Analysis and attention based on value

To move further toward value-based care, organizations need to make greater use of data analytics. To improve outcomes for patients at lower cost with higher patient satisfaction, data needs to be collected and analyzed. There are many ways to do this and I will point out a few, including some that I use.

Healthcare is awash in data. There is billing data, including diagnosis and procedure codes. There is data based on payer claims. There is much more detailed data in EHRs, like test results. These data are clinical data. Financial data is also important. Much of this data can be found in financial software like QuickBooks and patient management software.

To make effective use of this data, it is necessary to organize, graph, and investigate significant changes. Each of these requires the use of statistics, either at an elementary level to spot trends or at an advanced level to see if there are real and significant changes that indicate the provider is moving toward providing better care at lower cost. The examination and analysis of data collected from patients and grouped by patient characteristics is called population-level health management.

One approach to using analytics to improve patient health is to examine and track key indicators for chronically ill patients, typically the patients who incur the highest cost in a practice. For example, a practice may collect and analyze the A1c levels of all patients diagnosed with diabetes. This data can be collected each month and a mean and standard deviation determined. A month-by-month chart using an Excel spreadsheet will help visually indicate any trends that are occurring. If over a period of several months there is an upward trend, steps can be taken to break the trend. This could be making use of a nurse coordinator to help patients better manage their diabetes. Engineering statistics control charts could be used to better analyze whether the trends are real or due to random fluctuations that are normal in any data collection over a period of time. Statistical t-tests could also be used to determine if the changes are really significant or not.

It can be very helpful to graph the collected data such as A1c levels, plot the means over time as above, and present them in a Dashboard with brief descriptions and analysis. These boards can be shared in a practice to drive improvement. This can be very effective in making improvements with a care team. I recently listened to an NPR podcast of ‘The Hidden Brain’ that described how a hospital in Pittsburgh improved handwashing by caregivers before entering patient rooms. The rate hovered around 10% for a long time despite repeated education of caregivers. The hospital then began displaying monthly handwashing statistics on a dashboard that everyone could access and view. The administration focused the caretakers’ attention on the boards. Handwashing rates quickly improved to 90% and stayed there. The images had a significant impact on providers’ awareness of handwashing.

Analytics can also be used to improve patient satisfaction scores. The Medical Group Management Association (MGMA) provides a very good patient satisfaction survey for its members. I have adapted it to different providers depending on their demands. The survey covers 36 basic questions and ends with “Would you recommend the provider to others?”, a very good final question. I also add demographic questions. Providers can use the survey with patients and track performance in five areas: appointment making, quality of front-desk and billing staff, ease of communication, doctor visits, and facility condition. The goal is to have as high a composite score as possible for each of these areas. With advanced analytics you can discover more to improve satisfaction. It is possible to identify which of the questions has the greatest impact on the last question. It has been found that patients who are most likely to refer friends and family to a practice are often the most satisfied. So finding out which of the many questions have the most impact on this can help identify which areas need improvement. Analysis of which questions have the most impact should be carried out on a regular basis, as the ones with the most impact can change over time.

Of course, a dashboard should be created to inform staff of survey results every month or so. This will help incentivize staff to perform even better. If management likes, they can break down the results by staffing area or vendor to help identify where people can improve. Individual coaching can be used to help staff make improvements. However, boards should never be used to reprimand staff. Whether individual boards are shared with other staff depends on how well the staff function as a team, how supportive they are of each other.

The MGMA collects a great deal of data from its members through surveys. It then provides data for benchmarking to its members, some for free and some for costs. Providers who participate in their surveys can often get the results for free. In a recent article in his monthly publication Access provided a dashboard of data on the most profitable independent practices. It found, contrary to common sense, that those organizations with the highest average costs per physician FTE were also the most profitable. Some of the data on the dashboard is as follows:

Median Total Medical Revenue per Physician FTE = $1,169,542

Median total operating cost per physician FTE = $630,680

Median Compensation and Total Physician Benefit = $462,722

Median total support staff per physician FTE = 5.12

Article, Designing the practice of the future, found in the March 2017 issue of the journal, also provided the baseline data for all multi-specialty practices. This data, along with other article data and article analysis, can help providers develop long-term strategies to improve their practice’s profitability.

As you can see, there are many ways to use data and data analytics to improve outcomes at your care site. It is important to identify what data to collect and analyze to achieve the most significant impact. Like I said, care providers are inundated with data and trying to analyze it would be a waste of time and energy. Those just starting out with advanced analytics should start with a few data projects and expand as time goes on. Smart use of data and analytics can have a significant impact on the care you provide and your outcomes.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top