Getting it right first time: national survey of surgical site infection 2019

Abstract

Introduction

Surgical site infections (SSIs) are associated with increased morbidity and mortality. Deep SSI, or prosthetic joint infection (PJI), is associated with revision surgery involving longer operative times with higher infection rates, longer length of stay (LoS) and high costs in addition to the catastrophic effect on the patient. The surveillance of SSI is important for patient decision making, identification of outliers for support and maximising focussed improvement. This paper reports the findings of the second Getting it Right First Time (GIRFT) national SSI survey for orthopaedic and spinal surgery.

Methods

Data were submitted prospectively by 67 orthopaedic units and 22 spinal units between 1 May 2019 and 31 October 2019. For a patient to be included, they had to present with SSI within the study period and within 1 year of the index procedure.

Results

A total of 309 SSIs were reported from primary and revision, total hip, knee, shoulder, elbow and ankle replacements, and 58 SSIs were reported from lumbar spine single level discectomy or decompression, lumbar spine single-level instrumented posterior fusion, posterior cervical spine decompression and instrumented fusion and posterior correction of adolescent idiopathic scoliosis. SSIs rates have remained low compared with the 2017 survey. There were variations in SSI rates by procedure, with primary shoulder replacement reporting the lowest (0.4%) and revision shoulder replacement the highest (2.5%) rates.

Conclusions

The authors recommend that the elective surgical restart following the COVID-19 pandemic provides a unique opportunity for all units to implement a full SSI prevention bundle to minimise the risk of infection and improve patient outcomes.

Introduction

Surgical site infection (SSI) is a form of hospital-acquired infection at the site of recent surgical intervention. It provides a consistent and significant burden of morbidity among patients undergoing elective orthopaedic and spinal surgery. Deep SSI in arthroplasty surgery, also known as prosthetic joint infection (PJI), remains a key focus of research due to its potentially catastrophic effect on individual patients and healthcare providers.

The devastating consequences of PJI have led to decades of research into both cause and prevention. Infection rates in early arthroplasty as high as 11% have been reported.¹ Use of modern infection control procedures has been shown to reduce SSI rates in primary total hip replacement (THR) to closer to 2%.² Getting it Right First Time (GIRFT) data published in 2015 demonstrated that significant variation in rates of deep infection exists across the NHS, and vary between individual units, some remaining as high as 5%.³ PJI is associated with increased length of stay (LoS), further operations and a significantly increased cost. Revision surgery for infection involves longer operative time, with higher infection rates compared with revision for aseptic loosening.⁴ GIRFT estimates each PJI will cost around £100,000.³

The surveillance of SSI is important for patient decision making, identification of outliers for support, the study of outcomes following treatment and maximising focussed improvement in prevention. Since 2004, orthopaedic SSI surveillance and data submission to Public Health England (PHE) has been mandatory for all NHS trusts in England as part of the broader Surgical Site Infection Surveillance Service (SSISS).⁵ Currently only primary hip and knee replacement are included in these data. Some studies have questioned validity with orthopaedic specific data, particularly due to the broad nature of definitions used, which are not always relevant to orthopaedic practice.^6,7

The UK National Joint Registry (NJR) has been collecting data since 2003 with data submission mandatory from NHS organisations since 2011. Frequent publication of NJR data allows assessment of the revision rate for infection, which can be viewed by procedure, by unit and by surgeon. Danish registry data suggest arthroplasty for revision in hip arthroplasty is occurring in <1% of cases whereas UK NJR data suggest infection rates following primary TKR and revision TKR of 0.34% and 1.4%, respectively.^8,9 Registry data are also useful to detect late PJI. However, as a measure of prevalence of SSI, use of data where infection is reported only at the point of revision is likely to significantly underreport, as patients and surgeons can elect to use other treatment options and the compliance with the recording reoperations is known to be less reliable than primaries.

GIRFT launched its SSI programme in 2017 with its first national survey, including data from 95 trusts across England. Data were collected from ten elective arthroplasty procedures, and a mean SSI rate of just 0.6% was identified.¹⁰ Due to limitations in data collection, authors warned against using the data as observational representation.¹¹ However, the report highlighted the need for greater awareness and more rigorous SSI surveillance mechanisms.

