POEMs

				Using the search form

If you prefer to view this article in PDF form, click here.

Which rosacea treatments are most effective?

Clinical question What treatments are effective for rosacea?

Bottom line Effective treatments for rosacea include topical metronidazole, benzoyl peroxide 5%/erythromycin 3% gel, benzoyl peroxide 5%/clindamycin 1% gel, benzoyl peroxide alone, azelaic acid, and sodium sulfacetamide10%/sulfur 5%. Oral tetracycline was significantly better than placebo by physician assessment, but not by patient assessment. (Level of evidence = 1a)

van Zuuren EJ, Gupta AK, Gover MD, et al. Systematic review of rosacea treatments. J Am Acad Dermatol. 2006;56(1):107-115.

Synopsis These investigators thoroughly searched multiple databases—including MEDLINE, EMBASE, The Cochrane Registry of Clinical Trials, Science Citation Index, and reference lists—and consulted with experts. They also searched unpublished literature through correspondence with authors and pharmaceutical companies. Two reviewers independently performed searches and assessed articles for eligibility. Disagreement was resolved by consensus discussion. From a total of 71 possible clinical trials, the authors included 29 randomized trials meeting appropriate criteria for high quality (8) and intermediate quality (21). Fourteen trials used adequate blinding to treatment allocation, and 17 used intention-to-treat analysis. Only data on outcome measures from trials on topical metronidazole, topical azelaic acid, and oral tetracycline could be pooled. The primary outcome measure, quality of life, was not assessed in any of the studies, and only a few studies assessed the participant’s own opinion regarding rosacea severity. The following medications were significantly superior to placebo: topical metronidazole, benzoyl peroxide 5%/erythromycin 3% gel, benzoyl peroxide 5%/clindamycin 1% gel, benzoyl peroxide alone, azelaic acid, and sodium sulfacetamide10%/sulfur 5%. Oral tetracycline was significantly better than placebo by physician assessment, but not by patient assessment. There was no significant difference in efficacy between topical metronidazole and azelaic acid or between topical metronidazole and oral tetracycline. Rilmenidine and permethrin were not significantly better than placebo.

Clinical factors can predict severe CAP in adults

Clinical question Can clinical data collected at the time of diagnosis of pneumonia predict which patients will develop severe disease?

Bottom line A simple scoring system using clinical data can identify patients with community-acquired pneumonia (CAP) who will develop sepsis or require mechanical ventilation. This scoring system needs prospective validation. (Level of evidence = 1a)

Espana PP, Capelastegui A, Gorordo I, et al. Development and validation of a clinical prediction rule for severe community-acquired pneumonia. Am J Respir Crit Care Med. 2006;174(11):1249-1256.

Synopsis In this study, the authors evaluated 1,776 consecutive patients who had received a diagnosis of CAP in a single emergency department to develop a clinical prediction rule that can identify patients who may develop severe disease (septic shock, severe sepsis, or disease requiring mechanical ventilation). The authors then retrospectively validated the prediction rule on a subset of these same patients, as well as on 1,121 patients from four other hospitals. They excluded patients with an expected terminal event in the subsequent 30 days. The prediction model was more accurate than other rules (American Thoracic Society, British Thoracic Society, and the Pneumonia Severity Index) applied to the same patients. To apply the model, the following points are assigned: 13 points for a pH level of less than 7.30; 11 points for systolic BP less than 90 mm Hg; 9 points for a respiratory rate greater than 30 breaths per minute; 6 points for a PaO₂/FiO₂ ratio of less than 250 mm Hg; and 5 points each for a BUN level greater than 30 mg/dL, altered mental status, patient age at least 80 years, and multilobar/bilateral lung affectation on chest x-ray. Severe pneumonia developed in fewer than 3% of patients whose score was less than 10; in 10% of those whose score was 10 to 19; in 40% of those whose score was 20 to 29; and in more than 50% of those whose score was 30 or higher. This scoring system was developed in a Spanish center where the authors note that the intensive care unit is underused.

