A national database study has compared short-term outcomes after total knee arthroplasty in people with type 2 diabetes who had active prescriptions for tirzepatide or semaglutide before surgery. After matching the two groups on key baseline factors, the researchers found no statistically significant differences in medical complications at 90 days, surgical complications at 180 days, or emergency-department visits and readmissions. The analysis is a useful comparative snapshot, but it does not establish that either medicine caused a particular postoperative outcome.
The study addresses a practical research question. Both tirzepatide and semaglutide are increasingly present in records of adults with type 2 diabetes who undergo joint replacement. Yet direct comparisons between the two around the time of surgery have been sparse. The authors used a retrospective research-network design to compare people already recorded as receiving one or the other, rather than assigning a treatment in a clinical trial.
How the comparison was built
The researchers used the TriNetX research network and identified adults with type 2 diabetes who had primary total knee arthroplasty between 1 June 2022 and 31 December 2024. Eligibility required an active prescription for semaglutide or tirzepatide within 90 days before the procedure. The database analysis then used 1:1 propensity-score matching to make the two cohorts more comparable.
Matching included age, sex, race, body-mass index, glycated haemoglobin, comorbidity burden and use of other diabetes medicines. That process produced 415 matched pairs, or 830 people in total. It is an important methodological step because prescription groups can differ before an operation in ways that also affect later complications. Matching reduces measured differences, but it cannot remove every unrecorded difference between patients.
The outcomes were assessed through 90 and 180 days. They included medical and surgical complications, as well as emergency-department visits and readmissions. The authors estimated odds ratios with 95% confidence intervals using logistic regression. This is an observational comparison of coded clinical data, not a randomised head-to-head trial of tirzepatide and semaglutide.
What the authors found
After matching, the study found no statistically significant difference in 90-day medical complications. The odds ratio was 1.122, with a 95% confidence interval from 0.736 to 1.710 and a p value of 0.591. At 180 days, surgical complications were also statistically similar between the cohorts, with an odds ratio of 1.632, a confidence interval from 0.845 to 3.152 and a p value of 0.141.
The authors also report statistically similar individual-event results for myocardial infarction, stroke, pneumonia, sepsis, pulmonary embolism, deep-vein thrombosis, acute kidney injury, urinary-tract infection, surgical-site infection, periprosthetic joint infection, wound dehiscence, mortality and revision. For example, the acute-kidney-injury estimate was 0.867, with a 95% confidence interval from 0.417 to 1.801. The broad intervals are important: they show that several event estimates remain imprecise.
Emergency-department visits and readmissions were likewise comparable. At 90 days, the reported odds ratio was 0.775, with a confidence interval from 0.570 to 1.052. At 180 days it was 0.887, with a confidence interval from 0.672 to 1.170. In both cases, the interval crosses 1, so the data do not support a statistically significant difference between the matched groups on those outcomes.
Why an observational result needs restraint
The paper’s main value is comparative: in this matched dataset, the authors did not detect a difference between tirzepatide and semaglutide across the defined short-term postoperative endpoints. It does not prove that the two agents are interchangeable in every surgical setting. The data come from routine records and depend on the completeness of coding, prescription information and outcome capture across participating organisations.
Low event counts further limit precision for several individual outcomes, as the authors note. An estimate can be statistically non-significant because two groups are genuinely similar, but it can also be non-significant when the available data cannot distinguish a meaningful difference from random variation. The confidence intervals, not only the p values, are therefore central to reading this paper.
The study is also limited to adults with type 2 diabetes undergoing primary total knee arthroplasty within the stated time window. It does not answer questions about other operations, adolescents, different metabolic conditions, or longer-term outcomes. It also cannot determine whether a change in medication status, timing or perioperative management would alter an individual’s outcome.
Research context, not a protocol
Tirzepatide is a dual GIP/GLP-1 agonist, while semaglutide is a GLP-1 receptor agonist. Their different pharmacology makes direct comparisons scientifically useful, but product or research context must not be mistaken for clinical direction. The Tirzepatide research buyer guide is a research-material resource and does not interpret this study as a perioperative protocol.
Other incretin-related work should likewise be evaluated on its own evidence. Retatrutide is a distinct research compound, and the Retatrutide research reference hub describes its separate research context. Neither source changes the study’s direct comparison of tirzepatide and semaglutide after knee arthroplasty.
Peptra Labs materials sit within a Research Use Only boundary. Readers looking for broad background can consult Best Research Peptides 2026, but a site overview is not a substitute for study-specific evidence.
The measured conclusion is straightforward. In 415 matched pairs, the authors found similar short-term complication and healthcare-utilisation profiles after total knee arthroplasty. The results offer a current database signal for further study, while their observational design and imprecise estimates for some events mean that they should not be read as a universal clinical recommendation.
References
- Wu KA, Choudhury A, Wu JA, et al. Postoperative outcomes after total knee arthroplasty in type 2 diabetes mellitus patients receiving semaglutide versus tirzepatide: a propensity score-matched national research network analysis. The Knee. 2026;63:104639. PubMed
- Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behavioral Research. 2011. Article
- U.S. National Library of Medicine. Type 2 diabetes mellitus. MedlinePlus
The products referenced on this site are supplied for laboratory research use only. They are not medicines and are not intended for human or veterinary use. This article summarises published research for informational purposes and is not medical advice. Statements about third-party studies belong to their authors.
