Adult Contractures in Burn Injury: a Burn Model System National Database Study

Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021

Adult Contractures in Burn InjuryA Burn Model System National Database Study

Jeremy Goverman, MD, Katie Mathews, BA, Richard Goldstein, PhD, Radha Holavanahalli, PhD, Karen Kowalske, MD, Peter Esselman, MD, Nicole Gibran, MD, Oscar Suman, PhD, David Herndon, MD, Colleen M. Ryan, MD, Jeffrey C. Schneider, MD ORIGINAL ARTICLE

Adult Contractures in Burn Injury: A Burn Model System National Database Study

Jeremy Goverman, MD,*†‡§ Katie Mathews, BA,† Richard Goldstein, PhD,† Radha Holavanahalli, PhD,║ Karen Kowalske, MD,║ Peter Esselman, MD,¶

Nicole Gibran, MD,¶ Oscar Suman, PhD,# David Herndon, MD,# Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 Colleen M. Ryan, MD,*†‡§ Jeffrey C. Schneider, MD*†‡§

As the overall survival rate for burn injury has improved, increased emphasis is placed on postburn morbidity and the optimization of functional and cosmetic outcomes. One major cause of morbidity and functional deficits is that of joint contractures. The true incidence of postburn contractures and their associated risk factors remains unknown. This study examines the incidence and severity of contractures in a large, multicenter, burn population. The associated risk factors for the development of contractures are determined. Data from the National Institute on Disability and Rehabilitation Research Burn Model System database, for adult burn survivors from 1994 to 2003, were analyzed. Demographic and medical data were collected on each subject. The primary outcome measures included the presence of contractures, number of contractures per patient, and severity of contractures at each of nine locations (shoulder, elbow, hip, knee, ankle, wrist, neck, lumbar spine, and thoracic spine) at time of hospital discharge. Regression analysis was performed to determine predictors of the presence, severity, and numbers of contractures, with P < .05 used for statistical significance. Of the 1865 study patients, 620 (33%) developed at least 1 contracture J Burn Care Res at hospital discharge. Among those with at least one contracture, the mean is three (3.38) contractures per person. The shoulder was the most frequently contracted joint (23.0%), followed by the elbow (19.9%), wrist (17.3%), ankle (13.6%), and knee (13.4%). Most contractures were mild (47.2%) or moderate (32.9%) in severity. Statistically significant predictors of contracture development were male sex, black race, Hispanic ethnicity, medical problems, neuropathy, TBSA grafted, and TBSA burned. Predictors of the severity of contracture included male sex, black race, medical problems, neuropathy, TBSA grafted, and TBSA burned. Predictors of the number of contractures included male sex, medical problems, flash burn, neuropathy, TBSA burned, and TBSA grafted. Similar to a previous single-center study on postburn contractures, approximately one third of the patients with an eligible burn injury requiring autografting developed a contracture at hospital discharge. It is likely that these contractures develop despite early therapeutic interventions such as positioning and splinting; therefore, the challenge to the burn community remains, to identify new and better prevention strategies. (J Burn Care Res 2017;38:e328–e336)

From the *Surgical Services, Sumner Redstone Burn Center, Massachusetts General Hospital, Boston; †Department of Physi- cal Medicine and Rehabilitation, Spaulding Rehabilitation Supported by the National Institute on Disability and Hospital, Boston, Massachusetts; ‡Harvard Medical School, Rehabilitation Research grant H133A120034. Boston, Massachusetts; §Shriners Hospitals for Children®-Boston, Supplemental digital content is available for this article. Direct ║ Massachusetts; Department of Physical Medicine and Rehabili- URL citations appear in the printed text and are provided in tation, University of Texas Southwestern Medical Center, Dallas; the HTML and PDF versions of this article on the journal’s ¶Department of Surgery, University of Washington Medicine Web site. Regional Burn Center, University of Washington, Seattle; and Address correspondence to Jeffrey C. Schneider, MD, Surgical #University of Texas Medical Branch, Shriners Hospitals for Services, Sumner Redstone Burn Center, Massachusetts General Children, Galveston. Hospital; Department of Physical Medicine and Rehabilitation, Copyright © 2016 by the American Burn Association Spaulding Rehabilitation Hospital; Harvard Medical School; 1559-047X/2016 and Shriners Hospitals for Children®-Boston, Massachusetts. DOI: 10.1097/BCR.0000000000000380 Email: [email protected].

