# How Many People to Treat to Help One

By Nora Bissett, Pricing Editor (https://glpwatchdog.org/authors/nora-bissett) | Published: 2026-08-20 | Updated: 2026-10-31

Canonical page: https://glpwatchdog.org/research/how-many-people-to-treat-to-help-one

> Trials report benefit as a percentage, which sounds impressive and tells you nothing about your own odds. FLOW published the number needed to treat instead: 22, 13 and 17 over three years.

## Key points

- NNT over 3 years to prevent one kidney outcome: 22 (ASCVD), 13 (heart failure), 17 (high CV risk).
- Most trials publish a relative reduction instead, which always sounds larger.
- Several subgroup intervals cross 1, as expected when you halve a trial's events.
- Interaction P values of 0.62, 0.40 and 0.99: the subgroups showed no sign of differing.
- These NNTs do NOT transfer to a weight-loss population. Everyone here had diabetes AND kidney disease.

Trials almost always report benefit as a percentage reduction, which sounds impressive and tells you nothing about your own odds. The FLOW cardiovascular analysis did something rarer. It published the **number needed to treat**, which is how many people take the drug for three years so that one of them avoids the outcome.[1] The answers were 22, 13 and 17.

## What number needed to treat means

A 26% reduction in risk is a ratio. It describes how much the odds shift, not how likely the event was to begin with. Cutting a 40% risk by a quarter is a very different proposition from cutting a 0.4% risk by a quarter, and the same percentage describes both.

Number needed to treat collapses that into one figure: treat this many people for this long, and one of them will avoid the event who otherwise would not have. It answers the question a patient is asking, and it is the statistic most papers leave out.

> **Low numbers are good.** An NNT of 13 means one person in thirteen benefits. For comparison, in a 2022 review for the US Preventive Services Task Force, statins taken by adults with no prior heart attack or stroke lowered the risk of a heart attack by 0.85 percentage points over six months to six years of follow-up, which works out to roughly 120 people treated for each heart attack avoided.[2] Context matters more than the raw figure, but 13 to 22 for a hard kidney outcome over three years is a strong showing.

## The trial and the subgroups

Everyone in FLOW had two diagnoses at once, type 2 diabetes and established kidney disease, and half of them drew weekly semaglutide at 1.0 mg. Its primary outcome combined serious kidney deterioration (a halving of filtration rate, filtration falling below 15, dialysis or transplantation) with kidney or cardiovascular death. This analysis split the trial by cardiovascular status at entry.

*FLOW primary kidney outcome by baseline cardiovascular status, with the number needed to treat over three years.[1]*

| Subgroup | With the condition | Without it | P interaction | NNT |
| --- | --- | --- | --- | --- |
| Atherosclerotic CV disease | HR 0.80 (0.63–1.02) | HR 0.74 (0.62–0.89) | 0.62 | **22** |
| Heart failure | HR 0.67 (0.49–0.93) | HR 0.79 (0.67–0.93) | 0.40 | **13** |
| High total CV risk (PREVENT ≥20%) | HR 0.73 (0.58–0.91) | HR 0.73 (0.49–1.08) | 0.99 | **17** |

All-cause death moved in the same direction throughout, with hazard ratios between 0.71 and 0.82 across every subgroup and no sign of heterogeneity.

## Why some intervals cross 1 and it does not matter here

Read down the hazard ratio columns and several intervals include 1: atherosclerotic disease at 0.63 to 1.02, all-cause death in the heart-failure group at 0.54 to 1.05. Taken alone, each of those says “not established.”

That is exactly what splitting a trial does. Halve the participants and you halve the events, which widens every interval regardless of whether the underlying effect changed. Subgroup intervals crossing 1 in a trial whose overall result was positive is the expected picture, not a warning sign.

