Irfan A. Dhalla, MD, MSc; Tara Gomes, MHSc; Zhan Yao, MD, MS; Jeff Nagge, PharmD; Navindra Persaud, MD, MSc; Chelsea Hellings, MSc; Muhammad M. Mamdani, PharmD, MA, MPH; David N. Juurlink, MD, PhD
Disclaimer: Dr. Dhalla had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The opinions, results, and conclusions reported here are those of the authors and are independent from the funding sources. No endorsement by the Institute for Clinical Evaluative Sciences or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred.
Acknowledgment: The authors thank Brogan Inc., Ottawa, Ontario, Canada, for the use of its Drug Product and Therapeutic Class Database.
Grant Support: This study was funded by a grant from the Ontario Ministry of Health and Long-Term Care to the Ontario Drug Policy Research Network, which is led by Drs. Mamdani and Juurlink. The project was supported by the Institute for Clinical Evaluative Sciences, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. Dr. Dhalla was supported by a New Investigator Award from the Canadian Institutes of Health Research.
Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M12-1038.
Reproducible Research Statement: Study protocol: Available from Dr. Dhalla (e-mail, firstname.lastname@example.org). Statistical code and data set: Not available.
Requests for Single Reprints: Irfan A. Dhalla, MD, MSc, St. Michael's Hospital, 30 Bond Street, Toronto, Ontario M5B 1W8, Canada; e-mail, email@example.com.
Current Author Addresses: Drs. Dhalla, Persaud, and Mamdani and Ms. Gomes: St. Michael's Hospital, 30 Bond Street, Toronto, Ontario M5B 1W8, Canada.
Drs. Yao and Juurlink and Ms. Hellings: Institute for Clinical Evaluative Sciences, G1 06, 2075 Bayview Avenue, Toronto, Ontario M4N 3M5, Canada.
Dr. Nagge: University of Waterloo, School of Pharmacy, 10 Victoria Street South, Kitchener, Ontario N2G 1C5, Canada.
Author Contributions: Conception and design: I.A. Dhalla, T. Gomes, J. Nagge, N. Persaud, M.M. Mamdani, D.N. Juurlink.
Analysis and interpretation of the data: I.A. Dhalla, T. Gomes, Z. Yao, J. Nagge, C. Hellings, M.M. Mamdani, D.N. Juurlink.
Drafting of the article: I.A. Dhalla, N. Persaud, D.N. Juurlink.
Critical revision of the article for important intellectual content: I.A. Dhalla, T. Gomes, Z. Yao, J. Nagge, N. Persaud, C. Hellings, M.M. Mamdani, D.N. Juurlink.
Final approval of the article: I.A. Dhalla, T. Gomes, Z. Yao, J. Nagge, N. Persaud, C. Hellings, M.M. Mamdani, D.N. Juurlink.
Statistical expertise: T. Gomes, Z. Yao, M.M. Mamdani.
Obtaining of funding: T. Gomes, M.M. Mamdani, D.N. Juurlink.
Administrative, technical, or logistic support: I.A. Dhalla, T. Gomes, C. Hellings.
Collection and assembly of data: I.A. Dhalla, T. Gomes, Z. Yao, C. Hellings, M.M. Mamdani, D.N. Juurlink.
Dhalla I., Gomes T., Yao Z., Nagge J., Persaud N., Hellings C., Mamdani M., Juurlink D.; Chlorthalidone Versus Hydrochlorothiazide for the Treatment of Hypertension in Older Adults: A Population-Based Cohort Study. Ann Intern Med. 2013;158:447-455. doi: 10.7326/0003-4819-158-6-201303190-00004
Download citation file:
Published: Ann Intern Med. 2013;158(6):447-455.
Some evidence suggests that chlorthalidone may be superior to hydrochlorothiazide for the treatment of hypertension.
To compare the effectiveness and safety of chlorthalidone and hydrochlorothiazide in older adults.
Propensity score–matched observational cohort study with up to 5 years of follow-up.
All individuals aged 66 years or older who were newly treated with chlorthalidone or hydrochlorothiazide and were not hospitalized for heart failure, stroke, or myocardial infarction in the prior year were eligible for inclusion. Each chlorthalidone recipient was matched to up to 2 hydrochlorothiazide recipients on the basis of age, sex, year of treatment initiation, and propensity score.