Contractures represent a major source of morbidity One of the first and largest publications demon- for patients recovering from burn injury. Postburn strating the epidemiology of burn contractures was contractures often have a devastating impact on the a retrospective review of 681 patients performed quality of life and one’s ability to perform activities by Dobbs and Curreri in 1972. They found a 28% of daily living.1 Although prevention remains the incidence of contractures, with the hand, elbow, and most rational approach, a significant percentage of shoulder being the most frequently involved joints. burn survivors will develop functionally limiting, TBSA and burn depth were noted to be associated postburn contractures even at the best burn centers with contracture development.16 The first review and despite aggressive, early physical and occupa- of prospectively collected data on burn contrac-

tional therapy. Surgical correction with skin grating, tures was performed in 2006 by Schneider et al in Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 adjacent tissue rearrangement, or local and distant which they reviewed single-center data of 985 burn flaps is typically required, yet carries a significant patients. The shoulder (38%) was identified as the rate of recurrence with often undesirable cosmetic most frequently contracted joint, followed by the outcomes. Furthermore, postburn contractures and elbow (34%) and knee (22%). Of the study popu- the need for surgical treatment, which often require lation, 39% demonstrated at least one contracture additional postoperative physical and occupational with a mean of three contractures per person. Most therapy, increase the financial burden of burn care contractures were noted to be mild (60%) or mod- and limit patients from returning to work. erate (32%). Statistically significant predictors of Contractures are defined as an inability to per- contracture development were length of stay, TBSA form full range of motion (ROM) of a joint.2 Burns grafted, and TBSA burned.17 Predictors of contrac- damage the skin and often the underlying soft tis- ture severity included graft size, amputation, and sue, muscle, and bone; therefore, burn-injured inhalation injury. Limitations of the study were patients are at risk for developing joint contractures. that it only examined four joints (shoulder, elbow, Postburn contractures ultimately result from a com- hip, and knee) and included data from only one bination of wound contracture, primarily caused burn center. Herein, we build on Schneider et al’s by fibroblasts and a natural physiological response review by examining nine major joints in a larger to wound healing, as well as from scar contracture, patient population and from five major burn cen- primarily caused by myofibroblasts.3 Some measure ters. Furthermore, this is the first and largest study of burn scar contracture is expected and may be on postburn contractures to examine prospectively considered within normal range; however, exces- collected data from a national, multicenter, burn sive contracture of scar, such as hypertrophic scar- outcomes database. We examine the incidence and ring (HTS), is considered pathological. Patients severity of contractures and determine the associ- who develop contractures face the possibility of ated risk factors for their development. additional medical problems, including functional deficits, and interference with wound and graft healing. Functionally, contractures can inhibit nor- METHODS mal ambulation and transfers, fine motor tasks, and Prospectively collected data from the National activities of daily living such as eating, bathing, and 4–8 Institute on Disability and Rehabilitation Research grooming. (NIDRR) Burn Model System (BMS) database, for A combination of factors likely contribute to post- adult burn survivors from 1994 to 2003, were ana- burn contracture formation: 1) injury-related fac- lyzed. The NIDRR BMS database comprises patients tors include the depth, extent, cause, and location of who meet at least one of the following current inclu- burn; 2) patient-related factors include genetic, race, sion criteria: skin color, age, sex, nutritional status, and compliance with therapy; and 3) treatment-related factors 1. Burn injury greater than or equal to 10% including the type and timing of wound closure, the TBSA, which required surgery for at least some wound bed, and prevention strategies utilized.3,9–12 portion of wound closure (defined as autograft- The cellular and biological basis of wound, scar, ing); age: 65 years and older. and skin graft contracture, as well as HTS, has been 2. Burn injury greater than or equal to 20% TBSA, extensively researched and well described.13,14 Fur- which required surgery for at least some por- thermore, a prolonged stay in an intensive care unit tion of wound closure; age: 0 to 64 years. (ICU), even in the absence of a burn injury, has been 3. Electrical high voltage/lightning injury, which shown to be associated with a 34% incidence of func- required surgery for at least some portion of tionally limiting joint contracture.15 wound closure.

