Q&A

BACKGROUND: The mortality risk is high in patients with diabetes who sustain myocardial infarction. Recent evidence¹ suggests that coronary artery bypass grafting (CABG) reduces long-term mortality better than percutaneous transluminal coronary angioplasty (PTCA), but the protective effect of CABG for patients with diabetes who have a future acute myocardial infarction is not known. This substudy of the Bypass Angioplasty Revascularization Investigation addresses this question.

POPULATION STUDIED: A total of 3603 patients who were severely symptomatic and who had angiographic evidence of multivessel coronary artery disease were identified at 18 sites in North America. Of this group, 1829 were randomized; the remainder underwent the revascularization procedure selected by their physician, usually CABG. At the beginning of the study, the mean age was 61.5 years, 73% were men, 64% had unstable angina, 41% had 3-vessel disease, 9% had congestive heart failure, and 19% had diabetes mellitus.

STUDY DESIGN AND VALIDITY: Patients were randomized to PTCA or CABG. Patients were re-evaluated with clinic visits and electrocardiograms between 4 and 14 weeks and then annually either in person (odd years) or by telephone (even years) for up to 5 years. After 5 years, annual follow-up was by telephone only. Only spontaneous Q-wave myocardial infarctions that occurred after revascularization were analyzed. Cumulative mortality data were collected for a mean of 5.5 years from the time of randomization. Kaplan-Meier analysis of cumulative mortality risk and Cox regression analysis of relative mortality risks were appropriately employed.

OUTCOMES MEASURED: The primary outcomes were death from all causes and death from Q-wave infarction. A secondary outcome was an estimate of relative mortality risk in those who underwent PTCA or CABG.

RESULTS: All-cause mortality at 5 years was higher in patients with diabetes than in those who did not have diabetes (20% vs 8%, P <.001; number needed to treat [NNT]=8). The corresponding mortality rates for those patients who underwent CABG were 18% and 7%, respectively, compared with rates of 25% and 8% in the PTCA group. Q-wave myocardial infarction occurred in 4.8% of patients during the 5-year follow-up; this rate was similar between CABG and PTCA groups. There was no significant difference in the estimated 5-year mortality risk after a spontaneous Q-wave infarction in patients who did not have diabetes who underwent either PTCA or CABG (30% vs 27%, P=ns). Among patients who had Q-wave infarction, there was no significant difference in mortality risk between patients who did and did not have diabetes who underwent CABG (17% vs 27%, P=ns). However, mortality after a Q-wave infarction was much higher in patients with diabetes who underwent PTCA instead of CABG (80% vs 17%, NNT=1.6). Compared with PTCA, CABG protected patients with diabetes who had a subsequent Q-wave infarction (relative risk of death=0.09; 95% confidence interval [CI], 0.03-0.29; P <.001). CABG also protected patients with diabetes who did not have an infarction (relative risk=0.65; 95% CI, 0.45-0.94; P=.02). The study was well designed, particularly considering the relatively rare occurrence of Q-wave infarction after revascularization. The randomization was not concealed and could have introduced selection bias in the results. The severity, duration, and control of diabetes were also not analyzed in this study.

BACKGROUND: The mortality risk is high in patients with diabetes who sustain myocardial infarction. Recent evidence¹ suggests that coronary artery bypass grafting (CABG) reduces long-term mortality better than percutaneous transluminal coronary angioplasty (PTCA), but the protective effect of CABG for patients with diabetes who have a future acute myocardial infarction is not known. This substudy of the Bypass Angioplasty Revascularization Investigation addresses this question.

POPULATION STUDIED: A total of 3603 patients who were severely symptomatic and who had angiographic evidence of multivessel coronary artery disease were identified at 18 sites in North America. Of this group, 1829 were randomized; the remainder underwent the revascularization procedure selected by their physician, usually CABG. At the beginning of the study, the mean age was 61.5 years, 73% were men, 64% had unstable angina, 41% had 3-vessel disease, 9% had congestive heart failure, and 19% had diabetes mellitus.

STUDY DESIGN AND VALIDITY: Patients were randomized to PTCA or CABG. Patients were re-evaluated with clinic visits and electrocardiograms between 4 and 14 weeks and then annually either in person (odd years) or by telephone (even years) for up to 5 years. After 5 years, annual follow-up was by telephone only. Only spontaneous Q-wave myocardial infarctions that occurred after revascularization were analyzed. Cumulative mortality data were collected for a mean of 5.5 years from the time of randomization. Kaplan-Meier analysis of cumulative mortality risk and Cox regression analysis of relative mortality risks were appropriately employed.

OUTCOMES MEASURED: The primary outcomes were death from all causes and death from Q-wave infarction. A secondary outcome was an estimate of relative mortality risk in those who underwent PTCA or CABG.

RESULTS: All-cause mortality at 5 years was higher in patients with diabetes than in those who did not have diabetes (20% vs 8%, P <.001; number needed to treat [NNT]=8). The corresponding mortality rates for those patients who underwent CABG were 18% and 7%, respectively, compared with rates of 25% and 8% in the PTCA group. Q-wave myocardial infarction occurred in 4.8% of patients during the 5-year follow-up; this rate was similar between CABG and PTCA groups. There was no significant difference in the estimated 5-year mortality risk after a spontaneous Q-wave infarction in patients who did not have diabetes who underwent either PTCA or CABG (30% vs 27%, P=ns). Among patients who had Q-wave infarction, there was no significant difference in mortality risk between patients who did and did not have diabetes who underwent CABG (17% vs 27%, P=ns). However, mortality after a Q-wave infarction was much higher in patients with diabetes who underwent PTCA instead of CABG (80% vs 17%, NNT=1.6). Compared with PTCA, CABG protected patients with diabetes who had a subsequent Q-wave infarction (relative risk of death=0.09; 95% confidence interval [CI], 0.03-0.29; P <.001). CABG also protected patients with diabetes who did not have an infarction (relative risk=0.65; 95% CI, 0.45-0.94; P=.02). The study was well designed, particularly considering the relatively rare occurrence of Q-wave infarction after revascularization. The randomization was not concealed and could have introduced selection bias in the results. The severity, duration, and control of diabetes were also not analyzed in this study.

BACKGROUND: The mortality risk is high in patients with diabetes who sustain myocardial infarction. Recent evidence¹ suggests that coronary artery bypass grafting (CABG) reduces long-term mortality better than percutaneous transluminal coronary angioplasty (PTCA), but the protective effect of CABG for patients with diabetes who have a future acute myocardial infarction is not known. This substudy of the Bypass Angioplasty Revascularization Investigation addresses this question.

POPULATION STUDIED: A total of 3603 patients who were severely symptomatic and who had angiographic evidence of multivessel coronary artery disease were identified at 18 sites in North America. Of this group, 1829 were randomized; the remainder underwent the revascularization procedure selected by their physician, usually CABG. At the beginning of the study, the mean age was 61.5 years, 73% were men, 64% had unstable angina, 41% had 3-vessel disease, 9% had congestive heart failure, and 19% had diabetes mellitus.

STUDY DESIGN AND VALIDITY: Patients were randomized to PTCA or CABG. Patients were re-evaluated with clinic visits and electrocardiograms between 4 and 14 weeks and then annually either in person (odd years) or by telephone (even years) for up to 5 years. After 5 years, annual follow-up was by telephone only. Only spontaneous Q-wave myocardial infarctions that occurred after revascularization were analyzed. Cumulative mortality data were collected for a mean of 5.5 years from the time of randomization. Kaplan-Meier analysis of cumulative mortality risk and Cox regression analysis of relative mortality risks were appropriately employed.

OUTCOMES MEASURED: The primary outcomes were death from all causes and death from Q-wave infarction. A secondary outcome was an estimate of relative mortality risk in those who underwent PTCA or CABG.

RESULTS: All-cause mortality at 5 years was higher in patients with diabetes than in those who did not have diabetes (20% vs 8%, P <.001; number needed to treat [NNT]=8). The corresponding mortality rates for those patients who underwent CABG were 18% and 7%, respectively, compared with rates of 25% and 8% in the PTCA group. Q-wave myocardial infarction occurred in 4.8% of patients during the 5-year follow-up; this rate was similar between CABG and PTCA groups. There was no significant difference in the estimated 5-year mortality risk after a spontaneous Q-wave infarction in patients who did not have diabetes who underwent either PTCA or CABG (30% vs 27%, P=ns). Among patients who had Q-wave infarction, there was no significant difference in mortality risk between patients who did and did not have diabetes who underwent CABG (17% vs 27%, P=ns). However, mortality after a Q-wave infarction was much higher in patients with diabetes who underwent PTCA instead of CABG (80% vs 17%, NNT=1.6). Compared with PTCA, CABG protected patients with diabetes who had a subsequent Q-wave infarction (relative risk of death=0.09; 95% confidence interval [CI], 0.03-0.29; P <.001). CABG also protected patients with diabetes who did not have an infarction (relative risk=0.65; 95% CI, 0.45-0.94; P=.02). The study was well designed, particularly considering the relatively rare occurrence of Q-wave infarction after revascularization. The randomization was not concealed and could have introduced selection bias in the results. The severity, duration, and control of diabetes were also not analyzed in this study.

