Introduction
Blood cultures (BCs) are the gold standard method
for the diagnosis of bacteremias. A BC isolate is considered to be
a contaminant when it is not pathogenic for a patient and is
introduced into the culture of the patient during either blood
collection or processing (1,2). Among others, contaminated BCs lead to
unnecessary administration of antibiotics, prolonged hospital stay,
unnecessary removal of central intravenous lines, increased work
load and cost.
In the present study we aimed to evaluate the effect
in the laboratory work load by stopping the processing
(identification, susceptibility testing) of possible contaminant
isolates by the application of a certain exclusion criterion.
Materials and methods
All BCs sent to our Laboratory from January 1, 2016
to December 31, 2017 were analyzed. All positive BCs were processed
according to standard protocols and identification and
susceptibility testing of the isolates were performed with Vitek2
(BioMerieux, France). Only the first isolate per species and
patient was included in the study. Since the majority of the
samples from the Orthopedic Department were pusses or non sterile
body fluids inoculated in BC vials, all isolates from this
particular Department were excluded from the analysis. As possible
contaminants (from now on referred to as contaminants) were
considered skin flora species including coagulase-negative
Staphylococci (CNS), Corynebacterium species (sp.),
Micrococcus sp., Bacillus sp. and
Propionibacterium acnes. All viridians group
Streptococci were always fully processed.
The two-year time period was divided into two parts:
period A from January 1, 2016 to September 30, 2016 (first 9
months) and part B from October 1, 2016 to December 31, 2017 (last
15 months). During the first period all isolates including all
contaminants were processed (identification/susceptibility testing)
whereas in the second period among contaminants only those meeting
a selection criterion were processed. This selection criterion was
the presence of the same contaminant in more than one BC pairs
drawn from the patient within the same day. After discussion and
agreement with the clinicians a series of Departments were excluded
from the application of the criterion including all Pediatric
Wards, all Wards handling cancer patients along with the Nephrology
and Neurology Departments. Moreover, in all cases of positive BCs
with more than one isolate, including contaminant and
non-contaminant isolates, all contaminants were fully processed and
included in the analysis.
A series of indices (INs) were calculated. In
particular: i) the Working Rate IN (WR) defined as the total
isolates divided to the total number of BCs submitted per month;
ii) the Contaminant Rate IN (CR) defined as the total number of
contaminants processed divided to the total number of BCs submitted
per month; iii) the CNS Rate (CNSR) defined as the total number of
CNS processed divided to the total number of BCs submitted per
month; and iv) the C/I Ratio defined as the total number of
contaminants processed per month divided to the total number of
isolates processed per month.
Results
Over the whole period of the two years a total of
31,579 bottles were admitted and the total isolates were 2,199
(total WR: 6.96%). The total contaminants of the period were 1,028
(total CR: 3.26%) including 941 CNS (total CNSR: 2.98%), 23
Bacillus sp., 33 Corynebacterium sp., 19
Micrococcus sp., and 12 Propionibacterium acnes. CNS
comprised 42.79% of the total 2,199 isolates, whereas
Bacillus sp. comprised 1.05%, Corynebacterium sp.
1.5%, Micrococcus sp. 0.86%, and Propionibacterium
acnes 0.55% of the total isolates. The overall C/I ratio of the
period was 46.75%.
The species distribution, the number of bottles
admitted, the number of isolates and contaminants processed along
with the WR, the CR, the CNSR and the C/I Ratio per month and as
total for period A and period B are shown in Tables I and II, respectively.
 | Table IResults of period A (January 2016 -
September 2016). |
Table I
Results of period A (January 2016 -
September 2016).
