Skin breakdown in children and high-frequency oscillatory ventilation

Skin breakdown in children and high-frequency oscillatory ventilation

1565 Skin Breakdown in Children and High-Frequency Oscillatory Ventilation Jeffrey E. Schmidt, MD, Richard J. Berens, MD, Mary B. Zollo, RN, Margaret...

647KB Sizes 0 Downloads 44 Views

1565

Skin Breakdown in Children and High-Frequency Oscillatory Ventilation Jeffrey E. Schmidt, MD, Richard J. Berens, MD, Mary B. Zollo, RN, Margaret C&l G.M. Weigle, MD ABSTRACT. Schmidt JE, Berens RJ, Zollo MB, Weisner M, Weigle CGM. Skin breakdown in children and high-frequency oscillatory ventilation. Arch Phys Med Rehabil 1998;79: 1565-9. Objective: To investigate the relationship of high-frequency oscillatory ventilation (HFOV) to skin breakdown on the scalp and ears in mechanically ventilated children. Study Design: Retrospective cohort study of 32 patients supported with HFOV paired with 32 patients supported with conventional mechanical ventilation (CV) in a pediatric intensive care unit (PICU). Results: By univariate analysis, more HFOV patients had skin breakdown than did the CV patients (53% vs 12.5%, p = .OOl); HFOV patients also had greater severity of illness (Pediatric Risk of Mortality scores), higher mortality, and longer durations of neuromuscular blockade, low systolic blood pressure, and time exposed to risk. Life table analysis demonstrated no difference in the rate of skin breakdown between HFOV and CV patients. Multifactorial analysis showed that only PICU time at risk was a risk factor for skin breakdown. Conclusions: HFOV was not an independent risk factor for the development of skin breakdown. PICU time at risk was the sole risk factor for the development of skin breakdown in all mechanically ventilated patients in the PICU. 0 1998 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation

H

IGH-FREQUENCY oscillatory ventilation (HFOV) is a mode of mechanical ventilation that has only recently become available as a treatment for acute respiratory failure in neonates and children. l-5The goals of HFOV are to improve gas exchange and reduce traumatic injury caused by mechanical ventilation.6-9During conventional mechanical ventilation (CV), gas exchange is biphasic and is dependent on tidal volume gas exchange. HFOV differs from CV in that gas exchange occurs continuously; there is no tidal volume gas exchange. Airway and alveolar opening is maintained at a minimum opening pressure (mean airway pressure) and gas exchange occurs primarily through diffusion across concentration gradients and is augmented by the oscillatory gas flow produced by the ventilator’s piston.‘O In this way, the trauma produced by peak inspiratory pressuresand volumes seen with CV is reduced with HFOV while maintaining adequate gas exchange. Although From the Department of Pediatrics, Critical Care Section, Children’s Hospital of Wisconsin and Medical College of Wisconsin, Milwaukee, WI. Submitted for publication March 6, 1998. Accepted April 21, 1998. Supported in part by the Dr. Elaine C. Kohler Fund for Healing, Milwaukee, WI. No commercial party having a direct fimncial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon my organization with which the authors are associated. Reprint requests to Jeffrey E. Schmidt, MD, Critical Care Section, St. Jude Children’s Research Hospital, 332 North Lauderdale, Memphis, TN 38105. @ 1998 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation 0003-9993/98/7912-4918$3.00/O

