Program Written Summary NEJM Journal Watch Volume 25, Issue 11 June 1, 2014 Digital Media $24.99 Audio CD $27.99 The following is a list of articles summarized on this issue of NEJM Journal Watch Audio General Medicine. Journal Watch Audio Program Info Accreditation Info Cultural & Linguistic Competency Resources WARFARIN MIGHT NOT PREVENT STROKE IN DIALYSISPATIENTS WITH AF Dialysis patients with atrial fibrillation have an elevated risk for stroke, but no randomized trials have shown that anticoagulation lowers this risk in these patients. Because dialysis patients were excluded from trials of novel anticoagulants, warfarin is the only anticoagulant available to them. In a population-based, retrospective study in the March18 issue of Circulation ( http:// dx.doi.org /10.1161/circulationaha.113.004777 ), researchers examined the effects of warfarin in 1600 hospitalized dialysis patients in Canada who had atrial fibrillation; half were prescribed warfarin within a month after hospital discharge. After adjusting for various demographic and clinical variables, warfarin was not associated with a lower risk for stroke, but was associated with an excess risk for bleeding. Although warfarin might lower the risk for stroke in nondialyzed patients with atrial fibrillation and chronic kidney disease ( www. jwatch .org /na33861 ), this study and another recent study ( www. jwatch .org /na32486 ) suggest that warfarin might cause more harm than benefit in patients on dialysis. An accompanying editorial notes ( http:// dx.doi.org /10.1161/circulationaha.113.007549 ) that two guidelines groups Kidney Disease: Improving Global Outcomes ( http://kdigo .org ) and the Canadian Cardiovascular Society ( http:// dx.doi.org /10.1016/ j. cjca.2011.03.011 ) no longer recommend warfarin for patients with atrial fibrillation who are on dialysis. HIGH DIETARY FIBER INTAKEIS ASSOCIATED WITH LOWERMORTALITY AFTER MI A high intake of dietary fiber is inversely associated with cardiovascular risk factors, but we do not know whether it lowers the risk for mortality after myocardial infarction. Using data from the Nurses Health Study and Health Professionals Follow-Up Study, researchers examined associations between post-MI dietary fiber intake and mortality. The analysis involved 2000 women and 1800 men who had no known cardiovascular disease, stroke, or cancer at baseline; the patients had survived a first MI during follow-up and had completed pre- and post-MI food frequency questionnaires. Median follow-up was about 9years. Findings appear on the website of the British Medical Journal ( http:// dx.doi.org /10.1136/bm j. g2659 ). Compared with the patients with the lowest post-MI fiber intake, those with the highest intake had a significantly lower risk for all-cause death. Plus, the patients with the greatest change in increased fiber intake from before their MIs to after their MIs had lower risks for cardiovascular-related and all-cause death. Notably, only a higher intake of cereal fiber (not fruit or legume fiber) was strongly associated with lower mortality. A higher intake of dietary fiber, especially cereal fiber, was associated with a lower risk for cardiovascular-related and all-cause death after myocardial infarction. So high fiber intake might be an effective secondary prevention treatment after MI, along with a low-fat diet, regular exercise, and medications like aspirin and statins. GUIDELINE WATCH: VITAMINAND MINERAL SUPPLEMENTS FOR PRIMARY PREVENTION OF CVDISEASE AND CANCER In a statement in the April15 Annals of Internal Medicine ( http:// dx.doi.org /10.7326/M14-0198 ), the United States Preventive Services Task Force recommends against vitamin E supplementation and notes that the evidence is not sufficient to support using multivitamins. Back in 2003, the Task Force issued statements on vitamin supplementation to prevent cardiovascular disease or cancer ( http:// dx.doi.org /10.7326/ 0003-4819-139-1-200307010-00013 ) and concluded that the evidence was not sufficient to recommend for or against vitamins A, C, or E; multivitamins with folic acid; or antioxidant combinations. It recommended against β-carotene supplementation, because evidence showed that it did not prevent cardiovascular disease or cancer and might confer an excess risk for lung cancer. For the 2014 recommendation statement, the Task Force reviewed studies published since 2003. The reviewers did not evaluate doses higher than the tolerable upper intake levels as determined by the Institute of Medicines Foodand Nutrition Board ( http://iom .edu / Activities/Nutrition /SummaryDRIs/~/ media/Files/Activity%20Files /Nutrition/DRIs/ ULs%20for%20Vitamins%20and%20Elements.pdf ) or supplementation in pregnant women or in adults with special nutritional needs. KEY POINTS: Substantial and consistent evidence shows that vitamin E supplementation has no effect on cardiovascular disease, cancer, or mortality. It is not recommended. Supplementation