We searched PubMed for relevant articles and extracted/evaluated the available evidence.

We identified 22 microbiological studies reporting data for 2384 Acinetobacter spp. (1906 Acinetobacter baumannii). Susceptibility of at least 90% of the Acinetobacter isolates to tigecycline (with an MIC breakpoint of susceptibility ≤2 mg/L) was noted in 9/18 studies reporting data on MDR Acinetobacter and in 7/15 studies reporting specific data on carbapenem-resistant Acinetobacter. In an additional study reporting data for both resistance categories, adequate susceptibility of Acinetobacter spp. was observed by one (broth microdilution) of the methods employed. The effectiveness of tigecycline for MDR Acinetobacter infections was evaluated in eight identified clinical studies, reporting retrospective data regarding 42 severely ill patients, among whom 31 had respiratory tract infection (in 4 cases with secondary bacteraemia) and 4 had bacteraemia. Tigecycline therapy (in combination with other antibiotics in 28 patients) was effective in 32/42 cases. In three cases, resistance to tigecycline developed during treatment.

Tigecycline showed considerable, though not consistent, antimicrobial activity against MDR (including carbapenem-resistant) Acinetobacter spp. However, data to support its clinical use, particularly for ventilator-associated pneumonia or bacteraemia, caused by these pathogens, are still limited.

We performed a systematic review and meta-analysis to evaluate the accuracy of the NRA for the detection of rifampicin- and isoniazid-resistant tuberculosis. We searched Medline PubMed (NCBI), Global Health-CAB, EJS-E (EbscoHost), ISI Web, Web of Science and IFCC and contacted authors if additional information was required. Fifteen studies met our inclusion criteria for rifampicin resistance detection and 13 for isoniazid. Of these, the majority of the studies used culture isolates on solid medium, four used culture isolates on liquid medium and three used sputum samples. We applied the summary receiver operating characteristic (SROC) curve to perform meta-analysis and to summarize diagnostic accuracy.

For rifampicin, the majority of the studies that applied NRA to isolates had a sensitivity and specificity >94% and for isoniazid, >92%. The three studies that applied NRA directly on sputum samples had a sensitivity and specificity that ranged between 88% and 100%. The SROC curve had an area of >0.99 for both drugs.

There is evidence that NRA is highly sensitive and specific for the rapid detection of rifampicin and isoniazid resistance in culture isolates. More evidence is required for the NRA applied directly on sputum samples, but preliminary results appear promising and show a good sensitivity and specificity. Additional studies are required in countries with a high prevalence of MDR-TB and also cost-effectiveness analysis in order to obtain a complete picture on the utility of this method for rapid drug resistance detection in tuberculosis.

The plasmid pAB5S9 was transferred by electroporation into Escherichia coli TG1. The resistance phenotype mediated by pAB5S9 was determined. Moreover, the plasmid was sequenced completely and analysed for its structure and organization of reading frames.

Plasmid pAB5S9 mediated resistances to phenicols, sulphonamides, streptomycin and tetracycline. The analysis of the 24.7 kb sequence revealed the presence of 20 predicted coding sequences (CDSs), which included the floR, sul2 and strA-strB resistance genes and a tetR-tet(Y) determinant. Approximately 7.5 kb of pAB5S9 showed 100% nucleotide sequence identity to three non-contiguous segments of the SXT element of Vibrio cholerae. Regions identical to SXT comprised the floR gene, flanked upstream by a complete and downstream by a truncated ISCR2 element, and the region of the sul2 and strA-strB genes. Other CDSs of pAB5S9 related to plasmid replication and partitioning, metabolic and gene regulation functions as well as conjugative transfer showed homology to sequences from diverse bacterial species, indicating a mosaic structure.

This study provides the first report of a floR-carrying plasmid in the genus Aeromonas and the first description of a tetR-tet(Y) determinant. The analysis of the multiresistant A. bestiarum strain indicates that strains of this species, some of which are opportunistic pathogens for fish, might also act as a resistance gene reservoir in the freshwater environment.