Weisner, RN,

HFOV is now an accepted primary mode of ventilation, the use of HFOV during our study period was limited to rescue therapy. Rescue therapy was defined as therapy for patients with acute respiratory failure who were unresponsive to CV and included HFOV, inhaled nitric oxide, surfactant therapy, and extracorporeal membrane oxygenation (ECMO). Several months after HFOV was introduced at our institution, we perceived that a higher percentage (probably the majority) of patients ventilated by HFOV experienced skin breakdown on the scalp and ears (hereafter referred to as skin breakdown) than patients ventilated with CV. We suspected that this was caused by a decreased frequency of repositioning these patients’ heads because of the rigidity of the ventilator circuit and fear of extubation. We also suspected that the oscillatory force of the ventilator exposed these patients to increased friction and shear. By anecdotal reports, these perceptions and suspicions were shared by other institutions. Approximately 26% of patients in our unit experience some degree of altered skin integrity, 3 1% of whom experience actual disruption of the skin membrane.‘i In a previous study” we reported that children with alterations in skin integrity tend to be mechanically ventilated longer, have longer pediatric intensive care unit (PICU) stays, and incur higher overall hospital costs than children who do not experience alteration in skin integrity. The only factors, however, that were significantly associated with risk for skin breakdown were white race and severity of illness, as measured by the Pediatric Risk of Mortality (PRISM) score.11l12It has been well documented in both the pediatric and adult literature that skin breakdown significantly increases the intensity of hospital care, the length of hospital stay, and overall hospital costs. In addition, skin breakdown often results in additional morbidity and permanent cjc-ing.ll,l3-22 The occurrence of skin breakdown has a complex pathophysiology, and multiple risk factors have been identified. Risk factors for skin breakdown can be extrinsic or intrinsic. Extrinsic factors include the duration and degree of pressure, shear, friction, and moisture. Intrinsic factors include hypotension, hypoperfusion, malnutrition, infection, anemia, and neurologic deficits (sensory loss, altered mental status, immobility). *4-16,23,24 Although 95% of pressure ulcers develop over the bony prominences of the lower extremities in adults, most pressureulcers in children occur on the occiput, proportionately the largest and heaviest bony prominence in children.17 No risk factors for skin breakdown have been identified that are unique to the pediatric population, including HFOV. This study investigated the effect of HFOV and other previously defined risk factors on the development of skin breakdown in children. MATERIALS AND METHODS We obtained approval from our hospital’s Research and Publications Committee/Human Rights Review Board. Arch

Phys

Med

Rehabil

Vol 79, December

1998

1566

SKIN

BREAKDOWN

AND

MECHANICAL

Patient Population The study was conducted in a tertiary care children’s hospital, which, at the time, supported a 14-bed PICU. The hospital’s referral area includes approximately 2,500,OOOpeople in metropolitan and rural settings. Approximately 1,000 patients are admitted to the PICU each year and about 70% of these require mechanical ventilation. Over the time period covered by this study, fewer than 0.5% of all ventilated patients were ventilated with HFOV. All patients ventilated with HFOV were initially ventilated with CV. The decision to institute HFOV was based on the inability to achieve adequate gas exchange on toxic conventional ventilator settings (ie, when patients qualified for ECMO support). HFOV was instituted as the first line of rescue therapy in an attempt to avoid using ECMO. On occasion, HFOV was also used to transition patients off ECMO support. Other modes of rescue therapy, specifically nitric oxide and partial liquid ventilation, were not in use during the period of this study. Although there were no established criteria for instituting HFOV, by definition of rescue therapy, the HFOV patients during our study period met ECMO criteria with an oxygenation index (01) of >25, where 01 = (mean airway pressure in cm H20) X (fraction of inspired oxygen) X lOO/(arterial oxygen partial pressure in mmHg). We identified all children who were ventilated with HFOVa for longer than 24 hours in the PICU at our institution from 12/l/91 to 6130193. A control patient was identified for each HFOV patient as the next patient admitted to the PICU treated with CV for longer than 24 hours. There were 64 patients in the study. Study Design We reviewed the medical records of all patients. Data were collected on each patient during the time at risk, which we defined as the time from PICU admission until either the time of skin breakdown or until the time of PICU discharge if there was no skin breakdown. The standard of nursing training and practice at this institution included repositioning immobile patients every 2 hours (with exceptions, such as patient instability, noted in the nursing record) and daily total body skin surveillance to identify areas of skin disruption. This information was part of a standardized record in the nursing chart and was available in the permanent medical record. We recorded the number of times the patient’s head was repositioned each day, the duration of sedative use, neuromuscular blockade, systolic blood pressure, vasoconstrictor use (dopamine, epinephrine, and norepinephrine), and time on HFOV and CV. PRISMI scores on admission to the PICU, lowest serum albumin values, and highest and lowest white blood cell counts during time at risk were recorded. In addition, we made note of each patient’s major diagnoses, the use of ECMO support, and postoperative status. For patients who developed skin breakdown, additional data collected from the medical record included the number and severity of lesions based on the scale by Maklebust.l* Interrater reliability could not be measured for this particular study; therefore, data analysis was performed based on only the presence or absence of skin breakdown. Data Analysis Data were analyzed using Stata, version 5.0.b Given the 32 patients in each of our groups, and our previous study showing 26% of PICU patients developing altered skin integrity, this study had 89% power to detect the (clinically estimated) doubling of the skin breakdown rate in patients supported with HFOV. Univariate analysis was used to examine the differences Arch

