Medical Fun Facts Podcast

MFF0064: Blood for culture

It’s Thursday 11 May 2017.

So tonight, I want to talk about one of my favourite diagnostic tests in microbiology. I want to talk about blood for culture.

When patients are critically ill with an infection, they may become bacteræmic, that is, they have bacteria in their blood. This is a little bit of a misnomer. In a normal human with a normally functioning immunological system, it’s not uncommon to find very small numbers of bacteria in the blood, even what we might regard as highly pathogenic bacteria. I can recall on two separate occasions in my training, two patients who had blood collected for culture because of protocol and not because they were necessarily very unwell who grew meningococci in their blood. Most medical practitioners get excited and pay attention when I call them and explain we’ve grown a meningococcus. On these occasions, however, the referring doctors were surprised because their patients were so well.

I’m old enough to remember in the 1980s when a read of the death notices in the morning newspaper was a reasonable predictor of positive blood cultures. They were the days when we used unsophisticated methods, the days of biphasic Castaneda bottles which contained a slant of solid agar which would be bathed twice daily on liquid phase broth. Twice a day, we would have to open an incubator and shake a tray of bottles to agitate the broth and get it to wash over the solid agar.

Times have moved on and without describing the radiometric days, we now have safe plastic bottles which remain sealed until they signal positive. We have bottles with an aerobic atmosphere, bottles with an anaerobic atmosphere and specially formulated bottles for babies and for mycobacteria.

The bottles now need to be loaded into as instrument as soon as possible after they’ve been inoculated with blood. It’s still true that blood volume is the most critical factor in blood culture microbiology. The media in the bottles relies on nutrients in the patient’s blood to complement the nutritional elements in the broth media itself. Too little blood is a big no no. The dilution effect is also important to reduce the effect of any antimicrobials the patient may have been administered.

Rather than shaking the bottles twice a day, the bottles are loaded into trays that are continuously agitated and the bottles are ‘read’ every ten minutes. In the base of each bottle is a disc which has a light shone onto it every ten minutes in the instrument. If bacteria are present in the bottle, carbon dioxide will be released which acidifies the media. The disc at the bottom of the bottle will change in colour as the pH changes and the reflected light changes in colour. Those changes are measured every ten minutes and algorithms in the instrument are designed so that at the right moment when the change indicates potential growth the instrument signals and a scientist needs to remove the bottle so that it can be sampled and processed.

Given the potential for high risk bacteria to be present, bottles are sampled with a syringe in a biological safety cabinet class II. In my opinion the best approach to processing positive bottles is to prepare a wet preparation, a smear for Gram’s stain and then inoculate various plates to get solid phase growth. There are methods available for immediate multiplex PCR against common pathogens and to also perform MALDI-TOF straight from a bottle.

A Gram’s stain and a wet preparation along with a little clinical information can go a long way to making a quick presumptive diagnosis while waiting for formal organism identification and antimicrobial susceptibility testing. For example, if we get a specimen from a patient with a high fever who complains of constipation and a recent surfing trip to Bali and in the wet preparation I see fast (but not darting) swimming bacilli, and in the Gram’s stain I see Gram-negative rods, I’d bet a body part of Salmonella Typhi.

One of the biggest problems with blood for culture is contamination with skin commensals. This is especially true for blood drawn in the emergency department and especially when the person collecting the blood does so through a cannula or a central venous line. No matter how often microbiologists explain that the preferred specimen is blood collected from a peripheral vein we still see specimens that have an inherently higher probably of contamination. The problem with contamination is that we’re often not sure when the bottle signals positive if the Gram-positive cocci in clusters or Gram-positive motile bacilli will be significant of not.

The reason why I like blood for culture is because a positive result is a critical value result and it means the need for interaction between the treating team and a specialist microbiologist. These telephone interactions provide an opportunity for sharing clinical advice, helping the patient and educating referring medical practitioners about microbiology and infectious diseases.

If you disagree with anything in these podcasts or if you would like to voice a different view, please feel free to write a comment. If I have said something incorrect I welcome correction. Please also feel free to share your comments on social media.

You can find me on Twitter and Facebook. Please follow me on Twitter and like my Facebook page.