USF Libraries
USF Digital Collections

Wheelchair positioning and pulmonary function in children with cerebral palsy


Material Information

Wheelchair positioning and pulmonary function in children with cerebral palsy
Physical Description:
Barks, Lee
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:


Subjects / Keywords:
Postural management
Dissertations, Academic -- Nursing -- Doctoral -- USF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


ABSTRACT: Background: In children with cerebral palsy (CP), poor trunk control fosters spinal deformity, pulmonary compromise (Canet, et al., 1998), increased health risks, and costs of long-term care (Braddock, 2001). Evidence links posture and pulmonary function, but influence of wheelchair parameters on pulmonary mechanics is unknown. Objectives: 1) Determine relative contribution of five wheelchair configuration parameters to improvement in pulmonary mechanics--total airway resistance (RAW), tidal volume, minute ventilation (MV), and deadspace to tidal volume ratio; 2) Describe recruitment and retention of school-aged children with CP; and 3) Discuss response of the participants to the protocol. Method: This within-subjects, descriptive study employed a sample of 8 school-aged children with CP and flexible spines who could not sit alone.In a single session, participants experienced five seating parameters manipulated in a Prairie wheelchair simulator: 1) left and right upper extremity supports; 2) left and right lateral trunk supports; 3) secured, level, derotated pelvis; 4) tilt in space; and 5) all four parameters. The Viasys Jaeger Impulse Oscillometry System and Respironics Non Invasive Cardiac Output monitor (NICO) measured the dependent variable, pulmonary mechanics, via Hans Rudolph facemasks. Spasticity (by Modified Ashworth Scale), patient characteristics, and medications were recorded. A process log captured participant recruitment and retention challenges and response to protocol. Results: Recruitment was challenging; retention was 50%. For this sample, despite lack of power, both RAW and MV improved with upper extremity and lateral trunk supports. Highest RAW was seen with total absence and total presence of the parameters, and secured, level pelvis.The data collection protocol was feasible for 50% of participants, none of whom could execute conventional measurement. Facemask and seating simulator acceptability were 75 %, improving with participant verbal communication ability. The facemask seal was vulnerable to tilted positioning; 75% of participants became fatigued. RAW measures differed from manufacturer's directions but were reliable. Conclusions: The Prairie seating simulator, Jaeger IOS, Respironics NICO, and Hans Rudolph facemasks effectively measured pulmonary mechanics as a function of wheelchair seating parameters in this sample. Upper extremity and lateral trunk supports most greatly reduced RAW, maintaining MV. Verbal children tolerated the procedure best.
Dissertation (Ph.D.)--University of South Florida, 2007.
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Lee Barks.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 115 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001920228
oclc - 187971015
usfldc doi - E14-SFE0002107
usfldc handle - e14.2107
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001920228
003 fts
005 20080107125131.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080107s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002107
1 100
Barks, Lee.
0 245
Wheelchair positioning and pulmonary function in children with cerebral palsy
h [electronic resource] /
by Lee Barks.
[Tampa, Fla.] :
b University of South Florida,
ABSTRACT: Background: In children with cerebral palsy (CP), poor trunk control fosters spinal deformity, pulmonary compromise (Canet, et al., 1998), increased health risks, and costs of long-term care (Braddock, 2001). Evidence links posture and pulmonary function, but influence of wheelchair parameters on pulmonary mechanics is unknown. Objectives: 1) Determine relative contribution of five wheelchair configuration parameters to improvement in pulmonary mechanics--total airway resistance (RAW), tidal volume, minute ventilation (MV), and deadspace to tidal volume ratio; 2) Describe recruitment and retention of school-aged children with CP; and 3) Discuss response of the participants to the protocol. Method: This within-subjects, descriptive study employed a sample of 8 school-aged children with CP and flexible spines who could not sit alone.In a single session, participants experienced five seating parameters manipulated in a Prairie wheelchair simulator: 1) left and right upper extremity supports; 2) left and right lateral trunk supports; 3) secured, level, derotated pelvis; 4) tilt in space; and 5) all four parameters. The Viasys Jaeger Impulse Oscillometry System and Respironics Non Invasive Cardiac Output monitor (NICO) measured the dependent variable, pulmonary mechanics, via Hans Rudolph facemasks. Spasticity (by Modified Ashworth Scale), patient characteristics, and medications were recorded. A process log captured participant recruitment and retention challenges and response to protocol. Results: Recruitment was challenging; retention was 50%. For this sample, despite lack of power, both RAW and MV improved with upper extremity and lateral trunk supports. Highest RAW was seen with total absence and total presence of the parameters, and secured, level pelvis.The data collection protocol was feasible for 50% of participants, none of whom could execute conventional measurement. Facemask and seating simulator acceptability were 75 %, improving with participant verbal communication ability. The facemask seal was vulnerable to tilted positioning; 75% of participants became fatigued. RAW measures differed from manufacturer's directions but were reliable. Conclusions: The Prairie seating simulator, Jaeger IOS, Respironics NICO, and Hans Rudolph facemasks effectively measured pulmonary mechanics as a function of wheelchair seating parameters in this sample. Upper extremity and lateral trunk supports most greatly reduced RAW, maintaining MV. Verbal children tolerated the procedure best.
Dissertation (Ph.D.)--University of South Florida, 2007.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 115 pages.
Includes vita.
Advisor: Audrey Nelson, Ph.D.
Postural management.
Dissertations, Academic
x Nursing
t USF Electronic Theses and Dissertations.
4 856


