Starting Antihypertensive Drug Treatment With Combination Therapy
2021; Lippincott Williams & Wilkins; Volume: 77; Issue: 3 Linguagem: Inglês
10.1161/hypertensionaha.120.12858
ISSN1524-4563
AutoresZhen‐Yu Zhang, Yu‐Ling Yu, Kei Asayama, Tine W. Hansen, Gladys E. Maestre, Jan A. Staessen,
Tópico(s)Pharmaceutical Economics and Policy
ResumoHomeHypertensionVol. 77, No. 3Starting Antihypertensive Drug Treatment With Combination Therapy Open AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyRedditDiggEmail Jump toSupplementary MaterialsOpen AccessReview ArticlePDF/EPUBStarting Antihypertensive Drug Treatment With Combination TherapyControversies in Hypertension - Con Side of the Argument Zhen-Yu Zhang, Yu-Ling Yu, Kei Asayama, Tine W. Hansen, Gladys E. Maestre, Jan A. Staessen Zhen-Yu ZhangZhen-Yu Zhang https://orcid.org/0000-0002-3785-7417 From the Research Unit Hypertension and Cardiovascular Epidemiology, KU Leuven Department of Cardiovascular Sciences, University of Leuven, Belgium (Z.-Y.Z., Y.-L.Y., K.A.) , Yu-Ling YuYu-Ling Yu https://orcid.org/0000-0002-8255-3770 From the Research Unit Hypertension and Cardiovascular Epidemiology, KU Leuven Department of Cardiovascular Sciences, University of Leuven, Belgium (Z.-Y.Z., Y.-L.Y., K.A.) , Kei AsayamaKei Asayama https://orcid.org/0000-0003-3365-0547 From the Research Unit Hypertension and Cardiovascular Epidemiology, KU Leuven Department of Cardiovascular Sciences, University of Leuven, Belgium (Z.-Y.Z., Y.-L.Y., K.A.) Department of Hygiene and Public Health, Teikyo University School of Medicine, Tokyo, Japan (K.A.) Tohoku Institute for Management of Blood Pressure, Sendai, Japan (K.A.) Research Institute Alliance for the Promotion of Preventive Medicine (APPREMED), Mechelen, Belgium (K.A., G.E.M., T.W.H., J.A.S) , Tine W. HansenTine W. Hansen https://orcid.org/0000-0003-2274-0352 Research Institute Alliance for the Promotion of Preventive Medicine (APPREMED), Mechelen, Belgium (K.A., G.E.M., T.W.H., J.A.S) Steno Diabetes Center Copenhagen, Capital Region of Denmark, Denmark (T.W.H.) , Gladys E. MaestreGladys E. Maestre https://orcid.org/0000-0001-5690-8833 Research Institute Alliance for the Promotion of Preventive Medicine (APPREMED), Mechelen, Belgium (K.A., G.E.M., T.W.H., J.A.S) Department of Neurosciences and Department of Human Genetics, University of Texas Rio Grande Valley School of Medicine, Brownsville, TX (G.E.M.) Alzheimer´s Disease Resource Center for Minority Aging Research, University of Texas Rio Grande Valley, Brownsville, TX (G.E.M.). , Jan A. StaessenJan A. Staessen Correspondence to: Jan A. Staessen, Non-Profit Research Institute, Alliance for the Promotion of Preventive Medicine, Leopoldstraat 59, BE-2800 Mechelen, Belgium. Email E-mail Address: [email protected] https://orcid.org/0000-0002-3026-1637 Research Institute Alliance for the Promotion of Preventive Medicine (APPREMED), Mechelen, Belgium (K.A., G.E.M., T.W.H., J.A.S) Biomedical Sciences Group, Faculty of Medicine, University of Leuven, Belgium (J.A.S.) Originally published10 Feb 2021https://doi.org/10.1161/HYPERTENSIONAHA.120.12858Hypertension. 2021;77:788–798is corrected byCorrection to: Starting Antihypertensive Drug Treatment With Combination Therapy: Controversies in Hypertension - Con Side of the Argument.The 2018 European Society of Cardiology/European Society of Hypertension1 and the 2020 International Society of Hypertension2 guidelines for the management of hypertension proposed that initial combination therapy with 2 antihypertensive agents in a single-pill combination (SPC) is preferred in most patients in need of blood pressure (BP) lowering treatment and should replace the long-standing concept of starting treatment with a single agent, rotating through antihypertensive drug classes, and next moving towards combining drug classes. By moving SPCs forward as the initial BP-lowering strategy, the European1 and International2 Societies of Hypertension Guideline Committees overlooked several principles in hypertension management: (1) understanding the pathophysiology of hypertension; (2) prioritizing