A systematic review of the clinical significance of nephrostomy urine cultures.

PURPOSE: There are few published reviews that have assessed the clinical utility of renal urine cultures following percutaneous nephrostomy (PCN). In this systematic review, we evaluated the published evidence of the clinical utility of nephrostomy urine cultures in the light of emerging antimicrobial resistance and need for stewardship. METHODS: We performed a systematic literature search and review for evidence on the utility and role of nephrostomy urine cultures, using Medline, Embase and PubMed. We looked for evidence to assess whether there is any utility in collecting renal urine for culture at the time of percutaneous nephrostomy (PCN) and if the culture results of nephrostomy urine and bladder urine are different. We studied outcomes of treatment based on nephrostomy culture results. We also examined the role of PCN cultures at the time of routine nephrostomy exchange. Finally, we assessed if doing a PCN leads to infection or pyelonephritis. RESULTS: From 94 studies initially identified, we finally selected two randomised clinical trials (RCT), six original articles and five detailed conference abstracts for the review. These studies suggest that PCN urine cultures are overall useful in clinical practice. They are useful in selecting appropriate antimicrobial treatment for urosepsis following upper urinary obstruction. There does not appear to be any advantage in performing PCN cultures at routine nephrostomy exchanges. Occasionally, PCN itself can lead to subsequent urosepsis. CONCLUSION: Nephrostomy urine cultures have utility in clinical practice and can help support treatment and antimicrobial stewardship.

Outbreak report of investigation and control of an outbreak of Panton-Valentine Leukocidin-positive methicillin-sensitive Staphylococcus aureus (PVL-MSSA) infection in neonates and mothers.

BACKGROUND: In January 2011, there was an outbreak of Panton-Valentine Leukocidin-positive methicillin-sensitive Staphylococcus aureus (PVL-MSSA) infection in a neonatal unit (NNU). We describe the investigation and control of an outbreak of PVL-MSSA infection in neonates. SETTING: Neonatal unit in West London. METHODS: We performed descriptive and analytical (case-control study) epidemiological investigations. Microbiological investigations including screening of MSSA isolates by PCR for the presence of the luk-PV, mecA and mecC genes and comparison of isolate with Pulsed field gel electrophoresis (PFGE). Control measures were also introduced. RESULTS: Sixteen babies were infected/colonised with the outbreak strain. Of these, one baby developed blood stream infection, 12 developed skin pustules and four babies were colonised. Four mothers developed breast abscesses. Eighty-seven babies in the unit were screened and 16 were found to have same PVL-MSSA strain (spa type t005, belonging to MLST clonal complex 22). Multivariate analysis showed gestational age was significantly lower in cases compared to controls (mean gestational age: 31.7 weeks v 35.6 weeks; P = 0.006). Length of stay was significantly greater for cases, with a median of 25 days, compared to only 6 days for controls (P = 0.01). Most (88%) cases were born through caesarean section, compared to less than half of controls. (P = 0.002). No healthcare worker carriers and environmental source was identified. The outbreak was controlled by stopping new admissions to unit and reinforcing infection control precautions. The outbreak lasted for seven weeks. No further cases were reported in the following year. CONCLUSIONS: Infection control teams have to be vigilant for rising prevalence of particular S. aureus clones in their local community as they may cause outbreaks in vulnerable populations in healthcare settings such as NNUs.

Differential rates of group B streptococcus (GBS) colonisation in pregnant women in a racially diverse area of London, UK: a cross-sectional study.

OBJECTIVE: To describe the epidemiology of maternal group B streptococcus (GBS) colonisation by racial group. DESIGN: Cross-sectional study. SETTING: Antenatal clinics in London North West University Healthcare NHS Trust. POPULATION: Pregnant women. METHODS: Group B streptococcus (GBS) colonisation status was recorded during a screening programme for the prevention of invasive early-onset GBS infection. Information regarding age, address, ethnicity, parity, mode of delivery, body mass index (BMI), and diabetes was routinely collected. Data were analysed by multivariable analysis. MAIN OUTCOME MEASURES: Association between GBS colonisation and putative risk factors. RESULTS: Overall, 29.1% (1836/6309) of the women were colonized with GBS. Multivariable analysis showed significantly higher colonisation among women of black African origin (39.5%; OR = 1.57) compared with white British women (27.4%), and lowest colonisation in women of South Asian origin (23.3%; OR = 0.8). Higher parity (≥2) was associated with higher colonisation (35.3%), with the odds of colonisation over 40% higher than for nulliparous women. Increasing BMI was associated with an incremental rise in colonisation from 23 to 35%. Colonisation was not associated with age, season or mode of testing. CONCLUSION: This study identified high maternal GBS colonisation rates in a racially and socially diverse population. The highest rates were seen in women of black African origin and also with higher parity and BMI. Further research is needed to understand the relationship between these factors and rectovaginal colonisation. TWEETABLE ABSTRACT: Study of group B streptococcus colonisation in pregnant women in London shows highest rates in black African women and those with high BMI and parity.

