Journal of Rehabilitation Research and Development

Vol. 40, No. 5, September/October 2003

Pages 433–442

Pressure ulcers in veterans with spinal cord injury: A retrospective

study

Susan L. Garber, MA, OTR, FAOTA, FACRM; Diana H. Rintala, PhD

Houston Veterans Affairs Medical Center, Baylor College of Medicine, Houston, TX

Abstract—Pressure ulcers are a major complication of spinal INTRODUCTION

cord injury (SCI) and have a significant effect on general
health and quality of life. The objectives of this retrospective Pressure ulcers are defined as lesions caused by unre-
chart review were to determine prevalence, duration, and lieved pressure, resulting in damage to the underlying tis-
severity of pressure ulcers in veterans with SCI and to identify sue. They usually occur over bony prominences and are
predictors of (1) outcome in terms of healing without surgery, classified as stages by the degree of tissue damage
not healing, or referral for surgery; (2) number of visits veter- observed [1]. The etiology of pressure ulcers is multidi-
ans made to the SCI outpatient clinic or received from home
mensional. Pressure, shear, friction, moisture, and poor
care services for pressure ulcer treatment; and (3) number of
nutrition contribute directly to the physiological etiology
hospital admissions and days hospitalized for pressure ulcer
of pressure ulcers. Other factors associated with the devel-
treatment. From a sampling frame of 553 veterans on the
Houston Veterans Affairs Medical Center SCI roster, 215
opment of pressure ulcers include immobility and psycho-
(39%) were reported to have visited the clinic or received social factors, such as inadequate personal and financial
home care for pressure ulcers (ICD-9 code 707.0 = decubitus, resources and noncompliance with acknowledged preven-
any site) during the 3 years studied (1997, 1998, and 1999). tive behaviors [2,3]. Pressure ulcers are a serious, life-
From this sample, 102 veterans met the inclusion criteria for long complication of spinal cord injury (SCI). Reliable
further analyses, 56% of whom had paraplegia. The duration of prevalence data have been difficult to obtain because of
ulcers varied greatly from 1 week to the entire 3-year time-
frame. Overall, Stage IV pressure ulcers were the most preva-
lent as the worst ulcer documented. Number and severity of Abbreviations: ANOVA = analyses of variance, ASIA = Amer-
ulcers predicted outcome and healthcare utilization. This study ican Spinal Injury Association, DHCP = Decentralized Hospital
illustrates the magnitude of the pressure ulcer problem among Computer Program, SCI = spinal cord injury, VAMC = Veterans
veterans with SCI living in the community. Reducing the prev- Affairs Medical Center.
alence of pressure ulcers among veterans with SCI will have a This material was based on work supported by the Depart-
significant impact on the Department of Veterans Affairs’ ment of Veterans Affairs Rehabilitation Research and
financial and social resources. Innovative approaches are Development Center of Excellence on Healthy Aging with
needed to reduce pressure ulcer risk in veterans with SCI. Disabilities, Houston Veterans Affairs Medical Center,
Houston, Texas.
Address all correspondence and requests for reprints to Susan
L. Garber, MA, OTR, FAOTA, FACRM; Houston Veterans
Affairs Medical Center (153), 2002 Holcombe Boulevard,
Key words: pressure ulcer, pressure ulcer prevalence, spinal Houston, TX 77030; 713-791-1414, ext. 5366; fax: 713-794-
cord injury, veteran. 8040; email: sgarber@bcm.tmc.edu.

