Journal watch: Recent studies of note

Recent studies about heart failure after noncardiac surgery, in-hospital PCI, and other topics.


Heart failure worsens outcomes after noncardiac surgery

Elderly patients with heart failure do worse after noncardiac surgery than those without this disorder, according to a study.

Previous research on preoperative risk and perioperative management for noncardiac surgery have focused on patients with coronary artery disease (CAD) or those at risk for ischemic events. However, the risk associated specifically with heart failure remains unclear. Researchers at Duke University used Medicare data to compare outcomes of 13 major noncardiac procedures in patients with heart failure, CAD, or neither. The study's main outcome measures were rates of operative mortality and 30-day readmission. The results appeared in the April issue of Anesthesiology.

A total of 159,327 patients underwent procedures between 2000 and 2004, most commonly hip and knee replacements. Overall, 18.4% of the procedures were done in patients with heart failure, 34.4% in patients with CAD and 47.2% in patients with neither condition. Among patients with heart failure and those with CAD, adjusted hazard ratios for death were 1.63 (95% CI, 1.52 to 1.74) and 1.08 (95% CI, 1.01 to 1.16) and adjusted hazard ratios for readmission were 1.51 (95% CI, 1.45 to 1.58) and 1.16 (95% CI, 1.12 to 1.20), respectively, compared with patients without heart failure or CAD. Risk for both outcomes was higher in patients with heart failure than in patients with CAD.

The authors acknowledged that their study had limitations, including a lack of detailed information about disease severity and treatment. However, they concluded that for the noncardiac procedures studied, patients with heart failure are more likely to have higher rates of operative mortality and hospital readmission. Improved guidelines for perioperative care in this population are needed, the authors wrote, while the author of an accompanying editorial called for additional studies focusing on heart failure patients.

PCI can be done safely at hospitals without on-site cardiac surgery

Percutaneous coronary intervention (PCI) can be done safely and successfully in facilities without on-site cardiac surgical back-up, if the programs are organized and highly skilled, a study has found.

Researchers analyzed data on 9,029 patients treated at 61 medical centers with off-site cardiac surgery programs, and 299,132 patients treated at 404 medical centers with on-site programs. The data came from PCI cases reported to the National Cardiovascular Data Registry from Jan. 1, 2004 to March 30, 2006. The study was reported in early April at a meeting of cardiovascular interventionists in Chicago.

About 41% of patients treated at off-site centers were having a heart attack at the time of PCI, compared with 29% of patients treated at on-site centers. The success rate at both types of centers was similar—94% for off-site centers vs. 93% for on-site centers—as were rates of overall complications (6.4% off-site vs. 6.3% on-site), emergency surgery (0.31% off-site vs. 0.37% on-site) and mortality with emergency surgery (13.6% off-site vs. 12.6% on-site). In-hospital mortality also didn't differ between the groups (odds ratio, 1.08 for on-site vs. off-site, P= 0.507).

Performing PCI at medical centers with off-site back-up is controversial, with clinical guidelines advising it be done for patients who are having a heart attack, but not those whose procedures are elective. Some medical centers argue that they need to offer elective PCI to stay profitable and keep staff skill levels high. The study's lead author noted in a release that the off-site centers in his studies were the cream of the crop, and his results “shouldn't be used to encourage the wild expansion of more off-site PCI programs.”

MC-1 ineffective in preventing heart attack after CABG

A recent study concluded that the drug MC-1 was not effective in preventing heart attack following high-risk coronary artery bypass graft (CABG) surgery.

In the study, 3,023 high-risk patients undergoing CABG at 130 centers in Canada, the U.S. and Germany received either MC-1 (250 mg per day) or placebo before bypass surgery and for 30 days afterward. After 30 days, the all-cause mortality was similar among patients in the MC-1 group and the placebo group. The researchers used a definition of MI of CKMB 100 ng/mL, or about 20 times the upper limit of normal, during and after CABG surgery.

The study was posted early online by the Journal of the American Medical Association in conjunction with its April release at the American College of Cardiology (ACC) conference in Chicago.

The findings were unexpected because a smaller earlier phase of the trial resulted in a reduction in incidences of mortality and stroke. The authors noted that the discrepancies between the two trials highlight the danger of interpreting positive but nonstatistically significant findings from relatively small, phase 2 trials.

