Reduction in the incidence of hospital-acquired MRSA following the introduction of a chlorine dioxide 275 ppm based disinfecting agent in a district general hospital

Abstract

Background

Methicillin-resistant Staphylococcus aureus (MRSA) and Clostridium difficile are major nosocomial pathogens whose control relies on effective antimicrobial stewardship and infection control practices. This study evaluates the impact of a chlorine dioxide-based disinfectant (275 ppm) on the incidence of hospital-acquired (HA) MRSA and HA-Clostridium difficile infection (CDI) in a district general hospital.

Methods

This study was carried out from November 2009 to September 2013. From November 2009 to October 2011 sodium dichloroisocyanurate was used for routine environmental disinfection. In November 2011, this was changed to a chlorine dioxide 275 ppm based disinfectant. This product was introduced into the hospital in a phased manner with intensive training on its use provided to all nursing, nursing auxiliary and hotel services staff. The effect of this change on the incidence of HA-MRSA and HA-CDI was assessed using segmented regression analysis of interrupted time series. In addition, the potential cost savings as a result of this intervention were assessed.

Results

The HA-MRSA trend from November 2009 to October 2011 significantly increased (p=0.006). Following the introduction of the chlorine dioxide-based disinfectant there was significant decrease in the HA-MRSA trend, with the monthly incidence being reduced by 0.003 cases/100 bed days (p=0.001), equating to an average of four cases per month after 12 months of use This resulted in an annual potential cost saving of £276 000. No significant effect on the incidence of HA-CDI was observed (coefficient −0.03; p=0.873).

Conclusion

This study highlights the importance of effective environmental inanimate surface decontamination in controlling the spread of MRSA and the potential cost savings that can be achieved through decreasing HA-MRSA rates.

Introduction

It is well evidenced that the hospital environment is a reservoir of resistant micro-organisms. Recently, attention has focussed on the vital role of environmental cleaning and disinfection in minimising the spread of these organisms to hospital patients where they can cause infection.^{1 2}

It has been demonstrated that Clostridium difficile can be transmitted from patient to patient via the orofaecal route and the spores can persist in the hospital environment for up to 5 months.³ In addition, methicillin-resistant Staphylococcus aureus (MRSA) can be shed to varying degrees from the skin scales of positive patients into the near patient environment, thereby contaminating inanimate objects such as the bed, the locker, the bed table and other patient care equipment in the vicinity. These objects become reservoirs for potential transmission through indirect contact via the hands of healthcare workers and subsequent direct contact with another patient, thus establishing a chain of transmission. This has been demonstrated in an intensive care unit (ICU) setting where a small, yet significant risk of acquiring MRSA from prior room occupants was observed.⁴ It has been reported that MRSA can persist for between 7 days and 7 months on environmental surfaces.⁵ This can be explained by the presence of biofilms, which render the organisms up to 1000 times less susceptible to biocides compared with their planktonic counterparts. As organisms in biofilms are also more resistant to physical removal through cleaning, their eradication poses a major infection control challenge.⁶

A range of disinfectant agents are available for use in hospitals. These include alcohol and chlorine releasing agents such as hypochlorites.⁷ Several studies have demonstrated the effectiveness of alcohol against MRSA. However, this product is ineffective against C. difficile spores. Consequently, it has been suggested that alcohol may select for the proliferation and spread of this latter organism.^8–10 Hypochlorites are commonly used in healthcare settings for environmental inanimate surface decontamination; however, at high concentrations (>500 ppm) they are corrosive to metals and can cause oesophageal, oropharyngeal and ocular irritation.⁷ Consequently, chlorine dioxide (ClO₂) has been used by some hospitals as an alternative. This agent has demonstrated effectiveness against a variety of multidrug-resistant organisms such as Pseudomonas aeruginosa, Acinetobacter baumannii, MRSA, methicillin-sensitive S. aureus and C. difficile.^11–13 The superior effectiveness of this agent both in aerosolised form and in solution over hypochlorites has been demonstrated in the laboratory setting. The use of aerosolised ClO₂ resulted in a significantly greater reduction in the number of MRSA colony-forming units (CFUs) isolated from stainless steel, fabric and plastic surfaces compared with sodium hypochlorite.¹² Supporting this, the addition of 100 ppm ClO₂ to a suspension of MRSA was shown to reduce the number of CFUs from 10⁷ to below the detection limit after a 15 s incubation period. This compares favourably to sodium hypochlorite, where 100 ppm of this agent had no effect on the number of MRSA CFUs after a 120 s incubation period; however, the strength of the disinfectant solutions used in this study were not reflective of that used in the hospital environment.¹¹ The effectiveness of ClO₂ against C. difficile spores in a laboratory setting has also been demonstrated.¹³ This agent compared favourably with hypochlorites by achieving the target 3 log₁₀ reduction in C. difficile spores after a 10⁶ challenge. This reduction was achieved after 1 min on both clean and dirty surfaces; however, hypochlorites only achieved a 3 log₁₀ reduction in C. difficile spores after 60 min on clean and dirty surfaces. As all of these studies were carried out in the laboratory setting, it is not possible to determine whether these results will be replicated in the hospital environment.