4. Burn injury of any size to critical area(s): face purposes of this analysis, a limitation in the ROM in and/or hands and/or feet and/or genitals, at least one plane of motion at a specified joint was which required surgery for at least some por- considered to be a contracture at that joint. Further- tion of wound closure. more, if more than one muscle action was limited at a joint, the severity of the most impaired muscle During the 10 years of this study, and the action at that joint was considered to represent the 20 years of the NIDRR BMS database, minor modi- severity of contracture at that joint. Data were also fications have been made to the inclusion/exclusion collected for the presence of ectropion, microstomia, criteria. These modifications can be found at http:// and nasolabial contractures. These contractures con- burndata.washington.edu/standard-operating-pro- tributed to analyses of frequency and number, but Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 cedures, and the complete detailed inclusion and not to severity, of contracture. exclusion criteria have been previously described.18 Primary outcome measures included the presence, severity, and number of contractures. Demographic Statistical Analysis (age, sex, and ethnicity [white, black, Hispanic, and The frequency and severity of contractures at each other]) and medical data were collected. If demo- joint, and number of contractures per patient at graphic data cannot be collected through patient hospital discharge, were calculated. Logistic regres- report, they are collected from the medical record. sion analysis was used to determine predictors of the Medical data included length of hospital stay, length presence of contractures. Ordered logistic regression of ICU stay, presence of concomitant medical prob- was used to determine the severity of contractures. lems (defined as medical problems that might alter Negative binomial regression, one form of a mul- the course of recovery from the burn, such as dia- tivariate analysis, was used to determine predictors betes, chronic obstructive pulmonary disease, heart of the number of contractures. Negative binomial disease, asthma), cause of burn , inhalation injury, regression was used because the variance assump- neuropathy, heterotopic ossification, amputation as tion underlying Poisson regression was violated. the result of the burn injury, and TBSA burned and The negative binomial regression model adjusts for grafted (recorded as a whole number percentile). this overdispersion. The potential predictors were Neuropathy required determination by a physician the demographic and medical data (as detailed pre- and included upper or lower extremity sensory and/ viously). Violations of statistical assumptions and or motor abnormalities or generalized peripheral goodness-of-fit test were analyzed. A P < .05 was neuropathy. We chose to report on all contractures used for statistical significance. In addition, the level recorded in the NIDRR database, except for that of of multicollinearity was evaluated via examination the hand. This included musculoskeletal contractures of condition numbers and variance decomposition (eg, joint) and nonjoint contractures. These areas proportions and determined to be non-significant. of interest included the shoulder, elbow, hip, knee, Further details with respect to the statistical logic ankle, wrist, neck, lumbar spine, and thoracic spine. of model building for this analysis can be found in Lumbar and thoracic data were only collected after Appendix 1, Supplemental Digital Content 1, at 1998. At the time of discharge from the hospital, http://links.lww.com/BCR/A44. specified areas were examined for a total of 15 stud- ied sites per subject (neck, lumbar, and thoracic were RESULTS unilateral). The subjects’ active ROM at each joint was measured using a goniometer and inclinometer The demographic and medical data of the study with a standardized technique.19 Multiple planes of population are presented in Table 2. One thousand motion (ie, flexion/extension) were investigated eight hundred sixty-five adult patients were included at each joint. The specific methodology for ROM in the analysis. Average age was 41.7 years, and aver- measurements at each joint is detailed in the Model age TBSA was 18.3%. The frequency and severity System for Burn Injury Rehabilitation National Data- of limitations by joint muscle action is presented in base Data Dictionary.20 Joint muscle action in each Table 3. Shoulder flexion (n = 438) and abduction plane is assigned a normal ROM based on physical (n = 272), knee flexion (n = 275), and elbow flexion examination conventions.19,21 As in a similar study by (n = 275) and extension (n = 299) were the most Schneider et al,17 each impaired joint muscle action frequent joint muscle action limitations. Using the is assigned a severity rating. Such ratings are deter- methodology described previously, contracture fre- mined by dividing the normal ROM value equally in quency and severity at each joint was tabulated and thirds (mild, moderate, and severe; Table 1). For the is presented in Table 4. The shoulder (23.0%) was

Table 1. Range of motion severity ratings by joint muscle action (degrees)

Contracture Severity

Joint Muscle Action Mild Moderate Severe Expected19,21

Shoulder Flexion 120 to 180 60 to 119 <60 180 Extension 32 to 50 16 to 31 <16 50 Abduction 120 to 180 60 to 119 <60 180 Adduction 32 to 50 16 to 31 <16 50 Hip Flexion 67 to 100 34 to 66 <34 100 Extension 20 to 30 10 to 19 <10 30 Abduction 26 to 40 13 to 25 <13 40 Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 Adduction 13 to 20 7 to 12 <7 20 Elbow Flexion 93 to 140 46 to 92 <46 140 Extension 140 to 93 −46 to 92 >−46 0 Pronation 53 to 80 26 to 52 <26 80 Supination 53 to 80 34 to 66 <26 80 Knee Flexion 100 to 150 50 to 99 <50 150 Extension −150 to 100 −99 to 50 >−50 0 Wrist Flexion 40 to 59 20 to 39 <20 60 Extension 40 to 59 20 to 39 <20 60 Radial DV 13 to 19 6 to 12 <6 20 Ulnar DV 20 to 29 10 to 19 <10 30 Ankle Dorsiflexion 13 to 19 6 to 12 <6 20 Plantarflexion 26 to 39 13 to 25 <13 40 Inversion 20 to 29 10 to 19 <10 30 Neck Extension 50 to 74 25 to 49 <25 75 Rotation (L and R) 53 to 80 26 to 52 <26 80 Lateral flexion (L and R) 30 to 44 15 to 29 <15 45 Lumbar Forward flexion 53 to 79 26 to 52 <26 60 Extension 40 to 59 20 to 39 <20 0 Lateral flexion 16 to 24 8 to 15 <8 25 Thoracic Forward flexion 16 to 24 8 to 15 <8 60 Rotation (L and R) 20 to 29 10 to 19 <10 30

