Journal watch: Recent studies of note

Survival benefit in infection after combination antibiotics, ceftaroline monotherapy for complicated skin and skin structure infections, and more.


Survival benefit seen with combination antibiotics in patients with high death risk

Combination antibiotics improve survival for patients with infections who have a high risk of death, but they may be harmful to low-risk patients, a meta-analysis found.

After searching several databases, researchers found 50 randomized or observational studies of antimicrobial therapy for serious bacterial infections that were potentially associated with sepsis or septic shock. The studies needed sufficient data to calculate an odds ratio (OR) between single versus combination antibiotic therapy. Combination therapy was defined as a combination of two or more antibiotics of different antimicrobial class and mechanism of action. Mortality and infection-related mortality were the primary outcomes, though “clinical failure” was acceptable when these weren't provided. Results were published in the August 2010 Critical Care Medicine.

A pooled OR indicated no overall mortality/clinical response benefit with combination therapy (OR, 0.856; 95% CI, 0.71 to 1.03; P=0.0943), but when datasets were stratified by monotherapy mortality risk, the most severely ill subset saw significant benefit (monotherapy risk of death >25%; OR of death, 0.51; 95% CI, 0.41 to 0.64). Of these datasets that could be stratified by the presence of shock/critical illness, the more severely ill group saw increased efficacy of combination therapy (OR, 0.49; 95% CI, 0.35 to 0.70; P=0.0001). Meanwhile, low-risk patients had a higher risk of death with combination therapy (OR, 1.53; 95% CI, 1.16 to 2.03; P=0.003), and meta-regression showed that the benefit of combination therapy depended only on the risk of death in the monotherapy group.

In contrast with previous meta-analyses, this one showed that combination antibiotics do lower mortality in patients with serious infections, but only in those at highest risk of death (i.e., a projected mortality/clinical failure rate of >25%), the authors said. Further, the analysis of the studies that could be split into septic shock and nonshock strata indicates the benefit may be restricted to septic shock cases. It also seems that combination therapy may increase the risk of death in patients whose mortality risk is 15% or lower. “The presence of septic shock may be a simple prospective method to identify patients who may benefit from combination therapy,” the authors concluded. “Alternately, severity of illness scoring systems could be used to identify patients most likely to benefit.”

Ceftaroline monotherapy is effective in treating complicated skin and skin structure infections

Ceftaroline is as effective as vancomycin plus aztreonam in treating complicated skin and skin structure infection (cSSSI), an analysis found.

Ceftaroline was evaluated as treatment for cSSSI in two identical phase 3 clinical trials, of which researchers did a pooled analysis. The combined randomized, controlled trials comprised 693 patients who received ceftaroline (600 mg every 12 hours) and 685 who received vancomycin plus aztreonam (1 g every 12 hours) for 5 to 14 days. Patients came from 111 study centers in Europe, Latin America and the U.S.; baseline characteristics of the treatment groups were comparable. Therapy was deemed complete based on investigator determination that all signs and symptoms of infection had resolved or improved so that no more antimicrobial therapy was needed. Each of the two trials met the primary objective of noninferiority in the clinical cure rate. The analysis was published in the Sept. 15, 2010 Clinical Infectious Diseases.

Clinical cure rates were similar for ceftaroline and vancomycin plus aztreonam in the clinically evaluable population (91.6% vs. 92.7%) and the modified intent-to-treat population (85.9% vs. 85.5%) population, as well as in patients infected with methicillin-resistant Staphylococcus aureus (93.4% vs. 94.3%). Clinical cure rates were similar between patients infected with single or multiple pathogens, across infection types, and between patients with common comorbidities such as diabetes mellitus and peripheral vascular disease. Rates of adverse events, serious adverse events, discontinuations because of an adverse event, and death were also similar between the treatment groups. There was a trend toward a higher clinical cure rate for patients with bacteremia who were treated with vancomycin plus aztreonam, but the significance of this is unknown due to the small number of cases, the authors said. Also, the efficacy of aztreonam was better than ceftaroline against Pseudomonas aeruginosa and Proteus mirabilis.

Limitations of the study include small numbers of black, Asian and elderly patients, and the exclusion of patients younger than 18 years old, the authors acknowledged. They concluded nonetheless that ceftaroline is well-tolerated and effective, and is a potential monotherapy alternative for treating cSSSI.

