Journal of Hospital Infection - Articles in Press

Healthcare-associated infection in Irish long-term care facilities: results from the First National Prevalence Study - Corrected Proof

M. Cotter, S. Donlon, F. Roche, H. Byrne, F. Fitzpatrick — 2012-02-06

Summary: Background: Prevalence of healthcare-associated infection (HCAI) and antimicrobial use in Irish long-term care facilities (LTCFs) has never been studied.Aim: To collect baseline data on HCAI prevalence and antibiotic use in Irish LTCFs to inform national LTCF policy and plan future HCAI prevention programmes.Methods: A prevalence study of HCAI and antibiotic use was undertaken in Irish LTCFs. Participation was voluntary. Data on HCAI risk factors, signs and symptoms of infection and antimicrobial use were collected prospectively on a single day in each institution.Findings: Sixty-nine Irish LTCFs participated and 4170 eligible residents were surveyed; 472 (11.3%) had signs/symptoms of infection (266, 6.4%) and/or were on antibiotics (426, 10.2%). A third of residents (1430, 34.3%) were aged ≥85 years and more than half disorientated (2110, 50.6%) with impaired mobility (2101, 50.4%). HCAI prevalence was 3.7% (range: 0–22.2%). The most common HCAI was urinary tract infection (UTI) (62 residents, 40% of HCAI). Presence of a urinary catheter was associated with UTI (P < 0.0000001). Antibiotics were prescribed for treatment (262 residents, 57.8%) and prophylaxis (182 residents, 40.2%) of infection. The most common indication for prophylaxis was UTI prevention (35.8% of total prescriptions). Fourteen (10.2%) residents on UTI prophylaxis had a urinary catheter. The most common indications for therapy included respiratory tract infections (35.1%), UTI (32.1%) and skin infection (21.8%).Conclusion: This study highlights the frequency of prophylactic antimicrobial prescribing and provides an important baseline to inform future preventive strategies.

Efficacy, efficiency and safety aspects of hydrogen peroxide vapour and aerosolized hydrogen peroxide room disinfection systems - Corrected Proof

T.Y. Fu, P. Gent, V. Kumar — 2012-02-06

Summary: Background: This was a head-to-head comparison of two hydrogen-peroxide-based room decontamination systems.Aim: To compare the efficacy, efficiency and safety of hydrogen peroxide vapour (HPV; Clarus R, Bioquell, Andover, UK) and aerosolized hydrogen peroxide (aHP; SR2, Sterinis, now supplied as Glosair, Advanced Sterilization Products (ASP), Johnson & Johnson Medical Ltd, Wokingham, UK) room disinfection systems.Method: Efficacy was tested using 4- and 6-log Geobacillus stearothermophilus biological indicators (BIs) and in-house prepared test discs containing approximately 106 meticillin-resistant Staphylococcus aureus (MRSA), Clostridium difficile and Acinetobacter baumannii. Safety was assessed by detecting leakage of hydrogen peroxide using a hand-held detector. Efficiency was assessed by measuring the level of hydrogen peroxide using a hand-held sensor at three locations inside the room, 2 h after the start of the cycles.Findings: HPV generally achieved a 6-log reduction, whereas aHP generally achieved less than a 4-log reduction on the BIs and in-house prepared test discs. Uneven distribution was evident for the aHP system but not the HPV system. Hydrogen peroxide leakage during aHP cycles with the door unsealed, as per the manufacturer’s operating manual, exceeded the short-term exposure limit (2 ppm) for more than 2 h. When the door was sealed with tape, as per the HPV system, hydrogen peroxide leakage was <1 ppm for both systems. The mean concentration of hydrogen peroxide in the room 2 h after the cycle started was 1.3 [standard deviation (SD) 0.4] ppm and 2.8 (SD 0.8) ppm for the four HPV and aHP cycles, respectively. None of the readings were <2 ppm for the aHP cycles.Conclusion: The HPV system was safer, faster and more effective for biological inactivation.