| Variables | January 2016 | February 2016 | March 2016 | April 2016 | May 2016 | June 2016 | July 2016 | August 2016 | September 2016 | Total Per A |
|---|
| CNS | 45 | 51 | 30 | 30 | 44 | 50 | 38 | 48 | 49 | 385 |
| Bacillus
sp. | 0 | 0 | 2 | 2 | 1 | 2 | 0 | 1 | 1 | 9 |
| Coryne
sp. | 0 | 0 | 1 | 2 | 1 | 3 | 2 | 3 | 3 | 15 |
| Micrococcus
sp. | 1 | 1 | 2 | 3 | 0 | 1 | 1 | 0 | 0 | 9 |
| Propionibacterium
acnes | 0 | 1 | 0 | 1 | 0 | 4 | 3 | 1 | 2 | 12 |
| Total
contaminants | 46 | 53 | 35 | 38 | 46 | 60 | 44 | 53 | 55 | 430 |
| Total bottles | 1,378 | 1,292 | 1,075 | 1,230 | 1,183 | 1,182 | 1,124 | 1,267 | 1,249 | 10,980 |
| Total isolates | 80 | 94 | 62 | 85 | 82 | 101 | 81 | 109 | 95 | 789 |
| Working rate (%) | 5.81 | 7.28 | 5.77 | 6.91 | 6.93 | 8.54 | 7.21 | 8.60 | 7.61 | 7.19 |
| Contaminant rate
(%) | 3.34 | 4.10 | 3.26 | 3.09 | 3.89 | 5.08 | 3.91 | 4.18 | 4.40 | 3.92 |
| CNS rate (%) | 3.27 | 3.95 | 2.79 | 2.44 | 3.72 | 4.23 | 3.38 | 3.79 | 3.92 | 3.51 |
| C/I ratio (%) | 57.50 | 56.38 | 56.45 | 44.71 | 56.10 | 59.41 | 54.32 | 48.62 | 57.89 | 54.50 |
| Table IIResults of period B (October
2016-December 2017). |
Table II
Results of period B (October
2016-December 2017).
| Variables | October 2016 | November 2016 | December 2016 | January 2017 | February 2017 | March 2017 | April 2017 | May 2017 | June 2017 | July 2017 | August 2017 | September 2017 | October 2017 | November 2017 | December 2017 | Total Per B |
|---|
| CNS | 33 | 32 | 22 | 32 | 39 | 38 | 45 | 45 | 43 | 43 | 38 | 38 | 40 | 36 | 32 | 556 |
| Bacillus
sp. | 0 | 1 | 0 | 0 | 1 | 2 | 0 | 1 | 0 | 3 | 1 | 0 | 3 | 2 | 0 | 14 |
| Coryne
sp. | 1 | 0 | 0 | 2 | 1 | 0 | 2 | 0 | 0 | 0 | 1 | 0 | 3 | 1 | 7 | 18 |
| Micrococcus
sp. | 1 | 1 | 0 | 1 | 1 | 1 | 1 | 2 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 10 |
| Propionibacterium
acnes | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Total
contaminants | 35 | 34 | 22 | 35 | 42 | 41 | 48 | 48 | 43 | 46 | 40 | 38 | 46 | 40 | 40 | 598 |
| Total bottles | 1,180 | 1,185 | 1,175 | 1,609 | 1,411 | 1,488 | 1,533 | 1,452 | 1,295 | 1,344 | 1,419 | 1,305 | 1,413 | 1,341 | 1,449 | 20,599 |
| Total isolates | 77 | 75 | 70 | 86 | 84 | 101 | 108 | 110 | 108 | 103 | 106 | 90 | 104 | 97 | 91 | 1,410 |
| Working rate
(%) | 6.53 | 6.33 | 5.96 | 5.34 | 5.95 | 6.79 | 7.05 | 7.58 | 8.34 | 7.66 | 7.47 | 6.90 | 7.36 | 7.23 | 6.28 | 6.84 |
| Contaminant rate
(%) | 2.97 | 2.87 | 1.87 | 2.18 | 2.98 | 2.76 | 3.13 | 3.31 | 3.32 | 3.42 | 2.82 | 2.91 | 3.26 | 2.98 | 2.76 | 2.90 |
| CNS rate (%) | 2.80 | 2.70 | 1.87 | 1.99 | 2.76 | 2.55 | 2.94 | 3.10 | 3.32 | 3.20 | 2.68 | 2.91 | 2.83 | 2.68 | 2.21 | 2.70 |
| C/I ratio (%) | 45.45 | 45.33 | 31.43 | 40.70 | 50.00 | 40.59 | 44.44 | 43.64 | 39.81 | 44.66 | 37.74 | 42.22 | 44.23 | 41.24 | 43.96 | 42.41 |
Discussion
BCs remain the major tool for the diagnosis of
bacteremias. Bacteremias are a significant cause of morbidity and
mortality. Mortality can reach 37% as previously reported (3). BC contaminations are not uncommon.
Despite the standard recommendation for 3% BC contamination rate
usually the reports are above this level and vary significantly
even reaching 12% (4,5). The high fluctuation in BC contamination
rate reported is partially due to the lack of uniformity since some
authors decide themselves whereas others rely on the contributing
microbiologist's decision on the contaminant status of an isolate
(4). Despite this, the contamination
rate is a significant indication of the quality provided by the
hospital health-care. In the present study the overall
contamination rate over the two-year period was found to be 3.26%,
which is higher from the analogous indice of 2.7% previously
reported from Greece (6). The
overall CNS rate was 2.98%. CNS comprised 42.79% of total isolates,
which is a slightly higher percentage than that found from a
previous study in Greece (35.9%) (7).
