References Type of Study Sample Intervention(s) Outcome Measures &
Length of Follow-up
Results Limitations
Allcock N, Wharrad H, Nicolson A. Interpretation of pressure-sore prevalence. J Adv Nurs.
1994;20(1):37-45.
Study data were collected by the infection control link nurses (ICLN) who all attended training sessions to inform them of the aims and methods of the study as well as receiving education on the grading system
803 patients on wards day of study
The training involved a revision of the structure of the skin followed by a discussion of the grading system and an opportunity to practice using the grading system using color slides of sores collected prior to the study
It was not possible to assess the accuracy of pressure-sore identification due to the resources available The reliability of the ICLNs pressure sore grading was estimated using a series of six pictures not previously seen by the ICLN
This showed that the distinction between grades 1 and 2, 3 and 4, and 5 and 6 were difficult The number correct for each grade was 49%, 47%, 46%, 32%, 36% and 42%
for grades 1-6 respectively If the grades are combined into three groups the grading of sores was much more accurate, grade1 & 2 98% correct, grade 3 & 4 68%
correct and grade 5 & 6 79% correct
This suggests that the grading scheme used may have had too many categories to be reliable, although the difficulties of grading from a
photograph, due to the difficulty in judging depth and other factors, needs to be taken into account
Beeckman D, Schoonhoven L, Boucqué H, Van Maele G, Defloor T. Pressure ulcers: e-learning to improve classification by nurses and nursing students. J Clin Nurs.
2008;17(13):1697-707.
Random assignment of convenience sample to e-learning of the EPUAP grading system for pressure ulcers. Control group received instruction in the classroom
Nursing students Qualified nurses from 7 hospitals, 7 nursing homes, 1 home care agency and 5 schools of nursing
Both groups received similar illustrations, tests and instruction
Both groups tested on 20 photographs of normal skin (1), blanchable erythema (1), stage I (3), II (3), III (3), IV (3), moisture lesions (3) and combined ulcers (3) (both moisture and pressure)
No difference in pretest between groups
Self-assessment of expertise of identifying ulcers
Reliability improved with training but was not perfect, problems distinguishing stages and IAD from pressure ulcers
E-learning and classroom learning led to same outcomes
Beeckman D, Schoonhoven L, Fletcher J, Furtado Kt, Gunningberg L, Heyman H, et al. EPUAP classification system for pressure ulcers:
European reliability study.
Classified 20 photographs of normal skin (1), blanchable erythema (1), stage I (3), II (3), III (3), IV (3), moisture lesions (3) and combined ulcers (3) (both moisture and pressure)
Pressure ulcers were often classified erroneously and only a minority of nurses reached a substantial level of agreement: 33% reached moderate agreement and only 5% reached high agreement with known grade or skin lesion. Grade 3 lesions were regularly classified as grade 2. Non-blanchable erythema was frequently assessed incorrectly as blanchable erythema. Furthermore, the differential diagnosis between moisture lesions and pressure ulcers appeared to be complicated Briggs S-L. How accurate
are RGNs in grading pressure ulcers? Br J Nurs.
2006;15(22):1230-4.
respondent fell in the group which achieved between 16-20 (80%) correct answers and the number in this group rose to four (7.7%) respondents in the post-test. The
33
system
What was the effect of education on the accuracy?
biggest increase was in the group that scored between 11-15 (55% to 75%) correct answers; this group rose from eight (15.3%) in the pre-test to 29 (55.7%) in the post-pre-test.
For those who scored between 6 and 11) correct answers, this group reduced from 23 (44.2%) in the pre-test to 18 (34.6%) in the post-pre-test.
The results of the study indicated the level of accuracy of pressure ulcer grading was poor in the pre-test but markedly improved in the post-test.
Defloor T, Schoonhoven L.
Inter-rater reliability of the EPUAP pressure ulcer classification system using photographs. J Clin Nurs.
2004;13(8):952-9.
Development and validity testing of a series of
photographs of pressure ulcers and moisture lesions using the EPUAP to establish a "gold standard" for accuracy of diagnosis
7 researchers, 20 staff members and 17 pressure ulcer nurses
Photographs were shown to the second group and then differences were measured
Deviations in expert scores occurred in 5.9% of the cases, but 33.3% of the differences was only one grade of difference from the gold standard.
In 3.2% of the cases the deviation is more than two grades and in 7.3%
of the cases the lesion is classified as an incontinence lesion. Conclusion was that photographs were accurate and could be used to measure accuracy of pressure ulcer grading.