evidence from randomized clinical trials above observational studies and expert opinion; and (3) giving consideration to the cost-effectiveness of antihypertensive drug treatment and the sustainability of health care. This article addresses these points. Sources of information included (1) guidelines issued by European,1,3,4 American,5–7 International,2,8,9 and British10–12 Expert Committees, published between 19998 and 2020,2 summarized in Table S1 in the Data Supplement; (2) a PubMed search ran on May 5, 2020, without limitations with as search terms in the abstract or title "hypertension" combined with "fixed combination" OR "hypertension" combined with "single" and "costs"; (3) the placebo-controlled trials of antihypertensive drug treatment, as identified from the reference lists of 5 systematic literature reviews,13–17 of which 2 were published by the Blood Pressure Lowering Trialists' Collaboration14,16; (4) 3 randomized controlled trials of usual versus intensive BP control18–20; and (5) the retail costs of antihypertensive drugs on the Belgian market (https://www.bcfi.be).Tailoring Antihypertensive TreatmentIn the early 1970s, Laragh's group coined the terms low-renin, normal-renin, and high-renin hypertension by relating plasma renin activity to the daily sodium excretion.21 Under normal conditions, plasma renin activity increases with sodium restriction but decreases with higher BP.21 Although an imperfect generalization, low-renin hypertension is characterized by volume expansion and its high-renin counterpart by increased peripheral resistance,22,23 and are indications to start BP-lowering treatment with a diuretic as opposed to an inhibitor of the renin-angiotensin system or vasodilator.24 The activity of the renin system decreases with advancing age22 and is lower in Blacks compared with Whites.25–27 These pathophysiological principles explain why guidelines, with the exception of the 2018 European1 and the 2020 International2 guidelines recommend to start antihypertensive drug treatment with ACE (angiotensin-converting enzyme) inhibitors or ARBs below age 55 and with thiazide diuretics (TDs) or dihydropyridine CCBs (calcium-channel blockers) in older patients and in Blacks across the adult age range. Isolated systolic hypertension, which in its initial course is not associated with increased peripheral resistance, but is caused by stiffening of the large arteries28 is an indication for TDs29 or CCBs.30,31 The 2020 International Society of Hypertension guideline2 supported the use of thiazide-like diuretics, that is, indapamide and chlorthalidone, rather than regular TDs (chlorothiazide and hydrochlorothiazide), based on a systematic review of 19 randomized clinical trials involving 112 113 patients.32 The observed benefits were mainly confined to thiazide-like diuretics rather than TDs with reductions in the risk of cardiac events (odds ratio, 0.78; P<0.001), heart failure (odds ratio, 0.57; P<0.001), and stroke (odds ratio, 0.82; P=0.016).32ACE inhibitors not only inhibit the generation of active angiotensin II, but also the inactivation of the vasodilator bradykinin, explaining their higher potency compared with ARBs and direct renin inhibitors and the recommendation to prescribe ARBs only in ACE inhibitor-intolerant patients.33,34 The involvement of sympathetic drive and the renin-angiotensin system in the cardiovascular and renal complications of hypertension and its comorbidities clarifies why guidelines1–12 unanimously recommend the use ACE inhibitors and ARBs in patients with diabetes or chronic kidney disease, and βBs (beta-blockers),35,36 ACE inhibitors, and ARBs in secondary prevention.Control Rates on Monotherapy Versus Combination TherapyWith as objective to estimate the proportion of patients with hypertension who can be controlled on monotherapy, we reviewed the placebo-controlled randomized clinical trials listed in systematic reviews of BP-lowering therapies13–17 as well as the trials of intensive versus usual BP control.18–20 We extracted control rates on monotherapy from the trial reports.Placebo-Controlled