A trust set up a programme to monitor surgical site infection that had developed after patient discharge to identify the risk factors and assess compliance with best practice
In this article
- Benefits of post-discharge surgical site infection surveillance
- Characteristics or a surveillance programme
- How to set up an active SSI surveillance programme
Authors Judith Tanner is chair of clinical nursing research, De Montfort University and University Hospitals of Leicester NHS Trust; Susan Davey and Islwyn Jones are infection prevention and control specialist practitioners; Julia Ball is head of nursing; Caroline Aplin is deputy ward sister; Jane Pickard is matron; Karen Weafer is lead nurse for tissue viability; all at University Hospitals of Leicester Trust
Tanner J et al (2011) Surveillance of surgical sites in primary care. Nursing Times; 107: 3, early on-line publication.
A surgical site surveillance programme with post-discharge follow-up and dedicated staff can have many benefits. This article describes the development of an active surveillance programme and offers practical advice on setting up such programmes.
Keywords: Surgical site infection, Post-discharge, surveillance
- This article has been double-blind peer reviewed
Rates of surgical site infections (SSIs) and SSI surveillance are becoming increasingly important in the NHS. In 2009, The House of Commons Public Accounts Committee (2009) criticised the absence of post discharge surveillance for SSIs, saying this made it impossible to know the scale of infections or the risks to patients. The US has already introduced punitive measures against hospitals whose patients develop SSIs; in the UK, if patients are readmitted with an infection within 30 days of discharge hospitals will not be paid for further treatment from 2011-12 (Turner and Powell, 2011).
The Leicester programme
Concern about the increased use of vacuum-assisted dressings among colorectal patients led us to implement an SSI surveillance programme. We used the surgical site infection surveillance service (SSISS) – the national SSI surveillance programme led by the Health Protection Agency (HPA, 2008) - for mandatory orthopaedic surveillance. The service collects data on inpatients, and some post-discharge data, through readmissions or outpatient visits. However, we were interested in SSIs treated in primary care which developed after patient discharge. We wanted to identify risk factors for these, and investigate compliance with best practice.
A five-month pilot of an active, post-discharge SSI surveillance programme began in 2008, for colorectal and breast surgery patients, with the surveillance nurse contacting each patient on days 10, 20 and 30 after discharge. If any SSIs were suspected, the nurse would visit the patient to observe the wound.
Infection rates were more than twice those published by the national surveillance programme, with almost half the SSIs in both colorectal patients and breast surgery patients identified in the community. This is supported by other studies using active post-discharge surveillance, which show that inpatient and readmission data underestimate the incidence of SSIs (Scottish Surveillance of Healthcare Associated Infection Programme, 2003).
Cost of SSIs
UK studies show additional costs of between £1,000 and £10,000 for each patient with an SSI (Tanner, 2009; Plowman, 1999), although few studies include primary care costs. Identifying the cost of treating an SSI is not part of a standard surveillance programme, but it provides valuable justification for funding programmes or demonstrating the cost-effectiveness of interventions to prevent SSIs. The simplest method of identifying the costs of treating patient with SSIs is to compare them with costs of treating patients with no infections. Our pilot study showed the average cost of treating a colorectal patient with an infection was £10,300.
The biggest cost was increased length of stay - infected patients stayed around eight days longer than non-infected patients. District nurse visits were the second largest cost; one patient with a leaking wound was visited at home twice a day for 30 days. Most surveillance programmes do not collect data from primary care, meaning the considerable cost of district nursing is not included when calculating the cost of SSIs.
The pilot programme cost £10,400 for five months, disputing the view that post-discharge surveillance is prohibitively expensive (Taylor, 2003). Staffing is the largest cost, but can be minimised by using band 3 healthcare assistants (HCAs) for data collection. Other costs are for phone calls, travel to visit patients at home and data analysis. Surveillance staff also need a base from which to work.
Setting up a post discharge programme
The following practical guide to setting up a post discharge surveillance programme is based on our experiences.
Representatives from relevant directorates and staff groups must be involved in setting up the surveillance programme so that data collected is credible, and can be used to inform practice. A steering group should include surgeons, anaesthetists, microbiologists and nurses, as well as representatives from theatres, infection control, tissue viability and the trust board. A representative from primary care can help facilitate communication between acute and primary care. A senior manager, such as the director for infection prevention and control (DIPC), should chair the meeting and a senior member of the infection control team is best placed to take on the daily management of the programme.