with β-carotene does not lower the risk for cardiovascular disease or cancer, and it confers an excess risk for lung cancer among patients at high risk for the disease (like smokers and people exposed to asbestos). Once again, β-carotene is not recommended. The evidence is not sufficient to determine whether supplementation with multivitamins or single vitamins or minerals (except vitamin E and β-carotene) lowers the risk for cardiovascular disease or cancer. Plus, the balance of the benefits and the harms of these types of supplements has not been established. In the current review, the Task Force evaluated additional supplements, including vitamin D, calcium, selenium, and folic acid. New evidence increases the certainty that vitamin E does not prevent cardiovascular disease or cancer. Vitamins and minerals are essential to overall health, but the evidence just is not there to suggest substantial benefits or harms from supplements in healthy adults who maintain a reasonably balanced diet. At this time, there is no simple alternative to healthy eating. These recommendations are not likely to impact the many billions of dollars spent every year on vitamin and mineral supplements. VITAMIN D BLOOD LEVELS, SUPPLEMENTATION, AND CAUSE-SPECIFIC DEATH Despite considerable research, the health benefits of vitamin D supplementation in the general population are still controversial. In a systematic review and meta-analysis on the website of the British Medical Journal ( http:// dx.doi.org /10.1136/bm j. g1903 ), researchers asked whether blood vitamin D levels and vitamin D supplementation were associated with the risk for death. One analysis involved 70 observational studies (with an average follow-up ranging from 0.3years to 29years) that involved 900,000 patients (with a median age of 63 and a median baseline blood 25-hydroxyvitamin D level of 20.7 ng/ mL). Overall, compared with the patients whose blood levels were in the top third, those whose levels were in the bottom third had significantly greater risks for death from cancer and cardiovascular disease, and from death due to all causes. For each 10 ng/ mL lower increment of 25-hydroxyvitamin D, the risk for all-cause mortality increased by 16%. Another analysis involved 20 randomized, placebo-controlled trials (with 30,000 older patients; average follow-up ranged from 0.4years to 6.8years) with data on the effect of vitamin D supplementation on all-cause mortality; 8 of the trials provided vitamin D 2 (with a dose range between 200 IU/ day and 4500 IU/ day), and the rest of the trials provided vitamin D 3 (with a dose range between 10 IU/ day and 6000 IU/ day). Vitamin D 3 supplementation significantly lowered the risk for mortality, but vitamin D 2 supplementation did not. In this analysis, the observational data showed an inverse association between blood 25-hydroxyvitamin D levels and death, but reverse causality is possible (that is to say, patients who are ill have low vitamin D levels rather than low vitamin D levels actually cause illness). In the randomized trials, vitamin D 3 supplementation modestly lowered the risk for all-cause mortality, but we do not know the optimal dose and duration of vitamin D 3 supplementation. So clinicians should not recommend widespread vitamin D supplementation. DIAGNOSTIC ALGORITHM FOR SUSPECTED UPPER-EXTREMITY DVT The rate of upper-extremity deep vein thrombosis has risen in conjunction with the more frequent use of central venous catheters. We have clear algorithms for diagnosing lower-extremity, but not upper-extremity, deep vein thrombosis. In a multicenter study in the April1 Annals of Internal Medicine ( http:// dx.doi.org /10.7326/M13-2056 ), a diagnostic algorithm was evaluated in 400 patients with suspected upper-extremity deep vein thromboses. The evaluation included calculating a clinical decision score, d -dimer testing, and ultrasound. The clinical score consisted of +1 point each for the presence of a central venous catheter or lead, localized pain, or unilateral edema and −1 point for a plausible alternate diagnosis. Scores of 1 or lower implied that upper-extremity deep vein thrombosis was not likely. Follow-up was 3months. Upper-extremity DVTs were verified in 100 patients. Half of the patients were assigned clinical decision scores of 0 or 1 (signifying that upper-extremity DVT was not likely); of those, half had normal d -dimer tests and did not undergo further testing or treatment none developed symptomatic deep vein thrombosis. The low-scoring patients with abnormal d -dimer tests underwent ultrasound: Ultrasound was negative in 70 of these patients and they did not get treatment; none developed symptomatic DVTs. Upper-extremity deep vein thromboses were diagnosed in 12 of the low-scoring patients. In those with higher scores, ultrasound detected no upper-extremity deep vein thrombosis in 80 patients; those patients underwent d -dimer testing and repeat ultrasound if the d -dimer result was abnormal, with a yield of 3 additional diagnoses of deep vein thrombosis. The rates of upper-extremity DVT were significantly lower in the patients with scores of 0 or 1 than in those with scores higher than 1. This algorithm for upper-extremity deep vein thrombosis (namely a clinical decision score, a d -dimer test, and ultrasound) is similar to strategies for diagnosing lower-extremity deep vein thrombosis. Given that this algorithm is successful and noninvasive, it could become a standard for clinical practice. INITIAL TREATMENT OF PATIENTSWITH ACUTE DVT: ATHOME OR IN THE HOSPITAL? It has been shown that treating patients with deep vein thrombosis at home is as safe and effective as treating them in the hospital ( http:// dx.doi.org /10.1002/ 14651858.cd003076.pub2 ). But does that finding apply to the real-world management of deep vein thrombosis? In the May Journal of Vascular Surgery ( http:// dx.doi.org /10.1016/ j. jvs.2013.11.091 ), researchers used an international (mostly European) prospective registry of patients with venous thromboembolism to address this question. The study involved 14,000 outpatients who were diagnosed with acute lower-extremity deep vein thromboses and initially treated with low-molecular-weight heparin or fondaparinux; about 30% were treated at home and the rest were admitted to the hospital. The patients who were treated at home were, on average, younger and less ill than were those who were hospitalized, but through propensity scorematching analysis, the researchers adjusted for differences and minimized confounding. They found that the 7-day and the 90-day rates of symptomatic pulmonary embolism, major bleeding, and death among those initially treated as outpatients were the same as, or lower than, the rates among those who were inpatients. Past randomized trials and this new large cohort study suggest that the initial management of deep vein thrombosis with subcutaneously injected drugs can be accomplished safely at home as long as the patients home situation and local healthcare system support that kind of treatment. New all-oral treatments for acute deep vein thrombosis ( www. jwatch .org /jw201012210000001 ) have lowered some of the barriers to outpatient treatment. UNRUPTURED BRAIN AVMs: TOTREAT OR NOT TO TREAT? Managing patients with arteriovenous malformations has been controversial. In a population-based observational study in the April23 and 30 issue of JAMA ( http:// dx.doi.org /10.1001/jama.2014.3200 ), researchers compared patients with unruptured brain arteriovenous malformations who underwent intervention (including any combination of surgery, embolization, or other interventions) and those who got conservative treatment. Median follow-up was 7years. Of the 200 patients studied, half underwent intervention and the other half got conservative treatment. The primary endpoint (namely, all-cause death or a handicap lasting for 2 or more years) happened significantly less often in the conservative-treatment group than in the intervention group during the first 4years of follow-up, but was similar after 4years. The secondary endpoint of nonfatal stroke or arteriovenous malformation-related death was also significantly lower in the conservative-treatment group. These findings support initial conservative therapy for patients with unruptured brain arteriovenous malformations. Although this study was not randomized, the results are largely concordant with the recent randomized ARUBA trial ( www. arubastudy .org and www. jwatch .org /na33011 ). Some experts will claim that these malformations are quite heterogeneous, and treatment should be individualized. But on a broad population level, this argument is getting harder to defend. Abnormal UTERINE BLEEDING IS COMMON DURING THE MENOPAUSAL TRANSITION So-called abnormal uterine bleeding is common in women of late reproductive age, but which bleeding patterns should be considered normal, and which are distinctly abnormal? In a study on the website of BJOG: An International Journal of Obstetrics & Gynaecology ( http:// dx.doi.org /10.1111/1471-0528.12768 ), researchers analyzed 10years of menstrual diaries from 1300 black, white, Chinese, and Japanese women living in the United States; at enrollment, they ranged in age from 42 to 52. Menses ranged in duration from 1day to 130days, and 430 women experienced their final menstrual period during the study. The cumulative percentage of women with at least one occurrence of 10 or more days of bleeding was about 90%, and the percentage with at least three occurrences was 80%. Two thirds of the women had at least three episodes of 6 or more days of spotting; one third had at least three episodes of 3 or more days of heavy bleeding. These three patterns were particularly common during the menopausal transition and in women with uterine fibroids. Compared with white women, black women were less likely to report menses of 10 or