Faculty Peer Review
Change to the established order, in politics as in medicine, is a slow and painstaking process. Across the Middle East and particularly in Syria and Libya, the initial excitement and intensity of popular uprisings has faded as opposition groups dig in to continue their threat to existing regimes. A series of articles in the major medical journals this week present evidence that similarly threatens to overturn our current way of doing things, but, as in politics, change is likely to be slow. Highlights include yet more evidence that beta blockers may be beneficial, not harmful, in patients with chronic obstructive pulmonary disease, as well as a dramatic study suggesting HIV transmission can be cut impressively by early initiation of HAART. And one study in this week’s JAMA highlights just how slow change to clinical practice can be, showing that despite the COURAGE trial and the strong evidence for optimal medical therapy as compared PCI in patients with stable coronary disease, few of our patients are taking the right medications.
Evidence continues to mount against the traditional view that beta blockers should be withheld in patients with underlying COPD. A retrospective cohort study in this week’s British Medical Journal used an NHS database of patients with COPD in Scotland to examine the effects of beta blocker use on all cause mortality, mortality from myocardial infarction and COPD alone and on hospital admissions for COPD exacerbations.  Beta blocker use was associated with a statistically significant 22% reduction in all cause mortality, and the mortality benefit persisted when controlling for known underlying coronary disease. Additionally, the mortality effect and a significant reduction in hospital admissions persisted when beta blockers were added stepwise to existing COPD treatment regimens, such as inhaled corticosteroids, and inhaled corticosteroids combined with long-acting beta agonists. No significant difference in effect was found for cardioselective versus non-cardioselective beta blockers. The proposed mechanism involves up-regulation of beta-2 receptors by chronic beta blockade, which may then improve the efficacy of beta-2 agonists. The benefit on mortality and hospital admission seems all the more impressive as beta blockers are probably a marker for sicker patients. As with other recent studies showing a mortality benefit for beta blockers in the COPD population, it remains unclear whether the observed benefits are derived from the treatment of undiagnosed silent coronary disease versus a physiologic effect on COPD itself. This study also failed to address non-fatal adverse effects in the beta blocker treated group such as bronchospasm.
A forthcoming study from the National Institutes of Health, reported this week in the New York Times, may change clinical practice in the treatment of HIV/AIDS patients.  The study – HPTN 052 – enrolled 1763 sero-disconcordant couples at 13 sites across Asia, Africa and America. The HIV positive partner – who was required to have a CD4 count of 350-550 at enrollment – was randomized to immediate initiation of HAART, or initiation of therapy only when the CD4 count fell below 250. Across five years of follow-up, the study found 27 episodes of HIV transmission (confirmed by testing to be the positive partner’s genotype) in the delayed treatment group as compared only one in the early therapy group. The delayed therapy group was given placebo HAART. These results seem to confirm that inducing a low level of viremia significantly reduces transmission risk. The study’s authors conceded that over 90% of the couples in the study were heterosexual. Given the significantly magnified risk of HIV transmission in a homosexual medium, it remains questionable whether the protective benefit would be as pronounced in this group, which still accounts for the majority of infections in the United States. The information released thus far also did not comment on the incidence of adverse effects from HAART or the potential for development of resistance by initiating treatment earlier, although assessing the latter would, of course, require a much longer period of follow-up.
Residents everywhere rejoice when the elective post-catheterization patient lands on their service overnight, as these patients typically demand little more than checking for distal pulses and ruling out a bruit. But a new study shows that these patients may benefit from closer attention while highlighting the barriers to translating published research into clinical practice. The paper, published in this week’s JAMA, finds that 4 years after the COURAGE trial – which showed that PCI offers no benefit in preventing MI or death in patients with stable CAD as compared optimal medical therapy (OMT) – fewer than half of patients with stable CAD are on such therapy (aspirin, a beta blocker and a statin) before PCI.  Before COURAGE, 43.5% of patients were taking OMT prior to PCI, and after the publication, a paltry 44.7% were optimized. Although this difference was statistically significant, both the overall rate of OMT and the minimal increment after the study’s release are notable. Patients discharged from the hospital after PCI fared somewhat better, with 63.5% receiving OMT on discharge prior to COURAGE, and 66.0% after. Compliance with aspirin therapy was highest. The authors were able to exclude patients with a contraindication to any of the OMT medications. The JAMA study raises important concerns about why the results of such a highly publicized and generally well regarded $33.5 million dollar study are not better reflected in clinical practice. The results also suggest room for improvement in systems and communication amongst healthcare teams.
