Frail elderly, Enterobacteriaceae infections, and more

Summaries from ACP Hospitalist Weekly.


Frail elderly account for majority of potentially preventable Medicare spending

Among subpopulations of high-cost Medicare beneficiaries, frail elderly patients account for the majority of potentially preventable spending, according to a recent study.

Researchers used a 20% sample of Medicare fee-for-service claims from 2012 to examine the proportion of total spending that was potentially preventable in high-cost subpopulations. Beneficiaries were considered “high cost” if they were in the highest 10% of total standardized individual spending. Six subpopulations were defined: nonelderly disabled, frail elderly, major complex chronic, minor complex chronic, simple chronic, and relatively healthy. Costs for avoidable ED visits were added to costs for inpatient and associated 30-day post-acute costs for ambulatory care-sensitive conditions to arrive at potentially preventable spending. The researchers then compared the amount and proportion of spending that was potentially preventable across high-cost subpopulations and by ambulatory care-sensitive condition. Results were published online Oct. 17, 2017, by Annals of Internal Medicine and appeared in the Nov. 21, 2017, issue.

Photo by Thinkstock
Photo by Thinkstock.

A total of 6,112,450 Medicare beneficiaries were included in the study. Overall, 17.9% were assigned to the nonelderly disabled group, 8.6% were assigned to the frail elderly group, 18.0% were assigned to the major complex chronic group, 27.8% were assigned to the minor complex chronic group, 18.0% were assigned to the simple chronic group, and 9.7% were assigned to the relatively healthy group. In the study year, 2012, 4.8% of Medicare spending was considered potentially preventable. High-cost frail elderly patients made up 4% of the Medicare population but accounted for 43.9% of overall potentially preventable spending ($6,593 per person), while high-cost elderly disabled persons and persons with major complex chronic conditions accounted for 14.8% ($3,421 per person) and 11.2% ($3,327), respectively. Potentially preventable spending in the frail elderly group was most commonly related to acute care visits for heart failure, bacterial pneumonia, long-term complications of diabetes, and dehydration.

The authors noted that although potentially preventable acute care episodes were defined by using well-established algorithms, some admissions may not have been truly preventable. In addition, they pointed out that their 30-day time frame may have missed some preventable spending and that presence of chronic disease may have been underestimated, among other limitations. However, they concluded that potentially preventable spending varies widely across Medicare subpopulations and that frail elderly patients accounted for the majority of potentially modifiable costs. “Therefore, as we continue to move toward value-based frameworks, interventions that focus on frail elderly patients may be particularly valuable,” the authors wrote.

An accompanying editorial said that many steps will be needed to provide high-value care for every high-need, high-cost older adult, including a faster shift toward value-based health care; new data collection, analysis, and treatment planning; and new models for payment and patient assignment. “This all needs to happen as our health system addresses the critical challenges of lack of universal access to health care, underinvestment in primary care, administrative inefficiency, and disparities in delivery of care,” the editorialists wrote. Studies like the current one “can help point the way to ‘coolable’ hot spots,” they noted. “The onus is now on organizations and systems to shift culture and learn to implement the care and contracting methods used by their ‘coolest’ peers.”

Short course of antibiotics worked as well as long for Enterobacteriaceae bacteremia

For patients with Enterobacteriaceae bloodstream infections, a short course of antibiotics led to similar outcomes as a longer course, a recent study found.

The retrospective cohort study included adult patients from three medical centers who had Enterobacteriaceae bacteremia treated with in vitro active therapy between 2008 and 2014. Patients who had received a short course of therapy (median duration, eight days) were matched 1:1 with those who had received a long course of therapy (median duration, 15 days), resulting in 385 matched pairs. Results were published online by Clinical Infectious Diseases on Oct. 8, 2017, and appeared in the January 2018 issue.

The study's primary outcome was all-cause mortality within 30 days of the end of antibiotic treatment, and it did not differ between the short- and long-course groups. Rates of recurrent bloodstream infections and Clostridium difficile infections within 30 days were also similar between groups. The short course of therapy also showed a slight protective effect against developing multidrug-resistant gram-negative bacteria within 30 days, but this was only a trend, not a statistically significant difference (odds ratio, 0.59; 95% CI, 0.32 to 1.09; P=0.09).