Changes observed in urine microbiology following replacement of long-term urinary catheters: need to modify UTI guidelines in the UK?

AIMS: Bacteria adherent to long-term urinary catheters (LTUC) may give misleading urine culture results. Guidelines in the USA recommend changing LTUC before urine collection to diagnose UTI and before commencing appropriate antimicrobial treatment. However, in the UK there is no such guidance. In this study, we evaluated differences in urine cultures before and after changing LTUC. METHODS: In a prospective study in a UK urology department, we made a quantitative and qualitative comparison between paired urines collected before and after catheter change in patients with LTUC. We measured culture growth on a four-point ordinal scale as nil, scanty (< 107 cfu/L), moderate (107-108 cfu/L) or heavy (> 108 cfu/L) and recorded the range of bacterial species isolated. Statistical analysis was by Wilcoxon matched-pairs test. RESULTS: Sixty-six patients (55 males, 11 females) took part in the study. Urines with no growth increased from 7/66 (11%) before change of catheter to 21/66(32%) after change of catheter. Cultures reported as heavy growth (> 108 cfu/L) reduced from 48/66 (73%) to 25/66 (38%) after catheter change (p < 0.001). Except for Pseudomonas spp., other organisms were isolated less frequently after catheter change. No Proteus spp. was isolated after catheter change. CONCLUSIONS: This study confirms that failure to change long-term catheters before collecting urine for culture may give misleading results. In the interest of accurate diagnosis and antimicrobial stewardship, UK guidelines should recommend changing long-term urinary catheters before collection of urine for culture.

Outcome of a screening programme for the prevention of neonatal invasive early-onset group B Streptococcus infection in a UK maternity unit: an observational study.

BACKGROUND: Against a background of failure to prevent neonatal invasive early-onset group B Streptococcus infections (GBS) in our maternity unit using risk-based approach for intrapartum antibiotic prophylaxis, we introduced an antenatal GBS carriage screening programme to identify additional women to target for prophylaxis. OBJECTIVES: To describe the implementation and outcome of an antepartum screening programme for prevention of invasive early-onset GBS infection in a UK maternity unit. DESIGN: Observational study of outcome of screening programme (intervention) with comparison to historical controls (preintervention). SETTING: Hospital and community-based maternity services provided by Northwick Park and Central Middlesex Hospitals in North West London. PARTICIPANTS: Women who gave birth between March 2014 and December 2015 at Northwick Park Hospital. METHODS: Women were screened for GBS at 35-37 weeks and carriers offered intrapartum antibiotic prophylaxis. Screening programme was first introduced in hospital (March 2014) and then in community (August 2014). Compliance was audited by review of randomly selected case records. Invasive early-onset GBS infections were defined through GBS being cultured from neonatal blood, cerebrospinal fluid or sterile fluids within 0-6 days of birth. MAIN OUTCOME: Incidence of early-onset GBS infections. RESULTS: 6309 (69%) of the 9098 eligible women were tested. Screening rate improved progressively from 42% in 2014 to 75% in 2015. Audit showed that 98% of women accepted the offer of screening. Recto-vaginal GBS carriage rate was 29.4% (1822/6193). All strains were susceptible to penicillin but 11.3% (206/1822) were resistant to clindamycin. Early onset GBS rate fell from 0.99/1000 live births (25/25276) in the prescreening period to 0.33/1000 in the screening period (Rate Ratio=0.33; p=0.08). In the subset of mothers actually screened, the rate was 0.16/1000 live births (1/6309), (Rate Ratio=0.16; p<0.05). CONCLUSIONS: Our findings confirm that an antenatal screening programme for prevention of early-onset GBS infection can be implemented in a UK maternity setting and is associated with a fall in infection rates.