MC-1, a naturally occurring metabolite of vitamin B6, is one of several agents being tested for prevention of complications associated with ischemia-reperfusion injury during cardiac surgery. This trial indicates either that MC-1 is ineffective for ischemia-reperfusion injury or that there are other, unknown factors causing the adverse outcomes, said a study author in an April 1 ACC news release.

The findings also suggest that heart attack is more common than suspected in CABG patients, said the news release. “Effective therapies to reduce perioperative morbidity and mortality are needed,” concluded the study authors, “but remain elusive.”

Home-based defibrillators did not reduce risk of death after MI

Having an automated external defibrillator (AED) at home did not reduce the risk of death for patients who had myocardial infarctions, a recent study found.

In the study, 7,001 patients who were not eligible for implanted cardioverter-defibrillators were randomly assigned to respond in one of two ways in the event of a sudden cardiac arrest. One group would follow the standard procedure of calling emergency services and performing CPR while the other group would use an AED before taking those steps. Overall, mortality did not differ significantly between the two groups.

Overall, 450 patients died: 228 of 3,506 patients (6.5%) in the control group and 222 of 3,495 patients (6.4%) in the AED group, and only 160 of the deaths were from cardiac arrest. Of these deaths, AEDs were used in 32 patients; 14 of these received an appropriate shock and four survived to hospital discharge. The study was published in the April 1 New England Journal of Medicine.

Researchers noted several factors that might explain the results. The overall mortality rate was low, which may have been due to the effectiveness of modern drug therapy. Also, partners or companions of patients who experienced sudden events often failed to use an available AED once faced with a stressful situation. The AEDs were effective when used, the authors said, but depended on someone witnessing the event and using the device. Overall, they concluded that placing the devices in homes does not appear to be an effective strategy.

In related news, the American Heart Association recently advised people who witness a heart attack to apply hands-only CPR, after dialing 911. Pushing hard and fast in the middle of the chest continuously can help save the life of a heart attack victim even without performing rescue breathing, said an April 3 AHA news release.

Only about a third of people who collapse from cardiac arrest get CPR help from bystanders, the AHA said. The new recommendations, published in the April issue of Circulation, are based on studies showing that hands-only CPR by an untrained bystander can be as effective as CPR with mouth-to-mouth resuscitation.

Septic acute kidney injury worsens outcomes in ICU patients

Septic acute kidney injury (AKI) occurs often in patients in the intensive care unit (ICU) and is associated with worse outcomes, according to a recent study.

Researchers set out to examine rates, severity and outcomes of sepsis, nonseptic AKI and septic AKI in critically ill patients, using retrospective data from the Australian and New Zealand Intensive Care Society Adult Patient Database. Clinical and laboratory data and outcomes were the main outcome measures. The results were published online April 10 by Critical Care.

Data were available for 120,123 patients admitted to 57 ICUs for more than 24 hours from Jan. 1, 2000, to Dec. 31, 2005. Of these patients, 33,375 (27.8%) had a diagnosis related to sepsis, and of this group, 14,039 (42.1%) had septic AKI. Patients with septic AKI had more acute illness, lower blood pressure, worse pulmonary function, higher heart rates, greater acidemia and higher leukocyte counts than patients with nonseptic AKI (P< 0.0001 for all comparisons), as well as more severe AKI. ICU and in-hospital mortality rates were also higher in septic AKI patients than in nonseptic AKI patients (19.8% vs. 13.4% and 29.7% vs. 21.6%, respectively), and length of stay was longer in the septic AKI group.

The authors concluded that septic AKI occurs often in the first 24 hours after ICU admission. Patients with septic AKI are usually sicker than those with nonseptic AKI and are more likely to die and have longer hospital stays. Although their study had limitations, such as its lack of inclusion of sepsis that developed once the first 24 hours had passed, the authors suggested that their results may be clinically relevant for management of septic AKI patients. Future studies, they wrote, should take septic AKI into account when determining subgroups for analysis of potential therapies.

Long-term outcomes similar for stenting versus endarterectomy

Outcomes at three years were similar in patients who received stents for carotid artery disease and those who underwent endarterectomy, according to a recent study.

The SAPPHIRE trial enrolled patients at 29 centers who were classified as being at high risk for surgical complications of endarterectomy and who had at least 50% stenosis of the luminal diameter with symptoms or at least 80% stenosis without symptoms. Patients were randomly assigned to undergo carotid endarterectomy or stenting. Similar outcomes for the two procedures at 30 days and one year have previously been reported from this trial. The secondary end point at three years was defined as death, stroke or myocardial infarction within 30 days postprocedure or ipsilateral stroke from 31 days to three years postprocedure. The results of the study, which was funded by the stent manufacturer, appeared in the April 10 New England Journal of Medicine.

