Eimear T. Brannigan

An analysis of the development and implementation of a smartphone application for the delivery of antimicrobial prescribing policy: lessons learnt

Objectives Smartphone usage amongst clinicians is widespread. Yet smartphones are not widely used for the dissemination of policy or as clinical decision support systems. We report here on the development, adoption and implementation process of the Imperial Antimicrobial Prescribing Application across five teaching hospitals in London. Methods Doctors and clinical pharmacists were recruited to this study, which employed a mixed methods in-depth case-study design with focus groups, structured pre- and post-intervention survey questionnaires and live data on application uptake. The primary outcome measure was uptake of the application by doctors and its acceptability. The development and implementation processes were also mapped. Results The application was downloaded by 40% (376) of junior doctors with smartphones (primary target user group) within the first month and by 100% within 12 months. There was an average of 1900 individual access sessions per month, compared with 221 hits on the Intranet version of the policy. Clinicians (71%) reported that using the application improved their antibiotic knowledge. Conclusions Clinicians rapidly adopted the mobile application for antimicrobial prescribing at the point of care, enabling the policy to reach a much wider audience in comparison with paper- and desktop-based versions of the policy. Organizations seeking to optimize antimicrobial prescribing should consider utilizing mobile technology to deliver point-of-care decision support. The process revealed a series of barriers, which will need to be addressed at individual and organizational levels to ensure safe and high-quality delivery of local policy at the point of care.

Homogeneity of antimicrobial policy, yet heterogeneity of antimicrobial resistance: antimicrobial non-susceptibility among 108 717 clinical isolates from primary, secondary and tertiary care patients in London

Objectives We examined the 4 year trend in antimicrobial susceptibilities and prescribing across levels of care at two London teaching hospitals and their multisite renal unit, and for the surrounding community. Methods Laboratory and pharmacy information management systems were interrogated, with antimicrobial use and susceptibilities analysed between hospitals, within hospitals and over time. Results A total of 108 717 isolates from 71 687 patients were identified, with significant differences (at P < 0.05) in antimicrobial susceptibility between and within hospitals. Across the 4 years, rates of ESBL-/AmpC-producing Enterobacteriaceae ranged from 6.4% to 10.7% among community isolates, 17.8% to 26.9% at ward level and 25.2% to 52.5% in critical care. Significant variations were also demonstrated in glycopeptide-resistant enterococci (ward level 6.2%–17.4%; critical care 21.9%–56.3%), MRSA (ward level 18.5%–38.2%; critical care 12.5%–47.9%) and carbapenem-resistant Pseudomonas spp. (ward level 8.3%–16.9%; critical care 19.9%–53.7%). Few instances of persistently higher resistance were seen between the hospitals in equivalent cohorts, despite persistently higher antimicrobial use in Hospital 1 than Hospital 2. We found significant fluctuations in non-susceptibility year on year across the cohorts, but with few persistent trends. Conclusions The marked heterogeneity of antimicrobial susceptibilities between hospitals, within hospitals and over time demands detailed, standardized surveillance and appropriate benchmarking to identify possible drivers and effective interventions. Homogeneous antimicrobial policies are unlikely to continue to be suitable as individual hospitals join hospital networks, and policies should be tailored to local resistance rates, at least at the hospital level, and possibly with finer resolution, particularly for critical care.

Forecasting carbapenem resistance from antimicrobial consumption surveillance: Lessons learnt from an OXA-48-producing Klebsiella pneumoniae outbreak in a West London renal unit

Highlights • We forecast the incidence rate of carbapenem resistance using antimicrobial usage data.• We assess the impact of an antimicrobial stewardship intervention.• Meropenem usage was highly correlated with the incidence of OXA-48-producing organisms.• While meropenem usage decreased significantly, amikacin usage increased in the renal unit.