Spiral CT detects early lung cancer, but use for screening is premature

Clinical question Can spiral CT screening detect small lung cancers that are potentially curable?

Bottom line The authors point to this study as evidence to support widespread use of spiral CT to screen for lung cancer in high-risk patients. However, there are important problems with this recommendation. The unintended negative consequences of screening are of particular concern with a disease like lung cancer. A patient’s continued tobacco use because he is reassured by a negative result increases his risk of heart disease and other malignancies. Two randomized controlled trials that look at all-cause mortality are underway. Also, while the authors point to the low cost ($200) of a scan, this does not include the cost of follow-up tests such as positron-emission tomography (PET), follow-up CT, and biopsies. Finally, the radiation associated with repeated CT scans (even “low-dose” scans) is not trivial and will cause some cancers even as it detects others early. (Level of evidence = 2b)

International Early Lung Cancer Action Program Investigators; Henschke CI, Yankelevitz DF, Libby DM, et al. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med. 2006;355(17):1763-1771.

Synopsis Screening programs are thought to be most effective for tumors that grow at an intermediate speed. Those that grow too fast are often untreatable by the time they are detected, while those that grow very slowly may never require treatment. Lung cancers are thought to fall into the former category, and studies have shown that sputum cytology and chest x-ray are not effective screening tools for this disease. In this study, 31,567 patients at high risk for lung cancer because of current or previous tobacco use (83%), occupational exposure (5%), or exposure to secondhand smoke (12%) were recruited. The participants underwent baseline spiral CT between 1993 and 2005. The 4,186 participants who were found to have at least one suspicious nodule went through a management algorithm that involved antibiotics for suspected infection, repeat CT in 3 months, or immediate PET followed by biopsy if the lesion remained suspicious. A total of 405 patients received a diagnosis of lung cancer. The remaining patients underwent a repeat screen 7 to 18 months later. Of this group, 1,460 had newly identified noncalcified nodules and 74 had lung cancer. Five patients received an interim diagnosis of lung cancer before completing their second screening examination. Of the 484 patients with a diagnosis of lung cancer, 412 had clinical stage I lung cancer defined as no nodal involvement, no metastases; and if there was more than one adenocarcinoma, they were all 30 mm or less in diameter. The estimated 10-year survival for all patients with screening-detected lung cancer was 80%; 92% for those with stage I disease. Eight patients chose not to be treated, and all died within 5 years after diagnosis. Unfortunately, we are not given any information about these patients, who may have chosen to forego treatment because of other major comorbidities.

D-dimer can identify high-risk group for extended anticoagulation

Clinical question Can the D-dimer level be used to guide decisions regarding the duration of anticoagulation for venous thromboembolism (VTE)?

Bottom line The duration of anticoagulation involves an informed decision, made by the patient and the physician, that balances harm, benefit, cost, and convenience. This study provides additional information to help guide that decision. After the recommended 6 to 12 months of anticoagulation, patients with an abnormal D-dimer result should consider extended and perhaps indefinite anticoagulation, while those with a normal D-dimer level should consider discontinuing anticoagulation. (Level of evidence = 1b)

Palareti G, Cosmi B, Legnani C, et al; PROLONG Investigators. D-dimer testing to determine the duration of anticoagulation therapy. N Engl J Med. 2006;355(17):1780-1789.