The primary outcome was a composite of death or hospitalization for heart failure, stroke, or myocardial infarction. Safety outcomes included hospitalization with hypokalemia or hyponatremia.
A total of 29 873 patients were studied. During follow-up, chlorthalidone recipients (n = 10 384) experienced the primary outcome at a rate of 3.2 events per 100 person-years of follow-up, and hydrochlorothiazide recipients experienced 3.4 events per 100 person-years of follow-up (adjusted hazard ratio, 0.93 [95% CI, 0.81 to 1.06]). Patients treated with chlorthalidone were more likely to be hospitalized with hypokalemia (adjusted hazard ratio, 3.06 [CI, 2.04 to 4.58]) or hyponatremia (adjusted hazard ratio, 1.68 [CI, 1.24 to 2.28]). In 9 post hoc analyses comparing patients initially prescribed 12.5, 25, or 50 mg of chlorthalidone per day with those prescribed 12.5, 25, or 50 mg of hydrochlorothiazide per day, the former were more likely to be hospitalized with hypokalemia for all 6 comparisons in which a statistically significant association was found. The results of other effectiveness and safety outcomes were also consistent with those of the main analysis.
Unmeasured differences in baseline characteristics or physician treatment approaches or an insufficiently large sample may have limited the ability to detect small differences in the comparative effectiveness of the drugs.
As typically prescribed, chlorthalidone in older adults was not associated with fewer adverse cardiovascular events or deaths than hydrochlorothiazide. However, it was associated with a greater incidence of electrolyte abnormalities, particularly hypokalemia.
Ontario Ministry of Health and Long-Term Care.
James J. DiNicolantonio, PharmD, James H. O'Keefe, MD, Carl J. Lavie, MD
Wegmans Pharmacy, Universtiy of Missouri, The Universtiy of Queensland
March 20, 2013
Chlorthalidone Not Hydrochlorothiazide Should be the Thiazide Diuretic of Choice
Dhalla and colleagues1 state “As typically prescribed, chlorthalidone in older adults was not associated with fewer adverse cardiovascular events or deaths than hydrochlorothiazide. However, it was associated with a greater incidence of electrolyte abnormalities, particularly hypokalemia.” This observational cohort study is contrary to a meta-analysis of 108 clinical trials, which indicated that reductions in potassium were considered equivalent between chlorthalidone and hydrochlorothiazide when used in the 12.5-25 mg dose range.2 Additionally, data from the he Multiple Risk Factor Intervention Trial indicate that chlorthalidone significantly reduces cardiovascular events (CVEs) compared to hydrochlorothiazide (p = 0.0016).3 Moreover, a network meta-analysis has shown that compared to hydrochlorothiazide, chlorthalidone significantly reduces congestive heart failure (23% risk reduction [RR], 95% CI, 2-39%, p = 0.032) and CVEs (21% RR, 95% CI, 12-28%, p < 0.0001), with a number needed to treat to prevent 1 CVE over 5 years being 27. The benefit of chlorthalidone over HCTZ for preventing CVEs could not be attributed entirely to a greater blood pressure lowering effect of chlorthalidone. In fact, these differences may perhaps be explained through chlorthalidone’s strong inhibition of carbonic anhydrase isozymes (a pleiotropic effect not shared by hydrochlorothiazide) and its ability to lower sympathetic nerve activity and epinephrine induced platelet aggregation, as well as inhibition of transforming growth factors, vascular endothelial growth factors, fibroblast growth factors and neuropilin 2.4
In summary, clinical trial evidence suggests that the potassium-lowering effects of these two “thiazide-diuretics” are not clinically different and that chlorthalidone should be prescribed over hydrochlorothiazide based on its long history of CVE reduction, a history not equally shared by hydrochlorothiazide.4,5
Alberto Donzelli, MD
ASL di Milano
March 25, 2013
In their Canadian observational cohort study Dhalla et al (1) concluded that, as typically prescribed in older adults, chlorthalidone is not (significantly) associated with fewer adverse cardiovascular events or deaths than hydrochlorothiazide (adjusted hazard ratio 0.93; 95% CI, 0.81-1.06), but is associated with a greater incidence of electrolyte abnormalities, and that hydrochlorothiazide is safer.