Phys

Med

Rehabil

Vol 79, December

1998

VENTILATION,

Schmidt

in the occurrence of skin breakdown, putative risk factors, and general characteristics between patients supported with HFOV and patients supported by CV, and between the patient groups defined by the presence or absence of skin breakdown. Differences between groups were assessedusing the unpaired student t test for continuous variables and the Wilcoxon rank sum test for ordinal variables. Differences in dichotomous variables were analyzed using the x2 test or, where expected frequencies were less than 5, the Fisher exact test. The log rank test was used to assessgroup differences while controlling for the effects of time. Statistical significance was defined by ap < .05. Multivariate analysis using the Cox proportional hazard mode125 was performed to identify the relative importance of factors potentially associated with skin breakdown. This model requires the assumption that factors entered into the model do not vary in importance over time. Race, age, admission PRISM score, and cardiac surgical status on admission are such non-time-varying factors, which were entered into the model ECMO use and HFOV use are time-varying covariates. As such, they were coded as absent (0) or present (1) and then entered into the model with a different value for each day at risk as: variable = (code for absent/present) . (day - [average days at risk for all patients studied]). RESULTS Univariate analysis. The characteristics of the HFOV and CV groups are compared in table 1. Only diagnoses that differed significantly between ventilator groups are listed in table 1. We also found that three of the HFOV patients with skin breakdown developed breakdown while on CV before HFOV support. These patients were included with the HFOV group for analysis, according to the rule of intention to treat. Table 2 compares the characteristics of the two groups of patients defined by the presence or absence of skin breakdown. Table 1: Patient Characteristics and Comparison Factors for Skin Breakdown Between HFOV (Mean f SEMI

of Potential Risk and CV Patients

HFOV (n = 32)

CV (n = 32)

pValue*

16232 7.4 i 8.1

36 i 72 10.8 i 13.8

19/13 2618

17115 2319

19

3

NS .OOOl

8

0

.005

12

24

,002

2

20

.OOOl

4.5 2 3.7 19 It 10

3.7 ? 2.3 11 %I0

NS ,002

15.5 i 11.7 .82 ir .55 Il.62 30

4.9 + 4.6 .I6 2 .40 .5 i 1.7

.OOOl .OOOl .044

17

4

,001

Demographics Age ho) Weight (kg) Sex (male/female) Race (white/nonwhite) Death Diagnosis Congenital diaphragmatic herniaiECM0 Surgery including cardiovascular surgeryt Congenital heart cardiovascular Variables

disease surgery

NS NS NS

after

of interest

Head repositioning PRISM score

(times/d)

PICU time at risk (d)S Neuromuscular blockade(d) Low blood pressure (h/d)§ Skin breakdown

* p > .05 listed as NS, not significant. t Postoperative. * PICU time at riskwas defined as the time from PICU admission until the time skin breakdown first occurred, or until the time of PICU discharge if no skin breakdown occurred. § Low blood pressure was defined as the PRISM scoring cutoff point, ie, a systolic pressure of 466mmHg (
SKIN

Table

2: Patient Characteristics Factors for Skin Breakdown Breakdown and Patients With

BREAKDOWN

AND

MECHANICAL

and Comparison of Potential Risk Between Patients With Skin No Skin Breakdown (Mean f SEMI Intact Skin (I?= 43)

Breakdown (n = 21)

p Value*

33 + 66 23/20 10.4 t 13

13 k 23 1318 6.5 ? 4.7

NS NS NS

33110

1615

NS

Demographics

Age (mo) Sex (male/female) Weight (kg) Race (white/nonwhite) Death Diagnosis Congenital

13

9

NS

2

6

,012

22

14

NS

18

4

NS

diaphragmatic

hernia/ECMO Surgery, including cular surgeryt Congenital heart cardiovascular Variables of interest Head repositioning

cardiovasdisease surgeryt

after

3.8 k 3.2

4.7 k 2.9

NS

PRISM score PICU time at risk (d)S

142 12 8.7 2 9.9

17 210 13.4 2 10.5

NS .OOOl

Neuromuscular blockade (d) Low blood pressure (h/d)§

.34 + .52 0.3 k 1.1

.78 i .60 0.9 + 2.5

.0068 NS

15

17

,001

HFOV

(times/day)

* p > .05 listed as NS, not significant. t Postoperative. $ PICU time at riskwas defined as the time from PICU admission until the time skin breakdown first occurred, or until the time of PICU discharge if no skin breakdown occurred. 5 Low blood pressure was defined as the PRISM scoring cutoff values, ie, a systolic pressure of <66mmHg (<12mo) or ~76 (212mo).

