Esacaras Journal of the American Geriatrics Society 2012 60 (9) 1603

7/27/2019 Esacaras Journal of the American Geriatrics Society 2012 60 (9) 1603

1/6

CLINICAL INVESTIGATIONS

Hospital-Acquired Pressure Ulcers: Results from the NationalMedicare Patient Safety Monitoring System Study

Courtney H. Lyder, ND,* Yun Wang, PhD,k Mark Metersky, MD,# Maureen Curry, MHA,

Rebecca Kliman, MPH,** Nancy R. Verzier, MSN, and David R. Hunt, MD

OBJECTIVES: To determine the national and state inci-dence levels of newly hospital-acquired pressure ulcers(PUs) in Medicare beneficiaries and to describe the clinicaland demographic characteristics and outcomes of theseindividuals.

DESIGN: Retrospective secondary analysis of the nationalMedicare Patient Safety Monitoring System (MPSMS)

database.SETTING: Medicare-eligible hospitals across the UnitedStates and select territories.

PARTICIPANTS: Fifty-one thousand eight hundred forty-two randomly selected hospitalized fee-for-service Medi-care beneficiaries discharged from the hospital between

January 1, 2006, and December 31, 2007.

MEASUREMENTS: Data were abstracted from theMPSMS, which collects information on multiple hospitaladverse events.

RESULTS: Of the 51,842 individuals in the MPSMS2006/07 sample, 2,313 (4.5%) developed at least one newPU during their hospitalization. The mortality risk

adjusted odds ratios were 2.81 (95% confidence interval(CI) = 2.443.23) for in-hospital mortality, 1.69 (95%CI = 1.611.77) for mortality within 30 days after dis-charge, and 1.33 (95% CI = 1.231.45) for readmissionwithin 30 days. The hospital riskadjusted main length ofstay was 4.8 days (95% CI = 4.75.0 days) for individualswho did not develop PUs and 11.2 days (95% CI = 10.1911.4) for those with hospital-acquired PUs (P < .001).

The Northeast region and Missouri had the highest inci-dence rates (4.6% and 5.9%, respectively).

CONCLUSION: Individuals who developed PUs weremore likely to die during the hospital stay, have generallylonger hospital lengths of stay, and be readmitted within30 days after discharge. J Am Geriatr Soc 60:16031608,2012.

Key words: pressure ulcer; adverse events; hospital-acquired pressure ulcer; patient safety; medical events

The development of pressure ulcers (PUs) has beenassociated with the quality of health care.1,2 Hence,higher rates of PU development (PUD) may signal overallpoor care by the healthcare system. The National QualityForum (NQF) created hospital Never Events in 2003

events that should never occur during hospitalization.1

TheNQF believes that development of Stage III or IV PUs aresuch events. In the 13 states that have incorporated NeverEvents, hospitals can be financially penalized for notreporting individuals who develop Stage III or IV PUs in atimely manner. The federal government has also identifiedPUD as a public heath concern.2 Most recently, the Cen-ters for Medicare and Medicaid Services (CMS) made asignificant policy decision not to pay for hospital-acquiredStage III and IV PUs.3

A major limitation of existing PU studies has beentheir small sample size (


2/6

PUs among hospitalized Medicare beneficiaries as a basisfor measuring improvement. Neither are there knownpublished Medicare studies that have reported on HAPUsagainst which hospitals can benchmark at the national orstate level. Given the increasing number of Medicarebeneficiaries being admitted to hospitals and theconcomitant potential for increases in PU incidence, aMedicare Patient Safety Monitoring System (MPSMS)

study capitalized on abstracted data from a large sampleof medical records of fee-for-service Medicare beneficiariesdischarged in 2006/07 to address these gaps inknowledge.4 The data for this study came from threenational databases: the CMS National Claim Historydatabase, the CMS Claims History database, and theMPSMS. The purpose was to explore the overall incidence,prevalence, and clinical characteristics associated withMedicare beneficiaries who developed HAPUs and todetermine rates in states and geographic regions.