DV, deviation; L, left; R, right.

the most frequently contracted joint, followed by the and TBSA burned, and female sex was a protective elbow (19.9%), wrist (17.3%), ankle (13.6%), and factor. Predictors for the number of contractures knee (13.4%). Most contractures were mild (47.2%) include male sex, medical problems, flash burn, or moderate (32.9%) in severity. Among the study population, 620 patients (33%) demonstrated at least Table 2. Demographic and medical characteristics of the 1 contracture at hospital discharge. In total, 2097 study population joints were contracted resulting in an average of Total number of study patients 1865 3 (3.38) contractures per person (among those Male (%) 77.5 with at least 1 contracture). The frequency of con- Age at injury (yr), mean (SD) 41.7 (15.5) Ethnicity (%) tractures is presented in Table 5. The frequency of Caucasian 69.0 microstomia (0.27%), ectropion (0.86%), and naso- Black 15.0 labial contractures (0.16%) were also determined. A Hispanic 10.0 multivariate regression analysis was used to identify Other 5.4 predictors of postburn contracture development Length of stay (d), mean (SD) 25.0 (24.0) (Table 6), contracture severity (Table 7), and num- Inhalation injury (%) 10.2 Percent TBSA burned, mean (SD) 18.3 (16.3) ber of contractures (Table 8). Statistically significant Causative factor (%) predictors of contracture development include male Fire/flame 59.5 sex, black race, Hispanic ethnicity, medical prob- Electrical 7.2 lems, neuropathy, TBSA grafted, and TBSA burned, Flash 5.3 whereas female sex was a protective factor. Predictors Scald 9.3 Grease 7.1 of the severity of contracture include male sex, black Other 11.3 race, medical problems, neuropathy, TBSA grafted,

Table 3. Severity and frequency of range of motion limitation by joint muscle action (degrees)

Severity

Joint Muscle Action Mild Moderate Severe Total

Shoulder (%) Flexion 248 (57) 168 (38) 22 (5) 438 Extension 15 (27) 10 (18) 31 (55) 56 Abduction 132 (49) 127 (47) 13 (5) 272 Adduction 3 (8) 2 (5) 32 (86) 37 Hip (%) Flexion 63 (68) 26 (28) 4 (4) 93

Extension 10 (18) 16 (28) 31 (54) 57 Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 Abduction 20 (40) 27 (54) 3 (6) 50 Adduction 3 (11) 9 (33) 15 (56) 27 Elbow (%) Flexion 205 (75) 57 (21) 13 (5) 275 Extension 270 (90) 29 (10) 0 (0) 299 Pronation 44 (61) 13 (18) 15 (21) 72 Supination 92 (54) 46 (27) 33 (19) 171 Knee (%) Flexion 156 (57) 104 (38) 15 (5) 275 Extension 108 (98) 0 (0) 2 (2) 110 Wrist (%) Flexion 69 (59) 39 (34) 8 (7) 116 Extension 82 (62) 39 (30) 11 (8) 132 Radial deviation 15 (58) 6 (23) 5 (19) 26 Ulnar deviation 21 (54) 16 (41) 2 (5) 39 Ankle (%) Dorsiflexion 34 (17) 56 (28) 111 (55) 201 Plantarflexion 48 (31) 24 (15) 83 (54) 155 Inversion 7 (21) 17 (52) 9 (27) 33 Eversion 7 (21) 19 (58) 7 (21) 33 Neck (%) Forward flexion 85 (89) 7 (7) 3 (3) 95 Right rotation 8 (32) 13 (52) 4 (16) 25 Left rotation 9 (36) 16 (64) 0 (0) 25 Extension 5 (13) 21 (55) 12 (32) 38 Right lateral flexion 21 (68) 6 (19) 4 (13) 31 Left lateral flexion 19 (61) 7 (23) 5 (16) 31 Lumbar (%) Forward flexion 1 (100) 0 (0) 0 (0) 1 Right lateral flexion 1 (50) 1 (50) 0 (0) 2 Left lateral flexion 2 (67) 1 (33) 0 (0) 3 Thoracic (%) Forward flexion 6 (75) 1 (13) 1 (13) 8 Right rotation 0 (0) 0 (0) 0 (0) 0 Left rotation 0 (0) 0 (0) 0 (0) 0 Total (%) 1809 (56) 923 (29) 494 (15) 3226