In a previous study, we created and characterized the G. lamblia WB C6 clone C4 resistant to nitazoxanide and metronidazole. In this study, using a microarray-based approach, we have identified open-reading frames (ORFs) that were differentially expressed in C4 when compared with its wild-type WB C6. Using quantitative real-time PCR, we have validated the expression patterns of some of those ORFs, focusing on chaperones such as heat-shock proteins in wild-type and C4 trophozoites. In order to induce an antigenic shift, trophozoites of both strains were subjected to a cycle of en- and excystation. Expression of selected genes and resistance to nitazoxanide and metronidazole were investigated after this cycle.

Forty of a total of 9115 ORFs were found to be up-regulated and 46 to be down-regulated in C4 when compared with wild-type. After a cycle of en- and excystation, resistance of C4 to nitazoxanide and metronidazole was lost. Resistance formation and en-/excystation were correlated with changes in expression of ORFs encoding for major surface antigens such as the variant surface protein TSA417 or AS7 (‘antigenic shift’). Moreover, expression patterns of the cytosolic heat-shock protein HSP70 B2, HSP40, and of the previously identified nitazoxanide-binding proteins nitroreductase and protein disulphide isomerase PDI4 were correlated with resistance and loss of resistance after en-/excystation. C4 trophozoites had a higher thermotolerance level than wild-type trophozoites. After en-/excystation, this tolerance was lost.

These results suggest that resistance formation in Giardia to nitazoxanide and metronidazole is correlated with altered expression of genes involved in stress response such as heat-shock proteins.

Mutants resistant to triclosan were selected from nine S. enterica serovar Typhimurium strains. Mutants were characterized by genotyping, mutagenesis and complementation of fabI and analysis of efflux activity. Fitness of triclosan-resistant mutants was determined in vitro and in vivo.

Three distinct resistance phenotypes were observed: low- (LoT), medium- (MeT) and high-level (HiT) with MICs of 4–8, 16–32 and >32 mg/L of triclosan, respectively, for inhibition. The genotype of fabI did not correlate with triclosan MIC. Artificial overexpression and mutagenesis of fabI in SL1344 each resulted in low-level triclosan resistance, indicating that FabI alone does not mediate high-level triclosan resistance in Salmonella Typhimurium. Active efflux of triclosan via AcrAB–TolC confers intrinsic resistance to triclosan as inactivation of acrB and tolC in wild-type strains and the triclosan-resistant mutants led to large decreases in triclosan resistance, which were reversed by complementation. Exemplars of each phenotype were evaluated for fitness in vivo; no fitness cost was seen and mutants colonized and persisted in chickens throughout a 28 day competitive index experiment.

These data show that triclosan resistance can occur via distinct pathways in salmonella and that mutants selected after single exposure to triclosan are fit enough to compete with wild-type strains.

The proteomes of three distinct but isogenic triclosan-resistant mutants were determined using two-dimensional liquid chromatography mass separation. Bioinformatics was then used to identify and quantify tryptic peptides in order to determine protein expression.

Proteomic analysis of the triclosan-resistant mutants identified a common set of proteins involved in production of pyruvate or fatty acid with differential expression in all mutants, but also demonstrated specific patterns of expression associated with each phenotype.

These data show that triclosan resistance can occur via distinct pathways in Salmonella, and demonstrate a novel triclosan resistance network that is likely to have relevance to other pathogenic bacteria subject to triclosan exposure and may provide new targets for development of antimicrobial agents.

MICs, resistance frequencies and mutant prevention concentrations (MPCs) were determined using quinolone-susceptible strains and first-step parC mutant strains of S. pneumoniae. Target preferences were tested by the following two methods: antibacterial activities against gyrA or parC mutants and in vitro enzyme assays for the determination of 50% inhibition (IC₅₀) values.