Lower blood pressure may slow hematoma growth after acute intracerebral hemorrhage

Achieving low blood pressure (BP) targets may slow growth of hematomas after acute intracerebral hemorrhage (ICH), a study reported.

The pilot phase of the Intensive Blood Pressure Reduction in Acute Cerebral Haemorrhage Trial (INTERACT) previously found that hematoma growth after acute stroke was lessened with rapid BP lowering. The investigators performed further analyses to try to determine what level of systolic BP was associated with the greatest decrease in hematoma growth.

Eligible patients had ICH confirmed at baseline by computed tomography, had a systolic BP of 150 to 220 mm Hg, and were able to begin blood pressure-lowering treatment within six hours of stroke onset. Those randomly assigned to the intensive treatment group had a systolic goal of 140 mm Hg by one hour after randomization that was to be maintained for seven days. Patients in the usual treatment group were treated according to the American Heart Association's 1999 guidelines, which recommended a systolic target of 180 mm Hg or less. A second CT scan was done 24 hours after presentation. The study results were published in the Nov. 1, 2010 Hypertension.

Four hundred four patients were recruited for INTERACT. Of these, 346 (85%) had CT scans at both baseline and 24 hours. Baseline systolic BP was not associated with hematoma growth at 24 hours. However, absolute and proportional hematoma growth had a strong, continuous association with on-treatment systolic BP (P=0.03 for trend for both). Patients with the biggest reduction in hematoma growth also had the lowest on-treatment systolic BP (median, 135 mm Hg). No association was seen between hematoma growth and diastolic BP.

The authors noted that their findings might not apply to patients with severe ICH and that INTERACT had limited power to define the effects of on-treatment systolic BP. Although lower systolic BP in the target range of 130 to 140 mm Hg seems to be beneficial after stroke, they wrote, its risks, such as induced ischemia, should also be considered. Also, it is unclear if there are clinical benefits from modest reduction in hematoma growth. The authors did not recommend changing existing guidelines yet for BP goals after stroke but called for more research in this area.

An accompanying editorial also addressed the study's limitations and reiterated the need for additional data. However, they wrote, INTERACT, when considered with other recent findings, “suggests that our current guidelines are too permissive, perhaps even harmful, and will have to be revised soon, recommending judicious BP reduction to lower targets in ischemic stroke and a more aggressive antihypertensive approach to acute ICH.”

Experts deem IVC filter placement appropriate in half of VTE patients

Inferior vena cava filters were appropriately placed about half the time in acute venous thromboembolism patients, a community-based study has found.

Researchers conducted a retrospective chart review of the medical records of patients from three tertiary care hospitals in Worcester, Mass., who were diagnosed with acute venous thromboembolism (VTE) and had an inferior vena cava (IVC) filter placed. Patients in whom an IVC was placed before the index VTE event were excluded.

Follow-up records were reviewed between one and three years from the index event, to ascertain the outcomes of recurrent VTE and bleeding complications. State and national mortality records were also reviewed annually to assess patients' vital statuses, 99% of which were obtained. Three experts in the field of thromboembolism independently reviewed the data regarding the rationale for IVC filter placement for each patient in whom one was placed, and each case was categorized as meeting or not meeting 2001 American College of Chest Physicians' guidelines. Results were published online in the September 13, 2010 Archives of Internal Medicine.

Of the 1,547 patients, 66% had deep vein thrombosis (DVT), 19% had pulmonary embolism (PE), and 15% had both. Of the whole group, 203 patients (13.1%) had an IVC filter placed during a hospital stay after diagnosis of acute VTE. The median duration of follow-up was 926 days. Patients who got an IVC filter were more likely to be male, older, and diagnosed with VTE after hospital admission for a different condition; to have had another hospitalization within the last three months; to require ICU care before the VTE diagnosis; and to have more comorbidities. They were also more likely to have a platelet count lower than 100 × 103/µL at the time of VTE diagnosis, to have been diagnosed as having recurrent VTE, to have had a recent episode of bleeding, and to have been diagnosed with both DVT and PE. Patients who got an IVC filter were more likely to die in the hospital (9.9% vs 5.1%; P=0.01); all-cause mortality was also higher in this group (P<0.001).

