The Diuretic Comparison Project (DCP): For hypertensive patients already on hydrochlorothiazide, switching to chlorthalidone does not improve outcomes

By Lucas Marinacci on

Key Points:

  • Thiazide type diuretics are first-line medications for hypertension (HTN); while chlorthalidone (CTD) has been shown to lower blood pressure to a greater extent than hydrochlorothiazide (HCTZ), whether this translates into improved cardiovascular outcomes is unknown.
  • The DCP trial enrolled Veterans Affairs (VA) patients aged 65 or older currently on HCTZ 25-50mg daily and randomized them to either continue taking HCTZ or to switch to an equivalent dose of CTD.
  • There was no difference between the two groups for the primary composite outcome of time to non-cancer death, stroke, myocardial infarction (MI), urgent revascularization, or heart failure hospitalization.
  • The DCP used a novel pragmatic “point of care” design in which the primary care providers (PCPs) managed the medications, no study staff were present at study sites, and the outcome data was collected passively through the electronic medical record (EMR) or national databases, thereby mitigating the cost of study execution.

Thiazide type diuretics are considered one of four acceptable first line medication classes for hypertension (HTN). CTD has been shown to have a longer half-life and to lower blood pressure more effectively than HCTZ, however the benefit in regards to cardiovascular outcomes is unknown.1 Sponsored by the VA Office of Research and Development, the purpose of the Diuretic Comparison Project (DCP) was to determine whether CTD is superior to HCTZ at preventing adverse cardiovascular outcomes in veterans with hypertension, using a novel randomized, parallel, open-label, pragmatic “point of care” study design.The results were presented today by Dr. Areef Ishani (VA Health Care System, Minneapolis).

The DCP included VA patients aged 65 and older already receiving HCTZ from a VA pharmacy at a daily dose of 25 or 50mg with the most recent recorded systolic blood pressure (SBP) >= 120 mmHg and with no recorded SBP < 120 mmHg in the last 90 days. Patients with impaired decision-making capacity, death expected within 6 months as per PCP assessment, lab evidence of hypokalemia in the past 90 days or hyponatremia in the past 30 days were excluded. Patients were then randomized to either continue HCTZ or replace it with an equivalent dose of chlorthalidone. The primary outcome was time to a composite of non-cancer death or first major cardiovascular event after randomization, defined as stroke, MI, urgent revascularization, or heart failure hospitalization. The secondary outcomes included time to event for each individual component of the primary outcome.


The pragmatic point of care study work-flow was as follows: first, the DCP consented VA PCPs to participate. Then the DCP identified patients on the PCP’s panel meeting the study criteria and mailed these patients the informed consent information and obtained their verbal consent over the phone. Once the patient agreed and the PCP assented, the central study group would then randomize the patient. Those randomized to HCTZ would have no change in care. For those randomized to chlorthalidone, the central study group would send an electronic order to discontinue HCTZ and start CTD to their PCP, who would then sign the order, and the CTD would be filled at the local VA pharmacy. All management after randomization was considered usual care and determined by the PCP. Outcomes were assessed using the VA electronic medical record and national databases without any further patient contact by the study team. There was no study staff present at any site at any point. The investigators claimed that the cost saving nature of this design is what was made the DCP feasible.


Overall, 6767 patients were randomized to continue HCTZ and 6756 to switch to CTD. The two groups were well-matched: both were 97% male, 15% Black, with 45% residing in rural areas. They both had mean starting SBP of 139 +- 14 mmHg, and 11% of each group had a prior MI or stroke. The majority of both groups were on more than one blood pressure medication. Over the 5 years following randomization, there was no significant difference in recorded blood pressure values, however patients randomized to CTD were more likely to be hospitalized for hypokalemia (1.5% vs 1.1% p=0.05). There was additionally no difference in the time to primary outcome of non-cancer death or major cardiovascular events between the two groups (Hazard Ratio [HR] 1.04, 95% Confidence Interval [CI] 0.94-1.16, p=0.40). In a pre-specified subgroup analysis, the authors found that patients without history of MI or stroke at randomization who were switched to CTD were more likely to experience the primary outcome (HR 1.12, 95% CI 1.00-1.26), whereas those with history of MI or stroke at randomization who were switched to CTD were less likely to experience the primary outcome (HR 0.73, CI 0.57, 0.94), with the unadjusted p-value for interaction equal to 0.002 and multiplicity adjusted p-value equal to 0.035. The clinical significance of this later finding is unclear and requires further study.


Major limitations included the potential selection bias in favor of HCTZ, as all enrolled patients had a provider decide they would benefit from HCTZ and were by definition tolerating a stable dose of HCTZ without significant side effects.  Furthermore, this study included almost exclusively men and only patients aged greater than 65, limiting its generalizability to women and younger patients.


In conclusion, the DCP found that switching HCTZ to CTD not reduce the incidence of major cardiovascular events. While a subgroup analysis identified a difference in the primary outcome according to history of MI or stroke, this is hypothesis generating and requires further study. Most importantly, the DCP demonstrated that a pragmatic, centralized, point-of-care design is a feasible method to answer clinical questions via a randomized trial that would otherwise be impossible to perform due to cost constraints.



  1. Flack JM, Adekola B. Blood pressure and the new ACC/AHA hypertension guidelines. Trends Cardiovasc Med 2020;30(3):160–4.
  2. Ishani A, Leatherman SM, Woods P, et al. Design of a pragmatic clinical trial embedded in the Electronic Health Record: The VA’s Diuretic Comparison Project. Contemp Clin Trials [Internet] 2022 [cited 2022 Nov 5];116. Available from: