MacDougall study finds severe sepsis alert speeds drug delivery

Sepsis occurs when the body responds to an infection with a mix of tissue-damaging inflammation and anti-inflammatory responses. This biological storm can lead to acute organ dysfunction (severe sepsis) and dropping blood pressure that does not respond to intravenous fluids (septic shock).

There are more than 750,000 cases of severe sepsis each year in the United States alone, accounting for 10 percent of all admissions to intensive care units. Severe sepsis can be lethal in 25 to 50 percent of patients.

A key to improving outcomes is diagnosing the conditions and then giving broad-spectrum antibiotics as soon as possible to treat the triggering infection. A recently published study led by UCSF School of Pharmacy researchers found that adding severe sepsis alert software to the electronic medical record system both significantly speeds the delivery of antibiotics in the emergency department (ED) and notably shortens subsequent hospital stays. The alert software:

  • analyzes inputs into a patient’s electronic medical record in real time
  • detects diagnostic criteria as they emerge
  • alerts providers to specifically assess patients for severe sepsis if key criteria are met

The paper—in the American Journal of Emergency Medicine—was senior-authored by Conan MacDougall, PharmD, MAS, a faculty member in the School’s Department of Clinical Pharmacy. Its lead author was Navaneeth Narayanan, PharmD, a former resident in the Infectious Disease specialty residency jointly offered by the School and UCSF Medical Center. Co-authors included Kendall Gross, PharmD, and Megan Pintens, PharmD, volunteer faculty in the Department of Clinical Pharmacy, as well as Christopher Fee, MD, faculty member in the UCSF School of Medicine’s Department of Emergency Medicine.

Journal citation: Narayanan N, Gross AK, Pintens M, Fee C, MacDougall C, “Effect of an electronic medical record alert for severe sepsis among ED patients,” American Journal of Emergency Medicine, October 14, 2015 doi:10.1016/j.ajem.2015.10.005 (Epub ahead of print).

The challenge

Numerous studies have found that time to drug delivery is a critical factor in effective treatment of severe sepsis and septic shock. For example:

  • A 2006 review of North American hospital medical records of more than 2,000 patients exhibiting septic shock found that patients who received antibiotics effective for the bacteria causing the infection within the first hour after the documented shock-defining low blood pressure (hypotension) had a significantly improved survival rate (about 80 percent).
  • In the same study, each hour that therapy was delayed over the ensuing six hours was associated with an average 7.6 percent decrease in survival.
  • A 2014 worldwide analysis of about 18,000 patients with severe sepsis and septic shock (from 2005 to 2010) also found a significant and steady increase in the probability of death with every hour that initial antibiotic therapy was delayed.

Thus, the latest (2012) guidelines from the Surviving Sepsis Campaign—a collaboration between leading international critical and intensive care organizations—recommend administering effective intravenous antibiotics “within the first hour of recognition of septic shock and severe sepsis.”

However, the campaign notes: “Some patients will not meet severe sepsis criteria on ED arrival. … Frequent observations for changes in vital signs will lead to early recognition and improved outcomes.”

In addition, “clinical manifestations of sepsis are highly variable,” according to a 2013 review article in the New England Journal of Medicine. They depend on the initial infection site, the bacteria involved, the pattern of acute organ dysfunction, and patient health. Indeed, sepsis frequently occurs among infants and the elderly, as well as patients already ill or with compromised immune systems. These patients may not always show the classic symptoms of fever and a high white blood cell count (leukocytosis), adding to difficulty in diagnosis.

sepsis Venn diagram

Diagram shows sepsis occurs when a blood borne infection, by bacteria or other pathogens (fungi, parasites, viruses, etc.), causes systemic inflammatory response syndrome (SIRS), which can also have non-infectious causes. Sepsis turns to severe sepsis when there is also acute organ dysfunction and to septic shock when dropping blood pressure does not respond to IV fluids.

MacDougall et al write that the value of having severe sepsis alert software in the electronic medical record is its ability “to piece together in real time multiple asynchronous variables” such as an abnormally rapid heartbeat (tachycardia) noted upon initial emergency room evaluation, a later report of leukocytosis from one lab test and a subsequent finding of elevated lactate (indicating that not enough oxygen is reaching cells, tissues, and organs).

The Severe Sepsis Best Practice Alert (SS-BPA) software “is more efficient than a provider who is dealing with multiple competing issues at any given time,” the researchers note. “The SS-BPA looks to decrease time to recognition of severe sepsis and septic shock.”

The research

Like many hospitals nationwide, in recent years the UCSF Medical Center’s emergency department has undertaken initiatives to improve severe sepsis treatment. This has included involving a diverse team of health care providers— including pharmacists—in screening and treatment, as well as undertaking intensive educational efforts. The latter includes real-time quality improvement feedback to those involved in cases that do not meet so-called bundles—diagnostic and treatment actions to be undertaken in a specific time frame—and monthly reports of bundle compliance.

MacDougall and his co-authors specifically looked at the before and after effect of adding SS-BPA software developed here to the electronic medical record for adult patients arriving at the UCSF Medical Center emergency department.

The SS-BPA primarily works by analyzing a patient’s medical record using an automated real-time set of instructions (algorithm) to detect severe sepsis and septic shock. The software alerts health providers accessing the patient’s medical record via a pop-up window in two instances:

  • When the patient meets two or more criteria for systemic inflammatory response syndrome (SIRS), such as abnormal body temperature, heart rate, breathing rate, and white blood cell count. (This syndrome differs from sepsis in that it can be caused by something other than an infection.)
  • When the patient meets two or more SIRS criteria and an additional sign is met for severe sepsis (organ dysfunction) or septic shock (fluid non-responsive hypotension).

The alert software’s pop-up initiates a protocol of both further diagnostics (e.g, drawing blood cultures to determine a specific infection) and immediate treatment—including giving broad-spectrum antibiotics within 60 minutes.

The study involved reviewing the medical records of 103 patients who came to the UCSF emergency department with severe sepsis or septic shock during a seven-month period in 2013 after the SS-BPA was implemented. Those records were compared with a control group of 111 patients meeting the same criteria over the 7-month period just before the alert software was launched in late 2012.

The results

The goal of the study was to determine if the software alert for potential severe sepsis cases improved outcomes for patients—primarily survival rates and, secondarily, reducing length of hospital stays and improving care.

While deaths were reduced after the software alert was implemented, the difference was not large enough to be statistically significant. (MacDougall and his co-authors say a larger patient group may be needed to see the effects of the intervention on mortality.)

However, the researchers found that the software alert significantly improved secondary patient outcomes and processes of care:

  • The median time from when a patient was diagnosed to the administration of antibiotics was reduced from 61.5 to 29 minutes.
  • The percentage of patients receiving antibiotics within the target 60 minutes after diagnosis increased from 48.6 percent to 76.7 percent.
  • The mean length of hospital stay for patients was reduced by one-third.

MacDougall and colleagues concede their study was unable to determine the relative impact of the software alert alone versus the provider education initiatives. However, they note that education efforts had begun during the control group period: “We believe that the SS-BPA, due to its real-time clinical use, is the primary driver of the change in clinician behavior.”


School of Pharmacy, Department of Clinical Pharmacy, PharmD Degree Program

About the School: The UCSF School of Pharmacy aims to solve the most pressing health care problems and strives to ensure that each patient receives the safest, most effective treatments. Our discoveries seed the development of novel therapies, and our researchers consistently lead the nation in NIH funding. The School’s doctor of pharmacy (PharmD) degree program, with its unique emphasis on scientific thinking, prepares students to be critical thinkers and leaders in their field.

The pharmacist’s roles in recog­niz­ing, treating severe sepsis

Q&A with faculty member Conan MacDougall, PharmD, MAS

Jacobson: To implement best practices in recognizing and treating severe sepsis, the AJEM paper cites “utilization of a multidisciplinary team including pharmacists.” What unique contributions can pharmacists make in improving recognition of severe sepsis?

MacDougall: The medications that are ordered may be an indicator of patients experiencing sepsis or those at high risk. For example, if large volumes of fluid are being administered to a patient, that might be an indicator of distributive shock [inadequate blood supply to tissues and organs despite normal output by the heart] that is associated with sepsis. This might set off a red flag, pointing out a potential sepsis patient.

Also, patients receiving immunosuppressive drugs are at higher risk of severe infection and may manifest fewer or different symptoms—for example, patients receiving some immunosuppressive agents may not display the typical response in their white blood cell count that is often seen in sepsis.

Jacobson: Since time is of the essence, how do pharmacists help with the rapid treatment of severe sepsis and septic shock?

MacDougall: I think the main role that pharmacists can play is in reducing the time from recognition of severe sepsis to receipt of antimicrobials. When severe sepsis is recognized, there are multiple next steps and decisions to be made—what antibiotic to choose? What is the dose? Which drug should be administered first? Meanwhile there are multiple other competing demands on the treating physician’s time and attention. By helping to create protocols and tailoring them to the individual patient in real time, pharmacists can offload both the cognitive and practical aspects of that work. So when the recognition is performed, and antibiotics are ordered—the rest happens automatically, thanks to the work of the pharmacist.

Jacobson: Are there risks in administering broad-spectrum antibiotics if a patient shows symptoms of severe sepsis but there is not yet a documented bacterial infection?

MacDougall: Absolutely. Sepsis is based on a known or suspected infection, and not all patients with symptoms of severe sepsis or septic shock have a bacterial infection. So these patients may receive antibacterial drugs when none at all are needed, which not only increases both their individual risks for adverse consequences such as allergic reactions or Clostridium difficile superinfection [an overgrowth infection from C. difficile after a broad-spectrum antibiotic reduces its bacterial competition] but also contributes to the overall decline in utility of antibacterials.

Like much in the area of infectious diseases, it comes down to a judgment of risks and benefits; in the case of severe sepsis and septic shock, the risks of inadequately treating the infected patients are considered to outweigh the potential harms associated with treating patients who do not require broad-spectrum antibacterial therapy.

Jacobson: During the severe sepsis protocol set in motion by the computer alert, you draw blood cultures. Why is finding the specific cause of the infection still important?

MacDougall: Without knowledge of the specific infecting pathogen, clinicians are forced to rely on their knowledge of what organisms typically cause that infectious syndrome. However, that list can be very long; moreover, it may not always be possible on initial presentation to classify the patient according to one specific infectious syndrome—in some cases a patient might have a lung infection or an central nervous system infection or a bloodstream infection—which yields a long list indeed of potential organisms, requiring broad-spectrum therapy which may require combinations of antibiotics. This increases the risks of toxicities from the different antimicrobials as well as the contribution to antibiotic resistance.

In some cases a narrower-spectrum drug may be more effective than a broader-spectrum one. Because cultures are less likely to yield the true infecting organism if they are obtained after antibiotics are administered, it is just as important to obtain the cultures as soon as possible after recognition of sepsis as it is to administer antibiotics rapidly.