An inclusion of ‘‘unclear’’ in the calculation will artificially raise the level of agreement
for difficult cases.
Defloor T, Schoonhoven L, Vanderwee K, Weststrate J, Myny D. Reliability of the European Pressure Ulcer Advisory Panel
classification system. J Adv Nurs. 2006;54(2):189-98.
examine the interrater and intra-rater reliability of classifying pressure ulcers according to the EPUAP classification system when using photographs of pressure ulcers and incontinence lesions
Phase 1: 65 photos were presented in a random order to 473 nurses who were participating at a Congress for Wound Care in Belgium, the Netherlands and Finland.
Phase 2: Eighty-six nurses from a university hospital in Belgium participated in the
A set of 56 photographs of normal skin, blanchable erythema, pressure ulcers and incontinence lesions was used.
If erythema was visible on a photograph, a second photograph was also shown. On this second photograph a transparent pressure disk was pressed onto the erythema so that the extent to which the
The multi-rater Kappa for the 473 nurses was 0. 37 (P < 0.001). If only the pressure ulcer photos (N = 37) were considered, the average Kappa and the weighted Kappas varied between 0.41 and 0.50. The average percentage agreement was 55.6%
and 15.7% of the photos were scored one grade too high or too low.
The average Kappa did not differ statistically significantly between the nurses who received training on
34
Length of Follow-up second phase. The
sequential intra-rater reliability was evaluated by presenting the same 56 photos twice with an interval of 1 month. On both occasions the photographs were presented in a different random order.
erythema was blanchable was visible. The photographs were presented once and the nurses were asked to classify the photographs as normal skin, blanchable erythema, non-blanchable erythema (grade 1 pressure ulcer), blister (grade 2 pressure ulcer), superficial pressure ulcer (grade 3), deep pressure ulcer (grade 4) or incontinence lesion. Also the option ‘do not know’
was given.
At the outset the definitions of grades of pressure ulcers, as categorized by the EPUAP, were presented to the participants in a standardized way.
pressure ulcers and those who did not. P values varied between 0.13 and 0.63.
Grade 1 was the most frequently wrongly classified grade (54.5%;
3091/5676 observations). In 7.9% of the observations nonblanchable erythema was confused with blanchable erythema, and in 19.6%
of cases with incontinence lesions.
The incontinence lesions were also frequently wrongly assessed (44.3%; 1676/3784). In 20.6% of the observations they were assessed as superficial pressure ulcer (grade 3), in 4.2% as blister (grade 2) and in 6.3% as deep pressure ulcer (grade 4).
An inclusion of ‘‘unclear’’ in the calculation will artificially raise the level of agreement for difficult cases.
Gajewski BJ, Hart S, Bergquist-Beringer S, Dunton N. Inter-rater reliability of pressure ulcer staging: ordinal probit Bayesian hierarchical model that allows for uncertain rater response. specification of n =15 PrUs and p =6 raters
To construct a model to estimate the reliability of rating PrUs from patients in NDNQI hospitals. Stages I to IV are ordinal and unstageable, because it combines III and IV (and in this case DTI) does not fit this numeric relationship
Wide variation in reporting unstageable PrU in hospitals
Classified DTI as unstageable
Hart S, Bergquist S, Gajewski B, Dunton N.
Reliability testing of the National Database of Nursing Quality Indicators pressure ulcer indicator. J Nurs Care Qual. 2006;21(3 (Print)):256.
256 individuals at 48 randomly sampled whether the wound was a pressure, venous, arterial, or diabetic foot ulcer.
For part 2, individual participants were asked to stage the pressure ulcer in each of the 17 photographs or to
A 3-part
criterion-referenced test was created in collaboration with experts in wound care and instrument design. High-quality digital pictures of ulcerous wounds were used in this study. Nineteen of the 25 pictures were obtained from and used with permission from the
Most raters correctly identified pressure ulcers from photographs of ulcerous wounds, but frequently misclassified the type of non-pressure ulcer wound. Variations in K agreement were found between certified and noncertified nurses.
Wound, continence, and/or ostomy care certified nurses had higher K values (/c = 0.66, SD = 0.18) than noncertified nurses (K = 0.54,
35
centers or teaching hospitals and 33%/
held Magnet status.
classify it as unstageable.
Participants were instructed to use NPUAP guidelines when staging the wound or classifying it as unstageable.
For part 3 of the test, participants were asked to read 5 scenarios and determine from each whether the patient's pressure ulcer was nosocomial or community acquired.