DC-159a and sitafloxacin exhibited potent antibacterial activities, low frequencies of mutant selection, low MPCs and narrow mutant selection windows against both quinolone-susceptible strains and first-step parC mutants of S. pneumoniae, compared with gatifloxacin, moxifloxacin and other quinolones tested. DC-159a and sitafloxacin showed relatively low MIC ratios against single gyrA or parC mutants relative to the wild-type strain and low IC₅₀ ratios against DNA gyrase and topoisomerase IV.

DC-159a and sitafloxacin demonstrated a more balanced dual-targeting activity than gatifloxacin, moxifloxacin and other quinolones tested. In addition, DC-159a and sitafloxacin have a lower propensity for selecting first- and second-step resistant mutants.

The antimicrobial activities of cross-linking agents and antimicrobials (five antibiotics and four antiseptics) were evaluated by the broth dilution method. The interactions between individual cross-linking agents and antimicrobials were evaluated using the chequerboard test against common skin pathogens, Staphylococcus aureus and Pseudomonas aeruginosa.

From the MIC determined, antibiotics were the most active, followed by the antiseptics and cross-linking agents. Calcium ions, which are commonly used to cross-link alginate, exhibited very weak antimicrobial activity and higher fractional inhibitory concentration than the other cross-linking agents. The use of calcium and gentamicin resulted in antagonism against S. aureus. In contrast, aluminium, zinc and copper ions exhibited higher antimicrobial activities but insignificant interactions with the antimicrobials.

Commonly used topical antimicrobials that are active against the skin pathogens S. aureus and P. aeruginosa could be potentially incompatible with calcium alginate dressings. Copper, zinc and aluminium ions are more suitable cross-linking agents for alginate as they do not show antagonism with the antimicrobials and could impart antimicrobial property to the resultant dressing.

The Surveillance Network (TSN) comprises 296 laboratories across the nine census regions of the USA. TSN laboratories reported the susceptibility data for six antimicrobials by isolate with source and other relevant data. Antimicrobial susceptibility data were analysed by individual drug resistance, multidrug resistance and geographical distribution of resistance phenotypes.

There were over 380 000 isolates of S. aureus tested and reported for the period 2005–07. Methicillin resistance was observed in 57.8% in 2007, with little change from 2005. There was little difference in rates of methicillin resistance between community and hospital strains, although strains from intensive care units (ICUs) tended to be slightly more resistant overall. Resistance to other antimicrobials was also reported. A regional variation in resistance rates was noted with the highest rates in the Central states and lowest in the New England and Mid-Atlantic regions. There was high activity observed with trimethoprim/sulfamethoxazole and gentamicin. Linezolid resistance was rare. Oxacillin resistance was similar among paediatric and elderly cohorts, whereas ciprofloxacin and clindamycin resistance was significantly (P < 0.01) more common in elderly patients when compared with both paediatric and adult populations. Less than a third of all isolates showed no resistance mechanism, 30.3%. Three distinct resistance phenotypes accounted for 46% of all resistant strains. Overall, there were more highly drug-resistant isolates from the ICU with four, five or six drug-resistant phenotypes accounting for over a third of all strains.

S. aureus has become methicillin-resistant in both the community and hospital settings; however, little change has been seen in the past 3 years. Multiresistant strains now are seen in all settings, but due to regional variation, empirical therapy should be guided by local susceptibility patterns. Currently, among the agents studied, only trimethoprim/sulfamethoxazole, gentamicin and linezolid exhibit susceptibility rates of >95%.

Gram-positive clinical isolates, collected between October 2004 and December 2005 from 36 hospital laboratories in 15 countries, were tested by broth microdilution using CLSI methodology.

