Syncope, latent TB, and perioperative aspirin

Summaries from ACP Hospitalist Weekly.


Many patients seen in ED for syncope diagnosed with other condition after admission

ED visits for syncope are associated with significant costs, whether or not patients are admitted, a Canadian study found.

The study included 51,831 consecutive patients who presented to an ED in Alberta, Canada, from 2006 to 2014 with a primary diagnosis of syncope. After being seen in the ED, 6.6% of them were hospitalized and discharged with a diagnosis of syncope, 8.7% of them were hospitalized and discharged with a primary diagnosis other than syncope, and 84.7% were discharged from the ED with a syncope diagnosis. The study was published in the November 2017 JACC: Clinical Electrophysiology.

Within 30 days, 2.4% of admitted syncope patients, 1.2% of admitted non-syncope patients, and 1.7% of ED-discharged patients had returned to the ED with syncope. At one year, those rates were 8.6%, 5.2%, and 4.8%, respectively. Thirty-day readmission rates were under 1% in all groups. Thirty-day mortality rates were highest for admitted non-syncope patients: 5.2% compared to 1.2% in admitted syncope patients and 0.4% in ED-discharged patients. Mortality rates at one year followed the same pattern, at 17.7%, 9.2%, and 3.0%, respectively. Per-patient costs (in 2016 Canadian dollars) were $29,940 in admitted syncope patients, $42,558 in admitted non-syncope patients, and $10,226 in discharged patients.

The study authors noted that although per-patient costs were lowest in the group released from the ED, total costs were higher for them than either of the other groups because they made up the majority of the patients. The ED-discharged patients were younger and had a significantly lower rate of comorbidities and mortality, suggesting that appropriate triage was occurring in the ED, the authors said.

Although the study found lower rates of hospitalization than previous studies of syncope care in the U.S. and Europe, further reductions are needed, according to the study authors. “A promising strategy to reduce hospitalizations for any country may be ED-based syncope units which consist of time-limited observation with telemetry, a multidisciplinary team and expedited access to cardiac testing,” they wrote. Recent guidelines that provide algorithms for triage, diagnosis, and diagnostic testing may also help, as well as improved discharge planning and patient education, the authors said.

The study was limited by its reliance on administrative codes, and it is uncertain if the findings are generalizable to other provinces and countries, the authors noted.

Self-administered treatment appears effective for latent TB

Self-administered treatment with once-weekly isoniazid and rifapentine could be an effective strategy for U.S. patients with latent tuberculosis, a study found.

Researchers at 12 outpatient tuberculosis clinics performed an open-label, phase 4 randomized clinical trial to compare rates of treatment completion and safety with once-weekly isoniazid and rifapentine administered by patients themselves with and without reminders versus direct observation. The trial, which was funded by the CDC and conducted by its Tuberculosis Trials Consortium, was designed as a noninferiority study with a 15% margin (a predetermined maximum difference between treatment approaches acceptable in considering them to be noninferior). The study drugs were provided by Sanofi. Nine of the study sites were in the U.S. and one each were in Spain, Hong Kong, and South Africa.

Patients 18 years of age and older in whom latent tuberculosis treatment had been recommended were randomly assigned to receive once-weekly isoniazid and rifapentine by direct observation, through self-administration with monthly monitoring, or through self-administration with weekly reminders via text message plus monthly monitoring. Treatment completion, defined as at least 11 doses within 16 weeks, was the primary outcome. Clinical documentation and pill counts were used to measure those in the direct observation group, while self-reports, pill counts, and medication event-monitoring systems (MEMS) were used to measure those in the self-administration groups. Adverse event rate was the main secondary outcome. The study results were published by Annals of Internal Medicine online Nov. 7, 2017, and appeared in the Nov. 27, 2017, issue.