BACKGROUND: Unfractionated and low-molecular-weight (LMW) heparins are popular and successful agents for preventing pulmonary embolism (PE) and deep vein thrombosis (DVT) after major surgery, but they are usually stopped at hospital discharge. The use of an antiplatelet agent such as low-dose aspirin might provide additional benefits. A meta-analysis of data from 8000 patients indicated several weeks of such therapy reduced the frequency of PE and DVT by 40% to 60%. The Pulmonary Embolism Prevention (PEP) trial of low-dose aspirin was a large randomized control trial (RCT) designed to confirm or refute these findings.

POPULATION STUDIED: This was a multicountry trial (Australia, New Zealand, South Africa, Sweden, and the United Kingdom) of patients with proximal femur fractures or undergoing hip or knee arthroplasty. Patients with a clear indication for aspirin (such as a recent myocardial infarction) or clear contraindication (such as an active peptic ulcer) were not eligible. Concurrent use of other thromboprophylactic agents and previous use of nonsteroidal anti-inflammatory drugs (NSAIDs) were allowed.

STUDY DESIGN AND VALIDITY: The PEP study was a double-blind placebo-controlled trial in which eligible patients who gave consent were randomly assigned to receive either 160 mg enteric-coated aspirin or placebo daily for 5 weeks, starting before surgery. The authors described the methods used to prevent researchers from knowing to which group the patient would be assigned (ie, concealed allocation). Patients were advised to avoid other NSAIDs during this time. A diagnosis of DVT required confirmation by venography or ultrasound; PEs were classified as definite or probable by an independent committee on the basis of a combination of clinical findings, angiogram, ventilation-perfusion scan, or venographic evidence of DVT. Follow-up assessed in-hospital morbidity and mortality and out-of-hospital mortality for the month following surgery. Analysis was by intention to treat. This was a large well-conducted trial involving at least 150 hospitals, with complete follow-up of 99.4% of enrollees. Randomization resulted in equal distribution on the basis of age and sex; no data on race or ethnicity were presented. The concurrent use of heparin agents was equally distributed between the intervention and control groups. Diagnosis of DVT, PE, MI, and cerebrovascular accident was made by a committee unaware of the subjects' group assignment.

OUTCOMES MEASURED: Follow-up was for cause of death through day 35 and for nonfatal events during the hospitalization (mean duration=16 days).

RESULTS: Approximately 13,000 patients with hip fracture were randomized, with 92% initiating treatment preoperatively or immediately postoperatively and 80% taking the assigned medication for the entire 35-day follow-up. Approximately 45% of the participants were receiving a form of heparin (similar numbers in both groups). Symptomatic DVT was confirmed in 1.03% of those assigned aspirin and 1.45% of those assigned placebo (P=.03; number needed to treat [NNT]=238). Definite or probable PE was confirmed in 0.69% of patients taking aspirin and 1.21% of those taking placebo (P=.002; NNT=192). For fatal PE, the NNT was 270 (P=.002). The overall mortality was identical in both groups. Aspirin had statistically significant benefits on total thromboembolic events in those patients taking unfractionated heparin or no heparin but not in those who received LMW heparin. Nonfatal and fatal cardiac ischemic events occurred more frequently in the aspirin group (1.57% vs 1.18%, P=.05). A total of 256 patients would need to receive aspirin for one additional event to occur (number needed to harm [NNH]=256). Fatal bleeding episodes were approximately equal in both groups; however, 2.95% of the patients assigned to aspirin required transfusion versus 2.35% of patients assigned to placebo (P=.04; NNH=167). The number of patients in the arthroplasty section of the trial was smaller (n=4000), and none of the comparisons between the intervention and placebo groups showed statistically significant results.

BACKGROUND: Unfractionated and low-molecular-weight (LMW) heparins are popular and successful agents for preventing pulmonary embolism (PE) and deep vein thrombosis (DVT) after major surgery, but they are usually stopped at hospital discharge. The use of an antiplatelet agent such as low-dose aspirin might provide additional benefits. A meta-analysis of data from 8000 patients indicated several weeks of such therapy reduced the frequency of PE and DVT by 40% to 60%. The Pulmonary Embolism Prevention (PEP) trial of low-dose aspirin was a large randomized control trial (RCT) designed to confirm or refute these findings.

POPULATION STUDIED: This was a multicountry trial (Australia, New Zealand, South Africa, Sweden, and the United Kingdom) of patients with proximal femur fractures or undergoing hip or knee arthroplasty. Patients with a clear indication for aspirin (such as a recent myocardial infarction) or clear contraindication (such as an active peptic ulcer) were not eligible. Concurrent use of other thromboprophylactic agents and previous use of nonsteroidal anti-inflammatory drugs (NSAIDs) were allowed.

STUDY DESIGN AND VALIDITY: The PEP study was a double-blind placebo-controlled trial in which eligible patients who gave consent were randomly assigned to receive either 160 mg enteric-coated aspirin or placebo daily for 5 weeks, starting before surgery. The authors described the methods used to prevent researchers from knowing to which group the patient would be assigned (ie, concealed allocation). Patients were advised to avoid other NSAIDs during this time. A diagnosis of DVT required confirmation by venography or ultrasound; PEs were classified as definite or probable by an independent committee on the basis of a combination of clinical findings, angiogram, ventilation-perfusion scan, or venographic evidence of DVT. Follow-up assessed in-hospital morbidity and mortality and out-of-hospital mortality for the month following surgery. Analysis was by intention to treat. This was a large well-conducted trial involving at least 150 hospitals, with complete follow-up of 99.4% of enrollees. Randomization resulted in equal distribution on the basis of age and sex; no data on race or ethnicity were presented. The concurrent use of heparin agents was equally distributed between the intervention and control groups. Diagnosis of DVT, PE, MI, and cerebrovascular accident was made by a committee unaware of the subjects' group assignment.

OUTCOMES MEASURED: Follow-up was for cause of death through day 35 and for nonfatal events during the hospitalization (mean duration=16 days).

RESULTS: Approximately 13,000 patients with hip fracture were randomized, with 92% initiating treatment preoperatively or immediately postoperatively and 80% taking the assigned medication for the entire 35-day follow-up. Approximately 45% of the participants were receiving a form of heparin (similar numbers in both groups). Symptomatic DVT was confirmed in 1.03% of those assigned aspirin and 1.45% of those assigned placebo (P=.03; number needed to treat [NNT]=238). Definite or probable PE was confirmed in 0.69% of patients taking aspirin and 1.21% of those taking placebo (P=.002; NNT=192). For fatal PE, the NNT was 270 (P=.002). The overall mortality was identical in both groups. Aspirin had statistically significant benefits on total thromboembolic events in those patients taking unfractionated heparin or no heparin but not in those who received LMW heparin. Nonfatal and fatal cardiac ischemic events occurred more frequently in the aspirin group (1.57% vs 1.18%, P=.05). A total of 256 patients would need to receive aspirin for one additional event to occur (number needed to harm [NNH]=256). Fatal bleeding episodes were approximately equal in both groups; however, 2.95% of the patients assigned to aspirin required transfusion versus 2.35% of patients assigned to placebo (P=.04; NNH=167). The number of patients in the arthroplasty section of the trial was smaller (n=4000), and none of the comparisons between the intervention and placebo groups showed statistically significant results.

BACKGROUND: Unfractionated and low-molecular-weight (LMW) heparins are popular and successful agents for preventing pulmonary embolism (PE) and deep vein thrombosis (DVT) after major surgery, but they are usually stopped at hospital discharge. The use of an antiplatelet agent such as low-dose aspirin might provide additional benefits. A meta-analysis of data from 8000 patients indicated several weeks of such therapy reduced the frequency of PE and DVT by 40% to 60%. The Pulmonary Embolism Prevention (PEP) trial of low-dose aspirin was a large randomized control trial (RCT) designed to confirm or refute these findings.

POPULATION STUDIED: This was a multicountry trial (Australia, New Zealand, South Africa, Sweden, and the United Kingdom) of patients with proximal femur fractures or undergoing hip or knee arthroplasty. Patients with a clear indication for aspirin (such as a recent myocardial infarction) or clear contraindication (such as an active peptic ulcer) were not eligible. Concurrent use of other thromboprophylactic agents and previous use of nonsteroidal anti-inflammatory drugs (NSAIDs) were allowed.

STUDY DESIGN AND VALIDITY: The PEP study was a double-blind placebo-controlled trial in which eligible patients who gave consent were randomly assigned to receive either 160 mg enteric-coated aspirin or placebo daily for 5 weeks, starting before surgery. The authors described the methods used to prevent researchers from knowing to which group the patient would be assigned (ie, concealed allocation). Patients were advised to avoid other NSAIDs during this time. A diagnosis of DVT required confirmation by venography or ultrasound; PEs were classified as definite or probable by an independent committee on the basis of a combination of clinical findings, angiogram, ventilation-perfusion scan, or venographic evidence of DVT. Follow-up assessed in-hospital morbidity and mortality and out-of-hospital mortality for the month following surgery. Analysis was by intention to treat. This was a large well-conducted trial involving at least 150 hospitals, with complete follow-up of 99.4% of enrollees. Randomization resulted in equal distribution on the basis of age and sex; no data on race or ethnicity were presented. The concurrent use of heparin agents was equally distributed between the intervention and control groups. Diagnosis of DVT, PE, MI, and cerebrovascular accident was made by a committee unaware of the subjects' group assignment.

