Complicated UTI, stress ulcer prophylaxis, and more

Summaries from ACP Hospitalist Weekly.

Early antibiotic therapy did not improve outcomes in complicated urinary tract infection

Early antibiotic treatment for complicated urinary tract infection (UTI) did not affect rates of treatment failure or mortality, a recent study found.

The retrospective cohort study included 981 patients with complicated UTI treated at hospitals in 20 countries in Europe and the Middle East in 2013 and 2014. The study's primary outcome was treatment failure, which occurred in 26.6% of the patients. All-cause 30-day mortality was a secondary outcome and it occurred in 8.7% of patients, most of whom had catheter-associated UTI. Results were published by Clinical Infectious Diseases on May 17.

Multivariable analysis found several factors to be significantly associated with treatment failure: ICU admission (odds ratio [OR], 5.07; 95% CI, 3.18 to 8.07), septic shock (OR, 1.92; 95% CI, 0.93 to 3.98), corticosteroid treatment (OR, 1.92; 95% CI, 1.12 to 3.54), bedridden (OR, 2.11; 95% CI, 1.4 to 3.18), older age (OR for one-year increase, 1.02; 95% CI, 1.007 to 1.03), metastatic cancer (OR, 2.89; 95% CI, 1.46 to 5.73), and catheter-associated UTI (OR, 1.48; 95% CI, 1.04 to 2.11).

Image by Getty Images
Image by Getty Images

Overall, patients who received early appropriate empirical antibiotic treatment had a lower risk of treatment failure compared to patients who did not (46% vs. 55.8%; P=0.014). On multivariable analysis, neither appropriate empirical antibiotic treatment nor days to starting antibiotics was associated with treatment failure or 30-day mortality. Appropriate empirical antibiotic treatment was more common among patients with pyelonephritis than in those with catheter-associated UTIs (110 of 171 patients [64.3%] vs. 116 of 270 patients [43%], P<0.005), but it was not associated with any improvement in treatment failure or 30-day mortality rates even in this subgroup.

Infection with Acinetobacter baumannii or Pseudomonas aeruginosa was associated with higher risk of treatment failure, and future research should investigate whether the response to treatment truly differs by pathogen, the study authors said.

The results showed no benefit of early appropriate empirical treatment on survival rates or other outcomes, the study authors concluded. “Physicians might consider supportive treatment and watchful waiting in stable patients until the causative pathogen is defined,” they wrote. The authors noted that other research has shown significant benefits from early appropriate empirical treatment for bacteremia, sepsis, and septic shock but that effects might differ in UTIs, particularly catheter-associated ones, “as it is often difficult to distinguish between symptomatic urinary tract infection and febrile illness from other source with asymptomatic bacteriuria, a very common finding.”

H2 blockers prevented clinically important GI bleeding better than PPIs in ICU patients

For stress-ulcer prophylaxis in critically ill adults, histamine H2-receptor antagonists (H2 blockers) are associated with a significantly lower risk of clinically important gastrointestinal (GI) bleeding than proton-pump inhibitors (PPIs), a recent study found.

Researchers retrospectively assessed clinically important GI bleeding episodes in 70,093 patients who had at least one risk factor for stress ulcers and had received a PPI or H2 blocker for three or more days between Jan. 1, 2008, and June 30, 2012. Risk factors included mechanical ventilation for more than 24 hours, coagulopathy, head injuries, major burns, sepsis, corticosteroid therapy (≥250 mg of hydrocortisone per day or equivalent), acute renal failure, hepatic failure, transplantation, neurological injuries, hypotension, surgery, trauma, and ICU length of stay greater than one week. Episodes of clinically important GI bleeding were defined by one entry of the ICD-9 code that included hematemesis, blood in stool, and unspecified bleeding.

Overall, 49,576 (70.7%) patients received PPIs and 20,517 (29.3%) patients received H2 blockers. The most common risk factor for stress ulcers was mechanical ventilation (60%), and more than 50% of patients received anticoagulants, antiplatelets, or NSAIDs during their ICU stay. There were 424 (0.6%) cases of new clinically important GI bleeding in the cohort, which yielded an incidence rate of 1.2 cases per 1,000 patient-days (95% CI, 1.08 to 1.31). After adjustment for potential confounders, the PPI group had a nearly twofold higher risk of clinically important GI bleeding than the H2 blocker group (hazard ratio, 1.97; 95% CI, 1.48 to 2.63).