Prevalence of healthcare device-associated infection using point prevalence surveys of antimicrobial prescribing and existing electronic data☆

This study extended a previously described method for the prevalence of healthcare-associated infection, based on point prevalence surveys of antimicrobial prescribing and electronic data, to estimate the prevalence of device-associated infections. In June 2009, the six-month point prevalence survey of antimicrobial prescribing was carried out in accordance with the European Surveillance of Antimicrobial Consumption Protocol. For patients receiving antimicrobials the presence of devices was recorded. A census on device use was carried out concurrently in the relevant hospitals. We selected patients receiving antimicrobials, started >48h after admission and who had a device, or who were without a device but were receiving antimicrobials for the treatment of bloodstream infection, urinary tract infection, or pneumonia. From existing positive microbiological and radiology reports, these patients were assessed for the presence of device-associated infection according to specified definitions. Of 1354 patients surveyed, 253 (19%) were receiving antimicrobial for treatment; of these, 189 also had devices and 172 (only 13% of all patients surveyed) needed individual assessment for the presence of device-associated infection. It took about 5min per patient to check electronic microbiology and/or radiology reports. Twenty-three patients met the criteria for device-associated infection. The prevalence of catheter-associated urinary tract infection, central-line-associated bloodstream infection, local vascular access infection, and ventilator-associated pneumonia was 3.9%, 3.1%, 3.8% and 11.6%, respectively. This is a simple method, which can be adopted in other hospitals, to estimate the prevalence of device-associated infection using pre-existing data.

Screening suspected cases for carbapenemase-producing Enterobacteriaceae, inclusion criteria and demand.

Journal of Infection - In Press.Proof corrected by the author Available online since lundi 22 juin 2015

Emergence and clonal spread of colistin resistance due to multiple mutational mechanisms in carbapenemase-producing Klebsiella pneumoniae in London

Carbapenemase-producing Enterobacteriaceae (CPE) are emerging worldwide, limiting therapeutic options. Mutational and plasmid-mediated mechanisms of colistin resistance have both been reported. The emergence and clonal spread of colistin resistance was analysed in 40 epidemiologically-related NDM-1 carbapenemase producing Klebsiella pneumoniae isolates identified during an outbreak in a group of London hospitals. Isolates from July 2014 to October 2015 were tested for colistin susceptibility using agar dilution, and characterised by whole genome sequencing (WGS). Colistin resistance was detected in 25/38 (65.8%) cases for which colistin susceptibility was tested. WGS found that three potential mechanisms of colistin resistance had emerged separately, two due to different mutations in mgrB, and one due to a mutation in phoQ, with onward transmission of two distinct colistin-resistant variants, resulting in two sub-clones associated with transmission at separate hospitals. A high rate of colistin resistance (66%) emerged over a 10 month period. WGS demonstrated that mutational colistin resistance emerged three times during the outbreak, with transmission of two colistin-resistant variants.

Management of ganciclovir-resistant cytomegalovirus retinitis in HIV infection in the era of antiretroviral therapy.

The incidence of cytomegalovirus retinitis has decreased significantly since the advent of antiretroviral therapy. However, it remains an important problem in both the developed and developing worlds. Furthermore, long-term antiviral suppression is associated with a significant increase in viral resistance. We present the case of a 46-year-old man who developed cytomegalovirus retinitis one year after being diagnosed with HIV. While he initially demonstrated an excellent clinical response to ganciclovir, his cytomegalovirus viral load remained persistently elevated. Over the subsequent years, his virus developed ganciclovir resistance with a concomitant deterioration in his visual acuity. He responded poorly to salvage therapy with foscarnet and cidofovir. This case highlights the ongoing difficulty of managing cytomegalovirus disease nearly two decades into the era of antiretroviral therapy and underlines the need to develop new treatment strategies.

Vancomycin therapy in secondary care; investigating factors that impact therapeutic target attainment.