The expert panel unanimously agreed that, in 51% of cases where an IVC filter was placed, it was appropriate. It was inappropriate in 26% of cases, they agreed. For the remaining 23%, two reviewers thought a filter was indicated, but a third disagreed. Being at increased risk for anticoagulant-related bleeding was the listed indication for placement in 35% of cases, and the experts agreed with filter placement in only one out of three of these cases, the authors noted. “This finding highlights the fact that guidelines that use implicit phrases such as ‘at risk for bleeding’ or ‘have a contraindication to anticoagulation’ are subjective….,” the authors noted. More research is needed to better define the indications for, and effectiveness of, IVC filter placement, they concluded.

C-reactive protein alone is sufficient to diagnose inflammatory disorders

C-reactive protein (CRP) appears superior to erythrocyte sedimentation rate (ESR) for diagnosing patients with suspected inflammatory disorders, a study found.

Researchers conducted a retrospective analysis of 5,777 patients at an academic hospital in Paris who had had CRP and ESR measured concomitantly. CRP was measured by the system of Beckman Coulter, LX20, with a normal value defined as 10 mg/L or less. ESR was measured by the Diesse Ves-matic automated system, and was considered normal when lower than or equal to the following thresholds for different age groups: 13 mm/h for men and 19 mm/h for women younger than 50 years; 18 mm/h for men and 23 mm/h for women age 50 to 59 years; 22 mm/h for men and 25 mm/h for women age 60 to 64 years; 25 mm/h for men and 30 mm/h for women age 65 to 69 years; and 30 mm/h for men and women age 70 years or older. ESR and CRP were in agreement when both values were normal or elevated, or in disagreement when either value was normal while the other was out of reference range. Results were published in the September 2010 American Journal of Medicine.

After one concomitant measurement per patient was randomly selected, researchers found ESR to be elevated in 58%, compared to 35% for CRP. The two measurements agreed in 67% of patients (both normal in 37%, both elevated in 30%). A disagreement was observed in 33%, with an elevated ESR and normal CRP in 28%, and a normal ESR and elevated CRP in 5%. The kappa coefficient was 0.38, showing poor agreement between the measurements; agreement was lower in patients with anemia and renal failure. On randomly sampled chart review, ESR was more often inaccurate (92% of the time vs. 8% for CRP). Specifically, 25 patients with elevated CRP and normal ESR had active inflammatory disease, thus a false-negative ESR. In 74 patients with elevated ESR and normal CRP, 32% had resolving inflammatory disorders, 28% disclosed a variable interfering with ESR measurement (false-positive ESR), 32% had unexplained discrepancies, and 8% had active inflammatory disease, thus a false-negative CRP.

Interpretations of ESR are often spurious, the authors concluded, since common associated conditions often influence its measure, including elevated serum gamma globulins and monoclonal gammopathy. Thus, CRP measurement should be used instead. “Concomitant measures of ESR and CRP should be avoided; assessment of discordant results may incur substantial costs and could even place patients at risk for invasive procedures,” they wrote.

Thigh-length stockings beat below-knee version as DVT prophylaxis after stroke

Thigh-length stockings are better than below-knee stockings for preventing proximal deep vein thrombosis in hospitalized, immobile patients after stroke, a study found.

Researchers randomized about half of 3,114 acute stroke patients to wear thigh-length stockings and half to wear below-knee stockings while in the hospital. Patients came from 112 hospitals in nine countries. Ultrasonographers, who were blinded, performed compression duplex ultrasonography in 1,406 patients (96% of survivors) in each treatment group between seven and 10 days after enrollment. A second scan between 25 and 30 days was performed in 643 patients in the thigh-length stockings group, and 639 patients in the below-knee stockings group. The main outcome was symptomatic or asymptomatic deep vein thrombosis (DVT) in the popliteal or femoral veins as detected on either scan. Results were published in the Nov. 2, 2010 Annals of Internal Medicine.

In the thigh-length stockings group, 6.3% of patients had a DVT, compared to 8.8% in the below-knee stockings group, an absolute difference of 2.5 (95% CI, 0.7 to 4.4 percentage points; P=0.007). The differences in DVT rate were largely due to a reduction in proximal, not distal, DVT. Treatment groups didn't differ significantly in number of pulmonary emboli, deaths or calf DVT, or in adherence. Sixty-one patients (3.9%) with thigh-length stockings had skin breaks compared to 45 (2.9%) with below-knee stockings (odds ratio [OR], 1.38; 95% CI, 0.93 to 2.04). Seventy-five percent of patients in both groups wore the stockings for 30 days or until discharge, death, or regained mobility. Thigh-length stockings were more likely to be removed due to concerns about skin or patients' comfort.

