Preparing and reporting data to a practice or its quality improvement (QI) team is one of the most important steps in data-driven improvement and one of your most important roles as a facilitator. In preparing to do this, you will need to ensure that the data you have collected are accurate. You will also need to make sure that you have summarized them and presented them in a way that makes it easy for members of the practice to understand them and where their performance falls in the context of other similar practices or patient populations. Finally, you will need to be prepared to respond to members’ questions and challenges of the data and to help them reflect on the findings and use the findings to stimulate meaningful action toward improving the practice.
Cleaning and Validating Data
An important step toward presenting data to your practice is preparation. Once the data have been collected and you have entered them into a database, you will need to review them for missing entries, internal inconsistencies, or out-of-range values (e.g., systolic BP of 1125). These need to be corrected or removed from the spreadsheet.
One way to clean and validate data is to manually check the data in your data collection spreadsheets. Look at the data and ask yourself: Is each number plausible? Does the sequence of dates make sense? Do any of the data elements conflict with each other?
Another method is to run frequencies using a statistical program. A number of good online training programs teach basic skills for working with data using statistical software programs. These programs can identify data outliers and inconsistencies.
You will need to talk to staff and clinicians at the practice to better understand the validity or other problems with the data you are collecting. With electronic health records (EHRs) in particular, there can be data elements filled with meaningless data, entered simply to fill a required field. Talk to practice staff to find out whether there are any “junk codes” where the data are not what they appear to be. Much of your early work setting up performance systems is likely to focus on getting data and fixing data so they are accurate.
It is also common for entry fields on EHRs to have been inadvertently mapped to the wrong variable labels in the underlying databases, which are used to generate reports on patient care and practice performance. These mistakes can be difficult to identify but can introduce significant errors into any patient and performance reporting. Clinicians and staff can alert you to areas where these mapping mistakes may exist. When results are inconsistent with what is expected, or seem “strange” to clinicians and staff, this should be a red flag to check for mapping errors.
Describing Your Methods
When preparing reports, be sure to include a description of the methodology. How was the patient sample generated? What time period was used? What were the search parameters? Were any potential respondents or data sources excluded and why? This information is essential for interpreting the results accurately.
Failure to provide sufficient detail when you report data to the team can make the data difficult to interpret and validate. Providing too much detail, on the other hand, can bury the team in data and make it difficult for them to make inferences based on the data (Gregory, et al., 2008). For each performance metric, you should clearly describe the methods you used to obtain the data, the exclusion and inclusion criteria, and the denominators and numerators used to generate percentages. Part of your job as a facilitator is to help practices organize their performance data so that it can be easily understood and so that it is actionable.
|A picture is worth a thousand words.|
A picture paints a thousand words and nowhere is this truer than when presenting performance data. Graphic displays of the data are extremely effective in reporting data to the QI team. Visuals allow people to absorb large amounts of data quickly. Spreadsheets can be programmed to generate visual displays of key system and clinical performance data quickly and efficiently, which can make generating performance reports easier for both you and your practices. Ideally, you will be able to work with the information technology (IT) manager at the practice to build these reporting processes into the EHR and practice management systems.
When developing reports, you should include both raw numbers and percentages on the graphic whenever possible to make them easy to interpret. Also include the total N for each summary statistic. Make sure that values are clearly labeled and legends provided. Data are most compelling when mapped over time through the use of trendlines. QI teams can use these data displays to monitor progress over time and make decisions about QI priorities, training for staff, and revision of processes based on these cumulative data.
Different graphics are effective for presenting different types of data. Data that represent a single point in time can be presented using static displays such as bar graphs and pie charts. Data from multiple time points designed to track trends or changes over time are best displayed in more dynamic formats such as run charts. When possible, use graphics to make the data more accessible to your practices. Table 9.1 presents an example.
Helping Practices Reflect and Act on Data
Many if not most times, the information systems in practices contain errors. Errors mapping data entered into an EHR to the database variables are frequent. Expect clinicians and other members of the practice to question the data you present to them. When this happens, it is important that you listen carefully to their discussion of the errors that they believe exist in the data. You will then work with clinicians and often their IT staff to correct these errors and the corresponding performance data. It is not unusual for a practice facilitator to spend a considerable amount of time during the early stages of working with a practice correcting mapping errors in EHRs and other data systems.
Once you have helped the practice correct these errors and can present the corrected data again, you will be able to engage members of the practice in a productive discussion of the findings. Often clinicians and staff believe that they are performing better than they actually are, so the data you present are likely to stimulate robust discussion. It is important that you not become defensive or take challenges from practice members as a personal attack. Instead, it can be helpful to see yourself as an “ally” in helping them to acquire, reflect upon, and then use these data to help them improve performance.
When presenting performance data to a practice for the first time, it can help to enlist a leader from the practice as the main presenter, or as a co-presenter with you. It can also help to come prepared with a series of questions designed to help members of the practice reflect on the data. Some useful questions to ask include:
- How accurately do you believe these data reflect your practice?
- Are there problems with the data that should be considered or corrected before use?
- What findings did you expect?
- What findings were a surprise?
- What do these data suggest to you regarding setting goals for improvement at your practice and prioritizing these goals?
Gregory B, Van Horn C, Kaprielian VS. Eight steps to a chart audit for quality – a simple chart review can help your group answer the question on everyone’s mind: “How are we doing?” Fam Pract Manag 2008 Jul-Aug;15(7):A3-A8.