I disagree with the conclusion, for the following reasons:
1) The most recent and comprehensive systematic review and network meta-analysis (2) of randomized controlled trials (RCTs) has clearly demonstrated that chlorthalidone significantly reduces all cardiovascular events more than hydrochlorothiazide: -21% in the drug adjusted analysis, -18% in the office systolic blood-pressure adjusted analysis (note that it is the same advantage shown by the most sponsored amlodipine versus hydrochlorothiazide in the celebrated trial ACCOMPLISH). And a network analysis is considered more valid than an observational study (3);
2) Although the hospitalizations for electrolyte abnormalities were more common with chlorthalidone, all-cause hospitalizations were identical (adjusted hazard ratio 1.00) (1);
3) When a low-dose chlorthalidone (12,5 mg/d) was compared with a dose of hydrochlorothiazide of similar potency (25 mg/d), the Table 4 (1) does not show differences in rates of electrolyte abnormalities. A meta-analysis of RCTs (4) also concludes that, within the low-dose range currently recommended (for chlorthalidone), reductions in potassium can be considered equivalent to those of hydrochlorothiazide;
4) In the Canadian cohort the mean starting dose with chlorthalidone (27,3 mg/d, aided by the fact that it is commonly available only in a 25 mg tablet) is excessive for most patients, all the more so for older patients and because there is not a clear gain in blood pressure reduction with a full dose of chlorthalidone, or even with a double dose, compared with a low-dose of 12,5 mg/d (4).
Thus, it is surprising that a study funded by the public health system recommends to continue to focus on hydrochlorothiazide, a less effective drug for cardiovascular outcomes, instead of recommending to shift to the more efficacious (and in many countries even cheaper) low-dose chlorthalidone. Of course, this should be preceded and accompanied by the prescription of a plant based (potassium rich) diet, able to naturally compensate most hypokalemias due to a low-dose diuretic, and measuring the electrolytes during the settling of the therapy. And, finally, avoiding to stress a sodium reduction of uncertain public health value (5).
Alberto Donzelli, Service of Education for Appropriateness and Evidence Based Medicine Local Health Unit, ASL of Milan, Italy
1) Dhalla IA, Gomes T, Zhan Y, Nagge J, Persaud N, Hellings C et al. Chlorthalidone Versus Hydrochlorothiazide for the Treatment of Hypertension in Older Adults. Ann Intern Med 2013;158:447-55.
2) Roush GC, Holford TR, Guddati AK. Chlorthalidone Compared With Hydrochlorothiazide in Reducing Cardiovascular Events. Hypertension 2012; 59:1110-7.
3) Jansen JP, Fleurence R, Devine B, Itzler R, Barrett A, Hawkins N. Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1. Value Health. 2011;14:417-28.
4) Ernst ME, Carter BL, Zheng S, Grimm RH. Meta-Analysis of Dose-Response Characteristics of Hydrochlorothiazide and Chlorthalidone: Effects on Systolic Blood Pressure and Potassium. Am J Hypertens 2010; 23:440-6.
5) Stolarz-Skrzypek K, Kuznetsova T, Thijs L, Tikhonoff V, Seidlerová J, Richart T, et al. European Project on Genes in Hypertension (EPOGH) Investigators. Fatal and nonfatal outcomes, incidence of hypertension, and blood pressure changes in relation to urinary sodium excretion. JAMA. 2011;305:1777-85.
Robert G. Badgett, MD
Kansas University School of Medicine - Wichita
April 8, 2013
Questions about outcome variables
I appreciate the author’s revisiting a long-standing conflict in the management of hypertension. However, I have several concerns about their analysis. First, the relative importance of the primary and secondary outcomes is not clear. The primary outcome is stated as death or hospitalization “for” a major cardiac event. The secondary outcome is hospitalization “with” an electrolyte abnormality. The primary outcome is important and clear, but can the authors provide more detail about the secondary outcome? For example, if a patient is hospitalized for appendicitis and the admitting physician lists hypokalemia as a secondary diagnosis because the potassium is 3.4 mg per dl – is this patient counted as having a secondary outcome in this study? This concern with data from the Ontario Drug Benefit Program was raised after their analysis of hyperkalemia following the RALES study.(1)
In response, the authors provided a re-analysis after limiting the diagnosis of hyperkalemia to the primary diagnosis. In the case of hyperkalemia, the re-analysis confirmed the original analysis. Would the authors of the current study please re-analyze their data after restriction of electrolyte abnormalities to the primary reason for admission?