Key demographic characteristics of patients with bacteriuria due to extended spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae in a multiethnic community, in North West London.

BACKGROUND: Infections with extended-spectrum beta-lactamase-producing Enterobacteriaceae (ESBLE) are encountered worldwide, particularly in certain regions of the world and in certain ethnic groups. Simple criteria for identification of patients likely to be infected with ESBLE may enable clinicians to select appropriate empirical antibiotics for treatment. The aim of this study was to explore the association between ESBLE bacteriuria and readily available key demographic characteristics (age, gender and ethnicity) in a multiethnic population. METHODS: In this cross-sectional observational study, we explored the association between ESBLE bacteriuria and age, gender and ethnicity in 134 831 patients who submitted urine specimens for culture during 2007-2009 in two multiethnic boroughs in London, UK. RESULTS: In multivariate analysis, the risk of ESBLE bacteriuria was higher in males (odds ratio, OR = 1.3) and in patients older than 60 years (OR > 2). Patients from an Asian ethnic group were significantly more likely than White British subjects to have ESBLE bacteriuria (Indians, OR = 2.7; Asians of any other background, OR = 2.4; and Pakistanis, OR = 1.8). In contrast, patients who were of white ethnic background other than Irish were 0.66 times less likely to have ESBLE bacteriuria than White British patients (p = 0.025). CONCLUSIONS: Our study shows that in our local multiethnic population, older patients (> 60 years), males and those of South Asian ethnicity were significantly more likely to have ESBLE bacteriuria than others. Knowledge of these simple and readily available demographic data can help identify groups of patients at risk of urinary tract infection (UTI) with ESBLE and may aid in choice of empirical antibiotics.

Antimicrobial use and antimicrobial resistance in nosocomial pathogens at a tertiary care hospital in Pune.

BACKGROUND: Resistance to antimicrobial agents is emerging in wide variety of nosocomial and community acquired pathogens. Widespread and often inappropriate use of broad spectrum antimicrobial agents is recognized as a significant contributing factor to the development and spread of bacterial resistance. This study was conducted to gain insight into the prevalent antimicrobial prescribing practices, and antimicrobial resistance pattern in nosocomial pathogens at a tertiary care hospital in Pune, India. METHODS: Series of one day cross sectional point prevalence surveys were carried out on four days between March and August 2014. All eligible in patients were included in the study. A structured data entry form was used to collect the data for each patient. Relevant samples were collected for microbiological examination from all the clinically identified hospital acquired infection cases. RESULTS: 41.73% of the eligible patients (95% CI: 39.52-43.97) had been prescribed at least one antimicrobial during their stay in the hospital. Beta-lactams (38%) were the most prescribed antimicrobials, followed by Protein synthesis inhibitors (24%). Majority of the organisms isolated from Hospital acquired infection (HAI cases) were found to be resistant to the commonly used antimicrobials viz: Cefotaxime, Ceftriaxone, Amikacin, Gentamicin and Monobactams. CONCLUSION: There is need to have regular antimicrobial susceptibility surveillance and dissemination of this information to the clinicians. In addition, emphasis on the rational use of antimicrobials, antimicrobial rotation and strict adherence to the standard treatment guidelines is very essential.

Impact of implementation of a novel antimicrobial stewardship tool on antibiotic use in nursing homes: a prospective cluster randomized control pilot study.