Guidelines designed to reduce rates of SSI have been developed alongside the developing evidence base. Broad guidance on operating theatre practices is available from The Centres of Disease Control (CDC), Healthcare Infection Society (HIS), the Association for Perioperative Practice (AfPP) and the National Institute for Health and Care Excellence (NICE).

Although as many evidence-based recommendations as possible are made, the lack of good quality evidence in many areas is acknowledged and, therefore, consensus guidance is produced with the aim of providing a quality service.¹² The British Orthopaedic Association have published an expectation for all consultant practice as well as indicating support for orthopaedic-specific recommendations from NICE and GIRFT.¹³ The GIRFT 2015 review on elective orthopaedic services made key recommendations in the pursuit of prevention of SSI,³ re-emphasised in the subsequent follow-up report.¹⁴ The NICE guidance published in June 2020 has also supported specifically the use of ultraclean air systems.¹⁵ The second GIRFT national SSI survey was launched in 2019 and we report here the specialty-specific data from orthopaedic and spinal surgery.

Methods

The 2019 GIRFT SSI programme built on the data collected in the 2017 survey. The survey questions were enhanced and formulated with input from the GIRFT clinical leads and the SSI Workstream of the National Wound Care Strategy Programme.

Hospitals were contacted via their regional GIRFT hub team and recruited for the study. A total of 67 trusts participated and submitted data prospectively between 1 May 2019 and 31 October 2019. Each participating trust received a data collection pack that included information regarding the survey, relevant SSI definitions and data collection tools.

Each participating trust nominated a SSI Trust Champion to coordinate data collection and appoint specialty leads in their trust. Using an online portal, each trust uploaded their own data. This formalised the process compared with the 2017 survey, which relied on doctors in training and spreadsheets shared by email. For a patient to be included, they had to present with a SSI within 1 year of the implant being inserted. The ten orthopaedic procedures included were elective primary and revision hip, knee, shoulder, elbow and ankle replacements. The four spinal procedures included were lumbar spine single-level discectomy or decompression (unilateral or bilateral), lumbar spine single-level instrumented posterior fusion (including interbody fusion), posterior cervical spine decompression and instrumented fusion and osterior correction of adolescent idiopathic scoliosis. These were chosen as they represented a relatively consistent set of higher-volume procedures with reported rates of postoperative infection. Following the survey period, the data were analysed by the GIRFT SSI team and compared with any available equivalent data from PHE for comparison.

The GIRFT SSI programme defines and classifies SSI according to PHE guidance.⁵ For the purposes of this report, all infections have been grouped into either a ‘deep’ or ‘superficial’ category based on CDC definitions in accordance with accepted orthopaedic practice.^16,17

To carry out a comparison with the 2017 survey, all trusts that had not participated in both surveys for the same procedures were excluded to allow us to ensure a fair comparison and to limit the variation in case mix and demographics. Due to the differences in methodology, the 2019 SSI rates were recalculated using the HES data in relevant trusts. Significance of results were assessed using Fisher’s exact test and odds ratio with 95% confidence intervals (CI) were calculated. If the CI interval range crossed 1, then change was not deemed to be significant.

LoS for SSIs patients was calculated from the mean LoS for a procedure from Hospital Episode Statistics (HES) provided by NHS Digital for the orthopaedic procedures. Primary admission excess bed days were then found by subtracting HES LoS from the SSI LoS recorded. Finally, avoidable bed days were then calculated by the sum of primary admission excess bed days and readmission bed days. Reference cost tables from 2017/2018 (with a 4.1% uplift for inflation to 2018/2019) published by NHS England were used to calculate the costs of a surgical reintervention to manage the SSI and the cost of excess bed days.¹⁸ Costs were taken at the highest Charlson Comorbidity Index category to reflect the additional complexity of reintervention for SSI. The costs of excess bed days were used as a measure of missed opportunity costs. The costs of reintervention and of excess bed days were combined to give an overall figure.