TrialsTable 1 lists the placebo-controlled trials from which the proportion of patients remaining on monotherapy could be extracted. These trials were published from 198537 until 2008.38 The first Medical Research Council Trial (age range, 35–64 years)37 and the Perindopril Protection Against Recurrent Stroke Study (age range, not reported)39 enrolled adults, but all other recruited older patients,30,31,38,40–43 including exclusively38 or a substantial proportion of octogenarians.41,42 Considering the patients randomized to active treatment, at 2 years, from 25.8%38 to 90.0%41,42 remained on a single drug and at 4 years from 48.0%40 to 87.0%.41,42 In the Hypertension in the Very Elderly Trial38 and in the Systolic Hypertension in Europe Trial,30 the study coordinating office emailed or faxed recommendations for intensification of treatment to the local investigators, whenever at a visit a patient was not at goal BP, resulting in a substantially smaller proportion of patients remaining on monotherapy in the placebo compared with the active treatment group (Table 1). In the first Medical Research Council Trial,37 at 4 years of follow-up, 70.0% of patients had attained the target BP, defined as a diastolic BP of <90 mm Hg. Thus, a substantial proportion of patients remained on monotherapy or reached goal BP on a single drug in the placebo-controlled trials listed in Table 1.Table 1. Patients Remaining on First-Line Drug Treatment in Placebo-Controlled Randomized Clinical TrialsTrial, publication yearRefAge, yBlood pressure, mm HgYDrugsNFollow-up, number of patients (%)EntryGoalYear 1Year 2Year 3Year ≥4HYVET, 20083883.6, ≥80173.0/90.8(≥160/<100)<150/<801.8PLAC1912100 (14.2)IND1933196 (25.8)MRC1, 19853752.0, 35–64161.3/98.2 (<200/90–109)NA/<904.9PLAC8654… (…)†… (…)… (…)… (…)‡… (≈39.9)… (≈41.9)… (≈43.9)… (≈45.9)§↓13.5/6.5*BDF4927… (≈81.9)… (≈77.4)… (≈74.3)… (≈72.3)‡… (≈63.1)… (≈67.1)… (≈68.1)… (≈70.1)§↓9.5/5.5*PROP4403… (≈89.9)… (84.9)… (≈82.4)… (≈79.9)‡… (≈61.9)… (66.4)… (≈67.8)… (70.4)§MRC2, 19924065.7, 65–74184.7/90.7 (160–209/ 160/>853.6PLAC845≈634 (88.0)≈150 (80.0)§↓10/NAPER1235≈998 (90.0)≈271 (87.0)§STONE, 19964366.4, 60–79168.5/97.7 (160–219/96–124)<160/<902.5PLAC815… (…)NIF817≈531 (65.0)Syst-China, 19983166.4, ≥60170.5/86.0 (160–219/ 150/NA3.0PLAC1141578 (60.4)348 (42.9)203 (30.9)58 (36.0)§↓20/NANIT1253832 (72.3)584 (62.7)374 (51.5)110 (53.7)§Syst-Eur, 19973070.3, ≥60173.8/85.5 (160–219/ 150/NA2.0PLAC2297693 (41.2)343 (27.8)178 (19.2)95 (13.9)§↓20/NANIT23981037 (59.0)597 (46.5)385 (39.3)216 (30.6)§Age: average age at randomization (age eligibility criterion). Blood pressure: the blood pressure data given are the average systolic/diastolic blood pressure at randomization (blood pressure eligibility criteria) and the goal blood pressure (required decrease in blood pressure). In MRC1, nurses doing screening did 2 sets of BP measurements on separate occasions, but to ensure their diagnostic categorization, the third entry BP was done by a physician. As a result, it took nearly 9 months for the entry BP to reach its lowest approximately stable level. Mean diastolic BP in women randomized to placebo continued to fall for 5 years. As reported in reference 37, only a third (N=2285) had no measurements of diastolic BP below 90 mm Hg at any follow-up visit. Follow-up: Data are the number of patients (percentage) remaining on first-line monotherapy. AT indicates atenolol (50–100 mg/d); BDF, bendrofluazide (10 mg/d); CAND, candesartan (8–16 mg/d); HTCZ/AM, hydrochlorothiazide/amiloride (25/2.5 mg/d); HYVET, Hypertension in the Very Elderly Trial; IND, indapamide (1.5 mg/d); MRC1, Medical Research Council Trial in Young Adults; MRC2, Medical Research Council Trial in Older Adults; N, number of patients randomized; NA, not applicable; NIF, nifedipine (20–60 mg/d); NIT, nitrendipine (10–40 mg/d); PER, perindopril (4 mg/d); PLAC, matching placebo; PROGRESS, Perindopril Protection Against Recurrent Stroke Study; PROP, propranolol (up to 240 mg/d); SCOPE, Study on Cognition and Prognosis in the Elderly; STONE, Shanghai Trial of Nifedipine in the Elderly; Syst-China, Systolic Hypertension in China Trial; Syst-Eur, Systolic Hypertension in Europe Trial.; and Y, median or average follow-up on randomized treatment.