By William Sullivan, DO, JD; Catherine A. Marco, MD; and Robert C. Solomon, MD, FACEP

ACEP members can request a review of questionable expert witness testimony regarding emergency medicine’s standards of care. Procedures for this review are found here.

Question:  Is this expert testimony valid: ordering blood cultures necessitates hospital admission and antibiotic administration?

The patient’s emergency department records were not provided for this review, so the patient’s symptoms, physical examination, lab testing, and emergency department treatment were obtained solely from the expert witness deposition.

According to the information provided, the patient was a 53-year-old male with a past medical history including hepatitis C and chronic anemia, who presented to the emergency department with weakness, fevers, vomiting, loose dark stools, and a rash. There were also notations that the patient had experienced “flulike symptoms” for almost a month. The patient had been prescribed a course of ciprofloxacin for a urinary tract infection, but discontinued it approximately 1-2 weeks prior to his emergency department visit due to appearance of a rash. During his evaluation in the emergency department, he was persistently tachycardic with a heart rate of 120. He was given IV fluids and multiple tests were performed, including blood cultures. Lab results showed that the patient was hyponatremic, had a stable hematocrit of 30, and a positive nasal swab for influenza. He had a normal WBC count with no left shift, and his lactate level, urinalysis, and chest x-ray were also normal. The following day, preliminary results of the blood cultures were positive for gram-positive cocci. A message was left on the patient’s voicemail that day instructing him to return to the hospital. He did not return until three days later.  He died from septic shock, DIC, and multisystem organ failure shortly thereafter.

The expert witness faulted the treating emergency physician for several issues. This review addresses the expert’s repeated assertions that because blood cultures were performed, the patient should have been admitted to the hospital and treated with intravenous antibiotics due to a suspicion of bacteremia.

Excerpts from the expert’s deposition testimony include the following:
“One would not order blood cultures and discharge a patient home with a suspicion for bacteremia,” although at the same time noting that bacteremia “sometimes resolves spontaneously.”

“If blood cultures are ordered, that means that bacteremia in the bloodstream is suspected. There is no test to prove that it exists immediately. So unless there is a reason to suspect that someone could have occult bacteremia, like the conditions I mentioned, the treatment is admission and intravenous antibiotics. Otherwise this happens [referring to the patient’s death from sepsis]. You don’t send otherwise relatively immunocompetent patients home with bacteremia. You treat them.”

When questioned by the defense attorney about whether the expert was assuming that because blood cultures were ordered, the treating physician must have suspected bacteremia, the expert replied:
“I think doctors order things that have a low probability of being positive for number of reasons depending upon the disease. But in the case of blood cultures for bacteremia, it’s the only reason you do blood cultures. And if you suspect bacteremia in a patient who is immunocompetent, the treatment is antibiotics.”

When asked if the facilities at which he works discharge patients with pending blood cultures, the expert stated:
“Under certain circumstances, we do discharge patients with pending blood cultures; infants occasionally who are suspected of having occult bacteremia; patients with HIV known to have low CD4 counts; patients who have febrile illness who are dialysis patients; and on certain occasions, certain types of immunocompetent cancer patients will have blood culture sent from the emergency department at discharge. Other than those patients, we do not do blood cultures on patients are discharged because the only reason to do blood cultures is to suspect bacteremia, and bacteremia requires intravenous antibiotics.”

After reviewing the expert’s testimony and available literature on this topic, the Standard of Care Review Panel concluded that the expert witness presented opinions that did not represent the standard of care for several reasons.

1. Blood Cultures

A.     The Review Panel was not able to find any studies providing definitive guidelines regarding when blood cultures should be ordered. In fact, one study specifically noted that “published guidelines do not clearly state when blood culture should be drawn.” (http://jacobiem.org/wp-content/uploads/2012/12/Blood-Cultures-Bacteremia-JAMA-2012.pdf. Accessed 6/3/16).  The study also noted that there is only a 2% pretest probability of bacteremia in febrile ambulatory outpatients and that isolated tachycardia was not helpful in improving the accuracy of diagnosing bacteremia.
B.     The Review Panel also found literature noting that 3.8% of blood cultures obtained in patients discharged from the emergency department return positive and that 86% of those cultures were positive for pathogens which would typically necessitate inpatient antibiotic therapy. Only 50% of the patients in this study responded to a call for a return visit (Patel G, Kohlhoff S, Tejani N, Sharma R. Evaluation of pediatric blood culture ordering practice in a post-pneumococcal vaccination era in emergency department. Accessed June 3, 2016.)
C.      With regards to the expert’s assertion that obtaining blood cultures mandates admission and antibiotic treatment, the Review Panel did not find any literature recommending such management and unanimously agreed that the statement did not reflect the standard of care in emergency medicine. As the expert noted, and as many sources confirm, most occult bacteremia resolves spontaneously without treatment. In addition, studies show that many blood cultures are falsely positive due to skin contaminants.
D.      Using the above data, mandatory admission and antibiotic treatment after obtaining blood cultures in febrile ambulatory outpatients would result in a 50-fold increase in hospital admissions for this patient demographic.
E.      Finally, the expert’s own testimony contradicted his statements about the standard of care. He noted that the facilities at which he works do discharge patients with pending blood cultures under some circumstances. The Review Panel noted that types of patients the expert inferred could be discharged pending blood cultures were at higher risk for bacteremia and sepsis than the general population since they included patients more likely to be immunocompromised such as those with HIV, cancer, and on hemodialysis.
                          

2. Risk of Bacteremia

The Review Panel was able to find studies showing that the initial 2 day mortality was 4.8% in patients with community acquired bacteremia compared to 2.0% in culture-negative patients (MRR 1.9%).  After the first 2 days, mortality rates for both groups were essentially equal.  The 30 day mortality of community acquired bacteremia is approximately 10% with no significant differences between patients having bacteremia and patient having negative culture results. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3128048/). However, these data include any patients who had blood cultures performed within two days of hospital admission, so the results do not necessarily apply to emergency department patients as presented in this case. This study also noted that as many as half of the cultures that were positive represent contaminants—organisms inoculated from the skin into culture bottles at the time of sample collection which do not represent bacteremia. These false-positive blood culture results can lead to unnecessary investigations and treatments.

3. Retrospective Bias

The Standard of Care Review Panel believed that the expert’s opinions in this case were influenced by retrospective bias. The expert repeatedly stated that the treating physician should have known that the patient was bacteremic because the culture results returned positive. However, the preliminary culture results did not return until the following day. There is no way that the treating physician could have known the culture results at the time of treatment. It was the consensus of the Standard of Care Review Panel that strict prospective analysis is of utmost importance when reviewing the management of any patient care.

Conclusion

It was the consensus opinion of the Review Panel that obtaining blood cultures does not mandate antibiotic treatment nor hospital admission.

This case was published in ACEP Now in May, 2017