Table 3

Evidence of effectiveness of quality management and quality system activities in nursing homes

Source	Setting and sample	Design	QA activity	Data collection and measurement instrument	Dependent variable	Type of variable	Key results
AHCPR = Agency for Health Care Policy and Research; OBRA = Omnibus Budget Reconciliation Act; CARE = Continuous Assessment Review and Evaluation scheme.
Quality management
Dimant, 1991 (USA)	Nursing home;189 beds	Longitudinal	Quality system: practice guidelines design processes indicator monitoring participation residents team approach employee involvement	questionnaires registration forms	medication errors pressure ulcers physical restraints incidence accidents satisfaction resident	Outcome	less than 0.1% medication errors and pressure ulcers physical restraints reduced by 50% improved mobility independence decreased incidence of accidents low prescriptions of psychoactive medication improved resident satisfaction
1 Process control based on standards
Levine et al, 1994 (USA)	Nursing facility; 816 beds	Post-test only	pressure ulcer prevention protocol indicator specific audit instrument for monitoring	documentation of all aspects of ulcer care data collection feedback weekly and monthly	Prevalence of nosomial ulcer	Outcome	low nosomial ulcer prevalence rate of 3.1% compared to 23% in USA nursing homes
Suntken et al, 1996 (USA)	Two extended care facilities (n=160 residents); 2 acute care facilities (n=97); 2 home health agencies (n=91)	Pre-test/post-test	AHCPR pressure ulcer prevention and treatment guideline	Trained registered nurses and team members assessed clients	Prevalence of pressure ulcer	Outcome	pre-test: 13 residents (21%) had 16 pressure ulcers post-test: 1 resident (2%) had 5 pressure ulcers
Moseley, 1996 (USA)	Two cohorts with 3149 and 5078 residents in nursing homes	Pre-test/post-test	Federal regulations: the OBRA catheterisation standard	data based on automated assessment database audit assessment data	Difference in percentages of pre-OBRA and post-OBRA residents catheterised	Process	1.3% of post-OBRA residents were catheterised long term v 8–15% pre-OBRA 1% of long term catheterisations were inappropriately catheterised at assessment
Semla et al, 1994 (USA)	485 bed nursing home	Post-test only	Federal regulations: the use of psychotropic drugs	12 month chart review of 107 residents	prescription of neuroleptic drugs dose of neuroleptic drugs	Process	in 75% of residents studied an attempt was made to lower or stop the dose of neuroleptic treatment: neuroleptics were discontinued in 45% with dementia and in 25% with psychiatric diagnosis
Shorr et al, 1994 (USA)	9432 residents in nursing homes	Longitudinal	Federal regulations: the use of psychotropic drugs	Assessment data of residents during 30 months	Neuroleptic drug use in days/100 days of residence	Process	neuroleptic drug use decreased from 23.9 to 17.5 days/100 days of residence (27%) one quarter of facilities had no change or increased usage, another quarter had decreases of 46% or more variations were affected by staffing patterns
2 Process improvement by quality improvement procedures
Fitzgerald et al, 1996 (USA)	9 nursing facilities	Post-test only	measurement system of quality assurance feedback benchmarking information	Quality indicators based on minimum data set resident assessments	15 patient-related measures, e.g. prevalence of falls, weight loss, tube feeding	Process/ outcome	6 scores below state average 3 scores above average 6 scores did not change
Phillips et al, 1997 (USA)	254 nursing homes; 2000 nursing home residents	Pre-test/ post-test	Resident Assessment Instrument (RAI) and care planning	Assessment of all residents at baseline and 6 months later by research nurses using the minimum data set (MDS)	functional status cognitive status psychosocial wellbeing	Outcome	Implementation of the RAI reduced the rate of decline in 7 of the 9 outcomes
Hawes et al, 1997 (USA)	254 nursing homes; >2000 residents	Pre-test/ post-test	RAI (standardised system to assist in assessment and care planning)	assessment of residents interviews with direct staff interviews with and observation of residents review of medical records	accuracy of information comprehensiveness of care plans use of physical restraints psychotropic drug use	Process	increased accuracy of information in medical records significant decline in physical restraints (37.4% to 28.1%) and indwelling urinary catheters (9.8% to 7%) use of antipsychotic drugs was unaffected
Fries et al, 1997 (USA)	254 nursing homes; 2088 nursing home residents	Pre-test/ post-test	Implementation of the RAI	Assessment of residents at baseline and 6 months later using the MDS	Presence/absence of: falls; pain decubitus vision problems stasis ulcer dental problems malnutrition dehydration	Outcome	lower prevalence of dehydration and stasis ulcer fewer residents declined in nutrition state and vision increase in prevalence of daily pain
