In the News

Stress tests for low-risk chest pain, statins in stroke patients, and more.


Stress tests to diagnose low-risk chest pain may be overused

Many patients hospitalized with low-risk chest pain undergo stress tests that may not be necessary, a recent study found.

Researchers performed a retrospective cohort study of 2,107 patients with chest pain who were admitted to observation status at Baystate Medical Center in Massachusetts over a two-year period. Patients with myocardial infarction before stress testing were excluded. Researchers used administrative records to determine patient demographics, stress test type and comorbidities. For those who had an abnormal stress test result, they performed a detailed review of the discharge summary to see if the patient underwent cardiac catheterization, revascularization and/or medication changes.

Researchers also reviewed records of a random sample of 17% of the study patients and recorded chest pain characteristics and—if available—physician documentation of pretest probability of coronary artery disease (CAD). They then estimated CAD probability as low (<10%), intermediate (>10% to <90%) or high (>90%). They considered stress testing appropriate for patients at the intermediate probability level. Results were published in the June 11 Archives of Internal Medicine.

Results showed stress testing was performed in 70% of patients. The most frequently ordered type was exercise nuclear stress (46.2%), then pharmacological nuclear stress (28.4%), then exercise stress (22.9%). Of all 1,474 tests, results were abnormal in 12.5%. Stress test ordering was associated with a patient being younger than age 70 (relative risk [RR], 1.12), a patient having private insurance (RR, 1.19), and lack of housestaff coverage (RR, 1.39). Most patients underwent stress testing regardless of pretest probability. To wit, 68% percent of patients with low pretest probability underwent stress testing, compared to 76% with intermediate probability and 86% with high probability. Pretest probability was a strong predictor of abnormal stress test results, with abnormal results in 5% of low probability patients, 13% of intermediate probability patients, and 25% of high probability patients (P for trend, 0.02).

Formal assessment of low-risk chest pain is rare, the authors concluded, and most patients undergo stress testing before discharge even though most are at very low risk for major cardiovascular events in the short term. Evidence-based guidelines on ordering stress tests for low-risk chest pain are needed, they said. “Until then, patients with chest pain could be routinely risk stratified using existing tools…and stress tests could be reserved for those with at least an intermediate probability of disease,” they wrote, which in this study would have reduced the total number of stress tests by 30%.

Guidelines released on lupus nephritis

The American College of Rheumatology released new guidelines recently on the screening, treatment and management of lupus nephritis, the first to specifically cover this topic.

According to the guidelines, 35% of adults in the U.S. with systemic lupus erythematosus have clinical signs of nephritis at diagnosis, and a total of 50% to 60% are estimated to develop nephritis in the first 10 years of the disease. African-Americans and Hispanics are more likely to develop nephritis than whites, and men are more likely to develop it than women.

To provide expert advice for practicing clinicians managing patients with this condition, the American College of Rheumatology convened a task force panel to review previous guidelines, perform a systematic review of the evidence, grade the strength of the evidence, and create clinical scenarios, which were then discussed and voted on to arrive at the final recommendations.

The task force panel made recommendations in the following categories:

  • renal biopsy and histology,
  • adjunctive treatments,
  • induction of improvement in patients with disease of increasing severity,
  • maintenance of improvement in patients who respond to induction therapy,
  • modification of therapies in patients who do not respond adequately to induction therapy,
  • identification of vascular disease in patients with systemic lupus erythematosus and renal abnormalities,
  • treatment of lupus nephritis in pregnant patients, and
  • monitoring activity of lupus nephritis.

The authors acknowledged that the guidelines are limited because panel members could not agree on definitions of some terms, including remission, flare and response, and also noted that no data are currently available to support specific recommendations on dosing steroids and tapering immunosuppressive drugs. They called for further research in these areas, as well as more studies on how new therapies for lupus can be used in patients with lupus nephritis.

The guidelines were published in the June Arthritis Care & Research.

Statins may boost short-term outcomes for stroke inpatients

Stroke patients who used statins before and during their hospital stay were more likely to return home than those who didn't use the drugs, a recent study found.

Researchers analyzed the medical records of 12,689 ischemic stroke inpatients from 17 hospitals in the Kaiser Permanente Northern California system between 2000 and 2007. They also examined electronic pharmacy records for details on use of statins before and during hospitalization. The primary outcome was discharge to home, discharge to an institution like a nursing home, or death in the hospital. Patients discharged to hospice were excluded from the analysis. Results appeared in the May 22 Neurology.

