Sepsis cultures, resuscitation score, and more

Summaries from ACP Hospitalist Weekly.

Study affirms recommendations to collect blood cultures before initiating antimicrobial therapy for sepsis

Initiating empirical antimicrobial therapy in patients presenting to the ED with severe manifestations of sepsis reduces the sensitivity of blood cultures drawn shortly after treatment initiation, a recent study found.

At seven EDs in the U.S. and Canada, researchers obtained blood cultures in patients who presented with severe manifestations of sepsis. Inclusion criteria were evidence of systemic inflammatory response syndrome and a presumed or confirmed source of infection. Overall, 3,164 adults were screened, and 325 (mean age, 65.6 years; 62.8% men) were enrolled in the study between November 2013 and September 2018. Participants had two sets of blood cultures drawn before starting antimicrobial therapy and had additional sets drawn after empirical antimicrobial administration. Results were published online on Sept. 17, 2019, by Annals of Internal Medicine and appeared in the Oct. 15, 2019, issue.

Image by Getty Images
Image by Getty Images

Repeat blood cultures were drawn a median of 70 minutes after initiation of antimicrobial therapy (interquartile range, 50 to 110 min). A total of 264 patients had postantimicrobial blood cultures drawn between 30 and 120 minutes after treatment initiation, nine had cultures obtained between 1 and 29 minutes after initiation, and 52 had cultures obtained between 121 and 240 minutes after initiation. Blood cultures taken before antimicrobial therapy were positive for one or more microbial pathogens in 102 of 325 (31.4%) patients, whereas those taken after antimicrobial therapy were positive for one or more pathogens in 63 of 325 (19.4%) patients (absolute difference, 12.0% [95% CI, 5.4% to 18.6%]; P<0.001). Sensitivity of the cultures was reduced at all times after antimicrobial administration, and the overall sensitivity of postantimicrobial blood cultures was 52.9% when preantimicrobial cultures were used as a reference standard.

The study authors noted that a small minority of screened patients were enrolled and that the initial protocol specified a time window of 120 minutes for obtaining postantimicrobial blood cultures, although the results were consistent among the entire study and per protocol populations. “Despite the importance of early antimicrobial administration in this patient population, our results support the Surviving Sepsis Campaign guidelines and suggest that preantimicrobial blood cultures should not be routinely deferred,” they concluded.

It is crucial to prevent unintentional treatment delays, especially since blood cultures can generally be obtained in a few minutes when standard equipment and qualified personnel are available, an accompanying editorial noted. “If more sensitive techniques do come into widespread use, we will likely be prompted to ask again if it is necessary to withhold antibiotics until testing is done,” the editorialists wrote. “Until then, we would advise clinicians to quickly obtain blood cultures and, immediately after that, to give patients the crucial antimicrobial agents they need to achieve the best possible hope of survival.”

Score may help inform DNR discussions after in-hospital cardiac arrest

The Good Outcome Following Attempted Resuscitation (GO-FAR) score can estimate at admission the likelihood of a patient surviving to discharge with good neurologic outcome after an in-hospital cardiac arrest and may help inform discussions about do-not-resuscitate (DNR) status, a study found.

Researchers performed a retrospective observational study at a level 1 trauma and academic hospital in Minneapolis to determine the utility of the GO-FAR score in predicting prognosis. All adults who were admitted to the hospital between January 2009 and September 2018 and had a first case of pulseless in-hospital cardiac arrest while hospitalized were included in the study. The GO-FAR score was calculated for each patient based on variables in the electronic health record at admission. The study's primary outcomes were survival to discharge and survival to discharge with good neurologic outcome (defined as alert, able to work, and lead a normal life, possibly with minor psychologic or neurologic deficits). Results were published Sept. 11, 2019, by the Journal of General Internal Medicine.