Assessment of administrative data for evaluating the shifting acquisition of Clostridium difficile infection in England - Corrected Proof

M-H. Jen, S. Saxena, A. Bottle, R. Pollok, A. Holmes, P. Aylin — 2012-02-06

Summary: Background: Little is known about the acquisition of Clostridium difficile infection (CDI) and whether it represents hospital- or community-acquired infection.Aim: To test the feasibility and value of using national hospital admissions data from Hospital Episode Statistics to examine trends in CDI in England.Methods: Hospital Episode Statistics from the period 1997/98 to 2009/10 were used. Time trends were analysed using two different denominators of hospital activity: total admissions and total bed-days. We explored the impact of sociodemographic factors, comorbidity and healthcare pathways on the risk of CDI.Findings: CDI rates per admission and per bed-days increased from 1997/98 to 2006/07, then decreased significantly by >50% from 2008/9 and 2009/10. This pattern was similar for patients regardless of probable source of infection but the proportion of probable community-acquired CDI cases rose steadily from 7% in 1997/98 to 13% in 2009/10. CDI rates were higher among older patients (odds ratio: >65 years, 10.9), those with more comorbid conditions (odds ratio for Charlson index: >5, 5.6), and among patients admitted as an emergency compared with elective admissions, but no relationship was found with deprivation score.Conclusion: Our findings support not only the falling trend in CDI found in the national mandatory surveillance scheme from the Health Protection Agency, but a growing proportion of CDI presenting on admission with no evidence of prior hospital exposure in the previous 90 days. We suggest that these may be community-acquired CDI cases.

Pre-operative skin preparation practices: results of the 2007 French national assessment - Corrected Proof

2012-02-06

Summary: Background: Pre-operative skin preparation, aimed at reducing the endogenous microbial flora, is one of the main preventive measures employed to decrease the likelihood of surgical site infection. National recommendations on pre-operative management of infection risks were issued in France in 2004.Aim: To assess compliance with the French national guidelines for pre-operative skin preparation in 2007.Methods: A prospective audit was undertaken in French hospitals through interviews with patients and staff, and observation of professional practice. Compliance with five major criteria selected from the guidelines was studied: patient information, pre-operative showering, pre-operative hair removal, surgical site disinfection and documentation of these procedures.Findings: Data for 41,188 patients from all specialties at 609 facilities were analysed. Patients were issued with information about pre-operative showering in 88.2% of cases [95% confidence interval (CI) 87.9–88.5]. The recommended procedure for pre-operative showering, including hairwashing, with an antiseptic skin wash solution was followed by 70.3% of patients (95% CI 69.9–70.8); this percentage was higher when patients had received appropriate information (P < 0.001). Compliance with hair removal procedures was observed in 91.5% of cases (95% CI 91.2–91.8), and compliance with surgical site disinfection recommendations was observed in 25,529 cases (62.0%, 95% CI 61.5–62.5). The following documentary evidence was found: information given to patient, 35.6% of cases; pre-operative surgical hygiene, 82.3% of cases; and pre-operative site disinfection, 71.7% of cases.Conclusion: The essential content of the French guidelines seems to be understood, but reminders need to be issued. Some recommendations may need to be adapted for certain specialties.

Delivering the infection control message: a communication challenge - Corrected Proof

C. Farrugia, M.A. Borg — 2012-02-06

Summary: Background: Hospital infection control teams (ICTs) worldwide face a constant challenge to deliver timely information to audiences with specific and diverse information needs.Aim: Our study investigated the communication flow between the ICT and healthcare workers (HCWs) at St Luke’s Hospital, Malta, using an exploratory descriptive research design.Method: Using a self-administered questionnaire, data were collected from a stratified random sample of nurses (N = 143) and doctors (N = 63) working within inpatient wards; a response rate of 97% was achieved.Findings: HCWs felt most comfortable receiving information from members of their same profession. Information transfer was mainly vertically up and down the hierarchy. Respondents preferred to receive information through educational activities (35%, N = 69) and through face-to-face contact (31%, N = 62). Electronic channels (e-mail and intranet) were ranked least preferable; however, only 41% (N = 81) had regular access to a computer system at work. The majority of respondents 91% (N = 181) trusted the information by the ICT and 60% (N = 118) regarded it as being consistent. Nevertheless, this did not guarantee constant compliance; 54% (N = 106) of respondents implemented IC measures only when they perceived a risk for their health. Greater presence of the ICT on the wards was recommended.Conclusion: Notwithstanding the electronic age, the study confirms that face-to-face contact between ICTs and HCWs remains the most effective way of disseminating IC information.