neuropathy, TBSA grafted, and TBSA burned, and (ABA) with early interventions of aggressive occu- again female sex was a protective factor. There were pational and physical therapy. Yet despite this fact, no violations of statistical assumptions or goodness- 33% of patients in this review developed at least one of-fit test for these analyses. contracture. There are a multitude of reasons for this relatively DISCUSSION high incidence of contracture. Regional burn centers often treat the most seriously injured, those This work represents one of the largest published with the most critical injuries, larger TBSA burned, review of burn contractures and the first to utilize and grafted, and are therefore at a statistically higher the NIDRR BMS national burn outcomes database. risk of developing contractures. In fact, inclusion We confirm previous studies indicating that post- in the NIDRR BMS database necessitates a rela- burn contractures remain a common morbidity after tively extensive injury that requires skin grafting. major burn injury. Burn centers contributing to the Larger burns are also more likely to cross multiple NIDRR BMS database are large, regional, U.S. burn joints and require multiple surgical interventions, centers verified by the American Burn Association which in turn require a period of postoperative

Table 4. Contracture severity and frequency by joint (degrees)

Contracture Severity

Joint Mild Moderate Severe Total (%)

Shoulder 231 190 61 482 (23.0) Elbow 263 106 47 416 (19.9) Hip 49 42 43 134 (6.4) Knee 160 104 17 281 (13.4) Ankle 29 72 185 286 (13.6)

Wrist 171 146 45 362 (17.3) Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 Neck 78 27 19 124 (5.9) Lumbar 3 1 0 4 (0.2) Thoracic 6 1 1 8 (0.4) Total (%) 988 (47.2) 687 (32.9) 419 (19.9) 2097

Each impaired joint muscle action is assigned a severity rating determined by dividing the normal range of motion value equally in thirds (mild, moderate, and severe).

immobilization for appropriate healing; all of which appears relatively low. Finally, it is always important has been associated with contracture development. to consider that the overall incidence of contracture A large burn injury typically requires an extended will be directly related to the number of joints/ length of ICU stay, which, alone, is associated with locations examined. a significant incidence of contracture.15 Clavet et Similar to Schneider et al’s 2006 single-cen- al examined all patients admitted to an academic ter study, the majority of contractures were mild hospital ICU for more than 2 weeks (nonburn (47.2%) or moderate (32.9%); however, the inci- injured patients) and found a 39% incidence of dence of severe contractures in the current review is contracture, of which 34% were functionally limit- more than twice than previously noted (20 vs 8%). ing. Although there are methodological differences This may in part be because of the inclusion of the between Clavet et al’s study and the current study ankle, in the current study, which had the highest with respect to the number of locations examined incidence of severe contractures (65% of ankle con- (10 vs 15) as well as the definitions of contracture tractures were severe). Also similar to the study by and functional limitations, the results are similar. Schneider et al was the fact that the shoulder and Therefore, the influence of a prolonged ICU stay elbow remained the two most commonly involved (>14 days) alone is likely to contribute significantly joints with shoulder flexion and abduction the most to the development of contractures in our patient common muscle actions for limited ROM. However, population. Interestingly, in our study, severe ankle the inclusion of wrist, ankle, and knee, in the current contractures were notable more common than mild study, were all more commonly involved than that and moderate contractures. Further analysis dem- of the hip. onstrated that patients with ankle contractures are Given the relatively high incidence of postburn more likely to have had an ICU stay (P < .001). contractures, it seems reasonable to ask whether our Furthermore, considering that the average length current prevention strategies are effective. The fre- of ICU stay in this study, for those patients who quency of contractures in the current study (33%) spent at least 1 day in the ICU, was 17.1 (SD, is similar to that found in the review by Dobbs and 20.8), our incidence of ankle contractures actually Curreri (28%), more than 40 years before, similar to that of Schneider et al’s study (39%) and even similar to Clavet et al’s study (39%) of non-burn patients Table 5. Contracture frequency (N = 1865) with prolonged ICU stay. It should be noted that Patients with at least one contracture (%) 620 (33.2) direct comparison between these studies, as well as Total number of contractures 2097 other published studies on contracture rates, is not Mean number of contractures per person if at least one 3.38 valid given differences in the study populations, num- Patients with only one contracture (%) 141 (7.6) ber of joints examined, and definitions of contracture Patients with two contractures (%) 147 (7.9) and functional deficit. However, similar contracture Patients with three contractures (%) 91 (4.9) rates in this study compared with historical studies Patients with four contractures (%) 92 (4.9) may yield two indications. First, the similar contrac- Patients with more than four contractures (%) 149 (8.0) ture rates may be because of less mortality due to

