s a way of identifying the organisms in the bloodstream, blood culture is a valuable method for health care practitioners. Blood cultures are an important investigation to help effective management for patients with severe infection/sepsis. A positive blood culture may indicate a conclusive diagnosis, allowing the individual organism to be targeted for therapy. However, false-positive results because of contamination can limit the utility of this important tool 1 . Owing to contamination, which happens when species that are not naturally present in a blood sample are grown in culture, false positives arise. For decades, contaminated cultures have been described as a

Author ?: e-mail: [email protected] problematic problem and continue to be a source of irritation for both clinical and laboratory workers. Clinicians must assess if the organism represents a clinically relevant infection associated with a high risk of morbidity and mortality or a false-positive result without any clinical effects in the face of a positive blood culture outcome. Contaminated samples increase the workload of the laboratory and can interrupt patient management or cause incorrect changes. This can prolong hospitalization of patients, increase the risk of harm, and increase health boards' costs. Current guidelines advocate a contamination rate of 2-3% is acceptable 2 . Emergency departments (EDs) are important locations for the diagnosis and management of bacteraemia 3 . Blood cultures are considered the "gold standard" for the diagnosis of bacteraemia. Emergency departments are networks that are especially vulnerable to a heavy burden of infected blood cultures due to the high turnover of workers, the need to collect cultures before resuscitation of critically ill patients, and the time pressure to acquire cultures before the first dose of antibiotics 4 . This study is to compare the contamination rate and risk factors of blood culture done in the emergency department and MICU/MHDU. The mean age of population in the ED culture arm was 51 years and in the MICU/MHDU was 46 years. There is male predominance in both the arms. The total males accounted to 61% and the females accounted to 39%.

Lung infection/Lung pathology is the most common admission diagnosis encountered in the study comprising of 193 cases (19.3%). However, in ED the most common admission diagnosis was AUFI comprising of 23% of total ED cases. There are no cardiac diseases in ED. Others includes neuroleptic malignant syndrome, Diphtheria infection, G6PD deficiency, post renal transplant, nephrotic syndrome, polymyositis, Liver Abscess, cardiac pathology, acute abdomen, toxicology, autoimmune diseases. The mean value of total leucocyte counts in ED arm is higher than that of MICU/MHDU. The mean serum albumin levels were lower in MICU/MHDU arm than ED arm. Hypoalbuminemia is observed in patients with positive culture growth in MICU/MHDU. The mean Hb levels were also lower in MICU/MHDU than ED arm. The most common site of poke for culture in ED is brachial 81% followed by dorsum of hand 14%. There was no arterial line in ED. The most common site of poke for culture in MICU/MHDU is Central line (35%) followed by Arterial line (33%). There was no EJV line in MICU/MHDU. Total contamination was 48. Most common contaminant was CoNS (10 cases). NF-GNB as contaminant were found 1 in each department. Total True contaminants were reported (20 cases) out of which 14 were in ED and 6 were in MICU/MHDU. The most common site of poke for contamination in ED was from the femoral (22.2%) and the least common site of contamination was dorsum of hand (1.28%). In MICU/MHDU, the most common site of contamination is from the Brachial (6.20%) and the least common is from the arterial line (4.16%). Arterial line blood culture sample was not done in Ed. No femoral and Dorsum of hand blood culture sample was taken in MICU/MHDU.

In our observation study it was found that there was no scrubbing of hand before the procedure in ED whereas scrub hand was found in 62% cases in MICU/MHDU. The gloves were worn in unsterile manner in 9 out of 100 cases of ED and overcrowding during venipuncture was found in 35 cases. The antiseptic used in ED was chlorhexidine (20% cases) and betadine (80%). The antiseptic used in MICU/MHDU was chlorhexidine in all the cases (100%). In ED, the antiseptic was allowed to dry in 32 cases only. The set used for blood culture was blood culture set (77%) and dressing set (20%) in ED. The blood culture set was used in 51% cases of MICU/MHDU and other sets in 49% cases. In ED, the volume collected was 5cc in 35% cases and 10 cc were collected in 64 % of cases. In MICU/MHDU, 10cc volume was collected in 98% of cases and in 2% cases < 5 cc was collected. In ED 95% of cases, were collected by EMT, 17 by interns and 10 by registrars. Where as in MICU/MHDU most of the cultures were taken by registrars (89%). In MICU/MHDU, 382 by registrar and 46 by interns. The blood culture was done in first attempt in 91% of cases of ED and 92% cases of MICU/MHDU. The blood culture procedure variables done in ED were not significant with contamination (p value >0.05). The blood culture procedure variables were not significant with contamination (p value >0.05).

IV.

This was a prospective study comparing the contamination rate and risk factor profile of blood culture done in the Emergency Department and MHDU/MICUs. The analysis contained a total of 998 cases. Out of which 570 were from ED and 428 were from MICU/MHDU. This first Indian studies looking at the rates of BCC in ED and MICU/MHDU to the best of our knowledge.

