Adherence to Pressure Ulcer Prevention Guidelines:

Implications for Nursing Home Quality

Debra Saliba, MD, MPH,*‡§ Lisa V. Rubenstein, MD, MSPH,* †‡ Barbara Simon, MA,*
Elaine Hickey, RN, MS ¶ Bruce Ferrell, MD,* Elaine Czarnowski, and Dan Berlowitz, MD, MPH ¶

OBJECTIVES: This study aims to assess overall nursing
home (NH) implementation of pressure ulcer (PU) preven-
tion guidelines and variation in implementation rates among
a geographically diverse sample of NHs.
DESIGN: Review of NH medical records.
SETTING: A geographically diverse sample of 35 Veter-
ans Health Administration NHs.
PARTICIPANTS: A nested random sample of 834 residents
free of PU on admission.
MEASUREMENTS: Adherence to explicit quality review
criteria based on the Agency for Healthcare Research and
Quality Practice Guidelines for PU prevention was mea-
sured. Medical record review was used to determine over-
all and facility-specific adherence rates for 15 PU guideline
recommendations and for a subset of six key recommen-
dations judged as most critical.
RESULTS: Six thousand two hundred eighty-three instances
were identified in which one of the 15 guideline recom-
mendations was applicable to a study patient based on a
specific indication or resident characteristic in the medical
record. NH clinicians adhered to the appropriate recom-
mendation in 41% of these instances. For the six key rec-
ommendations, clinicians adhered in 50% of instances.
NHs varied significantly in adherence to indicated guide-
From the *Center for the Study of Healthcare Provider Behavior and
†Department of Medicine, Veterans Affairs Medical Center, Greater Los
Angeles System, Los Angeles, California; ‡Geriatric Research, Education,
and Clinical Center, Veterans Affairs Medical Center, Sepulveda, California;
§Department of Medicine, University of California at Los Angeles, Los
Angeles, California; Center for Health Quality, Outcomes and Economic
Research, Bedford Veterans Affairs Medical Center, Bedford, Massachusetts;
and ¶Boston University School of Public Health, Boston, Massachusetts.
This research is based on a VA HSR & D funded study #CPG97-012 and
was presented at the Department of Veterans Affairs 19th Annual Health
Services Research and Development Services Meeting, February 2001,
Washington, DC.
Dr. Saliba is a Research Associate of the Veterans Affairs Health Services
Research and Development Services and a 2000 Pfizer/American Geriatrics
Society Foundation for Health in Aging Junior Faculty Scholar for Health
Outcomes in Geriatrics.
Address correspondence to Debra Saliba, MD, MPH, RAND Corporation,
1700 Main Street, Santa Monica, CA 90401. E-mail: saliba@rand.org
line recommendations, ranging from 29% to 51% overall
adherence across all 15 recommendations (P  .001) and
from 24% to 75% across the six key recommendations
(P  .001). Adherence rates for specific indications also
varied, ranging from 94% (skin inspection) to 1% (educa-
tion of residents or families). Standardized assessment of
PU risk was identified as one of the most important and
measurable recommendations. Clinicians performed this as-
sessment in only 61% of patients for whom it was indicated.
CONCLUSIONS: NHs’ overall adherence to PU prevention
guidelines is relatively low and is characterized by large
variations between homes in adherence to many recom-
mendations. The low level of adherence and high level of
variation to many best-care practices for PU prevention in-
dicate a continued need for quality improvement, particu-
larly for some guidelines. J Am Geriatr Soc 51:56–62,
2003.
Key words: nursing home; quality improvement; pressure
ulcers; guidelines
P ressure ulcer (PU) prevention is a key element of nurs-
ing home care quality with important implications for
patients and healthcare systems. For patients, the develop-
ment of a PU is a major source of morbidity that is associ-
ated with an increased risk of death.1–3 Systems or facilities
must employ additional resources (staff time and treat-
ments) if a patient develops a PU.4,5 In addition, families
and oversight agencies view the development of a new PU
as a sentinel event or indicator of potentially poor nursing
home (NH) quality.6,7
The importance of this condition to patients and pro-
viders national8,9 and organizational10 monitoring efforts
indicated. Although reported prevalence and incidence
rates vary, on average, 12% of NH residents have an ex-
isting PU,11 and about 6% of NH residents will develop a
new PU over a 6-month period.12 Efforts to provide NHs
with facility-level and national PU prevalence data repre-
sent one approach for stimulating quality improvement
(QI) activities.

