All 12 studies showed a reduced blood culture contamination rate after introducing Kurin Lock. Most of the studies were based in the US but there were 3 quality improvement studies based in NHS emergency departments. Most of the studies were in adults but there was 1 study in children. Although the evidence base was limited, the committee considered it plausible that Kurin Lock would lead to the positive outcomes associated with reduced blood culture contamination.
There is a lack of direct evidence of the downstream resource impact of using Kurin Lock. By reducing the blood culture contamination rate, the number of false positives should also be reduced. This is expected to have an impact on a patient's length of stay and antibiotic use. The clinical experts advised that there is uncertainty in the length of stay for people who have a blood culture taken and that many factors influence this. The length of stay data used in the economic model was from Skoglund 2019, based in the US. The key parameters were that a person with a true-negative blood culture result would have a hospital stay of 5 days, and a person with a false-positive result would have a hospital stay of 7 days. The clinical experts advised that there is uncertainty in the length of stay for people who have a blood culture taken, and that many factors influence this. One clinical expert explained that other test results and clinical information are routinely used to help decide if a blood culture result is contaminated. So, in their opinion only a small proportion of people with a false-positive blood culture result would have additional treatment and a longer hospital stay. In Alahmadi (2010) a retrospective case-control study design was used in which false-positive blood culture cases were matched with comparator cases. But the EAG and clinical experts noted that contaminated blood cultures were not all matched to comparator cases from the settings. The committee agreed that it is reasonable to assume that people in intensive care may be expected to have longer stays and higher daily stay costs compared with other settings, so the cost savings per contaminated blood culture may be overestimated. The committee agreed with the EAG's view that the Alahmadi (2010) study, which estimated longer hospital stays associated with false positives compared with Skoglund 2019, was not generalisable to a wider NHS setting because of the high proportion of people (42%) in intensive care. The committee agreed that the 2‑day difference from Skoglund (2019) may not represent NHS clinical practice, and that further evidence of the resource impact in the NHS should be generated.
Further evidence is also needed on the impact of false positives on antibiotic treatment. The clinical experts advised that most people with suspected sepsis would start antibiotic treatment before the results of the blood culture test were available and so reductions in antibiotic use are difficult to estimate. The committee understood that antimicrobial resistance is an issue within the NHS. It considered that using Kurin Lock may contribute to antimicrobial stewardships efforts if people who have started antibiotics before getting their blood culture results are able to stop the antibiotics. In cases where using unnecessary antibiotics can be avoided, patient outcomes will improve, for example by reducing risk of side effects of unnecessary treatment. The committee noted that data on staff adherence to blood collection protocols is also important to determine if this reduces over time or in busy periods, and the impact on blood culture contamination rates.
According to the clinical experts, Kurin Lock is easy to use and needs minimal training. The clinical experts agreed that Kurin Lock was appropriate for use in the secondary care blood culture sampling pathway, to reduce blood culture contamination rates. It could be used to collect peripheral blood culture samples. The experts advised that Kurin Lock may improve outcomes for patients by reducing the number of repeat blood culture samples. One clinical expert stated that they had reduced contamination rates using an alternative method. This method involves staff training, ensuring samples are not taken from peripheral cannulas and introducing an additional blood bottle, which is used to isolate the first few millilitres from the blood culture sample. The expert advised that this approach reduced blood culture contamination rates, but the change in practice (one additional step) needs to be regularly reinforced and may be time-consuming. Another expert had trialled this approach before using Kurin Lock and felt that Kurin Lock is easier to adopt, and from their experience works just as well with cannulas as venepuncture.
The high cost of Kurin Lock compared with standard blood culture collection is a barrier to using it in the NHS, unless there is better evidence showing its impact on resources. One clinical expert commented that his hospital considered the Kurin Lock cost too high, and it has explored using an alternative approach to reduce contamination rates. The committee understood that the cost saving from the economic modelling relied on the high device cost being offset by a reduction in resource use, including hospital length of stay, associated with fewer false-positive results. Further evidence generation is needed to show that these cost savings will be realised in NHS clinical practice.
