Microbial contamination of open‐but‐unused portions of wound dressings stored in home settings

Abstract

The purpose of this study was to investigate incidence rates and levels of microbial contamination in open‐but‐unused portions of wound dressings stored in home settings. Portions of wound dressings were collected at up to four home visits for 104 clients undertaking wound management within their home. A control sample and stored sample was collected on each home visit and sent for pathology testing to identify levels of microbial contamination. The stored open‐but‐unused wound dressings were managed according to a written protocol. Of the tested samples (n = 776), 6% of control samples and 7% of test samples had microbial contamination. From regression analysis, the stored samples were more likely to have microbial contamination than control samples, but results were not statistically significant. In comparing occasions of storage and handling across four home visits, after adjusting for sample group and dressing type, none of the home visit occasion regressions were statistically significant. In conclusion, storage of open‐but‐unused portions of wound dressings kept in home settings does not appear to increase the rate of microbial contamination compared to newly opened wound dressings.

Introduction

The safe and effective management of wound dressing products is important for clients and health care providers in the prevention of wound contamination. This minimises the risk of wound infection and promotes sound health care, fiscal and resource management. In community‐based nursing, where care is delivered in the client's home, it is a common practice for wound dressing products to be opened, a portion used and the remainder saved for a subsequent wound dressing change (open‐but‐unused). The reasons for utilising open‐but‐unused wound dressing portions has not been formally investigated, although it includes financial savings, supply issues and not wanting to ‘waste’ dressings. The argument against this practice is that there is potential for contamination of the wound dressing during the process of cutting, storage and handling; however, this is not underpinned by evidence. The lack of published research on the contamination of wound dressings highlighted the need to investigate this practice further.

Study aim

This study investigated whether storage of open‐but‐unused portions of wound dressings in the home setting was likely to increase the incidence rates or levels of microbial contamination when stored and re‐opened on subsequent occasions, in comparison to a newly opened wound dressing.

Literature review

The use of open‐but‐unused portions of wound dressings in health care settings has rarely been reported in the literature. Aras and Sussman 1 examined three types of ‘single use only’ packs of amorphous hydrogels used in a wound clinic. For 1 month, weekly samples were collected from the open tubes and a control tube (opened immediately prior to sample collection), which were cultured to test for microbial contamination. All samples were free of microbial contamination. Zwanziger and Roper 2 cultured the equipment used during wound dressing procedures stored in client's homes. They determined that the longer equipment was used, the greater the amount of bacterial contamination. This study predominantly examined equipment rather than dressings. Alqahtani and Lalonde 3 cultured 2 × 2 cm samples of ‘clean dressings’ (consisting of sanitary pads, panty liners and diapers) compared with sterile gauze and sterile abdominal pads. There were very minimal instances of microbial growth on all samples except for the diapers, where 15 of 20 samples grew Bacillus sp. The use of non‐sterile scissors to cut a wound dressing is also another factor for consideration. Lawson et al. 4 noted no statistical significance in infection rates in the healing of open surgical wounds by secondary intention in the 3 months before or after implementing the use of non‐sterile scissors versus sterile scissors to cut wound dressings. It was clear from the paucity of literature that this study was necessary to help underpin clinical practice.

Methods

Study population

The study was conducted within a large community nursing organisation in Adelaide, South Australia: Royal District Nursing Service (RDNS SA). RDNS SA is part of the Silver Chain Group – a not‐for‐profit organisation in Australia that provides in‐home and clinic‐based services to persons with a variety of nursing and home help needs. Wound management forms the largest component of RDNS SA services.

The participants for this study were RDNS SA clients receiving wound management in their home, who were assessed as being at low risk of wound infection and who had open‐but‐unused portions of wound dressings stored in their home as part of their wound dressing regimen. RDNS SA has had a documented procedure for managing open‐but‐unused dressings for wound management since 2008, with anecdotal evidence of no adverse client outcomes.

Study design

This study utilised a prospective, cohort design with a convenience sample. Ethics approval was obtained from the Silver Chain Human Research Ethics committee. The study was overseen by the Silver Chain Research department. One Research Nurse and one Research Assistant assisted in client recruitment, training of staff and sample collection. The Research Nurse recruited clients by attending nurses' meetings and reviewing the client data base to seek clients who were already using one of the wound dressing types investigated in this study.