Other factors associated with a higher risk of GI bleeding included male gender (hazard ratio [HR], 1.27; 95% CI, 1.04 to 1.54), acute renal failure (HR, 1.59; 95% CI, 1.28 to 1.97), receipt of sucralfate (HR, 3.25; 95% CI, 2.18 to 4.85), and receipt of an antiplatelet agent (HR, 1.35; 95% CI, 1.01 to 1.79). Having a surgical procedure or being a trauma victim was associated with a lower risk of clinically important GI bleeding (HR, 0.46; 95% CI, 0.25 to 0.84). The study was published online on May 29 by CHEST and appeared in the September issue.

The study authors noted limitations, such as the fact that exposure to PPIs or H2 blockers was not randomly assigned and that confounding factors related to the tendency of an ICU to preferentially prescribe a PPI may have been present. Despite these limitations, the findings “support the conclusions of cost-effectiveness studies that favor the use of [an H2 blocker] over a PPI for stress ulcer prophylaxis of at risk critically ill adults,” they concluded.

Biomarker for ACS offers better mortality risk prediction than GRACE score

A biomarker, amyloid-β (1-40) (Aβ40), may improve risk stratification in patients with non-ST-segment elevation acute coronary syndrome (NSTE-ACS), according to a retrospective study.

Researchers used data from two European cohort studies of patients hospitalized with NSTE-ACS: the Heidelberg study (n=1145) and, for validation, the Advantageous Predictors of Acute Coronary Syndrome Evaluation (APACE) study (n=734). Median follow-up was 21.9 months and 24.9 months, respectively. Results were published online by Annals of Internal Medicine on May 22 and appeared in the June 19 issue.

In both cohorts, Aβ40 was associated with mortality, even after adjustment for variables including age, sex, diabetes, high-sensitivity cardiac troponin T and C-reactive protein levels, revascularization, and ACS type (hazard ratio [HR] for 80th vs. 20th percentiles in Heidelberg cohort, 1.66 [95% CI, 1.06 to 2.61; P=0.026]; HR in APACE cohort, 1.50 [95% CI, 1.15 to 1.96; P=0.003]). This finding remained significant even after adjustment for a commonly used risk-stratification tool, the Global Registry of Acute Coronary Events (GRACE) score (HR for 80th vs. 20th percentiles in Heidelberg cohort, 1.11 [95% CI, 1.04 to 1.18; P=0.001]; HR for APACE cohort, 1.39 [95% CI, 1.02 to 1.88; P=0.036]).

Using Aβ40 to determine patients' risk correctly reclassified both cohorts compared to the GRACE score (net reclassification index, 33.4% and 47.1% for the Heidelberg and APACE cohorts, respectively; P<0.05). “The novel finding of this study is that a single measurement of Aβ40 at presentation with NSTE-ACS confers prognostic value and correctly reclassifies patients into risk categories for mortality and a composite of death or nonfatal [myocardial infarction] over the GRACE score, which is widely recommended by clinical guidelines for accurate risk stratification,” the authors said.

They noted that accurate early risk stratification is important for making decisions about which patients require urgent interventional treatment, so this could be a useful biomarker. The authors did caution that concentrations of Aβ40 may be affected by the antibody used in testing as well as storage and preparation of samples and that researchers will therefore need to determine reference values before the biomarker can be used in clinical practice. The authors also called for research to determine whether Aβ40 could be a therapeutic target to prevent ACS or improve outcomes after it has occurred.

Combination therapy versus aspirin alone associated with fewer ischemic events, higher hemorrhage risk after minor stroke or TIA

Patients with minor ischemic stroke or high-risk transient ischemic attack (TIA) were less likely to have a major ischemic event at 90 days but more likely to have a major hemorrhage with combination antiplatelet therapy than with aspirin alone, according to a recent study.