OUTCOMES MEASURED: Follow-up was for cause of death through day 35 and for nonfatal events during the hospitalization (mean duration=16 days).

RESULTS: Approximately 13,000 patients with hip fracture were randomized, with 92% initiating treatment preoperatively or immediately postoperatively and 80% taking the assigned medication for the entire 35-day follow-up. Approximately 45% of the participants were receiving a form of heparin (similar numbers in both groups). Symptomatic DVT was confirmed in 1.03% of those assigned aspirin and 1.45% of those assigned placebo (P=.03; number needed to treat [NNT]=238). Definite or probable PE was confirmed in 0.69% of patients taking aspirin and 1.21% of those taking placebo (P=.002; NNT=192). For fatal PE, the NNT was 270 (P=.002). The overall mortality was identical in both groups. Aspirin had statistically significant benefits on total thromboembolic events in those patients taking unfractionated heparin or no heparin but not in those who received LMW heparin. Nonfatal and fatal cardiac ischemic events occurred more frequently in the aspirin group (1.57% vs 1.18%, P=.05). A total of 256 patients would need to receive aspirin for one additional event to occur (number needed to harm [NNH]=256). Fatal bleeding episodes were approximately equal in both groups; however, 2.95% of the patients assigned to aspirin required transfusion versus 2.35% of patients assigned to placebo (P=.04; NNH=167). The number of patients in the arthroplasty section of the trial was smaller (n=4000), and none of the comparisons between the intervention and placebo groups showed statistically significant results.

BACKGROUND: Hypertension is associated with an increased risk of morbidity and mortality. This study was undertaken to evaluate if newer classes of antihypertensive medications reduce the risk of cardiovascular events compared with an active control (diuretics) that has demonstrated a reduction in long-term clinical trials.

POPULATION STUDIED: The Antihypertensive and Lipid-Lowering Treatment to prevent Heart Attack Trial (ALLHAT) includes a total of 42,448 participants recruited from 625 centers in the United States, Canada, Puerto Rico, and the US Virgin Islands. Participants in the trial were men and women 55 years or older with hypertension (systolic blood pressure [SBP] Ž140 mm Hg or diastolic blood pressure [DBP] Ž90 mm Hg, or took medication for hypertension) and at least 1 additional risk factor for coronary heart disease (CHD). Follow-up was scheduled for 6 years. The doxazosin and chlorthalidone groups included 24,335 patients (mean age=67 years; 53% men; 47% non-Hispanic white) who were followed for a median of 3.3 years.

STUDY DESIGN AND VALIDITY: Eligible patients were initially randomized (allocation assignment concealed) to one of the 4 drugs (chlorthalidone, amlodipine, lisinopril, or doxazosin) as a first-line agent and then other drugs (atenolol, reserpine, clonidine, or hydralazine) were added in an open-label fashion, if necessary. The treatment goals were a DBP < 90 mm Hg and an SBP < 140 mm Hg. Following the initial titration visits, participants were seen every 3 months during the first year, and every 4 months thereafter. The randomization process provided an equal percentage of patients previously treated for hypertension, atherosclerotic cardiovascular disease, type 2 diabetes, smoking, low high-density lipoprotein, low ventricular hypertrophy (LVH) by electrocardiogram, and LVH by echocardiogram. The chlorthalidone group and the doxazosin group lost similar percentages of patients to follow-up, 3.2% and 3.7%, respectively, at the time of this analysis. The percentage of patients at years 3 and 4 who were still taking chlorthalidone (or another diuretic) was 87% and 86%, respectively. Of the patients taking doxazosin, only 76% were taking the drug at year 3 and 75% at year 4. By year 4, 40% of the chlorthalidone group was receiving step-up therapy, while 47% of the patients taking doxazosin required additional medications. All of these efforts to provide fair comparisons and use of the intention-to-treat analysis make this a valid comparative study.

OUTCOMES MEASURED: Primary outcome measures were fatal CHD and nonfatal myocardial infarction (MI). Secondary outcome measures included all-cause mortality, stroke, and combined cardiovascular disease (CHD death, nonfatal MI, stroke, angina, coronary revascularization, congestive heart failure [CHF], and peripheral arterial disease).

RESULTS: There was no difference between the doxazosin and chlorthalidone groups in the primary outcomes of fatal CHD or nonfatal MI or the secondary outcome of all-cause mortality. The doxazosin group did, however, have a higher risk of stroke than those patients treated with chlorthalidone (4.2% vs 3.6%; relative risk [RR]=1.19; 95% confidence interval [CI], 1.01-1.40; P=.04; number needed to harm [NNH]=167) and an increased risk of combined cardiovascular disease (25.5% vs 21.8%; RR = 1.25; 95% CI, 1.17-1.33; P <.001; NNH=27). Cardiovascular disease subset analysis demonstrated statistically significant differences in angina, coronary revascularization, and CHF. The difference between the groups was most remarkable for CHF, because a doubling of events occurred with doxazosin compared with chlorthalidone (8.1% vs 4.4%; RR=2.04; 95% CI, 1.79-2.32; P <.001; NNH=27).

BACKGROUND: Hypertension is associated with an increased risk of morbidity and mortality. This study was undertaken to evaluate if newer classes of antihypertensive medications reduce the risk of cardiovascular events compared with an active control (diuretics) that has demonstrated a reduction in long-term clinical trials.

POPULATION STUDIED: The Antihypertensive and Lipid-Lowering Treatment to prevent Heart Attack Trial (ALLHAT) includes a total of 42,448 participants recruited from 625 centers in the United States, Canada, Puerto Rico, and the US Virgin Islands. Participants in the trial were men and women 55 years or older with hypertension (systolic blood pressure [SBP] Ž140 mm Hg or diastolic blood pressure [DBP] Ž90 mm Hg, or took medication for hypertension) and at least 1 additional risk factor for coronary heart disease (CHD). Follow-up was scheduled for 6 years. The doxazosin and chlorthalidone groups included 24,335 patients (mean age=67 years; 53% men; 47% non-Hispanic white) who were followed for a median of 3.3 years.

STUDY DESIGN AND VALIDITY: Eligible patients were initially randomized (allocation assignment concealed) to one of the 4 drugs (chlorthalidone, amlodipine, lisinopril, or doxazosin) as a first-line agent and then other drugs (atenolol, reserpine, clonidine, or hydralazine) were added in an open-label fashion, if necessary. The treatment goals were a DBP < 90 mm Hg and an SBP < 140 mm Hg. Following the initial titration visits, participants were seen every 3 months during the first year, and every 4 months thereafter. The randomization process provided an equal percentage of patients previously treated for hypertension, atherosclerotic cardiovascular disease, type 2 diabetes, smoking, low high-density lipoprotein, low ventricular hypertrophy (LVH) by electrocardiogram, and LVH by echocardiogram. The chlorthalidone group and the doxazosin group lost similar percentages of patients to follow-up, 3.2% and 3.7%, respectively, at the time of this analysis. The percentage of patients at years 3 and 4 who were still taking chlorthalidone (or another diuretic) was 87% and 86%, respectively. Of the patients taking doxazosin, only 76% were taking the drug at year 3 and 75% at year 4. By year 4, 40% of the chlorthalidone group was receiving step-up therapy, while 47% of the patients taking doxazosin required additional medications. All of these efforts to provide fair comparisons and use of the intention-to-treat analysis make this a valid comparative study.

OUTCOMES MEASURED: Primary outcome measures were fatal CHD and nonfatal myocardial infarction (MI). Secondary outcome measures included all-cause mortality, stroke, and combined cardiovascular disease (CHD death, nonfatal MI, stroke, angina, coronary revascularization, congestive heart failure [CHF], and peripheral arterial disease).

RESULTS: There was no difference between the doxazosin and chlorthalidone groups in the primary outcomes of fatal CHD or nonfatal MI or the secondary outcome of all-cause mortality. The doxazosin group did, however, have a higher risk of stroke than those patients treated with chlorthalidone (4.2% vs 3.6%; relative risk [RR]=1.19; 95% confidence interval [CI], 1.01-1.40; P=.04; number needed to harm [NNH]=167) and an increased risk of combined cardiovascular disease (25.5% vs 21.8%; RR = 1.25; 95% CI, 1.17-1.33; P <.001; NNH=27). Cardiovascular disease subset analysis demonstrated statistically significant differences in angina, coronary revascularization, and CHF. The difference between the groups was most remarkable for CHF, because a doubling of events occurred with doxazosin compared with chlorthalidone (8.1% vs 4.4%; RR=2.04; 95% CI, 1.79-2.32; P <.001; NNH=27).

BACKGROUND: Hypertension is associated with an increased risk of morbidity and mortality. This study was undertaken to evaluate if newer classes of antihypertensive medications reduce the risk of cardiovascular events compared with an active control (diuretics) that has demonstrated a reduction in long-term clinical trials.

POPULATION STUDIED: The Antihypertensive and Lipid-Lowering Treatment to prevent Heart Attack Trial (ALLHAT) includes a total of 42,448 participants recruited from 625 centers in the United States, Canada, Puerto Rico, and the US Virgin Islands. Participants in the trial were men and women 55 years or older with hypertension (systolic blood pressure [SBP] Ž140 mm Hg or diastolic blood pressure [DBP] Ž90 mm Hg, or took medication for hypertension) and at least 1 additional risk factor for coronary heart disease (CHD). Follow-up was scheduled for 6 years. The doxazosin and chlorthalidone groups included 24,335 patients (mean age=67 years; 53% men; 47% non-Hispanic white) who were followed for a median of 3.3 years.