Table 6. Logistic regression analysis of predictors of the presence of contractures

95% Confidence Variable Odds Ratio SE t P > |t| Interval

Female 0.784 0.061 −3.15 .002 0.674 to 0.912 Black 2.172 0.812 2.08 .038 1.045 to 4.518 Hispanic 1.461 0.247 2.24 .025 1.049 to 2.034 Other ethnicity (not white, black, or Hispanic) 0.820 0.221 −0.74 .460 0.484 to 1.389 Medical problems 1.377 0.145 3.04 .002 1.121 to 1.692 Neuropathy 1.697 0.201 4.46 <.001 1.345 to 2.141

TBSA burned 1.060 0.015 4.14 <.001 1.032 to 1.091 Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 TBSA grafted 1.067 0.015 4.67 <.001 1.038 to 1.097

improved medical and surgical management. Because The findings of the study by Kolmus et al corre- more patients are surviving larger burns, greater pos- spond to a review of the literature on static splint- sible morbidity is a result. Thus, a relatively stable ing for burns, by Schouten et al, in which no strong contracture rate (33% vs Dobbs and Curreri’s 28%) supporting evidence was identified, and we are could actually indicate improvements in prevention reminded of the pathophysiology of scarring/con- techniques, given that survival of major burns is more tracture, ie, that mechanical tension is a strong stim- likely. In addition, when a 33% rate of contracture ulus for hypertrophic and abnormal scar formation. in this current study is compared with more recent Static splinting results in scar tension. Tensile forces studies, such as Schneider et al’s 39% finding that have been shown to induce transdifferentiation of evaluated a fewer number of joints, improvement in fibroblasts into myofibroblasts and to down-regulate contracture prevention could again be a possibility fibroblast proapoptotic genes, favoring the abnormal given that more areas were included in this study and collagen deposition seen in HTS.13,14 For patients a less or stable rate is noted. Second, current preven- with burn-associated joint contractures, Godleski et tion strategies are mostly composed of static splinting al24 demonstrate significant improvements in ROM, and active/passive ROM exercises. The role of static with an intensive stretching protocol, in the absence splinting and the specifics of technique, however, of splinting. However, this study had a patient popu- remain controversial.22 Very few high-quality trials lation of only nine. Prolonged durations of mechani- on splinting for prevention of burn contractures are cal stretching have also been shown effective in available. The only randomized controlled trial on contracture reduction as described in the systematic splinting of adult burns, written by Kolmus et al,23 review by Furia et al.25 failed to show a benefit of static axillary splinting for Despite the fact that our knowledge of specific axillary burns. Adherence to splinting was low, 77% at evidence-based strategies for contracture preven- week 1 and 16% at week 12, as the majority of non- tion are still limited, the identification of high- compliant patients reported that their arm was mov- risk patients is still a clinically useful endeavor. In ing normally. In fact both groups showed significant this study, multivariate analyses were performed improvements in shoulder ROM, and the additional to determine predictive factors associated with benefit of splinting was not evident, suggesting that the presence, severity, and number of contractures the tailored exercise regimen utilized in the study was postburn injury. Statistically significant predic- mostly responsible for such improvements. tors included male sex, the presence of medical

Table 7. Ordered logistic analysis of predictors of the severity of contractures

Variable Odds Ratio SE t P > |t| 95% Confidence Interval

Female 0.769 0.762 −2.65 .008 0.633 to 0.934 Black 2.179 0.698 2.43 .015 1.163 to 4.083 Hispanic 1.288 0.175 1.86 .063 0.986 to 1.683 Other ethnicity (not white, black, or Hispanic) 0.823 0.190 −0.84 .401 0.523 to 1.296 Medical problems 1.356 0.674 6.14 <.001 1.230 to 1.495 Neuropathy 1.584 0.133 5.47 <.001 1.343 to 1.868 TBSA burned 1.075 0.140 5.51 <.001 1.047 to 1.102 TBSA grafted 1.036 0.008 4.87 <.001 1.022 to 1.051

