Diabetes kidney prevention: German model

A 2026 German cost-effectiveness model compared standard type 2 diabetes care with added empagliflozin, semaglutide, or tirzepatide for preventing kidney disease. Semaglutide produced the highest modeled net monetary benefit at the authors' chosen threshold. Tirzepatide produced the largest modeled health gain but cost too much to be cost-effective in the base case^[1]. These are simulated lifetime estimates, not outcomes observed in a head-to-head trial.

How the German model worked

The researchers built a Markov microsimulation from the perspective of Germany's statutory health system. A Markov model moves simulated patients between defined health states over time. Here, five states represented increasing kidney damage. Transition probabilities, treatment effects, costs, and quality-of-life values came from published literature rather than a new patient cohort.

The primary simulated cohort started at age 62.8 years, with another scenario starting at 50. Future costs and health outcomes were discounted by 3% per year. The model valued one additional quality-adjusted life year, or QALY, at €100,000^[1]. A QALY combines length and quality of life into one economic measure. The threshold is a modeling choice and not an official German reimbursement rule that settles every decision.

Semaglutide led on value in the model

Standard care generated 10.28 discounted QALYs and €66,343.64 in direct costs. Adding empagliflozin raised the estimate to 10.48 QALYs and €79,767.42. Added semaglutide produced 10.58 QALYs at €84,328.59. The corresponding net monetary benefits were €961,445.98 for standard care, €968,406.95 for empagliflozin, and €973,647.74 for semaglutide^[1]. Under those assumptions, semaglutide gave the best value of the compared strategies.

The result depended most on the quality-of-life value assigned to the type 2 diabetes state and on drug cost. That sensitivity finding matters because both inputs can change. Acquisition prices differ by contract and over time, while a single utility value cannot represent every patient's lived experience. A result that survives the authors' tested scenarios is still conditional on the range they chose to test.

Net monetary benefit converts health gains into euros at the selected threshold, then subtracts cost. It is useful because all strategies can be compared on one scale. It should not be mistaken for money paid to a patient or cash saved by a clinic. A positive figure also does not mean a treatment lowers spending. In this model every added medicine increased direct cost; the question was whether the value assigned to extra QALYs exceeded that increase.

Tirzepatide gained more health at higher cost

Added tirzepatide generated the largest modeled outcomes: 10.75 QALYs and 14.00 years of life expectancy, compared with 10.28 QALYs and 13.36 years under standard care. Total direct cost reached €155,992.59, more than double the standard-care estimate. Net monetary benefit fell to €919,972.95^[1]. More modeled health therefore did not mean better modeled value at the selected €100,000 threshold.

This finding should not be read as a clinical verdict against tirzepatide. Cost-effectiveness asks whether added health is worth added spending under a specified perspective and threshold. A lower price could change the answer. So could different kidney-transition estimates, longer follow-up, or a model that includes outcomes beyond the kidney states used here.

The statutory-health-system perspective also sets a boundary around the result. Patient time, unpaid caregiving, lost work, and broader social costs may matter but do not necessarily appear in direct medical spending. Including them can move an economic estimate in either direction. Results built for Germany should not be transferred unchanged to countries with different prices or treatment pathways.

The clinical evidence behind the assumptions

The model's semaglutide case has direct kidney-outcome support from FLOW. In that randomized trial, 3,533 people with type 2 diabetes and chronic kidney disease were followed for a median of 3.4 years. Weekly semaglutide reduced the risk of the composite primary kidney outcome by 24% relative to placebo, with a hazard ratio of 0.76^[2]. FLOW studied patients who already had chronic kidney disease. It did not compare semaglutide with tirzepatide or empagliflozin on German costs.

For tirzepatide, the kidney evidence cited in this area includes a post-hoc analysis of SURPASS-4. Tirzepatide slowed annual estimated glomerular filtration rate decline compared with insulin glargine and reduced a composite kidney endpoint, but the analysis was secondary to the original trial and many participants did not have advanced kidney disease at baseline^[3]. A post-hoc kidney analysis can inform a model, but it does not carry the same certainty as a dedicated kidney-outcomes trial.

Empagliflozin has randomized evidence from EMPA-REG OUTCOME. Incident or worsening nephropathy occurred in 12.7% of participants assigned empagliflozin and 18.8% assigned placebo among people with type 2 diabetes at high cardiovascular risk^[4]. That trial supports a kidney benefit for the SGLT2 inhibitor class, while the German paper translates published effects into one economic framework.

What the result does not prove

The model does not prove that semaglutide is the best second-line medicine for every German patient. It does not randomize people among the three drugs, measure adherence in routine care, or capture every adverse effect and cardiovascular outcome. It also does not establish that a modeled life-expectancy difference will appear at the same size in practice.

Treatment choices depend on kidney function, cardiovascular history, glucose control, weight goals, tolerability, contraindications, and access. The useful conclusion is narrower: with the paper's prices and assumptions, semaglutide and empagliflozin remained cost-effective across the tested scenarios, while tirzepatide's higher modeled drug cost outweighed its larger health gain. Readers can check the separate tirzepatide evidence page for approved uses and current regulatory context.