Approach to Positive Blood Cultures in the Hospitalized Patient: A Review

Abstract

The use of blood cultures as a diagnostic tool has increased over the years along with improvements in techniques and results. The diagnostic dilemma arises when blood cultures are positive and there is possibility of contamination. Hence obtaining blood cultures in the appropriate setting and the interpretation of blood cultures by the hospitalist is imperative to the management of the hospitalized patient.

Introduction

Blood cultures are an extremely important tool that is employed in the diagnosis and management of the hospitalized patient. The methodology and the techniques used in the collection of the specimens has certainly evolved over the years to provide more accurate results. Blood cultures are commonly obtained in patients with concerns for an underlying infectious process, whether it is bacterial or fungal. However at times they are also obtained unnecessarily by providers that are not aware of the indications for collection. Consequently, patients are being subjected to further testing, prolonged hospitalizations, and unnecessary antibiotic treatments leading to an increased expenditure with potential for patient harm. In the event of positive blood cultures, many factors need to be taken into account to appropriately decide the next steps of management. This review article focuses on the techniques of obtaining blood cultures, the indications as well as how to interpret the positive blood cultures.

Techniques

The process of obtaining blood cultures is crucial as the yield can be negatively impacted if one does not exercise caution. Use of sterile technique is important as the chances of obtaining contaminated samples can drastically increase if not followed. As there are normal bacterial commensals on the skin, there are chances of having blood cultures become contaminated with these organisms. Obtaining blood cultures can be technically challenging in patients with morbid obesity and difficult to access veins increasing the risk for skin contamination, ESRD patients with limited access for venipuncture sites, critically ill patients with prolonged hospital stay having poor peripheral access sites. Hence, it becomes challenging to distinguish between an active infection as opposed to a false positive result. This not only poses a diagnostic dilemma but can also give rise to unnecessary testing or repetition of blood cultures which is costly. By properly cleaning the site of venipuncture¹ and not obtaining samples from peripheral intravenous lines the opportunity for contamination can be minimized.^2,3 The Centers for Disease Control and Prevention (CDC) recommends that alcohol should be used to disinfect the skin overlying the site of venipuncture and then chlorhexidine is applied after which the site should be permitted to dry.⁴ The tops of the blood culture bottles should also be sterilized with the application of 70% isopropyl alcohol.⁴ Disinfecting agents such as iodine tincture, chlorhexidine gluconate, chlorine peroxide are more effective as compared to agents with povidone-iodine.^2,3 A key factor that is most important in yielding positive results and capturing bacterial or fungal bloodstream infections is the total volume of blood sample collected from a set of blood cultures.^2,3 Approximately 20–30 ml of blood is recommended to be collected per set of blood cultures in cases of adults whereas in children age and weight are factored in when deciding the optimal volume needed.⁵ The number of sets of blood cultures is also imperative to having a good yield and in adults the recommendation is to obtain at least two sets of blood cultures and even at times up to four sets (such as in suspected endocarditis).^6,7 In each set one of the bottles should be aerobic and one should be anaerobic.⁵ Many hospitalized patients have long term intravenous access in the form of central lines, peripherally inserted central catheter (PICC), ports or dialysis access lines. In these cases it is also imperative to obtain blood cultures from these sites in addition to peripheral venipuncture sites when evaluating for bloodstream infections. For catheter-associated blood stream infections (BSI), the catheter tip should be sent to the microbiology lab in addition to a blood culture from a peripheral venipuncture site.⁸ A 5 cm of the distal portion of the catheter is recommended to be sent for roll-tip cultures .⁸ Another important aspect of methodology is the timing of collection of the blood cultures, though this is not as important as the volume of blood that is obtained.⁵ Ideally blood cultures should be obtained a few hours apart and trying to obtain them during times of febrile episodes if possible, as bacteremia may be intermittent.⁵ Given the current technology, in most cases the blood cultures can be expected to show growth in 48 hours and monitoring for any growth beyond 5 days is rarely needed.^2,3 However prolonged incubation times as well as specific media may be needed in cases of infectious agents such as mycobacteria and certain fungi.⁵ A point to be noted is that in cases of candidemia, blood cultures may not be positive in up to 50% of cases.⁵ Hence if there is strong suspicion for this and the patient is not clinically improving then one should consider adding empiric anti-fungal coverage instead of completely relying on cultures to return as positive.

As described above, the process of collection of blood cultures is just as imperative as the interpretation of the results as it can significantly impact a clinician’s decision. Therefore key factors such as sterilization techniques, volume of specimen, timing of the sample, and avoiding the use of peripheral intravenous lines for obtaining blood cultures should be taken into account for the practice of collection of blood.

Indications

Blood cultures are used to aid in the diagnosis of patients with suspected sepsis secondary to either a fungemia or bacteremia. Ideally the collection of blood cultures should precede the initiation of appropriate anti-bacterial or anti-fungal therapy. Table 1 summarizes some of the common indications for obtaining blood cultures.