Prospective study on the effect of shirt sleeves and ties on the transmission of bacteria to patients - Corrected Proof

R.L. Weber, P.D. Khan, R.C. Fader, R.A. Weber — 2012-02-03

Summary: Costs associated with hospital-acquired infections lead to policies aimed at decreasing their incidence. Clothing restrictions are often implemented in response, but they are based on little scientific evidence. This study is a prospective, controlled investigation of the effect of shirt sleeves and ties on the transmission of bacteria from doctors to patients. Results show that wearing an unsecured tie results in greater transmission, but that sleeve length does not affect transmission rate. The design is a possible model for further controlled experiments to fill the evidence gap regarding the transmission of micro-organisms from healthcare workers to patients.

Are commercially available Geobacillus stearothermophilus biological indicators an appropriate standard for hydrogen peroxide vapour systems in hospitals? - Corrected Proof

J.A. Otter, S. Yezli — 2012-02-02

We read with interest the article by Pottage et al. evaluating the relative susceptibility of meticillin-resistant Staphylococcus aureus (MRSA) and Geobacillus stearothermophilus biological indicators (BIs) to a hydrogen peroxide vapour (HPV) system. Their principal finding, that catalase-positive MRSA is less susceptible to HPV than the metabolically inert G. stearothermophilus BIs, came as no surprise. A study published in 2009 showed that catalase-positive bacteria including MRSA were less susceptible to an HPV system than catalase-negative bacteria (Enterococcus faecium and E. faecalis) and Clostridium difficile spores tested at a similar inoculum. In fact, the catalase-positive Gram-negative bacteria (Acinetobacter baumannii and Klebsiella pneumoniae) were even less susceptible to HPV than MRSA. These findings were attributed to the activity of catalase.

Where do hands go? An audit of sequential hand-touch events on a hospital ward - Corrected Proof

S.J. Smith, V. Young, C. Robertson, S.J. Dancer — 2012-02-01

Summary: Background: Reservoirs of pathogens could establish themselves at forgotten sites on a ward, posing a continued risk for transmission to patients via unwashed hands.Aim: To track potential spread of organisms between surfaces and patients, and to gain a greater understanding into transmission pathways of pathogens during patient care.Methods: Hand-touch activities were audited covertly for 40 × 30 min sessions during summer and winter, and included hand hygiene on entry; contact with near-patient sites; patient contact; contact with clinical equipment; hand hygiene on exit; and contact with sites outside the room.Findings: There were 104 entries overall: 77 clinical staff (59 nurses; 18 doctors), 21 domestic staff, one pharmacist and five relatives. Hand-hygiene compliance among clinical staff before and after entry was 25% (38/154), with higher compliance during 20 summer periods [47%; 95% confidence interval (CI): 35.6–58.8] than during 20 winter periods (7%; 95% CI: 3.2–14.4; P < 0.0001). More than half of the staff (58%; 45/77) touched the patient. Staff were more likely to clean their hands prior to contact with a patient [odds ratio (OR): 3.44; 95% CI: 0.94–16.0); P = 0.059] and sites beside the patient (OR: 6.76; 95% CI: 1.40–65.77; P = 0.0067). Nearly half (48%; 37/77) handled patient notes and 25% touched the bed. Most frequently handled equipment inside the room were intravenous drip (30%) and blood pressure stand (13%), and computer (26%), notes trolley (23%) and telephone (21%) outside the room.Conclusion: Hand-hygiene compliance remains poor during covert observation; understanding the most frequent interactions between hands and surfaces could target sites for cleaning.

