Table 4.

Evidence of validity for practice sensitive quality indicators

Quality indicator	Evidence of quality indicator validity
Antipsychotic use without psychosis	CMS: percentage of long-stay residents who received an antipsychotic medication.
Decline in mood	CMS: percentage of residents who have depressive symptoms.
	A (non-supportive) paper [20], reported that they found notable under-reporting; although, they agreed this QI was useful for reporting because of the clinical importance of the domain.
Declining behavioral symptoms	There is little yet reported to support the validity of this indicator, however it is clinically importance, and associated with resident safety.
Delirium	There is little yet reported to support the validity of this indicator, however it is clinically importance, and associated with resident safety.
Fallen last 30 days	HQO: percentage of residents who had a recent fall.
	Some data [21] suggests RAI-MDS data on falls over longer intervals (e.g. falls in last 180 days) may be more accurate and also cautions that falls tend to be underreported in the MDS data compared to in the chart [22].
Feeding tube	There is little yet reported to support the validity of this indicator, however it is clinically importance, and associated with resident safety.
Indwelling catheter	CMS: residents who have/had a catheter inserted and left in their bladder.
	Found to have the highest level of validity and highly recommended for use by CMS and nursing homes [22].
Late loss ADL decline	CMS: percentage of long-stay residents whose need for help with daily activities has increased.
	HQO: percentage of residents with increasing difficulty carrying out normal everyday tasks.
Physical restraint use	HQO: percentage of residents who were physically restrained.
	CMS: percent of residents who were physically restrained.
Pressure ulcer	HQO: percentage of residents who had worsening pressure ulcer status.
	CMS: pressure ulcer prevalence.
Unexplained weight loss	CMS: percentage of long-stay residents who lose too much weight. One study [23] concluded that the RAI-MDS weight loss QI is able to discriminate differences in prevalence of weight loss between facilities, suggesting concurrent validity of the QI.
Urinary tract infections	CMS: percentage of long-stay residents with a urinary tract infection.
	Found to have the highest level of validity and highly recommended for use by CMS and nursing homes [22].
	One study [24] comparing the RAI-MDS data for urinary tract infection (UTI), with data arising from active prospective surveillance in LTC facilities (n = 16) concluded that the RAI-MDS overestimated the number of cases. However, suggestions to use more explicit definition to reduce false positives have been instituted in 2008.
Worsening pain	HQO: percentage of residents with pain that recently got worse.
	The RAI-MDS pain QI* has been found to accurately differentiate the prevalence of pain between facilities however it has been suggested that high pain prevalence scores were associated with more frequent pain assessment and appropriate pain-related care practices, as opposed to poor care quality [25].

ADL, activities of daily living; CMS, US Centre for Medicare and Medicaid Services; HQO, Health Quality Ontario; QI, quality indicator.

^*The new CMS QI for pain is based on the self-report item of the newer MDS 3.0, and not the MDS 2.0.