* The average placebo-corrected decrease in blood pressure on active treatment.† An ellipsis indicates that the data could not be extracted from the trial report.‡ Number of patients (percentage) remaining on monotherapy.§ Number of patients (percentage) reaching goal blood pressure.Intensive Versus Usual BP ControlOf the 3 trials18–20 comparing intensive with usual BP control, 219,20 reported on treatment status by randomization group. In the ACCORD Trial (Action to Control Cardiovascular Risk in Diabetes)19 and in SPRINT (Systolic Blood Pressure Intervention Trial),20 patients with a systolic BP of 130 mm Hg or higher and an increased cardiovascular risk were randomly assigned to a systolic BP target of <120 mm Hg (intensive treatment) or a systolic target of <140 mm Hg (usual treatment). In the type-2 diabetic patients randomized to intensive (N=2174) and usual (N=2208) BP control in ACCORD,19 after 1 year, the achieved systolic BP averaged 119.3 mm Hg on intensive treatment and 133.5 mm Hg in the control group; in SPRINT,20 these levels were 121.4 mm Hg (N=4683) and 136.2 mm Hg (N=4683), respectively. In ACCORD (median follow-up, 4.7 years),19 at 1 year, 174 (8.0%) and 265 (28.0%) of patients randomized to intensive and standard treatment were on monotherapy and at the last visit 184 (8.0%) and 553 (24.0%); in SPRINT (median follow-up, 3.3 years),20 these numbers at last follow-up were 493 (10.5%) and 1455 (31.1%), respectively.Evidence Supporting SPCsThe literature on SPCs focuses on efficiency, adherence (also known as compliance),44 persistence, and safety. Over time, these notions permeated to several,1–9 but not all,10–12 guidelines. What is the evidence?Randomized Clinical TrialsOur extensive literature review revealed only one randomized clinical trial comparing the efficacy and safety of a SPC with its components.45,46 The COACH Study (Combination of Olmesartan Medoxomil and Amlodipine Besylate in Controlling High Blood Pressure) was a double-blind trial, conducted at 172 clinical sites in the United States.45,46 Patients aged 18 years or older with a diastolic BP ranging from 95 to 120 mm Hg were randomized in equal proportions to combination therapy with olmesartan/amlodipine (daily doses, 10/5, 20/5, 40/5, 10/10, 20/10, or 40/10 mg) or monotherapy with olmesartan (10, 20, or 40 mg) or amlodipine (5 or 10 mg). Of 4234 patients, who entered the 2-week washout phase, 1940 (45.8%) were randomized (women, 45.7%; mean age, 54.0 years; mean entry BP, 164/102 mm Hg) and 1689 (87.1%) completed the 8-week trial. Predictably, each treatment modality, compared with placebo, produced dose-dependent decreases in systolic and diastolic BP and at each dose, combination therapy reduced BP more and achieved BP control more frequently (<140/<90 and <130/<90 mm Hg in diabetic patients) than the equivalent dose of the single-component drug. Limitations of the COACH trial were selection of patients (45.8% of those screened were randomized), the short washout (2 weeks) and follow-up (8 weeks), the highly predictable BP results,47 and the post hoc analysis of patients with isolated systolic hypertension.46The Simplified Treatment Intervention to Control Hypertension Study was a cluster-randomized trial, involving 45 family practices in Ontario, Canada and compared control rates of hypertension as achieved by a simplified treatment algorithm (experimental group) or following the Canadian Hypertension Program guideline (control group).48 The systolic/diastolic target BP was <140/<90 mm Hg and <130/<90 mm Hg in diabetic patients. The simplified treatment algorithm consisted of the following: (1) initial therapy with a low-dose ACE inhibitor/TD or ARB/TD SPC; (2) uptitration of the combination therapy to the highest dose; (3) addition and subsequent uptitration of a CCB; and (4) addition of a βB, α-blocker, or spironolactone. The proportion of patients achieving target BP at 6 months was higher in the