NPUAP Guidelines of the NPUAP and the AHRQ for pressure ulcer staging and expert opinion were used to assess and stage ulcers in each picture.
The first part of the test (PrU Identification) contained 7 pictures of ulcerous wounds: 3 venous ulcers, 2 pressure ulcers, 1 arterial ulcer, and 1 diabetic foot ulcer. The second part (PrU Stage) contained 18 pictures of pressure ulcers.
During pilot testing, this section contained 4 pictures of stage I pressure ulcers, 3 pictures of stage II pressure ulcers, 5 pictures of stage III pressure ulcers, 5 pictures of stage IV pressure ulcers, and 1 picture of an unstageable pressure ulcer.
The third part (PrU Source) contained 5 scenarios that described a patient's course of hospitalization.
Two versions of the test were created because assessment of the reliability of pressure ulcer staging from Web-based photographs was a new methodology and subject to 2 limitations relative to in-person assessment.
First, even high-quality digital images are 2-dimensional and will not provide all of the information available from direct observation.
Second, the images available to raters were
SD - 0.18). When data were collapsed to binary values (pressure ulcer/not pressure ulcer), K agreement was 0.84 (SD = 0.25), with little variation in K values between wound, continence, and/or ostomy certified nurses (p< = 0.92, SD = 0.15) and noncertified nurses (p = 0.82, SD = 0.27), indicating near perfect reliability. When the test did not include wound descriptors, the adjusted K agreement for pressure ulcer staging by nurses who were certified in wound, continence, and/or ostomy care was 0.66 (SE — 0.04) compared with 0.54 (SE = 0.03) for noncertified nurses. When the test included wound descriptors, the adjusted K value for those certified in wound, continence, and/or ostomy care was 0.83 (SE = 0.03) compared with 0.71 (SE = 0.02) for noncertified nurses.
Their conclusion states that the study results suggest that NDNQI data on pressure ulcers can be used as an indicator of nursing care quality and as a tool for guiding quality improvement initiatives.
36
Length of Follow-up dependent on the quality of
their computer monitors. To address the 2-dimensional limitation, the first test version contained a short narrative description (wound descriptor) of each pressure ulcer in part 2 of the test. This was done to provide contextual information about the wound ordinarily available to the rater through direct wound observation.
Specifically, information about wound size, depth, surface characteristics, and surrounding tissue were included in the narrative description that
accompanied the pressure ulcer photographs. While this additional information may have provided details typically available to an in-person rater, it might also have cued raters to consider wound characteristics that might not have been noticed without prompting.
Therefore, the second version of the test excluded wound descriptors in part 2 to examine their effect on the reliability of pressure ulcer staging and enhance the comparability of findings to previous research studies.
It was anticipated that the reliability results from the version containing wound descriptors would be higher than the picture-only
37
version, and that the 2 versions of the test would bracket the reliability resulting from bedside, in-person assessments.
Houwing RH, Arends JW, Canninga-van Dijk MR, Koopman E, Haalboom JRE.
Is the distinction between superficial pressure ulcers and moisture lesions justifiable? A clinical-pathologic study. Skinmed.
2007;6(3):113-7.
12 lesions with blanchable erythema classified as moisture lesions
Punch biopsy taken 5 mm from superficial skin defect in incontinence lesions and within stage I
4 of the lesions were due to ischemia and 8 were chronic irritation
Stages I through IV are associated ischemic pathology. Moisture lesions are associated with a chronic irritation
Nixon J, Cranny G, Bond S.
Pathology, diagnosis, and classification of pressure ulcers: comparing clinical and imaging techniques.
Wound Repair Regen.
2005;13(4):365-72.
A total of 143 patients consented to participate in the pilot study (April to July 1998) and the main study (September 1998 to May 1999), including 93 elective and 50 acute patients.
To detect differences in mean blood flow perfusion units) between clinical skin grades— Skin lesions classified as:
0 No skin changes 1a Redness to skin (blanching) 1b Redness to skin (nonblanching) 1bþ Redness to skin (nonblanching) plus one or more:
Pain Induration Heat Edema Discoloration (specify) 2 Partial thickness wound involving epidermis/dermis only 3 Full thickness wound involving subcutaneous tissue
4 Full thickness wound through subcutaneous tissue to muscle or bone
5 Black eschar
Skin on the buttocks and sacrum was assessed immediately
postoperatively and daily until discharge using a combination of the following clinical and physiological measures:
clinical skin assessment preoperatively,
postoperatively, and daily until discharge), laser Doppler imaging (1/2 hour and 1 hour postoperatively), and laser Doppler imaging of Grade 1b and Grade 1bþ skin areas observed during follow-up, where feasible.