During the study period, 403 patients had a cardiac arrest during their hospitalization. Most (65.5%) were men; mean age was 60.3 years. Overall, 33.0% of patients survived to discharge, and 17.4% survived with good neurologic outcome. Among patients whose GO-FAR score indicated that their likelihood of survival was below average, average, and above average, 5.3%, 22.5%, and 34.1%, respectively, survived to discharge with good neurologic outcome. Overall, 44.6% of in-hospital cardiac arrests occurred within the first 48 hours of hospitalization, and these patients were more likely to survive to discharge (39.4% vs. 11.6%; P=0.01) and survive to discharge with good neurologic outcome (23.3% vs. 10.8%; P=0.005) than those whose arrest occurred after hospital day 2.

The researchers noted that their study was retrospective, involved only one hospital, and had data only for patients who had undergone cardiopulmonary resuscitation (CPR) for their in-hospital arrest, among other limitations. However, they concluded that based on their results and those of previous GO-FAR studies, the GO-FAR calculator can provide information at admission to help guide discussions with patients about CPR. Patients whose GO-FAR score is 14 or above can be told that they have a very low likelihood of surviving to discharge without some neurologic impairment, the authors said.

“After eliciting the patient's values, providers can then give a recommendation for either a Full Code or DNR status, emphasizing that the DNR status does not itself imply any limits on use of life-sustaining treatments while the patient has a pulse,” they wrote. They also stressed that clinicians should address code status again after 48 hours of hospitalization, especially in patients with an unfavorable GO-FAR score who have opted for Full Code, since in-hospital cardiac arrest on hospital day 3 or later is associated with even lower likelihood of neurologically intact survival.

Antipsychotics not effective for prevention, treatment of inpatient delirium

Antipsychotics are not effective for the prevention or treatment of delirium in hospitalized adults, according to two recent reviews.

One review focused on use of antipsychotics for the prevention of delirium and included 14 randomized controlled trials. The included trials that compared haloperidol with placebo showed no differences between the two in delirium incidence or duration, hospital length of stay, or mortality. There was also little to no evidence on the effect of haloperidol on cognitive function, delirium severity, inappropriate continuation, or sedation. For second-generation antipsychotics, there was limited evidence supporting lower delirium incidence in the postoperative setting. Looking at harms of antipsychotics, the review found little evidence of neurologic harms with short-term use but did find potentially harmful cardiac effects.

The review authors noted that guidelines suggest not using antipsychotics for prevention of delirium in the ICU and have found insufficient evidence to recommend them in other settings. The results of this review differ from a prior one, which found benefits of haloperidol prophylaxis in critically ill patients, possibly due to the inclusion of newer research. The current review was limited by heterogeneity in the studies, and more research may be needed in the perioperative setting and other high-risk populations, the study authors said.

The other review evaluated antipsychotics for treatment of inpatient delirium. It included 16 randomized controlled trials and 10 observational studies and found no difference in sedation status, delirium duration, hospital length of stay, or mortality with haloperidol and second-generation antipsychotics versus placebo. When haloperidol and second-generation antipsychotics were compared to each other, there was no difference in delirium severity and cognitive functioning. In direct comparisons of second-generation antipsychotics, there was no difference in mortality and insufficient or no evidence for other outcomes. On harms of antipsychotics, the results were similar to the prevention trials—no neurologic harm, but potential cardiac effects.

“Our findings are also consistent with recent clinical practice guidelines that do not recommend routine use of antipsychotics for treating delirium,” the authors said. They noted that some outcomes, such as patient distress and function, and some patient subgroups, such as older adults and palliative care patients, were not rigorously evaluated by the included studies, so more trials are needed.

Both reviews and an accompanying editorial were published by Annals of Internal Medicine on Sept. 3, 2019, and appeared in the Oct. 1, 2019, issue. An accompanying editorial agreed with the need for more research but added that the evidence is already definitive enough for action. “With regard to use of antipsychotics for broad treatment of delirium, I believe the findings presented are sufficient to stop this clinical practice,” the editorial said. “In my opinion, a patient-centered bundled approach that 1) identifies delirium early, 2) evaluates and addresses underlying causes, 3) prevents complications, and 4) promotes functional recovery is most likely to succeed. Identifying which practices belong in this bundle, and how to deliver it in a standardized, high-quality, and sustainable way, should be a major focus of the next generation of delirium treatment research.”