experimental (N=802) than the control (N=1246) group (64.7% versus 52.7%; P=0.026). At 6 months, 82.8% of patients in the experimental group were on SPCs and 16.4% in the control group. However, no information on BP control beyond 6 months was provided.48A third randomized double-blind study evaluated the efficacy and safety of triple therapy with amlodipine/valsartan/hydrochlorothiazide for moderate or severe hypertension (systolic/diastolic BP, 145/100 mm Hg or higher).49 After a 1-week single-blind placebo run-in, patients were randomly assigned to valsartan/amlodipine/hydrochlorothiazide 320/10/25 mg, valsartan/hydrochlorothiazide 320/25 mg, valsartan/amlodipine 320/10 mg, or amlodipine/hydrochlorothiazide 10/25 mg with uptitration of these once daily SPCs from week 1 to week 3. Of the 4285 patients screened, 2271 (53.0%) were randomized (women, 44.7%; mean age, 53.2 years; mean entry BP, 169.9/106.5 mm Hg) and 2060 (90.7%) completed the 8-week trial. Triple therapy was significantly superior to all of the dual therapies in reducing BP (P 140/>90 mm Hg). The original protocol prespecified the time of the primary end point at the end of phase 2, namely, 32 weeks after randomization, at which time all patients were receiving the same therapy. The statistical analysis plan, published before the data lock and unblinding, introduced 2 hierarchical co-primary end points.50 The first was the reduction in the systolic home BP averaged over phases 1 and 2, testing for the superiority of initial combination therapy over monotherapy. The co-primary end point, to be tested only if the first hypothesis was confirmed, was the reduction in systolic home BP at week 32, a time point, when all participants were receiving the same treatment. Comparing initial monotherapy with initial combination therapy (Figure 1), the systolic/diastolic reductions in the home BP were 13.3/6.5 versus 21.9/12.1 mm Hg (end of phase 1), 20.1/10.7 versus 19.5/10.6 mm Hg at week 24 (midpoint of phase 2), 23.6/12.7 versus 22.0/11.9 mm Hg at week 32 (end of phase 2), and 24.5/13.9 versus 23.6/13.4 mm Hg at week 52 (end of study). By the end of phase 3, over 75% of participants in the initial monotherapy and combination therapy groups had attained the target home BP with no difference between groups at the end of either phase 2 or 3.24 Based on the redefinition of the primary end points,50 the PATHWAY-1 researchers reported the average BP results combining phases 2 and 3 and all study periods.24 They concluded that initial combination therapy achieved target BP in twice as many participants as initial monotherapy,24 whereas in fact starting from week 24 (Figure 1), home BP was similar irrespective of whether antihypertensive treatment was started with SPC or free SD combination therapy. In the context of the current debate, a relevant finding of the PATHWAY-1 trial was that the BP reductions induced by losartan and hydrochlorothiazide were greatest in the top and bottom plasma renin activity tertiles, respectively,24 an argument supporting an insightful rather than a simplistic initiation of antihypertensive drug therapy.Download figureDownload PowerPointFigure 1. Systolic home blood pressure (BP) by randomization group and follow-up duration. Data points are means. Vertical bars indicate 95% CI. Patients were randomized to initial monotherapy with losartan (LOS) 50–100 mg (N=151) or hydrochlorothiazide (HCTZ) 12.5–25 mg (N=150), crossing over at 8 weeks (switching to the alternative monotherapy), or initial combination treatment with losartan 50–100 mg plus hydrochlorothiazide 12.5–25 mg (N=304). In phase 2 (weeks 17–32), all patients received losartan 100 mg and hydrochlorothiazide 12.5 to 25 mg. In phase 3 (weeks 33–52), amlodipine with or without doxazosin could be added to achieve target BP. SBP indicates systolic BP. Reproduced from MacDonald et al24 with permission. Copyright ©2017, Wiley.Observational StudiesA common denominator of all observational studies was that they had a retrospective design. A meta-analysis published in 201151 summarized 12 studies published from 200052 until 