High level disinfection of a home care device; to boil or not to boil? - Corrected Proof

K.L. Winthrop, N. Homestead — 2012-01-25

Summary: We developed a percutaneous electrical transducer for home therapy of chronic pain, a device that requires high level disinfection between uses. The utility of boiling water to provide high level disinfection was evaluated by inoculating transducer pads with potential skin pathogens (Staphylococcus aureus, Mycobacterium terrae, Pseudomonas aeruginosa, Candida albicans) and subjecting them to full immersion in water boiling at 4200 feet elevation (95°C). Log10 reductions in colony-forming units (cfu) at 10min were 7.1, >6.3 and >5.5 for S. aureus, P. aeruginosa and C. albicans, respectively, but only 4.6 for M. terrae. At 15min the reductions had increased to 7.5, >6.8, >6.6 and >7.5cfu, respectively.

Effect of surface coating and finish upon the cleanability of bed rails and the spread of Staphylococcus aureus - Corrected Proof

S. Ali, G. Moore, A.P.R. Wilson — 2012-01-24

Summary: Background: Bacterial reservoirs in the near-patient environment are likely vectors of healthcare-acquired infection.Aim: To conduct a laboratory-based study to confirm a previous clinical finding of higher numbers of bacteria on plastic than on painted steel bed rails.Methods: Six different surfaces were inoculated with Staphylococcus aureus suspended in a range of synthetic soils. Aerobic colony counts and ATP bioluminescence were used to assess the efficacy of cleaning with microfibre cloths and antibacterial wipes. The ease with which S. aureus was transferred between fingertips and each bed rail was also investigated.Findings: Antibacterial wipes reduced bacterial numbers to below detectable levels on all rails but were less effective than microfibre cloths in removing organic debris. Surfaces that were comparatively easy to clean were more likely to transfer S. aureus on contact. If inadequately disinfected these rails could pose the greatest risk in terms of cross-transmission. In the absence of contaminating soil, bacterial transfer from fingertips to rail ranged from 38% to 64%. Transfer from rail to fingertip ranged from 22% to 38%. Surface material and rugosity were important factors in determining cleanability and transfer rate. However, the presence of organic soils affected bacterial transfer from all bed rails regardless of material or finish.Conclusion: Bed rails can become heavily contaminated. Regular wiping with antibacterial wipes could be a cost-effective means of maintaining low numbers of bacteria near to the patient. To minimize the risk of cross-transmission, cleaning protocols should be validated to ensure effective removal of microbial and non-microbial surface contamination.

Legionella pneumophila contamination in a steam towel warmer in a hospital setting - Corrected Proof

F. Higa, M. Koide, A. Haroon, S. Haranaga, T. Yamashiro, M. Tateyama, J. Fujita — 2012-01-24

Summary: For prevention of nosocomial legionellosis, environmental investigation to identify possible infectious sources is essential. An environmental study in a ward of our hospital revealed that a steam towel warmer was contaminated with legionella whereas no legionella was detected in tap water supplies and shower heads. Water in the apparatus may be a reservoir of legionella. We abandoned the use of all steam towel warmers in our hospital. Based on this finding, we recommend that steam towel warmers in hospital settings be avoided. Otherwise, the apparatus should be drained, cleaned and dried every day.

Blood-borne virus transmission in healthcare settings in Ireland: review of patient notification exercises 1997–2011 - Corrected Proof

S. Donohue, L. Thornton, K. Kelleher — 2012-01-24

Summary: A review of patient notification exercises (PNEs) carried out in Ireland between 1997 and 2011 to investigate potential exposure to blood-borne viruses (BBVs) in healthcare settings was undertaken to inform future policy and practice. A questionnaire was sent to key informants in the health services to identify all relevant PNEs. Structured interviews were conducted with key investigators, and available documentation was examined. Ten BBV-related PNEs were identified. Despite testing over 2000 patients, only one case of transmission was found. However, in-depth local investigations before undertaking the PNEs identified six cases of healthcare-associated transmission.