Patients who used statins before stroke hospitalization were more likely to be discharged to home (54.6% for statin users, 50% for statin nonusers) and less likely to die in the hospital (7.6% for statin users and 8.6% for statin nonusers; test for trend, P<0.001). Those who used statins before and during hospitalization were also more likely to be discharged to home (56.5% for statin users, 47.3% for statin nonusers) and less likely to die in the hospital (5.5% for statin users, 10.6% for statin nonusers; P<0.001). Conversely, patients who had their statins withdrawn in the hospital were less likely to be discharged home (39.1% for statin withdrawal, 54.9% for statin continuation) and more likely to die in the hospital (22.3% for statin withdrawal, 5.3% for statin continuation; P<0.001).

Limitations of the study include its observational design and the inability to determine some variables, such as National Institutes of Health Stroke Scale score and stroke subtype, both of which predict outcome, the authors noted. Still, the current study adds to the body of evidence on using statins in stroke by adding information related to timing. The study's data “argue that ischemic stroke patients should be treated with a statin at the time of stroke hospitalization,” they concluded.

Nighttime intensivist staffing may reduce death in some ICUs

Adding nighttime intensivist staffing to a low-intensity daytime staffing model was associated with lower mortality, a study found.

Researchers performed a retrospective study of 65,752 patients from 49 ICUs in 25 hospitals that used the Acute Physiology and Chronic Health Evaluation (APACHE) system, which collects clinical, physiological and outcome data on ICU patients, from 2009 through 2010. They linked a survey of ICU staffing practices with patient-level outcomes data from adult ICU admissions and used multivariate models to examine the relationship between nighttime intensivist staffing and in-hospital mortality. The researchers adjusted for daytime intensivist staffing, case mix and severity of illness. They also conducted a second, confirmatory analysis in a separate population-based cohort of hospitals.

Nighttime intensivist staffing was associated with a significant reduction in risk-adjusted, in-hospital death (adjusted odds ratio [OR], 0.62; P=0.04), when used in ICUs with low-intensity daytime staffing (n=22). There was no benefit of nighttime intensivist staffing in ICUs with high-intensity daytime staffing. A sensitivity analysis that included coverage by residents in the definition of nighttime intensivist staffing showed reduced in-hospital mortality with nighttime intensivists in both low-intensity ICUs (OR, 0.42; P<0.01) and high-intensity ICUs (OR, 0.47; P<0.01). Similar relationships between day- and nighttime staffing and mortality were seen in the verification cohort. Results were published in the May 31 New England Journal of Medicine.

Compared with non-intensivist clinicians, “nighttime intensivists may direct more timely resuscitation of patients in unstable condition, initiate appropriate medical therapies sooner, and adjust complex therapies more efficiently,” the authors wrote in explaining the results. Night intensivists are also more accessible to nursing staff and others for clarifying care plans, which could reduce errors, they noted. In addition, the results indicate that the presence of any physician, including a resident, in the ICU at night improves outcomes, and adding an intensivist to an ICU that already has a night physician may not be helpful.

“Blanket endorsement of 24-hour intensivist coverage is premature, although such coverage appears to be useful in some clinical settings,” the authors concluded.

Decisions about chest pain in ED shared successfully with patients

Low-risk patients with chest pain successfully used a decision aid to choose between stress testing on an observation unit or outpatient follow-up in a recent study.

The prospective trial involved 204 patients who came to an emergency department with chest pain and were judged to be low risk for acute coronary syndrome (ACS). They were randomized to either usual care or use of a decision aid, which included a 100-person pictograph depicting their pretest probability of ACS and the possible management options. The study was published in the May Circulation: Cardiovascular Quality and Outcomes.

Patients who viewed the decision aid showed greater knowledge about their condition on a survey given immediately after their visit (3.6 questions correct vs. 3.0; mean difference, 0.67; 95% CI, 0.34 to 1.0). They were also more engaged in decision making (26.6 vs. 7.0 on the OPTION score) and less likely to choose admission for observation and stress testing (58% vs. 77%; absolute difference, 19%; 95% CI, 6% to 31%). Researchers also found that the decision aid did not significantly affect patients' trust of their physicians and left them more satisfied with the decision-making process. No patients in the decision or control groups had major cardiac events after discharge.

This is the first trial of sharing decision making with patients on treatment of possible ACS, and it shows that such a practice may be feasible, the study authors said. Further research on sharing other cardiovascular decisions (such as whether to manage coronary disease with medication, percutaneous coronary intervention or bypass surgery) is warranted, they wrote.

An editorial noted that the trial was conducted in an integrated delivery system in which coordinated follow-up was relatively easy to arrange for patients who chose outpatient care. Thus, not all health care facilities would necessarily be able to replicate this system, although the advent of accountable care organizations might facilitate this. The editorialist also noted that the study showed that decision aids can be used in urgent situations, which past research has not addressed.