Serviços Personalizados
Journal
Artigo
Indicadores
- Citado por SciELO
- Acessos
Links relacionados
- Similares em SciELO
Compartilhar
Revista de investigación clínica
versão On-line ISSN 2564-8896versão impressa ISSN 0034-8376
Rev. invest. clín. vol.59 no.1 Ciudad de México Jan./Fev. 2007
Artículo original
Drug treatment of hypertension: compliance and adverse reactions in a cohort of hypertensive patients in a primary care setting
Tratamiento farmacológico antihipertensivo: Cumplimiento y reacciones adversas en una cohorte de pacientes hipertensos en atención primaria
Dolores MinoLeón,* Hortensia ReyesMorales,** María Eugenia GalvánPlata,*** Héctor PonceMonter,**** José Antonio PalmaAguirre,**** Dante Amato,* Albert Figueras*****
* Coordinación de Investigación en Salud. Instituto Mexicano del Seguro Social.
** Unidad de Investigación Epidemiológica y en Servicios de Salud, CMNSXXI, IMSS.
*** Departamento de Medicina Interna. Hospital de Especialidades, IMSS.
**** Unidad de Investigación en Farmacología, IMSS.
***** Fundación Institut Català de Farmacologia. Barcelona. España.
Correspondence and reprint request:
Dolores Mino León, M.D.
Economía No. 15,
Col. Copilco Universidad,
04360, México, D.F.
Phone: +(5255) 5659 4130; Fax: +(5255) 5761 0952
Email: minodmx@yahoo.com
Recibido el 28 de febrero de 2006.
Aceptado el 12 de diciembre de 2006.
ABSTRACT
Objectives. The primary was to assess the frequency of therapeutic noncompliance due to ADRs in a cohort of patients with recently diagnosed systemic hypertension. The secondary objectives were to evaluate the blood pressure control during the followup in the whole cohort and in patients who received nonsteroidal antiinflammatory drugs (NSAIDs).
Methods. A cohort of 73 recently diagnosed ambulatory hypertensive patients was followedup for 6 months. Validated questionnaires for identification of therapeutic scheme changes and ADRs were applied monthly, during each medical visit.
Results. Family physicians selected monotherapy in 79% of patients. The frequency of therapeutic noncompliance was 44%; noncompliance secondary to ADR was 7%. Systolic and diastolic blood pressure at the beginning of the study were 140 ± 15/90 ± 15 mm Hg for the whole cohort. At the end of the study the figures were 130 ± 11/85 ± 6 (p < 0.001). Patients receiving nonsteroidal antiinflammatory drugs (NSAIDs) had higher blood pressure levels than the groups of patients not receiving such kind of drugs (134 ± 10 vs. 128 ± 8 mm Hg, p = 0.025 and 88 ± 7 vs. 83 ± 5 mm Hg, p = 0.05).
Conclusions. The drugs used in the present study as monotherapy are considered acceptable choices for hypertension treatment. The frequency of therapeutic noncompliance was within the limits reported in the literature and the frequency of therapeutic noncompliance secondary to ADRs in this cohort was lower than that reported in the literature. Higher blood pressure was found in the group of patients receiving NSAIDs.
Key words. Antihypertensive drugs. Pharmacovigilance. Adverse drug reactions. Compliance. Drug utilization study.
RESUMEN
Objetivos. El objetivo primario fue evaluar la frecuencia de falta de cumplimiento terapéutico debido a la presencia de reacciones adversas a medicamentos (RAMs) con el uso de antihipertensivos; los objetivos secundarios fueron evaluar el control de la presión arterial durante el seguimiento y en aquellos pacientes que recibieron fármacos antiinflamatorios no esteroideos.
Métodos. Se integró una cohorte de 73 pacientes hipertensos de reciente diagnóstico, a los que se les vigiló durante seis meses. En cada visita médica mensual se les aplicaron cuestionarios validados, para identificar cambios en el esquema terapéutico y RAMs.