All eight patients with congenital diaphragmatic hernia received ECMO support and are therefore listed together in tables 1 and 2. Life table analysis. By log rank test, neither type of mechanical ventilation used for support of respiratory function was independently associated with skin breakdown when the effects of time at risk were controlled for, nor was any other measured variable (including ECMO, surgical status in general, cardiac surgical status in specific, white race, PRISM score of >lO, and age younger than 12 months). These findings are illustrated by the examples of three paired Kaplan Meier survival curves seen in figure 1. Such Kaplan Meier plots were inspected for each variable entered into the Cox proportional hazard model. Each downward step in any curve represents a change in the group’s probability of skin breakdown based on one or more actual occurrences of skin breakdown at that time at risk. These curves also graphically demonstrate the validity of the assumption (for the Cox proportional hazard model to follow) that the hazards associated with these factors are proportional over time; that is, the curves do not cross (true of the variables not included in the figure for clarity’s sake). The only exceptions to this are the crossings of the HFOV and CV curves, and of the white race and nonwhite race curves, each of which represents a single patient who developed skin breakdown at 17 and 3 1 days, respectively (note: the 95% confidence bands could not be calculated at these points). By log rank analysis, p values were >.05 for all comparisons. Multivariate analysis. The Cox proportional hazards model was used with time-varying and non-time-varying covariates. HFOV use and ECMO use were treated as time-varying covariates as defined above, and codes for surgical status, cardiac surgical status, white race, PRISM score of >lO (a moderately high level of severity), and age younger than 12

VENTILATION,

1567

Schmidt

months were treated as simple proportional (non-time-varying) factors. Factors were added all at once, and dropped if colinearity was detected by Stata. This model demonstrated none of the putative risk factors to be significantly associated with the development of skin breakdown. Baseline risk for skin breakdown for all patients studied is shown in figure 2. DISCUSSION We found a higher incidence of skin breakdown in patients ventilated with HFOV than in those ventilated with CV (53% vs 12..5%), as predicted. However, multivariate and life table analysis determined that PICU time at risk, not type of ventilation, was the most important risk factor for development of skin breakdown in the group of patients studied. Pressureulcers occur when pressure against the skin exceeds the pressure in skin capillaries, resulting in local ischemic hypoxia and then necrosis. Removal of cytotoxic waste is inhibited, further promoting tissue death. Kosiak19 showed that both high continuous pressures over a short duration as well as low pressures over a long duration could result in irreversible tissue damage. A pressure as little as 70mmHg continuously applied over 2 hours results in permanent tissue injury but a pressure as high as 240mmHg relieved every 5 minutes over 2 hours produces no changes. In adults, capillary perfusion pressure is about 30mmHg and pressures five to ten times greater have been measured over bony prominences in recumbent patients.16,26In children, this occurs over the occiput.r7 Studies identifying risk factors for skin breakdown in the pediatric population are limited but consistent with findings in adults. Patients with neurologic impairment (ie, spina bifida, cerebral palsy) have increased risk for pressure ulcers with increased degree of neurologic deficit.20,27 Both adults and children undergoing cardiopulmonary bypass are at greater risk for the development of occipital pressure ulcers and alopecia.17x21Age, type of congenital heart defect, length of time intubated, and length of PICLJ stay have been identified as risk factors for the development of occipital pressure ulcersI Gershan and Esterly22 described occipital pressure ulcers that resulted in scarring alopecia in hypoxemic, hypoperfused neonates. HFOV of humans is confined to the neonatal and pediatric population. Patients may be exposed to additional extrinsic risk factors (shear and frictional forces from oscillation) as well as any of the previously identified intrinsic risk factors. Sedation and neuromuscular blockade to maintain optimal positioning to accommodate the rigid oscillator tubing may add to the risk of ulcer formation. In clinical practice, the use of HFOV has often been reserved for patients with a greater degree of hypoxemia and severity of illness. During the time of this study, HFOV was reserved for rescue therapy. This may explain why we found significant differences in patient diagnoses by univariate analysis when comparing HFOV and CV groups. In our study, all eight patients with congenital diaphragmatic hernias also received HFOV and required ECMO support (table 1). Six of these eight patients developed skin breakdown (table 2). Although a diagnosis of congenital diaphragmatic hernia and support with ECMO was significantly associated with skin breakdown by univariate analysis, life table analysis and multivariate analysis demonstrated that these were not independent risk factors for skin breakdown. The primary objective of this study was to determine if HFOV was an independent risk factor for the development of skin breakdown. We found that HFOV was not a risk factor for skin breakdown and we found that the most important factor was time at risk, which supports the work by Escher-Neidig17 and is consistent with Kosiak’s time-pressure studiesi This Arch