JAGS 51:56–62, 2003

© 2003 by the American Geriatrics Society 0002-8614/03/$15.00
JAGS JANUARY 2003–VOL. 51, NO. 1
Best-practice guidelines represent another potential
stimulus and provide an important tool for improving out-
comes of care.13 Recognizing the prevalence of PUs and the
need for prevention tools, the Agency for Healthcare Re-
search and Quality (AHRQ (formerly AHCPR)) assem-
bled a national interdisciplinary panel to develop a set of
best-practice guidelines for PU prevention. The expert
panel developed clinical practice guidelines based, in part,
on their evaluation of more than 800 scientific manu-
scripts related to PU prevention. The AHRQ Clinical Prac-
tice Guidelines for Pressure Ulcer Prevention incorporated
this literature and scientific expertise to identify appropri-
ate and effective care for preventing PU development. 14,15
These guidelines have been disseminated on a national
level and continue to be recommended as the standard of
care for PU prevention.5,16,17 A systematic reassessment
performed in 2000 judged the AHRQ PU prevention
guidelines to be still valid.18 In addition, QI efforts to im-
plement the guidelines have been associated with de-
creased PU incidence.19,20
Given the organizational and national emphasis on
PU prevention, continuous feedback about PU prevalence,
and the availability of prevention guidelines, most NHs
may be systematically implementing best practices for PU
prevention. The decline in NH-reported PU rates seen
during the 1990s10,21 points to this possibility. It may be
that the current rates represent the best that can be achieved
with existing technology and case mix and that no further
QI efforts are needed. To ascertain whether this is the
case, it is important to know whether NHs are imple-
menting the guidelines for best-care practices and whether
NHs vary in performance. The current study therefore
aimed to assess overall NH implementation of clinical
practice guidelines for PU prevention and variation in im-
plementation rates among a geographically diverse sam-
ple of NHs.
NHs were viewed within a single healthcare delivery
system and payment structure: the Veterans Health Ad-
ministration (VHA). The VA has been monitoring and
providing its NHs with specific PU incidence data since
1991.10 This allowed for the examination of guideline ad-
herence for NHs with similar reimbursement, ownership,
and surveillance but with diverse and varied staffing, size,
location, and resident characteristics.
METHODS
AHRQ Guidelines
The guidelines for PU prevention contain best-practice
recommendations organized around four goals: (1) to
identify persons at risk and the specific risk factors that
place them at risk; (2) to preserve and enhance the ability
of tissue to tolerate pressure; (3) to protect against pres-
sure, friction, and shear; and (4) to employ education to
reduce PU occurence.14 The guidelines and supporting
documents did not include explicit tools for abstracting
relevant information about guideline adherence from the
medical record.
In a project conducted for AHRQ, Lauori et al. used
formal methods and testing to translate the AHRQ guide-
lines into explicit quality review criteria.22 These quality
review criteria were written as “IF . . . THEN . . .” state-
ADHERENCE TO PRESSURE ULCER PREVENTION GUIDELINES 57
ments. Each IF statement identified the group of residents
to whom the specific review criterion applied. Each THEN
statement specified the recommended intervention for that
group. These review criteria were adopted to assess guide-
line use in the current study’s population.
Chart Abstraction Tool
A form that was developed and validated for assessing
hospital compliance with the PU prevention guideline re-
view criteria was modified.22 The modification accounted
for differences between hospitals and NHs in average length
of stay, service mix, record organization, and levels of doc-
umentation. The resulting NH chart abstraction tool and
accompanying instructions recognize the interdisciplinary
nature of PU care23,24 and use multiple data sources (nurs-
ing, physician, consultant, nurse’s aide, dietitian and ther-
apy notes, and laboratory reports) for the same informa-
tion. The abstraction tool reviews the first 45 days of NH
admission. To enhance comparability across reviewers and
across facilities, the NH abstraction tool provides specific
parameters and time intervals for identifying resident char-
acteristics, risk factors, and indicated interventions out-
lined in the guidelines.
Pilot testing revealed that some recommendations
were unlikely to be obtained reliably from the planned
NH chart review. In most instances, this lack of reliability
resulted from the time frame for review or from a lack of
specificity in typical chart documentation for the IF or the
THEN statement. If this lack of specificity might lead to
misclassifying care as failing to meet an indication, that
guideline was excluded from review (Appendix 1).
Key Guideline Recommendations
In the current study, study clinicians identified key guide-
line recommendations by independently ranking the qual-
ity review criteria on importance and feasibility of mea-
surement with medical records. This additional ranking
was performed because all review criteria might not be
equally important or equally feasible to measure through
chart review.
Sample Selection
One hundred thirty VHA NHs potentially were available
to participate. Forty were chosen to represent a range of
sizes, geographic regions, urban versus rural location, and
reported PU rates. Nested sampling was performed to
identify residents in the selected NHs. Resident inclusion
criteria, from the VHA Patient Assessment File (PAF),
were new NH admission between January and December
1998, absence of PU on admission, and NH stay of 2 or
more weeks. From residents meeting these inclusion crite-
ria, a sample was selected to represent a range of PU risk
within each NH. Resident risk was estimated based on a
modified Norton Scale score constructed from the
PAF.14,25
Chart Abstractions
The medical administrative services at each study site pro-
vided NH records for abstraction. Study nurses used the
NH chart abstraction tool to abstract information from
each resident’s record. Ten percent of records were re-
abstracted to determine interrater reliability.
58 SALIBA ET AL. JANUARY 2003–VOL. 51, NO. 1 JAGS