Second, the relative frequency of the primary and secondary is important as the primary outcome was several times more frequent. When considered together, there was no difference in the frequency of hospitalization for any reason due to a trend for less primary outcomes among patients treated with chlorthalidone.In summary, I do not believe this study as published establishes a difference between the two diuretics because the primary outcome, which is clinically more important and much more frequent, tends to favor chlorthalidone.
I look forward to the authors’ providing a re-analysis or else more detail about the nature of patients with the secondary outcomes. However, even if a re-analysis retains significance, the concern with the relative frequencies of the primary and secondary outcomes will remain and likely be larger.
1. Goldfarb DS. Hyperkalemia after the publication of RALES. N Engl J Med. 2004;351:2448-50; author reply 2448-50. PMID: 15575065.
George Roush, Theodore R. Holford, Achuta K. Guddati
UCONN School of Medicine, Yale University, Massachusetts General Hospital
April 11, 2013
Chlorthalidone versus hydrochlorothiazide for the treatment of hypertension
The retrospective observational cohort study among the elderly by Dhalla and colleagues comparing chlorthalidone patients with hydrochlorothiazide (HCTZ) patients(1) departs methodologically from other studies of antihypertensives and health outcomes. There is no standardization in diagnosing cardiovascular events (CVEs). Secondly, the median follow up for chlorthalidone patients is only 255 days (8.4 months). The difference in CVEs between the 2 drugs does not appear until about 9 months (see their Figure 1), implying that half of the impact of chlorthalidone is not reflected in the hazard ratio (HR). Thirdly, the analysis incorporates total mortality into the primary outcome, something not generally done because these medications have no impact on non-cardiovascular deaths.
These problems all bias the HR towards 1.00, i.e., towards finding no difference in health outcomes between the two medications. In spite of the above, the HR from Dhalla et al. is 0.93, not different statistically from the HR of 0.79 (95% CI: 0.68, 0.92) from another retrospective observational cohort study which avoids these methodologic problems.(2) Pooling these two gives an HR of 0.87 (95% CI 0.78, 0.96), P=0.005, an underestimate of the true benefit of chlorthalidone relative to HCTZ.Within chlorthalidone patients and within HCTZ patients in the study by Dhalla et al., what are the HRs (95% CIs) for hospitalization for CVEs in those who became hypokalemic versus those with sustained normokalemia? An analysis of ALLHAT, the largest hypertension trial available with approximately 9,000 patients on chlorthalidone, concludes:
“Thus, for most patients, concerns about potassium levels should not influence the clinician’s decision about initiating hypertension treatment with low-moderate doses of thiazide diuretics (12.5-25.0 mg of C [chlorthalidone])”.(3)Dhalla et al. state incorrectly that our peer-reviewed network analysis(4) is “conflicting” with the network analysis presented by Psaty and colleagues in their Letter to the Editor(5) and that the latter analysis “compared chlorthalidone and hydrochlorthiazide”. Per their title (see below), Psaty et al. compared chlorthalidone-based with non-chlorthalidone-based diuretics and used just 3 trials for the non-chlorthalidone side of the network: indapamide, HCTZ + amiloride, and HCTZ + triamterene. Also, Psaty et al. did not include ALLHAT, ACCOMPLISH, ANBP2 and HDFP, 4 large trials strongly favoring chlorthalidone over HCTZ in network analysis.(4)
George. C. Roush, MD, MPH, FACP
UCONN School of Medicine & St. Vincent’s Medical Center
Bridgeport, CT 06606
Theodore R. Holford, PhD
Yale University School of Public Health
New Haven, CT 06520
Achuta K. Guddati, MBBS, PhD
Massachusetts General Hospital
Boston, MA 02114
Potential conflicts of interest: None disclosed.
1. Dhalla IA, Gomes T, Yao Z, Nagge J, Persaud N, Hellings C et al. Chlorthalidone versus hydrochlorothiazide for the treatment of hypertension: A population-based cohort study. Ann Intern Med. 19 March 2013;158(6):447-455.