Synopsis Although the risk of recurrent VTE gradually declines with time from the index event, the risk of bleeding complications remains relatively constant. Thus, while extended anticoagulation reduces the risk of recurrent VTE, it also exposes patients to a greater risk of bleeding problems. In this study, patients with an initial episode of idiopathic VTE (eg, not pregnant, no recent immobilization or surgery, no cancer, and no antithrombin deficiency or antiphospholipid antibody syndrome) who had completed at least 3 months of oral anticoagulation were identified. The most recent guidelines from the American College of Chest Physicians recommend at least 6 to 12 months of anticoagulation for these patients. Approximately half the study patients had 7 to 12 months of anticoagulation therapy, approximately one third had more than 12 months, and the rest had less than 6 months. The patients were told to stop taking their anticoagulants, and their D-dimer level was checked approximately 30 days later (the test is not accurate while taking vitamin K antagonists such as warfarin). Patients who had a normal D-dimer test result (n = 385) discontinued anticoagulation. If the test result was abnormal, the patient was randomly assigned to either continue anticoagulation with a target international normalized ratio of 2.5 or discontinue anticoagulation. All patients were followed for up to 18 months and were evaluated every 3 to 6 months. Any patient with signs or symptoms that indicated possible VTE was evaluated using standard protocols for deep vein thrombosis (DVT) and/or pulmonary embolism (PE), and final outcomes were adjudicated by a committee blinded to treatment assignment. The primary outcome was the number of adverse events, which included recurrent DVT, PE, or a major bleeding episode. In the intent-to-treat analysis, patients with a normal D-dimer result experienced 4.4 adverse events per 100 person-years. Those with an abnormal result who discontinued anticoagulation had 10.9 adverse events per 100 person-years, while those who continued anticoagulation had only 2.0 adverse events per 100 person-years. The differences between the group of patients with the abnormal D-dimer results who were not anticoagulated and the other two groups were statistically significant, while the difference between patients with normal test results and those with abnormal results who continued anticoagulation was not. There was only one major bleeding episode, which occurred in a patient who had an abnormal result and continued anticoagulation. There were three deaths, one in each group.

Intranasal zolmitriptan is effective for cluster headaches

Clinical question Is intranasal zolmitriptan effective in treating cluster headaches?

Bottom line Intranasal zolmitriptan is more effective than placebo in treating cluster headaches. (Level of evidence = 2b-)

Cittadini E, May A, Straube A, et al. Effectiveness of intranasal zolmitriptan in acute cluster headache: a randomized, placebo-controlled, double-blind crossover study. Arch Neurol. 2006;63(11):1537-1542.

Synopsis Ninety-two adult patients with cluster headaches were randomly assigned to receive intranasal zolmitriptan (5 or 10 mg) or placebo. Each patient treated one moderately severe headache with one study regimen in rotation for a total of three headaches each separated by at least 24 hours. The patients graded the severity of their headaches on a 5-point scale. They could use oxygen for rescue. The main outcome was resolution of the headache to no pain or mild pain. When measured 30 minutes after administration, placebo relieved headaches 23% of the time compared with 42% for 5 mg zolmitriptan (number needed to treat [NNT] = 5; 95% confidence interval, 3-18) and 61% for 10 mg zolmitriptan (NNT = 3; 2-5). These results may be inflated because the authors had complete data for only 75% of the patients. Zolmitriptan is more effective in patients with episodic cluster headaches than in those with chronic cluster headaches. The authors report no serious adverse events in this study.

Diet and exercise help to delay diabetes

Clinical question Do diet and exercise delay the development of diabetes in high-risk patients?

Bottom line Diet and exercise are effective in delaying the diagnosis of diabetes in patients at increased risk. (Level of evidence = 2b)

Lindstrom J, Ilanne-Parikka P, Peltonen M, et al; Finnish Diabetes Prevention Study Group. Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study. Lancet. 2006;368:1673-1679.

Synopsis In the Finnish Diabetes Prevention Study, 522 men and women—age 40 to 65 years and at high risk for developing diabetes—were randomly assigned either to a tailored diet-and-exercise regimen or to usual care. To be eligible, the patients had to have a body mass index greater than 25 along with impaired glucose tolerance. The original study lasted a median of 4 years. The cumulative incidence of diabetes in the intervention group was 11% compared with 23% in the control group (number needed to treat = 8; 95% CI, 6-16). In this report, the researchers provide 3 additional years of observations on the patients who had not developed diabetes by the end of the original study. No specific diet or exercise information was provided to the patients during this follow-up period. After a total of 7 years, 75 patients in the intervention group and 110 patients in the control group developed diabetes. The cumulative incidence rate was 4.3 per 100 person-years in the intervention group and 7.4 in the control group. The authors estimate that 22 patients would have to be treated with diet and exercise to prevent one patient per year from developing diabetes.