more days or spotting for 6 or more days, and Japanese women were less likely to report 3 or more days of heavy bleeding. Higher body-mass index was associated with a heavier flow. Although these findings depend on the study participants perceptions of bleeding amounts, the results approximate the kinds of complaints that clinicians hear every day. These findings should help clinicians and patients understand which bleeding patterns are within the range of normal and which are not. Persistent bleeding abnormalities demand diagnostic evaluation, especially because the incidence of endometrial hyperplasia peaks during the perimenopausal years. But these data suggest that an endometrial biopsy might not be necessary for a single episode of abnormal bleeding during this time. CARDIOSELECTIVE β -BLOCKERS GENERALLY ARE SAFE IN ASTHMA PATIENTS β-blockers can cause airway obstruction and even severe exacerbations in patients with asthma. The mechanism of action is presumed to be catecholamine antagonism at the pulmonary β 2 -receptor with resulting unopposed cholinergic tone causing bronchoconstriction. But back in 2002, a Cochrane review concluded that, in patients with mild-to-moderate asthma, cardioselective β-blockers did not confer substantial short-term respiratory effects and should not be withheld from asthma patients with heart disease ( http:// dx.doi.org /10.1002/14651858.CD002992 ). In a meta-analysis in the Aprilissue of Chest ( http:// dx.doi.org /10.1378/chest.13-1235 ), researchers systematically reviewed 30 studies with 550 adult patients to examine the effect of cardioselective and nonselective β-blockers on lung function and symptoms in patients treated for 1day to 7days. In a dose-responsive manner, selective β-blockers caused an average 7% absolute drop in forced expiratory volume in 1 second and attenuated β-agonist response by an average of 10%. Thirteen percent of the patients experienced decreases in FEV 1 of 20% or greater, and 3% of the patients experienced worsened asthma symptoms. On the other hand, nonselective β-blockers caused an average 10% drop in FEV 1 and attenuated β-agonist response by 20%. Eleven percent of the patients experienced absolute decreases in FEV 1 of 20% or greater, and 8% of the patients had worsened asthma symptoms. There were drug-specific differences: Celiprolol (which is not available in the United States) had the least effect; atenolol and metoprolol (which are selective agents) and labetalol and propranolol (which are nonselective agents) had sequentially greater effects. This analysis offers further support for using cardioselective β-blockers in patients with stable asthma who have clear indications for the drug, but nonselective β-blockers should be avoided, if possible. Because even cardioselective agents can reduce FEV 1 and can interfere with drug effectiveness, patients with unstable asthma should not take β-blockers. Clinicians need to be aware that a few patients with asthma will experience worsening asthma symptoms and substantial declines in lung function while taking the first few doses of cardioselective β-blockers. STEROIDS FOR COPD: LESSISPROBABLY MORE It has been suggested that relatively low-dose oral corticosteroids are as good as or better than high-dose parenteral steroids in hospitalized patients with exacerbations of chronic obstructive pulmonary disease ( www. jwatch .org /jw201006240000001 ), but we do not know whether these observational findings can be extrapolated to patients in the intensive care unit. In the May1 American Journal of Respiratory and Critical Care Medicine ( http:// dx.doi.org /10.1164/ rccm.201401-0058OC ), researchers evaluated 17,000 patients with exacerbations of chronic obstructive pulmonary disease who were admitted to ICUs at nearly 500 hospitals in the United States. Three quarters of the patients were older than 60 and 30% used tobacco. Nearly a third of the patients got noninvasive ventilation; 15% were intubated. Almost all of the patients got antibiotics and bronchodilators. Doses of methylprednisolone were categorized as either high (higher than 240mg) or low (240mg or lower), based on the total methylprednisolone dosage administered on the first or second day of hospitalization; two thirds of the patients got high doses. The patients in the two steroid groups were matched by propensity scoring. After adjusting for unbalanced covariates, the groups had similar in-hospital mortality. Compared with high-dose treatment, low-dose treatment was associated with shorter ICU and hospital lengths of stay, lower hospital costs, and shorter durations of mechanical ventilation. The low-dose patients were less likely to need insulin or to develop fungal infections. This study strongly suggests that a moderate dose of steroids is more than adequate to treat intensive care unit patients with severe exacerbations of chronic obstructive pulmonary disease. It may be reasonable to treat these