2. Dorsch MP, Gillespie BW, Erickson SR, Bleske BE, Weder AB. Chlorthalidone reduces cardiovascular events compared with hydrochlorothiazide: a retrospective cohort analysis. Hypertension. 2011; 57(4):689-94.
3. Alderman MH, Piller LB, Ford CE, Probstfield JL, Oparil S, Cushman WC et al. Clinical significance of incident hypokalemia and hyperkalemia in treated hypertensive patients in the antihypertensive and lipid-lowering treatment to prevent heart attack trial. Hypertension. 2012 May;59(5):926-33.
4. Roush GC, Holford TR, Guddati AK. Chlorthalidone compared with hydrochlorothiazide in reducing cardiovascular events: systematic review and network meta-analyses. Hypertension. 2012; 59(6):1110-7.
5. Psaty BM, Lumley T, Furberg CD. Meta-analysis of health outcomes of chlorthalidone-based vs nonchlorthalidone-based low-dose diuretic therapies. JAMA. 2004;292(1):43-4.
Paula T. Einhorn, MD, MS; William C. Cushman, MD; Paul K. Whelton, MB, MD, MSc
NHLBI, Bethesda, MD (PTE); Veterans Affairs Medical Center, Memphis, TN (WCC); Tulane University, New Orleans, LA (PKW)
April 16, 2013
Conflict of Interest:
Dr Einhorn is a full-time employee of the National Heart, Lung, and Blood Institute (NHLBI). The views expressed in this letter do not necessarily represent the views of the NHLBI, National Institutes of Health, or any other government entity. Drs. Einhorn and Whelton have no financial interests to disclose. Dr. Cushman has received honoraria from Novartis and Takeda. The ALLHAT Collaborative Research Group was supported by contracts NO1-HC-35130 and HHSN268201100036C with the National Heart, Lung, and Blood Institute. The ALLHAT investigators acknowledge study medications contributed by Pfizer, Inc., (amlodipine and doxazosin), AstraZeneca (atenolol and lisinopril), and Bristol-Myers Squibb (pravastatin) and financial support provided by Pfizer, Inc.
From the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack (ALLHAT) Collaborative Research Group: Need for Clinical Trials
To Editor: We write to clarify that the ALLHAT main results paper(1) included information on serum potassium levels and supplementation, by randomized treatment group (N=15,255-chlorthalidone; 9,048-amlodipine; 9054-lisinopril; 9061-doxazosin). ALLHAT also explored the clinical significance of incident hypo- and hyperkalemia(2). Similar to Dhalla et al.(3), in our practice-based trial, severe hypokalemia (K<3.2mml per L), whether or not clinically apparent, was relatively uncommon (3.2% at year-1 of follow-up in the chlorthalidone arm; median dose 19 mg per day).
Dhalla et al. report an incidence of hospitalizations with hypokalemia among those taking chlorthalidone of ~0.7% at year-1 (modal dose [m-dose]=25 mg per day) vs. ~0.4% for those taking hydrochlorothiazide (m-dose=12.5 mg per day). Most would consider this dose of chlorthalidone to be 4-fold more potent than the 12.5 mg per day m-dose of hydrochlorothiazide. Median duration of follow-up was <1 year, with <15% followed for 5 years.(3)The paper confirms that use of chlorthalidone in community practice is unusual. Of 654,918 patients eligible for matching, only 11,389 (1.7%) were treated with chlorthalidone. As expected, characteristics of the patients receiving chlorthalidone were different from those receiving hydrochlorothiazide. Table 1 and Supplement Tables 5-13 provide limited information about patients’ demographics, general health and care utilization, but lack descriptors of hypertension. In the absence of BP levels and numbers of antihypertensive drugs, we can only speculate (based on the m-dose of chlorthalidone and the use of 50 mg per day) that chlorthalidone might have been used for patients with more severe or drug-resistant hypertension --- suggesting that the main comparison groups (Table 2) differed clinically, including use of fixed-dose combinations. The potentially more relevant comparison of 50 mg per day of HCTZ vs. 25 mg per day of chlorthalidone (Table 4) selected a considerably different cohort (Supplement Table 10). ALLHAT, a randomized, double-blind clinical trial conducted in a diverse cohort of high-risk hypertensives (N=42,418) from diverse North American clinical settings, demonstrated that neither first-step antihypertensive therapy with the ACE-inhibitor lisinopril nor the CCB amlodipine nor the alpha-blocker doxazosin was superior to the diuretic chlorthalidone (12.5-25 mg per day)-based treatment in preventing cardiovascular or renal outcomes; chlorthalidone was superior to all in preventing heart failure and, compared with lisinopril and doxazosin, in preventing stroke.
