1. Annane D, Bellissant E, Briegel J, et al. Corticosteroids for severe sepsis and septic shock: a systematic review and metaanalysis. BMJ. 2004;329:480.
Severe sepsis is a common cause of ICU admissions and continues to have mortality rates as high as 30%. As research has evolved, the role of corticosteroids in sepsis has changed. Annane and colleagues systematically reviewed corticosteroid treatment for severe sepsis and septic shock. The authors found 15 randomized trials that evaluated their main outcome of all-cause mortality at 28 days. When these results were pooled there was no significant improvement in mortality (RR 0.98; 0.87 to 1.10). However, the analysis showed very significant heterogeneity, so the authors appropriately performed a sensitivity analysis based on an a priori hypothesis that low-dose, long-course steroids (≤ 300 mg hydrocortisone/day and ≥ 5 days) would provide greater benefit than short-course, high-dose corticosteroids. The sensitivity analysis of five studies with low-dose, long-course corticosteroids showed a clear reduction in 28-day mortality (RR 0.80; 0.67 to 0.95) with no heterogeneity. There was no statistically significant difference in GI bleeds, superinfections, and hyperglycemia.
This is a very well-done review, which included an exhaustive search for all available evidence. The authors conclude based on the findings of an individual study in the review that patients with septic shock should undergo adrenal insufficiency testing prior to starting empiric therapy with low-dose hydrocortisone (200-300 mg/day). Hydrocortisone should then be stopped if there is no evidence of adrenal insufficiency. However, none of the studies in this review were limited to patients with adrenal insufficiency. Also, although none of the studies individually showed a statistically significant improvement with corticosteroids, they all favored the steroid treatment group with RRs < 1. Although there is good evidence that septic shock patients with adrenal insufficiency should be treated with corticosteroids, it remains unclear if therapy should be generalized to all septic patients. The ongoing European CORTICUS trial should help answer some of these questions.
2. Baddour L.,Yu, V. and the International Pneumococcal Study Group. Combination antibiotic therapy lowers mortality among severely ill patients with pneumococcal bacteremia. Am J Resp Crit Care Med; 2004;170:440-444.
Three retrospective studies indicate that patients with pneumococcal bacteremia are less likely to die if treated with combination antibiotics. The combination of a beta-lactam and a macrolide appear to be especially protective. Based upon this soft data, multiple societies now recommend combination antibiotic therapy with a beta-lactam and a macrolide as first-line therapy for hospitalized patients with CAP. This large, multi-center, prospective observational study analyzed outcomes for all patients with pneumococcal bacteremia. Critically ill patients were eight times more likely to die than non-critically ill patients (mortality: 54.6% vs. 7.3%, p=0.0001). Combination therapy was defined as any two or more antibiotics used concurrently: multiple different regimens were used. The 14-day mortality difference between pneumococcal bacteremic patients receiving monotherapy vs. those receiving combination therapy was not significant (10.4 vs. 11.5%, p-value not disclosed) However, in critically ill patients, combination antibiotic therapy was associated with a marked decrease in mortality (23.4 vs. 55.3%, p=0.0015).
This study has significant limitations. It is not randomized, combination therapy was broadly defined as any two or more antibiotics, and 16% of critically ill patients received only one antibiotic, a major deviation from the standard of care. Nonetheless, this is the first prospective trial that attempts to ascertain which patients with pneumococcal bacteremia benefit from combination therapy. The marked mortality reduction in critically ill patients who received combination therapy reinforces the current recommendation that septic patients with pneumococcal bacteremia should receive combination therapy. However, these findings also underscore the fact that current guidelines for the treatment of an exceedingly common and serious disease are based upon weak retrospective data. Further randomized prospective trials are needed to determine which patients with pneumococcal pneumonia may actually benefit from combination antimicrobial therapy and what combination is most efficacious.