The steering group may need to meet frequently at the start of the initiative, but eventually it can be wound down and SSI reports incorporated into other monitoring processes.
Continuous or rolling surveillance
Conducting surveillance continuously among one group of patients allows SSI rates to be fed back to staff regularly; however, it might be useful to know SSI rates for several patients groups. A better use of resources is to conduct rolling surveillance - following up one group of patients then moving to another.
Assessing the effectiveness of interventions to reduce SSIs can be done by repeating the surveillance. There is no evidence to show SSIs are seasonal, so surveillance does not have to be repeated at the same time of year. Based on our experience, the surveillance period should include at least 100 patients; smaller numbers may not be representative and could give misleading SSI rates. The SSISS suggests a three-month minimum period for surveillance (HPA, 2008).
Selecting surgical procedures
Specify which surgical procedures to include in the surveillance as SSI rates vary between procedures (HPA, 2009). A surgical specialty with a suspected problem could be targeted, or you can focus on cardiac or joint replacement surgery where the consequences of SSIs are most costly. It might be beneficial to select one of the categories included in the SSISS programme, so that data can be submitted to SSISS (HPA, 2008). This will become increasingly important if surveillance becomes mandatory. Ideally surveillance should be conducted on all patients having that procedure during the surveillance period.
Following up all patients having a specific procedure will provide the most reliable data. However, this may not be possible where there is high patient throughput, so it may be necessary to follow up a random selection. In our experience a surveillance staff member can survey 60-70 patients a month.
Methods for following up patients after discharge include:
- Patient self-assessment questionnaires;
- Patient telephone calls;
- Direct observation in outpatient clinics;
- Notification by primary care staff.
Each of these methods has merits, but a review of them did not find one more effective overall (Petherick, 2006Trusts should choose the method, or combination of methods, most suitable to them.
Self-assessment questionnaires: After discharge, patients can be given a questionnaire to complete and return if they suspect an SSI, or are diagnosed with one in primary care. Questionnaires require the fewest resources, but the return rate can be low (Edwards, 2002). They also rely on patients’ ability to identify their own SSI; research shows conflicting findings on this (Petherick, 2006).
Telephone: Surveillance staff can telephone patients at home using a validated questionnaire to identify potential infections. This can produce more accurate findings, but is resource intensive; workload analysis at Leicester shows 15% of surveillance staff time is spent telephoning patients.
In the pilot programme and first six months of the permanent programme patients were contacted on days 10, 20 and 30 after surgery. However, an evaluation of the first six months of the programme found the majority of SSIs were identified between days 10 and 20 so we now contact patients on days 15 and 30 only.
Direct observation: Surveillance staff or surgeons could observe wounds at outpatient clinics. This requires less staff time and, if all patients visit outpatient clinics should capture all patients. However, this method depends on the timing of the appointment in relation to the 30-day cut off and when infections present. It is commonly thought that SSIs present 3-6 days after surgery, but extended patient follow-up has found presentations at nine and 10 days (Smith, 2004; Barrett, 2000). In our pilot, the average for both colorectal and primary breast surgery was 12 days (Tanner, 2009).
Notification by primary care: GPs or district nurses could notify the surveillance team when they treat a patient with an SSI. Although this may be the best long-term solution for capturing post-discharge data, it requires considerable training, communication and collaboration.
Staff who collect surveillance data are usually junior nurses or senior HCAs. Nurses need less training as they already have most of the required skills and knowledge, such as reading case notes, accessing hospital information systems and understanding of medical terminology. However, they may not find data collection challenging and turnover could be high; HCAs may need more training, but could be more likely to see a surveillance post as a career move.
A band 5 nurse conducted surveillance for our pilot. Analysis of her activities showed this level of clinical judgement was not required, so we employed band 3 HCAs for the permanent programme. However, a qualified nurse was appointed for ventilator-associated pneumonia (VAP) surveillance in intensive care as the pilot showed considerable clinical knowledge was required.
Surveillance staff identify each patient having a specific procedure from admission and theatre lists; they obtain basic data from hospital systems and additional information from patients’ operation notes, anaesthetic charts, medical notes, nursing charts or drug charts. A data sheet is made for each patient. Patients are followed up by the surveillance lead nurse until the 30-day cut off, or earlier if they develops an SSI or meet any of the exclusion criteria outlined in the SSISS protocol (HPA, 2008). Finally, the surveillance nurse notes whether the patient meets the defined criteria for an SSI.