The wound dressings included in the study were grouped into two broad categories: dry and moist. The dry dressings were SeaSorb^®, a sheet alginate, and Aquacel^®, a sheet gelling fibre. The moist dressings were Mepilex^®, a silicone‐coated foam, and Atrauman^®, a coated polyester mesh. Inclusion of both categories was important as differences in the composition of dressings, particularly related to the moisture content of the interface, might affect the likelihood of microbial contamination. Dressings were supplied and funded by RDNS SA.

G*Power 3.1.4, a priori analysis for a two‐group Chi‐square test, was used. The sample size was calculated at 766 wound dressing samples given the following parameters: odds ratio of 2·2, an α error probability of 0·05 and power = 0·95.

Inclusion criteria

Inclusion criteria were RDNS SA clients receiving wound management through three participating service delivery centres who were at low risk of wound infection and would normally have open‐but‐unused portions of wound dressings used and stored in their home as part of normal care. A client was determined to be at low risk of wound infection if they:

• did not have a current, known wound infection

• did not have a wound that entered a sterile body cavity

• had not undergone surgery in the past 2 weeks

• did not have a diagnosis of any multi‐resistant organisms

• were not significantly immunocompromised (i.e., not taking immunosuppressive or cytotoxic medications; not receiving radiation or chemo therapy; had controlled diabetes, defined as HbA1C <8% or average fasting blood glucose level < 6mmoL; did not have a diagnosis of an autoimmune disorder, such as inflammatory bowel disease, systemic lupus erythematosus or acquired immune deficiency syndrome; or did not have a known low leucocyte count).

• had a home environment where it was possible to appropriately store opened dressing products.

Exclusion criteria

Exclusion criteria were clients who required a surgical aseptic technique, who were attending an external appointment for wound review during the protocol period or who were likely to be discharged from RDNS SA during the protocol period.

Sample collection

A detailed, written protocol was developed to standardise the collection of samples for analysis. Sample collection was undertaken by Registered and Enrolled Nurses (n = 28) who were individually trained for consistent sample collection in the client's home by the Research Nurse or Research Assistant. Nurses trained in sample collection were required to sign a consent form declaring they had received training before they were eligible to collect samples. All nurses who collected wound dressing samples had worked for RDNS SA for at least 6 months and were familiar with the RDNS SA procedure for managing open‐but‐unused wound dressings stored in clients' homes. Informed, written consent was obtained from each client prior to participation. The client also gave permission for wound dressings to be stored in their home.

Purposive incremental sampling of dressings was based on the client's wound management plan without reference to the home setting. The method of sample collection was chosen to enable the study to take place during routine, scheduled nursing visits.

Figure 1 illustrates the protocol for managing samples. At each visit on different days, a sample was taken from a newly opened control wound dressing, and up to four samples were taken from one test wound dressing. Samples were cut from the wound dressings using one pair of sterile scissors supplied as part of the project kit. The nurse opened one end of the packaging and removed the scissors using the handles. The control sample was cut and collected first, followed by the test sample. Following the collection of samples, the scissors were replaced in their original packaging, which was sealed with tape and then placed into the re‐sealable bag that also contained the re‐sealed open‐but‐unused wound dressing products. At each subsequent use, the blades of the scissors were wiped before and after use with a large alcohol wipe and allowed to dry. At no time did the scissors or the samples come into contact with the client, blood, tissue or body fluids. The two samples (control and test) were placed into separate, labelled specimen pots, which were placed into a labelled, re‐sealable bag.

Figure 1.

Sample management..

All clients had at least one occasion of sample collection Sample collection continued at each subsequent visit, up to and including four separate occasions. In routine practice, only a small proportion of open‐but‐unused dressing would still be in use at the fourth home visit. Therefore, the number of clients included in each testing occasion steadily decreased up to the fourth testing occasion. The reasons for this included: use of all of the open‐but‐unused wound dressing, clinical improvement (no longer requiring nursing visits) and nursing scheduling (if there was no nurse trained in sample collection available).

