Journal watch: Recent studies of note

Recent studies of note.

Intensive insulin therapy doesn't improve outcomes in septic shock treated with hydrocortisone

Intensive insulin therapy is not warranted in patients with septic shock who have been treated with hydrocortisone, according to a study.

The COIITSS (Corticosteroids and Intensive Insulin Therapy for Septic Shock) study investigators performed a multicenter, 2 × 2 factorial, randomized trial to determine whether intensive insulin treatment helped improve outcomes in patients with septic shock who were treated with hydrocortisone, and whether additional therapy with fludrocortisone would be of benefit.

Patients were randomly assigned to receive continuous IV insulin infusion with hydrocortisone, continuous IV insulin infusion with hydrocortisone plus fludrocortisone, conventional insulin therapy with hydrocortisone, or conventional insulin therapy with hydrocortisone plus fludrocortisone. Patients in the hydrocortisone groups received a 50-mg bolus every six hours for seven days, while patients receiving fludrocortisone received 50-microgram tablets once daily for seven days. The primary outcome measure was in-hospital mortality. The study results were published in the Jan. 27 Journal of the American Medical Association.

Five hundred and nine adults were included in the study, 255 treated with intensive insulin (target glucose level, 80 to 110 mg/dL) and 254 treated with conventional insulin (target glucose level, 150 mg/dL or less). Overall, 117 patients in the intensive insulin group and 109 patients in the conventional insulin group died (45.9% vs. 42.9%; relative risk, 1.07; 95% CI, 0.88 to 1.30; P=0.50). More patients receiving intensive insulin had severe hypoglycemia, defined as a blood glucose less than 40 mg/dL, than patients receiving conventional therapy (16.4% vs. 7.8%; P=0.003). Mortality rates at hospital discharge were similar among the 245 patients who received fludrocortisone and the 264 who did not (42.9% vs. 45.8%; relative risk, 0.94; CI, 0.77 to 1.14; P=0.50).

The authors concluded that in-hospital mortality did not improve with intensive insulin therapy in this study population, and that adding therapy with oral fludrocortisone was of no significant benefit. An accompanying editorial questioned the statistical power of the trial, asserting that the study population was too small to lead to definitive conclusions, and called for larger national and international trials. “Given the huge global burden of conditions such as septic shock ... such international collaboration should and must be achievable,” the editorialist wrote.

Procalcitonin strategy reduces antibiotic exposure for ICU patients

Physicians may be able to lower the number of days ICU patients with bacterial infections are on antibiotics by using an algorithm based on procalcitonin levels, a study found.

In a prospective, parallel-group, open-label trial, adult patients from six hospitals were randomized to procalcitonin (311 patients) or control (319 patients) groups. Researchers were blind to subjects' assignment before randomization, but not after, and expected all patients would stay in the ICU for more than three days. Antibiotics in procalcitonin patients were started or stopped based on predefined cutoff ranges of procalcitonin levels, while antibiotic use in the control group was based on current guidelines. In both groups, however, individual physicians had the final say as to when treatment started or stopped, and could overrule study algorithms. Outcomes were death at 28 days and 60 days, and number of days without antibiotics at day 28. The study was published online Jan. 23 in The Lancet.

Nine patients were excluded from the study analysis. Procalcitonin patients had significantly more days without antibiotics than did the control group (14.3 days vs. 11.6 days; P<0.0001). The same held true for predefined subgroups, including those with community-acquired infection (3.3-day difference), hospital-acquired infection (2.3-day difference), ventilator-associated pneumonia (3.1-day difference) and immune-compromised patients (3.6-day difference). Mortality between the procalcitonin group was non-inferior to the control group at day 28 (21.2% vs. 20.4%; absolute difference, 0.8%; 90% CI, −4.6% to 6.2%) and at day 60 (30% vs. 26.1%; absolute difference, 3.8%; 90% CI, −2.1% to 9.7%). Fifty-three percent of procalcitonin patients weren't given algorithm-guided treatment, either because the algorithm was overruled by an individual physician or because a patient was discharged from the ICU. This group still had significantly more days without antibiotics than the control group, even when all patients whose treatment algorithms were overruled were excluded from analysis (absolute difference, 3.2 days; 95% CI, 1.2 to 5.1 days).

The procalcitonin group didn't have a shorter length of stay (LOS), but LOS can depend on many factors, and doctors may have wanted to monitor patients more carefully who'd been on a shorter course of antibiotics, the authors said. There was also no between-group difference in rates of emerging multidrug-resistant bacteria, but this may be due, in part, to the fact that a three-day reduction in antibiotic use wasn't enough to register a lower resistance-emergence rate, the authors said.