The committee and clinical experts agreed that there are usually higher blood culture contamination rates in emergency departments than other hospital wards. The EAG noted that the clinical evidence is based in emergency departments only and the economic modelling used an emergency department setting. In published literature, contamination rates of up to 9% were reported in NHS emergency departments. But clinical experts stated that it could be significantly higher. During consultation, the company stated that Kurin Lock is also likely to be cost saving in settings with a high bed-day cost, such as intensive care units. The EAG confirmed that the higher daily costs in settings such as intensive care will increase the likelihood of cost savings. Clinical experts advised that although daily costs are higher in intensive care, contamination rates are usually lower than in emergency departments. Also, length of stay in intensive care is more likely to be driven by the need for organ support. The experts also advised that while it is likely that the reduction in contamination rates seen in emergency departments are generalisable to other settings such as intensive care, the impact of the reduction in contamination rates on people in intensive care is less certain. The committee concluded that with the focus of the current evidence in emergency departments, and the associated uncertainty about the impact of Kurin Lock in intensive care units, the recommendation should only be in emergency department settings.
The company's base-case model used an emergency department setting. It used a daily ward stay cost taken from patient-level data from 1 NHS trust, which was described as a non-elective short-stay cost. This was applied as a daily cost for the duration of the patient stay. The EAG did not have access to this cost data and considered the daily costs very high compared with other economic models for guidance development. The EAG used a non-elective short-stay cost as the first day of stay cost, and then calculated excess stay costs for additional days using costs from NHS Cost Collection data. The committee agreed with the changes the EAG made to the company's base-case model. It considered that the lower daily hospital stay cost used by the EAG was appropriate. After consultation the EAG obtained patient level information and costing systems (PLICS) data from NHS Wales, which resulted in a lower stay cost compared with the company model, but was higher than the EAG model. The committee concluded that there is too much uncertainty, including around the assumption of bed days saved. Accurate information on the costing of the hospital stay and the length of stay would help reduce uncertainty in the economic modelling.
If the difference in length of stay for people with true-negative blood culture results and false-positive blood culture results is overestimated, then the cost saving is reduced. This could lead to Kurin Lock being cost incurring rather than cost saving. The EAG confirmed that there is no length of stay difference data directly related to Kurin Lock that can be used instead of the Skoglund (2019) values to reduce the uncertainty of the model results. Because the main driver is the length of stay difference, the committee was cautious in its interpretation of the base-case results. Assuming a length of stay difference of 0.5 days, Kurin Lock would need to be £10 to achieve a break-even cost when using EAG PLICS data for bed day costs.
There is uncertainty about the cost effectiveness of introducing Kurin Lock, because the sensitivity analysis showed that it can be cost saving or cost incurring. Probabilistic sensitivity analysis of the EAG model reported a 62% probability that Kurin Lock would be cost saving compared with standard blood culture collection. The committee considered that the lack of evidence on the resource impact from using Kurin Lock is a significant limitation of the economic model. Further evidence generation including information on the length of hospital stay from NHS hospitals using Kurin Lock could be used to revise the economic model.
The key uncertainties about using Kurin Lock in the NHS are related to its cost effectiveness. Sensitivity analysis showed that Kurin Lock can be cost saving or cost incurring depending on the parameters used, particularly around the length and cost of hospital stay. The economic model in the base case assumes a 2‑day difference in length of stay between people with true-negative and false-positive results, which may be overestimated. The clinical experts advised that length of stay and treatment after a blood culture test is complex and depends on many factors. The committee concluded that evidence generation alongside using Kurin Lock in the NHS would provide an opportunity to collect resource impact data that could inform economic modelling in a future review of the guidance. NICE expects to review the guidance in 3 years or sooner if evidence becomes available.