State Nursing Home Quality Improvement Programs: Site Visit and Synthesis Report. Impact of Missouri's Quality Improvement Programs on Quality of Care/Quality of Life


Missouri is far ahead of other states in terms of evaluating the impact of its quality improvement programs. Dr. Rantz uses MDS data to measure the change in quality indicators associated with the program and has published several journal articles that report these results. In addition, facilities that receive QIPMO visits provide feedback through an evaluation form that they are given at the end of each visit. Analysis of MDS data, suggests that even poor performers are doing better, either due to QIPMO or other changes that have occurred in the past few years.

QIPMO nurses are noting increased levels of MDS understanding and more sophisticated questions, suggesting that the information offered in the support groups is having a positive effect. Participants in the meetings find them to be an extremely valuable resource--a common theme is that knowing that they are not the only one facing particular issues is a major help. One coordinator said that if "they had this kind of support at my last job, they wouldn't have taken me out of the facility on a stretcher with oxygen." The MDS Coordinator at one nursing home the researchers visited participates in the support groups, which she describes as being "very helpful."

Since the implementation of QIPMO, there have been improvements in 16 quality indicators and a decline in only six measures. In addition to analyzing median quality indicator scores, the research team analyzes trends for the 90th and 95th percentile, so that the effectiveness of the program in improving outcomes for low-quality facilities can be understood. Improvement in the following measures has been noted:

  • Behavioral symptoms (for both high and low-risk residents)--Median low risk prevalence fell from 8.3 percent to 0 between 1998 and 2001; high-risk prevalence fell from 29.7 percent to 25.6 percent.

  • Prevalence of depression--Median fell from 12.9 percent to 9.1 percent. 95th percentile fell from 39.6 to 32.6 percent.

  • Prevalence of depression with no treatment--Median fell from 7.1 percent to 3.1 percent. 95th percentile fell from 29.7 to 16.7 percent.

  • Cognitive impairment--Median fell from 11.7 to 6.7 percent; 90th percentile fell from 27.9 to 21.4 percent.

  • Prevalence of occasional or frequent bladder or bowel incontinence without a toileting plan--Median fell from 26.7 to 21.3 percent; 95th percentile fell from 92 to 90 percent.

  • Fecal impaction--Median prevalence for the 90th percentile fell from 2.4 to 0.8 percent.

  • Dehydration--90th percentile fell from 2.2 to 1.6 percent; 95th percentile fell from 3.6 to 2.3 percent.

  • Prevalence of bedfast residents--Median fell from 3.1 to 2.3 percent; 90th percentile fell from 9 to 7 percent.

  • Decline in late loss ADLs for low-risk residents--Median fell from 11 to 10 percent; 90th percentile increased slightly, from 20 to 21 percent.

  • Decline in range of motion, overall and for both high and low-risk residents--The median fell from 5.7 to 5 percent; the 95th percentile fell from 26 to 21 percent.

  • Antipsychotic/hypnotic use--Median prevalence fell from 17.5 to 16.7 percent.

  • Hypnotic use more than two times in last week--The median prevalence fell from 2.1 to 1.9 percent; 95th percentile fell from 9.4 to 8.3 percent.

  • Prevalence of little or no activity--The median fell sharply, from 26.3 to 14.3 percent; the 90th percentile fell from 50 to 38 percent.

  • Pressure ulcers for both low and high risk residents--Overall median prevalence fell from 6.9 to 6.2 percent; for low-risk patients, the median fell from 2.6 to 0 percent; for high-risk residents, median prevalence fell from 14.3 to 11.5 percent.

(Note that, as a researcher grounded in solid evaluation skills, Dr. Rantz has not compared outcomes for QIPMO participants vs. non-participants, although the state is now asking her to do this. Such a comparison would confound programmatic effects vs. selection effects, due to the non-random selection of facilities into QIPMO.)

Several quality indicators have gotten worse in Missouri since the implementation of QIPMO, including behavior problems for high-risk residents, patients receiving nine or more medications, range of motion training/practice, and antipsychotics use in the absence of an appropriate diagnosis. Preliminary investigations suggest that these declines may reflect MDS coding issues rather than actual decline of care.

View full report


"statenh.pdf" (pdf, 822.35Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®