Following the control of an outbreak of C. difficile infection (CDI) in Antrim Hospital, the use of sodium dichloroisocyanurate (NaDCC) (when reconstituted forms a hypochlorous acid solution of 1000 ppm available chlorine) was replaced with a new ClO₂ 275 ppm based disinfectant for routine environmental inanimate surface decontamination.¹⁴ This change occurred in a phased manner between April and October 2011. The aim of this study was to determine the effect of the introduction of ClO₂ on the incidence of hospital-acquired (HA)-MRSA and HA-CDI.

Methods

Setting and study period

This study took place in Antrim Hospital, a large district general hospital with 426 beds and serving a population of ∼450 000. This project was approved by the Northern Trust Research Governance Committee as a quality improvement study.

Between November 2009 and April 2011, NaDCC 1000 ppm was used for routine environmental disinfection in all wards in Antrim Hospital. Between April and October 2011, this product was changed to ClO₂ 275 ppm. This was applied to surfaces and then left to air dry, or wiped after 5 min. Unlike NaDCC, it did not require washing off with detergent and water.

The change from NaDCC to ClO₂ was effective throughout the whole hospital from November 2011; however, the phased introduction of this product commenced in April 2011, whereby wards started using the product once all staff were trained in its correct use. In 1996, an ongoing programme of Infection Prevention and Control (IPC) training and IPC audits of clinical practice and environmental cleanliness were introduced as standard practice for the IPC Department within the Trust. All clinical areas were audited at least annually and more frequently, if there were issues causing concern, to ensure these issues were addressed. In addition, mandatory IPC training was required to be undertaken by all clinical staff. These programmes continued throughout the study period as part of the normal IPC Department work. Since 2008, monthly environmental cleanliness audits were carried out on each ward by hotel services staff to support the in-depth audits undertaken in the IPC programme. Between April 2011 and October 2011 a collaborative approach was employed between company representatives and the Trust IPC team to ensure that all staff received adequate training on the correct use of this new disinfectant. This involved setting up an implementation team who met on a monthly basis from the initial introduction of the ClO₂, and then at 2-monthly intervals once the product was in full use. The membership of the implementation team included a Consultant Microbiologist, Antimicrobial Pharmacist, Head of Pharmacy and Medicines Management, Lead IPC Nurse, Hotel Services Supervisor, the Trust Health and Safety Advisor and company representatives. The remit of this team was to identify and address any problems resulting from the use of this new disinfectant.

Training was delivered to nursing, auxiliary nursing and hotel services staff in the sluice room on each ward on the correct use of ClO₂. This was repeated at regular intervals until ∼75% of staff on each ward were trained. Once 75% of staff on all wards were trained, link trainers were identified who were responsible for training any new staff who came into the Trust or any staff who missed the initial round of training sessions.

Routine cleaning of the environment, clinical and patient care equipment was carried out once daily. All areas and pieces of equipment associated with patients with CDI were cleaned using detergent and water, followed by a disinfectant agent twice daily. All small pieces of equipment were cleaned using the disinfectant agent between patient use. Prior to November 2011, the disinfectant agent was NaDCC. From November 2011, this was ClO₂ 275 ppm.

A programme of antibiotic stewardship activities have been in place in this hospital since 2008 as a result of a CDI outbreak.¹³ These involved weekly audits on adherence to the antibiotic policy, restrictions on the use of fluoroquinolones and cephalosporins and audits on the appropriateness of the use of restricted antibiotics. The results of all audits were communicated to prescribers on a monthly basis. Throughout the study period there were no changes to these antibiotic stewardship activities or the frequency at which they were carried out.

