There are no significant differences in hemolysis, transfusion complications or survival when using Rh-negative versus Rh-positive low-titer whole blood (LTO-WB) for transfusions, according to researchers from the Red Duke Trauma Institute at Memorial Hermann-Texas Medical Center, in Houston.

In a prospective, single-center observational study performed from November 2017 to October 2019, researchers examined the safety of replacing Rh-negative with Rh-positive LTO-WB in trauma patients, regardless of recipient Rh status. A total of 637 severely injured patients were enrolled in the study. Of the group, 448 were treated with Rh-positive whole blood and 189 received Rh-negative whole blood.

Researchers monitored markers of hemolysis at three, 24 and 48 hours following admission. The only difference between the groups was total bilirubin level, which was elevated in the Rh-positive group at the 24- and 48-hour time points.

For transfusion-related complications, the incidence of transfusion-related circulatory overload was significantly higher in the Rh-negative group, while the Rh-positive patients had fewer ICU-free days and ventilator-free days. Of note, the Rh-positive group was potentially more severely injured than the Rh-negative group.

In subgroup analyses, no differences were observed in hemolysis markers when patients receiving Rh-negative LTO-WB were stratified by recipient Rh type. In patients receiving Rh-positive LTO-WB, the only significant difference was 48-hour serum potassium level. In similar analyses, no differences were seen in transfusion complications or outcomes across subgroups of patients receiving Rh-negative or Rh-positive LTO-WB.

Study limitations included lack of randomization and the observational design of the study. In addition, exclusion criteria were expanded to satisfy patient safety concerns when switching from Rh-negative to Rh-positive whole blood, which may be confounding.

“Clinicians considering a whole blood program or already providing this therapy should feel confident utilizing type O-positive whole blood instead of more limited type O-negative product. This will dramatically improve the availability of this product for hemorrhage resuscitation,” said presenting author Cameron McCoy, MD, an assistant professor of surgery at the University of Kansas Medical Center, in Kansas City.

“Next, centers must evaluate how to safely expand the use of type O-positive whole blood for all severely injured trauma patients,” Dr. McCoy said. “As we study this therapy more, I think restrictions based upon age and sex will be eliminated.”

Readmission Adds to Necrotizing Soft Tiissue Infection Disease Burden

Patients with necrotizing soft tissue infections (NSTIs) experience greater disease burden with hospital readmissions, according to the results of a new analysis.

Patients with NSTIs have a high risk for unplanned hospital readmission, but the factors that influence readmission are poorly understood. Researchers from Tulane University School of Medicine, in New Orleans, retrospectively investigated the incidence and factors associated with unscheduled readmission among patients with NSTIs.

Of the 82,738 patients diagnosed with necrotizing fasciitis, gas gangrene or Fournier’s syndrome, 30.3% were readmitted within 90 days. Of the readmitted patients, 79.8% had unscheduled readmission. The most common causes of readmission were septicemia, cellulitis, abscess and postoperative infection.

Patients with unplanned readmission were generally older, female, more likely to be diabetic, had higher Charlson Comorbidity Index score and were less likely to be obese.

In a multivariate analysis, factors associated with readmission were prolonged versus short length of stay, Medicaid versus Medicare as the primary payor, and leaving against medical advice. No demographic variables, clinical characteristics or interventions were associated with readmission.

Of the readmitted patients, 3.8% died during their hospitalization. Age older than 65 years and fragmentation of care caused by readmission to a different hospital were independently associated with mortality, while being obese favored survival.

Presenting author Eman Toraih, MD, an associate professor of medical genetics at Tulane University School of Medicine, said clinicians need to ensure that discharged patients have the resources they need so they will not require readmission. “This may mean ensuring that they do not only have a prescription for antibiotics, but also the means and resources to obtain antibiotics and take them as directed. In addition, these patients often have complicated wound care needs, and we must ensure that those needs are met after discharge.”

In the future, Dr. Toraih envisions the implementation of formalized protocols for patients being discharged after NSTIs. “These protocols may need to be directed toward patients that are at high risk of readmission, such as those that had a prolonged hospital stay. These protocols may help with access to antibiotics and wound care needs. In addition, these protocols should provide clear lines of communication to the discharging care team in case the patient has questions or problems after discharge. This may help prevent fragmentation of care in another hospital.”

Dr. Toraih added: “Our vision is to apply artificial intelligence to design mobile application or online risk assessment tools to predict the probability of readmission and complications for each patient.”

AI Risk Calculator Predicts Trauma Patient Mortality

An artificial intelligence–based calculator, the Trauma Outcomes Predictor (TOP), accurately predicts outcomes for trauma patients, researchers have reported.

“TOP is a novel, accurate, smartphone-accessible and user-friendly calculator for risk and outcome prediction in trauma patients,” said presenting author Lydia Maurer, MD, a general surgery resident at Massachusetts General Hospital, in Boston.

The calculator was developed using American College of Surgeons Trauma Quality Improvement Program (ACS TQIP) data from 2010 to 2016, and included clinical information from 747,249 patients (derivation cohort) describing patient characteristics, mechanism of injury, injury severity, injury score by body region, and clinical characteristics on presentation. Patients who died in the emergency department were excluded from model development.

Once developed, the TOP calculator was tested in a validation cohort consisting of 186,804 patients. The c-statistic for blunt mortality was 0.88, and for penetrating mortality it was 0.94.

The study authors also tested the performance of the TOP calculator across different types of complications. C-statistics for individual complications ranged from 0.69 to 0.84.

The calculator performed well across penetrating trauma patients; the c-statistics for cardiac arrest requiring CPR, severe sepsis and deep surgical site infection were 0.83, 0.82 and 0.81, respectively. In blunt trauma patients, the c-statistic was 0.84 for cardiac arrest requiring CPR and was 0.79 for acute respiratory distress syndrome.

Limitations of the study included potential biases in the source data and lack of information on clinician intervention.

“While TOP offers excellent predictive ability for mortality and several individual complications, it is not clear whether there is a point at which clinicians could intervene to favorably alter outcome,” Dr. Maurer explained. “A novel sister methodology, called Optimal Prescriptive Trees, is a machine learning–based method that promises to identify ‘intervention points,’ where clinicians can intervene to alter the patient outcome and decrease the risks of mortality and morbidity.”

In addition to identifying intervention points, the study authors are working toward integration with electronic medical record systems, Dr. Maurer said, explaining that “integration into the EMR is an important milestone, as it will allow the machine learning algorithms to learn directly from patient data through a ‘feedback loop’ and thus continue to improve.”

“TOP can be a useful adjunct to clinical judgment for counseling critically injured trauma patients and their families at the bedside,” Dr. Maurer said. “This type of AI-based technology can also be leveraged for nuanced, risk-adjusted quality benchmarking of trauma programs.”