The study authors noted that current Infectious Diseases Society of America guidelines suggest antibiotics for seven to 14 days for such patients but that this recommendation is based on low-grade evidence. Previous research has shown variability in prescribing practices and the need for high-quality evidence to standardize therapy for gram-negative bloodstream infections, the authors said. The current study was limited by its reliance on observational data, but the authors attempted to overcome that problem by using propensity score matching to create comparator groups that were similar in demographics, pre-existing medical conditions, source control measures, and severity of illness.

Based on the results, the authors concluded that a short course of antibiotic therapy yields similar clinical outcomes as longer courses for patients with Enterobacteriaceae bacteremia and that a shorter course may also protect to some extent against multidrug-resistant gram-negative bacteria. The results do not apply to patients with inadequate source control, because few such patients were included in the study, the authors noted.

ICU admission may not reduce mortality at 6 months in critically ill elderly patients

Critically ill elderly patients may not always benefit from systematic ICU admission, according to a recent French study.

Researchers performed a multicenter, cluster-randomized trial in which 24 hospitals in France were randomly assigned to use a program promoting systematic ICU admission in elderly patients or to follow standard practice. Patients were 75 years of age or older, did not have cancer, had preserved functional and nutritional status, and presented at the ED between January 2012 and April 2015.

In hospitals using the systematic strategy, ED and ICU physicians were asked to systematically recommend ICU admission during triage for all patients included in the study. In addition, the trial protocol was presented at each hospital by a member of the study's steering committee, ED physicians were required to systematically call the attending ICU physicians whenever they entered a patient into the trial, ICU physicians were required to systematically evaluate patients at the bedside, and ED and ICU physicians were required to make a joint decision on whether to admit patients to the ICU. Patients for whom admission was recommended were transferred to an ICU at another hospital if no ICU beds were available. ED and ICU staff met monthly, and booklets and posters outlined recommendations for ICU admission. In hospitals using standard practice, no recommendations were made about the ICU triage process and physicians at the bedside made the final decision about ICU admission.

The study's primary outcome was death at six months. ICU admission rate, in-hospital mortality rate, functional status, and quality of life at six months were secondary outcomes. Patients were followed until November 2015. Results of the study were published online Sept. 27, 2017, by JAMA and appeared in the Oct. 17, 2017, issue.

Overall, 3,036 critically ill patients ages 75 years and older were included in the study. Of these, 1,518 were cared for at hospitals in the systematic strategy group and 1,518 were cared for at hospitals in the control group. Patients' median age was 85 years, and 1,361 (45%) were men. Septic shock (13.6%), acute respiratory failure that required noninvasive ventilation (11.4%), severe pneumonia (8.2%), and cardiac insufficiency that required noninvasive ventilation (7.2%) were the most common reasons for hospital admission. Severity of illness at admission, as measured by the Simplified Acute Physiology Score III, was higher in the systematic strategy group.

At six months, patients in the systematic strategy group had a higher risk for death than those in the control group (45% vs. 39%; relative risk [RR], 1.16; 95% CI, 1.07 to 1.26), although their ICU admission rate was higher (61% vs. 34%; RR, 1.80; 95% CI, 1.66 to 1.95). Likelihood of ICU admission was higher in the systematic strategy group after adjustment for baseline characteristics (RR, 1.68; 95% CI, 1.54 to 1.82), and risk for in-hospital death was higher (RR, 1.18; 95% CI, 1.03 to 1.33). Also after adjustment, no significant increase in risk for death at six months was seen in the systematic strategy group (RR, 1.05; 95% CI, 0.96 to 1.14). The researchers found no significant difference between groups in functional status or quality of life at six months.

The recruitment period for the standard practice group was longer than that for the systematic strategy group, and blinding to group assignment was not possible, the authors noted. They also pointed out that data on withdrawal of life-sustaining therapies were not collected and that confounding might have occurred due to recruitment of sicker patients in the systematic strategy group, among other limitations. However, they concluded that their systematic program promoting ICU admission was associated with higher ICU admission rates and rates of hospital mortality but did not appear to affect mortality, functional status, or health-related quality of life at six months.

On the basis of their results, the authors said that systematic ICU admission of all critically ill elderly patients may not be warranted. “However, this study should not be interpreted as suggesting that no elderly patient should be admitted to the ICU,” the authors wrote. “Because of substantial uncertainty in outcomes among individual patients, there is a need to systematically and thoughtfully assess the potential benefits and harms of ICU admission for every elderly patient presenting with critical illness.”