Laboratory analysis

All samples were transported to the Flinders Medical Centre, SA, laboratory within 12 hours of collection. Samples were packaged according to National Pathology Accreditation Advisory Council guidelines. If samples were received after hours, they were stored at 4 °C. The total aerobic bacterial counts were carried out on 2 × 2 cm samples of the product type under test. Where the product contained more than one component, all components were represented in the sample (i.e., Mepilex has several components, including silicone, foam and film); 20 ml of 1% peptone (1% Bacteriological Peptone (Oxoid, Thebarton, Australia) and 0·5% sodium chloride (Ajax Finechem, Thebarton, Australia) water containing 0·1% Tween 80 (Lab Supply, Mascot, Australia) was added to the entire sample and shaken vigorously for 10 minutes at room temperature. Samples were vortexed for 30 seconds at room temperature; 200 μl of peptone water from the samples were spread on Tryptone Soya Agar (TSA; Oxoid) plates and incubated for 48 hours at 35 °C for bacterial culture. Another 200 μl was spread on Sabouraud Dextrose Agar (SDA; Oxoid) plates and incubated for 5 days at 22 °C for the culture of dermatophytes and other types of fungi. Plates were examined and the colony‐forming units (CFU) counted. Estimates of total count was the sum of the counts obtained on the two media (TSA and SDA).

Dilution and volumes plated were adjusted to ensure that the overlap of colonies was not significant. Number of CFU was counted per plate for each sample, and numbers of replicate plates were sufficient to give a reliable estimate of the total count. Each colony was counted at 50 CFU/dressing. Samples were stored for 1 month at 4 °C.

Statistical analysis

Descriptive statistics were reported for the analysis of pathology test results. Due to the large percentage of samples without microbial growth, zero‐inflated negative binomial regression analysis was used to determine the variables associated with the count of microbial growth incidence. Regression of microbial load in open‐but‐unused portions of wound dressings compared with newly opened wound dressings sampled concurrently was used to determine which opening of the open‐but‐unused wound dressing portions was most likely to become contaminated. A descriptive list of types of pathogenic microorganisms was also examined to see which microorganisms were most prevalent.

Results

In total, 104 clients were enrolled into the study. All clients had at least one sample taken and sent for analysis.

Total sample

In total, 776 samples were collected and analysed, with 387 control samples and 389 test samples; 729 (94%) of samples had zero microbial growth. Out of the 47 (6%) samples with microbial growth, 13 samples had levels of microbial growth >200 CFUs/g. For the total sample, the mean count was 22 CFUs/g, SD = 262, range 0–6358 CFUs/g.

Results: control (previously unopened) versus test (open‐but‐unused)

For both types of samples, there were minimal incidences of microbial growth. Test samples had a slightly higher incidence rate of microbial growth than control samples. Table 1 presents the microbial count (including mean, standard, range and confidence interval) and incidence of microbial growth for both control and test samples.

Table 1.

Counts of incidence of microbial contamination

	Number in sample, n (%)	Microbial count (CFUs/g)				Incidence of microbial growth
		Mean	SD	Range (min–max)	CI (95%) (lower, upper limits)	Incidence of zero counts, n (%)	Incidence rate of microbial growth, n (%)
Control	387 (49·9)	14·1	162·5	0–3148	−2·1, 30·4	366 (94·6)	21 (5·4)
Test	389 (50·1)	30·3	333·1	0–6358	−2·8, 63·5	363 (93·3)	26 (6·7)
Total	776 (100)

CFUs, colony‐forming units; CI, confidence interval.

Test samples across four different occasions

Samples were collected from the same home setting on up to four consecutive visits. Table 2 presents the microbial count and incidence of microbial growth in test samples across the four sample collections. The second opening had the highest incidence rate of microbial growth, whereas the fourth opening had the lowest rate. For those samples with microbial growth, the third opening had the highest microbial count, whereas the fourth opening had the lowest count. There was no systematic trend or changes across the four sample collections.

Table 2.