Analyses
Table 1. Sample Characteristics
The adherence proportion was calculated as the number of
times the intervention was performed (the THEN statement) Characteristic Value
divided by the number of times the intervention was indi-
cated (the IF statement). Overall and facility-specific rates Nursing homes N  35
of adherence were determined for the 15 PU-prevention Daily census, mean residents (range) 122 (48–263)
quality review criteria and for the subset of key quality re- Norton score* by facility, mean (range) 14.1 (13–16.2)
view criteria identified as having strongest evidence for PU Region
prevention and being most feasible to measure with medi- East n  10
Midwest n8
cal record review. For all of the quality review criteria,
South n  10
compliance was broadly defined. For example, the recom-
West n7
mendation for risk assessment was considered met if any
Rural n3
of the following were present: a formal scale or reported re- Residents N  834
sult, any mention of a risk assessment, or any mention that Age, mean  standard deviation
the resident was or was not at risk for PU. Similarly, a rec- (range) 72  10.9 (25–105)
ommendation was scored as implemented if the record Admission source, %
documented that it was planned, ordered, or performed. A Acute care 49
process was also scored as adhered to if a provider docu- Other hospital based 15
mented any refusal or resident/family preferences that would Home 30
preclude an intervention. Not indicated 6
Variation in NH performance was assessed with a Male, % 96
generalized linear model, testing the null hypothesis of no Ethnicity, %
difference in guideline adherence associated with admis- White 53.9
sion to a particular NH. The correlation between facility Black 15.6
adherence to the 15 guideline recommendations and ad- No data 28.0
herence to the key recommendations was assessed. Norton score, number 14 (high risk)
Generalized estimation equations were employed that (range)14,25 509 (5–20)
considered adherence rates nested within facility and rates *Norton score ranges from 5–20. Lower scores indicate higher risk for PU devel-
nested by resident to determine whether selected facility opment. A generally accepted cut-off for increased risk is 14 or less.
characteristics; i.e., size, region, or type (urban teaching,
urban nonteaching, rural) were associated with differences
in documented adherence rates. fied in which a recommendation applied to a study resi-
dent based on a specific indication in the medical record
(mean number of indications per resident  7.5). For the
RESULTS
six key recommendations, 3,085 indications were found in
Sample the reviewed records.
Of the 40 NHs approached to participate, 35 are included Guideline Adherence
in the current analyses. Five NHs were not included be-
cause a local principal investigator could not be identified NH clinicians adhered to the appropriate guideline recom-
mendation in 41% (2,604/6,283) of the instances where
(n  2), too few admissions met the inclusion criteria (n 
the 15 quality review criteria applied. Overall adherence
1), or institutional review and chart transfer were delayed
for the six recommendations judged as most critical was
(n  2). Ten percent of charts obtained for review were
1,542/3,085, or 50%. Table 2 shows the adherence pro-
excluded because chart review revealed a documented PU
portion for each review criteria. For individual guideline
at the time of admission. The NH records of 834 residents
recommendations, adherence rates ranged from 94% (skin
from 35 NHs underwent complete review (mean number
inspection) to 1% (education of residents or families).
of residents per site  24, range 13–30). Table 1 shows se-
NHs differed significantly in their implementation of
lected characteristics of the NHs and residents. Almost
the guideline. Facility adherence rates ranged from 29% to
half of the residents were admitted from an acute care hos-
51% for the 15 guideline recommendations (P  .001).
pital. The NH admissions had a wide range of Norton-
For the six key recommendations, the variation was even
based risk scores.
greater, ranging from 24% to 75% (P  .001). Facilities
that performed well on the six recommendations also
Guideline Review Criteria tended to do well on all 15; correlation between NH scores
Six guideline recommendations were identified as key mea- on the 15 recommendations and the six key guideline rec-
sures of NH guideline implementation. These are indicated in ommendations was 0.87. Column 3 of Table 2 shows the
Table 2, column 1. The American Medical Director’s Associ- range of facility adherence rates for each guideline recom-
ation also highlighted all six in their efforts to abbreviate and mendation (adherence rate for the NH with the lowest
adapt the AHRQ guidelines for use by NH care providers.26 performance and the rate for the NH with the highest per-
Interrater agreement for abstraction was 95%. A large formance on each recommendation) and the results of signif-
number of indications or items meeting the IF element of icance tests for whether being a resident in any given NH
the review criteria were documented in the NH charts. For was associated with the rate of adherence to a particular
the 15 quality review criteria, 6,283 instances were identi- recommendation. This tests for the ability to reject the null
JAGS JANUARY 2003–VOL. 51, NO. 1 ADHERENCE TO PRESSURE ULCER PREVENTION GUIDELINES 59