Phys

Med

Rehabil

Vol 79, December

1998

1568

SKIN

1

.5 0 1

BREAKDOWN

1 7-

I

MECHANICAL

VENTILATION,

Schmidt

cv

I

7 -l

.5

AND

f PRISM I 10

0 1

-3

White

Nonwhite

--x

-T-l

.5 0 0

20

40

Fig 1. Examples of KaplanMeier survival curves comparing the risk of skin breakdown on the scalp and ears versus time at risk for aatients arouoed by putative r&k fact&. %he heavy lines represent the calculated risk, the light lines represent the 95% confidence limits for that risk.

0

Days at Risk finding is strengthened by the fact that the three HFOV patients who developed skin breakdown before HFOV were still analyzed with the HFOV group. We speculate that any predictor of a long PICU stay would also serve as a predictor of skin breakdown. Interestingly, admission PRISM scores were not significantly higher in the patients with skin breakdown, which is consistent with Sachdeva’s finding that the PRISM score on admission is not a strong predictor of length of stay (unpublished data). This retrospective study had several shortcomings. First, although we believe the medical record documentation to be accurate, it is probably not as accurate as documentation

Y

I

I

collected prospectively. Second, the lack of a standardized method for skin assessment in the pediatric population may have produced variability in the recorded descriptions of skin breakdown. This prohibited us from making comparisons with regard to severity of skin breakdown and limited our analysis to only the absence or presence of skin breakdown. The assumption that mild skin changes are induced by the same risk factors as would induce frank ulceration may not be valid. Finally, our small sample size may not have allowed for the accurate testing of six potential risk factors in the multivariate (Cox proportional hazard) analysis. The common approach is to ensure that there are at least 10 subjects in the study for each variable included in the multivariate model, a criterion that we barely met. For example, we did not find race or cardiac surgery with cardiopulmonary bypass to be factors associated with skin breakdown, as other studies have found.11,28The sample size and diversity of patients also makes the calculation of overall baseline risk of skin breakdown over time (fig 2) somewhat risky to interpret, especially beyond the first 2 weeks of time at risk, when only 16 patients remained in the sample. The sample size of this study does provide sufficient power for us to conclude that HFOV is not a risk factor for the development of skin breakdown based on life table analysis and Cox proportional hazard testing. References

0-l

I 0

I 40

2b

Days At Risk Fig 2. Baseline patients studied.

risk

Arch

Rehabil

Phys

Med

of skin

breakdown

Vol 79, December

versus

1998

time

at risk,

for

Study Group. Randomizedstudy of high-frequencyoscillatory ventilation in infants with severerespiratory distresssyndrome.J Pediatr1993;122:609-19. 2. Arnold JH, Truog RD, ThompsonJE,FacklerJC. High-frequency oscillatory ventilation in pediatric respiratoryfailure. Crit Care Med 1993;21:272-8. 3. Arnold JH, Hanson JH, Toro-FigueroLO, GutierrezJA, Berens RJ, Anglin DL. Prospective,randomized comparison of high1. HIFI

I

all

SKIN

4. 5.

6. 7.

8.

9.

10. 11. 12. 13. 14. 15.