In total, 3206 isolates were collected. Telavancin had potent activity against Staphylococcus aureus and coagulase-negative staphylococci (MIC range ≤0.015 to 2 mg/L), independent of resistance to methicillin or to multiple drugs. Telavancin had particularly strong activity against streptococcal isolates (MIC range ≤0.001 to 0.5 mg/L), including penicillin-resistant and multiple drug-resistant Streptococcus pneumoniae and erythromycin non-susceptible β-haemolytic and viridans group streptococci. Telavancin also had excellent activity against vancomycin-susceptible enterococci (MIC₉₀ 0.5 mg/L), and although its MICs were elevated against VanA strains (Enterococcus faecalis MIC₉₀ 8 mg/L and Enterococcus faecium MIC₉₀ 4 mg/L), its MIC₉₀ was substantially lower than observed with available glycopeptides.

Telavancin has potent in vitro activity against contemporary Gram-positive clinical isolates from diverse geographic areas in Europe and Israel.

Isolation, identification and typing of the pathogens were performed by means of conventional methods. The antimicrobial susceptibilities of the genital mycoplasmas were determined with commercially available kits and evaluated according to the CLSI.

In 65 (47.44%) out of the 137 positive specimens, U. urealyticum was grown as a single pathogen, in 0.72% M. hominis was grown as a single pathogen and in 2.92% both urogenital mycoplasmas were grown. In the remaining specimens (48.90%), there was a mixed growth with other microbes. Of the isolated U. urealyticum strains, 87.4% and 98.2% were susceptible to tetracycline and doxycycline, respectively, 79.2% were susceptible to josamycin, 48.6% were susceptible to clarithromycin and 91.8% were susceptible to pristinamycin, while erythromycin, azithromycin, ciprofloxacin and ofloxacin proved to be inactive against most of the strains. M. hominis isolates were 100% susceptible to tetracycline, doxycycline and pristinamycin, while susceptibilities to the other antimicrobial agents varied mainly in the range of ‘intermediate’ or ‘resistant’. As results originating from similar studies from various countries are very controversial, the simplest way to avoid therapeutic failures would be the implementation of rational treatment regimens based on culture isolation and the in vitro determination of the antimicrobial susceptibility of genital mycoplasmas in each clinical case.

During 1998–2005, E. coli from urine samples of patients attending urology services were collected in three regions in The Netherlands: north-east (NE, n = 1084), west (W, n = 1064) and south (S, n = 1212). The antibiotic susceptibility was determined using microbroth dilution following CLSI guidelines. E. coli ATCC 35218 and ATCC 25922 were used as reference strains.

Amoxicillin resistance remained stable over time (37% to 47%), but was higher in the south (44%) compared with the other regions (40%; P < 0.02). Resistance to piperacillin increased from 4% (1998) to 32% (2005; P < 0.001), and resistance to fluoroquinolones increased from 6% to 13% (P < 0.01). Interregional differences were observed for resistance to piperacillin (NE 10%, W 12%, S 14%; P < 0.05) and to fluoroquinolones (NE 7%, W 13%, S 8%; P < 0.001). Trimethoprim ± sulfamethoxazole resistance remained stable (27% to 37%), as did that of nitrofurantoin (4% to 9%). The percentage of strains with multidrug resistance (resistance to three or more groups of antibiotics) for each region increased over time (P < 0.05).

Antibiotic resistance was fairly constant over time for most agents tested, except for piperacillin and the fluoroquinolones. Regional differences were observed for several compounds. National and regional surveillance of antibiotic resistance is important to keep therapeutic guidelines up-to-date and adequate for the treatment of resistant microorganisms.

MICs were determined by Etest for isolates giving positive double-disc synergy tests, and all isolates were screened by PCR and sequenced, respectively, for bla_TEM, bla_CTX-M, bla_SHV and bla_VEB genes and for qnr genes (qnrA, qnrB, qnrS), using specific primers.

The prevalence of ESBLs was 25/141 (17.7%) with 11, 9, 4 and 1 isolates testing positive for genes encoding CTX-M-15, CTX-M-3, SHV-12 and VEB-1, respectively. Two SHV-12 producers and one CTX-M-15 producer expressed QnrS1, one isolate produced CTX-M-15 and QnrB1 and one SHV-12 producer co-expressed QnrS1 and QnrB4. qnrA was not detected in our collection, and qnr alleles were not detected in non-ESBL-producing isolates.