OBJECTIVES: To evaluate the impact of 'Resident Antimicrobial Management Plan' (RAMP), a novel antimicrobial stewardship tool on systemic antibiotic use for treatment of infection in nursing homes (NHs). METHODS: A pilot cluster randomized control study was conducted in 30 NHs in London. Pre-intervention, we collected point prevalence data on antimicrobial use on three occasions and total antimicrobial consumption for a 12 week period. Post-intervention data were collected in the same manner and included assessment of compliance with RAMP in the intervention group (IG). RESULTS: The number of residents included was 1628 pre-intervention [825 IG/803 control group (CG)] and 1610 post-intervention (838 IG/772 CG). The corresponding pre- and post-intervention point prevalence of systemic antibiotic prescribing for treatment of infection was 6.46% and 6.52% in the IG [estimated prevalence ratio: 1.01 (95% CI: 0.81-1.25), P = 0.94] compared with 5.27% and 5.83%, respectively, in the CG [estimated prevalence ratio: 1.11 (95% CI: 0.87-1.41), P = 0.4]. Total antibiotic consumption was 69.78 defined daily doses/1000 residents/day (DRD) pre-intervention and 66.53 DRD post-intervention in the IG compared with 49.68 and 51.92 DRD, respectively, in the CG. There was a significant decrease of 4.9% (3.25 DRD) in the IG (95% CI: 1.0%-8.6%) (P = 0.02) compared with a significant increase of 5.1% (2.24 DRD) in the CG (95% CI: 0.2%-10.2%) (P = 0.04). Main indications for antibiotics were lower respiratory tract infections (34.1%), urinary tract infections (28.5%) and skin/soft tissue infections (25.1%). CONCLUSIONS: This pilot study demonstrated that use of RAMP was associated with a statistically significant decrease in total antibiotic consumption and has the potential to be an important antimicrobial stewardship tool for NHs.

Emergency hospital admissions attributable to infective complications of prostate biopsy despite appropriate prophylaxis: need for additional infection prevention strategies?

PURPOSE: To report the incidence of emergency admissions attributable to infective complications of transrectal ultrasound-guided prostate biopsy (TGB) and evaluate appropriateness of antimicrobial prophylaxis. METHODS: Retrospective cross-sectional study of patients undergoing TGB at the North West London Hospitals in 2009-2011. Demographic information of patients who had emergency admission within 30 days of TGB, length of hospital stay and microbiology results were obtained from the hospital's information system, medical records and laboratory information system. All patients received ciprofloxacin and amikacin prophylaxis. RESULTS: 1,419 TGB were performed in 1,276 patients. Forty-eight (3.3 %) patients had emergency admissions. Thirty-three (2.3 %) admissions were due to complications from TGB, while 15 (1 %) were for unrelated reasons. 30/33 (90 %, overall 2.1 %) of admissions from TGB complications were due to confirmed infections or systemic inflammatory response syndrome (SIRS). The rest were admitted with urinary retention. Admission rates due to TGB complications and infection/SIRS remained consistent over the 3 years (2009: TGB complications-2 %, TGB infection/SIRS-1.8 %; 2010: TGB complications-2.4 %, TGB infections/SIRS-2.2 %; 2011: TGB complications-2.6 %, TGB infection/SIRS-2.4 %; P > 0.05). All 11 cases with bacteraemia were caused by ciprofloxacin-resistant but amikacin-susceptible E. coli. CONCLUSIONS: We observed a consistent rate of emergency admissions for complications following TGB; 90 % of these were due to infections. Ciprofloxacin-resistant but amikacin-sensitive E. coli was isolated in all bacteriologically confirmed infections. These results suggest that infective complications of TGB cannot be altogether eliminated despite appropriate antimicrobial prophylaxis. Therefore, additional strategies for reduction in biopsy-related admissions due to infections have to be considered.

Can culture confirmation of gonococcal infection be improved in female subjects found to be positive by nucleic acid amplification tests in community clinics?

BACKGROUND: Use of nucleic acid amplification tests (NAATs), such as strand displacement assay (SDA, BD ProbeTec C trachomatis/N gonorrhoeae Amplified DNA Assay), for the detection of gonococcal infection in the community is controversial because of the possibility of false-positive results in low prevalence populations. AIM: To evaluate if culture confirmation of gonococcal infection can be improved for subjects found to be positive by BD ProbeTec in community clinics. METHODS: Two cervical swabs were collected for culture to confirm NAAT positive results in women aged over 16 years-a majority of whom were <25 years and asymptomatic. One swab was urgently transported (UTP) and processed in the laboratory within 2 hours whereas the other swab (RTP) was stored at 4 degrees C, transported at room temperature and processed 4-72 hours after collection depending on the time and day of collection. RESULTS: Altogether, 56 subjects with NAAT positive results were recruited into the study. Nine (16.1%) subjects who were culture negative were excluded from final analysis due to prior antibiotic treatment (4/9) or the culture having been taken more than 1 month after the NAAT was positive (4/9) or an incorrect specimen being received (1/9). Overall, 41/47 (87.2%) NAAT positive subjects were confirmed by culture. In total, 40/47 (85.1%) UTP swabs and 27/47 (57.4%) RTP swabs were positive (p<0.05). CONCLUSION: This study shows that culture confirmation in NAAT positive subjects in a community gonococcus screening programme can be significantly improved by urgent transportation to and processing of specimens in the laboratory.