2010.53 It compared health care costs, adherence, and persistence between groups of patients taking antihypertensive agents as SPCs versus free-equivalent SDs. The mean difference in the annual all-cause and hypertension-related health care costs was $1357 (CI, $778–$1935) lower in favor of SPCs than free SD combinations. Adherence, measured as the mean difference in medication possession ratio, was 8% higher in patients naive to prior antihypertensive drugs and 14% higher in non-naive SPC patients compared with their counterparts on free SD combinations. Persistence in the SPC groups was twice as likely as in the free SD combination groups (pooled risk ratio, 2.1 [CI, 1.1–4.1]). The authors hypothesized that improved adherence and persistence likely contributed to the lower health care costs in the SPCs groups via improved clinical outcomes. Of the 12 studies included in the meta-analysis,52–63 2 did not include a conflict of interest statement,54,56 10 were directly funded by the pharmaceutical industry,52,53,55,57–63 and 752,56–59,61,63 had one or more co-authors employed by drug companies having a commercial interest in SPCs.The early literature was almost unanimous in stating that SPCs, in comparison with SDs or free combinations of SDs, were more efficacious in lowering BP, increasing adherence and persistence, and lowering health care costs. In view of this exceptional consistency, we searched PubMed for publications with discordant results. We identified 10 studies,51,64–72 published between 201064 until 2020,73 of which the principal outcome measures, data sources consulted, the methods applied, and principal limitations are summarized in Table S2. Of the 10 studies,51,64–72 751,64–67,69,70 were directly supported by SPCs producers, 551,64–67 involved a subcontractor to these manufacturers, and 551,64,65,67,69 were co-authored by one or more industry employees. The study by Hong et al68 stands out, because it was a publication not supported by industry, in which none of the authors reported a conflict of interest. In this article, free SD combinations had average monthly drug costs similar to the respective SPCs, when SPCs were not generically available.68 However, free SD combinations were more expensive compared with generic SPCs.68A study published in 2020 without industry support,72 applied the 2014 to 2015 Medical Expenditure Panel Survey data, to assess the uses and expenses of antihypertensive drugs among American men and nonpregnant women, aged 18 or older, who had a diagnosis of hypertension. Multiple medications users were patients who used 2 or more antihypertensive medications each year, including SPCs or multiple free SD combinations, or who switched BP-lowering agents within or between classes. Among 10 971 hypertensive adults, 4759 (44.1%) were SD users and 6212 (55.9%) were multiple medication users. The average annual total cost for antihypertensive medications was $336 per person: $199 for SD users and $436 for multiple medication users. The average annual costs for each medication class were estimated at $438 for ARBs and $49 for TDs. Thus, users of multiple medications, including SPCs, incurred more than twice the expense than single medication users.72 When comparing classes of medications, the costs for ARBs were highest, whereas those for TDs were lowest (Figure 2), a trend still visible in the 2020 retail prizes of antihypertensive drugs on the Belgian market (Table S3).72Download figureDownload PowerPointFigure 2. Estimated average annual per capita expenses of each medication class (95% CI), expressed in US dollars based on the 2014–2015 Medical Expenditure Panel Survey Notes. ACEI indicates angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CCB, calcium-channel blocker; and TD, thiazide diuretics. Reproduced from Park et al72 with permission. Copyright ©2020, Elsevier.Several studies addressed the health-economic aspects of the use of SPCs versus SDs or free combinations of SDs,65–67,70,73 or triple versus dual SPCs.69 Data from the MarketScan Database 2006 to 2008 in the United States showed that SPC