Overall, 1,002 patients screened between September 2012 and April 2014 were enrolled in the study, and 964 were randomly assigned to treatment. Three hundred twenty-eight patients were randomly assigned to directly observed therapy, 321 were assigned to self-administered therapy, and 315 were assigned to self-administered therapy plus reminders. Median age was 36 years. Slightly fewer than half of the patients (48.1%) were women, and 77.2% were enrolled in the U.S. Overall rates of treatment completion were as follows: 87.2% in the direct-observation group, 74.0% in the self-administration group, and 76.4% in the self-administration plus reminders group. Among U.S. patients, treatment completion rates were 85.4%, 77.9%, and 76.7%, respectively, and self-administered therapy with no reminders was noninferior to direct observation. Completion rates with self-administered treatment were high in the U.S., Spain, and Hong Kong but low in South Africa. Rates of drug-related adverse events were similar across study groups: 7.1% in the directly observed group, 8.3% in the self-administered group, and 7.9% in the self-administered plus reminders group.

The authors noted that weekly text messages could not be sent to all patients assigned to the self-monitoring plus reminders group due to access issues and that only one of the study sites was located in a country with a high burden of tuberculosis, among other limitations. However, they concluded that completion rates for once-weekly treatment with isoniazid and rifapentine were high in three of the four study sites and that self-administered therapy with monthly monitoring could be an acceptable strategy for treating latent tuberculosis in the U.S., as well as in other countries and settings when directly observed therapy is not feasible.

In an accompanying editorial, a representative from the World Health Organization's Global TB Programme said the study indicates that self-administered treatment of latent tuberculosis should be promoted in some settings and that shared decision making between patients and clinicians will be needed to determine settings in which self-administration is likely to be successful. “Evidence-based incentives and approaches tailored to the specific needs of patients and their families need to be promoted as part of programmatic management of [latent tuberculosis infection],” the editorialist wrote.

Perioperative aspirin may improve outcomes after noncardiac surgery in patients with prior PCI

Compared to placebo, perioperative aspirin during noncardiac surgery may improve 30-day outcomes in patients with prior percutaneous coronary intervention (PCI), a study found.

Researchers conducted a post hoc subgroup analysis of 470 patients with prior PCI who were enrolled in POISE-2, a previous multicenter factorial trial that randomized 10,010 patients having noncardiac surgery to receive aspirin and clonidine, placebo and clonidine, aspirin and placebo, or placebo for both drugs. The subgroup analysis compared 30-day events among patients with previous PCI who received either aspirin or placebo (200 mg within four hours before surgery).

Participants in the PCI subgroup were ages 45 years and older and were recruited from 82 centers in 21 countries. The 30-day primary outcome was a composite of death or nonfatal myocardial infarction (MI). Results were published online on Nov. 13, 2017, by Annals of Internal Medicine and appeared in the Feb. 20, 2018, issue.

Of the 470 patients with prior PCI, 234 received aspirin and 236 received placebo. Fewer patients experienced the primary outcome with aspirin than placebo (14 patients [6.0%] vs. 27 patients [11.5%]; absolute risk reduction, 5.5% [95% CI, 0.4% to 10.5%]; hazard ratio, 0.50 [95% CI, 0.26 to 0.95]). In trial participants without prior PCI, the primary outcome did not significantly differ between those who received aspirin and those who received placebo.

The beneficial effect of aspirin in the PCI cohort was driven by a reduction in MI incidence compared to placebo (12 patients [5.1%] vs. 26 patients [11.0%]; absolute risk reduction, 5.9% [95% CI, 1.0% to 10.8%]; hazard ratio, 0.44 [95% CI, 0.22 to 0.87]; P=0.021 for interaction).

In terms of major and life-threatening bleeding, a composite secondary outcome, the effect in patients with prior PCI was uncertain (absolute risk increase, 1.3% [95% CI, −2.6% to 5.2%]; hazard ratio, 5.08 [95% CI, 0.59 to 43.56]). In the overall trial population, aspirin increased the risk of major bleeding compared to placebo (230 patients [4.6%] vs. 189 patients [3.8%]; absolute risk increase, 0.8% [95% CI, 0.1% to 1.6%]; hazard ratio, 1.22 [95% CI, 1.01 to 1.48]). There was no significant interaction for major bleeding in the prior-PCI versus no-prior-PCI subgroups.

The authors noted limitations to the subanalysis, such as the small number of events that occurred in the cohort of patients with prior PCI and resulting imprecision in the effect estimates. In addition, patients with recent stent implantation were excluded, and no formal threshold was set for cardiac biomarkers, which could have affected adjudication and incidence of MI, an accompanying editorial noted.