Did the medication actually do anything?

You start a new medication. A few weeks later a lab comes back, and the number has moved. It is tempting to draw the obvious conclusion: it worked, or it did not.

Sometimes that conclusion is right. Often it is built on one data point, drawn at the wrong time, compared against a baseline that was itself unusual. A before-and-after comparison is some of the most useful evidence you can bring to a follow-up visit, but only if you read it honestly.

Here is what to check before you believe it.

Was it long enough?

Every drug and every lab marker has its own clock, and the two clocks are not the same.

Some markers respond in days. Others reflect weeks or months of history, so a draw taken too soon mostly measures the time before you started:

• HbA1c reflects average blood glucose over roughly the previous three months, because it measures glucose attached to red blood cells that circulate for about that long. A draw four weeks after starting a diabetes medication is still mostly the old you.
• TSH after a thyroid hormone dose change is usually rechecked after six to eight weeks, the interval the American Thyroid Association guidelines describe for the level to settle.
• Lipids after starting or increasing a statin are rechecked at four to twelve weeks under the 2026 ACC/AHA dyslipidemia guideline.
• Ferritin and iron stores take months to rebuild. British Columbia’s iron deficiency guideline, for example, rechecks ferritin three to six months after starting supplements in people who are not anemic.

If you are not sure what the right interval is, ask the prescriber when they expect to see a change, and write the answer down. A “no change” result drawn before that date is not evidence of anything.

Is the change bigger than the noise?

Lab values bounce around in everyone, even when nothing is happening. Some of that is the assay. Most of it is you: hydration, time of day, recent meals, exercise, sleep, a minor infection, the season.

Researchers call this biological variation, and it has been measured for hundreds of analytes. The EFLM Biological Variation Database publishes the figures. For some tests, day-to-day swings of 10 to 20 percent in a healthy, stable person are ordinary. A change of that size after starting a medication might be the drug, or it might be Tuesday.

This is where a long history earns its keep. If you have ten results for the same test from before the medication, you can see your own spread. A new value that sits well outside it is interesting. A new value that sits inside it is not yet a signal, whatever direction it moved.

Was the “before” value normal for you?

Medications tend to get started right after an unusually bad result. That timing creates a trap.

An extreme value is partly real and partly a bad day. On the next draw, the bad-day portion tends to fade on its own, and the number drifts back toward your usual level whether or not anything else changed. Statisticians call this regression to the mean. It makes almost any treatment started after a spike look effective.

The fix is to compare against your typical level before the medication, not against the single worst value that prompted it. That is only possible if you have more than one “before” point.

Did anything else change at the same time?

A new prescription often arrives with a new resolve. People start a statin and also start walking. They begin a thyroid medication and also stop drinking. They switch labs because they moved.

Each of those can move a number on its own. Some common ones:

• Diet, weight, alcohol, and exercise changes
• A second medication or supplement started or stopped in the same window
• Fasting versus non-fasting draws, or morning versus afternoon
• A different lab, with a different assay and reference range
• An illness or injury near the draw date

None of this means the medication did nothing. It means the before-and-after chart is showing the combined effect of everything, and you should say so when you discuss it. Where you have a choice, changing one thing at a time makes the result much easier to read.

Did you actually take it?

This is the question people least like to ask, and the one that explains the most flat trend lines.

Missing doses is not a character flaw. It is the normal human condition. The World Health Organization’s review of the evidence put adherence to long-term therapy for chronic conditions at about 50 percent in developed countries. Doses get skipped on weekends, forgotten when travelling, stretched when the bottle runs low, and quietly dropped when a side effect shows up.

If you missed a week of doses before a draw, the result reflects that week. If you took the medication at a different time of day than intended, some markers will show it. Before concluding that something does not work, be honest about how consistently it went in.

Two habits help:

• Get reminded at the time of the dose. A reminder at the moment you are supposed to take something does more than a good intention in the morning. If you want reminders and a simple checklist of doses taken, the Chickadee Bandit Medication Tracker, a companion app from us, handles daily schedules for everyone in a household, with a nudge at each scheduled time and a heads-up to a family member when a dose is missed.
• Record the misses. When a stretch goes badly, write it down with the dates. A note that says “missed most doses, week of the 10th” turns a confusing lab result into an understandable one.

Reading the chart honestly

Put together, a fair reading of a before-and-after looks like this:

1. Several “before” values, so you know your usual level and spread.
2. At least one “after” value drawn once the marker has had time to respond.
3. A change larger than your own normal variation.
4. A note of anything else that changed in the same window, including missed doses.
5. Ideally, a second “after” value confirming the first.

If you have all five, you have real evidence to discuss. If you have fewer, you still have something worth raising, but frame it as a question rather than a verdict. And whatever the chart shows, do not stop, start, or change a dose on your own because of it. Bring it to the person who prescribed it.

Why we build this

This kind of reading needs your whole history in one place, with the medication dates on the same timeline as the lab values. No single portal gives you that.

In HealthViewer, you record each medication with its start and stop dates in Medications, and those dates appear as markers on your Trends charts. The Medication Effect report lays out every test before and after a medication’s start date, side by side. The Journal holds the context the numbers cannot, like a missed week or a new exercise habit, and shows it on the same chart. Everything stays in an encrypted file on your own device.

This post is general information, not medical advice. Monitoring intervals and expected responses vary by drug, dose, and person. Talk to your prescriber before changing how you take any medication.