Results

Arthroplasty surgery

A total of 67 arthroplasty units (from 145 NHS Trusts) contributed data, with 309 SSIs reported. Table 1 lists the total number of procedures performed during the survey and the associated infection rate for each procedure compared with 2017. This was then adjusted excluding all Trusts that had not participated in both surveys for the same procedure (Table 2). Infection rates were classified according to deep or superficial (Table 3). When the high-volume procedures were reviewed in both arthroplasty surgery and spinal surgery, SSI rates were comparable (Figure 1). However, there were units that represented significant outliers in each procedure.

Table 1 .

Infection rate per procedure with comparison with 2017 survey for orthopaedic arthroplasty surgery and spinal surgery. Raw figures before being adjusted to match Trusts to procedures

Procedure		2017			2019
		No. of SSIs reported	No. of procedures performed	Infection rate %	No. of SSIs reported	No. of procedures performed	Infection rate %
Knee replacement	Primary	47	8,968	0.6	142	17,657	0.8
	Revision	10	638	1.6	17	1,276	1.3
Hip replacement	Primary	52	8,603	0.6	114	18,682	0.6
	Revision	10	1,028	1.0	21	1,772	1.2
Shoulder replacement	Primary	6	888	0.7	8	1,850	0.4
	Revision	0	47	0	4	157	2.5
Elbow replacement	Primary	0	78	0	3	234	1.3
	Revision	0	<10	0	0	58	0
Ankle replacement	Primary	0	93	0	0	308	0
	Revision	0	<10	0	0	39	0
Lumbar spine single-level discectomy or decompression		8	1,674	0.5	31	3,442	0.9
Single-level instrumented posterior fusion (including interbody fusion)		12	443	2.7	17	868	2.0%
Posterior cervical spine decompression and instrumented fusion		3	170	1.8	8	353	2.3

SSI = surgical site infection

Table 2 .

Infection rate per procedure with comparison with 2017 survey for orthopaedic arthroplasty surgery and spinal surgery when adjusted to exclude all Trusts that had not participated in both surveys for the same procedure

			2017			2019
Procedure		No. of Trusts	SSI	Procedures	SSI rate (%)	SSI	Procedures	SSI rate (%)	Infection rate change (%)	OR	CI
Hip replacement	Primary	4	13	1,685	1	17	1,599	1	0.29%	1.38	0.63–3.1
	Revision	2	2	75	3	3	76	4	1.28%	1.5	0.17–18.4
Knee replacement	Primary	10	23	2,169	1	23	2,199	1	–0.01%	0.99	0.53–1.85
	Revision	1	2	61	3	1	60	2	–1.61%	0.5	0.01–9.9
Shoulder replacement	Primary	1	1	22	5	1	23	4	–0.20%	0.96	0.01–78.41
Lumbar spine single-level discectomy or decompression		1	5	444	1	6	422	1	0.30%	1.27	0.32–5.29
Single-level instrumented posterior fusion (including interbody fusion)		3	8	93	9	4	94	4	–4.35%	0.47	0.1–1.85
Posterior cervical spine decompression and instrumented fusion		1	3	10	30	1	12	8	–21.67%	0.23	0–3.5

CI = confidence interval; OR = odds ratio; SSI = surgical site infection

Table 3 .

SSI by depth of infection

	Deep	Superficial	Not specified	Total
Elective primary knee replacement	89	49	4	142
Elective revision knee replacement	14	3	–	17
Elective primary hip replacement	82	33	–	115
Elective revision hip replacement	18	2	–	20
Elective primary shoulder replacement	6	2	-	8
Elective revision shoulder replacement	4	0	-	4
Elective primary elbow replacement	2	1	-	3
Elective revision elbow replacement	-	-	-	0
Elective primary ankle replacement	-	-	-	0
Elective revision ankle replacement	-	-	-	0
Total	215	90	4	309

SSI = surgical site infection

Figure 1 .

Distribution of SSI rates for high-volume arthroplasty and spinal procedures. SSI = surgical site infection

The estimated cost of SSIs for each procedure is given in Table 4. The estimated average cost per patient of an SSI varied from £9,152 to £22,548 depending on the procedure type, with an overall average of £18,788 across all procedures and patients. The average cost per day in hospital varied from £774 to £1,430 depending on the procedure type, with an overall average of £1,011 across all procedures and patients.

Table 4 .

