Infection control has become an area of great concern and cause for discussion in veterinary medicine; as hospital-acquired infections (HAI) are constantly changing and mutating, biosecurity needs to be taken seriously. Yet unfortunately, in most practices, surveillance which is defined as ‘the ongoing, systematic collection, analysis, and interpretation of health data essential for planning, implementation, and evaluation’ (Gaynes et al, 2001: 295) does not take place (Benedict et al, 2008).
The majority of HAIs go unnoticed within practices (Stull et al, 2018). The awareness of an outbreak is often treated as a measure of effective infection control, thus providing a false sense of security, and failing to reduce the risk of further HAIs. ‘One cannot manage what one does not measure’ (Burgess and Morely, 2015: 235). Surveillance is required to prevent this from occurring, making it an integral part of the working practice, allowing an infection control programme to be based on fact rather than ‘feelings’.
Surveillance will allow for the actual statistics of a problem to be known. When surveillance is carried out correctly, it will lead to a reduction in HAIs, which will reduce complication rates for the patients, prevent damage to a practice's reputation and decrease the financial costs to both the owner and the veterinary practice. It is estimated that HAIs occur in 2.8–16% of veterinary patients (Stull and Weese, 2015), although as a result of a lack of surveillance taking place in veterinary practices, there is a significant disparity in these numbers.
What is infection control surveillance?
A surveillance programme has two overall goals: to monitor specific outbreaks, e.g. meticillin-resistant Staphylococcus pseudintermedius (MRSP); and establish a baseline threshold for HAI incidences. When the baseline is breached, action should be taken to mitigate the risk of a potential out-break. To achieve a baseline, a ‘normal’ rate of HAIs must be established, and only with constant surveillance can this be achieved. Without this, it is possible that a small outbreak of a HAI could ‘rumble-on’ un-noticed, along with the associated ongoing environmental contamination, increased patient morbidity and financial cost to the clients and possible zoonotic risks.
Key components of surveillance programmes
Surveillance programmes must have specific and comprehensive key components to be effective:
- To assess infection control measures within the veterinary practice. An estimate of the infection rate and pathogen shed should be established. Considering current research and the data provided, a baseline/critical limit can be produced. When this limit is exceeded, action can be triggered
- To monitor the compliance with infection control policies. This can take place by auditing specific biosecurity standard operating procedures (SOP) and looking at the management of patients within the practice
- The information obtained provides rational decision-making. The data acquired from surveillance can be used to create updated SOPs
- To ensure financial resources are used correctly. Providing constant active surveillance will result in a sizeable economical layout for the practice and can be time-consuming. Therefore, resources should be used to maximise efficiency. The surveillance control programme will need to decide when to trigger further investigation (Morley, 2004).
The importance of surveillance in human health was first recognised in a landmark paper by Haley et al (1985), who found that US hospitals with strong surveillance programmes had 32% lower HAI rates. Although there are no comparable studies in veterinary medicine, it would seem reasonable to assume that similar results could be expected. The style of surveillance programme will differ depending on the practice's size and needs. A large referral hospital with six theatres and five wards will need a more diverse programme compared with a small, one theatre, one ward practice.
Why are hospital acquired infections a problem?
HAIs, sometimes referred to as nosocomial infections, are defined in human medicine as an infection that was not present or incubating before the patient was admitted (Loeffler, 2012). These will typically occur within 48 hours of admission or within 30 days post-surgical discharge (Loeffler, 2012; Stull and Weese, 2015). Common examples include infectious bacterial diarrhoea, surgical site infections (SSI), urinary catheter-associated infections, intravenous and arterial catheter-associated infections and pneumonia (Loeffler, 2012).
Furthermore, HAIs can contribute to illness and even death in patients. In a large study by Benedict et al (2008), it was found that between July 2006 and July 2007, 82% of veterinary teaching hospitals had had at least one HAI outbreak and 45% had multiple outbreaks, leading to restricted admissions of patients and/or closure of the hospital in 32% of cases. This study did not evaluate the efficacy or rigour of individual infection control programmes, therefore providing no answers as to which methods work well. Importantly it is thought that up to 70% of HAIs in human patients are preventable with optimal infection control measures. Moreover, the financial impact of HAIs can be enormous; in American human hospitals, this accounts for $28–48 billion per annum (not including indirect costs incurred by patients); unfortunately, veterinary data in this area are limited (Stull and Weese, 2015).
Different surveillance methods
To gather, analyse and interpret data, a number of different surveillance methods, can be implemented. These include passive, active, targeted, and syndromic surveillance. Since there is no universal optimal surveillance method, most veterinary practices will use a combination of these (Burgess and Morley, 2015).
