Mark Wallace, Ph.D.
Hospitals make strides in key patient mortality index
Vanderbilt hospitals are succeeding in lowering risk-adjusted patient mortality rates.
Hospital patients typically have multiple documented diagnoses, complications and co-morbid conditions, and together these factors determine a patient's expected mortality score, which is a type of statistical prediction of the risk of death during admission.
Every inpatient gets a score and every patient's score figures in the calculation of a hospital's expected mortality. Vanderbilt Medical Center has made observed to expected mortality a top quality improvement measure, using as its comparison group 54 academic medical centers with Level 1 trauma centers and membership in the University Hospital Consortium (UHC).
With health care continually advancing, UHC annually recalibrates the formula for expected mortality.
Using the most recent version of UHC's formula, VMC mortality decreased from 93 deaths per every 100 expected deaths in calendar 2006, to 76 deaths per every 100 expected deaths in calendar 2007. In 2006, VMC ranked 18th out of 54 centers in terms of observed to expected mortality, and in 2007, VMC improved to fourth out of 54 centers.
“As we strive for improvement, we can gain encouragement from seeing how our diverse quality initiatives are bearing meaningful benefit for our patients,” said C. Wright Pinson, M.D., M.B.A., associate vice chancellor for Clinical Affairs and chief medical officer. “We want to be the very best in quality and we know quality has to always be our first goal. Our steady improvement in this vital quality metric is due to the focus, care and hard work of clinicians, nurses and staff throughout our hospitals.”
Along with reductions in inpatient mortality, part of the improvement in observed to expected mortality at VMC is due to more accurate patient care documentation, and thus more accurate expected mortality scores.
A staff of eight full-time VMC clinical documentation coordinators reviews inpatient charts and queries clinicians whenever documentation in the medical record appears inadequately explicit for purposes of retrospective quality assessment and accurate billing.