SHV-12, QnrS1, QnrB1 and QnrB4 were reported for the first time in Algeria. This study also described a co-expression of qnrS1 and qnrB4 by an SHV-12 producer isolate.

Four hundred and ninety-five Gram-negative anaerobic clinical isolates (296 Bacteroides fragilis group, 58 non-fragilis Bacteroides spp. and 141 Prevotella spp.) were prospectively recovered in six Greek hospitals. Moxifloxacin MICs were determined in comparison with those of penicillin, piperacillin/tazobactam, cefoxitin, imipenem, metronidazole and clindamycin.

Overall moxifloxacin MIC₅₀ and MIC₉₀ were 2 and 32 mg/L, respectively. Based on the current CLSI breakpoints (susceptible, ≤2 mg/L; resistant, ≥8 mg/L), almost half of the total isolates (49%) were non-susceptible to moxifloxacin (32% resistant; 17% intermediate). This was more evident among the non-fragilis Bacteroides species, where 47% of the isolates were resistant and 14% intermediate to moxifloxacin. Species variation was noticed, with the highest non-susceptible rates detected among Prevotella oralis (90%), Prevotella bivia (80%), Bacteroides thetaiotaomicron (75%), Bacteroides uniformis (70%) and Bacteroides capillosus (67%) species. Among the 19 (4%) isolates that were metronidazole non-susceptible (MIC ≥ 16 mg/L), only 4 (21%) were additionally non-susceptible to moxifloxacin.

High resistance rates to moxifloxacin among Bacteroides and Prevotella spp. were recorded, exceeding those previously reported in Europe and contraindicating its use as monotherapy for infections involving Gram-negative anaerobes without prior microbiological confirmation. For empirical usage, moxifloxacin should be combined with metronidazole in order to cover for these pathogens.

Four strains of Candida albicans susceptible to voriconazole were exposed to voriconazole and amphotericin B, either alone, simultaneously or sequentially in an in vitro kinetic model. Bolus doses resulting in voriconazole and amphotericin B concentrations of 0.005–5 and 2.5 mg/L, respectively, were administered. Antifungal-containing RPMI 1640 was eliminated and replaced by a fresh medium using a peristaltic pump, with a flow rate adjusted to obtain the desired half-lives. With two drugs tested, a computer-controlled dosing pump compensated for differences in the elimination rates. Using static time–kill methodology, one C. albicans strain was exposed to 5 mg/L voriconazole for varying durations followed by 2.5 mg/L amphotericin B after three repeated washes of voriconazole.

Voriconazole and amphotericin B treatment alone resulted in fungistatic and fungicidal activities, respectively. Simultaneous administration of voriconazole and amphotericin B resulted in fungicidal activity, whereas only fungistatic activity was observed when repeated doses of amphotericin B were administered sequentially after voriconazole at 24–96 h. The inhibition of the fungicidal activity of amphotericin B was voriconazole dose-dependent, but seemed to be recovered once the voriconazole concentration fell below the MIC. The fungicidal activity was quickly regained after the removal of voriconazole, irrespective of the duration of voriconazole pre-exposure.

Voriconazole inhibited the fungicidal effect of sequentially administered amphotericin B in a concentration- and time-dependent manner; the clinical significance of this needs further investigation.

We used six C. dubliniensis clinical isolates and the CD 36 type strain. MICs and MFCs of caspofungin were determined using the standard broth microdilution method with normal (10³ cells/mL) and elevated (10⁵ cells/mL) starting inocula in RPMI-1640 and antibiotic medium 3 (AM3). MIC was determined after 24 h, and plating for MFC determination was performed after 48 h. In time–kill tests, all strains were tested at 0.06–16 mg/L caspofungin concentrations in RPMI-1640 and AM3.

