The biggest challenge? Filling them out.
Only those who work in the hospital setting can appreciate how challenging it is to keep up with clinical quality reporting. Both clinicians and those working on the data side get it. CMS alone has over 150 clinical quality measures, and there is an alphabet soup of other regulators requesting their own metrics.
And then there are the surveys hospitals do to be considered for ranking. US News and World Report (USNWR) is among the most famous of these. USNWR routinely issues surveys on which it bases its famous “Best Hospitals for X” lists. Participation is not required. But it doesn’t hurt. Rankings in this context can impact consumer confidence. After all, when you seek a doctor for care, the first characteristic you probably think about is “high quality”. And you don’t take their word for it. You look to some other entity to validate that reputation.
Putting to the side, though, whether these rankings matter or should matter to patients or providers: the things they measure do matter to hospitals themselves. In general, hospitals want to do well in the same domains these surveys evaluate. The problem isn’t one of misaligned goals. It’s one of, well, paperwork.
Take the USNWR Pediatric Survey. RTI International manages administration of this survey for USNWR, which is comprehensive to say the least. The pediatric survey has twelve sections, each several pages long. While the survey can be completed electronically, it is not structured in a way that permits the questions to be handled in a streamlined way or to facilitate integration into the hospital’s existing databases or dashboards. But this is exactly what one would want to be able to do. Most hospitals would be delighted to integrate survey questions with their existing digital reporting to monitor performance on these metrics and guide quality improvement initiatives, while also streamlining response time. The challenge is how to get that done.
Background
New York-Presbyterian Hospital (NYP) is a world-renowned hospital system that ranks consistently among the top hospitals in New York City and in the nation. NYP routinely participates in USNWR surveys that rank quality of care.
The Challenge
NYP wanted to integrate its response process to the USNWR Pediatric survey with pediatric quality improvement initiatives, while also lessening the administrative burden of collecting and calculating responses for multiple campuses. To accomplish this, a few issues needed to be addressed. The first was that the survey needed to be electronically parsed, treating questions like data themselves for storage in a database. The second was that the questions needed to structured into a data model that could serve a dashboard. The third was that NYP’s expansive legacy code base for querying data to calculate question responses needed to be transformed to avoid massive reworking of historical work.
Solution
Over the years, NYP’s data team had already coded most survey questions as SQL queries. Only a small number of survey questions change each year, so the existing code base was too valuable to toss out. However, since it hadn’t been organized in a database, individual queries had no relationship to one another and were being handled in a partially manual way that was cumbersome. This code base was rapidly transformed by leveraging dynamic SQL to standardize the structure of individual question tables en masse into one large master table.
Questions were then grouped into “families”, enabling cross referencing. This enabled dashboard users to dynamically select the questions of interest based on the focal point of any given quality improvement initiative, and drill in and out of patient populations that composed the questions.
Results
NYP’s extensive USNWR pediatric survey legacy code base was able to be transformed into a data model to serve a dashboard within three months. A dynamic dashboard was mocked up and approved by clinical teams that could be used to guide quality improvement initiatives, integrating USNWR questions. Stored procedures ensure data can be batch refreshed as often as needed, replacing the need for manual updating. Each NYP campus can isolate questions according to priorities, question relationships, and according to its own patient population. All work was able to be completed with no disruption to the existing survey response process.
Delivered:
Want more solution inspiration? Return to the gallery to read more.