Table 1.

Common indications for blood cultures:⁹

Sepsis: suspected or confirmed	Septic arthritis: native joint or prosthetic joint infection
Osteomyelitis	Meningitis
Endocarditis	Peritonitis
Pneumonia	Fever of unknown origin
Catheter associated blood stream infections	Immunocompromised patients with fever and/or leukopenia/leukocytosis
Soft tissue infections:	if severe systemic manifestations, underlying malignancy, immunocompromised state or unusual predisposing factor (immersion injury, animal bites)

The white blood cell count is notoriously known to mislead clinicians and hence the clinical context and associated patient factors need to be considered. The presence of leukocytosis does not necessarily imply infection and nor does the absence of leukocytosis exclude it.¹⁰ Hence clinical scores such as the SIRS (Systemic inflammatory response syndrome) criteria are often used to help in assessing the need for blood cultures. Another set of clinical criteria that was coined by Shapiro et al knows as the Clinical Decision Rule, looks at variables that are categorized as major and minor and can be utilized when making the decision of whether or not to obtain a blood culture.¹¹ As per the criteria, blood cultures should be obtained if there is one major or two minor criteria that are being met.¹¹ See Figure 1.

Figure 1.

Clinical Decision Rule by Shapiro et al¹¹

Hence, the clinical scenario needs to be considered along with the recognizing the pretest probability when ordering blood cultures for patients.¹² This will help to limit the number of blood cultures that are obtained and thereby lowering the associated cost of additional evaluation of the positive blood culture. The criteria developed by Shapiro et al. does not take into account leukopenia or hypothermia which the SIRS criteria does factor in. The presence of either of these in a patient with concern for an underlying infection or immunocompromised state should prompt the collection of blood cultures as well. The rule created by Shapiro et al is found to be highly sensitive but like the SIRS criteria has a low specificity.¹² As no clinical criteria is infallible, a clinician has to use their own judgement when using such scoring systems to execute a decision.

At times blood cultures should be repeated after the initiation of antimicrobial therapy to ensure that the bacteremia is cleared of infection and subsequent blood cultures are negative. The table below summarizes some of the common indications for obtaining follow-up blood cultures. See Table 2.

Table 2.

Indications for Repeating blood cultures:⁹

Endocarditis: confirmed or suspected	Suspicion of central nervous system or intra-abdominal infection
Staphylococcus aureus bacteremia	Signs and/or symptoms suggestive of infection > 72 hours after start of therapy
Poor penetration at site of infection: e.g. infection of joint space, abscess	Cases of intravascular lines, prosthetic vascular grafts, pacemakers
Infection with multi-drug resistant organisms: confirmed or suspected	Bacteremia with unidentified primary source

In cases of gram negative bacteremia there is evidence now to suggest that follow up blood cultures may not be needed. A study conducted by Canzoneri et all revealed that obtaining repeat blood cultures in the setting of gram negative bacteremia did not improve overall management and that mortality in patients admitted to intensive care was not linked to positivity on follow up blood cultures.¹³

Interpretation of Positive Blood Cultures

Once blood cultures are obtained the next step is to carefully interpret the results taking into account the organism that is growing and combining it with the clinical picture. Bacteremia can be divided into two forms and it can be either continuous or intermittent form of bacteremia. Intermittent bacteremia is more prevalent as compared to the continuous and the organism sporadically enters thebloodstream.⁹ Common scenarios in which it occurs include procedures in which mucosal surfaces such as the gingiva, urethral or rectal mucosa are involved (e.g. sigmoidoscopy, cystoscopy, oral procedures).⁹ Other cases include bacterial infections of the joints, bone, lungs, deep soft tissues, central nervous system⁹. The continuous form of bacteremia is seen on cases such as infected endovascular grafts, endocarditis, infection of blood vessels (e.g. suppurative thrombophlebitis, endarteritis, infection of aneurysm).⁹

Bacteremia can also be classified based on whether it originates from within the vascular system or external to it.⁹ Infections that originate within the vasculature are known as intravascular infections and can occur due to mycotic aneurysms, infective endocarditis, infections of endovascular grafts, other intravascular devices, or suppurative thrombophlebitis.⁹ Infections that arise in infected organs and tissues enter the blood stream through the lymphatics and include infections such as those of the lungs.⁹

It is imperative that differentiation be made between a contaminant and a true bloodstream infection and at times this can be very challenging to do so. Institutional rates of contamination should not exceed 3% and if they do, then steps should be taken to minimize the rates.¹⁴

If blood cultures become positive > 72 hours after collection, depending on the organism, then contamination should be considered.⁹ However, giving antibiotics before obtaining blood cultures can delay or prevent the growth of the bacteria and the growth of fastidious micro-organisms themselves should also be considered.⁹ Common fastidious organisms include Kingella, Eikenella, Cardiobacterium, species of Haemophilus.⁹

The following microbiological organisms should always be considered as true infections and not as mere contaminants:¹⁵

• Staphylococcus aureus

• Group A streptococci

• Streptococcus pneumoniae

• Pseudomonas aeruginosa

• Enterobacteriaceae

• Bacteroidaceae

• Haemophilus influenzae

• Candida species

In cases of Staphylococcus aureus bacteremia or candidemia, infective endocarditis needs to be excluded with the use of an echocardiogram (ideally a trans-esophageal echocardiogram). For patients with candidemia, an ophthalmological exam needs to be performed to assess for features of endophthalmitis.