433
434

Journal of Rehabilitation Research and Development Vol. 40, No. 5, 2003

methodological problems such as use of different pressure December 31, 1999. Investigators accessed the veterans’
ulcer classification systems, multiple sources of data, and electronic medical records sequentially for pertinent data
varying methods of obtaining data, including direct obser- for the 3 study years. Of the records reviewed, 102 met
vation versus retrospective chart review [4–6]. However, the study inclusion criteria (diagnosis of SCI; treated for
approximately one-third of persons with SCI residing in a pressure ulcer, any stage, at the outpatient clinic or at
the community are reported to have pressure ulcers [7–8]. home; and verifiable data on pressure ulcer stage, dura-
Yarkony and Heinemann reported prevalence rates of tion, and treatment of pressure ulcers). Patients whose
8 percent at the first annual evaluation following rehabili- files did not contain the relevant information (severity of
tation within a Model System facility [9], 9 percent at the ulcer, treatment, outcome of treatment, no pressure ulcers
2-year follow-up, and up to 32 percent at 20 years postdis- documented despite an associated ICD-9 code) were
charge. Carlson et al. reported a 29 percent prevalence rate excluded from the final data set. More than half (113/
during acute care, 3 percent during rehabilitation, and 215) of the charts were excluded primarily because of the
17 percent during follow-up [10]. Despite these reports of extent of incomplete or missing pressure ulcer data in the
pressure ulcer prevalence, no data are readily available on medical record.
the duration of pressure ulcers, number of patient visits
(clinic or home care) for pressure ulcer management, Procedures
severity of the ulcers, or the outcomes of the ulcers in A retrospective review of the electronic medical
terms of healing, nonhealing, or the need for surgical inter- records was used to obtain data on veterans who had
vention. Furthermore, although data are readily available sought treatment for pressure ulcers in the SCI outpatient
on the cost of treating pressure ulcers in acute and long- clinic or from the home care program of the Houston
term care, a dearth of information is available on the costs VAMC. The Decentralized Hospital Computer Program
of treating pressure ulcers in persons with SCI [11–14]. (DHCP), a database that allows access to the medical
Reports indicate that the average cost to heal complex full- record of patients who receive treatment at the Houston
thickness pressure ulcers is estimated to be $70,000 [15]. VAMC, was the primary source of information obtained
This study determined the prevalence, duration, and for this study. The investigators reviewed all notes for each
severity of pressure ulcers in veterans with SCI and iden- patient for the 3 years studied. These included daily notes
tified predictors of (1) visits made to the SCI clinic or of nurses and therapists, physicians, and surgeons and the
home care visits veterans received for pressure ulcer notes associated with medical tests and X-rays. A pressure
treatment, (2) the number of hospital admissions for pres- ulcer tracking form was developed on which the following
sure ulcer treatment, and (3) the number of veterans with information was recorded:
SCI with ulcers that healed, did not heal, or were referred 1. The anatomical location and severity of each ulcer at
for surgery. These data provide baseline information for the initial visit for that ulcer.
assessing etiology, determining compliance with treat- 2. The number of visits for each ulcer.
ment interventions, and identifying factors associated 3. Treatment (topical dressings, debridement, referral
with nonhealing or recurrence of pressure ulcers in veter- for hospital admission or surgery).
ans with SCI. 4. Outcomes in terms of healing, nonhealing, or referral
for surgery.
5. Number of hospitalizations for pressure ulcers.
METHOD 6. Number of days hospitalized for pressure ulcers.
The investigators reviewed the records of all veterans
Sample who had an ICD-9 code of 707.0 (decubitus, any site). Of
Investigators obtained a computer-generated list of these records, 102 electronic charts met the criteria for
all patients on the Houston Veterans Affairs Medical review. The physicians whose patients’ charts were
Center (VAMC) Spinal Cord Registry during the 3 years reviewed gave written permission for the investigators to
studied. From this sampling frame of 553 veterans, review their patients’ medical records. The local Institu-
investigators were able to identify 215 (39%) treated for tional Review Board for Human Subjects Research and
a pressure ulcer (ICD-9 code 707.0 = decubitus, any site) the local Veterans Affairs Research and Development
during the 3-year period January 1, 1997, through Committee approved the study.
 435

GARBER and RINTALA. Pressure ulcers with spinal cord injury

Measures
Stages of Pressure Ulcers
Demographic and SCI-Specific Characteristics
Demographic information included the person’s age, I: Nonblanchable erythema of intact skin, not to be confused
race, caregiver, marital status, and whether or not the per- with reactive hyperemia.
son was diabetic. SCI descriptors included level of injury,
completeness of the injury as determined by the Ameri- II: Partial thickness superficial skin loss involving epidermis
and/or dermis that usually presents as an abrasion, blister,
can Spinal Injury Association (ASIA) Impairment Scale
or shallow crater.
[16], etiology (motor vehicle crash, gunshot wound, fall,
diving, or other), age at onset of SCI, and time since III: Full thickness skin loss with damage or necrosis of subcu-
onset of the SCI. taneous tissue that may extend down to, but not through,
underlying fascia, presenting as a deep crater with or
Prevalence and Characteristics of Pressure Ulcers without undermining of adjacent tissue.