Less intensive warfarin therapy less effective at preventing VTE after arthroplasty

A lower-than-standard target international normalized ratio (INR) was not as effective at preventing venous thromboembolism (VTE) or death among patients taking warfarin after a hip or knee arthroplasty, a recent study found.

The trial included patients ages 65 years or older who initiated warfarin for elective hip or knee arthroplasty at six U.S. medical centers. The patients were randomized in a two-by-two factorial design to a target INR of either 1.8 (low target) or 2.5 (standard target) and to either genotype-guided or clinically guided warfarin dosing. For the first 11 days of therapy, warfarin dosing was guided by a web application. Results were published in the Sept. 3, 2019, JAMA.

A total of 1,650 patients were randomized (mean age, 72.1 years; 63.6% women; 91.0% white), 96.8% of whom received at least one dose of warfarin and were included in the analysis. The study's primary outcome was a composite of VTE within 60 days and death within 30 days. Rates were 5.1% in the low-target group versus 3.8% in the standard-target group, a difference of 1.3% (one-sided 95% CI, −∞ to 3.05%, P=0.06 for noninferiority). This difference did not meet the study's prespecified noninferiority margin of 3% for the absolute risk of VTE.

The secondary endpoint of major bleeding occurred in 0.4% of the low-target group and 0.9% of the standard-target group, a difference of −0.5% (95% CI, −1.6% to 0.4%). The other secondary endpoint of an INR value of 4 or more occurred in 4.5% and 12.2% of the patients, respectively, a difference of −7.8% (95% CI, −10.5% to −5.1%).

The study authors concluded that an INR target of 1.8 did not meet the criterion for noninferiority or significantly reduce clinically relevant bleeding compared to the standard target of 2.5. However, they noted that the rate of the primary outcome was lower than anticipated, so the trial may have been underpowered. Practices in the study that may have enhanced safety included the use of the web application and genotype guidance for warfarin dosing.

The trial's results were similar to those of several previous studies, which also found that INR targets of 2 and above were associated with fewer VTEs after surgery, but more research might be helpful to identify whether some patients at high risk of bleeding could benefit from a lower INR target, the authors said. They noted that limitations of this study include its open-label design and its small proportion of minority patients.

Strategies for venous thromboembolism prophylaxis after knee replacement compared

To prevent deep venous thrombosis after total knee replacement, single prophylaxis strategies were more effective than combinations, and rivaroxaban was the most effective single strategy, according to a recent meta-analysis.

Image by Getty Images
Image by Getty Images

Researchers conducted a systematic review and meta-analysis of 25 randomized controlled trials published in 2017 or before to assess the relative efficacy and safety of venous thromboembolism prophylaxis strategies: 23 with deep venous thrombosis as an outcome, 12 with pulmonary embolism, and 19 with major bleeding. Results were published by The Lancet Haemotology on Aug. 20, 2019, and appeared in the Oct. 1, 2019, issue.

For deep venous thrombosis, the analysis found that single prophylaxis strategies were more effective than combination strategies. The top three interventions in the deep venous thrombosis meta-analysis were rivaroxaban (relative risk [RR], 0.12; 95% credible interval [CrI], 0.06 to 0.22]), followed by apixaban (RR, 0.15; 95% CrI, 0.07 to 0.26), and low-molecular-weight heparin (LMWH) at a high prophylactic dose for a standard duration (10 to 14 days; RR, 0.18; 95% CrI, 0.10 to 0.30).