Response: Providing strong evidence of nosocomial outbreak of hepatitis B virus infection - Corrected Proof

K. Burns — 2012-01-19

In response to the letter submitted by Pourkarim et al., I would like to provide further clarification regarding the extensive multi-disciplinary investigation which took place into a nosocomial outbreak of hepatitis B virus (HBV) infection in the south east of Ireland during 2005 and 2006. The investigation combined epidemiological and molecular methods to arrive at the conclusions described in the article.

Reply to: ‘Hand hygiene: are we trying to make the patient the fail safe system?’ - Corrected Proof

C.L. Packham — 2012-01-18

May I add to the exchange of letters regarding hand hygiene with an additional observation from someone who was himself a patient. When three years ago I was transferred from the emergency department to a ward, I was handed a card about ensuring that healthcare workers cleaned or sanitized their hands before touching me or any equipment that was being used on me. However, there was no guidance on my own personal hygiene standards. Yet some research has suggested that a significant number of cases of healthcare-acquired infection are caused by patients infecting themselves. Are we missing something here?

Nosocomial outbreak of Pantoea agglomerans bacteraemia associated with contaminated anticoagulant citrate dextrose solution: new name, old bug? - Corrected Proof

2012-01-16

Summary: We describe an outbreak investigation of Pantoea agglomerans bacteraemia associated with anticoagulant citrate–dextrose 46% (ACD) solution prepared in-house. A healthy man presented with septic shock during plasmapheresis for granulocyte donation. The solution used for priming and blood samples were sent for culture. Identification of the isolate to species level was performed by gyrB sequencing. Typing was performed by pulsed-field gel electrophoresis (PFGE). In total, eight cases were identified during a three-week period. P. agglomerans was also cultured from six ACD solution bags. Isolates from patients and ACD bags were identical by PFGE. All isolates were susceptible to ampicillin, cephazolin, gentamicin, ciprofloxacin, cefepime and imipenem.

Providing strong evidence of nosocomial outbreak of hepatitis B virus infection - Corrected Proof

S. Amini-Bavil-Olyaee, P. Maes, M. Van Ranst, M.R. Pourkarim — 2012-01-16

We read with great interest the article ‘Nosocomial outbreak of hepatitis B virus infection involving two hospitals in the Republic of Ireland’ by Burns et al. The authors describe how five patients were hospitalized and subsequently infected with hepatitis B virus (HBV). They offer detailed descriptive evidence and have identified similar serological (HBeAg-negative) and virological (genotype D) markers in the patients. All these indications resulted in a hypothesis that a nosocomial outbreak may have happened. The authors conducted phylogenetic analysis based on 678 bp of HBV S gene to confirm a common source of infection with limited reference sequences. In their study, the mode of transmission was not determined precisely and they were suspicious that reusable needle/blood tube-holding devices and poor practice could have played a role in viral transmission between the infected cases. In our view, these findings might not warrant such a definitive conclusion.

Catheter application, insertion vein and length of ICU stay prior to insertion affect the risk of catheter-related bloodstream infection - Corrected Proof

T.I.I. van der Kooi, J.C. Wille, B.H.B. van Benthem — 2012-01-13

Summary: Background: The Dutch PREZIES surveillance scheme for catheter-related bloodstream infection (CR-BSI) collects data on infection rates and related risk factors.Aim: To evaluate risk factors for CR-BSI.Methods: Hospitals collected data for intensive care units (ICU) or for the entire hospital. All short-term central venous catheters (CVC), including Swan-Ganz catheters, present for ≥48h were surveyed, except in cases when bacteraemia was present at insertion. CVCs were monitored until infection, removal or death for up to 28 days. Data were collected on 3750 CVCs and 29,003 CVC-days.Findings: Of the CVCs surveyed, 1.6% [95% confidence interval (CI) 1.2–2.0] resulted in CR-BSI, representing 2.0/1000 CVC-days (95% CI 1.6–2.6). Multi-variate analysis revealed that the length of ICU stay prior to CVC insertion, insertion in the jugular or femoral vein, and use of the CVC to deliver total parenteral nutrition increased the risk of CR-BSI, whereas use of the CVC to deliver antibiotics decreased the risk of CR-BSI.Conclusion: Attention to these risks has the potential to reduce the incidence of CR-BSI.