Apart from the major clinical impact of BC
contaminations and the relevant alteration in the clinical
treatment algorithms, the overall additional cost burden due to BC
contaminations is very high and is estimated as a median of
$8,720(8). Among others this cost is
due to increased length of stay (estimated as a median 4.5-day
increase in North America), increased microbiology laboratory
charges and increased antibiotic use (4,5). In the
first nine months of the present study all isolates including all
possible contaminants were fully processed thus allowing the
clinicians to decide themselves whether the presence of a possible
contaminant in the blood was of clinical importance or not. Over
the last fifteen months of the present study and in order to reduce
the microbiology laboratory cost possible contaminants were only
processed if they met the criterion of their presence in a second
BC set of the same patient within the same day. The implementation
of this criterion was easy, costless and most of all safe for the
patients since the clinicians were informed immediately and their
consensus was required.
In the present study, when comparing the two periods
a 26.02% reduction in the CR was found (from 3.92% in period A to
2.90% in period B). An analogous percentage was recorded in the
reduction of the CNSR (23.08%; from 3.51% in period A to 2.70% in
period B). The overall WR was reduced from 7.19% in period A to
6.84% in period B and furthermore, the C/I Ratio was reduced from
54.50% in period A to 42.41% in period B. The reduction in all INs
noted is of great value since it was achieved by the implementation
of a simple, easily applicable criterion and without any cost. If
more wards of the Hospital were to be included in the application
of the criterion an even higher reduction would be expected and
such an application is strongly advocated.
Acknowledgements
Not applicable.
Funding
No funding was received.
Availability of data and materials
Not applicable.
Authors' contributions
IKN and DS carried out the experiments. DAS, IKN and
DS wrote the manuscript with input from each author, and approved
the final manuscript.
Ethics approval and consent to
participate
Not applicable.
Patient consent for publication
Not applicable.
Competing interests
DAS is the Editor-in-Chief for the journal, but had
no personal involvement in the reviewing process, or any influence
in terms of adjudicating on the final decision, for this article.
The other authors declare that they have no competing
interests.
References
|
1
|
Hall KK and Lyman JA: Updated review of
blood culture contamination. Clin Microbiol Rev. 19:788–802.
2006.PubMed/NCBI View Article : Google Scholar
|
|
2
|
Altindis M, Koroglu M, Demiray T, Dal T,
Ozdemir M, Sengil AZ, Atasoy AR, Doğan M, Cicek AC, Ece G, et al: A
multicenter evaluation of blood culture practices, contamination
rates, and the distribution of causative bacteria. Jundishapur J
Microbiol. 9(e29766)2016.PubMed/NCBI View Article : Google Scholar
|
|
3
|
Long B and Koyfman A: Best clinical
practice: Blood culture utility in the emergency department. J
Emerg Med. 51:529–539. 2016.PubMed/NCBI View Article : Google Scholar
|
|
4
|
Hughes JA, Cabilan CJ, Williams J, Ray M
and Coyer F: The effectiveness of interventions to reduce
peripheral blood culture contamination in acute care: A systematic
review protocol. Syst Rev. 7(216)2018.PubMed/NCBI View Article : Google Scholar
|
|
5
|
El Feghaly RE, Chatterjee J, Dowdy K,
Stempak LM, Morgan S, Needham W, Prystupa K and Kennedy M: A
quality improvement initiative: Reducing blood culture
contamination in a Children's Hospital. Pediatrics.
142(e20180244)2018.PubMed/NCBI View Article : Google Scholar
|
|
6
|
Hadziyannis AS, Stephanou I, Dimarogona K,
Pantazatou A, Fourkas D, Filiagouridis D and Avlami A: Blood
culture results during the period 1995-2002 in a Greek tertiary
care hospital. Clin Microbiol Infect. 10:667–670. 2004.PubMed/NCBI View Article : Google Scholar
|
|
7
|
Falagas ME, Bakossi A, Pappas VD, Holevas
PV, Bouras A and Stamata E: Secular trends of blood isolates in
patients from a rural area population hospitalized in a tertiary
center in a small city in Greece. BMC Microbiol.
6(41)2006.PubMed/NCBI View Article : Google Scholar
|
|
8
|
Gander RM, Byrd L, DeCrescenzo M, Hirany
S, Bowen M and Baughman J: Impact of blood cultures drawn by
phlebotomy on contamination rates and health care costs in a
hospital emergency department. J Clin Microbiol. 47:1021–1024.
2009.PubMed/NCBI View Article : Google Scholar
|