3. Fernandez-Avilles F, Alsonso J, Castro-Beivas A, et al, on behalf of the GRACIA (Grupo de Analisis de la Cardiopatio Isquemica Aguda) Group. Routine invasive strategy within 24 hours of thrombolysis versus ischemia-guided conservative approach for acute myocardial infarction with ST-segment elevation (GRACIA-1). Lancet;2004;364:045-1053
In the last ACC/AHA guidelines for acute myocardial infarction (1999), angioplasty soon after thrombolysis is strongly discouraged (class III) because studies to date showed no significant difference in outcomes and propensity for adverse events, i.e. bleeding. However, as stated in these last recommendations, the benefits of early post-thrombolysis percutaneous transluminal coronary angioplasty (PTCA) with stenting should be reassessed as interventional strategies mature. In the current era of stents and newer antiplatelet therapies, it looks as if now is the time to take another look.
The GRACIA-1 was a randomized, multi-centered trial to assess early intervention following thrombolysis in ST-elevation myocardial infarction (STEMI). Five hundred patients were enrolled and randomized to receive either fibrin-specific thrombolytics alone in the conservative group or thrombolysis combined with early PTCA. The primary endpoint was a combined rate of death, non-fatal reinfarction, or ischemia-induced revascularization at 1 year. Pre-discharge PTCA and stenting in those demonstrating recurrent ischemia in the conservative group (n=52) was analyzed as a secondary endpoint. One-year follow-up was completed in 98% of patients. Both groups had similar major cardiac events at 30 days, with the one-year endpoint demonstrating no difference in mortality. Significant differences were seen in the primary endpoint at one year [23 (9%) in the invasive group vs. 51 (21%) in the conservative group (RR-0.44, 95% CI 0.28-0.70 p=0.00008)].
A major limitation of this study was the low-risk profile of its patients, and the fact that it was not powered to differentiate between mortality and reinfarction. It will be interesting to see if these results are replicated in future studies (CARESS-2005) and their clinical impact on the management of acute myocardial infarction in centers without interventional support.
4. Hitcho EB, Krauss MJ, Birge S, et al. Characteristics and circumstances of falls in a hospital setting: a prospective analysis. J Gen Intern Med.2004;19:732-739.
Falls in hospital patients are common and persistent problems in hospitals across the country, with significant impact on patient safety, morbidity, and health care cost. Previous studies have reported up to 30% of inpatient falls result in injury, with 4-6% resulting in serious injury. Most studies on falls are retrospective reviews. This paper detailed a prospective observational study designed to identify and analyze the patient characteristics or the circumstances that contribute to falls. Using an online adverse event reporting system, patient falls were identified. Once identified, additional data sources, including the patients’ medical records, electronic nursing records, and interviews with patient or family members and nurses, were used to collect information. A total of 183 patients fell during the study period, with 85% of the falls occurring in the patient’s room.
Half of the falls involved bowel or bladder elimination and were more likely to occur in patients over 65 years in age (adjusted odds ratio 2.4; 95% confidence interval 1.1 to 5.3). Most of the falls (59%) occurred during the evening/overnight. 58% of the patients who fell had received agents with central nervous system activity in the 24 hours prior to a fall. 12% of the patients who fell received sedative-hypnotics. Repeat fallers were more likely to be men than women (11/86 [13%] vs 4/97 [4%]; p=0.03). 42% of the falls resulted in some type of injury, ranging from pain/swelling (34%) to cardiac arrest/death (0.5%). In multivariate analysis, only elimination-related falls remained a significant predictor of being injured from a fall (aOR , 2.4; 95% CI 1.1 to 5.3). Interestingly, patients who were confused or disoriented were less likely to be injured than alert and oriented patients (crude OR, 0.5; 95% CI, 0.3 to 0.98). When examined by services, medicine and neurology had the highest fall rate at 6.12 falls per 1000 patient days and had the highest patient-to-nurse ratio of 6.5 and 5.3, respectively.
While specific independent risk factors for prevention of falls could not be determined in this study due to a lack of a control group, their findings are consistent with previously identified risk factors for falling, including weakness, poor cognitive status including medication-related changes, and altered elimination. Potential interventions to prevent falls include toileting schedules for high-risk elderly patients and review of medication lists to minimize centrally acting, psychotropic, or sedating agents.
5. Kagansky N, Knobler H, Rimon E, Ozer Z, Levy S. Safety of anticoagulation therapy in well-informed older patients. Arch Intern Med; 2004;164:2044-2050
Elderly patients are increasingly likely to require long-term anticoagulation. Despite this, physicians often withhold long-term anticoagulation due to perceived risk factors for bleeding, including age, comorbidity, and cognitive or functional impairment.