Over and above our concern with the doses of chlorthalidone and hydrochlorothiazide compared by Dhalla et al., the limitations of propensity scores and the inability to match for unknown variables in a cohort study support the need for clinical trials to resolve this and optimal drug combination comparisons(5).
1. The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA 2002; 288: 2981-2997
2. Alderman MH, Piller LB, Ford CE, Probstfield JL, Oparil S, Cushman WC, et al; Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial Collaborative Research Group. Clinical significance of incident hypokalemia and hyperkalemia in treated hypertensive patients in Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial. Hypertension.2012;59:926-33.
3. Dhalla IA, Gomes T, Zhan Y, Nagge J, Persaud N, Hellings C et al. Chlorthalidone versus hydrochlorothiazide for the treatment of hypertension in older adults. Ann Intern Med 2013;158:447-55
4. Ernst ME, Carter BL, Goerdt CJ, Steffensmeier JJ, Phillips BB, Zimmerman MB, et al. Comparative antihypertensive effects of hydrochlorothiazide and chlorthalidone on ambulatory and office blood pressure. Hypertension. 2006;47:352-8
5. Carter BL, Einhorn PT, Brands M, He J, Cutler JA, Whelton PK, Bakris GL, Brancati FL, Cushman WC, Oparil S, Wright JT,Jr, Working Group from the National Heart, Lung, and Blood Institute. Thiazide-induced dysglycemia: call for research from a working group from the National Heart, Lung, and Blood Institute. Hypertension. 2008; 52: 30-36.
Wright State University Dayton OH
April 20, 2013
Practitioners ignorance of chlorthalidone should be blamed not chlorthalidone
Dear Editor: We thank Dhalla et al.1 for bringing an important but forgotten hypertension medication to lime light. But I think the conclusion was misleading and data does not support the conclusion as stated. The majority of patients in the Chlorthalidone (CTDN) group (70%) were on 25 mg daily dose while similar percentage (67%) in HCTZ group were on 12.5 mg daily dose. This is a difference of four fold in drug equivalency (25 mg of CTDN is equivalent to 50 mg of HCTZ), and many studies confirmed that at such equivalency the incidence of hypokalemia is same with both drugs.Mean dose of initiation of treatment was 27.3 mg for CTDN and 18.3 mg for HCTZ; considering CTDN is at least twice as potent as HCTZ (in lowering BP), this is a difference of three fold in potency. In historic SHEP2 study and ALLHAT3 study maximum dose of CTDN used was 25 mg daily and in SHEP at least one third patient received 12.5 mg or less. JNC 7 recommendation for use of CTDN is 12.5 mg to 25 mg daily and for HCTZ 12.5 mg to 50 mg daily. So the doses used by the practitioners in this study were clearly well above the recommended doses.
There was significant difference in use of ACE inhibitor and ARB between the CTDN (34%) and HCTZ (46%) groups. This difference would account for some of the higher incidence of hypokalemia in CTDN group.Though the reference number 27 and 28 were cited as in agreement with the results of this study, the actual statement in those two references were:“Equivalence analysis using data from several studies suggests that the SBP reductions achieved with HCTZ and chlorthalidone cannot be considered equivalent within the low-dose range currently recommended. However, within this dosing range, reductions in potassium can be considered equivalent.”4“Meta-regression of the effect of thiazides on systolic BP showed a log-linear relationship with a potency series: bendroflumethiazide>chlorthalidone>hydrochlorothiazide. The estimated dose of each drug predicted to reduce systolic BP by 10 mm Hg was 1.4, 8.6, and 26.4 mg, respectively.”5 Median follow up was 255 days (100-873 days) for CTDN group and 398 days (123-1307days) for HCTZ group.