Resultados. Los médicos familiares emplearon monoterapia en 79% de los pacientes. La falta de cumplimiento terapéutico se presentó en 44%; el incumplimiento terapéutico secundario a RAMs se observó en 7% de los casos. En todos los pacientes la presión arterial sistólica y diastólica al inicio del estudio fue 140 ± 15/90 ± 15 mm Hg y al final del estudio las cifras fueron 130 ± 11 / 85 ± 6 (p < 0.001). Los pacientes que recibieron fármacos antiinflamatorios no esteroideos (AINEs) tuvieron cifras de presión arterial más elevadas que pacientes que no recibieron este tipo de fármacos (134 ± 10 vs. 128 ± 8, p = 0.025 y 88 ±7 vs. 83 ± 5 mm Hg, p 0.05).
Conclusiones. Los fármacos empleados en este estudio como monoterapia son fármacos aceptados para el tratamiento de la hipertensión. La frecuencia de falta de cumplimiento terapéutico se ubicó dentro de los límites descritos en la literatura y la falta de cumplimiento terapéutico secundario a RAMs fue menor que lo informado en la literatura. Se observaron cifras de presión arterial elevadas en pacientes que recibieron AINEs.
Palabras clave. Fármacos antihipertensivos. Farmacovigilancia. Reacciones adversas a medicamentos. Cumplimiento y estudio de utilización de medicamentos.
INTRODUCTION
Systemic hypertension is a risk factor for cardiovascular disorders, and renal damage. Clinical trials have demonstrated that an adequate use of antihypertensive drugs is associated with a reduction of stroke and coronary artery disease mortality.1 It has also been demonstrated that most of the antihypertensive drugs have similar efficacy regarding blood pressure reduction; therefore, when a pharmacologic treatment for hypertension is initiated, the associated medical disorders, as well as adverse drug reactions (ADRs) and cost should be taken into account for choosing the optimal treatment.2
In spite of the availability of antihypertensive drugs with demonstrated efficacy, the risks for stroke or cardiovascular death have not shown the expected reductions, which may be partly explained by therapeutic noncompliance.2 There are reports that 16 to 50% of the recently diagnosed hypertensive patients abandon the treatment within the first year. Additionally, a significant proportion of patients that continue on treatment after one year withdraw one or more doses by routine.2 By the fifth year, 54% of the patients have abandoned the treatment.3 A study from Spain showed that less than 50% of the treated hypertensive patients achieved good blood pressure control,4 and other studies from the same country have reported therapeutic noncompliance in 4071% of the hypertensive patients.5 In Mexico, there is few information regarding ADRs. The National Survey of Hypertension (Reencuesta Nacional de Hipertensión Arterial RENAHTA), carried out in 20032004, reported that from 52% of people suffering hypertension and receiving medication, only 20% could mention the name of such medication.6 In patients receiving antihypertensive treatment, some symptoms are considered as ADRs and lead to treatment withdrawal, but they are rather aggravated by the therapeutic noncompliance itself. An example of this is headache. The symptom may be a secondary effect of certain antihypertensive drugs, but it also may be a consequence of high blood pressure.2 ADRs in adult patients accounted for 13% of the office visits in a primary health care center in Spain.7 Usually the ADRs associated with antihypertensive treatment are mild, selflimited, and easily controllable; thus, they are not considered as important primary care or public health problem. Nevertheless, the physicians should carefully discuss with the patients the profile of ADRs that they may experience, and the risks associated with the lack of an appropriate antihypertensive treatment.8 Several reports from different settings have indicated that the cause of drug withdrawal in up to 30% of hypertensive patients was nontolerable ADRs.3
The Mexican Institute of Social Security (IMSS) is the largest health care system in Mexico, and it is a public system that provides care to forty million people and fourteen thousand family physicians staff its primary care services. Hypertension is one of the most frequent causes of visit at primary care; at this level of care, there is an essential list of antihypertensive drugs, that includes thiazide diuretics, βBlockers, ACE inhibitors, Calcium channel blockersnondihydropyridines, Calcium channel blockersdihydropyridines, α1Blockers, Central α2Agonists, and other centrally acting drugs.
The primary objective of the present study was to assess the frequency of therapeutic noncompliance due to ADRs in a cohort of patients with recently diagnosed systemic hypertension. The secondary objectives were to evaluate the blood pressure control during the followup in the whole cohort and in patients who received nonsteroidal antiinflammatory drugs (NSAIDs).