Microbial count and incidence of microbial growth in test samples across the four sample collections

	Number of test samples, n (%)	Microbial growth (CFUs/g)				Incidence of microbial growth
		Mean	SD	Range (min–max)	CI (95%) (lower, upper limits)	Zero counts, n (%)	Microbial growth, n (%)
First opening	103 (26·5)	12·8	89·1	0–870	−4·7 – 30·2	98 (95·1)	5 (4·9)
Second opening	101 (26·0)	28·5	133·2	0–1087	−2·2 – 54·8	90 (89·1)	11 (10·9)
Third opening	96 (24·7)	76·5	650·1	0–6358	−55·2 – 208·2	89 (92·7)	7 (7·3)
Fourth opening	89 (22·9)	3·0	19·3	0–161	−1·0 – 7·1	86 (96·7)	3 (3·4)
Total	389 (100)

CFUs, colony‐forming units; CI, confidence interval.

Dry dressing type versus moist dressing type

In total, 528 dry wound dressing samples and 248 moist dressing samples were collected. The dry dressing had a higher incidence rate of microbial growth than the moist dressing type. Table 3 presents the microbial count and incidence of microbial growth in both dry and moist dressing type samples.

Table 3.

Count of incidence of microbial contamination by dressing type

Type of dressing	Number of samples, n (%)	Microbial count (CFUs/g)				Incidence of microbial growth
		Mean	SD	Range (min–max)	CI (95%) (lower, upper limits)	Incidence of zero counts, n (%)	Incidence rate of microbial growth, n (%)
Dry	528 (68)	12·4	75·2	0–1086	5·97–18·84	492 (93·2)	36 (6·8)
Moist	248 (32)	43·3	450·6	0–6358	−13·05 – 99·67	237 (95·6)	11 (4·4)
Total	776 (100)

CFUs, colony‐forming units; CI, confidence interval.

Types of bacteria

Table 4 identifies the specific types of bacteria identified on samples with microbial growth of over 200 CFUs.

Table 4.

Types of bacteria identified on samples with microbial growth >200 colony‐forming units (CFUs).

Sample number	Type of sample	Dressing type	Microbial count CFUs/g	Bacteria type/s
072·3	Test	Moist	6358	Rothi aeria, viridans streptococci, Neisseria sp., Klebsiella pneumoniae, Staphylococcus aureus (sensitive), Neisseria perflava, Staphylococcus haemolyticus, Acinetobacter pittii
024·6	Control	Moist	3148	Staphylococcus aureus (MRSA), Escherichia coli , Staphylococcus simulans, Acinetobacter baumannii, Dermabacter hominis, Stenotrophomonas maltophilia, Staphylococcus caprae
018·2	Test	Dry	1087	Staphylococcus haemolyticus
071·1	Test	Dry	870	Gram‐positive bacillus (not able to identify further)
020·2	Test	Dry	652	Staphylococcus hominis
024·3	Test	Moist	432	Stenotrophomonas maltophilia, Escherichia coli
003·2	Test	Moist	361	Staphylococcus capitis
041·3	Test	Moist	250	Staphylococcus pettenkoferi
020·1	Test	Dry	217	Brevibacterium
019·8	Control	Dry	217	Kocuria rhizophila
057·2	Test	Dry	217	Micrococcus luteus
059·7	Control	Dry	217	Cladosporium
016·5	Control	Dry	217	Gram‐positive bacillus

Predictors of microbial contamination

Zero‐inflated negative binomial regression was used to identify the variables associated with the count of microbial growth incidence. Results (Table 5) showed that the stored samples were more likely to have microbial growth than control samples (b = 0·51), but results were not statistically significant (P = 0·17). In comparing the occasions of storage/handling across four home visits, after adjusting for sample group and dressing type, none of the home visit occasion regressions were statistically significant. The type of dressing (i.e., dry or moist) was not a statistically significant predictor of microbial contamination.

Table 5.

Zero‐inflated negative binomial regression results for microbial growth

	Regression coefficient, B	z statistic	P value	Confidence interval (95%)
				Lower limit	Upper limit
Stored versus control dressing	−0·22	−0·71	0·48	−0·81	0·38
Second opening versus first opening	−0·26	−0·65	0·52	−1·05	0·53
Third opening versus second opening	−0·08	−0·19	0·85	−0·91	0·75
Fourth opening versus third opening	0·26	0·55	0·58	−0·66	1·18
Moist versus dry dressing type	0·46	1·31	0·19	−0·23	1·16

Discussion

All samples (n = 776) tested were within the Therapeutic Goods Administration (TGA) recommended levels of acceptable microbial load for goods required to be clean (<10,000 microorganisms per gram) 5. Whilst ‘clean’ is no longer consistent with contemporary practice and terminology related to aseptic technique and wound dressing procedure, the results support that open‐but‐unused wound dressings can be used as part of an aseptic technique, which ‘aims to prevent pathogenic organisms, in sufficient quantity to cause infection, from being introduced to susceptible sites by hands, surfaces and equipment’ 6.