Table 2. Quality Review and Utilization Criteria Pressure Ulcer Prevention*

Overall Range of Facility

Adherence Adherence
(n compliant/ %
Criterion n indicated) (P-value)*†

1‡ IF a person is admitted to a NH, 61% 0–100

THEN an assessment of pressure ulcer (PU) risk should be performed. (507/834) (.0001)
2‡ IF a person remains in a NH, 17% 0–76
THEN reassessments of PU risk should be performed at periodic intervals. (73/424) (.0001)
(Current study measures whether risk is reassessed between Week 2 and Week 6).
3‡ IF a NH resident is at risk for PU, 94% 75–100
THEN a systematic skin inspection that pays attention to bony prominences should be (668/712) (.110)
performed at least once a day.
(Current study measures whether skin inspection is performed at least once).
4 IF a NH resident is incontinent of urine or stool, 10% 0–38
THEN skin cleansing should occur at the time of soiling. (56/559) (.0034)
(Current study measures whether skin cleansing is planned or performed).
5 IF a NH resident has dry skin, 40% 0–100
THEN the skin should be treated with moisturizers. (31/77) (.073)
6 IF a NH resident is incontinent and exposure to urine/stool is not controlled, 87% 60–100
THEN underpads, briefs, or topical agents that act as barriers to moisture should (486/559) (.302)
be used.
7 IF a NH resident is immobile or has altered level of consciousness, 32% 0–100
THEN exposure to friction/sheer forces should be minimized through proper (95/298) (.0001)
positioning, transferring and turning techniques and lubricants, protective films,
protective dressings, or protective padding.
(Study measures whether positioning, protective dressings, or education of resident
about repositioning is planned, ordered, or implemented.)
8 IF a NH resident has inadequate dietary intake, 86% 56–100
THEN approaches to improving intake should be considered or implemented. (301/349) (.117)
9‡ IF a NH resident is nutritionally compromised, 37% 0–67
THEN a plan for nutritional support or supplement should be implemented. (129/349) (.048)
10‡ IF a NH resident is confined to bed (immobility, remains in bed), 69% 0–100
THEN repositioning should be scheduled and performed every 2 hours. (37/54) (.019)
(Study measures whether repositioning is planned or implemented)
11 IF a NH resident is confined to bed, 11% 0–50
THEN positioning devices should be used to keep bony prominences from direct (6/54) (.798)
contact.
12 IF a NH resident is immobile and confined to bed, 17% 0–50
THEN devices that relieve pressure on heels should be used. (49/292) (.197)
13 IF a NH resident is immobile or has altered level of consciousness, 10% 0–50
THEN lifting devices or bed linens should be used to move the resident in bed. (29/298) (.097)
14‡ IF a NH resident is at risk for PU, 18% 0–65
THEN a pressure-reducing device (foam, static air, alternating air, gel, or water (128/712) (.0001)
mattress) should be used when lying in bed.
15 IF a NH resident is at risk for PU, 1% 0–8
THEN education addressing PU prevention should be provided to the resident and (9/712) (.860)
family caregiver.
*Range reports the highest- and lowest-scoring NH for the individual quality indicator.
†P-value gives significance test for the association between facility of residence and compliance for a each recommendation.

