Changes in coding of pneumonia and impact on the Hospital Readmission Reduction Program

Area of Science:

• Health Services Research
• Healthcare Policy
• Pulmonary Medicine

Background:

• The Hospital Readmission Reduction Program (HRRP) aimed to reduce 30-day hospital readmissions for conditions including pneumonia.
• Concerns have been raised that changes in diagnosis coding practices might artificially inflate reductions in readmission rates.

Purpose of the Study:

• To investigate whether shifts in diagnosis assignment explain the observed decreases in 30-day pneumonia readmissions following the implementation of the HRRP.

Main Methods:

• A retrospective cohort study analyzed Medicare data from 2008-2015 for HRRP-eligible hospitals.
• Readmission rates were compared using a "narrow" definition (pre-October 2015) and a "broad" definition (post-October 2015) including sepsis and aspiration pneumonia.
• Changes in readmissions were assessed across pre-HRRP, HRRP implementation, and HRRP penalty periods.

Main Results:

• Under the narrow definition, readmissions decreased by 1.07 pp during HRRP implementation but showed minimal change during the penalty period (-0.09 pp).
• Under the broad definition, readmissions decreased by 1.28 pp during HRRP implementation and also showed minimal change during the penalty period (-0.09 pp).
• The observed reductions in 30-day pneumonia readmissions were not attributable to changes in diagnosis coding.

Conclusions:

• Changes in the coding of inpatient pneumonia admissions do not account for the readmission reductions observed after the Hospital Readmission Reduction Program.
• The findings suggest that the decrease in readmissions is likely due to genuine improvements in care or other factors, rather than coding artifacts.