Time Series Studies gold standard for reviewing infection control data
Conference told randomised controlled trials ineffective for analysing infection statistics
Measuring the impact of infection prevention and control interventions means rejecting gold standard study methods in favour of more challenging approaches to gathering and analysing data, experts have warned.
Speaking at this week’s Infection Prevention 2012 conference in Liverpool, Dr David Jenkins, a consultant medical microbiologist at the University Hospitals of Leicester NHS Trust, said the future of effective infection control procedures was reliant on accurate data on current trends and the impact of individual interventions.
To know whether the actions you take are having an effect on preventing infection, you have to review the evidence base. This is a difficult thing to do because you are ultimately trying to demonstrate that success occurs when nothing happens
But he warned that randomised controlled trials (RCTs), which are seen as the gold standard study design for most medical literature, were not effective when reviewing infection prevention research.
He explained: “To know whether the actions you take are having an effect on preventing infection, you have to review the evidence base. This is a difficult thing to do because you are ultimately trying to demonstrate that success occurs when nothing happens.
“Unfortunately, big pharma companies are unlikely to fund large, expensive RCTs and these are very difficult to apply to infection prevention settings.”
Instead, he recommends using time series studies, which basically chart levels of infection over a particular period of time with details of interventions introduced along the way. This enables those reviewing the data to see which initiatives have a measurable impact.
Dr Jenkins said: “With time series studies you are building up a body of evidence and it is an acceptable alternative to RCTs when they are not applicable.”
An example of where this type of study has illustrated the impact of a particular intervention was the letter sent to health trusts by former Health Secretary, John Reid MP, demanding improvements to infection control measures at hospitals.
The study could demonstrate a measurable positive impact on infection levels as a direct result. This is the best evidence we have that this is the best way forward for monitoring data for infection control
Dr Jenkins said: “A time series study was underway at the time and when John Reid sent his letter and chief executives of hospitals realised they might lose their jobs unless they did something, it interrupted the series. The study could demonstrate a measurable positive impact on infection levels as a direct result. This is the best evidence we have that this is the best way forward for monitoring data for infection control.”
Any trust looking to begin a study is advised to use the ‘R’ statistical support package, a free programme available online that can help to co-ordinate and read data.
Dr Jenkins said: “Once you have collected data, you do need to play around with it and this can be quite an involved process. The R solution is free and is the best support available. It is not a point-and-click tool and you do have to type things in, but it is a good way of learning from the information collected.