Prevalence. Prevalence is a cross-sectional count of IV: Full thickness skin loss with extensive destruction, tissue
the number of cases (e.g., persons with SCI, pressure necrosis, or damage to muscle, bone, or supporting struc-
ulcers) that occur in a particular population within a spe- tures (e.g., tendon or joint capsule), and may be present
cific period of time [17,18]. In this study, prevalence was with undermining and sinus tracts.
based on the number of persons who visited the Depart-
ment of Veterans Affairs (VA) SCI outpatient clinic or Figure.
received home care for the treatment of a pressure ulcer Stages of pressure ulcers.
in the years 1997, 1998, and/or 1999.
Severity. Severity was defined as the stage (I to IV) and did not have surgery, or was surgically repaired. It is
of a pressure ulcer at its initial assessment [1,17,19]. not known to what extent surgically repaired ulcers
These stages are described in the Figure. If a person had reopened after discharge.
multiple pressure ulcers, the most severe ulcer was the
one that was tracked. Healthcare Utilization
Number of Ulcers. The total number of ulcers for Number of Visits. The number of SCI outpatient
each person who received treatment for pressure ulcers in clinic visits and/or home-health visits for the treatment of
the SCI outpatient clinic and/or from home-health per- pressure ulcers over the 3 years was calculated from
sonnel during the 3-year period was recorded. information in the medical record. SCI clinic notes and
home-health nurse’s notes were used to verify the reason
Duration of Ulcers. Duration was defined as the
for the clinic or home visit. These notes were specific
length of time the veteran had a pressure ulcer during the
with regard to the problem being addressed in either the
3 years of the study.
clinic or during the home visit.
Location of Ulcer. The anatomical locations of pres-
Hospitalization. Whether or not the patient had been
sure ulcers for this population were the sacrum, coccyx,
hospitalized for pressure ulcer treatment during the 3-year
right and left ischia, right and left trochanters, right and
period was noted.
left leg, right and left medial and lateral malleoli, and
right and left foot. Number of Admissions. The number of hospital
admissions for pressure ulcer treatment during the 3 years
Outcome of the study was calculated. Physician or nurses’ admis-
Outcome was defined as the result of pressure ulcer sion notes indicated the reason for the admission. Only
treatment during that year. We determined outcome by those hospital admissions for pressure ulcer treatment
reviewing the medical records from the time the ulcer were counted.
first appeared in the first year of the study to either heal- Days of Hospitalization. The number of days of
ing or the end of the study period. Outcomes were hospitalization across all admissions for pressure ulcer
described as ulcer healed without surgery, did not heal treatment was recorded.
436

Journal of Rehabilitation Research and Development Vol. 40, No. 5, 2003

Analysis were wide ranges of age, age at onset of SCI, and time
We analyzed the data in two phases. Initially, we ana- since onset of SCI (Table 2).
lyzed the data from only 1 year to determine if the study
was feasible in terms of availability of information, as Prevalence and Characteristics of Pressure Ulcers
well as to determine what variables should be used and
which would result in the greatest yield of information. Prevalence
As the study progressed, additional variables were identi- Of the 553 veterans on the roster of the SCI outpatient
fied and incorporated into the data set. clinic at the Houston VAMC, 215 (39%) were diagnosed
Descriptive statistics were obtained for all study vari- with a pressure ulcer (ICD-9 code 707.0 = decubitus, any
ables for the entire sample and for a subset of patients site) during the 3 years studied. Relevant data were
admitted to the hospital for pressure ulcer treatment. Fre- obtained from 102 charts and were the basis for the results
quency tables were constructed for categorical variables presented in this study.
(level and completeness of SCI, etiology, ethnicity, mari-
tal status, caretaker, ulcer severity, diabetes, and treat- Characteristics of Pressure Ulcers
ment outcome). Means, standard deviations, and ranges The stage of the most severe pressure ulcer for each
were calculated for all continuous variables (age, dura- person was recorded. Stage IV pressure ulcers were the
tion of SCI, number of ulcers, number of visits, number most prevalent (Table 1). Patients had an average of
of hospital admissions, and days hospitalized). We per- nearly four (median = 3) ulcers each (Table 2). Duration
formed analyses to identify relationships between demo- of pressure ulcers varied from 1 week to the entire
graphic and SCI-related variables and selected outcome 3 years studied. Pelvic ulcers (sacrum, coccyx, ischial
and use variables. Ranks of variables that were highly tuberosities, and trochanters) accounted for almost two-
skewed (i.e., number of ulcers, visits, hospitalizations, thirds of the worst ulcers reported. Ulcers on the feet
and days hospitalized) were used in the analyses. A chi- were the next most prevalent.
square analysis was performed when both variables were
categorical. For 2 × 2 tables, Fisher’s exact test was used Outcome
to determine significance. Analyses of variance The majority of the ulcers did not heal. Overall, the
(ANOVA) and t-tests were used when one variable was ulcers of 23 of the veterans healed, 54 did not heal, and
continuous and one variable was categorical. For 11 were surgically repaired. Outcome data on 14 veterans
ANOVA, posthoc analyses for multiple comparisons were unknown (Table 1). We found significant relation-
were performed. When both variables were continuous, a ships between outcome and (1) stage of ulcer, (2) rank of
correlational analysis was performed. We collapsed cate- number of ulcers, and (3) rank of number of clinic or
gorical data into fewer categories as necessary to avoid home visits. Ulcers that were more severe were less
very small cell sizes. For statistical significance, p < 0.05 likely to heal and more likely to be referred for surgery
was adopted in all analyses. (Stage II: 60% healed, 40% did not heal, none referred
for surgery; Stage III: 33.3 percent healed, 60 percent did
not heal, 6.7 percent referred for surgery; Stage IV:
RESULTS 9.8 percent healed, 74.5 percent did not heal, 15.7 per-
cent referred for surgery; chi-square = 15.1, p < 0.004).
Characteristics of Study Population Patients whose study ulcer healed had fewer total number
We reviewed a total of 102 charts of veterans with of ulcers than those whose ulcer did not heal (rank of
SCI and pressure ulcers. All the patients whose charts number of ulcers: healed = 41.9, not healed = 61.4 [raw
were reviewed were male. More than half had paraplegia means = 2.5 versus 5.2 ulcers]; overall F = 4.1, p < 0.02;
and over two-thirds had complete injuries (ASIA A). The posthoc test of healed versus not healed, p < 0.02).
most frequent cause of injury was a motor vehicle crash. Patients whose ulcer healed had fewer clinic or home vis-
More than half of the sample was Caucasian. Most were its than those whose ulcer did not heal (rank of number of
not married, and in almost half of the sample, the spouse visits: healed = 40.5, not healed = 64.0 [raw means = 3.3
or significant other was the primary caretaker. The major- versus 8.5 visits]; overall F = 6.4, p < 0.003; posthoc test
ity of the samples were not diabetic (Table 1). There of healed versus not healed, p < 0.002). Those referred
 437