With such limited follow up time (usual follow up period for HTN outcome studies are 4-5 years) one cannot expect enough events to show difference in CV (cardiovascular) outcome; yet there was a trend for better outcome in CTDN group (HR 0.93 with CI 0.81-1.06). Larger sample size or longer follow-up period was likely to show statistically significant benefit with CTDN as seen in recent network meta-analysis by Rousch et al.6 which concluded that CTDN is superior to HCTZ in preventing CV events and relative to HCTA+Z number needed to treat with CTDN to prevent one CV event over 5 years was 27 only. Network metanalysis result is more significant than observational cohort study even when strengthened with propensity scoring. Other recent analysis and editorials also agreed on the superiority of CTDN over HCTZ. There is accumulated ignorance and amnesia about CTDN over last 3 decades.6-10 While all major NIH studies used CTDN in the protocol since mid-1975, HCTZ was marketed as the prototype thiazide to be used. The reason for this unfortunate course in the history of hypertension treatment was recently described in an editorial by Norman Kaplan.10 This unfamiliarity with the drug created a situation whereby CTDN is not used at all or if used, wrong doses are prescribed. But many recent analysis and editorials have clearly pointed out that treatment of hypertension with HCTZ 12.5 mg or 25 mg does not really give the cardiovascular benefit that 12.5 to 25 mg CTDN does.
I think, instead of labeling CTDN as inferior or risky drug, the authors should have underlined the urgent need for more education of the practitioners on the proper use of this proven and valuable antihypertensive drug. Annals of Internal medicine being the premier journal for the internist, it can take an initiative to educate the internists on proper use of diuretic category of antihypertensive drugs.
1. Dhalla IA, Gomes T, Yao Z et al. Chlorthalidone Versus Hydrochlorothiazide for the Treatment of Hypertension in Older Adults: A Population-Based Cohort Study. Annuals of Internal Medicine. 19 March 2013;158(6):447-455. 1.
2. SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program (SHEP). JAMA.1991;265:3255-3264.
3. ALLHAT Officers and Coordinators for the ALLHAT Collaboration Research Group. Major outcomes in high-risk hypertensive patients randomized to Angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA .2002;288:2981-97.
4. Ernst ME, Carter BL, Zheng S et al. Meta-analysis of dose response characteristics of hydrochlorothiazide and chlorthalidone: effects on systolic blood pressure and potassium. American Journal Hypertension; 2010;23:440-6.
5. Peterzan MA, Hardy R, Chaturvedi N et al. Meta-analysis of dose-response relationships for hydrochlorothiazide, chlorthalidone, and bendroflumethiazide on blood pressure, serum potassium, and urate. Hypertension. 2012;59:1104-9.
6. Roush GC, Holford TR, Guddati AK. Chlorthalidone compared with hydrochlorothiazide in reducing cardiovascular events: systematic review and network meta-analyses. Hypertension. 2012;59:1110-7.
7. Saklayen MG. Which diuretic should be used for the treatment of hypertension? American Family Physician. 15 Aug 2008;78(4):444-446.
8. Dorsch MP, Gillespie BW, Erickson SR et al. Chlorthalidone reduces cardiovascular events compared to hydrochlorthiazide. Hypertension. 2011; 57:689-94.
9. Messerli FH, Bangalore S. Half a century of hydrochlorthiazide: facts, fads, fiction and follies. American Journal Medicine. 2011; 124:896-9.
10. Kaplan NM. Chlorthalidone versus Hydrochlorothiazide: a Tale of Tortoises and a Hare. Hypertension. 2011:58:994-5.
Irfan A. Dhalla, MD, MSc, Muhammad M. Mamdani, PharmD, MA, MPH, David N. Juurlink MD, PhD
St. Michael's Hospital
May 1, 2013
We conducted a comparative effectiveness study to test the hypothesis that chlorthalidone is superior to hydrochlorothiazide in older adults as routinely prescribed in Ontario. To our surprise, our data did not support this hypothesis.
As noted in our manuscript, the network meta-analysis completed by Roush et al. suggests that chlorthalidone may be more effective than hydrochlorothiazide in clinical trial settings (1). We also noted that additional data support the argument that chlorthalidone is superior to hydrochlorothiazide (2, 3).