This combination retrospective and prospective observational study evaluated the safety and quality of anticoagulation in elderly patients. Over a three-year span, 323 elderly patients were discharged on warfarin from a large Israeli hospital. These patients were frail and potentially “at risk”; 54% were older than age 80, 81% were uneducated, and 84% had low income. 47% were considered cognitively intact (according to MMSE) and only 34% were functionally independent. Using a multivariate analysis, the authors determined that only poor quality of anticoagulation education (OR: 8.83; 95% CI: 2.0-50.2), polypharmacy (OR: 6.14; 95% CI 1.2-42.4) and INR >3 (OR: 1.08; 95% CI 1.03-1.14) were associated with major bleeding.
This study of a “real world” population of frail, at-risk elderly patients provides important insights into risks for major bleeding resulting from warfarin therapy. Surprisingly, advanced age, cognitive impairment and markers for frailty did not confer elevated risk for major bleeding. Importantly, the study did not explicitly address patient fall risk, an often-cited reason for withholding anticoagulation therapy. Supratherapeutic anticoagulation is associated with a statistically but not clinically significant bleeding risk. Polypharmacy is increasingly unavoidable in elderly patients. However, appropriate anticoagulation therapy should be the standard of care. This study makes it clear that we need to focus our efforts on ensuring that elderly patients and their caregivers receive appropriate anticoagulation education prior to discharge from the hospital.
6. Prinssen M, Verhoeven E, Buth J, et al. A Randomized trial comparing conventional and endovascular repair of abdominal aortic aneurysms. NEJM. 2004;351:1607-18.
Endovascular repair of abdominal aortic aneurysm, which involves percutaneous introduction of a graft to bridge an aortic aneurysm, has been available since the early 1990s. This procedure was initially used in patients who were felt to be at high risk for the traditional open procedure. This selection bias may have affected the outcome of earlier studies. Therefore, the Dutch Randomized Endovascular Aneurysm Management (DREAM) trial was designed to assess outcomes in patients deemed fit to undergo either open or endovascular repair. The study was conducted at 24 centers in the Netherlands and four centers in Belgium. Three hundred and forty-five patients had abdominal aortic aneurysms of at least 5 cm and were considered eligible for either open or endovascular repair. Patients were then randomized to undergo one procedure or the other.
Surgeons experienced in both endovascular and open repair performed the procedures. Primary outcomes were 30-day mortality (defined as death during the primary hospital admission or within 30 days), severe complications, and the combination of the two. The operative mortality rate was 4.6% in the open-repair group and 1.2% in the endovascular repair group, with a risk ratio of 3.9 (95% CI, 0.9-32.9). The combined endpoint of mortality and severe complication was 9.8% in the open-repair group and 4.7% in the endovascular repair group, with a risk ratio of 2.1 (95% CI, 0.9 to 5.4). It should be noted that 90% of study patients were men. These findings indicate that endovascular repair of aortic aneurysm results in less short-term morbidity and mortality than open repair. Larger studies with longer follow up are indicated.
7. Spargias K, Alexopoulos E, Kyrzopoulos S. Ascorbic acid prevents contrast-mediated nephropathy in patients with renal dysfunction undergoing coronary angiography or intervention. Circulation.110(18)2837-42
In this study, a team of Greek researchers studied the use of vitamin C in patients with renal insufficiency (creatinine >1.2) undergoing coronary catheterization or intervention to determine if it prevented contrast-mediated nephropathy (CMN). Although the etiology of CMN is unknown, ischemic insults to the renal medulla and free radical release have been implicated. Since vitamin C has been shown in animal studies to attenuate nephropathy when administered with known renal toxins, a similar effect was postulated when administered with IV contrast.
