ACP J Club. 1998 Jul-Aug;129:A12. doi:10.7326/ACPJC-1998-129-1-A12
To prepare the annual update in general internal medicine for the American College of Physicians' Annual Meeting and ACP Journal Club, we met regularly and reviewed recent articles from major journals. We reviewed articles that are relevant to the problems we see in our general medicine practices. We focused on well-done studies that would actually change the way we practice. Some other articles were included because they addressed important clinical questions, although concerns about validity dictated that we would not care for patients differently on the basis of their results. We circulated our list to the 25 other general internists in the Section of General Internal Medicine at the University of Chicago and to the editors of ACP Journal Club to solicit suggestions. The 1997 articles we consider most important to practice today are grouped into 7 categories: HIV infection, hormone replacement therapy (HRT), osteoporosis, cancer, cardiovascular disease, asthma, and tobacco and alcohol use.
Several important papers on anti-retroviral therapy highlighted both promise and caution. Randomized controlled trials using 2 nucleoside analogues plus a protease inhibitor (1, 2) showed remarkable efficacy in increasing T-cell counts, reducing viral load, slowing the rate of progression to AIDS, and improving survival in patients with fairly advanced HIV infection. In 1 study (1), 80% of patients receiving indinavir, zidovudine, and lamivudine achieved a nondetectable viral load for the duration of the study. These trials provide strong support for the use of a 3-drug regimen for patients with significant viral activity. However, basic science research (3, 4) showed that even patients with suppression of viral load below the detection limit for up to 2.5 years have recoverable virus in peripheral blood mononuclear cells, although no evidence of drug resistance existed. Therapy for HIV infection is complicated and expensive but can provide dramatic clinical benefit, at least in study situations. Compliance with these complicated regimens is critically important for their continued benefit.
HIV-1 RNA levels continue to predict outcomes better than CD4+ counts, and the combination can predict progression to AIDS and death (5). Using viral load to monitor response to therapy improves care and is recommended in published treatment guidelines (6, 7). Prophylaxis of tuberculosis in anergic patients with HIV infection was shown to be of no significant benefit (8). Finally, 1 study showed a community-based prevention program to be quite successful by self-report in limiting risky behaviors (9).
Hormone replacement therapy
This year's update includes 2 studies that support the benefits of HRT (10, 11). A 15-year, case-control study of women in the Nurses' Health Study (10) showed that current hormone users had a lower relative risk for death (0.63) than did women who had never used hormones. Benefits of HRT were greatest in patients with coronary risk factors and outweighed the increase in mortality from breast cancer. Similar findings were reported from a cohort of postmenopausal women in the Midwest who were followed prospectively for 8 years (11). Furthermore, women with a first-degree relative who had breast cancer did not have a significant increase in their risk for breast cancer during the study period. Taken together, these studies show that HRT provides mortality benefits, even given the increased risk for breast cancer, and support the broadened use of HRT in postmenopausal women.
Studies published this year provide guidance about preventing osteoporosis in healthy older patients (12), women with existing fractures (13), and patients at increased risk because of underlying diseases (14). A randomized trial of healthy ambulatory men and women older than 65 years of age showed that calcium (500 mg) plus vitamin D (700 IU) prevented bone loss and decreased the incidence of non-vertebral fractures (12). For postmenopausal women with evidence of osteoporosis (low bone-mineral density and evidence of at least 1 vertebral fracture), a randomized trial (13) found that alendronate (5 mg) decreased the frequency of radiographic and clinical fractures compared with placebo for a period of 3 years. The number needed to treat (NNT) to prevent clinically apparent vertebral fractures for 3 years was 37. Women at risk for developing osteoporosis because of corticosteroid therapy may also benefit from use of etidronate therapy (14).
Several studies on screening and early detection of prostate and colon cancer were published during the past year. For prostate-specific antigen (PSA) screening, Catalona and colleagues (15) found a 22% prevalence of prostate cancer in asymptomatic men with PSA levels of 2.6 to 4.0 ng/mL. 83% of the cases were clinically significant by histopathologic criteria. A Swedish study (16) reported that 15-year prostate cancer-related mortality was only 6% for patients with well-differentiated tumors and 17% for patients with moderately differentiated tumors. However, the data are biased by the large proportion of incidental tumors that were found at the time of trans-urethral prostatic resection (47%). The College published several studies (17-19) that reviewed prostate cancer screening. These articles concluded that patient preference is vital to the decision about PSA testing and that patients should thoroughly discuss the issues with their physicians before being screened.
Rectal bleeding is disconcerting to patients, and new data reaffirm its importance. Helfand and colleagues (20) studied patients with rectal bleeding noted on review of systems. 23% of patients had a serious cause of bleeding, including 6.5% who had colon cancer and a third of these cases were proximal. Read and coworkers (21) reported that the incidence of proximal neoplasia was significant in patients with small adenomatous polyps (≤ 5 mm) discovered at the time of screening sigmoidoscopy. 29% of such patients had a proximal neoplasm, which was advanced in 6% of patients. Patients with rectal bleeding or small distal polyps warrant a full evaluation of the colon.