Effectiveness of short-term, enhanced, infection control support in improving compliance with infection control guidelines and practice in nursing homes: a cluster randomized trial.

In this prospective cluster randomized controlled trial we evaluated the impact of short-term provision of enhanced infection control support on infection control practice in nursing homes in South London. Twelve nursing homes were recruited, six each in intervention (300 residents) and control (265 residents) groups. Baseline observations of hand hygiene facilities, environmental cleanliness and safe disposal of clinical waste showed poor compliance in both groups. Post-intervention observations showed improvement in both groups. There was no statistical difference between the two groups in the compliance for hand hygiene facilities (P=0.69); environmental cleanliness (P=0.43) and safe disposal of clinical waste (P=0.96). In both groups, greatest improvement was in compliance with safe disposal of clinical waste and the least improvement was in hand hygiene facilities. Since infection control practice improved in intervention and control groups, we could not demonstrate that provision of short-term, enhanced, infection control support in nursing homes had a significant impact in infection control practice.

Urinary tract infection in hospitalized elderly patients in the United Kingdom: the importance of making an accurate diagnosis in the post broad-spectrum antibiotic era.

The increasing prevalence of Clostridium difficile infection (CDI) has prompted many hospitals in the UK to recommend the use of narrow-spectrum antibiotics targeted at the likely bacteria at the clinical site of infection instead of broad-spectrum antibiotics. An underpinning requirement of such a strategy is the need to make an accurate diagnosis. In elderly patients, diagnosis of urinary tract infection can be challenging due to the frequent presence of respiratory signs and difficulties in the collection of urine specimens. This leads to the use of broad-spectrum antibiotics. Clinicians should attempt to make an accurate diagnosis of respiratory and urinary infections, as this will be crucial in the choice of the appropriate narrow-spectrum antibiotics.

Confirmation of BD ProbeTec Neisseria gonorrhoeae reactive samples by Gen-Probe APTIMA assays and culture.

BACKGROUND: Use of nucleic acid amplification tests (NAATs), such as strand displacement assay (SDA), for the detection of gonococcal infection in low prevalence populations is controversial because of the likelihood of false positive results. Use of supplementary NAATs with alternative target sites has been recommended for confirmation of primary NAAT results. AIM: To evaluate if SDA reactive specimens for Neisseria gonorrhoeae, which were either culture positive or negative, can be confirmed by alternative target NAATs such as transcription-mediated assays (TMA). METHODS: SDA reactive specimens were retested by TMA using APTIMA Combo 2 (AC2) and APTIMA GC (AGC) assays. Two different methods of specimen preparation were used to test the specimens. In method A, residual extract after SDA was retested and in method B, the original clinical specimen was re-extracted in TMA medium and then retested. Cervical or urethral swabs were requested to confirm the SDA results by culture. RESULTS: By method A, 26/49 (53.1%) of SDA positive specimens were positive by AC2 and/or AGC; 14/27 (51.8%) culture confirmed SDA positive tests were positive by AC2 and/or AGC. By method B, 38/39 (97.3%) SDA positive results were confirmed by both AC2 and AGC. All the 25 culture confirmed SDA positive tests were confirmed by both AC2 and AGC; 5/6 SDA positive tests that were culture negative were confirmed by both AC2 and/AGC. CONCLUSION: Alternative target site NAATs, such as AC2 and AGC, can be used to confirm SDA positive results using the same clinical specimen. There is high concordance between the three NAATs.

Prevalence and risk factors for meticillin-resistant Staphylococcus aureus in adult emergency admissions--a case for screening all patients?