Estimated average LoS, readmissions, avoidable bed days and costs of SSIs for each procedure type

Procedure	Average LoS for primary procedure with SSI (days)	Average LoS for primary procedure all patients (days)	Number of patients	Excess bed days for SSI during primary procedure	Bed days for readmissions	Total avoidable bed days	Total cost of reintervention (very major hip procedure)	Total cost of avoidable bed day	Overall cost	Overall cost per patient
Elective primary knee replacement	4.8	3.8	142	142	2,164	2,306.0	£1,336,773	£1,012,970	£2,349,743	£16,547
Elective revision knee replacement	18	7.5	17	178.5	299	477.5	£160,036	£209,754	£369,790	£21,752
Elective primary hip replacement	6.3	3.8	114	285	1,925	2,210.0	£1,336,476	£1,026,259	£2,362,735	£20,726
Elective revision hip replacement	14	9.5	21	94.5	395	489.5	£246,193	£227,309	£473,502	£22,548
Elective primary shoulder replacement	4.6	3.2	8	11.2	159	170.2	£58,794	£94,215	£153,009	£19,126
Elective revision shoulder replacement	6.8	3.8	4	12	60	72.0	£29,397	£39,856	£69,253	£17,313
Elective primary elbow replacement	1.3	2.9	3	-4.8	24	19.2	£21,548	£5,908	£27,457	£9,152
Total	-	-	309	718 . 4	5,026	5,744 . 4	£3,189,217	£2,616,273	£5,805,490	-

LoS = length of stay; SSI = surgical site infection

Participating trusts reported their compliance with SSI bundles, local antibiotics guidelines and usage and follow up (Table 5).

Table 5 .

Survey question responses

Questions	2019			2017
	Yes %	No %	Responses	Yes %	No %	Responses
Does the trust have an existing surgical site infection prevention bundle?	70	30	60	62	38	26
Does the trust have local antibiotic prophylaxis guidelines?	92	8	61	96	4	25
Was surgery conducted in a laminar flow theatre?	97	3	307	95	5	155
Were antimicrobial prophylaxis given preoperatively?	93	7	307	51	49	155
Was follow up arranged postoperatively?	96	4	307	99	1	155

A causative micro-organism was found in 89% of cases (309 responses) and in 96% of cases antibiotics were given, for a mean duration of 35.3 days. In 74% of cases the SSI required reoperation and in 27% of cases this resulted in removal of the implant. The reoperation rates per procedure are shown in Figure 2.

Figure 2 .

The percentage of SSIs per procedure whose management included a reoperation. SSI = surgical site infection

Spinal surgery

In the survey of spinal surgery, 22 trusts were included with specialist spinal units. An SSI was recorded for 58/5,003 procedures (1.2%). Table 1 lists the total number of procedures performed in each unit during the survey and the associated infection rate for each procedure compared with 2017. Using HES data, excess LoS resulted in 951 avoidable bed days, with total cost of £726,856 equating to a cost of £12,532 per SSI. Procedure-level SSI data are listed in Table 6. Of the 22 units, 14 (64%) had an existing SSI prevention bundle, while 21/22 units (95%) had local antibiotic prophylaxis guidance.

Table 6 .

SSIs recorded by procedure and by type for spinal procedures

Procedure Type	Deep	Superficial	Total
Lumbar spine single-level discectomy or decompression (unilateral or bilateral)	17	14	31
Lumbar spine single-level instrumented posterior fusion (including interbody fusion)	9	8	17
Posterior cervical spine decompression and instrumented fusion	5	3	8
Posterior correction of adolescent idiopathic scoliosis	2	0	2
Total	33	25	58

SSI = surgical site infection

Table 7 provides SSIs reported by procedure. Of the 58 recorded SSIs, survey information was provided for 57 SSIs (98%). Antibiotic prophylaxis was given in 56/57 (98%) of cases and the procedure performed in a laminar flow theatre in 41/57 (72%) of cases. Postoperative antibiotics were uncommonly given (n = 8/57, 14%). Follow up was arranged in all cases, with a mean length of follow up of 2.8 months. The SSI was treated with antibiotics in 54/57 (95%) cases, with a mean length of 20.9 days. Of all SSIs, 6/57 (11%) developed sepsis, with 8/57 (14%) having a reported positive blood culture. A delayed discharge occurred in 18/57 (32%) of cases, with a mean LoS of the primary admission of 6.5 days. An unplanned readmission occurred in 39/57 (68%) with a mean LoS of 20.6 days, whereas a reoperation was performed in 34/57 (60%) of cases. In all cases of SSI involving an implant, the implant was not removed. Overall, 2/57 patients died (4%).