The study period was from November 2009 to September 2013. The incidence of HA-MRSA and HA-CDI during the 23 months after the introduction of ClO₂ (November 2011–September 2013) was compared with the incidence of HA-MRSA and HA-CDI during the period when NaDCC was in use (October 2009–October 2011), to determine whether this change had a significant impact on the incidence of infection or colonisation with these organisms.

Microbiology and pharmacy data

The monthly incidence of HA-MRSA and HA-CDI was obtained from the clinical microbiology information system. HA-MRSA cases were defined as anyone testing positive for MRSA >48 h after admission.¹⁵ HA-MRSA cases included positive tests from screening swabs and clinical samples. Only the first isolate of MRSA for each patient during their admission period was included in the analysis. The monthly incidence of HA-MRSA was adjusted per 100 occupied bed days to take account of hospital activity. All samples were processed according to routine microbiology procedures using BioMérieux MRSA differential media (BioMérieux S.A, Marcy l'Etoile, France). Suspicious colonies were further identified as MRSA by a positive coagulase test. Coagulase positive isolates and their antimicrobial susceptibility were identified using the Vitek 2 system (BioMérieux, France).

HA-CDI was defined as anyone diagnosed with CDI >48 h after admission, or <48 h since admission, but with an admission to Antrim Hospital in the previous 4 weeks.¹⁵ The microbiology laboratory identified C. difficile-positive cases using the C.DIFF QUIKCHEK COMPLETE kit (Techlab, USA), which detected the presence of toxin A and B and glutamate dehydrogenase (GDH). Samples that tested GDH-positive and toxin-negative or toxin-positive and GDH-negative underwent further PCR testing. Samples that were toxin-negative, GDH- and PCR-positive were classified as toxigenic C. difficile. For the purposes of this study, only toxin-positive cases were included. Repeat samples <28 days apart were excluded.

The monthly use of high-risk antibiotics for the spread of MRSA (fluoroquinolones, co-amoxiclav, piperacillin/tazobactam and clarithromycin) were obtained from the Pharmacy Information System.^{10 16} These were expressed as the number of Defined Daily Doses (WHO ATC V.10) and adjusted per 100 occupied bed days.¹⁷

The volumes of alcohol-based handrub, NaDCC-based and ClO₂-based disinfecting agents supplied to each adult inpatient ward in the hospital were obtained from the Pharmacy Information System. This was also adjusted to the number of litres, tablets and sachets per 100 occupied bed days.

Economic analysis

The costs associated with the use of the ClO₂ disinfectant agent and staff training were identified. The impact of the intervention 12 months after the introduction of ClO₂ was calculated using segmented regression analysis (described below) as described elsewhere.¹⁰ Following this, any reduction in HA-MRSA and HA-CDI cases, as a result of the intervention, was estimated and the average yearly cost savings were calculated; the cost of a MRSA case (colonised and infected) was assumed to be €8673 and the cost of a CDI case was estimated to be £4577.^{18 19}

Statistical analysis

The effect of the introduction of the ClO₂ disinfectant (275 ppm) on the incidence of HA-MRSA and HA-CDI was assessed using segmented regression analysis of interrupted time series.²⁰ This quasiexperimental technique allows the estimation of changes in the incidence of HA-MRSA and HA-CDI during the period of routine use of NaDCC (November 2009 to October 2011) compared with use of ClO₂ (November 2011 to September 2013). This technique measures both the immediate and long-term effects of an intervention and is recommended to evaluate the longitudinal effects of interventions in healthcare settings, in the absence of a contemporaneous control group.^{18 21} Data for the period between April 2011 and October 2011 were omitted from the analysis as the full effect of the intervention was not present during this period (ie, mixed use of both NaDCC and CIO₂). Data were coded as described elsewhere (0 before intervention; 1 post-intervention), and a p value of <0.05 was considered to be statistically significant.¹⁸ Trends in the use of high-risk antibiotics were analysed using linear regression. All statistical analyses were performed using the Statistical Package for the Social Sciences for Windows V.20.0 (SPSS, Chicago, Illinois, USA).

Results

From November 2009 to September 2013, 423 cases of HA-MRSA and 120 cases of HA-CDI were identified in the hospital. The average monthly incidence rate was nine cases (range 2–19) and three cases (range 0–9) for HA-MRSA and HA-CDI, respectively. Over the study period, 128 (30.27%) HA-MRSA cases were identified from clinical samples and 292 (69.03%) cases were identified from screening swabs. For three cases, the source could not be identified. Throughout the study period, there was no significant change in the use of high-risk antibiotics (p=0.0845).