GARBER and RINTALA. Pressure ulcers with spinal cord injury

Table 1. Table 2.
Characteristics of sample: Frequencies and percentages of categorical Means, standard deviations (SDs), and ranges of continuous study
variables for total samples (n = 102). variables for total sample (n = 102).
Variable Number Percent Variable Mean SD Range
SCI Level Age (yr) 51 12.35 25–82
Tetraplegia 45 44 Age at Onset of SCI (yr) 34 13.34 13–69
Paraplegia 57 56
Duration of SCI (yr) 18 11.29 1–54
ASIA Score
A 70 69
Number of Ulcers 4 3.47 1–18
B 15 15 Number of Visits 6 7.63 1–59
C 9 9
D 7 7
Unknown 1 1 for surgery did not differ from either of the other two
SCI Etiology groups (healed or not healed) for either number of ulcers
MVC 48 47
GSW 22 22
or number of visits (rank of ulcers = 51.6 [raw mean =
Fall 9 9 3.7]; rank of visits = 55.0 [raw mean = 6.3]).
Diving 3 3
Other 18 18 Healthcare Utilization
Ethnicity
Caucasian 57 56
Noncausasian 45 44
Number of Visits
Marital Status
The mean number of clinic or home visits made for
Married 39 38 pressure ulcer treatment was more than six visits per per-
Not Married 63 62 son (median = 4) (Table 2). Number of visits was signifi-
Caretaker cantly associated with the severity of the ulcer and the
Spouse/Significant Other 48 47 rank of the number of ulcers. Patients with Stage II ulcers
Attendant 22 22
Lives Alone 23 22
had fewer visits than those with Stage IV ulcers (rank of
Other 9 9 number of visits: Stage II = 39.3, Stage III = 60.0, and
Stage of Worst Ulcer Stage IV = 58.8 [raw means = 3.9 versus 6.3 versus
II 15 15 8.4 visits]; F = 3.32, p < 0.05). Patients who had more
III 16 16 ulcers were likely to have more visits (r = 0.70, p < 0.01).
IV 57 56
Unstaged 14 14
Location of Worst Ulcer
Hospitalization
Sacrum/Coccyx 14 14 Of the 102 veterans included in the study, over half
R/L Ischia 36 35 were admitted to the hospital for pressure ulcer treatment
R/L Trochanters 14 14 at least once during the 3 years of the study (Table 3). Five
R/L Feet/Ankle 26 26
Other 9 9
predictors of hospitalization were identified. These
Unknown 3 3 included (1) etiology (dichotomized as motor vehicle
Outcomes Over 3 Yr crashes versus all other etiologies), (2) ASIA impairment
Healed 23 22 level (dichotomized as A and B versus C and D), (3) sever-
Not healed 54 53 ity of study pressure ulcer (dichotomized as Stages II or III
Surgery 11 11
Unknown 14 14
versus Stage IV), (4) location of ulcer (categorized as isch-
Hospitalized
ium and trochanter versus sacrum and coccyx versus foot
Yes 57 56 and ankle), and (5) outcome (dichotomized as healed or
No 45 44 not healed, omitting those referred for surgery). Persons
Diabetes injured in motor vehicle crashes were less likely than per-
Yes 17 17 sons injured in other ways to be hospitalized for pressure
No 83 81
Unknown 2 2
ulcer treatment (45.8% versus 67.3%, chi-square = 4.7,
ASIA = American Spinal Injury Association (Standards for Neurological and
p < 03). Persons with ASIA impairment levels of A or B
Functional Classification of Spinal Cord Injury) were more likely to be hospitalized than those with levels
MVC = motor vehicle crash, GSW = gun shot wound, R/L = right/left of C or D (61.2% versus 31.3%, chi-square = 4.9,
438