STUDY SAMPLE

The MPSMS is a nationwide surveillance system designedto identify rates of specific adverse events within thehospitalized fee-for-service Medicare population. Anadverse event is an unintended harm, injury, or loss that ismore likely associated with an individuals interaction withthe healthcare delivery system than with diseases theindividual may have. The MPSMS determines the nationalrates for Medicare beneficiaries in the following adverseevent categories: central venous catheter (CVC)-associatedblood stream infections; CVC-associated mechanicaladverse events; CVC-associated blood stream infection(BSI) adverse drug events; HAPU- and ventilator-associatedpneumonia; hip and knee replacements; and postoperativerates of venous thrombolic event, cardiac events, and

pneumonia.4The MPSMS PU study uses secondary analyses of the

Hospital Payment Monitoring Program (HPMP) sample.5

The HPMP medical record sample is an existing database,selected randomly each month from the Medicare NationalClaims History File by CMS from a pool of approximately1 million fee-for-service Medicare beneficiary hospitaldischarges. For this study, the records of hospitaldischarges between January 1, 2006, and December 31,2007, totaling 51,842 Medicare fee-for-service inpatientdischarges across the 50 states, Washington, DC, PuertoRico, and the U.S. Virgin Islands, were selected for use.

Trained medical record abstractors collected

documentation describing individuals with HAPUs thatdeveloped during their index hospitalizations. The medicalrecord abstractors were trained on model charts. Interraterreliability (between principal investigator and medicalabstractors) was established at 90% before the medicalabstractors were allowed to abstract the medical records.

Charts of individuals with PUs present on admission(prevalence) were reviewed to determine whether newulcers (incidence) developed during the hospital stay. TheNational PU Advisory Panels Stage I (2001) and Stage IIto IV (1989) definitions of PUs were used to distinguish aPU from another potential skin injury.6,7 Thecomorbidities for PUs were defined as congestive heart

failure (CHF), chronic obstructive pulmonary disease

(COPD), cerebrovascular disease (CVD), diabetes mellitus,usage of corticosteroids, obesity, and smoking. These wereselected because of their association with PUs in theliterature. The PU diagnosis was determined based onphysician and nurse documentation in the medical recordduring admission (prevalence) or at any time duringhospitalization. Newly developed PUs (incidence) werealso monitored during hospitalization.

The abstractors determined whether an individualdeveloped a HAPU based on documentation of a HAPU inthe medical record or on the description of ulceration. Thelocations of PUs, newly acquired and those documentedon admission, were also recorded. To further distinguishHAPUs from PUs already present on admission, new PUsfound in the same body region as PUs present onadmission were not counted as HAPUs. Any individualwho developed at least one new PU during thehospitalization was included in the analyses. Participantdemographic information (e.g., age, sex, race),International Classification of Diseases, Ninth Revision(ICD-9) diagnoses (707), and other characteristics (e.g.,obesity, CHF, COPD as determined by ICD coding) wereobtained from the Medicare Enrollment Database.

The outcomes of interest were in-hospital mortalityrates, readmission or mortality during the 30 days afterdischarge, and hospital length of stay. The CMS NationalClaim History database was the source of in-hospitalmortality and readmission information. The MedicareEnrollment Database was the source of 30-day mortalityinformation.

STATISTICAL ANALYSIS

Descriptive and bivariate analyses were conducted toidentify participants baseline demographics and medical

diagnoses and to compare the observed differences incharacteristics and outcomes (mortality, readmission, andlength of stay) of participants who developed HAPUs andthose who did not. The chi-square test was used tocompare dichotomous and categorical variables, and thet-test was used to compare continuous variables. Thehierarchal generalized linear modeling (HGLM) approach810 was applied to assess the association betweenparticipant characteristics and development of HAPUs bymodeling the log-odds of HAPUs as a function ofparticipant demographic and clinical variables. Thisapproach was used to determine whether there was arelationship between the outcomes and development of

HAPUs after adjustment for participant characteristics.Two modeling steps were constructed for each

outcome. The first step was fitted to a model withoutadjusting for participant characteristics, and the secondstep was fitted to a model with adjustments for participantdemographics and medical conditions.11 To examinedifferences in HAPU rates across states and regions, allHGLMs were fitted with a random state-specific effect toaccount for within-state correlation of the observedadverse events and outcomes to distinguish betweenwithin-state variations from between-state variation. TheCMS regions were used to cluster states. A 95%confidence interval (CI) was calculated for each estimate

for the models. To take into account differences in

1604 LYDER ET AL. SEPTEMBER 2012VOL. 60, NO. 9 JAGS


3/6

Medicare fee-for-service volume between states, allHGLMs were weighted by a total number of dischargesfrom each state. All statistical analyses were conductedusing Stata version 8.0 (StataCorp., College Station, TX)and SAS version 8.12 (SAS Institute, Inc., Cary, NC).Hierarchical models were estimated using the GLIMMIXmacro in SAS.

