Explanation & Elaboration

10. Planning the study of the intervention

1. Outlines plans for assessing how well the intervention was implemented (dose or intensity of exposure)
2. Describes mechanisms by which intervention components were expected to cause changes, and plans for testing whether those mechanisms were effective
3. Identifies the study design (for example, observational, quasi-experimental, experimental) chosen for measuring impact of the intervention on primary and secondary outcomes, if applicable
4. Explains plans for implementing essential aspects of the chosen study design, as described in publication guidelines for specific designs, if applicable (see, for example, http://www.equator-network.org)
5. Describes aspects of the study design that specifically concerned internal validity (integrity of the data) and external validity (generalizability)

Example

"The specific aim of this project was to eliminate catheter-related blood stream infections (CRBSI) throughout intensive care units (ICUs) in Michigan. We used a multiple time series design. The study population was all Michigan ICUs willing to participate.

The primary outcome (dependent) variable was the rate of CRBSI...Hospitals were to adhere to the Centers for Disease Control (CDC) definition of catheter-related bloodstream infection during the study period...The quarterly rate of infection was calculated as the number of infections per 1000 catheter-days for each 3-month period. Quarterly rates were assigned to one of eight categories on the basis of when the study intervention was implemented: at baseline, during the implementation period, or during one of six 3-month intervals occurring up to 18 months after implementation.

...To understand compliance with the interventions, teams completed a monthly form, called the team check up tool that evaluated what activities they did during the prior month, the number of times they met with their team and senior leader, and the barriers they were facing. All teams participated in bi weekly conference calls to discuss the project." [29]

Elaboration

This study highlights the multiple tradeoffs that occur in designing and reporting a quality improvement study. This report states the design as a multiple time series in which all ICU sites willing to participate were included. While a cluster randomized design may have been more robust, it was not tenable because all the teams wanted the intervention and none wanted to be randomized to the control - a common desire when testing the implementation of known, effective interventions. Although not necessary to note this, alternative designs that were considered and deemed to be untenable should be noted in the limitations. The effectiveness of the intervention for eliminating catheter related blood-stream infections (CRBSI) was clearly known from prior studies [30], so this was an assessment of the effectiveness of the implementation of the interventions across many sites. The description of this should account for the size and scope of the implementation. This example was a large-scale implementation across a large geographic area, so the description includes mention of the study design ("multiple time series"), the operational definition of the main outcome measure ("the Centers for Disease Control (CDC) definition of catheter-related bloodstream infection"), and how data integrity was maintained (example not included). An improvement study at one location may include a very different, but equally precise, description of the plan of the study of the intervention.

There is often a struggle between collecting data that is scientifically sound yet feasible. Quality improvement studies often have limited resources and data collection may be voluntary. As such, this study notes the reduced quantity - but not quality - of data collected. The definition of CRBSI, the main outcome measure, was clear (although not included in the above example), but there were no data collected about secondary outcome variables such as the type of organism or adherence to the evidence-based recommendations. There was no valid and feasible way to do this as independent observation would have required placing an observer in the ICU for substantial amounts of time; a costly endeavor for central lines which are often placed at irregular intervals. The study also reported on data quality control including data definitions, training, and amount of missing data. Nevertheless, due to insufficient resources, it did not report more robust methods of data quality control such as duplicate data entry or inter-rater reliability.

This study demonstrates how the intervention was guided by the study goal: eliminate CRBSI. When designing any study, the team invariably faces tradeoffs. This example demonstrates how the aim, the design of the study, and the measures were identified and implemented across multiple sites in a large geographic area. Improvement designs are sometimes limited by the systems willing to make changes. There are many ways to design a quality improvement study, and the study aims guide that design and the analysis. When writing about the study of the interventions, be sure to integrate the SQUIRE publication guidelines with guidelines for reporting on the specific study design whether it is a qualitative epidemiologic study (STROBE) [4], a randomized controlled trial (CONSORT) [1], [2], or other design.

References

1. Boutron I, Moher D, Altman D, Schulz K, Ravaud P, Group* ftC. Extending the CONSORT Statement to Randomized Trials of Nonpharmacologic Treatment: Explanation and Elaboration. Annals of Internal Medicine. 2008;148:295-309.

2. Moher D, Schulz K, Altman D, Group* ftC. The CONSORT Statement: Revised Recommendations for Improving the Quality of Reports of Parallel-Group Randomized Trials. Annals of Internal Medicine. 2001;134:657-662.

4. von Elm E, Altman D, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. Annals of Internal Medicine. 2007;147:573-577.

29. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. The New England journal of medicine. 2006;355(26):2725-2732.

30. Berenholtz SM, Pronovost PJ, Lipsett PA, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Critical care medicine. 2004;32(10):2014-2020.

to top ^