Is that test necessary?: blood cultures in pneumonia

In response to a question in the American Society of Microbiology laboratory discussion group (DivC), I sent this message:

In fact, as Dr. Manasse suspects, blood cultures are the subject of the Joint Commission Core Measure PN-3b.

However, careful examination of the measure does not justify the use of blood culture tests in all Emergency Department patients with pneumonia.

Core Measure PN-3b "Blood Cultures Performed in the Emergency Department Prior to Initial Antibiotic Received in Hospital" is described at:
http://manual.jointcommission.org/bin/view/Manual/MIF0006

As you can see there, the measure is justified by "Published pneumonia treatment guidelines recommend performance of blood cultures for all inpatients to optimize therapy" and lists several published guidelines. The most recent citation is:

Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin. Infect. Dis. 2007;44 Suppl 2:S27-72
(http://www.guideline.gov/summary/summary.aspx?ss=15&doc_id=10560)

Mandell et al. write that "Recommendations for diagnostic testing remain controversial. The overall low yield and infrequent positive impact on clinical care argue against the routine use of common tests, such as blood and sputum cultures", however they add that "cultures may have a major impact on the care of an individual patient" and identify the individual patients for whom "pretreatment blood samples for culture...should be obtained" as those with:

• Intensive care unit admission
• Cavitary infiltrates
• Leukopenia
• Active alcohol abuse
• Chronic severe liver disease
• Asplenia (anatomic or functional)
• Positive pneumococcal Urinary Antigen Test result
• Pleural effusion

(Table 5 and Recommendation # 12)

Severe pneumonia that justifies blood culture testing, in addition to the criteria above, also includes patients with three or more of the following "minor" criteria:

• Respiratory rate >=30 breaths/min
• PaO2/FiO2 ratio =<250 li="">
• Multilobar infiltrates
• Confusion/disorientation
• Uremia (BUN level, >=20 mg/dL)
• Hypothermia (core temperature, <36>
• Thrombocytopenia (platelet count, <100>
• Hypotension requiring aggressive fluid resuscitation
• Hypoglycemia (in nondiabetic patients)
• Hyponatremia
• Unexplained metabolic acidosis or elevated lactate level

For the rest of patients with less severe pneumonia, recommendation #11 applies: "Routine diagnostic tests to identify an etiologic diagnosis are optional for outpatients with Community Acquired Pneumonia".

Thus, Denise Jones' Emergency Department (ED) physicians are only 33.3% correct in requesting blood cultures in patients with low oxygen saturation, since hypoxia should be accompanied by two other minor criteria to justify the test, according to the authorities cited by the Joint Commission.

The criteria listed above require a basic patient assessment that can be completed quickly in the ED (problem-oriented patient history, measurement of blood pressure, temperature and respiration rate, chest x-ray, CBC, urine antigen tests and serum metabolic profile plus alcohol and lactic acid levels) so that they will only slightly delay administration of antibiotics.

For these reasons, and armed with the information above, the laboratory could argue convincingly against the need to obtain a sample before deciding whether a blood culture will be needed.

In addition to the unnecessary work imposed on usually overwhelmed ED and laboratory staffs, the invasive nature of the procedure and the cost of supplies and equipment, there is the problem of false negatives: in samples kept at room temperature, some organisms might grow to exceed the CO2 detection threshold so that their presence will not be detected by the continuous CO2 monitoring system when they are eventually placed in the blood culture incubator. BioMérieux sent a BacT/Alert system notification to that effect a couple of years ago.

False positive results could also be more frequent under these circumstances. I have noticed that often, blood culture contamination rates are higher in the ED than in in-patient services (is that also your experience?) and I wonder whether sterility precautions are observed as scrupulously when the person drawing the sample is not sure that it will, in fact, be used.

We are all compelled to avoid waste and error in health care, especially now that the prospect of real reform is on the horizon. Laboratory efforts to fine-tune accreditation demands so that they make sense, are a valuable contribution. Denise: go ahead and fight that battle using the information above.