PATIENTS AND METHODS
A cohort of 73 recently diagnosed hypertensive patients from a family medicine clinic belonging to IMSS in Mexico City were followedup for 6 months. Clinical records were reviewed to identify patients diagnosed by the family physician as suffering high blood pressure, and receiving antihypertensive therapy; patients with a diagnosis of secondary hypertension registered at the clinical record were excluded.
All the patients were diagnosed as being hypertensive within 6 months before the recruitment. To be eligible, the patients should have the same medication prescription since the beginning of the antihypertensive therapy. Patients who underwent major changes in their therapy were not included, as well as patients physically or mentally unable to respond to the interview and patients who did not accept to participate in the study.
A convenience sampling was done.
Therapeutic noncompliance was assessed by specific interrogatory and by pillcount. Noncompliance was considered to be present when one or more questions of a validated questionnaire5 received affirmative responses. The clinical chart was reviewed in order to complete the information.
Noncompliance secondary to ADRs was defined as a discontinuation of the antihypertensive treatment due to the presence of a drugrelated clinical manifestation, as reported by the patient, irrespective of the discontinuation duration or of the number of suspended doses.
Before the cohort was assembled, a pharmacoepidemiological committee was integrated. The committee included five members (two internists, two pharmacologists, and one family physician), who determined the appropriateness of the treatment, ADRs due to medication (antihypertensive) and classified the event according to its severity by using the algorithm of Karch and Lasagna.9 Appropriateness of treatment was defined as compliance of validated clinical guidelines based on The sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure criteria.10 All family physicians working at the clinic received an educative program of pharmacovigilance, in order to update them for identification and report of probable ADRs.
During the first sixth months of the year 2000, the patients were selected and recruited by their treating primary care physicians or by a trained nurse who identified and invited them to participate in the waiting room of the primary care facility. Patients were interviewed by the nurse, who filled a questionnaire about family history, evolution of the illness, comorbidity, and the complete treatment received since the beginning of the therapy until the last medical visit. During the 6months followup, after each monthly visit to family physician, the nurse interviewed to the patients and registered their blood pressure, weight, antihypertensive drugs and dosages, other received drugs (particularly nonsteroidal antiinflammatory drugs [NSAIDs]), and relevant clinical data. The day prior to the visit, the patients were reminded by telephone to bring their remaining pills. Immediately after the medical visit, the patients were interviewed by a nurse who asked them in an open way about symptoms suggestive of ADRs during the last 30 days. The interviewer did not mention or suggested any specific symptom, but only asked the patients if they experienced any symptom that they attributed to a drug during that period. If a suspected ADR was present, the patients' actions, such as medical advice request, treatment suspension, selfmedication, or others, were recorded.
Patients were classified in one of the four following groups, according to the antihypertensive prescription that they received during the followup: 1) Patients who continued with the therapeutic scheme during the entire followup; 2) modification of the therapeutic doses during the followup; 3) and those with permanent discontinuation of medication.
ETHICS
The study protocol was approved by the Institutional Review Board. Primary care physicians were invited to voluntarily participate in the study. An oral consent of the patients to participate in the study was obtained; patients were informed about the followup characteristics and the planed procedures, but the objective of the study was not disclosed. During the study, when mild ADRs were identified by the committee, the attending physician was contacted and informed, and a recommendation for modifying the treatment was done. As defined in the study protocol, in case of moderate to severe ADR, the patients should be found for immediate treatment, but moderate or severe ADRs did not occur during the study.
STATISTICS
Frequencies and percentages were calculated for each antihypertensive therapeutic scheme, ADRs, noncompliance, causes of noncompliance, and drugdrug interactions. Blood pressure level was compared among the 4 described groups by ANOVA and Tukey's posthoc test. Additionally, the blood pressure level of patients receiving NSAIDs was compared with that of patients who did not receive NSAIDs by Student's t test. A twotailed p value < 0.05 was considered as statistically significant.
RESULTS
Table 1 shows some clinical and demographic characteristics of the 73 patients included in the cohort.
At the beginning of the followup 58 patients (79%) received monotherapy and only 15 (21%) received a combination of antihypertensive drugs (see table 2). At the end of the study the number of patients receiving a single drug dropped to 45 (62%), and the number of patients receiving combined treatment increased to 20 (27%); 8 patients (11%) suspended the therapy definitively.