In RPMI-1640, the MIC range was 0.06–8 mg/L. Trailing growth was observed regardless of the starting inoculum after 48 h, but not after 24 h. In AM3 regardless of starting inoculum, MICs were 0.03 mg/L. After 48 h, trailing was not detected; two isolates grew at a concentration of 8 mg/L using 10⁵ cells/mL as the starting inoculum [paradoxical growth (PG)]. All MFCs in RPMI-1640 and AM3 were >8 and ≤0.12 mg/L, respectively. In AM3, all but a single isolate showed PG in the MFC tests. Time–kill tests confirmed the results obtained by MFC tests both in RPMI-1640 and AM3.

In vitro activity of caspofungin against C. dubliniensis depended on the starting inoculum and medium used. Using AM3 eliminated trailing from MIC determinations but not PG in MIC, MFC and time–kill tests.

An in vitro model of a C. albicans (ATCC 3153 or ATCC 66396) biofilm associated with 100% silicone catheters was used. The effectiveness of the antifungal treatment was assayed against biofilms aged 12 h or 5 days, after exposure to caspofungin (2 mg/L) or micafungin (5 mg/L) for 12 h. The durability of the reduction in the biofilm metabolic activity was investigated (1–3 days after echinocandin treatment). The efficacy of caspofungin and micafungin was determined by evaluating a significant decrease (P < 0.0001) in the metabolic activity of biofilm yeasts.

The results showed that the tested antifungal agents used as lock solution significantly (P < 0.0001) reduced the metabolic activity of C. albicans, whatever the biofilm maturation stage (12 h or 5 days old biofilms). The reduction in the metabolic activity of biofilm yeasts was maintained, even after 48 h.

These data suggest that caspofungin (2 mg/L) and micafungin (5 mg/L) could represent good candidates for the reduction or control of fungal biofilms associated with silicone medical devices, as part of an antifungal lock. They were able to induce a significant and persistent reduction in the yeast metabolic activity of intermediate and mature biofilms, 12 h and 5 days old, respectively, when used as catheter lock solutions.

Using time–kill studies at a wide range of concentrations of ABI-0043, we evaluated the killing activity against four clinical isolates of S. aureus over 24 h. An integrated pharmacokinetic/pharmacodynamic area measure was applied to all cfu data and was fitted to a Hill-type mathematical model to evaluate pharmacodynamics.

Bacterial killing for ABI-0043 occurred rapidly and in a concentration-dependent manner. Bactericidal activity was achieved within 4 h at ≥16x MIC against all isolates. Bacterial reductions were greatest at ≥64x MIC against MRSA and MSSA isolates, as a >4 log₁₀ cfu/mL reduction was observed as early as 2 h, and sustained throughout 24 h. The pharmacodynamics of ABI-0043 was well described by a Hill-type model, with a steep sigmoidicity constant and a low EC₅₀ against all isolates.

ABI-0043 displayed rapid and sustained bactericidal activity against S. aureus clinical isolates. ABI-0043 represents a promising antistaphylococcal agent to combat serious S. aureus infections. Further, pharmacokinetic, pharmacodynamic and in vivo studies are warranted to determine its ultimate place in antibacterial therapy.

Seventy-seven pharmacokinetic profiles (saquinavir/ritonavir 1000/100 mg twice daily, n = 34; 1600/100 mg once daily, n = 26; 2000/100 mg once daily, n = 17) from five studies were combined, presented as twice- and once-daily percentiles (P10–P90) and compared. At percentiles where trough concentrations fell below the alleged minimum effective concentration (MEC; 100 ng/mL), the length of time below MEC was determined.

Saquinavir concentrations were below MEC at P10 for 0.7 h for twice-daily saquinavir/ritonavir when compared with 8.6 and 6.6 h for 1600/100 and 2000/100 mg once daily, respectively. At P25, 1600/100 mg once daily produced suboptimal concentrations for 5.5 h in contrast to 0.5 h for 2000/100 mg once daily.