Table 8. Negative binomial regression of predictors of the number of contractures

Variable IRR SE t P > |t| 95% Confidence Interval

Female 0.773 0.041 −4.87 <.001 0.697 to 0.857 Length of hospital stay 1.008 0.007 1.05 .293 0.993 to 1.022 Presence of medical problems 1.281 0.070 4.52 <.001 1.151 to 1.427 Electricity 1.303 0.261 1.32 .186 0.880 to 1.928 Flash 1.128 0.043 3.18 .001 1.047 to 1.215 Scald 1.264 0.433 0.68 .494 0.646 to 2.472 Grease 1.221 0.128 1.90 .057 0.994 to 1.500 Other ethnicity (not white, black, or Hispanic) 0.908 0.052 −1.70 .089 0.813 to 1.015 Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 Neuropathy 1.284 0.136 2.36 .018 1.044 to 1.580 TBSA burned 1.049 0.013 3.93 <.001 1.024 to 1.075 TBSA grafted 1.073 0.008 9.38 <.001 1.057 to 1.089

problems, neuropathy, TBSA grafted, and TBSA We did not explore measures of quality of life or burned. Therefore, male patients with concomitant functional outcome. Future research may examine medical issues and more extensive, deeper burns the effect of contracture severity on quality of life and are most at risk for developing more contractures function, as well as the correlation between contrac- and contractures of worse severity. These findings tures and return to work outcomes. Furthermore, are consistent with Schneider et al’s 2006 study of details of physical and occupational therapy sessions, burn contractures, in which TBSA grafted was also splinting techniques, time to full active ROM, time found to be a predictor of presence, severity, and of immobilization, and time to complete wound clo- number of contractures. Schneider et al’s publica- sure were not recorded in the NIDRR database and tion found other factors, typically associated with are all valuable data points for future analyses. All severe burn injury, to be significant predictors as study centers were ABA-verified burn centers and, as well, such as inhalation injury, length of stay, and such, have physical and occupational therapy services amputation. The current study, however, did not in place as required for this verification.26 Details on corroborate these specific predictors. Nevertheless scar management techniques were also not available, by identifying such factors, health care professionals, but should be included in future studies, as should and researchers can best target prevention strategies other potential risk factors for HTS, such as race.27,28 in our search for the most effective approach. Furthermore, determining the cumulative effects of There are a number of limitations of this study. multiple contractures on limb function is a future First, these data were collected from 1994 to 2003. direction for further analysis. Finally, Richard et al29 To the extent that burn care has changed over this has expanded our understanding of burn location as time period, the results of this study may not accu- it relates to contracture formation using the concept rately reflect current burn practice. As stated pre- of cutaneous functional units. For example, burns viously, the criteria for inclusion into the NIDRR of the abdomen can contribute to contractures at BMS database select those with more severe burns the shoulder as a result of skin recruitment and the and, therefore, may not be representative of all burn availability of skin in certain areas of the body. The patients and burn centers. In addition, and in con- extent to which one can attribute burns of differ- currence with a previous trial, we chose to evaluate ent anatomical locations to contracture development the rate of contracture at one time point—hospital at specific locations, however, is an underdeveloped discharge. Postburn contractures may improve, or concept that deserves further investigation. Future worsen, over time depending on a number of factors investigations of postburn contracture may benefit including therapy, patient compliance, and develop- from the utilization of this concept of cutaneous ment of HTS and tension forces across a joint. In functional units. addition, the presence of various confounding factors at the time of discharge, such as pain, weakness, CONCLUSION and influence of medication, may have influenced ROM measurements. Therefore, it is possible that Approximately one third of burn patients admitted the contracture rates reported here do not necessary to one of the BMS ABA-verified burn centers with represent the rate of postburn contracture found a severe burn injury exhibited a contracture at dis- over time. charge despite aggressive physical and occupational

therapeutic interventions. Similar to previous find- 15. Clavet H, Hébert PC, Fergusson D, Doucette S, Trudel G. ings, risk factors for the development of postburn Joint contracture following prolonged stay in the intensive care unit. CMAJ 2008;178:691–7. contractures include male sex, the presence of medi- 16. Dobbs ER, Curreri PW. Burns: analysis of results of physical cal problems, neuropathy, TBSA grafted, and TBSA therapy in 681 patients. J Trauma 1972;12:242–8. burned. 17. Schneider JC, Holavanahalli R, Helm P, Goldstein R, Kowalske K. Contractures in burn injury: defining the prob- lem. J Burn Care Res 2006;27:508–14. 18. Klein MB, Lezotte DL, Fauerbach JA, et al. The National REFERENCES Institute on Disability and Rehabilitation Research Burn Model System database: a tool for the multicenter study of 1. Leblebici B, Adam M, Bağiş S, et al. Quality of life after the outcome of burn injury. J Burn Care Res 2007;28:84–96. burn injury: the impact of joint contracture. J Burn Care Res