There are a number of factors that are likely to affect the risk of contamination of wound dressings. These include: the characteristics of the home environment; the client's health status and understanding of their care; the type and characteristics of the wound/s; the type, characteristics and packaging of the wound dressing; and the techniques used in handling and storage of wound dressings. Home environments vary considerably, and an inspection of the client's surroundings will guide the nurse as to whether a suitable clean storage space and equipment is, or could be, available for storage of dressings. RDNS SA has had written guidelines governing the use of open‐but‐unused wound dressings for several years. This includes a risk assessment of the client, the wound and the environment in which the care occurs. Where the risk is low, open‐but‐unused portions of wound dressings are managed aseptically and stored in the original packaging, which is then sealed with tape and placed within a dedicated, re‐sealable plastic bag within the home for single client use. Open‐but‐unused wound dressings are not used or stored if the risk is not deemed low.

Despite the widely reported practice of using open‐but‐unused portions of wound dressings, little information has been reported to inform the likely safety of this practice. To some extent, state and federal regulatory guidelines direct the safe use of wound dressing products within Australia. The clarification of the definition of ‘single use’ was an important consideration for the development of this study as most wound dressings in Australia are labelled as ‘single use only’ – identified by the international symbol on the packet. There was concern that the premise of this study to test and use portions of an opened wound dressing at subsequent wound dressing changes constituted ‘reuse’. An investigation of the parameters of ‘single use’ and ‘reuse’ in relation to relevant state health and federal bodies revealed:

• ‘Use’ of an item is defined as: ‘the device has been placed into a wound or body cavity and comes into contact with blood or body fluids’ 7.

• ‘Reuse’ of an item is defined as: ‘the repeated use or multiple use of any medical device which has undergone some form of reprocessing (cleaning, disinfection or sterilisation) between each episode of use’ 7.

• ‘Open‐but‐unused’ is defined as referring to: ‘a [Single Use Device] SUD where the packaging has been damaged or opened but the device not used and/or did not come into contact with blood, tissue or body fluids’ 7.

Wound dressings in this study did not come into contact with a wound or body cavity, nor did they have any contact with blood or body fluids and were not reprocessed or repackaged, and therefore were not used or reused. Use of language and appropriate terminology is important: using an open‐but‐unused wound dressing is not the same as reuse of a wound dressing, and the terms should not be used interchangeably. Re‐packaging involves transfer of an item to new or alternative packaging. Any open‐but‐unused portions of dressing were replaced into their original packaging. This procedure ensured that all relevant information from the original packaging remained available to the nurse.

Maintaining asepsis is imperative when performing a wound dressing procedure. Australia has implemented the principles and procedures of Aseptic Technique such as Aseptic Non Touch Technique (ANTT^®). This is underpinned by the Australian Guidelines for the Prevention and Control of Infection in Healthcare 6. In practice, Aseptic Technique is 1 of 10 required standards in the National Safety and Quality Health Service Standards against which health care providers are assessed and accredited 8. The Australian Guidelines for the Prevention and Control of Infection in Healthcare 6 define asepsis as ‘freedom from infection or infectious (pathogenic) material’ and Aseptic Technique as aiming ‘to prevent pathogenic organisms, in sufficient quantity to cause infection, from being introduced to susceptible sites by hands, surfaces and equipment’. It is further acknowledged that it is not possible to achieve a sterile technique in a typical health care setting due to the multitude of organisms in the atmosphere 6. In relation to this study, it was therefore recognised that open‐but‐unused portions of wound dressings will not be, or needed to be, sterile but needed to have levels of microbial contamination that would not be sufficient to cause infection.