Table 7 .

Survey results by procedure where SSI occurred (n = number of survey responses for each category)

Category	Metric	Lumbar spine single level discectomy or decompression (unilateral or bilateral) n = 31	Lumbar spine single level instrumented posterior fusion (including interbody fusion) n = 17	Posterior cervical spine decompression and instrumented fusion n = 7	Posterior correction of adolescent idiopathic scoliosis n = 2
Primary surgery	Pre-operative antibiotic prophylaxis given	97%	100%	100%	100%
	Laminar flow theatre used	71%	71%	86%	50%
Postoperative care	Postoperative antibiotic prophylaxis given	6%	35%	0%	0%
	Mean length of course (days)	24.0	3.7	N/A	N/A
	Planned follow up (months)	2.5	2.5	5.1	2.0
SSI management	Antibiotics given	97%	94%	86%	100%
	Mean length of course (days)	19.4	26.8	12.8	21.0
SSI Consequences	Sepsis	13%	6%	14%	0%
	Positive blood culture	13%	6%	29%	50%
	Delayed discharge	29%	24%	57%	50%
	Mean LoS (days)	4.0	7.6	8.1	29.5
	Unplanned readmission	81%	59%	43%	50%
	Reoperation	48%	71%	71%	100%
	Implant removed if present	N/A	0%	0%	0%
	Mean readmission LoS (days)	13.0	35.2	37.0	15.0
	Mortality	0%	12%	0%	0%
Reoperation	Open drainage	73%	73%	60%	50%
	Drainage and washout	20%	27%	20%	50%
	VAC	0%	0%	20%	0%
	Lumbar drain	7%	0%	0%	0%

Of the 33 SSI cases where reoperation data were collected, open drainage and washout was most commonly performed (n = 31/33, 94%), while vacuum-assisted closure and a lumbar drain were uncommonly used (n = 1/33 for both, 3%). A causative organism was recorded in 47/58 SSIs (81%) due to seven cases where no growth was recorded and four cases with missing data. Of these, 29/47 (62%) were purely Gram positive, 7/47 (15%) purely Gram negative and 11/47 (23%) SSIs were polymicrobial. Specifically, Staphylococcus aureus was cultured in 8/47 cases (17%) and enterococcus cultured in 7/47 cases (15%) as the two most common organisms.

Discussion

Infection rates

Lack of consensus over diagnosis and definition of PJI has led to difficulty in drawing conclusions on combined data. PHE definitions for SSI are used less frequently in the orthopaedic literature due to the focus on PJI in most cases. In an attempt to clarify in more recent years, the Musculoskeletal Infection Society (MSIS) has produced criteria that are reassessed annually via international consensus meeting.¹² With this consideration, all survey data referencing rates of SSI in arthroplasty surgery should be treated with an appropriate degree of caution.

More data are available for hip and knee arthroplasty given that these procedures are performed in larger numbers; in comparison, rates of infection in shoulder, ankle and elbow may differ considerably due to their lower volume. This may explain in part why an increase in infection rate was detected in revision shoulder replacements compared with the 2017 survey when only 157 procedures are recorded in the current dataset. Similarly, the absence of any reported SSIs in primary and revision ankle as well as revision elbow replacements is most likely caused by the variability in results from low volume of reported procedures. Equally, incomplete reporting or loss to follow up are more expected than there being a truly lower SSI rate in arthroplasty techniques, which are less well established in comparison with hip and knee procedures. Regardless, once the unit and procedures were matched we found no significant difference in rates between the two surveys.

Reviewing the spinal infection rates in our study, there is a clear difference in rate of infections between the different procedures especially when comparing lumbar with cervical surgery. This illustrates the need to measure infections rates for each procedure rather than benchmarking across an entire specialty or spinal department. Interestingly, only 60% of spinal SSIs were reported to have reoperation compared with 73% of SSIs in arthroplasty. This marks the different approach in strategy in spinal surgery where the same selection of temporary devices for staged revision does not exist and where the strategy is often to suppress until fusion is achieved, as removal of implants and revision surgery does not have such favorable outcomes.