Two hundred and twenty patients electively admitted for coronary catheterization or intervention were randomized to placebo or to receive 3 grams of vitamin C 2 hours prior to the dye load followed by 2 gm that night and the morning after. The two groups were balanced for presence of hypertension, diabetes, tobacco use, and use of commonly prescribed medications. Intravenous hydration was given to all at variable rates depending on left ventricular ejection fraction. The catheterization team chose the type of contrast agent. Creatinine was measured at baseline (up to 3 months prior to dye load) and 2-5 days afterwards. CMN was defined as a rise in total serum creatinine by 0.5 mg/dL or by 25%. Using these criteria, 9% of the vitamin C group developed CMN as opposed to 20% in the controls (NNT 9; 95% CI: 5-53). Logistical regression was performed to account for differences in age, initial serum creatinine, or both, and always yielded significant differences. The treatment group was slightly older, had higher rates of isoosmolar contrast administration (reno-protective), and had slightly higher baseline creatinines, so the true protective effect was thought even higher than reported.
This study is limited by small sample size, an overwhelmingly male patient base (90%), and the suspected presence of atherosclerotic heart disease in all patients. The amount and rate of delivery of contrast was not noted. The vitamin C preparation used was not described, implying that all preparations were bioequivalent. No mention was made of whether these results could be extrapolated to other procedures involving contrast. Given the low cost and high tolerability of vitamin C, however, further studies would certainly be welcome.
8. Yadav J, Wholey M, Kuntz R, et al. Protected carotid-artery stenting versus endarterectomy in high-risk patients. NEJM.2004;351:1493-1501.
Carotid angioplasty has been available for the past decade, but, due to possible complications such as plaque embolization and stent compression, the procedure had been limited to patients unsuitable for surgical endarterectomy. Recently, the introduction of crush-resistant nitinol stents and emboli protection devices has made the procedure safer. This study was designed to test the hypothesis that angioplasty and stenting is not inferior to endarterectomy. Three hundred and thirty-four patients at 29 centers had either a symptomatic carotid lesion greater than 50% or asymptomatic lesion greater than 80% and were deemed to be suitable candidates for either carotid endarterectomy or angioplasty. Patients were then randomized to one of the treatment groups. Surgeons, cardiologists, or interventional radiologists performed angioplasty. The primary endpoint was a composite of death, stroke, or MI within 30 days of the procedure, or death or stroke between 31 days and one year.
Secondary endpoints were the need for revascularization and cranial-nerve palsy at one year. During the first month after procedure, 4.8% of patients in the stent group and 9.8% of patients in the surgical group reached the primary endpoint. In the subsequent year, 12.2% of stent patients and 20.1% of surgical patients reached the primary endpoint (absolute risk reduction of 7.9, CI 0.7-16.4 and NNT 14). Cranial nerve palsy at one year was 0% vs. 4.9% in the stent vs. surgical group. Need for revascularization at one year was 0.6% vs. 4.3% for stent vs. surgical group. Although this study was designed to show the non-inferiority of angioplasty vs. carotid endarterectomy in patients deemed suitable for either procedure, it actually was statistically significant for superiority of stenting. It also implied a lower complication rate in terms of cranial nerve palsy and the need for revascularization at one year made the procedure safer. This study was designed to test the hypothesis that angioplasty and stenting is not inferior to endarterectomy. Three hundred and thirty-four patients at 29 centers had either a symptomatic carotid lesion greater than 50% or asymptomatic lesion greater than 80% and were deemed to be suitable candidates for either carotid endarterectomy or angioplasty. Patients were then randomized to one of the treatment groups. Surgeons, cardiologists, or interventional radiologists performed angioplasty. The primary endpoint was a composite of death, stroke, or MI within 30 days of the procedure, or death or stroke between 31 days and one year.
Secondary endpoints were the need for revascularization and cranial-nerve palsy at one year. During the first month after procedure, 4.8% of patients in the stent group and 9.8% of patients in the surgical group reached the primary endpoint. In the subsequent year, 12.2% of stent patients and 20.1% of surgical patients reached the primary endpoint (absolute risk reduction of 7.9, CI 0.7-16.4 and NNT 14). Cranial nerve palsy at one year was 0% vs. 4.9% in the stent vs. surgical group. Need for revascularization at one year was 0.6% vs. 4.3% for stent vs. surgical group. Although this study was designed to show the non-inferiority of angioplasty vs. carotid endarterectomy in patients deemed suitable for either procedure, it actually was statistically significant for superiority of stenting. It also implied a lower complication rate in terms of cranial nerve palsy and the need for revascularization at one year.