Basal Rate Testing: How to Check Your Pump Settings

What a basal test is actually testing

Basal insulin has one job: hold glucose steady when you are not eating. Everything else — carb ratios, correction factors, pre-bolus timing — is layered on top of that assumption. If the basal rate is wrong, every calculation built on it inherits the error, and you end up adjusting ratios to compensate for a problem that is not in the ratios.

A basal test removes the other variables. You fast, you take no bolus insulin, and you watch what glucose does on basal alone. If it stays flat, the rate is right for that window. If it drifts, the rate is wrong — and the direction and timing of the drift tell you which segment to change.

This is why it comes first. Testing carb ratios before basal is settled produces results that cannot be interpreted.

Test one segment at a time

The day is split into windows and each is tested separately, on different days. Trying to fast through the whole day in one go is both unpleasant and less informative, because a drift late in the day may be the tail of something that started earlier.

Start with the overnight segment. It is the longest uninterrupted fasting window most people have, it is where the dawn phenomenon shows up, and it is the one that most often explains a morning number nobody can account for.

The drift rule

The commonly used threshold is a change of about 30 mg/dL (1.7 mmol/L) across the fasting window.

• Flat within ±30 mg/dL (±1.7 mmol/L) — the basal rate for that segment is doing its job. Leave it alone.
• Rises by more than 30 mg/dL — not enough basal. The usual adjustment is an increase of around 10%, starting roughly two hours before the rise begins.
• Falls by more than 30 mg/dL — too much basal. Reduce by around 10%, again starting about two hours before the fall.

The two-hour offset is the part people miss. Subcutaneous insulin does not act the moment it is delivered, so a rate change made at the time you observe the drift lands too late to prevent it. You are adjusting the rate that was running upstream of what you are seeing.

Change one segment at a time, then retest. Adjusting two windows at once makes the next test uninterpretable, because a change in the earlier segment alters the glucose you start the later one with.

What makes a test unreadable

A basal test is only valid under fairly narrow conditions, and most failed tests fail before they start.

Insulin still on board. Begin at least four to five hours after your last rapid-acting bolus. A correction still working will pull glucose down and look exactly like too much basal. Our insulin on board calculator shows what is still active.

A high-fat meal beforehand. Fat and protein delay gastric emptying and can push glucose up hours later, into what you believe is a clean fasting window.

Exercise, alcohol, illness or a poor night’s sleep. All four move glucose independently of basal insulin, and alcohol in particular can cause delayed lows many hours later.

Starting from an out-of-range number. Beginning above roughly 250 mg/dL (13.9 mmol/L) or below 80 mg/dL (4.4 mmol/L) makes the result meaningless — and starting low is unsafe. Postpone.

A site problem. An infusion set on its third day, or one placed over scar tissue, can absorb erratically. A single strange test that contradicts three consistent ones is more likely a site issue than a basal issue.

When to stop the test

Stop and treat if glucose falls below roughly 70 mg/dL (3.9 mmol/L), or if you feel hypoglycaemic at any level. A test abandoned for a low is not a wasted test — it has told you the basal rate in that window is too high, which is the answer you were looking for.

Stop as well if glucose climbs above about 250 mg/dL (13.9 mmol/L), or if ketones appear. Take a correction and treat the segment as needing more basal, then retest another day.

Repeat each segment two or three times before acting on it. A single night is a data point; a pattern across several is a finding. If you use a CGM the trace makes this far easier to read than fingersticks, though the same interpretation rules apply.

Limitations & Safety Notes

• A basal test is only interpretable while fasting, with no bolus insulin still active and no exercise during the window.
• One test is not a result. Repeat each segment two or three times before changing a rate, and change only one segment at a time.
• Do not test during illness, after alcohol, or when starting from an out-of-range glucose — all three produce misleading curves.
• Overnight testing carries real hypoglycemia risk while you are asleep. Agree the protocol, check times and stop criteria with your care team beforehand.
• Erratic absorption from an old or scarred infusion site mimics a basal problem. Rule out the site before changing settings.

Frequently Asked Questions