250 Manuscript: 1109 M AN US CR IP T AC CE PT ED ACCEPTED MANUSCRIPT 3 Dear editor, We read with interest the article by Valencia-Rey a nd colleagues who investigated the role of vancomycin for empirical therapy in coagulase negat ive staphylococcal blood stream infections. 1 For vancomycin therapy outside of critical care there a re limited data describing vancomycin pharmacokinetics. We undertook a retrospective inve stigation of dosing of vancomycin in the noncritically ill patients managed across three Univer sity hospitals in London. Using patient TDM data, the aim was to build a population pharmacokinetic m odel to estimate population parameters and identify key factors associated with target attainm e t. Routinely collected data from two prospective hospi tal wide audits of vancomycin therapy in the noncritical care setting was included. Data on patient demographics, infection parameters, biochemical results, antimicrobial treatment, and TDM data was extracted for analysis. Patients on renal replacement therapy were excluded. Age, gender, eth nicity, total body weight (TBW), ideal body weight (IBW), lean body weight (LBW), body mass ind ex (BMI), glomerular filtration rate (GFR; Modification of Diet in Renal Disease [MDRD] formul a), and creatinine clearance (CrCL; estimated using the Cockcroft-Gault equation using TBW and IB W) were all collected or calculated for investigation as model covariates. An NPAG populati on pharmacokinetic algorithm embedded in the program; Pmetrics within R (LAPKB, CA, USA) , was used to estimate the population pharmacokinetics of vancomycin. Comparison between oneand two-compartment models w re performed and covariates were tested within the model. Statistical significance was asse s ed by comparison of twice the log-likelihood values against a chi-squared distribution with the appropriate number of degrees of freedom, depending on the number of parameter values within each model. Population parameters were estimated and assessed. Individual patient 24-hour steady state AUC was calculated. Given a paucity of minimum inhibitory concentration (MIC) data, the European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoints were us d to estimate the mean MIC for staphylococcal infections. For estimates of individual AUC:MIC ratio, an MIC of 1mg/L was selected as the M AN US CR IP T AC CE PT ED ACCEPTED MANUSCRIPT 4 breakpoint. A target AUC:MIC ratio ≥400 was defined as appropriate –. Monte-Carlo simulation was performed (1000 patients) and probability of ta rget attainment (PTA) estimated for eight dosing regimens commonly used in clinical practice at diff erent MIC values (0.25 to 8). Following this, simulations were performed for different weight ran ges (45-75kg & 75-150kg) to investigate the effect of weight on AUC:MIC ratios across dosing re gimens (n=2000). Statistical analysis was performed in SPSS 22.0 (IBM, NY, USA). Ethical app roval was not required for this study as only routinely collected data were included. 79 patients were identified on vancomycin therapy. 30 individuals receiving vancomycin therapy for known or suspected staphylococcal infections were s el cted for inclusion. Of the 49 patients excluded, 42 had covariate data missing and 7 were on renal replacement therapy. Median age (range) of the included subjects was 60 (21-87) yea rs, with the majority being male (18/30, 60%). Table 1 summarises the key population parameters and indica tions for therapy. Median (IQR) GFR was 75.85 (38.8-107.8)ml/min/1.73m . Patients received a median (range) of 1500mg (500 -30 mg) vancomycin 24-hourly and had a median (range) of 3. 5(1-6) TDM samples taken during the observation period. Local target TDM plasma concent ration (10-15mg/L or 15-20mg/L in severe or deep seated infections) was reached in 26/30 (87%) of cases. Three of the four individuals not reaching these targets were obese (BMI >30kg/m ), with the fourth classified as overweight (BMI = 26kg/m). A two-compartment model produced optimal observed-v rsus-predicted fit for the individual data extracted giving an r 2 of 0.9. TBW and GFR were included as covariates to the model using an allometric scaling and linear association, respecti v ly. These significantly improved the model (-2log likelihood = 602.4 to 590.2 and AIC = 612.9 to 600. 8). Population parameter estimates and individual steady state 24-hour area under the curve (AUC) est imations are described in Table 1. Overall, mean (SD) AUC:MIC ratio was 471 (163). 19/30 (63%) pati ents had AUC:MIC ratio ≥400. Of these, 3/19 (16%) had AUC values >700, which has been associate d with greater risk of toxicity . Of the 11/30 (37%) not meeting the target AUC:MIC ratio, 5/11 (4 5%) were obese. On comparison of obese versus M AN US CR IP T AC CE PT ED ACCEPTED MANUSCRIPT 5 non-obese subjects, the AUC:MIC ratio were statisti cally significant (mean (SD) = 320 (74) versus 509 (157); p<0.01). The GFR of those failing to attain a AUC:MIC rat io ≥400 were similar to those attaining the target (mean (SD) = 72 (36) vs. 79 ( 47);p=0.70). Monte-Carlo simulation and PTA estimation for a ran ge of dosing schedules of vancomycin from 500mg 24-hourly to 2000mg 12-hourly were performed (Figure 1). The PTA target was an AUC ≥400 simulated for a range of MIC values (from 0.25 to 8.0). Simulation of PTA demonstrated that for dosing regimens commonly used within our hospit al guidelines, doses of 1000mg twice daily are required for greater than 80% probability of attain ing a AUC:MIC ratio of >400 when the MIC is 1mg/L. Moreover, with an MIC of 2mg/L only 2000mg t wice daily would achieve greater than 80% PTA. Analysis of the effect of TBW on the AUC:MIC o f patients simulated across a range of dosing regimens at TBW 45-75kg (n=1000) and 75-150kg (n=10 00) was performed. Median (IQR) AUC:MIC was 464 (254-879) and 391 (217-748), respec tiv ly for both groups (p<0.01). In conclusion obesity was associated with significa nt likelihood of failing to attain AUC:MIC ratio of >400 in this real-world cohort of non-critical care patients from secondary care. This finding was replicated on the Monte-Carlo simulation of high ve rsus low TBW individuals receiving a range of dosing regimens. With approximately 20% of UK in-pa tients now classified obese, this observation warrants further investigation to define the direct clinical implications of significantly lower targe t attainment on patient outcomes. Furthermore, for d ifficult to treat infections with an organism with an MIC of 2mg/L or greater, vancomycin may not be a n effective agent, given its low likelihood of target attainment at the majority of dosing strateg ies modelled and the potential risks of toxicity at higher doses. The significant differences in AUC:MI C target attainment reported within this study highlights the need to conduct in-depth analysis of similar populations to define the direct clinical implications of significantly lower target attainme nt on patient outcomes. Estimation of AUC and incorporation of pharmacodynamic indices, such as A UC:MIC may be an intermediate method of optimising therapy for these individuals. However, in the long term, alternative approaches to dosing high risk individuals is required with computer ass isted approaches, such as iterative learning contro l M AN US CR IP T AC CE PT ED ACCEPTED MANUSCRIPT 6 providing a potential solution to dose optimisation in such cases . Research must focus on investigating novel pharmacodynamic vancomycin targ ets and combination therapy to better describe the effects of potentially synergistic regimes. We now plan to undertake larger prospective work to investigate direct patient outcomes with observed d ifferences in PK-PD parameters within obese populations in order to review standard intravenous vancomycin dosing recommendations in infections.