In preparation for the Fall

As the number of confirmed cases seems to be decreasing (as shown by the May 4th CDC figure) and severity appears not to exceed that of seasonal influenza, public attention to the 2009 H1N1 flu wanes. Laboratorian's attention should not: as mentioned earlier, parallels with the 1918 pandemic predict worse outbreaks in the Fall. In my opinion, the following laboratory diagnostic tools should be in place in the U.S. by then: (1) a sensitive molecular test for virus detection and (2) a specific serum/plasma antibody test for retroactive diagnosis.

The CDC developed the amplified RNA test in an amazingly short time. However, it has been limited to local Public Health laboratories. Despite those laboratories' heroic efforts, this limitation precludes its widespread use. The first test offered outside Public Health labs, was only made available by Quest Diagnostics two days ago. However, this test is only offered in one location in the U.S. and still requires confirmation by local Public Health laboratories. Further, important details of sample collection and estimates for the clinical sensitivity of the molecular test are not yet available.

I was personally involved in two potential school events which begged a rapid sensitive test. I described the first experience below (another teacher in that school's district has since been confirmed as suffering from the swine flu). Last Friday, I was asked advice in a third-grade incidence, where six out of 30 students in a single classroom in a school in New York City fell ill with suggestive symptoms. On Monday half of the students in the class, including five of the six, were absent. The symptoms (fever, malaise, headache, muscle aches) of the sixth, a nine year old girl whose pediatrician prescribed Tamiflu (oseltamivir) on Saturday, resolved within hours of starting treatment. Today, a week later, most of the sick children have returned, and the whole class is visiting the local Zoo. If the incident was, in fact, due to the flu, this group of students has disseminated the virus widely. Had one of the sick students been tested within the first day and found to host the H1N1 virus, the school might have decided to take containing measures, including possible closure, despite the CDC advise of not to close "unless there is a magnitude of faculty or student absenteeism that interferes with the school's ... ability to function."

While we should no longer worry that the children who fell ill last week could spread the virus, others in the incubation phase of this potential H1N1 outbreak, might. In this scenario, the finding of specific antibodies in the blood of convalescent students, could help with containment efforts and in the diagnosis of contacts. Such antibody test would also be used as an important epidemiological tool.

Method and sample collection for the '09 H1N1 confirmation test

Dan Jernigan from the CDC is quoted by GenomeWeb, as informing that the CDC's rRT-PCR Swine Flu Panel has been distributed to 78 qualified labs throughout all 50 states in the US, as well as labs in 130 countries, and 10 Department of Defense labs in the US. Because of the shortage of reagents, the CDC is working with other vendors to get new molecular assays cleared.

The current CDC rRT-PCR test uses TaqMan probes designed by the CDC and is run on Applied Biosystems' 7500 Fast Dx Real-Time PCR instrument (the brochure is a 31.4Mb download). Last October, this instrument/CDC rRT-PCR Flu Panel combination received 510(k) clearance for sub-typing of influenza viruses in qualified laboratories.

The "flocked swab" recommended for sample collection is explained in Copan's web site, which also offers a Twit, a set of video instructions and an animation explaining nasopharingeal sample collection.

Dr. E. J. Baron, Director of the Clinical Microbiology and Virology Laboratories at Stanford U. Med. Center, informed that these Copan swabs are available from Copan (Cat # 503CS01), Becton Dickinson (Cat # 220252) and Diagnostic Hybrids (which offers several configurations).

More links

New web-based resources addressing the 2009 H1N1 virus pandemic are available:

PubMed added useful links to its homepage, including one to publications recently added and another to GeneBank's H1N1 sequences. An FDA page devoted to testing reiterates that "no influenza tests are sufficiently sensitive to rule out an influenza infection, and clinicians must consider clinical and epidemiological criteria along with laboratory testing results."

The CDC's page devoted to clinical laboratory issues, lists links to other pages in the CDC site as well as a link to the Association of Public Health Laboratories' homepage, from where you can access their H1N1 page. Three links to pages in the WHO site ("Protocol for Antiviral Susceptibility Testing by PyrosequencingExternal Web Site Policy", "Sequencing Primers and Protocol" and "CDC Protocol of Realtime RTPCR for Swine Influenza A(H1N1)"), refer to sequencing and molecular testing.

The U.S. Government Pandemic Flu site now addresses the 2009 H1N1 virus.