Maintaining asepsis using an Aseptic Technique underpinned the protocol and the study. All nurses in the study organisation had been trained in asepsis and Aseptic Technique. Within this study, asepsis was further reinforced by having an explicit, documented procedure, and all nurses who participated received additional training in the study protocols. As 28 nurses were involved in sample collection, this potentially increased practitioner variability, which therefore reduced bias that might have occurred if only one nurse collected samples. Additionally, the use of open‐but‐unused wound dressing portions has been regular clinical practice in the study setting for many years; therefore, the conceptual framework and procedures were not new.

Findings

This study demonstrated that by following a documented protocol, the use of open‐but‐unused portions of wound dressing products for a wound dressing procedure maintained asepsis and did not increase the risk of a client developing a wound infection, compared to using a newly opened wound dressing product. The level of microbial contamination of open‐but‐unused wound dressing portions remained very low and within the recommendations of the Australian Therapeutic Goods Administration 5. This study therefore provided support for use of open‐but‐unused wound dressing portions in low‐risk clinical and environmental situations for clients undergoing wound dressing procedures in their home.

Considerations in application of the findings

The incidence of wound infection was not measured as part of this study. The investigators and other nursing staff did not report a concerning rate of wound infections when using open‐but‐unused wound dressings. The wound dressings used in the study were not antimicrobial, indicating that the wounds included were not determined to have a bacterial load that required treatment. Additionally, as wound infection can result from multiple variables, reporting outcomes based on only one variable (use of open‐but‐unused wound dressings) might have incorrectly attributed this practice to the development of wound infection.

It is acknowledged that wound dressings used in this way must be appropriate to be cut into portions. Manufacturer's recommendations must be verified to determine the suitability for cutting. Furthermore, using the dressing size most appropriate for the wound size is recommended. A wound dressing much larger than required is avoided based on the ability to cut it into smaller portions. If there is a suitable dressing size that will avoid the need to cut it, this is always chosen.

Risk assessment and management is an ongoing process. If there is a change to the client, wound or environmental risk, a change in practice will be necessary. For example, if a previously shallow wound has undergone surgical debridement and there are now deeper structures exposed, use of open‐but‐unused wound dressings would be inappropriate.

The use of open‐but‐unused wound dressings in clinical practice has the potential to result in significant fiscal savings. Wound dressings have become increasingly advanced, with a corresponding increase in their cost. Using open‐but‐unused wound dressing portions can have a significant cost saving on wound dressing expenditure. Scarce fiscal resources can be utilised for other initiatives and to support other aspects of practice. For organisations where clients are required to purchase their own dressings, use of open‐but‐unused wound dressings can ease the client's financial burden and might promote purchase of more appropriate wound dressings.

Limitations

This study demonstrated some very promising results in support of using open‐but‐unused portions of wound dressings in clinical practice. However, there are several limitations, including:

• The study was undertaken in home care environments. Applicability to other care settings, such as residential care, general practice, clinic care and inpatient care, needs further exploration.

• Only four wound dressings were included in the study. Extrapolation of these findings to other wound dressings should be undertaken with caution.

• Open‐but‐unused wound dressings were used for a maximum of four consecutive dressing changes. The potential level of bacterial contamination for any subsequent use of portions of an open‐but‐unused wound dressing is unknown. Therefore, these outcomes demonstrate using up to four portions from a single wound dressing. Further studies are required to examine microbial counts if further portions are cut.

• Only aerobic bacteria were able to be identified. The presence of anaerobic bacteria and spores were not specifically tested for.

• Where there was some bacterial contamination of the open‐but‐unused wound dressing, potential sources of bacterial contamination were not examined. Potential sources of bacterial contamination included the air (including droplets), instruments used to handle the wound dressing, the packaging itself and inadequate re‐sealing of the packaging. Whilst protocols and training minimised the risk of bacterial contamination, each nurse participating in the study and collecting samples managed the procedure unsupervised. This might have resulted in some reduction to protocol adherence, which could have influenced results.

Conclusion

This study found that use of open‐but‐unused wound dressing portions in low‐risk clinical, wound and environmental situations using a standardised protocol resulted in no or very low levels of wound dressing bacterial contamination, making this a procedure that can be considered in clinical practice.

The investigators encourage other practitioners to replicate and expand on this study to build an evidence base regarding the use of open‐but‐unused wound dressing portions in clinical practice. This might include larger studies, studies using different types of wound dressings and studies in other clinical settings.