> **The column that matters is the interaction P, and here it is reassuring in the opposite direction from usual.** At 0.62, 0.40 and 0.99, these tests find no evidence that the subgroups behave differently from one another. The right conclusion is that the drug worked and the split found nothing, not “it works in heart failure and maybe not in atherosclerotic disease.”

That is the mirror image of a case we covered in [the statistic that asks if groups differ](https://glpwatchdog.org/research/the-statistic-that-asks-if-groups-differ), where an interaction P of .06 sat under a table that looked like a clean split. Same test, opposite reading, and the pair is more instructive than either.

## Why you cannot carry these numbers elsewhere

> **An NNT is not a property of a drug. It is a property of a drug in a population.** Every participant carried diabetes *and* failing kidneys together, a group already close to the outcome being counted. In a lower-risk population, the same drug with the same relative effect produces a much larger NNT, because there are far fewer events available to prevent.

So these figures say something precise and narrow: for people with diabetes and kidney disease, treating between 13 and 22 for three years prevents one serious kidney outcome. Applied to someone taking a GLP-1 for weight without kidney disease, they would be fabricated. We are not going to extrapolate them and nobody else should either. The same limit applies to the death results quoted as proof of longevity; see [are GLP-1s longevity drugs?](https://glpwatchdog.org/research/glp1s-as-longevity-drugs).

The dose matters too: semaglutide 1.0 mg, the type 2 diabetes strength, not the 2.4 mg used for weight management. For the kidney evidence more broadly, see [GLP-1s and your kidneys](https://glpwatchdog.org/research/glp1-and-your-kidneys).

## Frequently asked questions

### What is number needed to treat?

How many people must take a treatment, for a stated length of time, for one of them to avoid the outcome being measured. Lower numbers mean more people benefit. It answers a patient's question in a way a percentage reduction does not.

### What were the numbers in FLOW?

To prevent one primary kidney outcome over three years: 22 in the group with atherosclerotic cardiovascular disease, 13 in the group with heart failure, and 17 in those at high total cardiovascular risk without established disease.

### Do those numbers apply to someone taking a GLP-1 for weight loss?

No. Every participant carried diabetes and kidney disease together, which puts them at far higher risk than a typical reader. The same drug in a lower-risk population produces a much larger number needed to treat, because there are fewer events available to prevent.

### Some of the confidence intervals cross 1. Is that a problem?

Not here. Splitting a trial into subgroups halves the events and widens every interval. The test that matters is whether the subgroups differ from each other, and at P values of 0.62, 0.40 and 0.99, they showed no sign of differing.

### Which dose was used?

Semaglutide 1.0 mg weekly, the dose used for type 2 diabetes, not the 2.4 mg strength used for weight management.

## References

1. Tuttle KR, Bakris GL, Baeres FMM, et al. Kidney and Survival Benefits of Semaglutide in Diabetes With Chronic Kidney Disease: FLOW Trial Cardiovascular Subgroup Analyses. Journal of the American College of Cardiology. 2026. PMID: 42233552. https://pubmed.ncbi.nlm.nih.gov/42233552/
2. Chou R, Cantor A, Dana T, et al. Statin Use for the Primary Prevention of Cardiovascular Disease in Adults: Updated Evidence Report and Systematic Review for the US Preventive Services Task Force. JAMA. 2022. PMID: 35997724. DOI: 10.1001/jama.2022.12138. https://pubmed.ncbi.nlm.nih.gov/35997724/

---

Cite as: GLP Watchdog, Nora Bissett, "How Many People to Treat to Help One", published 2026-08-20, updated 2026-10-31. https://glpwatchdog.org/research/how-many-people-to-treat-to-help-one

Canonical HTML page: https://glpwatchdog.org/research/how-many-people-to-treat-to-help-one

Medical disclaimer: Nothing here is medical advice. We report on this subject; we do not treat anyone. Whether to begin a medication, supplement, or treatment, whether to adjust it, and whether to stop are questions for a clinician licensed to answer them, who knows your history.