19. Norkin CC, White DJ. Measurement of joint motion: a Downloaded from https://academic.oup.com/jbcr/article/38/1/e328/4568945 by guest on 28 September 2021 2006;27:864–8. guide to goniometry. Philadelphia: FA Davis; 1985. 2. Mosby’s dictionary. 6th ed. St. Louis: Mosby Inc.; 2002. 20. Model system for burn injury rehabilitation national data- 3. Kwan P, Hori K, Ding J, Tredget EE. Scar and contracture: biological principles. Hand Clin 2009;25:511–28. base data dictionary. Version 8.0. 2001. Available at: http:// 4. Palmieri TL, Petuskey K, Bagley A, Takashiba S, Greenhalgh burndata.washington.edu/sites/burndata/files/files/ DG, Rab GT. Alterations in functional movement after axil- BMS_DataDictionary_ArchivedVariables.pdf. Accessed May lary burn scar contracture: a motion analysis study. J Burn 8, 2006. Care Rehabil 2003;24:104–8. 21. Magee DJ. Orthopedic physical assessment. Philadelphia: 5. Kowalske KJ, Voege JR, Cromes GF Jr., et al. The relation- Saunders; 2002. ship between upper extremity contractures and functional 22. Schouten HJ, Nieuwenhuis MK, van Zuijlen PP. A review outcome after burn injury (abstract). Proc Am Burn Assoc on static splinting therapy to prevent burn scar contracture: 1996;28:55. do clinical and experimental data warrant its clinical applica- 6. Helm P, Herndon DN, Delateur B. Restoration of function. tion? Burns 2012;38:19–25. J Burn Care Res 2007;28:611–4. 23. Kolmus AM, Holland AE, Byrne MJ, Cleland HJ. The ef- 7. Engrav LH, Covey MH, Dutcher KD, Heimbach DM, fects of splinting on shoulder function in adult burns. Burns Walkinshaw MD, Marvin JA. Impairment, time out of 2012;38:638–44. school, and time off from work after burns. Plast Reconstr 24. Godleski M, Oeffling A, Bruflat AK, Craig E, Weitzenkamp Surg 1987;79:927–34. D, Lindberg G. Treating burn-associated joint contracture: 8. Bayat A, McGrouther DA, Ferguson MW. Skin scarring. results of an inpatient rehabilitation stretching protocol. BMJ 2003;326:88–92. J Burn Care Res 2013;34:420–6. 9. Jones T, McDonald S, Deitch EA. Effect of graft bed on 25. Furia JP, Willis FB, Shanmugam R, Curran SA. Systematic long-term functional results of extremity skin grafts. J Burn review of contracture reduction in the lower extremity with Care Rehabil 1988;9:72–4. dynamic splinting. Adv Ther 2013;30:763–70. 10. Tredget EE, Levi B, Donelan MB. Biology and principles of 26. American College of Surgeons. Chapter 14: Guidelines for scar management and burn reconstruction. Surg Clin North the operation of burn centers. In: Resources for optimal care Am 2014;94:793–815. of the injured patient 2014. 6th ed. Committee on Trauma, 11. Ladak A, Tredget EE. Pathophysiology and management of American College of Surgeons. the burn scar. Clin Plast Surg 2009;36:661–74. 27. Sood RF, Hocking AM, Muffley LA, et al. Race and melano- 12. Harrison CA, MacNeil S. The mechanism of skin graft con- cortin 1 receptor polymorphism R163Q are associated with traction: an update on current research and potential future post-burn hypertrophic scarring: a prospective cohort study. therapies. Burns 2008;34:153–63. J Invest Dermatol 2015;135:2394–401. 13. Armour A, Scott PG, Tredget EE. Cellular and molecular 28. Butzelaar L, Soykan EA, Galindo Garre F, et al. Going into pathology of HTS: basis for treatment. Wound Repair Regen surgery: risk factors for hypertrophic scarring. Wound Repair 2007;15(Suppl 1):S6–17. Regen 2015;23:531–7. 14. Engrav LH, Garner WL, Tredget EE. Hypertrophic scar, 29. Richard RL, Lester ME, Miller SF, et al. Identification of wound contraction and hyper-hypopigmentation. J Burn cutaneous functional units related to burn scar contracture Care Res 2007;28:593–7. development. J Burn Care Res 2009;30:625–31.