The prescribed drug combinations (see table 3) were considered as appropriate in 86.6% of the patients at the beginning, and in 80% at the end of the study; in 4 cases there was not indication (heart failure) for such therapeutic scheme: a combination of a betablocker and an ACEinhibitor (2 patients) and a combination of four drugs including a betablocker, an ACEinhibitor, a calcium channel antagonist, and chlorthalidone (2 patients).
Therapeutic noncompliance was recorded in 32 patients (43.8%) but only in 5 of them, it was due to the occurrence of an ADR. When the group of noncompliant patients was compared with the group of compliance, no differences in sociodemographic variables (age, sex or literacy), treatment (duration of antihypertensive treatment, type of medication prescribed, number of ADRs, potential interactions), and blood pressure level were found.
Twelve patients temporarily discontinued the antihypertensive treatment because they did not get their medication (the pharmacy was short of supplies). Nine of them resumed the original therapeutic scheme, two modified the original scheme, and one permanently discontinued the treatment. These 12 patients were taking captopril (8 patients), chlorthalidone (2) and metoprolol (2).
Table 4 shows the classification of patients according to levels of blood pressure and type of treatment. At the beginning of the study, 20.5% was controlled, from optimal to normal high, while at the end, this percentage improved to more than fifty percent. Regarding type of treatment, combination of medication did not increased from baseline to final evaluation. At the end of the study, 46 patients (63%) remained with monotherapy, and 20 (43.5%) were in bad control of hypertension.
Table 5 shows the 21 ADRs detected during the six months of the follow up. All the ADRs were related only to five different drugs. Thirteen out of 21 ADRs were attributed to an ACE inhibitor (captopril [11] or enalapril [2]). All the suspected ADRs were identified by the nurses during the interview after the medical visit, but only 4 of them were detected and reported by the attending physicians during the medical visit. All the ADRs were classified as mild. According to the Karch and Lasagna algorithm,9 20 ADRs were classified as "probablyrelated" and one as "doubtfullyrelated" to the suspected drug.
Antihypertensives were prescribed concomitantly with another drug with well known potential drugdrug interaction risk in 31 patients (42.5% of the study sample). The most frequently prescribed combination of this kind was an ACE inhibitor with a NSAID (24 patients); in up to 13 of them captopril was prescribed with diclofenac.
Systolic and diastolic blood pressure at the beginning of the study was 140 ± 15 / 90 ± 15 mm Hg (mean ± standard deviation [SD]) respectively, for the whole cohort. At the end of the study the figures were 130 ± 11/85 ± 6. The differences were statistically significative (p < 0.001). When classified in subgroups according to the treatment, patients who continued with the therapeutic scheme during the entire followup (n = 45), those with modification of the therapeutic doses during the followup (n = 20), and those with permanent discontinuation of medication (n = 8), showed similar blood pressure levels during the study (129 ± 13 / 84 ± 6, 134 ± 9 / 87 ± 7 and 128 ± 8 / 85 ± 6 mm Hg, n.s.) (mean ± SD systolic and diastolic blood pressure). Patients receiving NSAIDs (n = 28; 38.3%) had significantly higher blood pressure than patients not receiving such kind of drugs (n = 45; 61.7%) (134 ± 10 mm Hg versus 128 ± 8 mm Hg, p= 0.025 and 88 ± 7 mm Hg versus 83 ± 5 mm Hg; p = 0.05).
