HCTZ vs Chlorthalidone -- A Win for Practicing Doctors and Science

John M. Mandrola, MD


November 05, 2022

U.S. experts favor chlorthalidone over hydrochlorothiazide (HCTZ) in the first-line treatment of hypertension. Yet practicing doctors are approximately 20-fold more likely to choose HCTZ over chlorthalidone. 

During the opening late-breaking clinical trials session here at the American Heart Association (AHA) Scientific Sessions 2022 in Chicago, we learned that practicing doctors had it right.

The Diuretic Comparison Project

The Diuretic Comparison Project (DCP) set out to compare cardiovascular outcomes with the two agents in patients treated with hypertension.

One rationale for comparing these drugs in a head-to-head fashion is the tension between what experts recommend for hypertension and what doctors actually prescribe. A more important one is that even a small difference in outcomes in a condition as common as hypertension would be a big finding.

The U.S. experts cite three main reasons for favoring chlorthalidone: it was the drug used in the seminal outcomes trials of hypertension; some (but not all) observational studies and meta-analyses favor chlorthalidone over HCTZ, and chlorthalidone has a longer half-life and may have favorable pleiotropic effects. (The European Society of Cardiology guidelines don't make a distinction between the two agents.)

Randomizing in the EHR

DCP is not your normal randomized controlled trial (RCT). Primary investigator Areef Ishani, MD, and colleagues in the Veterans Administration (VA) system used a centralized process that was embedded within standard care.

They first used the electronic health record (EHR) to find eligible patients: those taking HCTZ who had a systolic blood pressure greater than 120 mm Hg. They then called both patients and their clinicians to obtain consent. One group continued to receive HCTZ and the other was switched to chlorthalidone.

The primary outcome was the first occurrence of a major cardiovascular outcome (stroke, myocardial infarction [MI], noncancer death, hospitalization for heart failure, or urgent coronary revascularization). Consistent with a VA population, enrolled patients were mostly male and had a mean age of 72 years. 

The results were clear: after 5 years of follow-up, there was no difference in the primary major adverse cardiovascular event  outcome (hazard ratio, 1.04; 95% CI, 0.94 – 1.16). No component of the primary endpoint differed significantly. Hypokalemia occurred in 6% of the chlorthalidone group vs 4.4% of the HCTZ group—a difference that met statistical significance. Potassium levels below 3.1 mEq/L were significantly more common in the chlorthalidone group.

The results held up in most of the prespecified subgroups except one. In the more than 12,000 patients without a history of MI or stroke, there was a 12% increase in the primary endpoint with chlorthalidone, but in the 1700 patients with a history of MI or stroke, patients assigned to chlorthalidone had a 27% reduction in the primary endpoint. This interaction met the threshold of statistical significance.  


The results confirm the current practice of favoring HCTZ. No difference in hard outcomes plus the higher rate of hypokalemia in the chlorthalidone arm means that HCTZ is the preferred agent.

If you are tempted to make much of the MI/stroke history interaction, first read the ISIS-2 trial and its subgroup finding based on astrological sign. Subgroup findings in trials with null primary endpoints should be considered noise, not signal, until proven otherwise.

A naysayer might argue that because DCP won't change their practice (I already use HCTZ ), that it's no big deal.

Even a cursory survey of the history of medicine would render this as wrong thinking. It is absolutely critical to confirm our beliefs about standard care in proper trials.

In their rationale paper, the investigators wrote that more than 50 million prescriptions were filled for HCTZ in 2016. If DCP had shown a benefit to chlorthalidone, it would have been a massive reversal of practice. Practicing doctors got this one right, but chalking it up to wisdom would be ahistorical.

The other reason to laud the DCP trial is how it was accomplished. In fact, the methods may be its biggest contribution to medical science.

This was a pragmatic trial of patients who were going to take one of the thiazide-like diuretics for hypertension. The investigators changed the decision by simply embedding randomization into regular clinical practice.

I think about this almost every day in clinic when I treat a condition in which there are many choices and clear equipoise. What if we had a structural framework to randomly assign this patient to a therapy? In a few years we would know which one is better. Making such processes the norm would surely advance knowledge.

Pragmatic trials aren't perfect. There will be more noise than in a carefully controlled RCT that enrolls highly selected patients. But the trade-off of having more useful results may be worth it.

Whenever we try to apply results of a typical RCT, we have to consider how similar our patient is to the trial participants. When a trial is embedded in normal practice, we might be more confident in translating its results.


I look forward to the discussion of DCP when it is published.

It was a fun trial to cover because it not only confirms a current practice , but also has the potential to provide a way forward for better aligning practice with evidence. And what could be more important than that?

John Mandrola practices cardiac electrophysiology in Louisville, Kentucky, and is a writer and podcaster for Medscape. He espouses a conservative approach to medical practice. He participates in clinical research and writes often about the state of medical evidence. 

Follow John Mandrola on Twitter

Follow | Medscape Cardiology on Twitter

Follow Medscape on Facebook, Twitter, Instagram, and YouTube



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.