Previous studies using procalcitonin to reduce antibiotic exposure in the ICU were underpowered, while this study's sample was calculated to have enough power to exclude a 10% between-group mortality difference, the authors noted. The current study also comprised patients who were critically ill with a wide range of infections—40% had septic shock, and two-thirds needed mechanical ventilation—suggesting the treatment strategy could apply to most non-surgical ICU patients, including the immunocompromised, the authors concluded.

NSAID does not reduce post-surgical pericardial effusion

Diclofenac neither reduced the size of pericardial effusions nor prevented late cardiac tamponade in patients after cardiac surgery, according to a small study in France.

Clinically insignificant pericardial effusion occurs in 50% to 85% of patients within a few days of cardiac surgery. Cardiac tamponade occurs in about 1% to 2% of patients, and may develop slowly and without clear-cut clinical signs. Most tamponade occurs more than seven days after surgery, by which time patients often have been discharged from the hospital.

While nonsteroidal anti-inflammatory drugs (NSAIDs) are widely prescribed and guidelines and reviews suggest their use, no study has ever shown efficacy for this purpose. Although usually given for only a short time, they can cause serious adverse effects such as myocardial infarction, acute heart failure, acute renal failure and upper gastrointestinal tract bleeding or perforation.

To assess whether diclofenac is effective in reducing post operative pericardial effusion volume, researchers supported by the French Society of Cardiology conducted a multicenter, randomized, double-blind, placebo-controlled study recruiting from five postoperative cardiac rehabilitation centers. They reported results in the Feb. 2 Annals of Internal Medicine.

Of the 5,455 patients screened, 196 patients met inclusion criteria, which included a moderate to large pericardial effusion more than seven days after cardiac surgery. The patients were randomized to 50 mg of diclofenac or placebo twice daily for 14 days. Researchers administered transthoracic echocardiography upon admission and repeated it at 14 days after surgery, or earlier if they suspected tamponade.

The primary end point was mean change from baseline in the grade of pericardial effusion after 14 days of treatment. Secondary end points included frequency of tamponade, at least a 1-grade decrease in effusion, and mean change in millimeters of the width of effusion.

The initial mean pericardial effusion grade was 2.58 (SD, 0.73) for the placebo group and 2.75 (SD, 0.81) for the diclofenac group. The groups showed similar mean decreases from baseline after treatment (−1.08 grades [SD, 1.20] for the placebo group vs. −1.36 [SD, 1.25] for the diclofenac group). The mean difference between groups was insignificant at −0.28 grade (95% CI, −0.63 to 0.06 grade; P=0.105). Eleven cases of late cardiac tamponade occurred in the placebo group and nine in the diclofenac group (P=0.64). These differences persisted after adjustment for grade of pericardial effusion at baseline, treatment site and type of surgery. No patients had recognized gastrointestinal hemorrhage. The mean increase in hemoglobin level was 0.46 mmol/L (SD, 0.62) in the placebo group versus 0.170 mmol/L (SD, 0.63) in the diclofenac group (P=0.002), which is consistent with subclinical gastrointestinal bleeding in patients receiving diclofenac.

Based on the results, the editors recommended that doctors stop prescribing NSAIDs for post-surgical pericardial effusion because they have no or only small benefit. Although the study suggests NSAIDs are not useful for this condition, researchers acknowledged that the study was underpowered to detect small beneficial effects from diclofenac or to evaluate adverse clinical events from the drug.

Higher copayments lead to more hospitalizations

Increasing copayments for outpatient care increases elderly patients' use of hospitals and ups overall health care expenditures, a study found.

The longitudinal research compared expenditures by enrollees in Medicare plans that increased copayments for ambulatory care with those in similar matched control plans that made no changes. About 900,000 patients from 36 plans were included in the study, which ran from 2001 to 2006. In the plans that increased copayments, the average payments nearly doubled: from $7.38 to $14.38 for primary care and from $12.66 to $22.05 for specialty care.

In the year after the cost increases, the increased-cost plans had 19.8 fewer outpatient visits per 100 enrollees than the control (no-increase) plans. However, the plans that increased copayments also had 2.2 additional annual hospital admissions per 100 enrollees, 13.4 more inpatient days per 100, and 0.7% more enrollees who were hospitalized. The effects were magnified among enrollees living in areas of lower education and income and among patients who had hypertension, diabetes or a history of myocardial infarction. The study was published in the Jan. 28 New England Journal of Medicine.

The findings indicate that cost-sharing initiatives for Medicare patients may have an adverse effect on both health and overall spending, study authors concluded. Even using the bounds of the study's 95% CIs, they calculated that additional expenditures for hospital care would definitively outweigh savings from the copayment increase. These results were found despite high-spending enrollees leaving the higher-copayment plans after the change, and increases in the copayment for inpatient care in those plans—two factors that should have pushed inpatient expenditures down.