The author of an accompanying editorial said that although the study had several inherent limitations, it was well conducted and its results may have implications for clinicians and policymakers regarding ICU access in elderly patients, especially in France. The editorialist also pointed out that several questions still need to be answered, including whether and how any of the trial patients were harmed by ICU care, what criteria should be used to audit ICU use, and whether there are beneficial practices in general wards and harmful practices in ICUs that can be promoted and eradicated, respectively. In addition, for countries like the U.S. where ICU beds are in much greater supply than in France, the trial's findings “certainly support an argument for close examination of ICU admission decisions, with the potential to safely reduce ICU beds, care, and costs,” the editorialist wrote.

Anticoagulation didn't benefit patients with afib secondary to ACS, pulmonary disease, or sepsis

Patients who received anticoagulants because they were diagnosed with atrial fibrillation (AF) while hospitalized for acute coronary syndrome (ACS), acute pulmonary disease, or sepsis did not see benefit from anticoagulants, according to a recent study.

The retrospective cohort study include 2,304 Canadian patients ages 65 years or older who were hospitalized with a primary diagnosis of ACS, acute pulmonary disease (defined as chronic obstructive pulmonary disease, pneumonia/influenza, pulmonary embolism, or pleural effusion), or sepsis from 1999 to 2015. All were diagnosed with new-onset AF as a secondary complication. Results were published online Sept. 27, 2017, by JACC: Clinical Electrophysiology.

A third of the patients were discharged on anticoagulation. Over a three-year follow-up period, receiving anticoagulation was not associated with a significant reduction in ischemic stroke in any of the primary diagnosis groups (adjusted odds ratio [OR], 1.22 [95% CI, 0.65 to 2.27] for ACS, 0.97 [95% CI, 0.53 to 1.77] for pulmonary disease, and 1.98 [95% CI, 0.29 to 13.47] for sepsis). Anticoagulation was associated with a significantly increased risk of bleeding in patients with acute pulmonary disease (OR, 1.72; 95% CI, 1.23 to 2.39). There was a trend toward higher bleeding risk in ACS patients (OR, 1.42; 95% CI, 0.94 to 2.14) and no effect in sepsis patients (OR, 0.96; 95% CI, 0.29 to 3.21).

The results show that clinicians were less likely to use anticoagulation in patients with secondary AF, compared to those with primary AF, among whom other studies have shown anticoagulation rates of 46% to 60%, the authors observed. This study's patient population had high rates of comorbidity and thus high risk for both bleeding and stroke (approximately 45% to 60% with a HAS-BLED score ≥3 and approximately 60% to 65% with a CHADS2 score ≥2). They had a higher risk for bleeding (3.6 to 6.2 per 100 person-years) than for stroke (1.1 to 1.6 per 100 person-years). “One possibility for the lack of observed benefit in anticoagulation is that risk of stroke and bleeding may be different in secondary AF compared to primary AF,” the authors suggested.

The study was limited by inability to determine whether AF was transient, the overall low incidence of stroke, and infrequent use of direct-acting oral anticoagulants, among other factors. Still, the results show that it is unclear whether anticoagulation benefits patients who develop AF secondary to ACS, acute pulmonary disease, or sepsis and support individual assessment rather than routine use of anticoagulants in this patient population, according to the authors.

Four factors predict risk of DVT when it's suspected in hospitalized patients

An algorithm based on four variables identified inpatients with suspected deep venous thrombosis (DVT) who had low enough risk of a clot that ultrasonography might be unnecessary, a study found.

The prospective cohort study included 2,960 adult inpatients who underwent lower-extremity ultrasonography for suspected DVT at one academic medical center. A derivation cohort included the 1,135 patients treated from Nov. 1, 2012, through Dec. 31, 2013, while the validation cohort included 1,825 patients treated from Jan. 1, 2014, through May 31, 2015. Results were published by the Journal of General Internal Medicine on Sept. 15, 2017.

The primary outcome of the study was proximal DVT, which was found in 11.6% of the patients, and the secondary outcome, any DVT, occurred in 20.4% of patients. Using recursive partitioning, researchers identified four variables that predicted DVT and assigned each of them a point value: history of DVT (six points), active cancer (one point), hospital stay of six or more days before imaging (one point), and age 46 years or older (one point). Patients with no points were considered low risk and those with any points were considered high risk.

The overall incidence of proximal DVT in the validation cohort was 2.9% in the low-risk patients and 12.0% in the high-risk patients, and incidence rates of any DVT were 5.2% and 21.0%, respectively. The model's area under the curve for risk of proximal DVT was 0.73, and its sensitivity was 98.1% for proximal or any DVT. The results show that a setting-specific risk stratification model has greater predictive accuracy than the original Wells criteria for DVT, although this model was 4% less efficient than the Wells criteria (with efficiency defined as the proportion of patients in the low-risk group).