‡Key criteria.

NH  nursing home; PU  pressure ulcer.

hypothesis that NH of residence did not matter for predict-
ing the rate of adherence to each guideline recommenda-
tion. For five of the six key guideline review criteria, the
NH adherence proportion for each criterion varied signifi-
cantly among NHs. One key guideline review criterion
with high overall adherence rates, skin assessment at least
once, did not show a significant difference among NHs.
Of the facility characteristics considered in the nested
generalized estimation equations, Midwest location revealed
a borderline trend for being associated with increased over-
all adherence (P  .0549) to the 15 guideline recommenda-
tions. The other specific facility characteristics failed to
show statistical significance, but the number of facilities
being compared on any particular characteristic may
60 SALIBA ET AL.
have limited the ability of these analyses to detect a rela-
tionship between facility characteristics and documented
adherence.
DISCUSSION
Explicit NH review criteria revealed low overall rates of
documented adherence to best-care practices for PU pre-
vention. In this geographically diverse sample of NHs
within a large healthcare system, overall adherence to 15
national PU prevention recommendations and a subset of
six key recommendations was low, and variance in adher-
ence to quality standards among facilities was high. For
the six key care processes, overall adherence was 50%,
and no NH had adherence above 75%. Guideline adher-
ence varied across care processes. Although skin inspec-
tion achieved the highest rate of implementation, stan-
dardized risk assessment—a critical step for systematically
targeting subsequent assessments and interventions26–29—
was performed in only 61%. These results point to the
need for additional interventions to improve PU preven-
tion practices and suggest that current PU rates do not rep-
resent the best that can be achieved.
These rates rely on quality review criteria for which
the definition of adherence was set at a minimal level. This
level was selected to allow for the possibility that docu-
mentation may not detail all care processes that are per-
formed. In addition, this study did not rely on one data
source to indicate compliance. Instead, the record review
was designed to capture the same data across different data
sources to ensure that any documented care was captured.
Although comparable NH studies were not available,
a recently published study of PU prevention in community
hospitals measured five of the six recommended care pro-
cesses identified in this study as key (daily skin assessment,
use of PU-reducing surfaces, assessment of risk, scheduled
repositioning, and nutritional consult for persons at risk).30
For three recommendations (skin assessment, scheduled
repositioning, and nutrition consult), the reported rates of
adherence were similar to those that this study found in
NHs. This study found higher adherence rates than those
reported in hospitals for two recommended care processes:
use of pressure-reducing surfaces (NH  18%, hospital 
8%) and assessment of risk (NH  61%, hospital  23%).
These differences may stem from differing definitions of
the adherence proportion or may indicate real differences
between the two settings in clinicians’ orientation toward
performing initial assessment and using support surfaces.
The low overall adherence rates across settings and varia-
tion across care processes indicate that systematic adop-
tion of guidelines has not occurred.
It is not entirely surprising that guideline dissemina-
tion and external surveillance/feedback of PU rates have
not resulted in the systematic implementation of best
prevention practices in NHs. First, preventive care prac-
tices may tend to be underused in general. The NHs’
rates of adherence to best practices for PU prevention are
similar to rates seen in the community for other preven-
tive care practices.31 Second, given the interdisciplinary
nature of PU prevention,14,24,26 attention to multiple ele-
ments is required for organizations to systematically im-
port the guidelines into daily practice. In the area of
staffing resources, for example, experience, distribution,
JANUARY 2003–VOL. 51, NO. 1 JAGS
task division, and baseline knowledge affect implementa-
tion and will vary across facilities. Thus the approaches
for implementing the guidelines would need to be adapted
to each NH’s staffing resources. 23 Overall, increased at-
tention needs to be focused on implementing the guide-
lines. Although QI activities focusing on PU prevention
can improve implementation,19,20,32 such QI efforts can
be made more efficient by identifying specific areas of
weakness.24
It is notable that this study found significant varia-
tion in adherence to the guidelines even among NHs with
common ownership and reimbursement structures. What
does this interfacility variation mean? It has potential im-
plications for NH care and highlights an opportunity for
guiding improvement. For residents, it means that the
NH they enter makes a difference in the overall PU pre-
ventive care they receive. For health systems and facili-
ties, the patterns of compliance and variation that are
documented may provide a basis for informed quality im-