RESULTS

Participant Characteristics

Fifty-one thousand eight hundred forty-two dischargeswere included in the final study sample for the combinedyears 2006 and 2007 (Table 1). The HAPU incidence ratewas determined to be 4.5% (2,313/51,842), and PU preva-lence on admission was 5.8% (2,999/51,842). Of the 2,999individuals who entered the hospital with a PU, 16.7%(502/2,999) developed at least one new PU at a differentlocation during their hospitalization. The majority of par-ticipants who developed at least one PU were nonwhiteand aged 75 to 84. These participants had significantlyhigher rates of CHF, COPD, CVD, diabetes mellitus, anduse of corticosteroids during hospitalization. Obesity wassignificantly associated with HAPUs (Table 1).

The majority of HAPUs were located on the coccyx orsacrum (41%), followed by the hip and buttock region(23%) and the heels (23%). The stages of HAPUs couldnot be determined because of the wide variability in docu-mentation of description and staging by clinicians. Partici-pant characteristic such as age, diagnosis (cancer, CHF,COPD, CVD, diabetes mellitus), and presence of obesitywere all significantly associated with HAPUs (P < .001;Table 2).

Overall, the nationwide HAPU incidence rate was

4.5% (95% CI = 4.1

4.7%). Variance across the nationwas statistically significant (Figure 1). The between-state

variance was 3.2% (standard error 1.2%), the weightedrates drawn from HGLM ranged from 3.1% (95%CI = 2.54.1%) to 5.9% (95% CI = 5.17.0%). The oddsof developing PUs if treated in a state 1 standard deviation(SD) above the national average relative to the odds ofdeveloping PUs if treated at a state 1 SD below thenational average were 1.43. Higher incidence rates werenoted in the Northeast and Missouri. The five states withthe lowest HAPU rates were Wisconsin (3.1%), Alabama(3.3%), Tennessee (3.7%), Puerto Rico (3.7%), and NorthCarolina (3.8%), and the five states with the highestHAPU rates were New York (5.2%), Missouri (5.3%),New Jersey (5.3%), Massachusetts (5.5%) and Pennsylva-

nia (5.9%). Rates in Wisconsin and Alabama were statisti-cally significantly lower than the national average.

Table 1. Participant Characteristics

Characteristic

Total

(N = 51,842)

Participants with

Pressure Ulcers

(n = 2,313)

Participants

without Pressure

Ulcers (n = 49,529) P-Value

Age, mean standarddeviation

73.3 13.0 78.0 11.2 73.2 13.0


4/6

Mortality, Readmission, and Length of Stay

The mortality data revealed that the development of HAPUswas significantly associated with higher in-hospital mortal-ity (11.2%) and mortality within 30 days after discharge(15.3%). HGLMs were fitted to further understand theassociation between the development of new HAPUs andhospital outcomes (Table 3), mortality and hospital lengthof stay (Table 4). Participants with HAPUs were signifi-cantly more likely to be readmitted within 30 days after dis-charge (odds ratio (OR) = 1.33, 95% CI = 1.231.45), and

were more likely to have died in the hospital. The risk-adjusted ORs were 2.81 (95% CI = 2.443.23) for in-hospi-tal mortality and 1.69 (95% CI = 1.611.77) for mortalitywithin 30 days after discharge. Participants who developedHAPUs had significantly longer hospital lengths of stay(11.6 10.1 days) than those without (4.9 5.2 days).

DISCUSSION

The data from this study revealed that HAPUs remain asignificant problem for hospitalized individuals. Althougha 4.5% HAPU incidence rate and 5.8% prevalence rate onadmission were found, no other large Medicare studies

were located to place these findings into context of the cur-rent literature. The between-state incidence rate variance

was 3.2% (P < .001). Thus, to benchmark appropriately,it is imperative for hospitals to be cognizant not only of

the national incidence rate, but also of their own statesrates. Several states mandate that hospitals track and

Figure 1. Pressure ulcer incidence rates according to Centers for Medicare and Medicaid Services Regional Map.