DISCUSSION
In the present study, the proportions of patients receiving 1 and 2 antihypertensive drugs were similar, but the proportion of patients receiving more than 2 drugs (11%) was more than twofold higher as compared with a recent study that included a sample of more than 5,000 hypertensive patients.11 The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure recommend the addition of a second drug from a different class when a single drug fails to achieve the goal of control of blood pressure. The initial recommendation drug is thiazide diuretics, but is known that most of patients will need combination of medication to control blood pressure.12 However, it is interesting to highlight that up to 16% of the patients receiving a combination of drugs after 6 months of followup in the present study, received a nonrecommended combination in the literature.11,13
It has been documented that fixeddose combinations usage avoids irrational combinations.11 The drugs used in the present study as monotherapy are considered acceptable choices for hypertension treatment;14 however, the proportion of patients receiving monotherapy at the end of the study was higher than 50%, whereas it is accepted that monotherapy allows good control for only 3050% of the patients.15 The high proportion of patients receiving monotherapy and remained in bad blood pressure control at the end of the study, reinforce the relevance of using combined drugs when monotherapy is not enough in achieving the goal.12
Noncompliance is related to many factors, and definition of an appropriate cutpoint is very difficult, because clinical outcomes are related to multiple variables and not only to the antihypertensive treatment itself. Our decision to classify as noncompliance secondary to ADR those cases in which some of the drugs were discontinued, even transitorily, allowed us to identify certain patients' misconducts that may potentially affect the results of the therapy. We were expecting that with the use of these strict criteria our proportion of patients with noncompliance would be higher than that reported in the literature,16 but this did not occur; it is possible that closely monitoring during the followup, and the short period of observation accounted for the low frequency of ADRrelated noncompliance (5 out of 73 patients; 7%) seen in the present study. The trend for treatment abandon after longer followup periods was not assessed, but it may be important as suggested by other studies.3 An important finding in this study was the temporarily treatment discontinuation in 12 patients (16% of the study sample) due to lack of drugs.
Another finding of the present study is the low rate of ADRs identification by the physicians. They only reported 4 of the 21 ADRs identified by the research nurse. Some authors have reported similar results; in a study of hospitalized patients, the physicians failed to detect and report 57% of the patients whose admission cause was an ADR to drugs;17 moreover, in the same study there was an association between the severity of the ADR and the probability of detection, and mild and moderate ADRs were underrecognized. Another report showed that 68% of the physicians had suspected one or more ADRs, but they did not report them, particularly because the ADR was very well known or mild, thus reporting it was not worthy; 25% of the physicians in the same study never had diagnosed an ADR, and 30% of them did not know how to report them.18 The issue regarding who should be responsible for detecting and reporting the ADR to antihypertensive drugs remains controversial.18 In developing countries where the healthcare institutions have limited personnel and resources, this responsibility usually corresponds to the physicians, instead pharmacists. This was the case of the clinical setting in which our study was carried out.
In the present study only mild ADRs were observed. In addition, most ADRs were observed in patients treated with captopril, a drug with a wellknown toxicity profile. Although there have been reported that side effects attributed to antihypertensives drugs do not represent an important problem in primary care, or public health problems,7 this idea is controversial.19
About one third of the patients in this study were taking NSAIDs; this point is important because, patients that receiving NSAIDs had significantly higher blood pressure levels than patients who did not receive this kind of drugs; this finding has been previously reported.20,21 The mechanism by which these drugs increase blood pressure and interfere with antihypertensive drugs is not fully understood, but it has been reported that they may reduce the efficacy of betablockers, diuretics, and ACE inhibitors because of inhibition of the prostaglandin metabolic pathway, enhancing vasoconstriction.20
Since the beneficial effect of a strict blood pressure control has been widely demonstrated, opportune detection of therapeutic noncompliance, use of inappropriate drug combinations or potential drugdrug interactions (e.g., antihypertensive NSAIDs) should be a priority for the healthcare team, in order to identify potential failures in drug treatment. This is especially important in chronic asymptomatic health conditions such as hypertension, in which lack of control can cause severe complications.
In conclusion, the results of this study underscore the importance of implementing actions directed to strengthen the appropriate utilization of drugs and the guidelines' adherence, as well as the reinforcement of the pharmacovigilance program in primary care settings. Additionally, it is necessary to consider the convenience of designing education programs for the patients, in order to improve compliance and the identification of possible ADRs; for these programs, the role of the primary care nurse is relevant, as part of the health care team supporting the physician's recommendations and helping in the followup of patients with chronic diseases.
SPONSORSHIP
This study was supported by grant IMSS FP 038/ 1327 from the Instituto Mexicano del Seguro Social.