The mean age in our study in ED is 51.3 years and MICU/MDHU is 46.4 years. A similar study by Choi et al had shown a mean age of 67 years in ED and 65years in general ward 5 . As life expectancy in India is less when compared to Singapore, the mean age in our study is less than the study done by Choi et al at Singapore 5 .

Our study shows a slight male predominance which is in contrast to Choi et al study where there is female predominance 5 . This might be because of the high female sex ratio (1:1.04) in Singapore when compared to India 6 .

The most common comorbidity in our study is diabetes accounting for 339(34%) of cases. Choi et al also showed diabetes as the most common comorbidity accounting for 163/400(40.8%) cases 5 . There is a positive association of diabetes with culture growth in our study and study by Lee et al. 7 . The mean hemoglobin in this study in ED was 11.82mg/dl which is almost equivalent to the mean hemoglobin in Choi et al study which was 12.2 mg/dl 5 . The mean total leucocyte counts in ED were higher (14.7 x109/L) when compared to Choi et al study (11.6x109/L) as most of our cases presented with high fever 5 .

The mean serum albumin in our study is 3.4 gm/dl which is slightly lower than Choi et al study which was 3.6gm/dl 5 . The total rate of contamination in our study done in ED and MICU/MHDU was 4.8%. In this study, the rate of contamination is lower in ED (4.4%) when compared to MICU/MHDU (5.4%).

A Similar study done by Choi et al showed blood culture contamination rates were higher in ED comprising 4% when compared to general wards (0.5%) 5 . In a study by Ramirez et al showed a blood culture contamination rate in ICU decreased from 23% to 13 % by using an education-based intervention 8 .

Raja et al studied 11000 patients over 2 years period showed that the contamination rates were higher in ICU (31%) when compared to ED (20%) 9 . The Bentley et al study also found that BCC rates were higher in ED (4.74 percent), which they were able to reduce to 2 percent within a year with a simple and clear checklist and rationalizing equipment to help and not detract from this approach with a specifically specified preferred technique 2 . Self WH et al in their study was able to reduce the BCC rates from 4.3% to 1.7% by following a standardized, sterile process for culture collection using chlorhexidine skin antisepsis, sterile gloves, sterile drapes, and checklists 10 .

In our study, the growth is seen in 191/998 (19.1%) cases. Of which growth in ED is 108 (18.9%) and in MICU/MHDU is 83(19.4%). A higher percentage of growth in ED may be because of more number of patients in this arm. A study done by Choi et al also had near similar growth in ED (17.5 %) 5 . A similar study done in ICU by Ramirez et al showed a culture growth of 31% (12). The most common contaminant found during this study was CoNS which was similar to most of the studies 7,9 .

The blood culture procedure variables were not significant with contamination (p-value >0.05). But according to the study by Lee et al in Taiwan, there was a strong correlation between blood culture contamination rates and the degrees of ED crowding (P.001) 7,11 .

A study done by Kim et al on blood culture contamination stated that the contamination rate was 0.5% in routine sterile gloving and 0.9% in optional sterile gloving with a significant P-value. Wearing a sterile glove in an aseptic manner before venepuncture may reduce blood culture contamination 12 . Various studies on the BCC rate among different antiseptics showed no significant difference among the antiseptics used 13 . Weinstein at el. study suggests that iodine tincture and chlorhexidine tincture are equivalent antiseptic agents for skin antisepsis in patients who require blood cultures 14 they can also cause true bloodstream infections. Due to its clinical effects, this distinction is of practical significance because it can avoid the unfair use of antibiotics and the development of antimicrobial resistance. More importantly, the inability to ascertain and treat true bacteremia can prove costly to the patient, more so if the patient is critically ill or immunocompromised. A clue to the significance of CoNS-positive blood cultures is the number of positive cultures, thus more the number of positive cultures, the higher the chances of it being true bacteremia. However, this is not feasible if before beginning the patient on antimicrobial agents, only a single culture sample is collected. Quantitative blood cultures (QBCs) can aid interpretation. QBCs are cumbersome and not very feasible. On the other hand, the time-to-positivity (TTP) of blood cultures after loading in the automated systems like BacT/ Alert may be a useful surrogate test for bacterial density and interpretation of the significance of CoNS isolated from positive blood cultures 15 .

V.

Blood culture contamination is a common clinical problem and often leads to both adverse impacts on health care and costs. We identified a low contamination rate among blood cultures collected in the adult ED at our hospital 4.4% when compared to MICU/MHDU (5.4%). We researched the process of blood culture collection and found inconsistent methods for culture collection with recurrent breaches in aseptic technique in ED. As we know ED frequently experiences high patient volumes and crowding and that leads to making things do as soon as possible and in that way, many lapses in protocol happen and that leads to degraded performance of blood cultures, both increasing the rate of contamination and decreasing the diagnostic yield.