The model would be easy to implement in a computerized clinical decision support system, and its ability to identify an inpatient subgroup with a less than 3% risk of proximal DVT might allow some patients to safely avoid ultrasonography, the study authors said. “While we feel that a <3% chance is a reasonable, acceptable risk, we also believe that the ultimate decision regarding imaging should be made by patients and their physicians,” they wrote. The generalizability of the findings to other settings was also unclear, the authors noted.

AHA updates practice standards for electrocardiographic monitoring

The American Heart Association (AHA) recently released a scientific statement on electrocardiographic monitoring of hospitalized patients.

Photo by Thinkstock
Photo by Thinkstock.

The statement updated practice standards published by the AHA in 2004 and was developed by experts in general cardiology, electrophysiology, interventional cardiology, hospital medicine, and alarm management. It was published online by Circulation on Oct. 3, 2017, and appeared in the Nov. 7, 2017, issue. The recommendations are graded according to their level of evidence support, but the authors noted that many of the recommendations are based on limited data.

The updated statement offers recommendations for indication and duration of electrocardiographic monitoring by patient population, including (among others) patients with chest pain, coronary artery disease, or arrhythmias; patients who are post-myocardial infarction; patients who have undergone percutaneous coronary intervention (PCI), diagnostic coronary angiography, or a major cardiac intervention; patients with other cardiac conditions, including acute decompensated heart failure; and patients hospitalized for noncardiac conditions, including surgery, stroke, pneumonia, or drug overdose.

In addition to the specific clinical recommendations, the statement includes four other sections that:

  • provide an overview of arrhythmia, ischemia, and QTc monitoring;
  • discuss organizational aspects of monitoring, including alarm management, education of staff, and documentation;
  • offer recommendations on implementation of practice standards; and
  • call for additional research.

The authors noted that among other changes, one significant difference from the 2004 practice standards was a downgrading of the recommendation for continuous ST-segment monitoring for certain patients at risk of myocardial ischemia (those in the early phase of acute coronary syndrome, those who have chest pain or angina-equivalent symptoms in the ED, those who had nonurgent PCI with suboptimal angiographic results, or those with possible variant angina caused by coronary vasospasm), “not because we doubt the potential for this technology but because recent literature demonstrates a very serious problem with false and nonactionable alarm signals.”

Ceftazidime-avibactam may be ‘reasonable alternative’ to colistin when treating CRE infections, study finds

Using ceftazidime-avibactam rather than colistin to treat patients with carbapenem-resistant Enterobacteriaceae (CRE) infections may yield better outcomes, according to a recent study.

Researchers enrolled 137 patients with Klebsiella pneumoniae carbapenemase-producing CRE infections initially treated with either ceftazidime-avibactam (n=38) or colistin (n=99) who participated in a previous multicenter study, Consortium on Resistance Against Carbapenems in Klebsiella and other Enterobacteriaceae (CRACKLE).

The most common infections among participants were of the bloodstream (n=63, 46%) and respiratory tract (n=30, 22%), and most patients received additional anti-CRE treatments. Nearly all patients (n=133, 97%) were infected with Klebsiella pneumoniae, while the other four (3%) were infected with Enterobacter species.

The researchers analyzed the efficacy of each treatment by comparing patients' disposition 30 days after treatment initiation. They adjusted all analyses for confounders with covariates including the Pitt bacteremia score, type of infection, and creatinine level ≥2 mg/dL at time of first positive culture. Results were published online on Sept. 4, 2017, by Clinical Infectious Diseases and appeared in the January 2018 issue.

Patients who received ceftazidime-avibactam had significantly lower adjusted all-cause hospital mortality 30 days after treatment initiation compared to those who received colistin (9% vs. 32%; difference, 23% [95% bootstrap CI, 9% to 35%]). Furthermore, 30 days after starting treatment, the adjusted probability of a better outcome on ceftazidime-avibactam versus colistin was 64% (95% CI, 57% to 71%), according to desirability of outcome ranking analyses.

The authors noted limitations to the study, such as its observational design, the possibility of unmeasured confounders, and the relatively small number of participants. They concluded that ceftazidime-avibactam may be a “reasonable alternative” to colistin when treating patients with CRE infections. The results must be confirmed in a randomized controlled trial, they wrote, noting that unfortunately no such data on ceftazidime-avibactam are “imminently expected.”