REFERENCES
1. He J, Whelton P. Selection of initial antihypertensive drug therapy. Lancet 2000; 356: 19423. [ Links ]
2. Flack JM, Novikov SV, Ferrario CM. Benefits of adherence to antihypertensive drug therapy. Eur Heart J 1996; 17 (Suppl A): 1620. [ Links ]
3. Cohen JS. Adverse drug effects, compliance, and initial doses of antihypertensive drugs recommended by the Joint National Committee vs. the Physicians' Desk Reference. Arch Intern Med 2001; 161: 8805. [ Links ]
4. GascónCánovas J, SaturnoHernández P, LlorEsteban B, SauraLlamas J, SánchezOrtuño M, VicenteLópez I, et al. Evaluación y mejora de la adhesión terapéutica en los pacientes hipertensos. Aten Primaria 2001; 28: 61519. [ Links ]
5. Piñeiro F, Gil V, Donis M, Orozco D, Pastor R, Merino J. Validez de 6 métodos indirectos para valorar el cumplimiento del tratamiento farmacológico en la hipertensión arterial. Aten Primaria 1997; 19: 3725. [ Links ]
6. Rosas PM, Lara EA, Pastelón HG, Velasquez MO, Martínez RJ, Méndez OA, et al. Reencuesta Nacional de Hipertensión Arterial (RENAHTA): Consolidación Mexicana de los factores de Riesgo Cardiovascular. Cohorte Nacional de Seguimiento. Arch Cardiol Mex 2005; 75: 96111. [ Links ]
7. Joyanes A, Higueras LM, De León JM, Sanz E. Análisis de las reacciones adversas detectadas en un centro de atención primaria. Aten Primaria 1996; 17: 2627. [ Links ]
8. Nies AS. Adverse reactions and interactions limiting the use of antihypertensive drugs. Am J Med 1975; 58: 495503. [ Links ]
9. Karch FE, Lasagna L. Toward the operational identification of adverse drug reactions. Clin Pharmacol Ther 1977; 21: 24754. [ Links ]
10. Oviedo MA, Espinosa F, Olivares R, Reyes H, Trejo JA. Guía clínica para el diagnóstico y tratamiento de la hipertensión arterial. Rev Med IMSS 2003; 41(Supl): S15S26. [ Links ]
11. Ruzicka M, Leenen FH. Monotherapy versus combination therapy as first line treatment of uncomplicated arterial hypertension. Drugs 2001; 61: 94354. [ Links ]
12. Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. The seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. JAMA 2003; 289: 256072. [ Links ]
13. Sequeira RP, Jassim KA, Damanhori AH, Mathur VS. Prescribing pattern of antihypertensive drugs by family physicians and general practitioners in the primary care setting in Bahrain. J Eval Clin Pract 2002; 8: 40714. [ Links ]
14. Gasse C, Stieber J, During A, Keil U, Hense HW. Population trends in antihypertensive drug use: results from the MONICA Augsburg Project 1984 to 1995. J Clin Epidemiol 1999; 52: 695703. [ Links ]
15. Mateo C, Gil A, Sevillano ML, Barutell L, Lorenzo A, Pérez de Lucas N, et al. Calidad de la prescripción de fármacos antihipertensivos en un área de salud. Aten Primaria 2000; 25: 3027. [ Links ]
16. Materson BJ, Reda DJ, Cushman WC, Massie BM, Freis ED, Kochar MS, et al. The Department of Veterans Affairs Cooperative Study Group on Antihypertensive Agents. Single drug therapy for hypertension in men. A comparison of six antihypertensive agents with placebo. N Engl J Med 1993; 328: 91421. [ Links ]
17. Dormann H, CriegeeRieck M, Neubert A, Egger T, Geise A, Krebs S, et al. Lack of awareness of communityacquired adverse drug reactions upon hospital admission. Dimensions and consequences of a dilemma. Drug Saf 2003; 26: 35362. [ Links ]
18. Hasford J, Goettler M, Munter KH, MüllerOerlinghausen B. Physicians' knowledge and attitudes regarding the spontaneous reporting system for adverse drug reactions. J Clin Epidemiol 2002; 55: 94550. [ Links ]
19. Ambrosioni E, Leonetti G, Achille C, Pappelli A, Trimarco B, Zanchetti A. Patterns of hypertension management in Italy: results of a pharmacoepidemiological survey on antihypertensive therapy. J Hypertens 2000; 18: 16919. [ Links ]
20. Chamontin B, Amar J. Interactions médicamenteuses avec les antihypertenseurs. Thérapie 1995; 50: 2216. [ Links ]
21. Conlin PR, Moore TJ, Swartz SL, Barr E, Gazdick L, Fletcher C, et al. Effect of indomethacin on blood pressure lowering by captopril and losartan in hypertensive patients. Hypertension 2000; 36: 4615. [ Links ]