Researchers at 269 international sites in 10 countries assigned patients with minor ischemic stroke or high-risk TIA to receive clopidogrel (loading dose of 600 mg on day 1, then 75 mg/d) plus aspirin (50 to 325 mg/d) or placebo plus the same dosage range of aspirin. In a time-to-event analysis, the primary efficacy outcome was a composite of major ischemic events at 90 days, that is, ischemic stroke, myocardial infarction, or death from an ischemic vascular event. Study results were published May 16 by the New England Journal of Medicine and appeared in the July 19 issue.

Image by Getty Images
Image by Getty Images

Overall, a total of 4,881 patients were enrolled in the study, 2,432 of whom received combination therapy and 2,449 of whom received placebo plus aspirin. Most of the patients, 82.8%, were in the U.S. Major ischemic events occurred in 121 and 160 patients in each group (5.0% and 6.5%, respectively; hazard ratio [HR], 0.75; P=0.02), most commonly within the first week after the initial stroke or TIA. Ischemic stroke, a secondary outcome, occurred in 112 patients in the combination therapy group and 155 patients in the aspirin group (4.6% vs. 6.3%; HR, 0.72; P=0.01). Twenty-three patients in the combination therapy group and 10 patients in the aspirin group developed major hemorrhage (0.9% vs. 0.4%; HR, 2.32; P=0.02). The trial was stopped early by the data monitoring board due to excess hemorrhage risk in the combination therapy group and evidence of treatment efficacy.

The researchers noted that their trial did not include patients who had moderate to severe stroke or cardioembolic stroke or those who were candidates for thrombolysis or thrombectomy and that aspirin doses were determined by the clinical judgment of each site's investigator, among other limitations. However, they concluded that combination therapy with clopidogrel and aspirin was associated with a lower risk of the composite outcome but a higher risk of major bleeding versus aspirin alone over the initial 90-day period. They estimated that for every 1,000 similar patients treated with clopidogrel plus aspirin for 90 days, approximately 15 ischemic events would be prevented and five major hemorrhages would be caused.

The author of an accompanying editorial noted that although the study results might suggest no net benefit with combination therapy, most of the events that were prevented were ischemic stroke, “the most common and arguably most important of the outcomes after a TIA or minor stroke,” and most of the bleeding complications were nonfatal. He also pointed out the timing of the outcomes. “Most of the benefit regarding stroke prevention occurred in the first week of treatment with the combination, whereas most of the bleeding occurred later,” he wrote.

The take-home message for clinicians from this and related research, the editorialist said, is that aspirin plus clopidogrel reduces risk for recurrent ischemic stroke during a high-risk period immediately after a TIA or noncardioembolic ischemic stroke but that it should be used only for the first three weeks before transitioning to monotherapy. He noted that dual therapy should not be recommended for patients for whom follow-up and adherence are uncertain and may not be advisable in patients with an uncertain diagnosis of TIA or in patients at increased risk for bleeding.

BNP predicts mortality, even in patients without heart failure

B-type natriuretic peptide (BNP) levels predicted mortality risk better than some traditional risk factors, even in patients without heart failure, a recent study found.

Researchers used data from the Vanderbilt University Medical Center electronic health record to identify 30,487 patients who had a first plasma BNP measurement between 2002 and 2013. Their median age was 63 years, 50% were men, 17% were black, and 38% were diagnosed with heart failure. Follow-up continued through 2015, and results were published in the May 15 Journal of the American College of Cardiology.

Over 90,898 person-years of follow-up, 31% of the patients without heart failure and 53% of those with heart failure died. BNP levels were lower in patients without heart failure than in those with heart failure (median, 89 pg/mL [interquartile range, 34 to 238 pg/mL] vs. 388 pg/mL [interquartile range, 150 to 940 pg/mL]; P<0.0001). However, the risk for death according to BNP level was similar regardless of whether patients had heart failure. For example, a BNP level of 400 pg/mL was associated with a three-year risk for death of 21% (95% CI, 20% to 23%) in patients with heart failure and 19% (95% CI, 17% to 20%) in those without.