Journal of Rehabilitation Research and Development Vol. 40, No. 5, 2003

Table 3. p < 0.04). Persons with Stage II or III ulcers were less
Characteristics of patients admitted to hospital: Frequencies and per-
likely to be hospitalized than persons with Stage IV ulcers
centages of categorical variables (n = 57).
Variable Number Percent
(35.5% versus 71.9%, chi-square 11.0, p < 0.001). Persons
whose study ulcer was on the ischium or trochanter were
SCI Level
Tetraplegia 23 40 most likely to be hospitalized. Fifty patients had their
Paraplegia 34 60 worst ulcers in either the ischia or trochanters. Of these
ASIA Score patients, 38 (76%) were hospitalized. Persons with ulcers
A 43 75
B 9 16 on the sacrum or coccyx were next most likely to be hospi-
C 3 5 talized. Fourteen patients had their worst ulcers on the
D 2 3
sacrum or coccyx. Of these patients, seven (50%) were
SCI Etiology
MVC 22 39 hospitalized. Twenty-six patients had their worst ulcers on
GSW 15 26 their feet and/or ankles; of these, eight (31%) were hospi-
Fall 8 14
Diving 2 3
talized (chi-square = 15.0, p < 0.001). Patients whose
Other 10 17 ulcers healed without surgery during the 3 years of the
Ethnicity study were less likely to be hospitalized for pressure ulcer
Caucasian 31 54
treatment than those whose ulcer did not heal (30.4% ver-
African-American 23 40
Hispanic 3 5 sus 68.5%, chi-square = 9.6, p < 0.003).
Marital Status
Married 19 33 Number of Admissions
Not Married 38 67
Number of admissions per patient averaged just over
Caretaker
Spouse 27 47 two admissions (median = 2) (Table 4). Almost 30 per-
Attendant 16 28 cent were admitted three or more times. These patients
None 14 25
did not differ significantly from the total sample in demo-
Worst Ulcer
II 3 5 graphic or SCI-specific factors. No predictors of number
III 8 14 of admissions were identified. The characteristics of the
IV 41 72
Unstaged 5 9
hospitalized patients are displayed in Tables 3 and 4.
Location of Worst Ulcer
Sacrum/Coccyx 7 12 Number of Days Hospitalized
R/L Ischia 27 47 The mean number of days hospitalized for pressure
R/L Trochanters 11 19
R/L Feet/Ankle 8 14 ulcer treatment was 150 days (median = 125) (Table 4).
Other 4 7 The only predictor of the rank of number of days hospital-
Total Admissions ized was the rank of number of hospitalizations (r = 0.57,
1 26 46
2 14 25 p < 0.001).
3 11 19
4 2 3
5 3 5 Table 4.
6 1 2 Means, standard deviations (SDs), and ranges of continuous study
Outcomes Over 3 Yr variables for patients admitted to hospital for pressure ulcer treatment
Healed 7 12 (n = 57).
Not Healed 37 65 Variable Mean SD Range
Surgery 10 18
Unknown 3 5 Age (yr) 53 11.73 28–78
ASIA = American Spinal Injury Association (Standards for Neurological and Age at Onset of SCI (yr) 36 12.57 19–70
Functional Classification of Spinal Cord Injury) Duration of SCI (yr) 17 11.25 1–52
MVC = motor vehicle crash
GSW = gun shot wound Number of Admissions 2 1.25 1–6
R/L = right/left Days Hospitalized 150 142.87 2–786
 439

