the device's selected time zone. The portal card covers the three most recent finalized lab documents, regardless of elapsed time. The lab slip represents one specimen collected on June 10. These are a continuous device window, a document-selection rule, and a single event. Placing them on one page would not turn them into a common timeline.
Write the window in words: what starts it, what ends it, and what qualifies for inclusion. If the report does not state the rule, label it unknown. Do not fill the gap with a guess based on graph width.
Route stop 4
Read Completeness Before Reading the Average
An average describes available values, not values that were never recorded. A device summary may call completeness sensor use, active time, data captured, wear time, or another defined term. The denominator matters: percent of calendar time, expected readings, or active sessions are not necessarily the same. Look for gaps in the plot and for text describing minimum data requirements. A smooth summary card can conceal an uneven pattern of missingness.
In the educational device example, 82% of expected readings are present. The missing periods occur mainly overnight on four dates. That pattern could matter more than the overall percentage because the absent times are not evenly distributed. The guide does not invent the values that would have appeared. It marks those periods as no data. Filling gaps with a straight line would visually imply observations that do not exist.
Ask two questions before using any aggregate: how much is missing, and when is it missing? Then check whether the report describes exclusions, warm-up periods, errors, or manual entries. Completeness can affect confidence in a picture, but it does not explain why data are absent.
Route stop 5
A Worked Example
Illustrative data, not patient results.Educational example. A reader sees a fictional headline card that says Average 142 mg/dL. The large type looks complete, but its linked note says June 1–14, device time; 82% of expected readings captured; manual entries excluded; summary generated June 16. A dashed gap symbol appears on the chart but is defined only in the legend. Reading the card alone would miss the observation window, data coverage, exclusion rule, and version date.
The reader rewrites the statement without interpretation: The report displays an average of 142 mg/dL for available included readings from June 1 through June 14, with 82% of expected readings captured; manual entries were excluded. Illustrative data, not patient results. This sentence is longer because it carries the boundaries with the value. It still does not say whether the average is desirable, explain missing nights, or compare the metric with A1C.
Next, the reader matches the dashed symbol to no data, checks that the y-axis uses mg/dL, and records that the report was generated two days after the window. The result is a faithful description, not a clinical conclusion.
| Boundary | Fictional note | Why it travels with the average |
|---|---|---|
| Unit | mg/dL | Defines the scale |
| Window | June 1–14, device time | Defines included dates and clock |
| Coverage | 82% of expected readings | Shows that the series is incomplete |
| Exclusion | Manual entries excluded | Defines part of the calculation set |
| Version | Generated June 16 | Identifies the report copy |
Route stop 6
Separate Measured, Calculated, and Imported Fields
Reports may place measured and calculated quantities side by side. A laboratory A1C is a measured assay result. Estimated average glucose, or eAG, is calculated from A1C when reported that way. A device average is calculated from included device readings. A portal trend may import results from several sources. Similar labels do not make these values direct substitutes. Find words such as measured, calculated, estimated, derived, imported, self-entered, or corrected.
The fictional portal card includes one manually entered document and two imported final reports. Its note says the trend line connects displayed entries for navigation and does not interpolate laboratory values between dates. That sentence limits what the line means. Without it, a sloping segment could look like continuous measurement. The visual joins document points; it does not show the path between them.
Record provenance, which means where a field came from. If a source is unclear, do not assign one. Keep the original label and use the source system's help material. When clinical decisions depend on a distinction, bring the original documents to a qualified healthcare professional.
Route stop 7
Route Method, Specimen, and Version Notes Correctly
A lab note may identify specimen type, handling, assay method, interference, or a condition under which a result needs caution. A device note may identify software version, sensor model, compatible data sources, rounding, or an algorithm change. A portal note may describe imports and refresh timing. Keep each note with its originating document. A device limitation should not be pasted onto a laboratory assay, and a laboratory footnote should not automatically be applied to a portal-generated trend.
Look for amendment and generation dates. A newer report can use a changed calculation or corrected source data even when the visible headline is similar. Preserve both versions and compare the notes before comparing numbers. If a vendor explains that an algorithm changed, find the exact effective version and metric affected; do not assume all historical displays were recalculated.
Official NIDDK guidance explains important A1C limitations, including factors involving red blood cells and hemoglobin. That supports careful reading of relevant notes. It does not let a reader decide that an interference occurred in an individual case.
Route stop 8
What It Does Not Tell You
Small print can define a metric without establishing that the metric is accurate for every purpose. It does not reveal missing values, validate a device against a laboratory, diagnose a condition, or explain an individual's pattern. A report generated successfully is not necessarily complete. A high coverage percentage can coexist with strategically important gaps. A laboratory reference interval is not automatically a personal target, and an app's color band is not a medical instruction.
The seven-heading method is a reading tool, not a substitute for manufacturer instructions, laboratory guidance, or professional care. When two reports disagree, preserve their units, windows, source, methods, and versions instead of choosing the more attractive display. Ask the issuing organization about its definitions and a qualified healthcare professional about clinical meaning.
- Carry units and time windows with every copied value.
- Describe missing data as missing; do not invent the line.
- Keep calculated, measured, and imported values distinct.
- Attach each limitation to the field it qualifies.
- Treat version details as part of the evidence trail.