Figure 1 illustrates the incidence of HA-MRSA and the use of the ClO₂-based disinfectant throughout the study period. The segmented regression analysis outlined in table 1 represents the effect of the intervention on the incidence of HA-MRSA 12 months after its implementation. Prior to the introduction of the ClO₂-based disinfectant a significant increase in the HA-MRSA trend of 0.004 cases/100 bed days (p<0.01) was observed. Following the introduction of ClO₂ (275 ppm), a significant decrease in the HA-MRSA post-intervention slope (p=0.001) was observed. Modelling the level effect of the intervention after 12 months showed a decrease in the incidence of HA-MRSA by 0.095 cases/100 bed-days/month (coefficient=−0.095, 95% CI −0.04 to −0.15, p value <0.001).

Figure 1.

Incidence of hospital-acquired-methicillin-resistant Staphylococcus aureus (HA-MRSA) throughout study period November 2009–September 2013 and the use of the chlorine dioxide 275 ppm disinfectant agent (HA-MRSA; use of ClO₂ 275 ppm).

Table 1.

Estimates of parameters in segmented regression model assessing the effect of introducing a chlorine dioxide (275 ppm) disinfectant agent on changing hospital-acquired -methicillin-resistant Staphylococcus aureus incidence rate in Antrim Hospital

Term	Coefficient	95% CIs	p Value
Constant	0.059	0.03 to 0.08	<0.0001
Pre-intervention slope	0.004	0.002 to 0.006	0.006
Level change 12 months after introduction of ClO2-based disinfectant	−0.095	−0.04 to −0.15	0.001
Post-intervention slope	−0.003	−0.005 to −0.001	0.001

No significant change in the HA-CDI trend after the introduction of the ClO₂ 275 ppm based disinfectant was observed (coefficient: −0.03; p=0.873). Furthermore, no significant change in the level of HA-CDI was observed 12 months after the introduction of ClO₂ (coefficient: −1.31; p=0.221). The total cost for the intervention, after its implementation for 1 year was £52 000, which included the cost of the product and training provided by the company. Considering the most conservative estimates of HA-MRSA incidence rate reduction (ie, 0.04 cases/100 bed days/month; the minimal impact estimated by the lower confidence limit outlined in table 1—‘level change 12 months after introduction of ClO₂-based disinfectant’), and an average monthly bed occupancy in the study site hospital of 10 000, the intervention contributed to an average reduction of approximately four HA-MRSA cases per month which equates to an annual reduction of 48 cases. Taking into account an estimated cost of an MRSA case of €8673 (£6845), it is predicted that the intervention would contribute to a reduction in the annual costs associated with HA-MRSA by approximately £276 000.

The results of the weekly environmental cleanliness audits carried out by ward staff showed between 94% and 95% adherence to infection control guidelines throughout the study period.

An attempt was made to assess the effect of ClO₂ 275 ppm on pieces of clinical equipment such as patient vital sign monitors, thermometers and volumetric infusion pumps, between August 2010 and November 2012, using the Mann–Whitney U test. There was no significant difference between the number of faults reported for these items when NaDCC and ClO₂ 275 ppm were in use (p=0.286; 0.836; 0.248 respectively).

Discussion

Decreasing environmental inanimate surface contamination can reduce the risk to patients of acquiring nosocomial pathogens; however, previous studies have highlighted that in practice the effect of disinfectant agents is dependent on contact time, concentration and the presence of organic load.^22–27 Furthermore, it has been demonstrated that cleaning with detergent and water followed by a disinfectant containing 1000 ppm chlorine failed to remove MRSA from the hospital environment.²⁸ Possible reasons for this could be inadequate exposure times of the product to the environment and the presence of organic material not visible to the naked eye, resulting in a suboptimal effect. In the present study, we have demonstrated that changing from a disinfectant containing NaDCC to a ClO₂-based agent resulted in a significant decrease in the monthly incidence of HA-MRSA. This decrease in HA-MRSA may be in part due to the intensive staff training on the use of the ClO₂-based disinfectant and also due to the reported rapid action of ClO₂ compared with NaDCC 1000 ppm.¹² As the NaDCC-based cleaning agents previously used in the hospital required washing off with detergent and water, these agents may have been washed off the surfaces prior to exerting their full effect. This may explain the observed decrease in the incidence of HA-MRSA following the introduction ClO₂ 275 ppm.