STUDY DESIGN AND VALIDITY: Eligible patients were initially randomized (allocation assignment concealed) to one of the 4 drugs (chlorthalidone, amlodipine, lisinopril, or doxazosin) as a first-line agent and then other drugs (atenolol, reserpine, clonidine, or hydralazine) were added in an open-label fashion, if necessary. The treatment goals were a DBP < 90 mm Hg and an SBP < 140 mm Hg. Following the initial titration visits, participants were seen every 3 months during the first year, and every 4 months thereafter. The randomization process provided an equal percentage of patients previously treated for hypertension, atherosclerotic cardiovascular disease, type 2 diabetes, smoking, low high-density lipoprotein, low ventricular hypertrophy (LVH) by electrocardiogram, and LVH by echocardiogram. The chlorthalidone group and the doxazosin group lost similar percentages of patients to follow-up, 3.2% and 3.7%, respectively, at the time of this analysis. The percentage of patients at years 3 and 4 who were still taking chlorthalidone (or another diuretic) was 87% and 86%, respectively. Of the patients taking doxazosin, only 76% were taking the drug at year 3 and 75% at year 4. By year 4, 40% of the chlorthalidone group was receiving step-up therapy, while 47% of the patients taking doxazosin required additional medications. All of these efforts to provide fair comparisons and use of the intention-to-treat analysis make this a valid comparative study.

OUTCOMES MEASURED: Primary outcome measures were fatal CHD and nonfatal myocardial infarction (MI). Secondary outcome measures included all-cause mortality, stroke, and combined cardiovascular disease (CHD death, nonfatal MI, stroke, angina, coronary revascularization, congestive heart failure [CHF], and peripheral arterial disease).

RESULTS: There was no difference between the doxazosin and chlorthalidone groups in the primary outcomes of fatal CHD or nonfatal MI or the secondary outcome of all-cause mortality. The doxazosin group did, however, have a higher risk of stroke than those patients treated with chlorthalidone (4.2% vs 3.6%; relative risk [RR]=1.19; 95% confidence interval [CI], 1.01-1.40; P=.04; number needed to harm [NNH]=167) and an increased risk of combined cardiovascular disease (25.5% vs 21.8%; RR = 1.25; 95% CI, 1.17-1.33; P <.001; NNH=27). Cardiovascular disease subset analysis demonstrated statistically significant differences in angina, coronary revascularization, and CHF. The difference between the groups was most remarkable for CHF, because a doubling of events occurred with doxazosin compared with chlorthalidone (8.1% vs 4.4%; RR=2.04; 95% CI, 1.79-2.32; P <.001; NNH=27).

BACKGROUND: Previous studies have shown the median hospital stay for AMI is 9 days. Shortening stays could reduce health care costs. The Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries-1 (GUSTO-1) study showed that patients without cardiac complications 4 days after thrombolysis had 30-day mortality rates of 1%. The authors took this a step further: If the mortality rate for a 4-day uncomplicated hospitalization was low, would a 3-day stay yield similar results? This clinical decision analysis determined whether reducing hospital stays to 3 days would be economically feasible for patients receiving thrombolytics for AMI.

POPULATION STUDIED: GUSTO-1 was an international randomized control trial in which 4 different thrombolytic regimens were compared for more than 41,000 people. The authors of this study used data from the 22,000 people who had uncomplicated 72-hour hospital courses after thrombolysis; “uncomplicated” was defined as the absence of death or serious cardiovascular events.

STUDY DESIGN AND VALIDITY: The authors of this decision analysis compared the cost per year of life saved by discharging patients without complications at 72 hours instead of 96 hours after thrombolysis. The researchers determined the number of patients who survived a treatable ventricular arrhythmia after 3 days and assumed that these patients would have died in an outpatient setting. The primary costs considered were the additional nursing and bed charges for the fourth day. Ancillary costs, such as laboratory testing and stress testing for risk stratification, were assumed to have occurred at a comparable frequency and cost in the outpatient setting. Costs were obtained from reasonable sources with appropriate discounting (discounting means that a benefit in the future is less highly valued than a benefit in the present).

OUTCOMES MEASURED: The primary outcome was cost per life-year saved. Other outcomes included the rates of death, ventricular arrhythmia, and other cardiovascular events.

RESULTS: The rate of ventricular arrhythmia for all 41,000 GUSTO-1 subjects dropped sharply after 48 hours. Of the 22,000 patients with uncomplicated courses at 72 hours, 16 had a subsequent serious ventricular arrhythmia (1 event per 1400 patients). Three patients with arrhythmias died, while 13 survived at least 24 hours. Patients who suffered a ventricular arrhythmia between days 3 and 4 had a 1-year mortality rate of 9.5%, 4 times that of those without arrhythmia. The authors concluded that hospitalizing a patient without complications for 4 days added an average of 0.006 years of life (slightly more than 2 days). This extra hospital day costs $105,629 per year of life saved. The sensitivity analysis showed that the range of costs is between $66,000 and $184,000 per year of life saved.

BACKGROUND: Previous studies have shown the median hospital stay for AMI is 9 days. Shortening stays could reduce health care costs. The Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries-1 (GUSTO-1) study showed that patients without cardiac complications 4 days after thrombolysis had 30-day mortality rates of 1%. The authors took this a step further: If the mortality rate for a 4-day uncomplicated hospitalization was low, would a 3-day stay yield similar results? This clinical decision analysis determined whether reducing hospital stays to 3 days would be economically feasible for patients receiving thrombolytics for AMI.

POPULATION STUDIED: GUSTO-1 was an international randomized control trial in which 4 different thrombolytic regimens were compared for more than 41,000 people. The authors of this study used data from the 22,000 people who had uncomplicated 72-hour hospital courses after thrombolysis; “uncomplicated” was defined as the absence of death or serious cardiovascular events.

STUDY DESIGN AND VALIDITY: The authors of this decision analysis compared the cost per year of life saved by discharging patients without complications at 72 hours instead of 96 hours after thrombolysis. The researchers determined the number of patients who survived a treatable ventricular arrhythmia after 3 days and assumed that these patients would have died in an outpatient setting. The primary costs considered were the additional nursing and bed charges for the fourth day. Ancillary costs, such as laboratory testing and stress testing for risk stratification, were assumed to have occurred at a comparable frequency and cost in the outpatient setting. Costs were obtained from reasonable sources with appropriate discounting (discounting means that a benefit in the future is less highly valued than a benefit in the present).

OUTCOMES MEASURED: The primary outcome was cost per life-year saved. Other outcomes included the rates of death, ventricular arrhythmia, and other cardiovascular events.

RESULTS: The rate of ventricular arrhythmia for all 41,000 GUSTO-1 subjects dropped sharply after 48 hours. Of the 22,000 patients with uncomplicated courses at 72 hours, 16 had a subsequent serious ventricular arrhythmia (1 event per 1400 patients). Three patients with arrhythmias died, while 13 survived at least 24 hours. Patients who suffered a ventricular arrhythmia between days 3 and 4 had a 1-year mortality rate of 9.5%, 4 times that of those without arrhythmia. The authors concluded that hospitalizing a patient without complications for 4 days added an average of 0.006 years of life (slightly more than 2 days). This extra hospital day costs $105,629 per year of life saved. The sensitivity analysis showed that the range of costs is between $66,000 and $184,000 per year of life saved.

BACKGROUND: Previous studies have shown the median hospital stay for AMI is 9 days. Shortening stays could reduce health care costs. The Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries-1 (GUSTO-1) study showed that patients without cardiac complications 4 days after thrombolysis had 30-day mortality rates of 1%. The authors took this a step further: If the mortality rate for a 4-day uncomplicated hospitalization was low, would a 3-day stay yield similar results? This clinical decision analysis determined whether reducing hospital stays to 3 days would be economically feasible for patients receiving thrombolytics for AMI.

POPULATION STUDIED: GUSTO-1 was an international randomized control trial in which 4 different thrombolytic regimens were compared for more than 41,000 people. The authors of this study used data from the 22,000 people who had uncomplicated 72-hour hospital courses after thrombolysis; “uncomplicated” was defined as the absence of death or serious cardiovascular events.

STUDY DESIGN AND VALIDITY: The authors of this decision analysis compared the cost per year of life saved by discharging patients without complications at 72 hours instead of 96 hours after thrombolysis. The researchers determined the number of patients who survived a treatable ventricular arrhythmia after 3 days and assumed that these patients would have died in an outpatient setting. The primary costs considered were the additional nursing and bed charges for the fourth day. Ancillary costs, such as laboratory testing and stress testing for risk stratification, were assumed to have occurred at a comparable frequency and cost in the outpatient setting. Costs were obtained from reasonable sources with appropriate discounting (discounting means that a benefit in the future is less highly valued than a benefit in the present).

OUTCOMES MEASURED: The primary outcome was cost per life-year saved. Other outcomes included the rates of death, ventricular arrhythmia, and other cardiovascular events.