Lead surveillance nurse
A senior infection control nurse should act as the surveillance lead. This role involves managing and providing clinical support to surveillance staff, performing quality control on data entry, writing reports and assisting with rapid surveillance feedback. The lead nurse also advises clinical directorates of areas where practice can be improved.
Feedback to staff is effective in reducing SSI rates (Guebbels, 2006), particularly to surgeons, anaesthetists and nurses working in theatres and surgical wards. A ‘top down’ approach is the simplest method to feedback surveillance data, complemented by monthly SSI rates posted on ward notice boards or in theatre coffee rooms. The information contained in rapid feedback only requires simple analysis and should be presented in a format where individual staff are not identifiable.
This pilot surveillance programme enabled us to evaluate our practice and led to major changes throughout the trust. These included the implementation of a trustwide surveillance service and clinical practice interventions which saw SSI rates fall by a third. It is unlikely that any improvements in practice would have occurred without evidence to show the true rates and cost of surgical site infection.
Box 1. Changes to practice after the pilot
The programme began in 2009, employing 4.5 whole time equivalent staff. Surveillance is conducted on surgical site infections (SSIs), central line associated bloodstream infections in renal haemodialysis and intensive care patients, and ventilator-associated pneumonia.
A care bundle to reduce SSIs helped cut the number of colorectal SSIs by a third.
Pre-operative patient information was amended to stress the need to stop smoking, a risk factor for breast surgery patients. We also work with community smoking cessation advisors to reduce patient smoking preoperatively.
Clinical protocols indicating when to aspirate a seroma in breast surgery patients reduced the aspiration rate. Wound drains are now removed a day earlier, reducing length of stay.
Box 2. Data collection advice
SSI identification questionnaire
A validated questionnaire focusing on each of the criteria used to define an SSI should be used. This can be found in the surgical site infection surveillance service protocol (HPA, 2008). Questionnaires can be completed by patients, or by staff in telephone interviews or at outpatient appointments.
Data collection sheet
Data collection sheets provided by the SSISS (HPA, 2008) are designed to identify the incidence of SSIs, rather than the associated risk factors. For more detailed analysis, or to identify compliance with good practice, extra data fields are needed, but this can increase the time needed for follow-up.
Fields on the database should mirror questions on the data collection sheets and SSI identification questionnaire. If the database is held on a shared drive, surveillance staff can enter data wherever they are in the hospital. The lead surveillance nurse should check the database regularly for transcription errors or blank fields.
Barrett E et al (2000) An evaluation of post discharge methods for surgical site infection: Phase II. Final report.London: Public Health Laboratory Service.
Edwards P et al (2002) Increasing response rates to postal questionnaires. British Medical Journal; 324:1183
Geubbels E et al (2006) Reduced risk of surgical site infection through surveillance in a network. InternationalJournal for Quality in Health Care; 18:2,127-133.
Health Protection Agency (2008) Protocol for the Surveillance of Surgical Site Infection: Surgical Site Infection Surveillance Service. London: HPA.
Health Protection Agency (2009) Healthcare Associated Infections in England: 2008-2009 Report. London: HPA.
Turner D, Powell T (2011) NHS Commissioning. London: House of Commons Library.
House of Commons Public Accounts Committee (2009) Reducing Healthcare associated Infection in Hospitals in England. London: The Stationary Office Limited.
Petherick E S et al (2006) Methods for identifying surgical wound infection after discharge from hospital. BMCInfectious Diseases; 6:170
Plowman R P (1999) The Socioecomic Burden of Hospital Acquired Infection. London:Public Health Laboratory Service.
Smith R L et al (2004) Wound infection after elective colorectal resection. Annals of Surgery; 239: 599-605.
Scottish Surveillance of Healthcare Associated Infection Programme (2003) Surveillance of surgical site infection. Glasgow: Scottish Centre for Infection and Environmental Health.
Tanner J et al (2009) Post discharge surveillance to identify colorectal surgical site infections rates and related costs. Journal of Hospital Infection; 72:243-250
Taylor E (2003) Telephone call contact for post discharge surveillance of surgical site infections. A pilot, methodological study. Journal of Hospital infection; 55:1, 8-13