In the GIRFT spinal unit previsit questionnaires, very few trusts (20%) were able to provide data on their SSI rates for spinal surgery.¹⁹ This suggests that the majority of trusts undertaking spinal surgery are not routinely monitoring their SSI rates. Consequently, most trusts are not submitting data to the PHE SSI surveillance service. Currently, there is mandatory SSI surveillance in hip replacements, knee replacements, hip fracture fixation and reduction of long bone fractures. SSI is voluntary for a further 13 categories, including spinal surgery. The GIRFT spinal services National Specialty Report recommends it should become a mandatory requirement for trusts to participate in surveillance of SSI in postspinal surgery.¹⁹

Grouping all primary arthroplasty together for the purposes of describing infection rate should be avoided. Similarly, revision procedures are known to have higher rates of infection than primary. It is for this reason that the GIRFT programme recommends the centralisation of low-volume procedures through regional networks and agreed care pathways, whether primary arthroplasty such as ankle and elbow replacement or revision arthroplasty. The NJR has demonstrated that there is a link between the volume of procedures performed and revision rates.

SSI and PJI prevention

The evidence base for infection prevention strategy is evolving rapidly. A variety of individual factors have been studied with varying degrees of success. An individual surgeon and healthcare provider must consistently keep abreast of new research developments while assessing the implications of any additional cost or wider impact of changes in practice.

Systematic review and meta-analysis of patient-related risk factors that have been shown to increase risk of deep infection include among such factors diabetes (higher risk with poor glycaemic control), obesity, malnutrition, smoking, rheumatoid arthritis, depression, steroid use and previous joint surgery.^20,21 Age and alcohol intake have no significant increase in risk. Given that several of these factors may be present in combination, it is vital that an individual patient is aware that their own risk may be higher than the national or departmental average as part of fully informed consent.

In addition, evidence exists for lower infections rates when key surgical pathway factors are used. These include preoperative S. aureus screening, management of disease-modifying medication, perioperative antibiotic prophylaxis, presurgical skin preparation, minimising operating theatre traffic and avoidance of aggressive anticoagulation.¹⁹

Use of ultraclean air systems or laminar flow is a good example of a costly intervention. Following Charnley’s case series data suggesting the benefits of his ultraclean air techniques,²² Lidwell et al published their landmark Medical Research Council trial in 1987, demonstrating a significant reduction in infection rates with use of clean air systems.²³ Several observational studies using large-scale registry data have since disputed this conclusion,^24,25 and, concerningly, based on these and similar studies the World Health Organization (WHO) made a conditional recommendation that laminar flow ventilation should not be used to reduce the risk of SSI for patients undergoing total joint arthroplasty.²⁶

Despite the large case numbers involved, these registry studies under-report infection rates by up to 40% due to their relatively short follow up (6 months) and inability to identify those infected patients who do not undergo revision surgery.²⁷ In addition, these studies have not accounted for confounding factors including surgeon or trust volume, patient factors and use of other SSI prevention measures.

NICE guideline 157 supports the recommended use of ultraclean air ventilation but notes limitations in evidence.¹⁵ As with many infection prevention strategies, the potential for benefit is felt to be worth the cost of universal implementation. Using a ‘bundle approach’ to standardise care has been shown to reduce overall risk.²⁸ Using a combination of national guidance and international and local consensus, a trust can produce its own ‘bundle’ aimed at reducing SSI in arthroplasty surgery.

GIRFT

Deep infection in orthopaedics is extensively researched and one contributing factor is the significance of the effect on the patient’s journey. As demonstrated, an SSI will increase LoS and is likely to result in reoperation as well as incurring significant financial costs for the NHS. The unmeasured sequelae of this are likely to include protracted recovery time, the effect on a patient’s general health and wellbeing as well as the increased pressure to the healthcare provider of prolonged follow up, further investigation and treatment.