patients with less than 240mg of methylprednisolone (for example, 80mg to 160mg), but we need a randomized trial to determine the optimal dosing and duration of steroids ( http:// dx.doi.org /10.1164/rccm.201403-0568ED ). FOCUSED SCREENING STRATEGYFOR PROSTATE CANCER Although screening men for elevations in serum prostate-specific antigen levels might lower mortality from prostate cancer ( www. jwatch .org /jw200903180000001 ), widespread screening is costly and results in substantial overtreatment. For each prostate cancer death prevented during 11years of follow-up in a recent study, about 1000 men were screened, and 30 prostate cancers were diagnosed ( www. jwatch .org /jw201203140000001 ). Might a focused approach be more effective? To answer that question, researchers in Sweden evaluated how baseline serum PSA levels at the age of 60modified the effects of screening. Their findings appear on the website of the British Medical Journal ( http:// dx.doi.org /10.1136/bm j. g2296 ). The screened group consisted of 1800 men who participated in the screening arm of the Gothenburg randomized prostate cancer screening trial; the unscreened group consisted of 1200 men who participated in the Malmö Preventive Project (in whom PSA levels were measured retrospectively using stored blood samples). The distributions of PSA levels at the age of 60 were similar in the two groups and were less than 2 ng/ mL in 70% of the patients. After 15years of follow-up, the risks for a diagnosis of prostate cancer, metastasis, and death increased with higher baseline PSA levels. For 60-year-old men with PSA levels lower than 2 ng/ mL, screening yielded no benefit. But for those with PSA levels of 2 ng/ mL or higher, screening significantly lowered prostate cancerspecific mortality; for each prostate cancer death prevented during 15years of follow-up, only 23 men would be screened, and 6 prostate cancers would be diagnosed. The results of this study suggest that, rather than widespread annual prostate-specific antigen screening, clinicians should focus on high-risk men (those with PSA levels of 2 ng/ mL or higher at the age of 60). This strategy would lower prostate cancerspecific mortality and result in less testing ( http:// dx.doi.org /10.1136/bm j. g2559 ). Low-risk men (namely, those with PSA levels lower than 2 ng/ mL at the age of 60 which is nearly three quarters of the men in this study) could be exempted from subsequent screening. These results should be confirmed by randomized trials. SCREENING FLEXIBLE SIGMOIDOSCOPY AND COLONOSCOPY ARE ASSOCIATED WITH LESS COLORECTAL CANCERAND LOWER MORTALITY How effective is screening with flexible sigmoidoscopy or colonoscopy in preventing colorectal cancer and colorectal cancerassociated death? To find out, researchers reviewed randomized, controlled trials as well as observational studies from the United States, Canada, and Europe. The findings of their meta-analysis appear on the website of the British Medical Journal ( http:// dx.doi.org /10.1136/bm j. g2467 ). In four randomized trials, the researchers evaluated screening flexible sigmoidoscopy vs . no endoscopy in more than 400,000 patients. In intent-to-screen analyses, sigmoidoscopy significantly lowered colorectal cancerrelated mortality; pooled estimates of risk reduction were 20% for colorectal cancer incidence and 30% for colorectal cancerrelated mortality. For distal colorectal cancer, sigmoidoscopy lowered incidence by 30% and lowered mortality by 50%, but had a minimal effect on proximal colorectal cancer rates. The effects were even stronger in per protocol analyses. In six observational studies in which colonoscopy was compared with no endoscopy, colonoscopy was significantly associated with a lower overall incidence of colorectal cancer and colorectal cancerrelated death. Plus, colonoscopy was associated with lower risks for proximal and distal colorectal cancers and death from these cancers. This meta-analysis convincingly shows that screening flexible sigmoidoscopy lowers distal colorectal cancer incidence and mortality and strongly suggests that screening colonoscopy is protective against developing and dying from proximal and distal colorectal cancers. Of course, adherence to screening with these tests might be challenging: In the four randomized trials of sigmoidoscopy, participation in the intervention ranged from 50% to 90%. GUIDELINE WATCH: SCREENING OLDER ADULTS FOR COGNITIVE IMPAIRMENT In a statement on the website of the Annals of Internal Medicine ( http:// dx.doi.org /10.7326/M14-0496 ), the United States Preventive Services Task Force finds that the evidence is inconclusive to support or reject screening older patients for cognitive impairment. The prevalence of dementia increases from 5% among people in their 70s to 25% among people in their 80s. Mild cognitive impairment does not affect the activities of daily living and is more difficult to quantify on a population basis