Wheelchair Positioning and Pulmonary Function in Children with Cerebral Palsy by Lee Barks A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Nursing University of South Florida Major Professor: Audrey Nelson, Ph.D. Jason Beckstead, Ph.D. Mary Evans, Ph.D. William Lee, Ph.D. Date of Approval: July 5, 2007 Keywords: posture, respiration, breat hing, seating, postural management Copyright 2007, Lee Barks


Dedication For my family and those who I am proud to call my friends, You are all my heroes; you are very strong and very kind, and the value of your love and support is beyond measure.


Acknowledgments With deepest gratitude, I acknowledge th e members of my dissertation committee Dr. Audrey Nelson, Committee Chair and my mentor these past five years, your leadership and vision are extraordinary. Th ank you for steady, reasoned, calm guidance, and unbelievable email responsiveness, as well as enthusiasm, encouragement, and countless unexpected favors. Dr. Jason Beckst ead, for always having higher expectations of me than I did, and for so many patient explanations of statistical concepts. Dr. Bill Lee, for encouragement, teaching, and schol arliness in biomedical engineering, and especially for telling me I rea lly could have been an engineer. Dr. Mary Evans, for your candid confession that you are still a curious sc ientist, for guidance in the intricacies of our academic program, for unflagging cheerfuln ess, and for encouraging me to remain true to my original goals. I also thank Tric ia Holtje for her encouragement and wisdom over these past five years, as well as her unique skills and practical knowledge of research documentation, and Dean Patricia Burn s, who saw potential in me. I have been richly blessed to learn from each of you. Grateful acknowledgement is exte nded to Elaine M iller, editor of Rehabilitation Nursing for permission to reprint Chapters On e and Two; to Shriners Hospitals for Children, Tampa, for participation as a st udy site; and to the Rehabilitation Nurses Foundation, who funded this study.


i Table of Contents List of Tables iii List of Figures iv ABSTRACT v Chapter One: Introduction 1 Problem Statement 5 Purpose 5 Research Questions 6 Definition of Terms 6 Significance to Nursing 8 Research Progress 12 Construct One: Genesis of Deformity/Scoliosis 14 Construct Two: Scoliosis/Respiratory Dysfunction 15 Construct Three: Seated Position/Pulmonary Function 17 Seven Gaps in the Research 20 Conceptual Framework 22 Research Design 25 Objectives 26 Purpose 26 Research Questions 26 Variables 27 Power 42 Inclusion/Exclusion Criteria 43 Setting 44 Procedures 44 Approvals 44 Human Subjects Protections 45 Participant Recruitment 45 Data Collection 46 Data Analysis 50 Chapter Four: Results 52 Sample Characteristics 52 Verbal Ability and Motor Control 55 Findings 56