RESULTS: The rate of ventricular arrhythmia for all 41,000 GUSTO-1 subjects dropped sharply after 48 hours. Of the 22,000 patients with uncomplicated courses at 72 hours, 16 had a subsequent serious ventricular arrhythmia (1 event per 1400 patients). Three patients with arrhythmias died, while 13 survived at least 24 hours. Patients who suffered a ventricular arrhythmia between days 3 and 4 had a 1-year mortality rate of 9.5%, 4 times that of those without arrhythmia. The authors concluded that hospitalizing a patient without complications for 4 days added an average of 0.006 years of life (slightly more than 2 days). This extra hospital day costs $105,629 per year of life saved. The sensitivity analysis showed that the range of costs is between $66,000 and $184,000 per year of life saved.

BACKGROUND: The current recommendation for management of uncomplicated pyelonephritis is a 14-day course of ciprofloxacin.¹ Studies have shown this to be 90% effective.² Few studies have evaluated shorter courses of therapy for pyelonephritis. The purpose of this study was to compare the effectiveness of 7 days of ciprofloxacin with 14 days of trimethoprim-sulfamethoxazole (TMP-SMZ) for treatment of acute uncomplicated pyelonephritis in an outpatient setting.

POPULATION STUDIED: The study participants were premenopausal women aged 18 years or older with a clinical diagnosis of pyelonephritis, defined as flank pain or costovertebral angle tenderness, fever, and pyuria. Patients were excluded if they had abnormal renal function (creatinine >2.7), severe sepsis, urologic abnormalities, persistent vomiting, or if they were immunocompromised, had diabetes, were admitted to the hospital, or were pregnant or lactating.

STUDY DESIGN AND VALIDITY: A total of 378 patients were randomized to receive either 7 days of ciprofloxacin 500 mg twice daily or 14 days of TMP-SMZ 800/160 mg twice daily. In both groups, managing physicians had the option to treat patients with an initial dose of intravenous antibiotic, if clinically indicated (400 mg of ciprofloxacin in the ciprofloxacin group or 1 gram of ceftriaxone in the TMP-SMZ group). Blood cultures and a urine culture were obtained by either clean catch or catheterization before therapy was initiated. Urine cultures were repeated on days 3 to 5 of treatment. Patients were evaluated at 4 to 11 days and 22 to 48 days following treatment to assess for clinical cure and to repeat urine cultures.

OUTCOMES MEASURED: Primary study outcomes were bacteriologic and clinical cures at a visit 4 to 11 days post-therapy, as determined by urine culture and signs and symptoms, respectively. Secondary outcomes included bacteriologic and clinical responses at the visit 22 to 48 days post-therapy, adverse drug events, and a health resource analysis.

RESULTS: At 4 to 11 days post-therapy, the efficacy analysis showed that patients treated with ciprofloxacin had a better bacteriologic cure rate than the patients treated with TMP-SMZ (99% vs 89%, P=.004; number needed to treat [NNT]=10) and a better clinical cure rate (96% vs 83%, P=.002; NNT=7.6). Escherichia coli represented 90% of cultured organisms, of which 18% were resistant to TMP-SMZ, and <1% were resistant to ciprofloxacin. When the analysis was done using only organisms sensitive to TMP-SMZ, the efficacy rates were similar. An initial IV dose was associated with a greater cure rate in the TMP-SMZ group, but not in the ciprofloxacin group. In the intention-to-treat analysis, clinical cure rates for ciprofloxacin were better than TMP-SMZ at 22 to 48 days post-therapy (82% vs 72%; 95% confidence interval [CI], 0.01-0.19; NNT=10). Benefits were also seen when comparing ciprofloxacin with TMP-SMZ for bacteriologic cure (84% vs 74%; 95% CI, 0.01-0.19; NNT=10). Adverse events occurred in 24% of the ciprofloxacin and 33% of the TMP-SMZ group; 6% of patients taking ciprofloxacin and 11% taking TMP-SMZ discontinued the drug. The cost per cure was $615 for ciprofloxacin compared with $770 for TMP-SMZ; however, this study did not have enough power to detect a statistically significant difference.

BACKGROUND: The current recommendation for management of uncomplicated pyelonephritis is a 14-day course of ciprofloxacin.¹ Studies have shown this to be 90% effective.² Few studies have evaluated shorter courses of therapy for pyelonephritis. The purpose of this study was to compare the effectiveness of 7 days of ciprofloxacin with 14 days of trimethoprim-sulfamethoxazole (TMP-SMZ) for treatment of acute uncomplicated pyelonephritis in an outpatient setting.

POPULATION STUDIED: The study participants were premenopausal women aged 18 years or older with a clinical diagnosis of pyelonephritis, defined as flank pain or costovertebral angle tenderness, fever, and pyuria. Patients were excluded if they had abnormal renal function (creatinine >2.7), severe sepsis, urologic abnormalities, persistent vomiting, or if they were immunocompromised, had diabetes, were admitted to the hospital, or were pregnant or lactating.

STUDY DESIGN AND VALIDITY: A total of 378 patients were randomized to receive either 7 days of ciprofloxacin 500 mg twice daily or 14 days of TMP-SMZ 800/160 mg twice daily. In both groups, managing physicians had the option to treat patients with an initial dose of intravenous antibiotic, if clinically indicated (400 mg of ciprofloxacin in the ciprofloxacin group or 1 gram of ceftriaxone in the TMP-SMZ group). Blood cultures and a urine culture were obtained by either clean catch or catheterization before therapy was initiated. Urine cultures were repeated on days 3 to 5 of treatment. Patients were evaluated at 4 to 11 days and 22 to 48 days following treatment to assess for clinical cure and to repeat urine cultures.

OUTCOMES MEASURED: Primary study outcomes were bacteriologic and clinical cures at a visit 4 to 11 days post-therapy, as determined by urine culture and signs and symptoms, respectively. Secondary outcomes included bacteriologic and clinical responses at the visit 22 to 48 days post-therapy, adverse drug events, and a health resource analysis.

RESULTS: At 4 to 11 days post-therapy, the efficacy analysis showed that patients treated with ciprofloxacin had a better bacteriologic cure rate than the patients treated with TMP-SMZ (99% vs 89%, P=.004; number needed to treat [NNT]=10) and a better clinical cure rate (96% vs 83%, P=.002; NNT=7.6). Escherichia coli represented 90% of cultured organisms, of which 18% were resistant to TMP-SMZ, and <1% were resistant to ciprofloxacin. When the analysis was done using only organisms sensitive to TMP-SMZ, the efficacy rates were similar. An initial IV dose was associated with a greater cure rate in the TMP-SMZ group, but not in the ciprofloxacin group. In the intention-to-treat analysis, clinical cure rates for ciprofloxacin were better than TMP-SMZ at 22 to 48 days post-therapy (82% vs 72%; 95% confidence interval [CI], 0.01-0.19; NNT=10). Benefits were also seen when comparing ciprofloxacin with TMP-SMZ for bacteriologic cure (84% vs 74%; 95% CI, 0.01-0.19; NNT=10). Adverse events occurred in 24% of the ciprofloxacin and 33% of the TMP-SMZ group; 6% of patients taking ciprofloxacin and 11% taking TMP-SMZ discontinued the drug. The cost per cure was $615 for ciprofloxacin compared with $770 for TMP-SMZ; however, this study did not have enough power to detect a statistically significant difference.

BACKGROUND: The current recommendation for management of uncomplicated pyelonephritis is a 14-day course of ciprofloxacin.¹ Studies have shown this to be 90% effective.² Few studies have evaluated shorter courses of therapy for pyelonephritis. The purpose of this study was to compare the effectiveness of 7 days of ciprofloxacin with 14 days of trimethoprim-sulfamethoxazole (TMP-SMZ) for treatment of acute uncomplicated pyelonephritis in an outpatient setting.

POPULATION STUDIED: The study participants were premenopausal women aged 18 years or older with a clinical diagnosis of pyelonephritis, defined as flank pain or costovertebral angle tenderness, fever, and pyuria. Patients were excluded if they had abnormal renal function (creatinine >2.7), severe sepsis, urologic abnormalities, persistent vomiting, or if they were immunocompromised, had diabetes, were admitted to the hospital, or were pregnant or lactating.

STUDY DESIGN AND VALIDITY: A total of 378 patients were randomized to receive either 7 days of ciprofloxacin 500 mg twice daily or 14 days of TMP-SMZ 800/160 mg twice daily. In both groups, managing physicians had the option to treat patients with an initial dose of intravenous antibiotic, if clinically indicated (400 mg of ciprofloxacin in the ciprofloxacin group or 1 gram of ceftriaxone in the TMP-SMZ group). Blood cultures and a urine culture were obtained by either clean catch or catheterization before therapy was initiated. Urine cultures were repeated on days 3 to 5 of treatment. Patients were evaluated at 4 to 11 days and 22 to 48 days following treatment to assess for clinical cure and to repeat urine cultures.

OUTCOMES MEASURED: Primary study outcomes were bacteriologic and clinical cures at a visit 4 to 11 days post-therapy, as determined by urine culture and signs and symptoms, respectively. Secondary outcomes included bacteriologic and clinical responses at the visit 22 to 48 days post-therapy, adverse drug events, and a health resource analysis.