Passive surveillance
Passive surveillance uses data collected from a secondary source, for example, diagnostic laboratory samples. This is an inexpensive and straightforward data collection method but has limitations. A well-known example of this is the Veterinary Medicines Directorate (VMD) Suspected Adverse Reaction Surveillance Scheme (SARSS), which records adverse reactions to veterinary medicines. In 2011 it received an average of one report per 10 veterinary practices, suggesting gross underreporting, leading to the reliability of SARSS being questioned (O'Neill et al, 2014). The shortcomings of passive surveillance have also been high-lighted in a study by Turk et al (2015), where only 65% of surgical site infections (SSIs) were detected by passive surveillance of patients' hospital records.
Active surveillance
The primary purpose of active surveillance is to identify outcomes via a collection of a primary source of data. However, this method can be time-consuming and expensive. One example of active surveillance using primary source data would be owner surveys to assess the rate of surgical wound-associated SSIs. In the previously mentioned study by Turk et al (2015), active surveillance using an owner survey 30 days postoperatively revealed a significant number of infections (35%) that would have been missed with passive surveillance alone. Another study mirrored these results where 27.8% of SSIs would have gone unnoticed without active postoperative surveillance (Garcia Stickney and Thieman Mankin, 2018).
Targeted surveillance
Targeted or risk-based surveillance focuses on answering a specific question in relation to a condition or disease, and can be more cost-effective than active surveillance. An example of this is the targeted surveillance for Brucella melitensis; to maintain disease-free status, 5% of sheep and goats over 6 months old must be tested annually (Defra, 2011). This method can offer effective early detection of disease in the large animal setting.
Syndromic surveillance
Unlike the other methods, syndromic surveillance involves non-specific disease indicators, such as inflammation around a catheter site, rather than measuring specific disease. Therefore, data collection is easier, quicker, and less expensive, although less specific.
Ruple-Czerniak et al (2013) assessed syndromic surveillance in four veterinary referral practices. Data were collected using a questionnaire with six syndromes available for each patient in critical care (e.g. gastrointestinal signs). These established risk factors associated with HAIs concluded that syndromic surveillance could work well across multiple hospitals and disciplines. Dórea et al (2011) also found that syndromic surveillance allowed early detection of HAIs and improved HAI protocols. However, as with other forms of passive surveillance, HAI rates can be underestimated. A critical function of syndromic surveillance is that it can reveal a rise in baseline HAI levels, resulting in a trigger for active surveillance.
Cross centre surveillance
Veterinary practices will often act on theoretical rather than evidence-based threats because of the lack of data collected (Burgess and Morley, 2015). This problem is heightened by the absence of published studies in veterinary medicine regarding which surveillance methods are most effective in different settings. The veterinary industry does not, at present, facilitate cross-hospital level surveillance, which is standard practice in human hospitals (e.g. the NHS annual HAI reports).
Although this level of data sharing does not take place, using data from laboratories, patient records, and veterinary practices may allow this to be possible in the future. Currently, organisations such as the Small Animal Veterinary Surveillance Network (SAVSNET), a voluntary scheme set up by BSAVA and Liverpool University, are researching antibiotic resistance and zoonotic infections (University of Liverpool, 2020). Also, the Royal College of Veterinary Surgeons (RCVS) has set up the vetAUDIT website, a free online auditing tool, currently looking into small animal neutering, canine cruciate registry, postoperative outcomes, and antimicrobial resistance.
Similarly, large groups of veterinary corporations in the UK may one day make it conceivable for surveillance data to be shared across practices. However, for the time being, this level of surveillance is expensive, complex, and requires engineering, making it challenging to implement. Therefore, the surveillance occurring at present is primarily in individual veterinary practices.
Considerations for development of a surveillance programme
An infection control ambassador should be appointed; depending on the size and structure of the practice, this may take different forms:
- An individual person
- A registered veterinary nurse (RVN)/veterinary surgeon team
- A committee with individual allocated responsibilities. This person or team's role would be to oversee the development and implementation of infection control policies, collate and analyse the data provided, produce reports including infection rates that should be made available to all staff (Stull et al, 2018).
When a surveillance programme is developed, it is essential to consider how the data will be collected systematically and what information should be gathered, avoiding potential data collection hazards. These hazards can occur when inappropriate recording occurs or adverse events are not recognised. Data should be collected by a trained staff member creating methodical and well thought out records with a good understanding of medical conditions (Lee et al, 2007). Staff involved in infection control may benefit from attending continuing professional development (CPD) sessions in these areas to help and enable them to provide CPD (training and strategies) to the rest of the practice.
These vital components should be included during the development of a surveillance programme:
- Specific objectives should first be identified, and the most valuable areas to assess determined (e.g. catheter site infections in the intensive care unit or wound problems reported on clinical records)
- The optimal surveillance method must be identified, e.g. passive, active, or combined
- Exact parameters to be measured should be defined, e.g. laboratory-confirmed infections, record wound breakdowns, etc
- A baseline should be established
- How data is being collected should be clear
- A reporting method must be implemented, and a written SOP produced to ensure the surveillance is carried out effectively (Morley, 2004; Burgess and Morley, 2015). See Figure 1 for the key components to be considered when developing a programme (Turk et al, 2015; Gerrard, 2016; Garcia Stickney and Thieman Mankin, 2018).