Diploma in Hospital Infection Control - Corrected Proof

2012-01-11

We are pleased to announce that the following candidates have been awarded the Diploma in Hospital Infection Control by examination or Accreditation of Prior Experiential Learning (APEL).

A large exposure to Brucella melitensis in a diagnostic laboratory - Corrected Proof

2012-01-11

Summary: Background: Brucella species are easily transmitted by aerosols and can be acquired in the laboratory.Aim: To report the management of a large exposure to Brucella melitensis that occurred over six days in a hospital diagnostic laboratory.Methods: Fifty-one exposed staff were managed according to Centers for Disease Control and Prevention guidelines. A further 96 non-exposed laboratory staff were tested for seroprevalence. Testing was carried out using the Brucella sp. serum agglutination test.Findings: Twenty-seven people had high-risk exposure and 24 had low-risk exposure. High-risk staff were offered post-exposure prophylaxis. Twelve (44.4%) agreed to this, of whom eight (66.7%) completed the course. Overall compliance with serological follow-up at baseline, 2, 4, 6 weeks and 8 months was 45.9%. Despite this poor compliance there were no clinical brucellosis cases and no seroconversion in the 47.1% of staff tested at 8 months. Brucella sp. seroprevalence among all staff tested was 3/147 (2.0%).Conclusion: Lack of experience with Brucella spp. and lack of policies for handling potentially hazardous organisms contributed to this prolonged exposure. As compliance with current recommendations may be poor, the optimum frequency of serological follow-up and target groups for prophylaxis should be reassessed. Laboratories in low- or non-endemic areas must prepare for potential isolation of Brucella spp. The impact of human brucellosis in Malaysia requires further study.

Prevalence survey of healthcare-associated infections in Argentina; comparison with England, Wales, Northern Ireland and South Africa - Corrected Proof

2012-01-09

Summary: Background: Prevalence surveillance methodology is the systematic observation of the occurrence and distribution of healthcare-associated infections (HCAIs) so that appropriate actions can be taken.Aim: The objectives of a prevalence survey with an international validated methodology were to determine the prevalence of HCAIs for the first time in Argentina, and to provide data which could be used for international benchmarking.Methods: In 2008, an HCAI prevalence survey was carried out in 39 hospitals in seven of 23 provinces in Argentina, with methodology identical to that employed by the Hospital Infection Society in the third prevalence survey of HCAIs in acute hospitals in the British Isles. Data collected were processed and analysed at the Northern Ireland Healthcare-Associated Infection Surveillance Centre at Belfast.Findings: A total of 4249 patients were surveyed; 480 of these had at least one HCAI, resulting in a prevalence of 11.3% of patients. Male prevalence was 13.6% and female 9.0%. The most common HCAIs were pneumonia (3.3%), urinary tract infection (3.1%), surgical site infection (2.9%), primary bloodstream infection (1.5%), and soft tissue infections (1.2%). Among the 1027 patients who underwent surgery, the prevalence of surgical site infection was 10.2%. The prevalence of meticillin-resistant Staphylococcus aureus was 1.1%, accounting for 10.0% of all HCAI isolates. The results for Argentina show higher HCAI rates compared with corresponding findings for England, Wales, Northern Ireland and South Africa.Conclusion: This survey will contribute to the prioritization of resources and help to inform Departments of Health and hospitals in the continuing effort to reduce HCAIs.

