Feature / Infection control

Minimising surgical site infections clearly benefits patients, but understanding the financial costs of infections can strengthen the case for prevention. Steve Brown reports

There are frequent claims that higher quality should mean lower costs, as right-first-time services lead to reduced lengths of stay, less reworking and fewer readmissions.

But sometimes these claims are based on theory rather than hard evidence – improve the services and then hope the savings fall out downstream. However, Plymouth Hospitals NHS Trust has looked to put some hard figures on the financial benefits of eliminating surgical site infections (SSIs).

SSIs are the third largest category of healthcare-associated infections (HCAIs), after respiratory and urinary tract infections, accounting for nearly 16% of all HCAIs. According to the National Institute for Health and Care Excellence (NICE), at least 5% of patients undergoing a surgical procedure develop an SSI. These have a significant impact on patients – in extreme cases, patients can die from an SSI. But they can also prolong length of stay and lead to readmissions and avoidable treatment.

So there are clear quality improvements to be had from eliminating these infections. A national surveillance programme looks to monitor the level of infections and encourage providers to ensure they implement all the best practice steps in NICE clinical guidelines to prevent SSIs. However, there is limited understanding of the financial costs of SSIs – or in fact an actual misunderstanding of what the real and significant costs are.

Plymouth recognised this deficit back in 2011 and set out to fill the gap by matching comprehensive surveillance data for SSIs to its detailed patient-level cost data.

There were two key drivers. The trust already feeds back anonymised infection rates to individual consultants – seen as a key part of its programme to reduce infections. But it wanted to underline the potential for improvements by linking this to the costs of those infections (not just the average cost of an infection based on modelling elsewhere).

Business case

The information was also seen as important for business cases. ‘If you know what it costs, you also know what you can save and you can justify investment in prevention,’ says Dr Peter Jenks, consultant microbiologist and director of infection prevention and control at the trust.

The NICE guidelines set out good practice steps such as maintaining patient temperature and equipment to warm intravenous fluids

and blood products, all of which require upfront investment. ‘It can be very powerful if you can say to your finance director that this is what we can save and use local data about infections and actual costs/potential savings,’ adds Dr Jenks.

Plymouth’s SSI surveillance programme is already more comprehensive than those of many other providers. Providers are only required to undertake one quarter’s worth of surveillance a year in one of four orthopaedic procedures (hip, knee, neck of femur repair or reduction of long bone fracture) although Public Health England also encourages wider voluntary submissions in other categories.

Plymouth runs a full-time surveillance programme in all the areas covered by the national scheme and believes this has itself led to improvements. For example, inpatient and readmission sternal (or chest) wound infections for cardiac surgery have fallen from 4% to just over 2% since the programme began.

However, the trust wanted to take this comprehensive database and link it with the more detailed patient-level costs that it has been producing for a few years. There have been attempts to establish SSI costs before, but Dr Jenks suggests many of them are flawed, based on modelling rather than actual costs incurred in specific cases.

There is no healthcare resource group for SSIs and so costs to date have typically been calculated using the increased length of stay and an average per day rate. However, several studies have come up with a range of different answers for the increased cost related to SSIs. These stretch from an average of £469 (in the NICE clinical guideline) to £10,523 for SSIs associated with colorectal surgery. The range of costs is so wide that business cases could succeed or fail based on the estimate used.

‘We wanted to use our patient-level costing and information system to get a more accurate cost. Rather than, say, taking an increased length of stay of 10 days and multiplying this by an average £300 per day, we wanted to find the genuine amount of increased cost, taking account of the drugs, the dressings and the specific blood products used,’ says Dr Jenks.

Length of stay is still the key driver, but using average costs or excess bed day rates misses the fact that SSI patients could be high users of expensive services.

‘By examining our patient-level cost data, you realise SSI patients spend a greater amount of time in critical care beds, which have a much higher cost per day than in a general ward. They might also have repeat visits to theatre, again with significant expenses,’ says Sarah McQuarry, Plymouth’s finance systems and costing accountant.

Across all 14,300 surgical episodes included in the study, the trust found 733 SSIs – a rate of 5.1%. The median additional length of stay attributable to SSI for all 19 surgical categories was 10 days. A total of 4,694 bed days were lost over the two-year period. On the financial side, the median additional cost attributable to SSI was £5,239 and the aggregate extra cost over the study period was just under £2.5m.

‘What we have shown is that some of the previous cost data, based on modelling, was pretty flawed – hence the wide range of figures. We also show that the NICE figures represent a significant underestimate of the real costs,’ says Dr Jenks. In fact, the NICE estimate of £469 is less than 10% of Plymouth’s calculated cost. On this basis the total national costs of £57m (estimated by NICE on the basis of its £469 figure) could be out by a factor of 10.

Perhaps of even greater interest is the variation from specialty to specialty. The study found that median costs in fact ranged from £21,500 for gastric surgery SSIs to just £1,500 for SSIs related to breast surgery. Additional SSI-related length of stay broadly mirrored this with gastric surgery SSIs adding a median 29 days to the stay, while breast surgery SSIs added just three days.

But the patient-level cost data also enabled Plymouth to understand where these costs were actually arising. Across all surgical categories, 11% of the additional cost was incurred in operating theatres, with 24% on the general ward and 10% in critical care areas. Medical staff were another big contributor at 18%, while diagnostics, dressings and drugs accounted for a further 12%.

But again the cost profile was different across the surgical categories. While ward and staffing costs clearly rose in line with length of stay, other areas experienced more localised effects. So critical care cost was particularly high for SSI episodes in cardiac, cranial, gastric and bile liver and pancreatic surgery. This was due to the longer initial patient stay or return to critical care following wound debridement.

Drug costs were particularly pronounced in cardiac, orthopaedic, spinal and complex intra-abdominal surgery, where prolonged courses of antimicrobial therapy were frequently used to treat wound infections.

Delayed cost hit

An interesting quirk with the payment system at the time of the study meant that the trust did not feel the full pain of the £2.5m additional cost. Factoring in the opportunity costs of eliminating all SSIs – income received for SSIs, for example through increased length of stay, readmissions or higher per diem critical care charges – the trust would only have been £694,000 better off if it had prevented the SSIs.

However, with the introduction of changes to the payments system – and particularly the no payment for readmissions policy – this situation has changed and Plymouth is likely to be absorbing much closer to the full cost of SSIs without supporting payment.

‘We now know what SSIs cost and regardless of income and penalties, every single surgical category has money to save by reducing SSIs,’ says Dr Jenks. He believes the work underlines the benefits of SSI surveillance on a wide basis and that trusts still taking a sampling approach should step up their surveillance activity.

To date the trust has stopped short of reporting back the cost data to clinicians and management teams, although this is the clear next step alongside actually publishing infection rates clinician by clinician.

‘We have run a pilot report with a surgical team bringing the costs in with infection rates, but it is clearly a sensitive issue,’ says Dr Jenks. ‘The feedback was largely positive, but we need to build that feedback into how we take the reports forward. We are keen to publish as widely as we can but there need to be caveats around what a high rate means. A surgeon may have a higher rate because they are operating on a difficult group of patients with high risk factors. We need to be very explicit about things like that.’

As national calls grow to start using patient cost data to drive service improvement, Plymouth appears to be ahead of the game. And by demonstrating the financial benefits of quality improvements, there should be plenty of organisations looking to follow suit.