Important advances were published on coronary artery disease (CAD), hypertension, stroke, and venous thromboembolism. Patients having noncardiac procedures randomized to atenolol compared with placebo had a 60% relative risk reduction and a 12% absolute risk reduction in 2-year mortality (22). Treatment was started before surgery and stopped at discharge.
2 important papers compared revascularization to medical therapy in patients with angina. The Asymptomatic Cardiac Ischemic Pilot (ACIP) trial (23) found lower mortality in patients having revascularization than in those receiving medical therapy (4.7% vs 12.2%). 76% of the patients had multivessel disease, and all patients had silent ischemia on ambulatory monitoring. However, the applicability to patients not evaluated with ambulatory monitoring is unclear. The Randomized Intervention Treatment of Angina (RITA-2) (24) trial failed to show a benefit for revascularization in patients with single-vessel disease, confirming the results of the 1992 Angioplasty Compared to Medicine (ACME) trial (25).
The Systolic Hypertension in the Elderly Program (SHEP), which studied diuretic-based, stepped care of isolated systolic hypertension, showed a protective effect in preventing congestive heart failure (26). The relative risk for fatal and nonfatal heart failure was 0.51 in patients receiving diuretics compared with placebo (NNT = 48) after 4.5 years of follow-up. Benefits were particularly pronounced in patients with a history of myocardial infarction. This study adds to current knowledge that diuretics reduce the incidence of stroke and major cardiovascular events of all types in this patient population.
For stroke care, an important randomized trial compared the use of aspirin and heparin in patients presenting with acute ischemic stroke (27) and found that the use of aspirin for 14 days led to fewer recurrent strokes and a reduction in the rate of death or dependency at 6 months' follow-up. Heparin did not show similar benefits. A major meta-analysis (28) reported that lowering cholesterol with HMGcoA reductase inhibitors reduced the overall risk for stroke and total mortality in patients with known CAD; the overall odds ratio for stroke was 0.68 and for all deaths was 0.78. This meta-analysis supports the use of cholesterol-lowering drugs for secondary prevention of CAD and stroke in patients with known coronary disease.
Venous thromboembolism was the topic of 3 important studies. A randomized trial (29) showed that patients with a second episode of deep venous thrombosis (DVT) or pulmonary embolism (PE) should receive anticoagulation therapy indefinitely to prevent further recurrences. In addition, 2 randomized trials (30, 31) showed that low-molecular-weight heparin was as safe and effective as unfractionated heparin in patients with acute PE. Low-molecular-weight heparin has not been shown to be superior to unfractionated heparin, but it seems to be a suitable alternative to standard therapy with unfractionated heparin for DVT and PE.
D-dimer levels may be useful to rule out PE in outpatients. A prospective study (32) of emergency department patients found a sensitivity of 99%, specificity of 41%, negative predictive value of 99%, and negative likelihood ratio of 0.02 for a D-dimer level < 500 µg/L. If these results are confirmed, D-dimer levels < 500 µg/L would virtually rule out the diagnosis of PE.
2 randomized trials provided useful approaches to asthma control. Low-dose theophylline plus low-dose inhaled steroids seemed to be as effective as higher doses of inhaled steroids in preventing the symptoms of moderate asthma (33). In addition, the use of inhaled long-acting β-agonists provided added benefit to inhaled glucocorticoids in preventing asthma symptoms and severe and moderate exacerbations (34).
Counseling patients for substance abuse
3 important studies underscored the role that physicians can play in helping patients reduce alcohol and tobacco use. A randomized controlled trial (35) showed that two 15-minute visits with community-based primary care physicians decreased problem drinking, and a meta-analysis (36) of 12 randomized trials showed that brief advice from a physician improved drinking habits in heavy drinkers of alcohol.
A randomized trial (37) of bupropion compared with placebo for smoking cessation received attention from the smoking public. This study showed that bupropion, 150 mg twice daily, in conjunction with counseling visits, resulted in higher rates of smoking cessation than placebo with counseling visits (NNT = 9).
Our group of 4 general internists was impressed by how important, yet challenging, it is to keep up with the medical literature. We found that biweekly meetings were very helpful in discussing new studies and questioning each other about whether these findings would change our practices. We strongly encourage others to consider regular meetings with their colleagues as a method of keeping up-to-date.
Many therapies and studies that may become important are not covered in this update. For example, studies of treatment for Alzheimer disease, including α-tocopherol (38), selegiline (38), ginkgo extract (39), and donepezil (40), were published in 1997. We do not think that the results of these trials were definitive enough to change our current clinical practice and hope that more convincing evidence will emerge soon. In addition, an important case series (41) suggested an association between fenfluramine-phentermine and unusual cardiac valvular abnormalities and led to recall of the drug.