Rapid influenza antigen testing recommendations

On Friday, the CDC posted recommendations for the use of rapid influenza testing in the assessment of patients suspected of having the SOIV H1N1 virus.

The Festschrift Symposium in Honor of Peter Jatlow, MD takes place in New Haven, CT. today. Dr. Jatlow, former Chairman of the Department of Laboratory Medicine at YaleUniversity, Director of Laboratories at Yale-New Haven Hospital, and an authority on the clinical chemistry of addiction, presided over the training of numerous promotions of clinical pathologists, many of whom will be attending.

A bird's eye view

This week's The Economist's insightful editorial on the pandemic follows the magazine's birds-eye view style. It correctly points to the next Winter as the main worry. That is, next month for countries in the southern hemisphere!. It contrasts the couple of hundred confirmed deaths so far attributed to the swine flu to the 30,000 deaths caused by the seasonal flu every year in the U.S., acknowledging that the "lack of proper tests" prevents an accurate mortality count in Mexico.

Will enough of the "proper" tests be available north of the border?

The "common" flu test: the rapid antigen test, itself in short supply these days, might not be of much help. The amplified molecular test is now available in Public Health labs, but these facilities could not possibly handle the demand. The Economist's editorial starts with the saying that no battle-plan survives contact with the enemy. In the plan to deal with the upcoming pandemic somehow we forgot to recognize that the thousands of hospital and private laboratories in the U.S., if equipped with the proper technology could, in short notice, change to the new primers and detection probes to fit the emerging flu virus recombinant and, within days, be ready to make a crucial contribution to containment efforts by providing the "proper" test.

Update: the CDC just issued a page listing Internet social tools related to the swine flu epidemic. Other web resources had been listed here.

A negative rapid flu test result is worthless

The teacher of a first grade class in a school in a New York City suburb yesterday fell ill with chills, high temperature and muscle aches. The class is made up primarily by children of immigrant families, including ten from Mexico. Three of the children in the class had been out for a few days because of a disease characterized by a sore throat. The teacher was seen at a local Emergency Department (ED) and, after a rapid flu test, was told that she "did not have the swine flu".

What is wrong with this scenario?: (1) the null rule-out value of the rapid antigen test for the swine flu was not recognized, and (2) the patient epidemiological circumstances did not seem to have been addressed by the health care facility (even though they were aware of the risk to health care workers and held the patient in isolation while waiting for the test results).

The school nurses were well aware of the need to quickly rule out an outbreak, but had no idea whom to call for advice. I tried to help by calling the NY State's Swine Flu Hotline, but they referred me to a New York City physician resource, even though the school is outside the city limits. There was no information or contact links prominently listed on the CDC, State or County Health Department web sites for institutions to obtain rapid advice in situations where an outbreak is suspected.

Hopefully, this particular teacher's illness will end up being caused by something else, and the event will just be one of the numerous false alarms that will be triggered by the public concern about the disease. But the epidemic needs just one such institutional exposure incident to continue expanding in our midst. Institutions need instant clear guidance from local authorities in order to determine the need for timely and appropriate containment measures. Right now, this guidance is not easy to find.

Testing for Swine Flu

New York City (NYC) now has the largest number of confirmed cases in the U.S. The NYC Department of Health recommends testing only when people experience severe symptoms or fall sick in clusters. A negative rapid antigen, immunofluorescence or viral culture test does not rule out swine flu, according to the Centers for Disease Control and Prevention (CDC). Conversely, a suspected patient who tests positive for influenza A in a rapid antigen or immunofluorescent test, is considered a probable case of swine influenza A.

Confirmation requires a positive viral culture or a positive result of the molecular real-time polymerase chain reaction test (RT-PCR) for influenza A, B, H1 and H3 available in State Public Health Laboratories. This confirmatory test is indirect since even though the swine flu virus is type A(H1N1), the results are negative for H1. The direct confirmation test is currently only available at the CDC.

The nasal/pharyngeal swab used for the rapid flu antigen test might not be appropriate for the RT-PCR test. The later requires synthetic swab materials, viral transport media and transport at refrigerated or freezing (- 70 C) temperatures.