org355: Ozempic for children

From The Guardian

1. What is the study design?

2. What are eligibility criteria?

3. What was the primary outcome definition? And secondary?

4. What were the comparison groups?

5. Calculate and interpret a relative risk figure.

6. Calculate and interpret a NNT (number needed to treat).

7. Do you believe that a causal relationship has been confirmed?

Re: Ozempic for children

by | ISHITA GUPTA - Friday, 19 May 2023, 4:26 AM

1.Study design: intervention study/Individual randomized (not mentioned though) control trial
2. Eligibility criteria: children aged between 12 and 18yrs, clinically obese, assuming had parental consent
3. Primary outcome : Proportion of participants no longer being classified as obese
And secondary: Proportion of participants with improvement of at least one BMI category
4. Comparison groups: Intervention: children aged between 12 and 18 received one dose of 2.4mg of semaglutide a week across 68 weeks in addition to healthy lifestyle counselling
Control: children aged between 12 and 18 received a placebo for the same period of time healthy lifestyle counselling.
5. relative risk
I 134 45% 60.3 61 0.455224 3.388889
C 67 12% 8.04 9 0.134328
Children in the intervention group had more than thrice the likelihood of loosing weight to be classified as non-obese if they received semaglutide compared to placebo.

6. NNT (number needed to treat)- NOT SURE OF THIS, SEEK ADVISE
1/(RR-1)*po RR=3.4, po=proportion of outcome in control group=.12
1/(3.4-1)*0.12 =4
7. Do you believe that a causal relationship has been confirmed?
Bradford Hill criteria.
Temporality is there as children were obese and the does of drug was given later to observe weight loss
Magnitude of RR is high, unlikely to be due to chance (need to see CIs)
Biological plausibility, that drug suppressing appetite thus reducing weight.

However,
Need to ascertain adjustment for confounders.
Bias in the study
Blinding
Random allocation
Loss to follow-up ?

Thank you.

Re: Ozempic for children

by | Sujit Rathod - Friday, 19 May 2023, 5:12 PM

Terrific responses Ishita!

The NNT is 1/(risk difference).

Also worth noting is that Bradford Hill are _guidelines_, not criteria. B-H are a set of tools one can use to consider causation in a structured way.

Re: Ozempic for children

by | ANNAAT VAN DER WALT - Friday, 19 May 2023, 1:51 PM

1. What is the study design?

A placebo-controlled intervention study. - RCT.

2. What are the eligibility criteria?

Clinically-classified Obese Adolescents, between 12 and 18 years old.

3. What was the primary outcome definition? And secondary?
Primary outcome: Percentage change in BMI,
Secondary outcome: Reduction in body weight

4. What were the comparison groups?
The intervention(treatment) group: One dose of 2.4mg of semaglutide weekly for 68 weeks.
The control group: A placebo weekly for 68 weeks. (About a year and a half)

5. Calculate and interpret a relative risk figure.
Relative Risk Reduction = 0.74 - 0.19 = 0.55 = 55% ≈ 50%

6. Calculate and interpret an NNT (number needed to treat).
NNT = 1/Risk Difference = 1/0.55 = 1.8 ≈2
A broad interpretation is that for every 2 adolescents treated with semaglutide in the study, 1 adolescent improved at least one BMI category. – About half of the patients treated with semaglutide benefitted.

7. Do you believe that a causal relationship has been confirmed?
Yes, causality has been confirmed.
A randomised placebo-controlled trial design has the best chance of demonstrating causality.

Confirming causality:

The obesity was pre-existing and the BMI reduction was achieved after a year and a half on semaglutide compared to the placebo.

The sample size should be sufficiently large to demonstrate that the BMI reduction did not happen by chance alone. Statistically, a confidence interval excluding zero (Null value for the relative risk reduction) would be a confirmation.

To establish causality, demonstrate that nothing else accounts for the observed relationship. Baseline differences and confounders should have been accounted for by randomisation. The balance between the semaglutide arm and the placebo arm is assumed. Also, both arms received similar lifestyle guidance.

Blinding could limit performance bias. (Investigators potentially giving differential treatment to the trial arms resulting in an over or underestimation of the treatment effect)

Regards,

Annaat