provement. Arguably, competing demands for limited re-
sources may impede guideline implementation. However,
the wide variation across NHs in documented compliance
indicates that some NHs may be performing better than
others. NHs that have implemented the guidelines might
serve as models for NHs that have not yet done so. Those
NHs that have lower rates of implementation could pro-
vide information on barriers to implementation or rea-
sons for omission.
Limitations
This study assesses quality through medical record review.
Not all resident characteristics or care processes are docu-
mented in the record. If the NH record does not identify
that a problem exists, then the resulting underdetection of
instances where a care process should be applied may lead
to overestimates of adherence. However, a large number
of documented indications or resident risk factors on
chart review were found. In addition, assessment itself is
identified as a care process for which facilities are held
accountable.
The medical record may also under- or overestimate
adherence because day-to-day care may not be documented
accurately. Care that is provided routinely many times per
day may be most susceptible to inaccurate documentation
in medical records.33 This possibility was considered while
identifying the guideline review criteria to be measured,
and adherence was measured to include any mention of a
plan or implemented care. Multiple data sources for evi-
dence of adherence were also searched. In other settings,
this approach to measuring care processes has shown cor-
relation between quality scores and outcomes.34–36
It was not possible to measure all of the guideline rec-
ommendations with medical record review, particularly
those specific to chair-bound individuals. It is possible that
some NHs may be doing a better job in implementing
those care processes not measured.
This study was conducted in VHA NHs. It remains
possible that care in nongovernment NHs would differ.
Community NHs began to collect the Resident Assessment
Instrument (RAI)8 in 1991, but it has only recently has
been imported into VA NHs. However, VA PU prevalence
rates of 7.7% for Stage 2 or worse ulcers in 19933 com-
JAGS JANUARY 2003–VOL. 51, NO. 1
pared favorably with 1993 post-RAI community rates of
8.5%.11 In addition, it is not clear whether the mandatory
collection of RAI data obviates the need for other formal
assessment and QI tools;37–39 one large evaluation of the
effect of the RAI on community NH care practices failed
to show an improvement in preventive skin care practices
associated with the implementation of the RAI.40
Finally, although the guideline was developed based
on expert consensus and the scientific literature, no one
has conducted rigorous randomized controlled trials that
test the outcomes associated with implementing the entire
guideline set in a variety of NHs.5,41 Future studies should
assess the effect of guideline compliance on preventing
pressure ulcer development. Until such work is performed,
these guidelines represent the summation of the current
understanding of best preventive practices and provide the
best available approach for improving care.5,17
CONCLUSIONS
The high level of variation and low level of adherence ob-
served in NHs for this important and common clinical
condition indicate a continued need for QI. The findings
also identify specific guideline areas in which general im-
provement across all NHs is needed and others in which
some NHs excel and could serve as models for improve-
ment. These models need to consider the interdisciplinary
nature of NH care and the need to systematically incorpo-
rate preventive care into assessments and daily care rou-
tines. In addition, future studies should examine care vari-
ations in PU preventive care among NHs with different
organizational and reimbursement structures.
REFERENCES
1. Allman RM. Pressure ulcers among the elderly. N Engl J Med 1989;320:
850–853.
2. Sugarman B. Pressure sores and underlying bone infection. Arch Intern Med
1997;147:553–555.
3. Berlowitz DR, Brandeis GH, Anderson J et al. Effect of pressure ulcers on the
survival of long-term care residents. J Gerontol A Biol Sci Med Sci 1997;52A:
M106–M110.
4. Xakellis GC, Frantz R. The cost of healing pressure ulcers across multiple
health care settings. Adv Wound Care 1996;9:18–22.
5. Rochon PA, Minaker KL. Agency for Health Care Policy and Research
guidelines for pressure ulcer prevention: Improving practice and a stimulus
for research. J Gerontol 1993;48:M3–M5.
6. Zimmerman DR, Karon SL, Arling G et al. Development and testing of nurs-
ing home quality indicators. Health Care Financ Rev 1995;16:107–128.
7. Medicare. Nursing Home Compare [On-Iine]. Available: www.Medicare.gov/
Nhcompare/Home.asp Accessed March 2001.
8. Morris J, Hawes C, Fries B et al. Designing the national resident assessment
instrument for nursing homes. Gerontologist 1990;30:293–307.
9. Morris JN, Nonemaker S, Murphy K et al. A commitment to change. Revi-
sion of HCFA’s RAI. J Am Geriatr Soc 1996;45:1011–1016.
10. Berlowitz DR, Halpern J. Evaluating and improving pressure ulcer care: The
VA experience using administrative data. JT Comm J Qual Improv 1997;
423:424.
11. Fries BE, Hawes C, Morris JN et al. Effect of the National Resident Assess-