Table 3. Association Between Hospital Outcomes and Pressure Ulcer (PU) Development

Outcome Total With PUs Without PUs P-Value

Mortality, n (%)Within 30-days after discharge 2,551 (4.0) 353 (15.3) 2,198 (4.4)


5/6

report their PU rates to their state departments of health,but it is often difficult for hospital staff to obtain access torelevant benchmarks for comparison between likeminded hospitals within the state. Thus, this studyprovides a glimpse of PU incidence and presence at thestate and national level for Medicare beneficiaries. Mostprevalence studies use a 1-day period to determine com-munity-acquired PU rates for the hospital, with some

studies reporting hospital PU prevalence rates of as high as15%.1214

The large difference between the 15% reported in theresearch literature and the 5.8% found in the current studymay be because, when a 1-day prevalence is used, individu-als with longer hospital lengths of stay are more likely to becounted. Another explanation for this difference is that themajority of prevalence studies reported are completed dur-ing hospitalization, versus the current study, which countedon PUs on admission and HAPUs. The findings from theMPSMS have also demonstrated that individuals withHAPUs have much longer hospital lengths of stay. Thus, a1-day prevalence study (often used by hospitals to deter-mine prevalence rates) may not be an accurate reflection ofindividuals who developed HAPUs during their stay.

The findings of the current study support current liter-ature that suggests that specific chronic diseases such asCHF, COPD, CVD, diabetes mellitus, obesity, and use ofcorticosteroids may increase vulnerability to the develop-ment of HAPUs.15,16 Thus, individuals entering the hospi-tal with a constellation of these conditions should beidentified as being at higher risk for developing HAPUs.

The majority of participants developed HAPUs overthe coccyx and sacrum, hips and buttocks, and heels.These findings support the analyses by the NationalPressure Ulcer Advisory Panel (NPUAP).14 Obesity wasassociated with HAPUs. Chronic impairment of systematic

perfusion that occurs in individuals who are obese fre-quently results in chronic skin and wound problems.17

Blood supply to fatty tissue may not be adequate toprovide appropriate oxygen and nutrition. This could befurther compounded if the individuals dietary input doesnot include essential vitamins and nutrients.

HAPUs also have been shown to be an important riskfactor associated with mortality. The MPSMS PU findingsstrengthen the body of research that suggests that individu-als with HAPUs have higher mortality in the hospital andwithin 30 days of discharge. Therefore, hospitals shouldidentify individuals at high risk for HAPUs and implementpreventative interventions on admissions accordingly. It

could be argued that, because of the good PU preventionprograms being implemented in hospitals throughout theUnited States, rates of 4.5% might be acceptable.18,19

These findings also suggest that HAPUs may developindependent of good care being provided.20,21 Thus, someHAPUs may be unavoidable, as suggested by the NPUAP22

and other national associations.23,24

There were several limitations inherent in the MPSMSPU study. Retrospective abstraction of medical records is alimitation in data collection because it is possible that HA-PUs could have been present and simply not recorded oradequately described. In several cases, the clinician did notdocument PUs, yet based on clinical characteristics and the

location and description of the ulcer, the research team

decided to count these as HAPUs. There is a slight possi-bility (highly unlikely) that these were not HAPUs. Carewas taken to exclude all other chronic ulcers based onmedical diagnoses in the medical records and using ICD-9coding (e.g., diabetic ulcers, venous stasis ulcers, arterialulcers). To distinguish individuals with HAPUs from thosewith PUs already present on admission, no PUs, new orold, found in the same body region as a PU already pres-

ent on admission were counted as HAPUs. This approachmay have led to undercounting the number of individualswho developed new HAPUs during the hospital stay.

The inability to capture the staging of HAPUs wasanother limitation. The variability of documented stagingof HAPUs in the medical record made this variable unreli-able. Staging of PUs using the NPUAP staging system is notrequired. Given the new CMS Hospital-Acquired Condi-tions, under which hospitals will no longer be reimbursedfor Stage 3 and 4 PUs unless they were identified uponadmission, accuracy in staging of these ulcers may improve.That is, clinicians may need to be reeducated on staging ofPUs using the 2007 NPUAP staging system because correctstaging of PUs is necessary to make appropriate choices intheir management of the PUs. Although several undesirableoutcomes were associated with the development of HAPUs,no assertion is made that there was a causal relationshipbetween the development of HAPUs and these outcomes.Nonetheless, describing these associations is important tohelp identify high-risk individuals. For example, eventhough it cannot be stated that HAPUs directly contributedto the longer hospital length of stay, these findings suggestthat an individual with a prolonged hospital length of stayis at greater risk of developing HAPUs and that both addfinancial burden to the hospital. Hospitals should use theMSPMS PU data as one point of comparison, but hospitalsmay have different rates because of differences in patient

acuity, so the data from the current study may or may notindicate that institutional improvement in PU prevention iswarranted. The data support that multiple chronic condi-tions or any combination of these conditions may contrib-ute to the development of HAPUs. Perhaps these conditionslead to readmission and HAPUs are merely a surrogate foridentifying very sick people rather than being a causativefactor for readmission. Thus, data on the incidence andprevalence of PUs must not simply be collected; under-standing of the underlying diseases that may lead toHAPUs must be gained.