This increase in mortality was observed whether patients' elevated BNP level was found in an acute care or outpatient setting. Higher BNP level was the strongest predictor of mortality risk among patients without heart failure, even in multivariate models including traditional risk markers such as age, renal function, diabetes, vital signs, left ventricular mass, and left ventricular ejection fraction. The authors noted that the latter two factors were found to be strongly associated with higher BNP levels. “Thus, finding an elevated BNP level in a patient without [heart failure] may warrant additional investigation, including assessment of cardiac structure and function,” the authors said.

They cautioned that as an observational analysis, the study is susceptible to residual confounding, and that other potentially predictive factors, such as C-reactive protein and troponin, were not compared to BNP. Variability in treatment, such as high diuretic doses for patients with acute heart failure decompensation, may also have affected BNP levels, they noted.

An accompanying editorial called the results “stimulating” and offered several possible explanations for the prognostic value of BNP in patients without heart failure. Elevated BNP levels could be a sign of asymptomatic cardiac disease, an effect of noncardiac conditions such as sepsis or chronic obstructive pulmonary disease, or a sign of vascular aging. Based on the study's findings, the editorialists recommended more extensive cardiovascular evaluation for patients found to have BNP levels above 35 pg/mL in outpatient settings or above 100 pg/mL in acute care. They noted that the next question for researchers will be how to treat patients who have an elevated BNP level without heart failure.

Encouraging use of subcutaneous opioids reduces IV opioid prescribing, pilot study finds

A pilot intervention that promoted subcutaneous opioid administration substantially reduced IV opioid prescribing while improving patient-reported pain control on one hospital unit.

The 28-bed adult general medical unit adopted a standard of practice preferring oral and subcutaneous (for patients unable to take pills) over IV opioids, although prescribers' orders were not restricted. The intervention also included education for prescribers and nursing staff to increase awareness and use of subcutaneous opioid administration.

Researchers compared opioid use between a six-month control period (n=287 patients) and a three-month intervention period (n=127 patients). The primary outcome was the number of IV doses administered per patient-day. Secondary outcomes included total parenteral and overall opioid doses per patient-day, parenteral and overall opioid exposure per patient-day, and daily rate of patients receiving parenteral opioids. Researchers also measured pain scores on a 0 to 10-point Likert scale during the first five days of hospitalization and compared them between groups. Results were published online on May 14 by JAMA Internal Medicine and appeared in the June issue.

The control period included 4,500 patient-days, and the intervention period included 2,459 patient-days. During the intervention, 65% of parenteral opioid doses were administered subcutaneously, compared with fewer than 1% during the control period. IV opioid doses decreased by 84% from the control to the intervention period (0.39 vs. 0.06 doses per patient-day; P<0.001).

Doses of all parenteral opioids decreased by 55% after the intervention (0.39 vs. 0.18 doses per patient-day; P<0.001), and mean daily parenteral opioid exposure decreased by 49% (5.67 vs. 2.88 morphine-milligram equivalents [MMEs] per patient-day). The daily rate of patients receiving parenteral opioids decreased by 57% (14% vs. 6%; P<0.001). In addition, overall opioid doses decreased by 23% (0.95 vs. 0.73 doses per patient-day; P=0.02), and mean daily overall opioid exposure decreased by 31% (9.11 vs. 6.30 MMEs per patient-day).

There were no significant differences in patients' mean reported pain scores for hospital days 1 through 3, but there was significant improvement in the intervention group on day 4 (4.82 vs. 3.75; difference, −1.07; 95% CI, −1.80 to −0.34; P=0.004) and on day 5 (4.65 vs. 3.59; difference, −1.06; 95% CI, −1.84 to −0.27; P=0.009).

The study authors noted limitations, such as how the intervention was a pilot project conducted on a single unit at one academic medical center. In addition, prescribers in specialties other than internal medicine did not participate in the study, which may limit generalizability, and pain scores may not have captured discomfort associated with subcutaneous administration, they noted.

The authors added that the results occurred without any changes to the electronic health record, which could potentially increase its effects. “This intervention should be scalable to larger populations of inpatients, and future directions may include study in other areas of hospital practice, including the surgical, emergency, obstetric, pediatric, and critical care environments,” they wrote.