In this study, we found that the overall average cost of SSI across all procedures and patients was £18,788. This represented only direct costs of treatment received in hospital as an inpatient or day-case to manage the SSI and is likely to be a significant underestimate of the total SSI costs across the wider health service, including outpatient, primary and community care.

Garfield et al recent identified the 5-year cost of SSI in THR requiring revision to be £41,633 per patient, approximately double our figure for the immediate costs of hospital care.²⁹ However, their analysis was directed towards a higher cost patient group who all underwent revision, rather than our analysis, which recognises real-world practice in which an infected joint is not always treated with revision.

Another cost to the health system is the delay in treating the patient who is on the waiting list for their operation while a SSI is treated for an additional period of time. The cost in terms of delayed treatment captured by lost quality of life is important. In this respect, there is also the costs of quality of life to the patient with the SSI, so this is a double burden. Just looking at primary THR, we can see that using the national average LoS of 3 days the SSIs in THRs in this study alone prevented 737 additional THRs being performed, assuming all other resources required were available. With a similar number of missed procedures in the other joints, it is clear there is a significant opportunity cost in terms of waiting lists with higher SSI rates.

In recognition of the magnitude of costs, SSI prevention is a key priority for GIRFT, which has reported the global costs as estimated to be as high as £100,000 per SSI for TKR and THR.¹⁴ Table 8 outlines the updated recommendations issued in the 2020 update of the orthopaedic report.

Table 8 .

GIRFT recommendations (GIRFT February 2020)¹⁴

• Ring-fenced beds • Strictly enforced theatre discipline with limitations to team numbers and theatre traffic • Dedicated experienced scrub staff with appropriate supervision of trainees • Dedicated orthopaedic theatres with laminar flow • Awareness of the operating procedure for the whole theatre team

GIRFT = Getting It Right First Time

Following initial publication of recommendations in 2015, significant progress has been made across the country in key criteria such as the SSI prevention bundle, which is now implemented in 70% of orthopaedic units compared with 62% in 2017.

In response to the COVID-19 pandemic, GIRFT has worked in partnership with the London Region to transform elective surgical care, by developing 29 pathways across seven surgical Specialties, setting the standards expected in terms of performance and outcome at the top decile of the GIRFT metrics for the individual specialty. Further ‘Hub sites’ have been encouraged to improve outcomes, efficiency and resilience for elective recovery as we act to reduce the increased waiting lists that have resulted from COVID-19. This programme has now been commissioned and is being rolled out nationally. In orthopaedics and spinal surgery, a key component has been to ensure that best practice is implemented across all clean sites. With the opportunity for change presented by the pandemic, we anticipate a continued improvement in compliance with preventative recommendations and a corresponding fall in SSI rates in the future.

Limitations

Trust inclusion was voluntary and therefore selection bias may be present. GIRFT methodology required appointment of a Trust Champion and individual specialty leads with an aim of improving the quality of data collection. However, an element of reporting bias is likely to be present, leading to possible under-reporting. It is important to recognise that true rates of PJI may be higher due to the 1-year postoperative time limitation. Other potential causes of under-reporting include loss to follow up including patients presenting to another trust with their complication or postoperative mortality for another cause.

Conclusions

Within the limitations of the data collection methods for this survey, rates of SSI and associated PJI remain low. Despite this, several key recommendations from multiple professional bodies are yet to be implemented universally. The authors recommend that the elective surgical restart following the COVID-19 pandemic provides a unique opportunity to raise the standard of patient care for all units. This will ensure not only better patient outcomes but also ‘best value’ for the NHS. This can be achieved by implementing the following guidance:

• •introduction of a SSI prevention bundle to standardise practice across orthopaedic surgery, to include recommendations from GIRFT (referenced in Table 8);
• •antibiotic prophylaxis, wound irrigation, intracavity lavage and antibiotics before wound closure in accordance with NICE;
• •appointment of specialty clinical infection lead to be aware of updates to literature and best practice guidance.

GIRFT team

Allison Beal	Project lead
Jamie Day	Chief information officer
William K Gray	Senior research associate
Gary Scott	Project analyst
Jody Walton	Project administrator
Lisa Paget	Project administrator
Gillian Salter	Project coordinator
Anne-Marie Ridgeon	Project coordinator
Melanie Proudfoot	Head of communications