A community health worker programme did not prevent diabetes in prediabetes

Design

Pragmatic cluster-randomised trial

Population

Adults aged 18–75 with HbA1c 5.7–6.4% in two New York primary care systems

Primary outcome

Incident diabetes up to 6 months after the 12-month programme

Effect

5.9% vs 4.3%; risk difference 1.6 points (95% CI −3.7 to 4.1)

This pragmatic cluster-randomised trial assigned primary care teams at a Veterans Affairs system and a public hospital in New York to a 12-month community health worker programme or usual care. Adults aged 18 to 75 with an HbA1c of 5.7–6.4% received health action plans, motivational interviewing, education and community referrals.

In the six months after the programme, diabetes developed in 5.9% of the intervention group and 4.3% of controls, a difference that was not significant. HbA1c, weight, blood pressure and lipids did not differ. The intervention group had more primary care visits (78% vs 63%) and more emergency department visits (35% vs 29%).

Community health workers are central to Indian primary care through ASHAs and similar roles, and they clearly improve engagement. This trial shows that engagement alone, delivered without a structured lifestyle curriculum, did not change glucose in a year.

The structured diabetes prevention programmes that do work — intensive, curriculum-based, with weight targets — differ from this lighter model. The lesson is about content and intensity, not about abandoning health workers.

• Refer people with prediabetes to a structured, curriculum-based lifestyle programme where one exists, rather than general support.
• Set a concrete weight-loss and activity target with each patient, since these drive prevention.
• Use community health workers to bring patients back for review, not as the whole intervention.
• Recheck HbA1c at least yearly in prediabetes, whatever programme the patient is in.

Why it matters

It separates getting patients through the door from changing what happens to their glucose.

Don't overread it

A null result for this light-touch model does not show that intensive diabetes prevention programmes fail.

The statistics, in plain English

The risk difference of 1.6 points had a 95% confidence interval from −3.7 to 4.1, which includes both a small benefit and a small harm, so the trial shows no effect rather than proving harm. Follow-up for diabetes was only six months after the programme, which may be too short to see a difference.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.