DISCUSSION
The GIRFT programme is funded by the United Kingdom Department of Health. It aims to improve patient care through highlighting unwarranted variations in clinical practice. Clinical outcome data are presented to clinicians highlighting the performance of their trust relative to all trusts in England. Where unwarranted variation in outcomes is identified, clinicians and managers are engaged to identify ways in which outcomes could be improved. Ultimately, the approach relies on decision-makers within individual trusts recognising that their outcomes may be sub-optimal and then working constructively to improve. The use of data to benchmark performance and to identify subsequent improvements is key to this approach. Our study investigates the impact of the GIRFT programme on three specific recommendations from the GIRFT National Report in urology.
Use of day-case surgery for TURBT was increasing steadily prior to GIRFT visits, but the rate of increase more than doubled after the GIRFT visits. Based on our analysis, the GIRFT programme has had a significant impact, reducing use of hospitals beds for overnight stay. Our team have previous reported outcomes for TURBT performed as day-case surgery to be at least as good as those for TURBT performed with an overnight stay.10 Whilst not all patients undergoing TURBT will be suitable for day-case surgery, with some requiring an extended period of monitoring due to co-morbidity, frailty or disease severity, the increased use of day-case surgery for TURBT is encouraging. Nevertheless, with a number of trusts performing over half of all TURBT procedures as day-case, there appears to be significant scope for greater use of day-case surgery for many trusts.
For stent use to manage ureteric stones, we identified a significant temporal trend towards greater ureteric stent use prior to GIRFT visits and a significant change in trend after visits. We recognise that use a ureteric stent will be the best treatment option for some patients presenting with urinary tract stones, notably those with associated sepsis, and in these cases its use is clearly appropriate. However, complications associated with stenting are relatively common.11 Consequently, its use purely as a short-term measure, where other evidence-based, definitive treatment options are indicated and available, should be discouraged.12During the same period, use of ureteroscopy or ESWL increased significantly after GIRFT visits. Such changes are likely to reduce the need for return admissions and so represent an efficient use of resources but are also likely to improve patient quality of life by offering definite treatment at an earlier time point.
Although there was a significant trend towards longer waiting times for male bladder outflow tract surgery after a presentation with urinary retention prior to GIRFT visits, which was not apparent afterwards, this change was not significant, suggesting any impact of GIRFT visits on waiting times was limited. The reasons why this recommendation was not implemented to the same extent as the other recommendations investigated are likely to be multifactorial, complex and vary from trust to trust. However, reducing surgery waiting times against a backdrop of increase pressure on services, due to factors such as demographic ageing and increasing morbidity of the background population, is likely to be challenging.13 Indeed, the steady and significant increase in outflow tract surgery waiting times during the baseline period provides evidence of this. Compared to the other recommendations investigated, reducing waiting times may be a more intractable problem that the changes to practice and service organisation required to move to a day-case model for TURBT or to reduce stent use for ureteric stones. Moreover, these other recommendations have the potential to either reduce costs, free-up beds or reduce human resource requirements. Reducing waiting times may only be achievable though increasing bed usage and staff numbers. Nevertheless, the GIRFT methodology is based on engagement with key stakeholders at a local level to help understand how recommendations can best be implemented within a given setting.14 Ongoing support is offered to trusts to help them implement recommendations. Further investigation of barriers and facilitators to implementation of GIRFT recommendations is needed. In other areas of healthcare, where barriers and facilitators to implementation of guidance has been investigated, resource constraints have been identified as a key barrier15, 16 and engagement with clinicians locally, peer support and use of data to change perceptions were found to be facilitators.17
ITSA is a potentially powerful technique for the analysis of the impact of an intervention.18 It is particularly useful where a randomised controlled trial is not practical, and this is often the case when evaluating the impact of a new healthcare policy or quality improvement programme.19 However, one of the major threats to the validity of ITSA is that it is often not certain that it was the intervention of interest that caused any observed change. In our study this is less of an issue, since the visits to individual trusts were spread over 20 months, with each trust having its own intervention point. For the same reason the impact of seasonal factors is likely to be small, and the lack of any significant autocorrelation supports this view. A second threat to ITSA validity is changes in the way data are recorded over time.19 In our case, this is also unlikely to be a major problem since data are recorded by experienced data coders and there have been no major changes in the codes used over the time period covered. As such, our findings are likely to be relatively robust.
One further limitation of the ITSA approach we used is that there is no comparator group that received no intervention. We are unable to comment on how stent use would have changed without our interventions.
In summary, we present evidence that GIRFT programme recommendations around increasing rates of day-case surgery for TURBT and reducing stent use and increasing use of ureteroscopy or ESWL for ureteric stones have had an impact on clinical practice across England. The recommendation to reduce waiting times for male bladder outflow surgery have, as yet, had no significant impact on clinical practice. Further research should look to investigate reasons for good and poor adoption of recommendations and clinical guidelines within health services with a view to establishing the most effective methods of influencing practice. Although the GIRFT methodology may have helped to remove a number of barriers to effective implementation of changes to clinical practice, better understanding how the methodology can be adapted to address more intractable issues is needed.