Regardless of the limitations, the MPSMS PU studyhas major strengths. To the best of the knowledge of the

authors, this is the first study to use data abstracted directlyfrom medical records to assess HAPUs in hospitalizedMedicare beneficiaries at the national and state levels. Withthe use of an evidence-based algorithm developed bynationally recognized experts, the MPSMS is the largestdatabase of its kind. The PU findings have important clini-cal and public health implications. MPSMS PU findingsclearly guide clinicians to anatomical sites where PUs aremost likely to develop and identify critical characteristicsand conditions that increase the risk of HAPUs in Medicarebeneficiaries. The findings suggest that these individuals aremore likely to die or be readmitted within 30 days. It wassurprising that the Northeast region had higher rates

of HAPUs. The aggressive prevention and recognition

JAGS SEPTEMBER 2012VOL. 60, NO. 9 HOSPITAL-ACQUIRED PRESSURE ULCERS 1607


6/6

programs occurring in the Northeast region may lead tohigher rates being reported than in other regions, whichmay explain this in part.2527 Regardless, with these data,it is now possible for hospitals to benchmark their Medi-care beneficiaries HAPU rates at the national and statelevels.

ACKNOWLEDGMENTS

The analyses upon which this publication is based were per-formed under Contract 5002006-CToo2C, entitled Utili-zation Quality Control: Quality Improvement Organizationfor the State of Connecticut, sponsored by CMS, Depart-ment of Health and Human Services. The authors assumefull responsibility for the accuracy and completeness of theideas presented. This article is a direct result of the HealthCare Quality Improvement Program initiated by CMS,which has encouraged identification of quality improvementprojects derived from analyses of patterns of care and there-fore required no special funding on the part of the contrac-tor. Ideas and contributions to the authors concerningexperience and engaging with issues presented are wel-

comed. The authors would like to thank the MPSMS teamat Qualidigm for its support and contributions. We specifi-cally thank Nancy Verzier, RN, MSN, CPHQ, and MichaelPineau, RN, MS, for their excellent management of theproject and Nancy Morse for manuscript preparation.

Conflict of Interest: There are no financial, personal,potential conflicts of interest in the conduct of the study orin the manuscript development. Dr. Lyder had full accessto all of the data in the study and takes responsibility forthe integrity of the data and the accuracy of data analysis.M. Meterskys employer is remunerated for his qualityinsurance and safety work with Qualidigm.

Author Contributions: Study concept and design:

Courtney Lyder, Mark Metersky, Nancy Verzier, DavidHunt. Acquisition of subjects and/or data: Courtney Lyder,Maureen Curry, Nancy Verzier. Analysis and interpreta-tion of data: Courtney Lyder, Yun Wang, Mark Metersky,Maureen Curry. Preparation of manuscript: CourtneyLyder, Yun Wang, Mark Metersky, Maureen Curry,Rebecca Kliman, Nancy Verzier, David Hunt.

Sponsors Role: The content of the publication doesnot necessarily reflect the views or policies of the Depart-ment of Health and Human Services, nor does mention oftrade names, commercial products, or organizations implyendorsement by the U.S. government.

REFERENCES1. Pressure Ulcer Stages Revised by NPUAP [on-line]. Available at http://

www.npuap.org/pr2.htm Accessed May 31, 2010.

2. National Quality Forum. Serious Reportable Events in Healthcare: A Con-

sensus Report. Washington, DC: National Quality Forum, 2002.

3. Medicare Program; Changes to the Hospital Inpatient Prospective Payment

Systems and Fiscal Year 2008 Rates [on-line]. Available at http://www.cms.

hhs.gov/AcuteInpatientPPS/downloads/CMS-1533-FC.pdf Accessed October15, 2010.

4. Healthy People [on-line]. Available at http://www.healthypeople.gov/Docu-

ment/HTML/Volume1/01Access.htm Accessed October 15, 2010.5. Hospital Payment Monitoring Program: Monitoring and Reducing Mis-

souris Payment Error Rate [on-line]. Available at http://www.primaris.org/

sites/default/files/resources/HPMP/hpmp%20info%20sheet.pdf Accessed

September 20, 2011.