In this study we describe the prevalence and frequency of risk factors for meticillin-resistant Staphylococcus aureus (MRSA) colonization in emergency admissions during a one-year MRSA screening programme. Overall, 7801/13 826 (56.4%) adult emergency admissions by 6469 patients were screened for MRSA. Of those screened, 670/7801 (8.6%) admissions by 433 patients (6.7%) were colonized with MRSA. Risk factors studied were previous hospital admission, previous MRSA colonization and residence in care homes. Patients with any risk factor (MRSA positive: 366/3952 (9.3%) vs MRSA negative: 67/2450 (2.7%), P<0.001), advanced age (mean ages for MRSA positive and negative patients were 74.4 and 56.2 years, respectively (P<0.001)), and increasing frequency of hospital admissions (P<0.001), patients from care homes (41/184, 22.2%) and previous MRSA colonization (232/1855, 12.5%) had a significantly higher MRSA colonization rate. The cost of the screening programme was nearly 24,500 pounds. It was noticed that there was a substantial reduction in hospital-acquired MRSA and MRSA bacteraemia during the study period. The study was not designed to establish whether this reduction was causally associated with the screening programme. We conclude that there is a high MRSA colonization rate among emergency admissions, especially those with risk factors. Using a selective risk factor based screening strategy more than 60% of the patients would have to be screened but still 3952/6469, 67 (15.5%) MRSA carriers would have been missed. Screening of all emergency admissions to detect MRSA colonization is preferable to selective screening, relatively inexpensive, and might reduce the MRSA colonization rate of hospital-acquired MRSA and MRSA bacteraemia among emergency admissions.

Interaction between methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-sensitive Staphylococcus aureus (MSSA).

In this paper we describe the in vitro interaction between three strains of methicillin-sensitive Staphylococcus aureus (MSSA) [NCTC 11561 and two strains derived from patients (PMSSA)] and endemic methicillin-resistant Staphylococcus aureus (EMRSA) 1, 3, 15 and 16. Mixed bacterial cultures of MSSA and EMRSA were incubated and subcultured after one, two and seven days. A proportion of MRSA in 50 randomly selected colonies was assessed. All strains of EMRSA (EMRSA 1, 3, 15 and 16) outgrew and virtually eradicated MSSA (NCTC 11561) after 24 h. The interaction between strains of PMSSA and the various strains of EMRSA was variable. PMSSA strain 1 was almost completely outgrown by EMRSA 1 and EMRSA 3 after seven days. Similarly there was a substantial increase of EMRSA 1 and 3 when tested against PMSSA strain 2. EMRSA 15 increased modestly against both strains, but EMRSA 16 failed to increase in proportion against either of the strains. We conclude that there is a complex interaction between various strains of EMRSA and MSSA. This interaction may have an important bearing on colonization of patients with MRSA.

Clostridium difficile-associated diarrhoea in patients with community-acquired lower respiratory infection being treated with levofloxacin compared with beta-lactam-based therapy.

The aim of the study was to compare the incidence of Clostridium difficile-associated diarrhoea (CDAD) following treatment of community-acquired lower respiratory tract infection (LRTI) in hospitalized patients with levofloxacin or a beta-lactam-based therapy. Nine hundred and thirty-eight patients were included in the prospective open-labelled "randomized" trial. This included 490 patients treated with levofloxacin and 448 patients treated with beta-lactams such as cefuroxime or amoxicillin. The overall incidence of CDAD was 3.8%. There was a lower incidence of CDAD (P < 0.01) in the patients treated with levofloxacin (levofloxacin group) (11/490; 2.2%) compared with patients treated with beta-lactams (beta-lactam group) (25/448; 5.6%), particularly with cefuroxime (cefuroxime group) (21/229; 9.2%; P < 0.0001). There was no significant difference (P = 0.6) in the incidence of CDAD between patients treated with levofloxacin or amoxicillin (amoxicillin group) (4/219; 1.8%). Patients in the cefuroxime and amoxicillin groups had a significantly longer duration of treatment than patients in the levofloxacin group. Although previous antibiotic therapy was a significant risk factor for CDAD in each of the groups, previous antibiotic therapy or admission to specific wards in the hospital were not confounding factors when comparing the groups. The levofloxacin group had a significantly shorter duration of hospitalization (mean 11.7 days; P < 0.01) compared with the beta-lactam group (mean 13.3 days), especially compared with the cefuroxime group (mean 16 days; P < 0.0000001). The amoxicillin group (mean 10.5 days) had a shorter duration of stay compared with the levofloxacin group, but this was not found to be statistically significant. Patients with CDAD had a longer duration of hospital stay than those without CDAD (CDAD +ve 25.8 days; CDAD -ve 11.9 days; P < 0.0000001). In conclusion, levofloxacin is less likely to be the cause of CDAD and is associated with a shorter duration of hospital stay compared with beta-lactam-based therapy for LRTI.