Infection control ‘undercover’: a patient experience - Corrected Proof

S.J. Dancer — 2012-01-09

As a National Health Service microbiologist and former editor of the Journal of Hospital Infection, most of my professional time is spent managing nosocomial infection. Becoming a patient was not part of the job plan until a sudden accident on holiday compelled hospital admission. It was an opportunity to experience infection control in the real world and one that could not be ignored. Bedbound, and ‘undercover’, this patient witnessed the ward-based impact of infection control in hospitals in two countries.

Outcome of central venous catheter-related bacteraemia according to compliance with guidelines: experience with 91 episodes - Corrected Proof

2012-01-09

Summary: Background: Infection is a major complication associated with the use of central venous catheters. Guidelines for medical management of catheter-related bacteraemia have been published, but no study has assessed the appropriateness of physician practices.Aim: To assess medical practices in cases of central venous catheter-related bacteraemia (CRB) in a university hospital.Methods: Cases were recorded over a period of 12 months and their management was evaluated. All cases of positive blood cultures based on central venous catheter sampling were analysed, and episodes of CRB were determined in this group of patients. Medical management and patient outcome were analysed independently by two physicians.Findings: In all, 187 cases of positive blood culture were recorded and 91 cases of CRB were analysed. Systemic antimicrobial therapy was optimal in 56% of the episodes. In 51 episodes, catheter salvage was attempted, for 29 with an indication in agreement with the guidelines but without antibiotic-lock therapy in 20 episodes. The overall medical management was appropriate in 41.8% of the episodes. The overall cure rate was 72.5%. CRB-related death occurred in 5.5% of the episodes. Cure was associated with guideline compliance (P = 0.03) and with adaptation of systemic antimicrobial therapy (P < 0.01). Conservative treatment success was associated with compliance with the guidelines for the indication (P = 0.01).Conclusion: Medical management of CRB did not closely adhere to international guidelines. CRB outcome was associated with the appropriateness of this management, particularly when conservative treatment was attempted.

Final comment on letter by Allegranzi et al. - Corrected Proof

G. Kampf, C. Ostermeyer — 2011-12-12

We thank our respected colleagues for their interest and comments on our work and are grateful to learn that both World Health Organization (WHO)-recommended formulations will probably be changed in order to improve their efficacy, which has also been indicated by Suchomel et al.

Do staffing and workload levels influence the risk of new meticillin-resistant Staphylococcus aureus acquisitions in a well-resourced intensive care unit? - Corrected Proof

F. Kong, D. Cook, D.L. Paterson, M. Whitby, A. Clements — 2011-11-28

Summary: Background: There is limited knowledge of existing relationships between new meticillin-resistant Staphylococcus aureus (MRSA) acquisitions, staffing and workload levels in Australia.Aim: The temporal relationship between staffing and workload levels and new MRSA acquisitions was explored in an intensive care unit (ICU) of an Australian metropolitan tertiary hospital to determine whether these variables could be used as predictors of new MRSA acquisitions.Method: Sixty-one ICU patients with new laboratory-confirmed MRSA acquisition (colonizations or infections) were identified from January 2003 to December 2006. Univariate logistic regression models were constructed with the occurrence of one or more new MRSA acquisitions in the ICU in a given week as the binary outcome, and multiple staffing and bed occupancy variables (aggregated by week) as predictors, entered at lag intervals from zero to six weeks. A multivariate logistic regression used backward stepwise elimination, retaining variables with P < 0.20. A receiver operator characteristic (ROC) analysis cross-validated the model with data from January to December 2007.Findings: The final model contained two variables: total nursing hours (per 100) at a one-week lag [odds ratio: 0.90; confidence interval (CI): 0.90–0.91; P = 0.04] and the ratio of elective cancellations to elective admissions at a two-week lag (not significant). The area under the ROC was 0.46 (CI: 0.25–0.67).Conclusion: ICU staffing and workload indicators did not predict prospectively risk of new MRSA acquisition in any given week, possibly because of the ICU’s strict staffing policy, low overall MRSA acquisition rate and good compliance with infection control procedures.