Statistical Process Control Charts
Conditions
Brief summary
Hospitals collect copious amounts of data to share with their board for quality assurance and improvement purposes. The way these data are presented can influence board members' decisions. For example, time-series charts highlight the highest and lowest data but do not clarify whether those data lie outside expected or 'common cause' variation. Statistical process control charts make this clarification and, in so doing, guide quality assurance and improvements in a more targeted fashion. A previous study showed that data suitable for presentation as a control chart are seldom presented in that format. A training intervention called 'Making Data Count' was created to improve the uptake of statistical process control charts by hospitals in England. The current study will use a randomized design to evaluate whether the intervention increases control charts in hospitals that were not early adopters of the training intervention. This study is a parallel cluster randomized trial (with baseline-line measurements) across 20 National Health Service (NHS) hospitals in England. The hospitals will be randomly split into two groups. One group will be scheduled to experience the training intervention, and the other group will be placed on a waiting list to experience the training later. The primary analysis will compare the difference in the use of control charts between waitlist and intervention hospitals (adjusting for pre-intervention use) reported with 95% confidence intervals. A qualitative thematic analysis of feedback forms will be conducted. The present research will evaluate the impact of the training intervention on the use of control charts. The results will apply to institutions that are not early adopters of this training intervention.
Interventions
BEHAVIORALMaking Data Count Training intervention
The Making Data Count training intervention was designed by NHS-Improvement to improve knowledge about SPC charts and to increase their uptake. Training sessions are tailored for two sets of attendees: board members and data analysts. Board member and analyst training sessions are delivered as close as possible in time, typically within the same month. Training sessions for board members are usually delivered over about one-and-a-half hours and focus more heavily on the benefits of control charts compared to other charts. Training sessions for analysts are usually delivered over three hours and focus more heavily on the structure and interpretation of the individual and moving range charts (X-mR charts).
Sponsors
University of Birmingham
University of Warwick
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
NONE
Masking description
Participants cannot be blinded as this is a behavioural intervention. Some of the researchers can be blinded. Specifically, it is not possible to blind reviewers R1 and R2 to the board papers group or time-period, as those researchers are extracting charts from the published board papers. R3 and R4 will be blinded and will be instructed to inform R1 if those researchers are unblinded at any point, and whether/how often those researchers are unblinded will be reported in the final manuscript.
Eligibility
Sex/Gender
ALL
Healthy volunteers
Yes
Inclusion criteria
* NHS Hospital * Placed on a red-list for predominantly using R-A-G or two-point data presentations in their board papers.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Change in the proportion of SPC charts out of all quality and safety charts from baseline to five months post intervention | One month pre-intervention and then five months post-intervention. | Change in the proportion of SPC charts (the numerator) out of all charts indicated in each board paper (the denominator) |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Chart types | One month pre-intervention and then five months post-intervention. | counts of whether the charts are time series charts, between subjects charts, time and between charts, or other (e.g., pie) |
| SPC elements | One month pre-intervention and then five months post-intervention. | counts of whether the charts include recalculated control lines, highlight special cause variation, use R-A-G colouring, and whether the accesses are labeled. |
| SPC supporting text | One month pre-intervention and then five months post-intervention. | counts of whether the charts include recalculated control lines, highlight special counts of whether SPC charts are explained in supporting text, regarding where control lines are set, reasons for variations, and suggestions for improvements. |
Other
| Measure | Time frame | Description |
|---|---|---|
| Feedback forms | immediately post intervention | optional forms participants fill out after training to improve future sessions. Intervention group only. |
Countries
United Kingdom
Outcome results
None listed