GARBER and RINTALA. Pressure ulcers with spinal cord injury

DISCUSSION ulcer did not heal. However, persons referred for surgery
were no different from individuals in the healed and not
The occurrence of pressure ulcers is among the most healed groups with respect to number of ulcers or number
common long-term secondary medical complications in of visits.
persons with SCI [19]. Methodological problems have Healthcare utilization for treatment of pressure ulcers
limited the reliability of data describing the prevalence of among persons with SCI has not been well studied. In
pressure ulcers in this population. In this retrospective this study, patients received an average of over six clinic
chart-review study, almost 39 percent of the veterans on or home visits for pressure ulcer treatment in a 3-year
the SCI roster at the one VAMC were treated for a pres- period. Furthermore, more than half of the persons whose
sure ulcer during a 3-year time frame. This figure is con- charts were reviewed were admitted to the hospital for
sistent with a number of earlier studies in which pressure ulcer treatment at least once during the 3 years
prevalence ranged from 17 to 33 percent in persons with of the study and almost one-third had been admitted three
SCI residing in the community [10,6]. Rish and col- or more times. The average number of days hospitalized
leagues conducted a 25-year morbidity and mortality was over 150 days. It has been estimated that 1 day of
study of veterans with SCI [20]. They found that the hospitalization on the SCI unit of a VAMC costs $1,000
majority of morbidity problems and the most frequent (personal communication with Administrative Officer on
cause of death were sepsis associated with genitourinary SCI Care Line). Therefore, the estimated average cost per
and pressure ulcer sequelae. Recurrence also is a major patient for hospitalization for pressure ulcer treatment is
problem for veterans with SCI. Niazi and colleagues $150,000. This does not cover the costs associated with
reported that in their sample of 176 veterans with SCI surgical intervention.
and a history of one or more pressure ulcers [21], 35 per- As with the predictors of outcome, several of the pre-
cent experienced a recurrence regardless of whether they dictors of hospitalization for pressure ulcers were
had received surgical or nonsurgical treatment. Higher expected (ASIA A or B impairment level, severity of
recurrence rates occurred in patients who smoked or who study pressure ulcer, and healed versus nonhealed during
had diabetes or cardiovascular disease. In a study of 48 outpatient treatment). In addition, hospitalization
veterans with SCI who had surgery to repair their pres- occurred more frequently for ischial and trochanteric
sure ulcers, postoperative complications were high (40%) ulcers than for ulcers in other anatomical areas. However,
and 79 percent experienced ulcer recurrence or new ulcer one unforeseen finding was that etiology of SCI (motor
development [22]. Schryvers and colleagues conducted a vehicle crash versus all others causes) was associated
retrospective chart review of patients with SCI who had with hospitalization for pressure ulcer treatment. Persons
had pressure ulcer surgery between 1976 and 1996 [23]. injured by motor vehicle crashes were less likely to be
Recurrent admissions were reported for 54 percent of hospitalized; however, this group was younger at the time
their sample. Recurrent severe ulcers were reported for of the study and had had their SCI for a shorter duration
12 percent of the total sample. Factors that were associ- than those injured in other ways. Another unexpected
ated with recurrence included unemployment, low level finding was that neither outcomes nor healthcare utiliza-
of education, drug or alcohol abuse, and poverty. tion was associated with whether or not the patient had
Almost three-fourths (69%) of the sample in the cur- diabetes. Future studies might explain these observations.
rent study had complete SCI (ASIA A), indicating the Inconsistent, inaccurate, and/or missing data in the
absence of sensation and motor function below the neu- medical records were the main limitations of this study.
rological level, including the sacral segments S4 and S5 These factors reflect a lack of continuity in the documen-
[16]. This finding is consistent with the literature that tation process that may have an effect on clinical care and
contends that increased immobilization (decreased motor follow-up. Appropriate documentation has been reported
function) and lack of sensory feedback are the leading to have several major benefits to patients and healthcare
causes of pressure ulcers [24,25]. providers. Among these are (1) tracking an increase or
The predictors of outcome (healed, not healed, or decrease in the number of patient visits, (2) determining
surgery) identified in this study are intuitive. Persons areas of high risk for skin breakdown for the individual
whose worst ulcer healed had less severe ulcers, fewer and other similar veterans, (3) monitoring the effective-
ulcers, and fewer clinic or home visits than those whose ness of treatment over time, and (4) performing cost anal-
440

