Trauma programs and their registries are very good at abstracting and compiling a wide variety of data points on admitted trauma patients. They are not so great at recognizing readmissions after the original event. And they completely fail to capture (or at least link) readmissions to another hospital after the initial injury.

The US federal government implemented a Nationwide Readmissions Database (NRD), which provides information on patient readmissions nationally across all payors and the uninsured. This is extremely important data which provides interesting data about a population that is normally very difficult to identify.

A multidisciplinary group at Johns Hopkins analyzed the NRD for trauma patients age 15 or greater over a six year period. Patients were excluded if they were transferred, died during the initial hospitalization, or were admitted to a low trauma volume hospital (<100 patients per year). Readmissions within 1, 3, and 6 months were analyzed and statistical tools were applied to help identify predictors of readmission.

Here are the factoids:

• A total of 3 million trauma patients were identified, with 93% blunt, 6% penetrating, and 1% burns
• Readmissions were 10% within 1 month, 20% in 3 months, and 26% within 6 months (!)
• These numbers remained relatively constant across all three mechanisms
• Predictors of readmission, with odds ratios, included:
  • male gender (1.15)
  • lowest income quartile (1.04)
  • number of comorbidities (1.17)
  • leaving the hospital AMA (2.32)
  • initial admission to a private hospital (1.17)

The authors concluded that understanding these factors provides an opportunity for quality improvement and offers implications for hospital benchmarking.

Here are some questions for the authors and presenter to consider in advance to help them prepare for audience questions:

• Is there any indication in the data about why patients were readmitted? This would offer even more specific information to help focus quality improvement efforts.
• Can you make any specific recommendations at this point as to how to begin to identify patients who have a higher potential to be readmitted? The odds ratio for income was not very high. Gender and AMA were more predictive, but apply to quite a few of our patients.
• Can you speculate about why readmission risk increases for private hospitals? This is not intuitive to me.

This is another excellent and provocative paper from this group. I’m looking forward to hearing the nitty gritty during the presentation.

Reference: Unplanned readmission after traumatic injury: a long-term nationwide analysis. EAST 2019, Quick Shot Paper #27.