Here, we provide substantive data that indicate once-daily saquinavir, in particular 1600/100 mg, is not as robust as the twice-daily regimen based on a population of UK patients; this raises concern over late or missed doses. However, pharmacokinetic data can only ever be a guide to the impact on long-term efficacy.

A retrospective cohort study of patients receiving continuous vancomycin infusion as outpatient parenteral antibiotic therapy (OPAT) was performed. The likelihood of developing nephrotoxicity (≥50% increase in serum creatinine from baseline) was evaluated in relation to demographic variables, underlying co-morbidities, infectious disease diagnoses, concomitant drug exposures and vancomycin concentration. Logistic regression was used to determine the association of various variables. Classification and regression tree analysis was used to determine the most significant breakpoint for continuous variables.

We examined 102 adult patients between January 2004 and June 2007. The mean ± SD age, baseline serum creatinine and steady-state vancomycin concentration were 48.2 ± 17.6 years, 78.0 ± 32.5 µmol/L and 15.5 ± 10.8 mg/L, respectively. The majority of the patients (66.7%) were treated for bone and joint infection. The cumulative incidence of nephrotoxicity was 15.7%. Nephrotoxicity was found to be associated with hypertension [odds ratio (OR) 5.302 (95% confidence interval (CI) 1.159–24.246), P = 0.031], exposure to aminoglycosides [OR 6.594 (95% CI 1.026–42.385), P = 0.047], loop diuretics [OR 8.123 (95% CI 1.449–45.528), P = 0.017], and steady-state vancomycin concentration ≥28 mg/L [OR 21.236 (95% CI 2.687–167.857), P = 0.004].

We have identified independent risk factors for nephrotoxicity in patients receiving continuous infusion vancomycin in OPAT. A serum steady-state vancomycin concentration ≥28 mg/L markedly increases the risk.

In this population-based cohort study in Northern Denmark, we included 1079 women who had a live birth or a stillbirth after the 20th week of gestation and who redeemed at least one prescription for fluconazole during the first trimester. The reference cohort comprised 170 453 pregnant women who redeemed no fluconazole prescription during pregnancy. The women were identified through the Danish Medical Birth Registry. Data on drug use, birth outcome and covariates were extracted from population-based healthcare databases. We used logistic regression to estimate the prevalence odds ratio (POR) for congenital malformations after fluconazole exposure, while adjusting for maternal smoking, parity, maternal age and concurrent prescriptions for antiepileptics or antidiabetics.

Among 1079 women who filled a fluconazole prescription during the first trimester, 797 (74%) received a total of 150 mg of fluconazole, 235 (22%) received 300 mg of fluconazole, 24 (2%) received 350 mg of fluconazole and 23 (2%) received 600 mg of fluconazole. These women gave birth to 44 (4.1%) children with congenital malformations. The 170 453 women without fluconazole prescriptions gave birth to 6152 (3.6%) children with congenital malformations. For congenital malformations overall, the adjusted POR associated with the first-trimester fluconazole use was 1.0 (95% confidence interval: 0.8–1.4).

We found no overall increased risk of congenital malformations after exposure to short-course treatment with fluconazole in early pregnancy.

The study was undertaken using a consensus development panel to which the HFMEA^TM process was applied. Key stakeholders in the local OPAT process were invited to join the HFMEA^TM team with the aim of describing and agreeing (defined as 100% participant agreement) an OPAT map, its sub-processes and potential OPAT system failures.

The HFMEA^TM team identified 6 processes, 67 sub-processes and 217 possible failures over the course of four meetings. Key areas identified in the OPAT map concerned identifying and checking patient suitability for an OPAT service, involvement of a multidisciplinary team and robust communication channels.

An OPAT map was developed, which may serve as a practical model for other organizations setting up a similar service.

Pathogens were grouped into seven categories based upon the antibiotic susceptibility and epidemiological characteristics. Colonization of the upper respiratory tract was modelled in the BN and depended—in additional steps—on (i) duration of admission and ventilation, (ii) previous culture results and (iii) previous antibiotic use. A database with 153 VAP episodes and their microbial causes was used as reference standard. Appropriateness of antibiotic prescription, with fixed choices for pathogens predicted, was determined.

