{"id":4261,"date":"2026-10-12T08:32:27","date_gmt":"2026-10-12T08:32:27","guid":{"rendered":"https:\/\/nhdbio.com\/?p=4261"},"modified":"2026-10-12T08:32:27","modified_gmt":"2026-10-12T08:32:27","slug":"evidencia-cardiovascular-sobre-el-cambio-de-semaglutida-a-tirzepatida-con-escalada","status":"publish","type":"post","link":"https:\/\/nhdbio.com\/es\/semaglutide-escalation-tirzepatide-switch-cv-evidence\/","title":{"rendered":"Aumento de dosis de semaglutida o cambio a tirzepatida: c\u00f3mo interpretar los datos cardiovasculares"},"content":{"rendered":"
\n
\"Thematic
Editorial illustration, not an experimental photograph.<\/figcaption><\/figure>\n

"Cardiovascular risk lower by 6%" can easily make people think a certain drug is better in all situations. In fact, this study compared a very specific treatment path: how people who were already using semaglutide adjusted afterward.<\/p>\n

\n

The starting point of the study is that everyone was already using a drug<\/h2>\n

The COMPETE SWITCH CV analysis announced on September 29 organized insurance claims data for 636525 adults with type 2 diabetes. They were all using semaglutide 1 milligram at the start.<\/p>\n

Afterward, some maintained the original regimen, some increased the semaglutide dose, and some switched to tirzepatide. The study focused on the two paths of dose increase and switching, not on people who had never used the drug choosing a drug for the first time.<\/p>\n<\/section>\n

\n

What events are meant by cardiovascular events?<\/h2>\n

The major adverse cardiovascular events here include all-cause death, myocardial infarction, and stroke. These are different indicators from how much blood sugar decreased or how much weight was lost.[1]<\/a><\/p>\n

If a treatment performs better on weight, that does not automatically mean it must also perform better on cardiovascular outcomes in another population. When evaluating a study, first look at what it actually measured.<\/p>\n<\/section>\n

\n

A 6% relative reduction does not mean 6 fewer events per 100 people<\/h2>\n

The company announcement said that increasing the dose to semaglutide 2 milligrams was associated with lower event risk, reporting a relative difference of about 6%. A relative difference needs to be understood together with the original likelihood of the event in order to grasp the actual impact.[1]<\/a><\/p>\n

Give a purely mathematical example: if an event originally had a risk of 10%, a relative reduction of 6% would make it 9.4%, a difference of 0.6 percentage points. This is only an explanation of the calculation, not the actual incidence rate in this study.<\/p>\n<\/section>\n

\n

Why can't retrospective data yet determine the winner?<\/h2>\n

Doctors do not randomly decide who increases the dose and who switches drugs. Blood sugar control, weight changes, past medical history, cost, and tolerability may all affect the choice, and these people may already have been different.<\/p>\n

Statistical adjustment can handle some recorded differences, but it cannot guarantee that all differences are eliminated. Therefore, the safer statement is that "the results of these two paths are associated," not that "switching drugs caused the risk to rise."<\/p>\n<\/section>\n

\n

When reading comparative news, first check three conditions<\/h2>\n

First look at the population: here it is people with type 2 diabetes who were already using semaglutide. Then look at the comparison method: here it is retrospective records. Finally, look at the endpoint: here it focuses on cardiovascular events, not only on weight.<\/p>\n

If any of these three conditions changes, the answer may change. Putting this result side by side with another obesity weight-loss trial and ranking them based on a single percentage is usually unfair.<\/p>\n<\/section>\n

\n

If the study has a very large number of people, can selection bias be ruled out?<\/h2>\n

A large sample usually makes estimates more precise, but it cannot automatically make the two groups the same. Even with hundreds of thousands of people, if the reasons doctors chose to switch drugs are related to disease risk, the comparison may still be affected.<\/p>\n

You can record both the size of the difference and the study design. A small number with a precise estimate does not mean causation has been proven; a large number with a wide uncertainty range also needs to be understood carefully.<\/p>\n<\/section>\n