quality indicators in nursing homes (MDS-based)

Introduction

Facility quality indicator (QI) profile

• use of QIs started 07/01/99
• developed for
  • regulatory purposes (original purpose)
  • facility quality improvement programs
  • consumer & payer information
• 24 QIs derived from 2 page minimum dataset (MDS) 2.0 quarterly form
• each QI represents a potential problem with a care process
  • not a direct measure of quality
  • only after the care process is evaluated can a determination of quality be made
• the MDS assessments used are annuals, quarterly updates & significant change assessments
  • admission assessment are not used as they reflect care received prior to nursing home administration
  • Medicare day 14 assessments are used (with 5th day assessment as the reference)
• federal & state site visits are mandated every 15 months but in practice occur ~every 12 months
  • the site visits are use to confirm the reporting data
  • survey findings are used to determine how many stars are given to a facility on the publically-reported quality rating scale
  • deficiencies my prompt corrective action, financial penalties, limits on admission, &/or facility closure
• facility QI data is available only to facility & surveyors
  • some QI-related information is available to the public on the internet; comparing facility data with state & national data

Quality Indicators:

• some QIs are prevalence QIs, with data coming only from the most recent assessment
• some QIs are incidence QIs, which compare the most recent assessment with the one before it looking for the presence of a condition not present before
• 4 QIs are risk-adjusted to account for differences in residents across facilities
  • these produce 3 separate measures
    • occurrence without regard to risk
    • occurrence in those with risk factors (high risk)
    • occurrence in those without risk factors (low risk)
• 4 QIs are sentinal health events - all require investigation
• many QIs are clinically linked to each other
• some QIs depend on MDS input (11, 12, 15 {see below})
  • accuracy depends upon MDS assessment & documentation
• to obtain definitions for the QIs, see:

Quality Indicators (CHSRA)

1) new fractures          (I)
2) falls                  (P)
3) behavioral symptoms    (P, RA)
4) depression symptoms    (P)
5) No antidepressant      (P)
6) 9+ medications         (P)
7) Cognitive impairment   (I)
8) B/B incontinence       (P, RA)
9) No toilet plan         (P)
10) Indwelling catheters   (P)
11) fecal impactation      (P, S)
12) UTIs                   (P)
13) weight loss            (P)
14) tube feeding           (P)
15) dehydration            (P, S)
16) Bedfast                (P)
17) ADL decline            (I)
18) ROM decline            (I)
19) antipsychotics         (P, RA)
20) anxiolytics/hypnotics  (P)
21) hypnotics > 2X/week    (P)
22) physical restraints    (P)
23) minimal activity       (P)
24) pressure ulcers        (P, RA, S)

* P: prevalence

I: incidence

RA: risk-adjusted

S: sentinel

Calculation of QIs

• for each QI, the % score is the proportion of residents in a facility who have the problem compared with the corresponding proportions in every other facility in the state

Problems* with QIs

• accuracy & reliability depend upon accuracy of MDS assessments
• selection bias - depends on resident characteristics
• ascertainment bias - some facilities better at identifying some conditions
• validity issues with some QIs - unclear link to outcomes & whether care processes can affect, or whether QIs can differentiate facilities with poor care from those with good care
• possibility of gaming the system with some QIs
• risk adjustment inadequately done
• small sample sizes or other factors make some QIs unstable

* QI re-evaluation report due summer 2003 new QIs to be incorporated into MDS 3.0 (2004)

Use of QIs for survery preparation

• surveyors use this data to focus surveys: expect review of
  • any flagged QIs
  • QIs in which the facility is 75% or higher
  • sentinel events
• 50% of surveyor's samples come from this list
• higher chance of review of residents with multiple QIs checked
• CMS will continue to use current QIs until new system is in place

Use of QIs for quality improvement

• despite the problems with QIs, the are based on the most comprehensive & current data set of clinical information in nursing homes (MDS) & are an excellent place to look for quality of care problems at the facility level
• QIs should supplement, not replace other monitoring systems present in the nursing home (generally some overlap)
• retrospective approach
  • not to improve quality for individual residents
  • detect patterns of care which can be improved for groups of residents
• one way to do this is to track the individual QIs on a monthly basis & graph both the facility % (absolute risk) & % vs state group (relative risk) & review monthly at QA meetings
  • QIs investigated
    • QI % is consistently higher than the state levels or facility goals
    • unexplained variation
    • a trend is detected
• time frames to trend
  • monthly
  • quarterly
  • semiannual
  • annual
    • gives more accurate picture of true rate when smaller time frames have small numbers of events
    • even with monthly time frames, MDS scores represent MDS done that month which may reflect a longer time period

After selecting a QI for further investigation

• select sample of residents (form quarterly or semiannual report) from resident level QI summary
  • residents from every unit
  • residents with similar patterns of flagged QIs
  • residents with few flagged QIs
  • residents with many flagged QIs
• sample should be varied & contain at least 5 residents, with more for high prevalence conditions
• check QI definition - consider data abstraction form
• check accuracy of information (including data input errors) against other sources
• was the condition assessed appropriately?
• review care process & policy & procedure
• was the condition care-planned?
• were the care processes & care plan implemented properly
• was effectiveness of the care plan evaluated & changes made as needed?
• improve care processes using standard CQI techniques

Sentinel events:

• pay careful attention to definitions & how nursing is doing the assessment (eg. impaired transfer or bed mobility, MDS item Ga or b, which may differentiate high from low risk residents with pressure ulcers
• update diagnosis list with every assessment

MDS-based quality indicators:

• rely on self-reported MDS data
• only some are risk adjusted
• need to be more individualized due to heterogeniety of nursing home residents

Quality measures reported by CMS: see Nursing home compare

More general terms

• quality indicator (QI); quality measure

Additional terms

• Center for Health Systems Research & Analysis (CHSRA)
• continuous quality improvement (CQI) or total quality management (TQM)
• minimum data set (MDS) for nursing home residents
• Nursing home compare
• qualtity of care in nursing homes