One hundred and seven VAP episodes were monobacterial and 46 were caused by two pathogens. Using duration of admission and ventilation only, areas under the receiver operating curve (AUC) ranged from 0.511 to 0.772 for different pathogen groups, and model predictions significantly improved when adding information on culture results, but not when adding information on antibiotic use. The best performing model (with all information) had AUC values ranging from 0.859 for Acinetobacter spp. to 0.929 for Streptococcus pneumoniae. With this model, 91 (85%) and 29 (63%) of all pathogen groups were correctly predicted for monobacterial and polymicrobial VAP, respectively. With fixed antibiotic choices linked to pathogen groups, 92% of all episodes would have been treated appropriately.

The BN models' performance to predict pathogens causing VAP improved markedly with information on colonization, resulting in excellent pathogen prediction and antibiotic selection. Prospective external validation is needed.

A qualitative study using focus group discussions explored the usability and applicability of local antibiotic guidelines together with possible supportive measures. The sample included 22 physicians, deliberately divided between internal medicine (59.1%) and surgery (40.9%), and levels of experience (59.1% residents; 40.9% supervisors). Focus groups were conducted within one specific subgroup. Analysis was carried out using a framework analysis approach.

General acceptance of local guidelines was high but clear differences were present between subgroups with different desires and requirements from guideline contents. Opposing views were present towards supportive measures, especially multidisciplinary collaboration. Guideline distribution and accessibility appeared to be confusing, resulting in delayed application. An important supplementary barrier was the need to collect the guideline personally. Supervisors in their role as opinion leaders were mentioned as highly influential towards residents' practice.

Locally developed hospital guidelines experience the same barriers as other guidelines. Within one hospital, prescribers have to be seen as a number of different target groups instead of a homogeneous population. For an optimal effect, interventions will have to consider these differences. Also, in order to improve local guideline use and antibiotic consumption, supervisors have to be aware of how their role as opinion leaders can influence residents. Lastly, active guideline distribution and promotion remains critical to ensure efficient guideline use. Future research should focus on how to adapt interventions to these different target groups.

Data (2003–04) from 26 departments in 6 hospitals were retrieved. Defined daily doses (DDD)/100 bed-days were calculated for total antibiotic use and by antibiotic class. Patterns identified were correlated with 15 patients’ and departmental variables by univariate and multivariate analyses.

Total antibiotic consumption differed by a factor of 2.3 (115 DDD/100 bed-days to 49.1 DDD/100 bed-days) between the highest and lowest consuming departments. Antibiotic classes differed by a factor of 22.8 for macrolides, a factor of 20 for piperacillin/tazobactam, a factor of 17 for carbapenems, a factor of 13.3 for quinolones, a factor of 9 for vancomycin, a factor of 6.8 for amoxicillin/clavulanate, a factor of 6.6 for aminoglycosides, a factor of 5.3 for penicillins and a factor of 2.8 for cephalosporins. Even among departments within hospitals, there was a difference of up to 1.5-fold for total use and antibiotic class differences ranged between 2.5- and 7.2-fold for third- and fourth-generation cephalosporins, despite similar Charlson scores and other patient variables. In the multivariate analysis, hospital affiliation and rate of 1 day hospitalization were the only significant variables predicting total antibiotic use, contributing 43% and 7.3%, respectively, to the variance. By antibiotic class, controlling for hospital affiliation, patients with neutropenia, lower respiratory tract infections and assisted ventilation were the most common significant contributors, ranging from 3.5% for quinolones to 7.7% for piperacillin/tazobactam.

Patterns of antibiotic use vary widely among internal medicine departments in Israel, which cannot be explained by objective parameters related either to patients or wards. Ongoing monitoring and guideline formulation are needed to regulate antibiotic prescription.