Table 4.

Studies of RAI-related structural variables [4,5,21,33,34,35,36,37,38].

Author	Setting	Sample size	Data sources	Design	Measures	Main findings
Hawes, 1997 [33]	SNFs	254 SNFs	2 resident cohorts (>2000); 10 states	Quasi-experimental probability-based sample	Completeness, accuracy-care plans; medical records	Increased medical record accuracy; Completeness of care plan
Bernabei, 1997 [34]	RAI training sites	9 countries	Staff participating in training sessions	Descriptive	RAI training sessions: purpose, length, content	Greatest variation in training between US and other countries
Hansebo, 1998 [35]	Sweden	3 elder care facilities; 50 nursing staff	Nursing staff trained in RAI	Cross-sectional survey	Staff views of RAI/MDS and care quality	Most staff reported positive improvement in care quality with RAI/MDS
Ettinger, 2000 [36]	428 Iowa SNFs	236 directors of nursing (DON)	DON surveys	Cross-sectional survey	DON perceptions of utility of dental section	76% viewed MDS section as useful; 9% used to identify dental needs
Jogerst, 2001 [37]	Geriatric MD practices	472 MDs	Internist and family MDs	Cross-sectional survey	% time reviewing MDS and CCP; how used; attitudes about MDS	11% reviewed all MDS and 21% partially. 19% did not review CCP; 56% had negative or derogatory attitudes
Parmelee, 2009 [38]	VA NH care units	289 NHs; 259 VA NH staff	34 DONs, 96 MDS RNs, 97 nurse managers; 19 medical directors; others 13	Mixed methods:	Accuracy. Usefulness, utility for quality improvement	78.4% rated as very accurate or accurate; 85.7% rated MDS as useful; 85.7% rated QIs as very or somewhat useful.
Abt report to CMS 2015 [21]	SNFs	Pilot survey SNFs	RAI/MDS documents	Retrospective descriptive	Evidence of adherence to MDS 3.0 reporting requirements, RN role; accuracy	99% compliance with mandated RN participation; 2.2% of MDS noncompliant with required timelines; MDS assessment/medical record discrepancies ranged 0.8% to 25.5%.

Shaded rows indicate studies using mixed methods.