Study limitations include imperfect sonographer blinding, which could have biased detection; lack of screening logs; and no central verification of negative results or 100% source data verification, the authors noted. The results indicate that thigh-length stockings are “probably more effective” than below-knee stockings, they added, and should be the preferred option, “at least until robust evidence indicates that below-knee stockings are both equally effective and more effective than no stockings.”

Hospice deaths easier on cancer patients, caregivers than ICU deaths

Cancer patients who die in an intensive care unit have more distress and worse quality of life at the end of life than those who die at home, according to a prospective study.

The study included 342 advanced cancer patients and their caregivers. The patients' quality of life at the end of life was assessed by caregiver report within two weeks after the patients' death. Almost 59% of the patients in the study died at home with hospice, while 7.5% died at home without hospice, 25.5% died in a hospital and 8.4% died in the ICU. The study was published online Sept. 13, 2010 by the Journal of Clinical Oncology.

Mean quality-of-life scores were lowest among patients who died in the ICU (5.0) or hospital (5.3) and highest among patients who died at home with hospice (6.6) or without hospice (7.3). The ICU patients also had the lowest scores on physical comfort and psychological well-being (all P≤0.03). The finding that patients at home without hospice had the best quality-of-life scores was surprising, the study authors noted. It may be because these patients were particularly well-adjusted and did not need hospice services, or that they received services that weren't reported.

The study also looked at the impact of place of death on caregivers, comparing their mental health at baseline to six months after the patient's death. Caregivers of the ICU patients were more likely to develop post-traumatic stress disorder than those whose family members died in hospice at home (21% [4 of 19] vs. 4.4% [6 of 137]). The caregivers were also at higher risk for prolonged grief disorder if their patients died at the hospital instead of in home hospice (21.6% [8 of 37] vs. 5.2% [4 of 77]). The findings indicate that caregivers' ICU experiences may be traumatic, the study authors concluded.

The authors cautioned that the study is limited by its observational design. Confounding factors, including the patients' clinical status near death, may have determined the place of death. Future research should look at this issue, and the question of why home deaths result in better quality of life. The subject is increasingly important as patients with advanced cancer receive increasingly aggressive care at the end of life, the authors said. Interventions to reduce terminal hospitalizations and increase hospice use may increase quality of life for patients and decrease psychiatric illness in caregivers, they concluded.

Neuromuscular blocker helps ARDS patients

Early administration of a neuromuscular blocking agent improved outcomes for patients receiving mechanical ventilation to treat acute respiratory distress syndrome (ARDS), a study found.

The multicenter double-blind trial included 340 intensive care unit patients who had presented in the past 48 hours with severe ARDS. The patients were randomized to either placebo or 48 hours of treatment with the neuromuscular blocker cisatracurium besylate. The results were published in the Sept. 16, 2010 New England Journal of Medicine.

The crude 90-day mortality of the groups was not significantly different. However, after adjustment (for the baseline ratio of partial pressure arterial oxygen to the fraction of inspired oxygen [PaO2:FIO2], plateau pressure, and the Simplified Acute Physiology II score), patients in the cisatracurium group had a reduced risk of 90-day mortality (hazard ratio 0.68; 95% CI, 0.48 to 0.98, P=0.04). Further analysis found that the mortality benefit was confined to patients with a PaO2:FIO2 of less than 120.

The cisatracurium group also had significantly more ventilator-free days and days spent outside the ICU. The rate of ICU-acquired paresis did not differ significantly between the groups, leading study authors to conclude that early administration of the drug improved 90-day survival and ventilator use without increasing muscle weakness.

The researchers acknowledged that the study was limited by its size, its inclusion of only severe ARDS patients, and use of only this specific agent. The mechanism by which the drug worked is still uncertain, noted the authors of both the study and an accompanying editorial. The results are interesting because they run counter to the developing critical care paradigm of less intervention (e.g., fewer transfusions, less bed rest, less sedation) being better for patients, the editorialist said.

However, further research replicating and expanding these findings is needed before the use of neuromuscular blockers early in the course of severe ARDS can be adopted in clinical practice, the researchers cautioned. Remaining questions include the optimal duration of therapy and whether other drugs in the class would have the same effect, according to the editorial.