The meta-analyses for pulmonary embolism and major bleeding had highly uncertain results, but LMWH ranked first for both outcomes. For pulmonary embolism, with a standard prophylactic dose for 28 to 35 days, the RR was 0.02 (95% CrI, 0.00 to 3.86), while for bleeding, with a low prophylactic dose for 10 to 14 days, the odds ratio was 0.08 (95% CrI, 0.00 to 1.76). The study authors said that on this outcome “no clear conclusion can be made other than what is biologically plausible (eg, that no prophylaxis and mechanical prophylaxis strategies should have the lowest risk of major bleeding).”

The authors noted that British guidelines have recommended combination prophylaxis strategies for patients undergoing elective total knee replacement surgery. “These findings call into question the value of combined prophylaxis in this population, which needs to be further examined in head-to-head randomized controlled trials,” they wrote.

An accompanying editorial comment also noted that this meta-analysis, although it was used to inform guidelines, is no substitute for randomized controlled trials. The editorialists also recommended that prophylaxis decisions be individualized based on patients' risks for venous thromboembolism or bleeding. “Until we navigate between safety, simplicity, cost-effectiveness, and practicality, especially in busy ward settings, the search for the most appropriate form of venous thromboembolism prophylaxis continues,” the comment concluded.

Cognitive biases affect decisions about whether to go to a SNF after discharge

The decision about whether a patient should go to a skilled nursing facility (SNF) after discharge is influenced by cognitive biases of both patients and clinicians, a recent study found.

To understand how cognitive biases affect decisions regarding postacute care in SNFs, researchers conducted a secondary analysis of 105 semi-structured interviews with patients (all >65 years of age and discharged to SNFs after hospitalization), their caregivers, and multidisciplinary frontline clinicians. Interviewees were gathered from three hospitals and three SNFs (some Veterans Health Administration facilities, some not) in a single metropolitan area. Results were published by the Journal of Hospital Medicine on Aug. 21, 2019, and appeared in the January issue.

Authority bias and halo effect were the most prevalent biases identified in patients' reports. For example, veterans expressed confidence in the health care system providing direction on whether to go to a SNF and which to choose, while nonveteran patients trusted their health care team's perspectives on the decisions. Among interviewed clinicians, framing was the strongest bias revealed. “Clinicians most frequently described the SNF as a place where patients could recover function (a positive frame), explaining risks (eg, rehospitalization) associated with alternative postacute care options besides the SNF in great detail,” the study reported. Word choice was also a component of framing, with the use of “rehab” instead of “nursing home” to avoid negative connotations with the latter.

To a lesser extent, the study also found examples of default/status quo bias and anchoring bias in decision making about SNFs. Patients and caregivers tended to favor SNFs with which they had previous experience, and they would anchor on certain issues to make the decision, such as the cost of care or a facility characteristic, such as the presence of cats.

Cognitive biases do not always have a negative impact, noted the study authors. “For example, clinicians may be trying to act in the best interests of the patient when framing the decision in terms of regaining function and averting loss of safety and independence,” they wrote. However, based on current evidence, it's not certain what postdischarge setting is best for many patients. Information about cognitive biases “should be integrated into the development of decision aids to facilitate informed, shared, and high-quality decision-making that incorporates patient values,” the authors said.

In the meantime, “these findings may serve as a call to discharging clinicians to modulate language and ‘debias' their conversations with patients about care after discharge,” they wrote. “As postacute care undergoes a transformation spurred by payment reforms, it is more important than ever to ensure that patients understand their choices at hospital discharge and can make a high-quality decision consistent with their goals.”

Admission lab values used to develop model to predict in-hospital mortality

Admission laboratory values can predict in-hospital mortality and could be used as a marker of severity of illness, a recent study found.

The retrospective cohort study used data from 38,367 adult non-trauma patients admitted to the University of Maryland Medical Center between November 2015 and November 2017. Results of laboratory tests—hemoglobin, platelet count, white blood cell count, urea nitrogen, creatinine, glucose, sodium, potassium, and total bicarbonate (HCO3)—ordered within 24 hours of hospital admission were compared to in-hospital mortality, and a multivariable logistic regression model was constructed.