RESULTS: At 4 to 11 days post-therapy, the efficacy analysis showed that patients treated with ciprofloxacin had a better bacteriologic cure rate than the patients treated with TMP-SMZ (99% vs 89%, P=.004; number needed to treat [NNT]=10) and a better clinical cure rate (96% vs 83%, P=.002; NNT=7.6). Escherichia coli represented 90% of cultured organisms, of which 18% were resistant to TMP-SMZ, and <1% were resistant to ciprofloxacin. When the analysis was done using only organisms sensitive to TMP-SMZ, the efficacy rates were similar. An initial IV dose was associated with a greater cure rate in the TMP-SMZ group, but not in the ciprofloxacin group. In the intention-to-treat analysis, clinical cure rates for ciprofloxacin were better than TMP-SMZ at 22 to 48 days post-therapy (82% vs 72%; 95% confidence interval [CI], 0.01-0.19; NNT=10). Benefits were also seen when comparing ciprofloxacin with TMP-SMZ for bacteriologic cure (84% vs 74%; 95% CI, 0.01-0.19; NNT=10). Adverse events occurred in 24% of the ciprofloxacin and 33% of the TMP-SMZ group; 6% of patients taking ciprofloxacin and 11% taking TMP-SMZ discontinued the drug. The cost per cure was $615 for ciprofloxacin compared with $770 for TMP-SMZ; however, this study did not have enough power to detect a statistically significant difference.

BACKGROUND: National Institutes of Health guidelines for the treatment of obesity recommend a 10% weight loss from initial bodyweight. Pharma-co therapy may help achieve this goal. However, some agents have been associated with serious side effects and inadequate long-term effectiveness.

POPULATION STUDIED: The authors recruited 796 obese men and women with a body mass index (BMI) of 30 to 44 kg per m² from 17 primary care practices. They excluded patients who had any clinically significant chronic conditions; had recently changed smoking status; were taking drugs known to affect body weight; or with substance abuse, eating disorders, or reported weight loss of 4 kg or more. The study population was 90% white and 78% women.

STUDY DESIGN AND VALIDITY: This was a randomized double-blind placebo-controlled multicenter trial. Eligible subjects were prescribed a reduced energy diet (1200-1500 kcal/day) at the beginning of a 4-week single-blind placebo run-in period. The 635 patients who were at least 75% compliant by placebo pill count were randomized to receive placebo, orlistat 60 mg 3 times daily, or orlistat 120 mg 3 times daily for 1 year. Allocation was adequately concealed (personal communication) and intention-to-treat analysis was reported. Calories were increased by 300 per day for those patients who continued to lose weight at the end of the first year, and therapy was continued for another year. Patients watched behavior modification videos and received written weight management materials. Patients were asked to increase physical activity, but compliance was not monitored. Physicians assessed body weight, vital signs, and adverse events 10 times during the trial. At the end of 2 years, 57% of the placebo group and 46% of patients in each of the orlistat groups had withdrawn from the study.

OUTCOMES MEASURED: Change in body weight was the primary end point. Secondary end points were blood pressure, adverse drug effects, and serum lipid, glucose, and insulin levels.

RESULTS: At the end of 1 year more patients in the orlistat group lost at least 5% of their initial body weight than patients taking the placebo (50.5% vs 30.7%, P <.001; number needed to treat [NNT]=5.1). At the end of 2 years, this trend continued (34.3% vs 24.1%, P=.02; NNT=9.8). At the end of 1 year, more patients in the orlistat group lost 10% or more of their initial body weight than the patients taking the placebo (28.6% vs 11.3%, P <.001; NNT=5.8). The trend continued at 2 years (18.6% vs 6.6%, P=.001; NNT=8.3). Body weight lost was 1.65±0.62 kg in the placebo group and 5.02±0.73 kg in the orlistat group at the end of 2 years.

BACKGROUND: National Institutes of Health guidelines for the treatment of obesity recommend a 10% weight loss from initial bodyweight. Pharma-co therapy may help achieve this goal. However, some agents have been associated with serious side effects and inadequate long-term effectiveness.

POPULATION STUDIED: The authors recruited 796 obese men and women with a body mass index (BMI) of 30 to 44 kg per m² from 17 primary care practices. They excluded patients who had any clinically significant chronic conditions; had recently changed smoking status; were taking drugs known to affect body weight; or with substance abuse, eating disorders, or reported weight loss of 4 kg or more. The study population was 90% white and 78% women.

STUDY DESIGN AND VALIDITY: This was a randomized double-blind placebo-controlled multicenter trial. Eligible subjects were prescribed a reduced energy diet (1200-1500 kcal/day) at the beginning of a 4-week single-blind placebo run-in period. The 635 patients who were at least 75% compliant by placebo pill count were randomized to receive placebo, orlistat 60 mg 3 times daily, or orlistat 120 mg 3 times daily for 1 year. Allocation was adequately concealed (personal communication) and intention-to-treat analysis was reported. Calories were increased by 300 per day for those patients who continued to lose weight at the end of the first year, and therapy was continued for another year. Patients watched behavior modification videos and received written weight management materials. Patients were asked to increase physical activity, but compliance was not monitored. Physicians assessed body weight, vital signs, and adverse events 10 times during the trial. At the end of 2 years, 57% of the placebo group and 46% of patients in each of the orlistat groups had withdrawn from the study.

OUTCOMES MEASURED: Change in body weight was the primary end point. Secondary end points were blood pressure, adverse drug effects, and serum lipid, glucose, and insulin levels.

RESULTS: At the end of 1 year more patients in the orlistat group lost at least 5% of their initial body weight than patients taking the placebo (50.5% vs 30.7%, P <.001; number needed to treat [NNT]=5.1). At the end of 2 years, this trend continued (34.3% vs 24.1%, P=.02; NNT=9.8). At the end of 1 year, more patients in the orlistat group lost 10% or more of their initial body weight than the patients taking the placebo (28.6% vs 11.3%, P <.001; NNT=5.8). The trend continued at 2 years (18.6% vs 6.6%, P=.001; NNT=8.3). Body weight lost was 1.65±0.62 kg in the placebo group and 5.02±0.73 kg in the orlistat group at the end of 2 years.

BACKGROUND: National Institutes of Health guidelines for the treatment of obesity recommend a 10% weight loss from initial bodyweight. Pharma-co therapy may help achieve this goal. However, some agents have been associated with serious side effects and inadequate long-term effectiveness.

POPULATION STUDIED: The authors recruited 796 obese men and women with a body mass index (BMI) of 30 to 44 kg per m² from 17 primary care practices. They excluded patients who had any clinically significant chronic conditions; had recently changed smoking status; were taking drugs known to affect body weight; or with substance abuse, eating disorders, or reported weight loss of 4 kg or more. The study population was 90% white and 78% women.

STUDY DESIGN AND VALIDITY: This was a randomized double-blind placebo-controlled multicenter trial. Eligible subjects were prescribed a reduced energy diet (1200-1500 kcal/day) at the beginning of a 4-week single-blind placebo run-in period. The 635 patients who were at least 75% compliant by placebo pill count were randomized to receive placebo, orlistat 60 mg 3 times daily, or orlistat 120 mg 3 times daily for 1 year. Allocation was adequately concealed (personal communication) and intention-to-treat analysis was reported. Calories were increased by 300 per day for those patients who continued to lose weight at the end of the first year, and therapy was continued for another year. Patients watched behavior modification videos and received written weight management materials. Patients were asked to increase physical activity, but compliance was not monitored. Physicians assessed body weight, vital signs, and adverse events 10 times during the trial. At the end of 2 years, 57% of the placebo group and 46% of patients in each of the orlistat groups had withdrawn from the study.

OUTCOMES MEASURED: Change in body weight was the primary end point. Secondary end points were blood pressure, adverse drug effects, and serum lipid, glucose, and insulin levels.

RESULTS: At the end of 1 year more patients in the orlistat group lost at least 5% of their initial body weight than patients taking the placebo (50.5% vs 30.7%, P <.001; number needed to treat [NNT]=5.1). At the end of 2 years, this trend continued (34.3% vs 24.1%, P=.02; NNT=9.8). At the end of 1 year, more patients in the orlistat group lost 10% or more of their initial body weight than the patients taking the placebo (28.6% vs 11.3%, P <.001; NNT=5.8). The trend continued at 2 years (18.6% vs 6.6%, P=.001; NNT=8.3). Body weight lost was 1.65±0.62 kg in the placebo group and 5.02±0.73 kg in the orlistat group at the end of 2 years.

BACKGROUND: Osteoarthritis is a major problem in primary care, but its optimal management is unclear. Nonsteroidal anti-inflammatory drugs (NSAIDs), analgesics, and intra-articular corticosteroids have significant toxicity and do not alter disease progression; thus, many patients have turned to glucosamine and chondroitin.

POPULATION STUDIED: The authors of this meta-analysis reviewed 15 randomized double-blind placebo-controlled trials of glucosamine and chondroitin (both together and separately) for patients with osteoarthritis of the knee or hip. Trials were included if they were longer than 4 weeks in duration and used an accepted osteoarthritis outcome measure. A total of 1710 patients were enrolled; no information was provided on age, sex, severity of osteoarthritis, or other treatments, making generalizability to a typical family practice difficult.