Implementing a surveillance programme
Ideally, multiple surveillance methods would be used when setting up a programme. Benedict et al (2008) found in one study veterinary practices most commonly relied on passive surveillance (via the collection of clinical data). As this method tends to underestimate the rate of HAIs, it should not be solely relied on. Nonetheless, it can be used to allow for the baseline of infection rates to be determined; it is also a cheap and easy way for practices to collect data. Once this baseline is breached, active surveillance can commence. See Figure 2 for an example of how a critical limit breach would be displayed.

Using the example in Figure 2 to implement syndromic surveillance, a clear wound classification system must be published for staff to reference quickly (Table 1). Staff need to be educated in how to record the findings and the method used to log data, ensuring the accuracy of the results.
Table 1. Postoperative wound classification use in practice
Postoperative check reference | Auditing codes | Description |
---|---|---|
Postoperative check 1 | WC01 | Normal, clean, no inflammation, no self-induced trauma (SIT), no erythema, no suture irritation, no dehiscence, normothermia |
Postoperative check 2 (minor complications) | WC02 | Clean contaminated, slight inflammation, slight SIT, some skin suture irritation, some dermatitis, seroma formation possible. No pharmaceutical intervention required |
Postoperative check 3 (moderate complications) | WC03 | Contaminated, inflammation, erythema, slight dehiscence not requiring surgical intervention, pyrexia, SIT, skin suture irritation, discomfort, seroma formation. Pharmaceutical intervention required |
Postoperative check 4 (considerable complications) | WC04 | Dirty, infection starting, erythema, SIT, pyrexia, pain, skin suture irritation, decreased blood supply to wound possible. Pharmaceutical intervention required |
Postoperative check 5 (major complications) | WC05 | Infected, major dehiscence, odour, SIT, pain, exudate, ischaemia, necrosis, orthopaedic failure. Surgical intervention required |
Once a case has been established by using active/syndromic surveillance, laboratory culture and sensitivity testing can be used to identify the pathogen causing the infections. These findings may result in a practice conducting an audit to investigate the cause further.
Surveillance and audit can sometimes get used interchangeably (Hay, 2006). However, an audit is a form of targeted active surveillance and is defined as ‘a quality improvement process that seeks to improve patient care and outcome’ (Kerrigan, 2017: 332). Therefore, once surveillance has recognised a problem, the infection control committee may trigger an audit to determine why this happens. For example, looking closely into why catheter site infections are happening — are the catheters being changed regularly, etc?
Using different forms of surveillance and auditing, a greater understanding of a particular HAI problem can be recognised. This will ensure the practice acts accordingly and allows further investigation of causes and modifications of protocols to reduce ongoing risk. It will also enable the practice to reduce the cost of unnecessary testing, as only targeted cases are being sampled.
The end result of surveillance — auditing
Once a problem has been isolated using surveillance, auditing can take over, such as auditing the cleanliness of wards. However, the auditing process can sometimes be unsuccessful because of a lack of training, resources, time pressures, and poor methodology (Hay, 2006).
In a study, Malik et al (2003) found that when three cleanliness audits using visual assessment, adenosine triphosphate (ATP) bioluminescence and microbiologic sampling were carried out, 90% of areas were marked as acceptably clean with a visual inspection; yet when using ATP values, none of the regions were acceptable, and with microbiological testing, only 10% were adequate. Thus, most practices will notice: if an audit is not carried out correctly, the wrong results may be obtained.
Until recently, the veterinary profession did not have any guidelines on auditing standards, unlike the NHS, which could make auditing problematic for practices. The RCVS have now produced a clinical auditing template, which can be found on the RCVS website, ensuring the process is more accessible to all practices (RCVS Knowledge, 2018). Unfortunately, there are still no guidelines for surveillance and limited research available in the veterinary profession, leading to research being taken from the human sector. Nevertheless, this should not stop the profession from improving its infection control programmes.
Conclusions
When carrying out a surveillance programme, objectives must be set out, and a data collection system should be constructed. There is no standardised way for collecting the data; therefore, practices must establish their own method. The surveillance of HAIs is not the job of one person, and everyone within the practice must be on-board, understanding the aims and importance of the programme. This can be a time-consuming project, but it will reduce HAIs and subsequently improve patient outcomes, if carried out correctly.
KEY POINTS
- Surveillance is often the missing link in infection control.
- Combined surveillance methods work better than just one method.
- An infection control committee or person should be appointed.
- Correct surveillance will allow a practice to understand the extent of hospital acquired infections in order to implement strategies, protecting them both financially and reputationally.