The study found that patient age and all laboratory tests predicted mortality with good discrimination (c=0.79). Patients with abnormal HCO3 levels or leukocyte counts at admission were twice as likely to die during their hospital stay as patients with normal results. The model showed good calibration and fit for predicting mortality (Hosmer-Lemeshow test=13.9; P=0.18). Results were published by the Journal of General Internal Medicine on Aug. 20, 2019.

The findings show that admission laboratory tests could be used to categorize patients by severity of illness, potentially in place of diagnosis-related groups, according to the study authors. “An admission laboratory–based method is more accessible, user-friendly, and more objective. Additionally, as it is not diagnosis-specific, it does not depend on the accuracy of ICD-10 coding in hospitals, which continues to pose challenges,” they said.

Although other research has investigated the predictive value of laboratory tests in certain inpatients, this study is the first the authors know of to predict hospital-wide severity of illness using admission labs, they said. It should be validated in other hospitals, and limitations include that in-hospital death was fairly rare among the studied patients, which affected the model's calibration.

“In conclusion, our findings suggest admission laboratory tests can predict severity of illness in a simple, widely available, and diagnosis non-specific manner assisting both clinicians with patient prognosis and healthcare researchers with adjusting for severity of illness in clinical studies,” the authors wrote.

Intensifying hypertension regimen at discharge associated with readmissions, adverse events

Intensifying the medication regimens of older patients with hypertension at hospital discharge was associated with worse outcomes, a recent study found.

The retrospective cohort study included 4,056 patients ages 65 years and older with hypertension who were hospitalized in Veterans Health Administration national health system facilities from Jan. 1, 2011, to Dec. 31, 2013, for common noncardiac conditions (mean age, 77 years; 97.7% men). In half of the patients, antihypertensive medications had been intensified at discharge; they were propensity matched with an equal number of patients who had not received intensification. Results were published by JAMA Internal Medicine on Aug. 19, 2019.

The patients receiving intensification had higher risk of readmission (hazard ratio [HR], 1.23 [95% CI, 1.07 to 1.42]; number needed to harm [NNH], 27 [95% CI, 16 to 76]) and serious adverse events (HR, 1.41 [95% CI, 1.06 to 1.88]; NNH, 63 [95% CI, 34 to 370]) within 30 days of discharge. At one year, the groups had similar rates of cardiovascular events (HR, 1.18 [95% CI, 0.99 to 1.40]) and no significant difference in systolic blood pressure (mean, 134.7 mm Hg vs. 134.4 mm Hg; difference-in-differences estimate, 0.6 mm Hg [95% CI, −2.4 to 3.7 mm Hg]).

“These results suggest that intensification of outpatient antihypertensive regimens during hospitalization for noncardiac conditions should generally be avoided in older adults,” the study authors said. They noted that in exploratory subgroup analyses, medication-related adverse events and readmissions were increased with intensification only in the patients who had well-controlled blood pressure before admission, suggesting that this practice should be particularly discouraged in these patients. However, even for patients with elevated blood pressure at baseline, intensification at discharge may be unnecessarily risky, the authors added.

The lack of difference between groups in systolic blood pressure at one year suggests that either outpatient clinicians or patients themselves were reversing the intensification of therapy at some point, the authors speculated. “Shifting practice from intensifying antihypertensive regimens during hospitalization to communicating concerns about patients' long-term [blood pressure] control to outpatient practitioners for close follow-up may provide a safer treatment path for patients,” they said.

The study was limited by the risk of confounding inherent to observational studies, the authors noted. It could be that unmeasured patient differences were associated with medication intensification or that the physicians who intensify antihypertensive medications differ from those who don't in other ways. In addition, because the study was conducted in the Veterans Health Administration health system, it included more men and patients with more comorbidities than the general U.S. population.