Cold autoimmune hemolytic anemia secondary to atypical pneumonia.

A 44-year-old Caucasian woman presented with a 10-day history of dry cough, exertional dyspnea, myalgia, and fever. Symptoms persisted despite having completed a course of amoxicillin and erythromycin. She was an ex-smoker of 15 pack years living in a rented accommodation with mold and with bird exposure, having a canary as a pet. On presentation she was hemodynamically stable and afebrile with a normal respiratory rate, an oxygen saturation of 96% on air, and sinus tachycardia of 98 beats per minute. There were fine crackles throughout the lung fields and right upper zone reduced air entry with increased vocal resonance. The spleen tip was palpable. Chest radiography showed a focal area of patchy consolidation in the right upper lobe. Her blood count on admission showed a raised white cell count of 16 3 10/L and platelet count of 494 3 10/L. She was anemic with a hemoglobin concentration of 84 g/L, an MCV of 77 fL, and a reticulocyte count of 176 3 10/L. Renal and liver function (including total bilirubin) were normal, with a raised C-reactive protein of 34 mg/L (normal range <6) and a high lactate dehydrogenase of 639 iu/L (125–243). An automated differential count suggested monocytosis but a manual differential count showed neutrophils 8.3 3 10/L, lymphocytes 6.1 3 10/L, and monocytes 0.6 3 10/L. Her blood film showed red cell agglutinates, polychromatic macrocytes, nucleated red blood cells, and atypical lymphocytes. Some of the lymphocytes were large with increased cytoplasmic basophilia (left Image). Others had plasmacytoid features with an eccentric nucleus and a Golgi zone (right Image). There were occasional examples of erythrophagocytosis. A direct antiglobulin test was positive for C3d. On the basis of the clinical and the hematological findings, cold-antibody-induced hemolytic anemia secondary to Mycoplasma pneumoniae infection was suspected. This was confirmed by demonstration of both immunoglobulin (Ig) G and IgM antibodies, indicative of a recent infection. Polymerase chain reaction (PCR) for Epstein–Barr virus (EBV) was negative, as were laboratory tests for multiple other organisms. A computed tomography (CT) scan showed “tree in bud” airspace consolidation and associated mediastinal lymphadenopathy. The patient made a rapid recovery with intravenous co-amoxiclav and oral clarithromycin. Subsequent imaging showed complete resolution of radiographic abnormalities. The combination of cold autoimmune hemolytic anemia and atypical lymphocytes suggests either EBV or mycoplasma infection. As in this patient, consideration of the clinical as well as hematological features can indicate the correct diagnosis.