6. 1989 NPUAP Pressure Ulcer Stages [on-line]. Available at http://www.

npuap.org/positn6.htm Accessed September 20, 2011.

7. Ankrom MA, Bennett RG, Sprigle S et al. Pressure-related deep tissue

injury under intact skin and the current pressure ulcer staging systems. Adv

Skin Wound Care 2005;18:3542.

8. Austin PC, Tu JV, Alter DA. Comparing hierarchical modeling with tradi-

tional logistic regression analysis among patients hospitalized with acutemyocardial infarction: Should we be analyzing cardiovascular outcomes dif-

ferently?. Am Heart J 2003;145:2735.

9. DeLong E. Hierarchical modeling: Its time has come. Am Heart J2003;145:1618.

10. Christiansen CL, Morris CN. Improving the statistical approach to health

care provide profiling. Ann Intern Med 1997;127:764768.

11. Coyle YM, Battles JB. Using antecedents of medical care to develop valid

quality of care measures. Int J Qual Health Care 1999;2:512.

12. Robinson C, Gloeckner M, Bush S et al. Determining the efficacy of a pres-

sure ulcer prevention program by collecting prevalence and incidence data:

A unit-based effort. Ostomy Wound Manage 2003;49:4446. 4851.

13. Whittington K, Patrick M, Roberts JL. A national study of pressure ulcer

prevalence and incidence in acute care hospitals. J Wound Ostomy Conti-nence Nurs 2000;27:209215.

14. Cuddigan J, Ayello EA, Sussman C. Pressure Ulcers in America: Prevalence,

Incidence, and Implications for the Future. Reston, VA: National Pressure

Ulcer Advisory Panel, 2001.15. Lyder CH, Preston J, Grady J et al. Quality of care for hospitalized medi-

care patients at risk for pressure ulcers. Arch Intern Med 2001;161:1549

1554.

16. Allman RM, Goode PS, Patrick MM et al. Pressure ulcer risk factors

among hospitalized patients with activity limitations. JAMA 1995;273:865

870.

17. Krasner DL, Kennedy-Evans KL et al. Bariatric wound care: Common

problems and management strategies. Bariatric Times 2006;3:2627.

18. Shahin EM, Dassen T, Halfens RG. Incidence, prevention and treatment of

pressure ulcers in intensive care patients: A longitudinal study. Int J Nurs

Study 2009;46:413421.

19. Gunningberg L, Lindholm C, Carlsson M et al. Reduced incidence of pres-

sure ulcers in patients with hip fractures: A 2-year follow-up of qualityindicators. Int J Qual Health Care 2001;13:399407.

20. Langemo D, Brown G. Skin fails too: Acute, chronic and end-stage skin

failure. Adv Skin Wound Care 2006;9:206211.

21. Black JM, Edsberg LE, Baharestani MM et al. Pressure ulcers: Avoidableor unavoidable? results of the national pressure ulcer advisory panel con-

sensus conference. Ostomy Wound Manage 2011;57:2437.

22. Pressure Ulcers: Avoidable or Unavoidable? Results of the National Pres-

sure Ulcer Advisory Panel Consensus Conference [on-line]. Available at

http://npuap.org/OWM2011_Black_0.pdf Accessed February 27, 2012.

23. Avoidable versus Unavoidable Pressure Ulcers [on-line]. Available at

http://www.wocn.org/resource/resmgr/docs/wocn-avoidable-unavoidable_p.pdf

Accessed February 27, 2012.

24. European Pressure Ulcer Advisory Panel [on-line]. Available at http://www.

epuap.org/ Accessed February 27, 2012.25. Partnering for Prevention A Pressure Ulcer Prevention Collaborative Project

[on-line]. Available at http://www.nursingcenter.com/pdf.asp?AID=762003Accessed September 24, 2011.26. Massachusetts Pressure Ulcer Collaborative [on-line]. Available at http://

www.patientcarelink.org/uploadDocs/1/Statewide-Pressure-Ulcer-Collabora-

tive_Fact-Sheet.pdf Accessed September 24, 2011.

27. Preventing Pressure Ulcers in Hospitals [on-line]. Available at http://www.

ahrq.gov/research/ltc/pressureulcertoolkit/putoolkit.pdf Accessed September

24, 2011.

1608 LYDER ET AL. SEPTEMBER 2012VOL. 60, NO. 9 JAGS

Documents

Esacaras Journal of the American Geriatrics Society 2012 60 (9) 1603