Mor et al, 1997 (USA)	268 nursing facilities; 4196 residents	Pre-test/ post-test	Implementation of the RAI	research nurse reviewed records, interviewed staff and observed residents resident assessment data	Transitions during interval of 6 months: hospital admission, nursing home transfer, discharge to home	Process	Decline in hospitalisation: 15.9% in 1990 and 10.9% in 1993
Chambers et al, 1996 (USA)	16 nursing homes; 138 residents	Pre-test/post-test	nurse audit facilitator audit projects	satisfaction questionnaire geriatric depressions score	satisfaction resident depression score	Outcome	satisfaction levels improved in 6 out of 11 aspects mean geriatric depression score fell from 4.73 to 4.25
Challiner, 1997 (UK)	Nursing home; 28 residents	Pre-test/ post-test	CARE scheme: clinical audit package examining 9 aspects of geriatric nursing care	Audit form filled in for each resident and audit meetings to evaluate the audit forms	Policy with regard to urinary/faecal continence drug use falls/accidents pressure sores environment and equipment aids and adaptations medical role preserving autonomy	Process	increased documentation of residents' goals policy for incontinence, increased monitoring policy for and less evidence of clinical problems more thorough assessment of individual falls pressure sores monitored daily instead of weekly no change in equipment staff trained in use of hoist improved recording in care plans
Dickinson et al, 1997 (UK)	15 long term care facilities; 337 and 258 residents	Pre-test/post-test	CARE scheme: auditing the care of residents, discussing results, setting objectives	Audit form filled in for all residents and again 8 months later	Policy with regard to: urinary/faecal continence drug use falls/accidents pressure sores environment and equipment medical role preserving autonomy	Process	42% of facilities developed policies for care (32% already had policies) standard of policies became better, fewer missing components (reduced from 38% to 19%) no change in policies of optimising drug use 72% had reached the care standard for residents (increase of 4% in drug use and environment) no change in policies for urinary continence
Mohide et al, 1988 (Canada)	60 nursing homes; 1525 residents	Controlled trial	service of a quality assurance consultant working through a quality assurance cycle for two indicator conditions	record review assessment forms	Change score of: urinary incontinence constipation potential skin breakdown hazardous mobility	Outcome	improvement in the conditions of hazardous mobility and constipation was greater in exp. group no differences for the hidden conditions: potential skin breakdown and urinary incontinence
Gustafson, 1992 (USA)	60 nursing homes	Controlled trial	Two different facility and resident assessment processes	Quality of care assessment by a research team using interviews and observations	overall quality of care detection of problems recidivism of problems unnotified problems cost of methods	Process/outcome	assessment method helps better to detect problems and at the same time to reduce the costs of the survey process many process improvements will not occur without external review
3 Human resources management
Richmond et al, 1996 (USA)	Long term care psychiatric facility; 773 incidents of disruptive behaviour	Pre-test/post-test	Training on prevention and management of disturbed behaviour (critical moment) and monthly feedback of results	Self-developed criteria to determine whether patients were appropriately assessed and managed	use of seclusion use of restraints	Process	one-to-one verbal intervention followed by medication mainly used alternative seclusion hours decreased by 31% restraint hours decreased by 47%
Avorn et al, 1992 (USA)	12 nursing homes	Controlled trial	Educational programme to reduce potentially excessive use of psychoactive drugs	evaluation drug use follow up evaluation	Index of psychoactive drug use	Process	Decline of 27% in experimental group v 8% in controls
Ray et al, 1993 (USA)	4 nursing homes	Controlled trial	Educational programme for physicians, nurses, staff	evaluation drug use	Total number of days of drug use	Process	Drug use decreased by 72% in experimental group v 13% in controls
Schnelle et al, 1993 (USA)	7 nursing homes; 340 residents	Pre-test/post-test	Staff training toileting programme	random hour check of residents for incontinence toileting chart	Prevalence of incontinence in residents	Outcome	Significant reduction in incontinence from 43% to 21%
4 Involvement of residents
Cherry, 1993 (USA)	210 nursing homes	Post-test only	Nursing home ombudsman who can visit residents and resolve disputes on their behalf	Two sources of secondary data: (1) collected during inspection, e.g. staffing patterns, resident status; (2) annual survey of ombudsman programmes	bedsores catheterisations two infection measures sum of compliance with eight standards	Process/outcome	significant support for the hypothesis that the existence of ombudsman programmes can have an impact on patient related outcomes no relationship was found between presence of ombudsman and compliance with standards
5 Quality assurance documents
No studies were found