estimates range anywhere from 3% to 40% in people 65 or older. In the current review, the Task Force updates its recommendation from 2003 ( http://annals .org / article.aspx ?articleid=716466 ) on screening for dementia (but not mild cognitive impairment) in community-dwelling asymptomatic older adults. KEY POINTS: With regard to the risks for mild cognitive impairment or dementia, Age is the strongest predictor for cognitive impairment. Other risk factors include cardiovascular disease, head trauma, learning disabilities, depression, alcohol abuse, physical frailty, low education level, and never being married. Some data suggest that specific lifestyle factors can lower the risk for cognitive impairment. Examples include an adherence to a Mediterranean diet or a diet high in fruits and vegetables, regular exercise, and cognitive engagement. In terms of screening tests, the most commonly used test is the Mini-Mental State Examination, with a sensitivity and a specificity of nearly 90% for detecting cognitive impairment. In terms of interventions, many pharmacologic and nonpharmacologic interventions for mild-to-moderate dementia yielded statistically significant benefits that were of unknown clinical importance. The Task Force concluded that the evidence is not sufficient to make a recommendation one way or the other on screening for mild cognitive impairment or early dementia. Unfortunately, this update shows that little progress has been made since the Task Force last reviewed the topic back in 2003. ANXIOLYTICS AND HYPNOTICS ARE ASSOCIATED WITH EXCESS MORTALITY Using anxiolytic and hypnotic drugs, like benzodiazepines and the so-called Z-drugs (like zolpidem [trade name: Ambien and generics] and zaleplon [trade name: Sonata and generics]), is associated with adverse effects, including daytime fatigue, accidents, and falls ( www. jwatch .org /jw201303280000002 ). To find out whether patients taking anxiolytic or hypnotic drugs are at an elevated risk for death, researchers in the United Kingdom retrospectively studied 35,000 primary care patients 16 or older who were first prescribed anxiolytics, hypnotics, or both between 1998 and 2001 and who were matched by age, sex, and practice site with almost 70,000 patients for whom these drugs were not prescribed. Average follow-up was 7.6years. Findings appear on the website of the British Medical Journal ( http:// dx.doi.org /10.1136/bm j. g1996 ). After adjusting for age and multiple potential confounders, there was an excess risk for mortality among users of benzodiazepines, Z-drugs, and other anxiolytic or hypnotic drugs during the first year after they were prescribed. For all three drug classes, higher drug doses were associated with a higher risk for death in a doseresponse pattern. In this study, using benzodiazepines, Z-drugs, and other anxiolytic and hypnotic drugs was associated with an elevated risk for death; plus, there was a doseresponse effect. Notably, many patients in this study were prescribed drugs from more than one class (for example, 18% were prescribed both benzodiazepines and Z-drugs). Although confounding is possible, the results of this study add to growing evidence that these drugs are associated with substantial harm. GLUCOSAMINE FOR KNEE OA: ANOTHER NEGATIVE STUDY Back in 2010, a meta-analysis of randomized trials showed no clinical benefit from glucosamine in patients with knee or hip osteoarthritis ( www. jwatch .org /jw201010070000002 ). Even so, many patients take glucosamine, and some clinicians encourage its use because a few small studies suggested potential benefit and because there seem to be no adverse effects associated with it. In a study in the Aprilissue of Arthritis and Rheumatology ( http:// dx.doi.org /10.1002/art.38314 ), researchers evaluated the effect of Regenasure glucosamine hydrochloride ( www. cargillfoods .com /na/en/products/ health-promoting-ingredients /bone-joint-health/regenasure-glucosamine/index.jsp ) on joint structure (as evaluated by magnetic resonance imaging) and cartilage degradation (as evaluated by the biomarker urinary C-terminal crosslinking telopeptide of type II collagen). The researchers randomized 200 patients with an average age of 52 and mild-to-moderate knee pain (typical of osteoarthritis) to either 1500 mg/ day of glucosamine or placebo for 6months; about 70% of the patients had radiographic features of osteoarthritis. Compared with placebo, glucosamine did not protect against the progression of cartilage damage on MRI, lower the urinary levels of the collagen biomarker, or improve symptoms. This study adds to the evidence against glucosamine for treating patients with knee osteoarthritis. Although 30% of the patients did not have radiographic evidence of osteoarthritis, the study population was typical of patients in the community who take glucosamine for the pain that they (or their clinicians) presume to be related to knee osteoarthritis. This study was