ment Instrument on selected health conditions and problems. J Am Geriatr
Soc 1997;45:994–1001.
12. Brandeis GH, Berlowitz DR, Hossain M et al. Pressure ulcers: The Minimum
Data Set and the Resident Assessment Protocol. Adv Wound Care 1995;8:
18–25.
13. Katz PR, Ouslander JG. Clinical practice guidelines and position statements: The
American Geriatrics Society approach. J Am Geriatr Soc 1996;44:1123–1124.
14. Agency for Health Care Policy and Research Panel for the Prediction and
ADHERENCE TO PRESSURE ULCER PREVENTION GUIDELINES 61
Prevention of Pressure Ulcers in Adults. Pressure Ulcers in Adults Prediction
and Prevention. Clinical Practice Guideline Number 3 (AHRQ publication
no. 92–0047). Rockville, MD: Agency for Health Care Policy and Research,
Public Health Service, U.S. Department of Health and Human Services, 1992.
15. Bergman-Evans B, Cuddigan J, Bergstrom NJ. Clinical practice guidelines:
Prediction and prevention of pressure ulcers. Gerontol Nurs 1994;20:19–26.
16. Bennett RG, O’Sullivan J, DeVito EM et al. The increasing medical malprac-
tice risk related to pressure ulcers in the United States. J Am Geriatr Soc
2000;48:73–81.
17. American Geriatrics Society Clinical Practice Committee. Pressure ulcers in
adults: Prediction and prevention. J Am Geriatr Soc 1996;44:1118–1119.
18. Shekelle PG, Ortiz E, Rhodes S et al. Validity of the Agency for Healthcare
Research and Quality Clinical Practice Guidelines: How quickly do guide-
lines become outdated? JAMA 2001;286:1461–1467.
19. Xakellis GC. Quality assurance programs for pressure ulcers. Clin Geriatr
Med 1997;13:599–606.
20. Xakellis GC, Frantz RA, Lewis A et al. Cost-effectiveness of an intensive
pressure ulcer prevention protocol in long-term care. Adv Wound Care 1998;
11:22–29.
21. Berlowitz DR, Bezerra HQ, Brandeis GH et al. Are we improving the quality
of nursing home care: The case of pressure ulcers. J Am Geriatr Soc 2000;48:
59–62.
22. Laouri M, Mittman B. Developing Quality and Utilization Review Criteria
for Pressure Ulcers in Adults: Phase II Final Report (Rand Publication no.
PM-403-AHCPR). Santa Monica, CA: RAND, 1995.
23. Kartes SK. A team approach for risk assessment, prevention and treatment of
pressure ulcers in nursing home patients. J Nurs Care Qual 1996;10:34–45.
24. Levine JM, Simpson M, McDonald RJ. Pressure sores: A plan for primary
care prevention. Geriatrics 1989;44:75–76,83–87,90.
25. Exton-Smith AN, Norton D, McLaren R. Pressure sores:. A study of factors
concerned in the production of pressure sores and their prevention. In:
Norton D, McLaren R, Exton-Smith AN, eds. An Investigation of Geriatric
Nursing Problems in the Hospital. London: Churchill Livingstone, 1962. pp
194–236.
26. American Medical Directors Association. Pressure Ulcers Clinical Practice
Guideline. Columbia, MD: American Medical Directors Association, 1996.
27. Bergstrom N, Braden B. A prospective study of pressure sore risk among in-
stitutionalized elderly. J Am Geriatr Soc 1992;40:747–758.
28. Bergstrom N, Braden B, Kemp M et al. Multi-site study of incidence of pres-
sure ulcers and the relationship between risk level, demographic characteris-
tics, diagnoses, and prescription of preventive interventions. J Am Geriatr
Soc 1996;44:22–30.
29. Braden BJ, Bergstrom N. Predictive validity of the Braden scale for pressure
ulcer risk in a nursing home population. Res Nurs Health 1994;17:459–
470.
30. Lyder CH, Preston J, Grady JN et al. Quality of care for hospitalized Medicare
patients at risk for pressure ulcers. Arch Intern Med 2001;161:1549–1554.
31. Schuster MA, McGlynn EA, Brook RH. How good is the quality of health
care in the United States? Milbank Q 1998;76:517–563.
32. Moody BL, Fanale JE, Thompson M et al. Impact of staff education on pres-
sure sore development in elderly hospitalized patients. Arch Intern Med
1988;148:2241–2243.
33. Saliba D, Schnelle J. Indicators of the quality of nursing homes residential
care. J Am Geriatr Soc 2002;50:1421–1430.
34. Asch SM, Kerr EA, Lapuerta P et al. A new approach for measuring quality
of care for women with hypertension. Arch Intern Med 2001;161:1329–
1335.
35. Kahn KL, Rogers WH, Rubenstein LV et al. Measuring quality of care with
explicit process criteria before and after implementation of the DRG-based
prospective payment system. JAMA 1990;264:1969–1973.
36. Kravitz RL, Laouri M, Kahan JP et al. Validity of criteria used for detecting
underuse of coronary revascularization. JAMA 1995;274:632–638.
37. Kramer J. Merging regulations and recommendations: How can the best out-
comes be achieved? J Gerontol Nurs 1998;24:5–6.
38. National Pressure Ulcer Advisory Panel. Position Statement. Minimum Data
Set-2 (MDS-2) and Skin Ulcer Assessment [On-line]. Available: www.npuap.org/
positn3.htm Accessed March 2001.
39. Ouslander JG. The resident assessment instrument (RAI): Promises and pit-
falls. J Am Geriatr Soc 1997;45:975–976.
40. Hawes C, Mor V, Phillips CD et al. The OBRA-87 nursing home regulations
and implementation of the Resident Assessment Instrument: Effects of pro-
cess quality. J Am Geriatr Soc 1997;45:977–985.
41. Thomas DR. Issues and dilemmas in the prevention and treatment of pres-
sure ulcers: A review. J Gerontol A Biol Sci Med Sci 2001;56A:M328–M340.
62 SALIBA ET AL. JANUARY 2003–VOL. 51, NO. 1 JAGS