On the other hand, agents such Viridans group streptococci or enterococci can be either contaminants or actual infections.⁹ For microbiological agents such as Cutibacterium/Propionibacterium acnes, Bacillus species, coagulase-negative staphylococci as well as species of Corynebacterium it becomes much more challenging to differentiate a contaminant from an infection.^16–18 Evaluation for infective endocarditis needs to be kept in mind as well for cases of patients with prosthetic cardiac valves, intra-cardiac or intra-vascular devices and positive blood cultures with these organisms.⁹

Recent decades have seen an increase in rate of contamination that has been attributed to drawing cultures through indwelling catheters, improved lab techniques to detect minuscule quantities of microorganisms, changing specimen collection techniques aimed at minimizing needlestick injuries.¹⁹ Additional cost incurred due to hospitalization, repeat blood cultures, therapy with IV antibiotics, imaging to evaluate the source not to mention emotional burden, waste of time and resources, add a significant financial burden to both the patient and the institution.

Factors that impact the possibility of a microorganism detected in blood being a contaminant include the cultured species, number of positive blood cultures obtained from different sources in a set, positive cultures in different sets, time to positivity, quantity to growth per culture sample. Positive blood cultures in a patient with clinical features of infection increases the likelihood of the blood culture being significant. Conversely, the rate of false positive blood culture increases in a patient with very low likelihood of bacteremia. Chang et al. found that in emergency departments with high volume of patients, those with critical illness, end-stage renal disease and the elderly were more likely to have blood culture contamination.²⁰ They attributed these findings to the time restriction and urgency of obtaining blood cultures in critically ill, and potential for antimicrobial resistance in the skin commensal of ESRD and elderly patients.

For bacteremia due to catheter- associated blood stream infections (BSIs), the diagnosis is usually made by exclusion of other sources of primary infection⁵. There are two major steps in the diagnosis with the first being that bacteremia has to be confirmed and then proving that the source of the bacteremia is the catheter itself.⁵

There are three main approaches used in the diagnosis of the catheter-associated BSIs:⁵

1. Time to positivity

   Blood cultures are obtained from a peripheral vein and from the catheter/port simultaneously. If an identical organism grows from both sites, with the blood culture from the catheter/port turning positive > 2 hours prior to the peripheral blood culture then this usually indicates a catheter associated infection.²¹

2. Quantitative Blood cultures

   Simultaneous blood culture are obtained from the catheter/port and one from a peripheral site. If an identical organism grows from both sites with the one obtained from that catheter/port having a five-fold increased growth in comparison to the peripheral culture then this favors a catheter associated infection.^22,23

3. Catheter tip or segment cultures

   The method developed by Maki et al involves obtaining a blood culture from peripheral venipuncture site and obtaining about a 5 cm length tip of the distal portion of the catheter after removal .⁸ However this is able to diagnose the microbial agents that are growing on the external surface of the catheter (hence producing a biofilm that does not allow for penetrance of antibiotics) and not the ones presence in the lumen itself.⁵ There are other methods that involve an endoluminal brush or a technique referred to as ‘vortexing of the catheter tip’ that may detect organisms in the lumen itself.⁵

If a catheter- associated bloodstream infection is diagnosed then ideally the intravascular device (catheter or port) needs to be removed. However in some cases, salvage with systemic and antibiotic lock therapy is used to prevent intravascular access from being removed (especially in cases of ports).

Conclusion

Blood cultures are a very useful tool in the diagnostic kit that a hospitalist carries and also one that is commonly utilized. Hence the proper use of this is crucial to the management of the hospitalized patient. From the indications to the process of collection of blood cultures, a clinician should be aware of what this entails. Once the results of positive blood cultures are obtained then one should be cautious on how to interpret them. Further steps need to be carried out accordingly and one should not overlook a true infection from an uncommon organism and nor should a contaminant receive superfluous work-up. The consequences of further blood draws, unnecessary treatment with patient harm, increased consultations, additional diagnostic tests, prolonged hospitalizations among many others can all be prevented. Therefore, blood cultures can predispose to dilemmas regarding management and hospitalists need to be cognizant of how to interpret them.