Journal of Rehabilitation Research and Development Vol. 40, No. 5, 2003

ysis of treatment [4]. French and Ledwell-Sifner of consistent and reliable data in some patient records
developed a flow sheet to help nurses to more thoroughly, reflects possible inaccuracies. A total of 625 visits were
objectively, and consistently assess a patient with pres- made to treat 400 pressure ulcers. At a cost of approxi-
sure ulcers [26]. This approach was more successful in mately $250 per outpatient visit, this amounts to over
preventing and managing pressure ulcers than a standard- $156,000. The average number of hospitalization days for
ized teaching protocol. To attempt to meet standards of pressure ulcer treatment was 150 days, with an average
quality of patient care, Cardi developed the “Clinical cost of $150,000 per patient hospitalized. In this study,
Competency Tool for Documentation of Pressure Ulcer 57 patients were hospitalized for pressure ulcer treatment,
Prevention and Management” [27]. costing approximately $8,550,000 over 3 years. This find-
The National Pressure Ulcer Advisory Panel ing evidently shows that the costs of treating pressure
(NPUAP) believes pressure ulcer incidence should be ulcers in persons with SCI are astounding. Innovative
considered an indicator of quality care in healthcare insti- approaches, especially for individuals with SCI living in
tutions [4,28]. By tracking these indicators of care, clini- the community, are needed to reduce a person’s pressure
cians and administrators will be documenting patient ulcer risk. Consistent and more reliable documentation is
progress, healthcare utilization, and successful treatment one mechanism that may result in better treatment out-
approaches. comes and quality of life for persons with SCI and pres-
The cost-effectiveness of pressure ulcer prevention sure ulcers.
versus treatment has been studied in the acute and long-
term-care patient populations. Aggressive and relatively
low-cost preventive measures were found to reduce sig- REFERENCES
nificantly pressure ulcer incidence in a long-term-care
setting [29]. In a small study that included persons with 1. Bergstrom N, Allman RM, Carlson CE, Eaglstein W,
Frantz RA, Garber SL, et al. Pressure ulcers in adults: pre-
paraplegia, implementation of the Agency for Health
diction and prevention. Clinical Practice Guideline Number
Care Policy and Research’s (AHCPR’s) clinical practice
3, Agency for Health Care Policy and Research, Publica-
guideline on prevention of pressure ulcers in adults tion No. 92-0047. Rockville, MD: Agency for Health Care
resulted in lower costs than current practice at the facility Policy and Research, Public Health Service, U.S. Depart-
in which they were being treated [1,30]. Similar studies ment of Health and Human Services; May 1992.
designed to assess the effect of pressure ulcer prevention 2. Krouskop TA, Noble PC, Garber SL, Spencer WA. The
programs in persons with SCI may reduce the incidence effectiveness of preventive management in reducing the
and recurrence of pressure ulcers, improve care, and occurrence of pressure sores. J Rehabil Res Dev 1983;20:
reduce the staggering costs of treatment. 74–83.
A potential indirect outcome of consistent documenta- 3. Lloyd EE, Baker F. An examination of variables in spinal
cord injury patients with pressure sores. SCI Nursing 1986;
tion will be the development and evaluation of innovative
3(2):19–22.
and cost-effective outpatient and home-health programs
4. National Pressure Ulcer Advisory Panel. Pressure ulcer
that identify persons at highest risk for pressure ulcers and prevalence, cost and risk assessment: Consensus develop-
that prevent and/or reduce the frequency of pressure ulcer ment conference statement. Decubitus 1989;2:24–28.
occurrence or recurrence in veterans with SCI. 5. Gawron CL. Risk factors for and prevalence of pressure
ulcers among hospitalized patients. J Wound Ostomy Con-
tinence Nurs 1994;21:232–40.
CONCLUSIONS 6. Gallagher SM. Outcomes in clinical practice: Pressure
ulcer prevalence and incidence studies. Ostomy/Wound
Manag 1997;43(1):28–40.
This study illustrates the magnitude of the pressure
7. Fuhrer MJ, Garber SL, Rintala DH, Clearman R, Hart KA.
ulcer problem among veterans with SCI living in the com- Pressure ulcers in community-resident persons with spinal
munity. In this study, 39 percent of all patients on the SCI cord injury: prevalence and risk factors. Arch Phys Med
roster at the Houston VAMC were reported to have visited Rehabil 1993;74:1172–77.
the clinic or received home care for pressure ulcers within 8. Vidal J, Sarrias M. An analysis of the diverse factors con-
a 3-year time frame. Although these data are consistent cerned with the development of pressure sores in the spinal
with prevalence rates reported in other literature, the lack cord-injured patient. Paraplegia 1991;29:261–77.
 441