BREAKDOWN

AND

MECHANICAL

frequency oscillatory ventilation and conventional mechanical ventilation in pediatric respiratory failure. Crit Care Med 1994;22: 1530-9. Johnston D, Hochmann M, Timms B. High-frequency oscillatory ventilation: initial experience in 22 patients. J Paediatr Child Health 1995;4:297-301. Clark RH, Gerstmann DR, Null DM Jr, deLemos RA. Prospective randomized comparison of high-frequency oscillatory and conventional ventilation in respiratory distress syndrome. Pediatrics 1992;89:5-12. Froese AB, Bryan AC. High-frequency ventilation, Am Rev Respir Dis 1987;135:1363-74. Hamilton PP, Onayemi A, Smyth JA, Gillan JE, Cutz E, Froese AB. et al. Comnarison of conventional and high-freauencv oscillatory ventilation: oxygenation and lung pathology. i Appi Physiol 1983;55:131-8. McCulloch PR, Forkert PG, Froese AB. Lung volume maintenance prevents lung injury during high-frequency oscillatory ventilation in surfactant-deficient rabbits. Am Rev Respir Dis 1988;137: 1185-92. Bond DM, Roese AB. Volume recruitment maneuvers are less deleterious than persistent low lung volumes in the atelectasisprone rabbit lung during high-frequency oscillation. Crit Care Med 1993;21:402-12. Pedley TJ, Corieri P, Kamm RD, Grotberg JB, Hydon PE, Schroter RC. Gas flow and mixing in the airways. Crit Care Med 1994;22: S24-36. Zollo MB, Gostisha ML, Berens RJ, Schmidt JE, Weigle CGM. Altered skin integrity in children admitted to a pediatric intensive care unit. J Nurs Care Qua1 1997;11:62-7. Pollack MM, Ruttimann UE, Getson PR. Pediatric risk of mortality (PRISM) score. Crit Care Med 1988;16:1110-6. Quigley SM, Curley MAQ. Skin integrity in the pediatric population: preventing and managing pressure ulcers. J Sot Pediatr Nurs 1997;1:7-18. Brook 1. Microbiologic studies of decubitus ulcers in children. J Pediatr Surg 1991;26:207-9. Galpin J, Chow A, Bayers A, Guze L. Sepsis associated with decubitus ulcers. Am J Med 1976;61:346-50.

VENTILATION,

1569

Schmidt

16. Herman H, Rothman K. Prevention care and treatment of pressure (decubitus) ulcers in ICU patients. J Intens Care Med 1989;4: 117-23. 17. Escher-Neidig J, Kleiber C, Oppliger A. Risk factors associated with pressure ulcers in the pediatric patient following open heart surgery. Prog Cardiovasc Nurs 1989;4:99-106. 18. Maklebust J. Pressure ulcers: etiology and prevention. Nurs Clin North Am 1987;22:359-77. 19. Kosiak M. Etiology of decubitus ulcers. Arch Phys Med Rehabil 1961;42:19-29. 20. Harris M, Banta J. Cost of skin care in the myelomeningocele population. J Pediatr Orthop 1990;10:355-61. 21. Lawson N, Mills N, Ochsner J. Occipital alopecia following cardiopulmonary bypass. J Thorac Cardiovasc Surg 1976;71: 342-7. 22. Gershan L, Esterly N. Scarring alopecia in neonates as a consequence of hypoxaemia-hypoperfusion. Arch Dis Child 1993;68: 591-3. 23. Schubert V. Hypotension as a risk factor for the development of pressure sores in elderly subjects. Age Ageing 1991;20:255-61. 24. Shea J. Pressure sores: classification and management. Clin Orthop 1975;112:89-100. 25. Fisher L, van Belle G. Biostatistics: a methodology for the health sciences. New York: John Wiley & Sons, Inc.; 1993. 26. Rondorf-Klym K, Langham D. Relationship between body weight, body position, support surface, and tissue interface pressure at the sacrum. Decubitus 1993;6:22-30. 27. Okamato G, Lamers J, Shurtleff D. Skin breakdown in patients with myelomeningocele. Arch Phys Med Rehabil 1983;64:20-3. 28. Jericho M, Ryan P, Carvalho M, Bukvich J. Pressure ulcer risk factors in an ICU population. Am J Crit Care 1995;4:361-7. Suppliers a. Sensormedics 3100A; Sensormedics, 22705 Savi Ranch Parkway, Yorba Linda, CA 92887. b. Stata Corporation, 707 University Drive East, College Station, TX 77840.

Arch

Phys

Med

Rehabil

Vol 79, December

1998