Statins are typically known for their potential toxic effects on the liver, but an interesting case control study in the American Journal of Gastroenterology suggests a more friendly interaction between the two.  This study used a government health service database to identify 1,227 cases of newly diagnosed liver cancer in Taiwan between 2005 and 2008 and matched them with controls. After controlling for a number of potential confounders, including hepatitis B and C virus infection, cirrhosis and diabetes, the ever-use of a statin was associated with a reduction in the incidence of liver cancer nearly by half, with an odds ratio of 0.53, confidence interval 0.41-0.69. Interestingly, when the population was stratified by cumulative statin dosing, the protective effect of statin therapy for liver cancer lost its statistical significance at higher statin doses. The authors attributed this finding to sample size limitations. Several mechanisms for the protective effect of statins were proposed, including disruption of HCV RNA replication and interference with cellular replication pathways. The study suffered from the many limitations of case control studies including the inability to control for certain confounders and the risk for confounding by indication if patients with liver disease were less likely to be prescribed statins. However, given the favorable side effect profile of these medications and the potential physiologic basis for a protective effect, further study may be warranted.
Any resident will agree that pulmonary embolus (PE) continues to be in fashion as an element in the differential diagnosis for anything from asymptomatic tachycardia to pleuritic chest pain. This is likely due, at least in part, to our enhanced ability to detect PE with computed tomography pulmonary angiography (CTPA). But an Archives of Internal Medicine study now suggests that this improved diagnostic sensitivity may do more harm than good.  Soylemez-Wiener and colleagues examined trends in the incidence, mortality and case fatality in PE in the period before and after CTPA was introduced. They used the National Inpatient Sample and Multiple Cause of Death databases. In the period after CTPA, the incidence of PE rose an impressive 81%, from 62 to 112 per 100,000 patients. But despite the dramatic rise in cases detected, mortality decreased only minimally. In fact, in the 20 years before CTPA, PE mortality decreased by 970 deaths per 100,000 cases, but after CTPA, mortality decreased merely 0.4 deaths per 100,000 (although this decrease was still significant with p = 0.02). The authors suggest that CTPA has allowed us to detect more nonfatal emboli that may not require treatment. The authors also report that in the post-CTPA era, in-hospital presumed complications of anticoagulation, including GI bleeding, increased by 71% (p<0.001) on the whole, which they suggest is mostly due to treatment of the newly found pulmonary emboli. The study serves to reinforce the importance of using clinical decision rules to establish pre-test probability prior to initiating diagnostic testing. The study also suggests that our clinical decision rules should be adapted to identify patients not just with any pulmonary embolus, but rather those emboli likely to have clinical significance.
The ways in which clinicians translate published evidence into clinical practice is a complex and multifactorial process driven by personal bias and experience, institutional culture and a consideration of the costs and benefits for an individual patient. Are we too slow to change our practices? The JAMA article might certainly suggest so. But, as in our very own political system, the concepts of moderation, compromise and a purposefully slow pace of change may in some ways be protective to patients – a notion supported by the Archives of Internal Medicine study on CTPA.
Dr. Ingham is a 1st year resident at NYU Langone Medical Center
Peer reviewed by Barbara Porter, MD, section editor, Clinical Correlations
Image courtesy of Wikimedia Commons
1. Short, Philip et al. Effect of Beta Blockers in treatment of chronic obstructive pulmonary disease: a retrospective cohort study. British Medical Journal. 2011; 324; 2549-2558. http://www.bmj.com/content/342/bmj.d2549.full
2. McNeil, Donald. Early Therapy for HIV Said to Cut Spread. The New York Times. May 13, 2011. http://www.nytimes.com/2011/05/13/health/research/13hiv.html
3. Borden et al. Patterns and Intensity of Medical Therapy in Patients Undergoing Percutaneous Coronary Intervention. Journal of the American Medical Association. 2011; 305(18); 1882-1888. http://jama.ama-assn.org/content/305/18/1882
4. Chiu et al. Statin Use and the Risk of Liver Cancer: A Population-Based Case-Control Study. American Journal of Gastroenterology. 2011; 106; 894-898. http://www.nature.com/ajg/journal/v106/n5/abs/ajg2010475a.html
5. Wiener et al. Time Trends In Pulmonary Embolism in the United States. Archives of Internal Medicine. 2011; 171(9); 831-837. http://archinte.ama-assn.org/cgi/content/short/171/9/831