patients (N=382 476) fared better over a 6-month period than their counterparts on free SD combinations (N=197 375).73 The analyses were adjusted for the baseline characteristics of the selected patients, a reason why this article73 was excluded from the 2011 meta-analysis.51 SPC patients had higher medication possession rates (+11.6%), fewer all-cause hospitalizations (−23.0%), and emergency room visits (−13.0%). SPC patients showed greater reductions in post-therapy initiation in all-cause medical costs ($208 [CI, −$302 to −$114]), but larger increases in hypertension-related prescription costs (+$53 [CI, +$51 to +$55]).73 Similarly, in a study conducted in UK general-practice, hospitalization costs validated up to 2011 were lower in SPC patients compared with free SD users (N=9929 versus 18 665; £62 versus £112; P<0.001), whereas drug costs were higher (£126 versus £78; P<0.001), resulting in similar mean annual management costs in the 2 groups (£192 versus £192).66All observational reviewed above (Table S2) had a retrospective design and were, therefore, vulnerable to overt and hidden sources of bias, for which analyses did not account. Particularly, most studies had no information on the severity of hypertension at the time of initiation or adjustment of BP-lowering treatment, higher BP being an indication for SPCs or multiple drugs, or on the patients' health insurance status as determinant of the out-of-pocket costs and adherence (Table S2). Data on health behaviors, patients' lifestyle, and use of over-the-counter drugs were unavailable. In several analyses, there was a remarkable imbalance between SPC and free SD combination users,65–67,73 indicating selection bias in the patients being prescribed SPCs versus free SD combinations or in data extraction from the claims databases by researchers. Medication possession rate, although an objective measure, but only in settings with a closed pharmacy system,44 is an ambiguous concept. Although there is moderate association between claims for filled prescriptions and measured drug levels44 or prevention of adverse health outcomes,74 claims databases do not ensure that the medication was taken as prescribed. Moreover, information from claims databases disfavors free SD drug combinations, because in their publications investigators selected the SD with the worse adherence,73 or when 2 or more SDs were prescribed, probabilities of nonadherence were multiplicative, not additive. Furthermore, the claims data used for the health-economic analyses were collected for payment purposes rather than for research. A diagnostic code on a medical claim is no proof for the presence of disease, because diagnoses might be incorrectly coded or included as a rule-out criterion rather than as an actual disease. All reviewed health-economic studies only accounted for direct health care costs, disregarding patient values,75 and out-of-pocket costs.71 A follow-up duration ranging from 6 months64 to 5 years66 is not representative of the life course of hypertension. No study measured adverse health outcomes in a prospective manner (Table S2). Transitions between health states applied in Markov modeling were not directly measured, but extrapolated,69,70 introducing arbitrariness in selecting data sources best fitting the hypothesis to be proven.Narrative ReviewsOf 7 reviews on the use of SPCs,76–82 published from 200976 until 2019,82 6 were written with direct financial support from SPC manufacturers,77–82 3 included co-authors employed by these manufacturers,77,79,82 2 involved a for-profit company running the literature search77 or providing assistance in writing the text,78 and 1 article's co-author received research support from a company marketing SPCs.76Take Home MessagesTable 2 lists the major limitations of the recommended policy to initiate antihypertensive treatment using SPCs in most patients.1,2Table 2. Take Home MessagesIssueSummary of the literatureRCTsLack of RCTs comparing the long-term efficiency, adverse effects, and cost-effectiveness of initiating antihypertensive drug treatment with
Referência(s)