Data were available at three years for 260 patients (77.8% of the cohort), 85.6% of the stenting group and 70.1% of the endarterectomy group. Forty-one patients in the stenting group and 45 patients in the endarterectomy group developed the secondary end point (cumulative incidence, 24.6% vs. 26.9%; absolute difference, -2.3% [95% CI, -11.8% to 7.0%]). Fifteen strokes occurred in each group; of these, 11 and 9, respectively, were ipsilateral.

The authors concluded that no difference in long-term outcomes exists for high-risk patients receiving stents or undergoing endarterectomy for carotid artery disease. However, they cautioned that their results apply only to this subgroup and should not be used to guide treatment in low- or moderate-risk patients.

Shorter hospital stays not always better for patients with PE

Length of stay for pulmonary embolism (PE) varies by hospital, and patients discharged before four days may do worse than those hospitalized slightly longer, according to a recent study.

Swiss and U.S. researchers used the Pennsylvania Health Care Cost Containment Council database to identify patients discharged with a primary diagnosis of PE from January 2000 through November 2002. The study's objectives were to determine whether certain hospital and patient factors were related to time to discharge and whether length of stay was related to 30-day postdischarge mortality. The results appeared in the April 14 Archives of Internal Medicine.

Overall, 15,531 patients discharged from 186 acute care Pennsylvania hospitals were included in the study. Disease severity was classified by using the Pulmonary Embolism Severity Index (PESI). The median length of stay was six days overall and varied among and within hospitals in all groups of patients, regardless of disease severity. Black patients, more severely ill patients and patients who did not have private health insurance were more likely to have longer hospital stays. Patients who were discharged after four or fewer days had significantly higher postdischarge mortality rates than those hospitalized for five to six days and those hospitalized for seven or eight days (odds ratios, 1.55 [95% CI, 1.21 to 2.00] and 1.32 [CI, 1.00 to 1.74]).

The authors speculated that lengths of stay for PE might vary by and within hospitals because of differences in treatment protocols and physician expertise, and that the link between worse outcomes and shorter length of stay could indicate that physicians are inappropriately discharging high-risk patients. They called for increased use of prognostic models such as PESI as well as development of specific discharge criteria to help appropriately select PE patients for discharge.

An accompanying editorial called the study results “concerning” and suggested that shorter lengths of stay could also be detrimental in other high-risk conditions. However, the editorialists pointed out that the study could have been affected by the exclusion of all in-hospital deaths and by the lack of knowledge of cause of death in patients who died soon after early discharge. On the basis of the study results, it is reasonable, they wrote, for physicians to use evidence-based methods to assess illness severity and mortality risk in PE patients and to ensure that high-risk patients are stable, can tolerate anticoagulant therapy, and have a coordinated care plan in place before being discharged. “For any potentially lethal condition, we need evidence-based criteria for discharge similar to those criteria that have been developed for other diseases, such as pneumonia,” the editorialists concluded.

Clinical uncertainty can lead to worse outcomes in dyspnea

Clinical uncertainty in emergency department (ED) diagnoses of dyspnea can lead to worse outcomes, a recent study found.

Although patients often present to the ED with dyspnea, it can be difficult for physicians to determine its cause. Some studies have suggested that the use of biomarker testing can improve diagnostic accuracy for ED patients with acute decompensated heart failure (ADHF). Researchers from Massachusetts General Hospital and Harvard Medical School set out to determine the relationship between clinical uncertainty in ED diagnoses of dyspnea and outcomes, as well as whether diagnostic accuracy improves when amino-terminal pro-B type natriuretic peptide (NT-proBNP) testing is combined with clinical judgment.

The authors performed a retrospective analysis of data from 592 patients who participated in the ProBNP Investigation of Dyspnea in the Emergency Department (PRIDE). Physicians were asked to estimate the likelihood that patients had ADHF. Those who gave an estimate of 20% or lower or 80% or higher were considered to be clinically certain, while those who gave an estimate of 21% to 79% were considered to be clinically uncertain. The study results were published in the April 14 Archives of Internal Medicine.

In 185 patients (31%), physicians were clinically uncertain whether ADHF was present. Of this group, 103 (56%) had ADHF. Length of stay was increased in patients with uncertain compared with certain diagnoses (6.6 days vs. 5.4 days), and clinical uncertainty was found to independently predict death (hazard ratio, 1.88 [95% CI, 1.02 to 2.25]) and death or rehospitalization (hazard ratio, 2.18 [95% CI, 1.71 to 2.49]) by one year. NT-proBNP testing was found to improve clinical accuracy when used alone or in combination with clinical judgment.