There are several potential explanations why our results yield a different conclusion. First, in the aforementioned network meta-analysis, most chlorthalidone patients were prescribed either 12.5 mg or 25 mg per day. In contrast, the mean dose of chlorthalidone at initiation was 27.3 mg per day in our study. Second, the patients included in the network meta-analysis were all enrolled in randomized controlled trials and may have received superior counseling and monitoring than the patients included in our study. Clinical trial environments rarely accurately reflect real world populations and practices. Third, our primary outcome included all-cause mortality. Although Roush et al. argue that hydrochlorothiazide and chlorthalidone have “no impact on non-cardiovascular deaths,” we respectfully disagree. Either medication can, for example, contribute to orthostatic hypotension, which can lead to falls that are ultimately fatal.
DiNicolantonio et al. argue that the “potassium-lowering effects of … [the two drugs] … are not clinically different.” A recent meta-analysis suggests otherwise (4).
Badgett raises several legitimate concerns about the hypokalemia and hyponatremia outcomes in our study. Although we appreciate his suggestion that we re-analyze our data to include only the cases in which hypokalemia was listed as the primary outcome, such an analysis would be severely underpowered in this instance. It also merits emphasis that studies such as ours that use administrative data are likely to underestimate the occurrence of clinically relevant electrolyte abnormalities (5), and that there is no reason to suspect differential coding for patients treated with one drug or the other.
Donzelli implies that we claimed that hydrochlorothiazide to be the safer drug. In fact, we stated that “it may be reasonable to conclude that hydrochlorothiazide is safer than chlorthalidone in elderly patients at typically prescribed doses.” Donzelli also incorrectly states that we found no difference in electrolyte abnormalities in our post-hoc comparison of chlorthalidone 12.5 mg daily versus hydrochlorothiazide 25 mg daily. Although there were only a small number of patients included in this analysis, admission to hospital with hypokalemia occurred more frequently among patients treated with chlorthalidone, as noted in Table 15 of the Supplement.
In response to Einhorn et al., we did not mean to imply that the ALLHAT main results paper did not include information about serum potassium, only that the network meta-analyses that compared chlorthalidone with hydrochlorothiazide did not report this information. Einhorn et al. also state that chlorthalidone may have been prescribed to patients with more severe hypertension in our study. Although we cannot definitively exclude this possibility, the dose difference is likely related to the fact that chlorthalidone as a single agent is only available on the public formulary in 50 mg tablets (and in combinations only at a dose of 25 mg).
We strongly agree with Donzelli that it would be prudent to start with a lower dose of chlorthalidone than is typically the case in Ontario. We also agree with Einhorn et al. that our findings support the need for clinical trial with clinically relevant endpoints comparing the two drugs.
1. Roush GC, Holford TR, Guddati AK. Chlorthalidone compared with hydrochlorothiazide in reducing cardiovascular events: systematic review and network meta-analyses. Hypertension 2012; 59(6):1110-1117.
2. Dorsch MP, Gillespie BW, Erickson SR, Bleske BE, Weder AB. Chlorthalidone reduces cardiovascular events compared with hydrochlorothiazide: a retrospective cohort analysis. Hypertension 2011; 57(4):689-694.
3. Ernst ME, Carter BL, Goerdt CJ, Steffensmeier JJ, Phillips BB, Zimmerman MB et al. Comparative antihypertensive effects of hydrochlorothiazide and chlorthalidone on ambulatory and office blood pressure. Hypertension 2006; 47(3):352-358.
4. Peterzan MA, Hardy R, Chaturvedi N, Hughes AD. Meta-analysis of dose-response relationships for hydrochlorothiazide, chlorthalidone, and bendroflumethiazide on blood pressure, serum potassium, and urate. Hypertension 2012; 59(6):1104-1109.
5. Gandhi S, Shariff SZ, Fleet JL, Weir MA, Jain AK, Garg AX. Validity of the International Classification of Diseases 10th revision code for hospitalisation with hyponatraemia in elderly patients. BMJ Open 2012; 2(6).
to gain full access to the content and tools.
Learn more about subscription options.
Register Now for a free account.
Cardiology, Endocrine and Metabolism, Geriatric Medicine, Nephrology, Hypertension.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only