STUDY DESIGN AND VALIDITY: The authors’ search included MEDLINE, the Cochrane Controlled Trials Registry, rheumatology meeting abstracts, citation lists from review articles, and consultation with experts. Two trained reviewers independently examined outcomes, quality, and financial sponsorship of the articles. Disagreements were resolved by discussion. A pooled effect size was calculated by dividing the difference in mean outcomes between treatment group and control group by the standard deviation of the outcome value in the control group. The resulting effect measure combines pain outcomes with disability outcomes. Publication bias was addressed by funnel plots and regression of effects on the inverse of study variance.

OUTCOMES MEASURED: The primary outcome measured was improvement in symptoms at 4 weeks, measured by either a pain scale or a disability index. The article did not address other outcomes important for osteoarthritis, such as quality of life, cost, side effects, or radiographic progression of osteoarthritis.

RESULTS: Interrater reliability of the reviewers was excellent. The quality of the available studies was relatively poor, with only 2 employing intention-to-treat analysis. None of the studies reported independent funding from a governmental or not-for-profit organization, and there was evidence of a publication bias toward positive trials. Glucosamine showed an effect size of 0.44 (95% confidence interval [CI], 0.24-0.64), and chondroitin had an effect size of 0.96 (95% CI, 0.63-1.3). An effect size of 0 indicates equivalency with placebo, with less than 0.2 indicating a small effect, 0.2 to 0.8 a moderate effect, and greater than 0.8 a large effect. The studies were heterogenous; one trial with chondroitin seemed to be significantly different from the others. Excluding that trial weakened but did not eliminate the pooled effect of chondroitin. Shorter duration of treatment was associated with a lower effect size. Poorer-quality studies showed a larger effect than better-quality studies.

BACKGROUND: Osteoarthritis is a major problem in primary care, but its optimal management is unclear. Nonsteroidal anti-inflammatory drugs (NSAIDs), analgesics, and intra-articular corticosteroids have significant toxicity and do not alter disease progression; thus, many patients have turned to glucosamine and chondroitin.

POPULATION STUDIED: The authors of this meta-analysis reviewed 15 randomized double-blind placebo-controlled trials of glucosamine and chondroitin (both together and separately) for patients with osteoarthritis of the knee or hip. Trials were included if they were longer than 4 weeks in duration and used an accepted osteoarthritis outcome measure. A total of 1710 patients were enrolled; no information was provided on age, sex, severity of osteoarthritis, or other treatments, making generalizability to a typical family practice difficult.

STUDY DESIGN AND VALIDITY: The authors’ search included MEDLINE, the Cochrane Controlled Trials Registry, rheumatology meeting abstracts, citation lists from review articles, and consultation with experts. Two trained reviewers independently examined outcomes, quality, and financial sponsorship of the articles. Disagreements were resolved by discussion. A pooled effect size was calculated by dividing the difference in mean outcomes between treatment group and control group by the standard deviation of the outcome value in the control group. The resulting effect measure combines pain outcomes with disability outcomes. Publication bias was addressed by funnel plots and regression of effects on the inverse of study variance.

OUTCOMES MEASURED: The primary outcome measured was improvement in symptoms at 4 weeks, measured by either a pain scale or a disability index. The article did not address other outcomes important for osteoarthritis, such as quality of life, cost, side effects, or radiographic progression of osteoarthritis.

RESULTS: Interrater reliability of the reviewers was excellent. The quality of the available studies was relatively poor, with only 2 employing intention-to-treat analysis. None of the studies reported independent funding from a governmental or not-for-profit organization, and there was evidence of a publication bias toward positive trials. Glucosamine showed an effect size of 0.44 (95% confidence interval [CI], 0.24-0.64), and chondroitin had an effect size of 0.96 (95% CI, 0.63-1.3). An effect size of 0 indicates equivalency with placebo, with less than 0.2 indicating a small effect, 0.2 to 0.8 a moderate effect, and greater than 0.8 a large effect. The studies were heterogenous; one trial with chondroitin seemed to be significantly different from the others. Excluding that trial weakened but did not eliminate the pooled effect of chondroitin. Shorter duration of treatment was associated with a lower effect size. Poorer-quality studies showed a larger effect than better-quality studies.

BACKGROUND: Osteoarthritis is a major problem in primary care, but its optimal management is unclear. Nonsteroidal anti-inflammatory drugs (NSAIDs), analgesics, and intra-articular corticosteroids have significant toxicity and do not alter disease progression; thus, many patients have turned to glucosamine and chondroitin.

POPULATION STUDIED: The authors of this meta-analysis reviewed 15 randomized double-blind placebo-controlled trials of glucosamine and chondroitin (both together and separately) for patients with osteoarthritis of the knee or hip. Trials were included if they were longer than 4 weeks in duration and used an accepted osteoarthritis outcome measure. A total of 1710 patients were enrolled; no information was provided on age, sex, severity of osteoarthritis, or other treatments, making generalizability to a typical family practice difficult.

STUDY DESIGN AND VALIDITY: The authors’ search included MEDLINE, the Cochrane Controlled Trials Registry, rheumatology meeting abstracts, citation lists from review articles, and consultation with experts. Two trained reviewers independently examined outcomes, quality, and financial sponsorship of the articles. Disagreements were resolved by discussion. A pooled effect size was calculated by dividing the difference in mean outcomes between treatment group and control group by the standard deviation of the outcome value in the control group. The resulting effect measure combines pain outcomes with disability outcomes. Publication bias was addressed by funnel plots and regression of effects on the inverse of study variance.

OUTCOMES MEASURED: The primary outcome measured was improvement in symptoms at 4 weeks, measured by either a pain scale or a disability index. The article did not address other outcomes important for osteoarthritis, such as quality of life, cost, side effects, or radiographic progression of osteoarthritis.

RESULTS: Interrater reliability of the reviewers was excellent. The quality of the available studies was relatively poor, with only 2 employing intention-to-treat analysis. None of the studies reported independent funding from a governmental or not-for-profit organization, and there was evidence of a publication bias toward positive trials. Glucosamine showed an effect size of 0.44 (95% confidence interval [CI], 0.24-0.64), and chondroitin had an effect size of 0.96 (95% CI, 0.63-1.3). An effect size of 0 indicates equivalency with placebo, with less than 0.2 indicating a small effect, 0.2 to 0.8 a moderate effect, and greater than 0.8 a large effect. The studies were heterogenous; one trial with chondroitin seemed to be significantly different from the others. Excluding that trial weakened but did not eliminate the pooled effect of chondroitin. Shorter duration of treatment was associated with a lower effect size. Poorer-quality studies showed a larger effect than better-quality studies.

BACKGROUND: Previous meta-analyses have demonstrated that b-blockers reduce morbidity (symptoms, left ventricular function, and rate of hospitalization) and mortality (cardiovascular and total) in patients with congestive heart failure (CHF). The authors of this meta-analysis explored differences in effectiveness between vasodilating and nonvasodilating agents and defined differences in outcomes related to the underlying cause of CHF.

POPULATION STUDIED: A total of 5849 patients (79% men) were studied for a median duration of 6 months. Approximately 50% received b-blockers. Both vasodilating b-blockers (carvedilol and bucindolol) and nonvasodilating b-blockers (metoprolol, bisoprolol, and propranolol) were used for patients with primarily class II and III heart failure. Most patients also received angiotensin-converting enzyme inhibitors, diuretics, and digitalis.

STUDY DESIGN AND VALIDITY: The authors of this meta-analysis combined the results of 21 randomized controlled trials comparing the various b-blockers with placebo. All studies evaluated mortality on an intention-to-treat basis.

OUTCOMES MEASURED: The primary outcomes were total and cardiovascular mortality (including death due to CHF, sudden death, myocardial infarction, and other cardiovascular causes) and hospitalizations. Outcomes were evaluated on the basis of the use of vasodilating and nonvasodilating b-blockers and the underlying cause of CHF (ischemic heart disease vs nonischemic heart disease).

RESULTS: b-blockers as a group reduced total mortality in patients with CHF (10.6% vs 16.6%, relative risk [RR]=0.71; 95% confidence interval [CI], 0.63-0.80; number needed to treat [NNT]=17). A similar benefit was demonstrated in cardiovascular mortality (9.0% vs 13.7%, RR=0.71; 95% CI, 0.59-0.86; NNT=21). The reduction in cardiovascular mortality was mainly because of a significant reduction in death due to CHF and sudden death.

BACKGROUND: Previous meta-analyses have demonstrated that b-blockers reduce morbidity (symptoms, left ventricular function, and rate of hospitalization) and mortality (cardiovascular and total) in patients with congestive heart failure (CHF). The authors of this meta-analysis explored differences in effectiveness between vasodilating and nonvasodilating agents and defined differences in outcomes related to the underlying cause of CHF.

POPULATION STUDIED: A total of 5849 patients (79% men) were studied for a median duration of 6 months. Approximately 50% received b-blockers. Both vasodilating b-blockers (carvedilol and bucindolol) and nonvasodilating b-blockers (metoprolol, bisoprolol, and propranolol) were used for patients with primarily class II and III heart failure. Most patients also received angiotensin-converting enzyme inhibitors, diuretics, and digitalis.

STUDY DESIGN AND VALIDITY: The authors of this meta-analysis combined the results of 21 randomized controlled trials comparing the various b-blockers with placebo. All studies evaluated mortality on an intention-to-treat basis.