GARBER and RINTALA. Pressure ulcers with spinal cord injury

9. Yarkony GM, Heinemann AW. Pressure ulcers. In: Stover 21. Niazi ZB, Salzberg CA, Byrne DW, Viehbeck M. Recur-
SL, DeLisa JA, Whiteneck GG, editors. Spinal cord injury: rence of initial pressure ulcer in persons with spinal cord
Clinical outcomes from the model systems. Gaithersburg injuries. Adv Wound Care 1997;10:38–42.
(MD): Aspen Publishing; 1995. p. 100–19. 22. Goodman CM, Cohen V, Armenta A, Thornby J, Netscher
10. Carlson CE, King RB, Kirk PM, Temple R, Heinemann A. DT. Evaluation of results and treatment variables for pres-
Incidence and correlates of pressure ulcer development sure ulcers in 48 veteran spinal cord-injured patients. Ann
after spinal cord injury. J Rehabil Nurs Res 1992;1:34–40. Plast Surg 1999;43:572–74.
11. Frantz RA, Gardner S, Harvey P, Specht J. The cost of 23. Schryvers OI, Stranc MF, Nance PW. Surgical treatment of
treating pressure ulcers in a long-term care facility. Decubi- pressure ulcers: 20-year experience. Arch Phys Med Reha-
tus 1991;4:37–45. bil 2000;81:1556–62.
12. Alterescu V. The financial costs of inpatient pressure ulcers
24. Curry K, Casady L. The relationship between extended
to an acute care facility. Decubitus 1989;2:14–23.
periods of immobility and decubitus ulcer formation in the
13. Javitz HS, Ward MM, Martens L. Major costs associated
acutely spinal cord injured individual. J Neurosc Nurs
with pressure sores. J Wound Care 1998;7:286–90.
1992;24(4):185–89.
14. Xakellis GC, Frantz R, Lewis A. Cost of pressure ulcer
prevention in long-term care. J Am Geriatr Soc 1995;43: 25. Mawson AR, Siddiqui FH, Connolly BJ, Sharp J, Summer
496–501. WR, Biundo JJ. Sacral transcutaneous oxygen tension lev-
15. Braun J, Silvetti A, Xakellis G. What really works for pres- els in the spinal cord injured: risk factors for pressure
sure sores. Patient Care 1992;26:63–70. ulcers? Arch Phys Med Rehabil 1993;74:745–51.
16. American Spinal Injury Association. Int Stand Neurol 26. French ET, Ledwell-Sifner K. A method for consistent doc-
Funct Classif Spinal Cord Inj 1996;6:27. umentation of pressure sores. Rehabil Nurs 1991;16: 204–7.
17. Garber SL, Biddle AK, Click CN, Cowell JF, Gregorio- 27. Cardi M. A tool to document the competence of clinicians
Torres TL, Hansen NK, et al. Pressure ulcer prevention and to prevent and manage pressure ulcers. Decubitus 1992;5:
treatment following spinal cord injury: A clinical practice 46–49.
guideline for health-care professionals. Consortium for 28. Cuddigan J, Frantz RA. Pressure ulcer research: pressure
Spinal Cord Medicine Clinical Practice Guidelines, Para- ulcer treatment. A monograph from the National Pressure
lyzed Veterans of America; 2000. Ulcer Advisory Panel. Adv Wound Care 1998;11:294–300.
18. Langemo DK, Hanson D, Cathcart-Silberberg T, Hunter S. 29. Xakellis GC, Frantz RA. The cost-effectiveness of inter-
Prevalence of pressure ulcers in five patient care settings. J ventions for preventing pressure ulcers. J Am Board Fam
Enterostom Ther 1990;17:187–92. Pract 1996;9:79–85.
19. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ.
30. Hu TW, Stotts NA, Fogarty TE, Bergstrom N. Cost analy-
Long-term medical complications after traumatic spinal
sis for guideline implementation in prevention and early
cord injury: a regional model systems analysis. Arch Phys
treatment of pressure ulcers. Decubitus 1993;6:42–46.
Med Rehabil 1999;80:1402–10.
20. Rish BL, Dilustro JF, Salazar AM, Schwab KA, Brown
HR. Spinal cord injury: a 25-year morbidity and mortality Submitted for publication June 25, 2002. Accepted in
study. Mil Med 1997;162:141–48. revised form March 27, 2003.