Histogram was created of pixels at each level of tissue injury.
Similarness of pixel levels was compared. There was a difference between the levels, with grade iB (DTI?) the most different
38
Length of Follow-up Plus a minimum
sample size of 42 scans was estimated using pilot study data. This was based on an analysis of variance with 95 percent power at the 5 percent significance level.
Nixon J, Thorpe H, Barrow H, Phillips A, Andrea Nelson E, Mason SA, et al.
Reliability of pressure ulcer classification and diagnosis. J Adv Nurs.
2005;50(6):613-23.
A total of 378 paired assessments was undertaken by 116 nurses for the inter-rater reliability assessments during the period from December 2000 to February 2001
0 No skin changes 1a Redness to skin (blanching)
1b Redness to skin (non-blanching)
2 Partial thickness wound involving
epidermis/dermis only (i.e.
skin break or blister) 3 Full thickness wound involving subcutaneous tissue
4 Full thickness wound through subcutaneous tissue to muscle or bone 5 Black eschar
Agreement was measured between CRNs and WNs for the 2396 paired site assessments for all grades.
There were a total of 508 (21.2%) disagreements between CRNs and WNs: 419 were one grade different (such as 0/1a, 1a/1b and so on), 68 were two grades different [including 0 and 1b (21), 1a and 2 (46) and 3 and 5 (1)], and 21 were more than two grades different [including 0 and 2 (13), 0 and 3 (1), 1a and 3 (3), 2 and 5 (4)]
Russell LJ, Reynolds TM.
How accurate are pressure ulcer grades? An image-based survey of nurse performance. J Tissue Viability. 2001;11(2):67.
Thirty images were selected and classified with the EPUAP scale, establishing the gold standard through consensus of an expert panel. Twelve images were then used for the study in paper form.
These images were classified by 97 of 200 contacted nursing experts from EPUAP and Great Britain.
About 61.9% of the grades agreed with the gold standard, the mean deviation from the gold standard was 0.49 (standard deviation 0.25).
Stausberg Jr, Lehmann N, Krager K, Maier I, Niebel W. Reliability and validity of pressure ulcer diagnosis and grading: an image-based survey. Int J Nurs
No pressure ulcer (grade 9 in the analysis);
Pressure ulcer grade 1:
Localized erythema of the skin; Pressure ulcer grade 2:
Loss of skin; Pressure ulcer
Seven participants with 100 assessments each led to 700 grades.
Comparing every two participants with each other led to 21 pairs with 2100 single comparisons. Half of the results related to the foot/heel
39
Stud. 2007;44(8):1316-23. grade 3: Deep loss of skin,
muscles and tendons are visible and possibly affected; Pressure ulcer grade 4: Deep loss of skin with affection of bones.
region, the other half to the buttock/hip region (350 grades and 1050 single skin comparisons each).
In about one-third of the images, all participants agreed concerning the grade, in about 80% at least five participants agreed.
Verdu J. Can a decision tree help nurses to grade and treat pressure ulcers?
J Wound Care.
2003;12(2):45-50.
Random assignment of 66 nurses to a decision tree to classify pressure ulcers. Nurses randomly selected from 100 at the General University Hospital of Elche, Alicante, Spain. All had a similar level of experience and often cared for patients with chronic wounds.
At baseline participants were given a sealed envelope containing three clinical case studies, each of which briefly described the health status of a patient and his or her pressure ulcer, and included photographs of the site and aspect of the wounds.
Of the 66 nurses, 34 (51.5%) were assigned to the control group and 32 (48.5%) to the experimental Group
Case 1 is a stage I or II Case 2 is a stage III Case 3 is a stage IV with necrotic tissue Photos are blurred
Case I* Case II** Case III***
ConNo.
(%)
Exp.
No. (%)
Con No.
(%)
Exp. No.
(%)
Con No.
(%)
Exp. No.
(%) Accurate 29 (85.3) 25 (78.1) 18 (52.9) 21 (65.6) 21 (61.8) 14 (43.8) Inaccurate 5 (14.7) 7 (21.9) 16 (47.1) 11 (34.4) 13 (38.2) 18 (56.2)
There was no statistically significant difference between the grades selected by the two groups.
Overall accuracy rates were:
● Case I: 81.8% (54/66)
● Case II: 59.1% (39/66)
● Case III: 53.0% (35/66)