manufacturer supported. LUNG DISEASE IN METHOTREXATE-TREATED PATIENTS WITH RA Pulmonary manifestations (like interstitial lung disease, pleural effusions, and nodules) are common in patients with rheumatoid arthritis; the differential diagnosis includes infection, rheumatoid arthritis itself, and the adverse effects of the drugs that are used to treat rheumatoid arthritis. Drugs that have been implicated in causing pulmonary complications include the nonbiologic agents sulfasalazine, methotrexate, and leflunomide, as well as several biologic agents. Methotrexate has been implicated most often, but we do not know the incidence of methotrexate-related lung disease among patients with rheumatoid arthritis. In a meta-analysis in the Aprilissue of Arthritis and Rheumatology ( http:// dx.doi.org /10.1002/art.38322 ), researchers in Ireland evaluated the risk for pulmonary disease among nearly 9000 patients with rheumatoid arthritis who were treated with methotrexate. The researchers reviewed 20 randomized, controlled trials (with durations ranging from 6months to 2years) of methotrexate vs . other agents or placebo. Mortality from lung disease was no higher among the patients treated with methotrexate than among those in the comparator groups; even so, the methotrexate patients were at a slightly, but significantly, higher risk for all adverse respiratory events and total infectious adverse respiratory events. This study confirms the clinical suspicion that methotrexate is associated with a higher risk for pulmonary complications than are other agents used in treating patients with rheumatoid arthritis. In fact, methotrexate-induced pulmonary complications might be underrepresented in this study: The maximum follow-up was 2years, and events have been reported to happen after that. But rheumatologists and primary care clinicians can take comfort in the fact that the excess risk for pulmonary complications in methotrexate-treated rheumatoid arthritis patients was very small, and no deaths from lung disease were attributed to methotrexate. A limitation of this analysis is that we were not told how respiratory or pulmonary events were defined in these studies. CALORIC INTAKE HAS INCREASED AMONG STATIN USERS BUT NOT AMONG NONUSERS In the United States, from the 1970s through the 1990s, overall caloric intake increased among adults, and the proportion of calories they consumed from fat decreased; both types of intake plateaued during 1999 and 2000. To explore whether these trends in food intake differed between statin users and nonusers, researchers used data from the National Health and Nutrition Examination Survey to compare calorie and fat intake every 2years in about 30,000 adults between 1999 and 2010. Findings appear on the website of JAMA Internal Medicine ( http:// dx.doi.org /10.1001/jamainternmed.2014.1927 ). Statin use more than doubled, from 8% in 1999 to 17% in 2010. Body-mass index increased by 1.3kg/m 2 among the statin users, compared with 0.5kg/m 2 among the nonusers. At the start of the study, the statin users consumed about 200 fewerkcal /day than the nonusers and about 10 g /day less fat. By the end of the study, the statin users were consuming about 50morekcal /day and 3 g /day more fat than the nonusers. These findings suggest that starting a statin might give a patient a false sense of reassurance and weaken his or her dietary resolve. New or more intense approaches to dietary counseling might be needed in statin users to counteract this effect. INCRETIN-BASEDDIABETESDRUGS ARE NOT ASSOCIATEDWITH EXCESSRISK FOR ACUTEPANCREATITIS Several years ago, the Food and Drug Administration alerted clinicians to case reports of acute pancreatitis in patients treated with the incretin-based diabetes drugs exenatide (trade name: Byetta; a glucagon-like peptide-1 receptor agonist www. fda .gov /drugs/drugsafety/postmarketdrugsafetyinformationforpatientsandproviders/ucm124713.htm ) and sitagliptin (trade name: Januvia; a dipeptidyl peptidase 4 inhibitor www. fda .gov /Drugs/DrugSafety/ PostmarketDrugSafetyInformationforPatientsandProviders /DrugSafetyInformationforHeathcareProfessionals/ucm183764.htm ). In two studies on the website of the British Medical Journal , researchers asked whether incretin-based treatment raises the risk for acute pancreatitis in patients with type 2 diabetes. In the first study ( http:// dx.doi.org /10.1136/bm j. g2366 ), a meta-analysis of 50 randomized trials in which glucagon-like peptide-1 receptor agonists, dipeptidyl peptidase 4 inhibitors, or both were evaluated in 30,000 patients with an average age between 50 and 67 who were followed anywhere from 3months to 5years, the researchers found that only 40 patients experienced acute pancreatitis, and that incretin-based treatment was not associated with an elevated risk for pancreatitis. The same researchers also analyzed data from five observational studies (with 300,000 patients): In only one study (which was a case-control study of 2500 patients) was the use of exenatide or sitagliptin significantly associated with an increased risk for developing acute pancreatitis. In the second study ( http:// dx.doi.org /10.1136/bm j. g2780 ), researchers analyzed a population-based cohort of 70,000 patients in the United Kingdom and compared new users of incretin-based drugs with new users of sulfonylureas. (Sulfonylureas were chosen as the comparator because they are used at roughly the same stage of diabetes.) After an average treatment duration of 1.4years, there was, again, no association between incretin-based drugs and acute pancreatitis. These analyses suggest that incretin-based treatments for patients with type 2 diabetes are not associated with an excess risk for acute pancreatitis. All of the randomized trials that were involved in the meta-analysis were industry-funded, and none described criteria for diagnosing pancreatitis. Even so, these results should be reassuring to clinicians who prescribe (and to patients who use) these second-line drugs for type 2 diabetes. UNDERSTANDING PRIMARY NONADHERENCE TO PRESCRIBED MEDICATIONS Primary nonadherence to medications is defined as never filling a prescription; secondary nonadherence is defined as filling a prescription, but not taking the medication as prescribed. In a study in the April1 Annals of Internal Medicine ( http:// dx.doi.org /10.7326/M13-1705 ), researchers in Canada evaluated the incidence and characteristics of primary nonadherence using pharmaceutical data covering more than 15,000 patients who got about 40,000 new prescriptions from 130 primary care doctors from 2006 to 2009. Overall, a third of new prescriptions were not filled within 9months. About half of the prescriptions for headache or ischemic heart disease medications and a third of the prescriptions for depression medications were never filled; in contrast, only about 20% of the prescriptions for urinary tract infections were not filled. Older patients were more adherent than younger ones. Overall, 13% of the prescriptions were for a new drug in the same class as that in a past prescription: Adherence was higher for new medications within the same class as past medications (90%) than for new classes of medication (70%). Not surprisingly, cost played a significant role compared with the patients who were charged the maximum copayment, the patients who got free medications were 60% more likely to fill their prescriptions. Patient characteristics were correlated more closely with nonadherence than were characteristics of the prescribing clinicians. This study is notable for its focus on primary nonadherence. A third of prescriptions were never filled. These findings are concerning and likely apply to other countries, too. As noted by the researchers, identifying potential interventions is important; providing more follow-up care and lowering drug costs could also have a substantial positive effect. But it is vital to acknowledge that some prescribed medications are not necessary and may even be harmful; patients might have good reasons not to fill those prescriptions. VIRTUAL AUTOPSY WITH CTANGIOGRAPHY COMPARES FAVORABLY WITH MEDICAL AUTOPSY IN UNEXPECTED DEATHS For a lot of reasons, medical autopsies are not often performed to investigate unexpected deaths. In a prospective study in the April15 Annals of Internal Medicine ( http:// dx.doi.org /10.7326/M13-2211 ), researchers in Germany compared virtual autopsy ( www. jwatch .org /jw201202090000007 ) using computed tomography angiography with traditional autopsy in 50 hospitalized patients who died unexpectedly or within 2days of events necessitating cardiopulmonary resuscitation. Virtual autopsy consisted of CT without angiography followed by multiphase CT angiography of the arteries, veins, and circulation. Medical autopsy and histology were performed after virtual autopsy. Two internists reviewed the clinical records for diagnoses made before death and compared them with the findings of the virtual and the medical autopsies. Diagnoses without radiologic or morphologic correlates (like electrolyte abnormalities) were excluded from the analysis. Of the more than 300 diagnoses listed in clinical records, 90% were confirmed by virtual autopsy, and 80% were confirmed by medical autopsy. Virtual and medical autopsy together identified 16 new major diagnoses (like aortic dissection, pulmonary embolism, and coronary artery stenosis); 2 were identified by virtual autopsy alone (namely, tension pneumothorax and esophageal intubation), and 1 (myocardial infarction) was identified by medical autopsy alone. In this study, virtual autopsy with CT angiography compared favorably with traditional medical autopsy. But a known limitation of virtual autopsy is the detection of small metastases, so traditional autopsy is still preferred for oncology investigations.
Ratings and Reviews
To review this course, please login.Login