The authors determined that clinical uncertainty is associated with higher morbidity and mortality in ED patients with dyspnea, particularly those in whom ADHF is the underlying cause. Worse outcomes could be associated in part with incorrect treatment, delayed diagnoses, and premature hospital discharge, they wrote. Although their study had limitations, including its retrospective design, the authors concluded that combining NT-proBNP testing with clinical judgment has the potential to improve diagnostic accuracy, decrease costs and improve outcomes in this group of patients.

Quick screening test could help identify sleep apnea before surgery

A four-question screening test could help determine whether patients have obstructive sleep apnea (OSA) before surgery, according to a new study.

OSA is more common in patients undergoing surgery than in the general population and can lead to higher rates of perioperative complications and death. Because patients with OSA are often unaware that they have the disorder, a quick, simple screening instrument could help identify those at risk. Canadian researchers set out to develop and validate a concise screening test for OSA in surgical patients. The study results appeared in the May issue of Anesthesiology.

The researchers used the Berlin questionnaire (an existing 10-question screening tool), expert consensus, and review of the literature to develop the STOP questionnaire, which consisted of four yes/no questions:

  • Do you Snore loudly?
  • Do you often feel Tired, fatigued or sleepy during daytime?
  • Has anyone Observed you stop breathing during sleep?
  • Do you have or are you being treated for high blood Pressure?

Patients who answered “yes” to two or more questions were considered to be at high risk for OSA. The questionnaire was validated by comparison with the apnea-hypopnea index as determined on monitored polysomnography, the gold standard for OSA diagnosis. The STOP questionnaire's ability to predict risk improved in patients with high body mass index, male sex, age over 50 years and large neck circumference, all of which are known risk factors for OSA.

Although the authors noted that the STOP questionnaire should be validated in other settings, they consider it a useful tool for identifying surgical patients at high risk for OSA. “We hope that the evidence from our research will lead to the development of an overall care protocol for patients with OSA and the patients at high risk of having OSA,” the study's lead author said in a press release.

Hemoglobin-based blood substitutes linked to adverse outcomes

Hemoglobin-based blood substitutes (HBBSs) are linked to increased risk for death and myocardial infarction (MI), a recent study reported.

Researchers from the NIH performed a meta-analysis to examine the association between HBBSs and death or MI, as well as to determine why HBBS trials continued to be conducted despite concerns about the agents' safety. The authors identified 16 trials that met the inclusion criteria (therapeutic use of an HBBS in patients 19 years of age or older, with death or MI as an outcome) and involved five different HBBS products and 3,711 patients. The study and an accompanying editorial were released early online in late April by the Journal of the American Medical Association; both appeared in the May 21 print issue.

The authors' analysis found that 164 patients died in the HBBS group compared with 123 in the control group (relative risk, 1.30 [95% CI, 1.05 to 1.61]). Risk for MI was also higher in patients who received an HBBS (59 MIs vs. 16 MIs; relative risk, 2.71 [CI, 1.67 to 4.40]). Although the authors acknowledged that their meta-analysis had limitations, including lack of details on some study protocols, they concluded that HBBSs are related to significantly increased risk for death or MI and that this risk is present for all types of products and patients.

“The cumulative mortality analysis. . .indicates that prompt meta-analysis of the HBBS trials by the FDA most likely would have demonstrated significant risks by 2000,” the authors wrote. They noted that although sponsors of clinical trials are required to report results to the FDA regardless of whether the data are published, the FDA does not make these data public unless the product is approved or undergoes review by an advisory committee. The authors recommended that results of all clinical trials in humans should be made available to researchers in a timely manner, in part through increased use of clinical trial registries, and suggested that the FDA's practice of keeping data confidential during product development should be reversed.

The accompanying editorial pointed out that because results of some studies were made public long after enrollment had stopped, ethics boards did not have all of the available information needed to review study proposals, and patients and their proxies were not able to provide truly informed consent for studies involving HBBSs. The editorialists recommended against further phase 3 trials of HBBSs until the existing data are thoroughly and critically analyzed. “Until the mechanisms and potential toxicities of [hemoglobin-based oxygen carrier] products are better understood, patients cannot be placed at unacceptable risk,” they wrote.