Appendix 1. Guidelines Not Measured Through Chart Review

IF a nursing home resident is at risk for PU,
THEN massage over bony prominences should be avoided.
IF potential for improving mobility and activity status exists,
THEN rehabilitation efforts should be instituted if consistent with overall goals of therapy. (Maintaining current activity level, mobility,
and range of motion is an appropriate goal for most individuals.)
IF an intervention for PU prevention is implemented,
THEN outcomes of intervention should be monitored and documented.
IF the side-lying position is used in bed,
THEN avoid positioning directly on the trochanter.
IF a resident is at risk for PU,
THEN uninterrupted sitting in chair or wheelchair should be avoided.
IF a resident is chair bound
THEN pressure-reducing devices should be used. Do not use donut-type device.
IF a resident is chair bound,
THEN chair positioning should include consideration of postural alignment, distribution of weight, pressure relief, balance, and
stability.
IF a resident is at risk for PU,
THEN the head of the bed should be maintained at the lowest degree of elevation that is possible.
IF an education program is implemented,
THEN the program should include information on PU etiology, risk, risk assessment, skin assessment, support surfaces,
individualized skin care program, positioning, and documentation of important data.
IF an educational program is implemented,
THEN the program should: 1) identify specific roles and responsibilities of each person; 2) be appropriate to audience; 3) be updated
on a regular basis.
IF an educational program is implemented,
THEN it should be consistent with principles of adult learning and it should be evaluated.
PU  pressure ulcer.