OUTCOMES MEASURED: The primary outcomes were total and cardiovascular mortality (including death due to CHF, sudden death, myocardial infarction, and other cardiovascular causes) and hospitalizations. Outcomes were evaluated on the basis of the use of vasodilating and nonvasodilating b-blockers and the underlying cause of CHF (ischemic heart disease vs nonischemic heart disease).

RESULTS: b-blockers as a group reduced total mortality in patients with CHF (10.6% vs 16.6%, relative risk [RR]=0.71; 95% confidence interval [CI], 0.63-0.80; number needed to treat [NNT]=17). A similar benefit was demonstrated in cardiovascular mortality (9.0% vs 13.7%, RR=0.71; 95% CI, 0.59-0.86; NNT=21). The reduction in cardiovascular mortality was mainly because of a significant reduction in death due to CHF and sudden death.

BACKGROUND: Previous meta-analyses have demonstrated that b-blockers reduce morbidity (symptoms, left ventricular function, and rate of hospitalization) and mortality (cardiovascular and total) in patients with congestive heart failure (CHF). The authors of this meta-analysis explored differences in effectiveness between vasodilating and nonvasodilating agents and defined differences in outcomes related to the underlying cause of CHF.

POPULATION STUDIED: A total of 5849 patients (79% men) were studied for a median duration of 6 months. Approximately 50% received b-blockers. Both vasodilating b-blockers (carvedilol and bucindolol) and nonvasodilating b-blockers (metoprolol, bisoprolol, and propranolol) were used for patients with primarily class II and III heart failure. Most patients also received angiotensin-converting enzyme inhibitors, diuretics, and digitalis.

STUDY DESIGN AND VALIDITY: The authors of this meta-analysis combined the results of 21 randomized controlled trials comparing the various b-blockers with placebo. All studies evaluated mortality on an intention-to-treat basis.

OUTCOMES MEASURED: The primary outcomes were total and cardiovascular mortality (including death due to CHF, sudden death, myocardial infarction, and other cardiovascular causes) and hospitalizations. Outcomes were evaluated on the basis of the use of vasodilating and nonvasodilating b-blockers and the underlying cause of CHF (ischemic heart disease vs nonischemic heart disease).

RESULTS: b-blockers as a group reduced total mortality in patients with CHF (10.6% vs 16.6%, relative risk [RR]=0.71; 95% confidence interval [CI], 0.63-0.80; number needed to treat [NNT]=17). A similar benefit was demonstrated in cardiovascular mortality (9.0% vs 13.7%, RR=0.71; 95% CI, 0.59-0.86; NNT=21). The reduction in cardiovascular mortality was mainly because of a significant reduction in death due to CHF and sudden death.

BACKGROUND: Given the high cost of neonatal intensive care and the serious sequelae associated with preterm delivery, an inexpensive safe treatment that would reduce the risk of preterm delivery is appealing. One previous study demonstrated a small decrease in the rate of preterm delivery using oral erythromycin plus metronidazole in a population of women at high risk of preterm delivery (mostly women with a previous preterm delivery)¹; a small study of women at low risk of preterm delivery, however, found no benefit.²

POPULATION STUDIED: Women presenting for prenatal care were considered for randomization. Exclusions were reasonable and included women with: symptoms of vaginitis, contraindications to metronidazole, recent or anticipated antibiotic use, preterm labor, planned or current cervical cerclage, multifetal gestation, failure to have a postscreening visit before 24-weeks’ gestation, and unanticipated use of antibiotics between the screening and the postscreening visit. This was a group of fairly low-risk and adherent patients.

STUDY DESIGN AND VALIDITY: This was a randomized placebo-controlled double-blinded study. Women were considered to have BV if they had a vaginal hydrogen ion concentration higher than 4.4 and a positive Gram’s stain using reasonable criteria well described by the authors. Women with BV who were seen before 24 weeks’ gestation, had given consent, and had not used antibiotics since the screening were randomized to receive either 2 doses of 2 g of oral metronidazole 48 hours apart or a matching placebo. They were scheduled for a follow-up visit at least 2 weeks later and were given 2 more doses of the same active medication or placebo given previously. Randomization was well described, and concealment of allocation was adequate. One significant weakness of this study is the large number of women who screened positive for BV but were excluded from randomization (4604 out of 6540, including 2126 who had difficulty arranging follow-up).

OUTCOMES MEASURED: The primary outcome measures were the length of gestation and the rate of preterm labor.

RESULTS: No significant differences in the length of gestation or in birth weight were observed between the treatment and placebo groups. The number of preterm deliveries - 116 of 953 (12.2%) in the metronidazole group and 121 of 956 (12.5%) in the placebo group-make the likelihood of any clinically significant difference extremely small. Extreme prematurity, low birth weight, and very low birth weight were also equal between the 2 groups. Subgroup analyses for women with significant risk factors for premature delivery, such as previous preterm delivery and prepregnancy weight less than 50 kg (110 pounds), also failed to show any benefit of treatment.

BACKGROUND: Given the high cost of neonatal intensive care and the serious sequelae associated with preterm delivery, an inexpensive safe treatment that would reduce the risk of preterm delivery is appealing. One previous study demonstrated a small decrease in the rate of preterm delivery using oral erythromycin plus metronidazole in a population of women at high risk of preterm delivery (mostly women with a previous preterm delivery)¹; a small study of women at low risk of preterm delivery, however, found no benefit.²

POPULATION STUDIED: Women presenting for prenatal care were considered for randomization. Exclusions were reasonable and included women with: symptoms of vaginitis, contraindications to metronidazole, recent or anticipated antibiotic use, preterm labor, planned or current cervical cerclage, multifetal gestation, failure to have a postscreening visit before 24-weeks’ gestation, and unanticipated use of antibiotics between the screening and the postscreening visit. This was a group of fairly low-risk and adherent patients.

STUDY DESIGN AND VALIDITY: This was a randomized placebo-controlled double-blinded study. Women were considered to have BV if they had a vaginal hydrogen ion concentration higher than 4.4 and a positive Gram’s stain using reasonable criteria well described by the authors. Women with BV who were seen before 24 weeks’ gestation, had given consent, and had not used antibiotics since the screening were randomized to receive either 2 doses of 2 g of oral metronidazole 48 hours apart or a matching placebo. They were scheduled for a follow-up visit at least 2 weeks later and were given 2 more doses of the same active medication or placebo given previously. Randomization was well described, and concealment of allocation was adequate. One significant weakness of this study is the large number of women who screened positive for BV but were excluded from randomization (4604 out of 6540, including 2126 who had difficulty arranging follow-up).

OUTCOMES MEASURED: The primary outcome measures were the length of gestation and the rate of preterm labor.

RESULTS: No significant differences in the length of gestation or in birth weight were observed between the treatment and placebo groups. The number of preterm deliveries - 116 of 953 (12.2%) in the metronidazole group and 121 of 956 (12.5%) in the placebo group-make the likelihood of any clinically significant difference extremely small. Extreme prematurity, low birth weight, and very low birth weight were also equal between the 2 groups. Subgroup analyses for women with significant risk factors for premature delivery, such as previous preterm delivery and prepregnancy weight less than 50 kg (110 pounds), also failed to show any benefit of treatment.

BACKGROUND: Given the high cost of neonatal intensive care and the serious sequelae associated with preterm delivery, an inexpensive safe treatment that would reduce the risk of preterm delivery is appealing. One previous study demonstrated a small decrease in the rate of preterm delivery using oral erythromycin plus metronidazole in a population of women at high risk of preterm delivery (mostly women with a previous preterm delivery)¹; a small study of women at low risk of preterm delivery, however, found no benefit.²

POPULATION STUDIED: Women presenting for prenatal care were considered for randomization. Exclusions were reasonable and included women with: symptoms of vaginitis, contraindications to metronidazole, recent or anticipated antibiotic use, preterm labor, planned or current cervical cerclage, multifetal gestation, failure to have a postscreening visit before 24-weeks’ gestation, and unanticipated use of antibiotics between the screening and the postscreening visit. This was a group of fairly low-risk and adherent patients.

STUDY DESIGN AND VALIDITY: This was a randomized placebo-controlled double-blinded study. Women were considered to have BV if they had a vaginal hydrogen ion concentration higher than 4.4 and a positive Gram’s stain using reasonable criteria well described by the authors. Women with BV who were seen before 24 weeks’ gestation, had given consent, and had not used antibiotics since the screening were randomized to receive either 2 doses of 2 g of oral metronidazole 48 hours apart or a matching placebo. They were scheduled for a follow-up visit at least 2 weeks later and were given 2 more doses of the same active medication or placebo given previously. Randomization was well described, and concealment of allocation was adequate. One significant weakness of this study is the large number of women who screened positive for BV but were excluded from randomization (4604 out of 6540, including 2126 who had difficulty arranging follow-up).

OUTCOMES MEASURED: The primary outcome measures were the length of gestation and the rate of preterm labor.

RESULTS: No significant differences in the length of gestation or in birth weight were observed between the treatment and placebo groups. The number of preterm deliveries - 116 of 953 (12.2%) in the metronidazole group and 121 of 956 (12.5%) in the placebo group-make the likelihood of any clinically significant difference extremely small. Extreme prematurity, low birth weight, and very low birth weight were also equal between the 2 groups. Subgroup analyses for women with significant risk factors for premature delivery, such as previous preterm delivery and prepregnancy weight less than 50 kg (110 pounds), also failed to show any benefit of treatment.