The lack of effect of the ClO₂ 275 ppm based disinfectant against HA-CDI should be interpreted with caution due to the low baseline incidence of this infection, with an average incidence of three cases per month compared with nine HA-MRSA cases/month. In addition, most patients identified as C. difficile toxin-positive were placed in isolation rooms. Following the discharge of these patients, the isolation rooms underwent a terminal clean with ClO₂ 275 ppm, followed by a deep clean with steam and vaporised hydrogen peroxide for 2 h. Therefore, these measures would further minimise the risk of patient-to-patient transmission of this organism.

This study has identified the annual cost savings that could be attained as a result of this intervention to be approximately £276 000. The majority of this cost saving represents opportunity costs through the release of isolation rooms and reduced length of stay. The cost of an MRSA case did not include the cost of treating MRSA infections; therefore, for infected patients this may underestimate of the potential cost savings. The cost on the use of and training on ClO₂ 275 ppm was estimated to be £52 000 per year; however £9675 of this was for staff training. As training expenses were a one off cost during the implementation of this intervention, future cost savings in subsequent years of using ClO₂ 275 ppm should be greater.

There was no significant difference in the number of clinical equipment faults reported with ClO₂ (275 ppm) and NaDCC. As this assessment was carried out retrospectively and the suspected cause of the faults was poorly documented, firm conclusions cannot be drawn. A prospective study would be required to determine the true effect of the use of these infection control agents on all types of clinical equipment.

This study has the strength of using segmented regression of interrupted time series analysis. This analysis allowed the quantification of the contribution that environmental inanimate surface decontamination has made to the changing incidence of HA-MRSA and HA-CDI. In addition, this technique minimises the risk of selection bias as all patients admitted throughout the study period were included. Nevertheless, this study has some limitations. It was not possible to split the HA-MRSA or HA-CDI cases by ward and sample ribotyping was not undertaken; therefore, it is difficult to determine if new cases were the result of transmission between patients on the same ward. In addition, the definition of HA-MRSA did not take into account unknown carriers who did not meet the criteria for the hospital’s screening policy. Admission screening was carried out on all patients with a history of MRSA, patients admitted from residential and nursing homes, patients admitted to ICU, neonatal and renal units and when requested by the infection control department. The definition of HA-CDI did not take into account patients who were carriers of C. difficile (ie, patients who tested toxin A- and B-negative, but GDH- and PCR-positive). As these patients have been reported to have a higher level of skin and environmental contamination compared with non-carriers, the effects of ClO₂ against the spread of C. difficile may have been underestimated by not considering these patients.²⁹ It was not possible to adjust for differences in cleaning and antibiotic prescribing practices between pre-intervention and post-intervention periods; however, as the adherence to infection control guidelines for environmental cleaning remained between 94% and 95% throughout the study period, this suggests that any improvement in HA-MRSA rates were not due to improvements in general environmental cleanliness. In addition, there was no significant change in the use of high-risk antibiotics throughout the study period; therefore, it is unlikely that they may have differentially influenced the incidence of HA-MRSA or HA-CDI during the pre-/post-intervention periods.

In conclusion, the use of ClO₂ 275 ppm for environmental disinfection together with an intensive training programme on its use has resulted in a significant decrease in the incidence of HA-MRSA in our hospital. This has allowed us to achieve potential cost savings of up to £276 000 per year. This study highlights the importance of effective environmental inanimate surface decontamination in controlling the spread of MRSA and the potential cost savings that can be achieved through decreasing HA-MRSA rates.

Key messages.

What is already known on this subject

• Healthcare-acquired infections and antibiotic resistance are a major cause of morbidity and mortality in healthcare institutions.

• The surfaces in the hospital environment act as a reservoir for the spread of resistant organisms.

• Environmental decontamination with effective agents is vital to reduce the spread of these organisms in the ward environment.

What this paper adds

• This study has demonstrated the potential monthly reduction in hospital-acquired methicillin-resistant Staphylococcus aureus (MRSA) as a result of a more efficient environmental inanimate surface decontamination programme using chlorine dioxide 275 ppm.

• The potential cost saving as a result of the reduction in hospital-acquired MRSA has been identified.