ii Research Question One: Contribution of Wheelchair Parameters to Total Airway Resistance and Minute Ventilation 57 Qualitative Data from the Process Log 62 Discussion of Findings 68 Chapter Five: Discussion 70 Contribution of Each Parameter to Pulmonary Mechanics 70 Effects on Subject Recruitment, Retention, & Responses to Data Collection Protocol 71 Seating Simulator and Facemask Acceptability 71 Postural Instability and Unsupported Position 72 Usefulness of Pulmonary Measurement 72 Optimal Measurement Intervals 74 Participant Recruitment and Retention Rates 75 Differences in Total Airway Resistance (RAW) By Wheelchair Parameter 76 Study Limitations 77 Sample 77 Method 77 Recommendations for Future Research 78 Conclusions 79 References 81 Appendix A: Modified Ashworth Scale 87 Appendix B: Consent to Participate 89 Appendix C: Invitation to Participate 100 Appendix D: Instrument Specifications 102 Appendix E: Institutional Review Board Approvals 111 About the Author End Page


iii List of Tables Table 1 Definitions of Terms 7 Table 2 Variables and Measurement Plan 27 Table 3 Data Collection Protocol 47 Table 4 Sample Composition 53 Table 5 Study Completion, Age, and Verbal Ability 56 Table 6 Effect of Seating Condition on RAW Means (SDs) 62 Table 7 Process Log Themes and Data 63


iv List of Figures Figure 1 Wheelchair Seating/Pulm onary Mechanics Logic Model 23 Figure 2 Prairie Reflecti ons Seating Simulator 32 Figure 3 Respironics Non-Invasi ve Cardiac Output Monitor 35 Figure 4 Viasys Jaeger Impulse Oscillometry System 38 Figure 5 Child with Viasys Jaeger Impul se Oscillometry System Attached to Facemask 39 Figure 6 Four HRI Facemasks 41 Figure 7 Total Airway Respiratory Means by Seating Parameter (+ or SD), (n = 8) 59 Figure 8 Mean Minute Ventilation by Seating Parameter (+ or SD), (n = 8) 60 Figure 9 Scatterplot of RAW by Seating Parameter (n = 8) 61


v Wheelchair Positioning and Pulmonary Function in Children with Cerebral Palsy Lee Barks ABSTRACT Background: In children with cerebral palsy (CP), poor trunk cont rol fosters spinal deformity, pulmonary compromise (Canet, et al ., 1998), increased health risks, and costs of long-term care (Braddock, 2001). Evidence links posture and pulmonary function, but influence of wheelchair parameters on pulmonary mechanics is unknown. Objectives: 1) Determine relative contri bution of five wheelchair configuration parameters to improvement in pulmonar y mechanics--total airway resistance (RAW), tidal volume, minute ventilation (MV), and deadsp ace to tidal volume ratio; 2) Describe recruitment and retention of school-aged chil dren with CP; and 3) Discuss response of the participants to the protocol. Method: This within-subject s, descriptive study employed a sample of 8 school-aged children with CP and flexible spines w ho could not sit alone. In a single session, participants experienced five seating para meters manipulated in a Prairie wheelchair simulator: 1) left and right upper extremity supports; 2) left a nd right lateral trunk


vi supports; 3) secured, level, dero tated pelvis; 4) tilt in space; and 5) all four parameters. The Viasys Jaeger Impulse Oscillometry Syst em and Respironics Non Invasive Cardiac Output monitor (NICO) measured the depe ndent variable, pulmonary mechanics, via Hans Rudolph facemasks. Spasticity (by Modified Ashworth Scale), patient characteristics, and medications were r ecorded. A process log captured participant recruitment and retention challenges and response to protocol. Results: Recruitment was challenging; retenti on was 50%. For this sample, despite lack of power, both RAW and MV improved with upper ex tremity and lateral trunk supports. Highest RAW was seen with total absence and total presence of the parameters, and secured, level pelvis. The data collection protocol was feasib le for 50% of participants, none of whom could execute conventional measurement. Facemask and seating simulator acceptability were 75 %, improving with par ticipant verbal communication ability. The facemask seal was vulnerable to tilted positio ning; 75% of participants became fatigued. RAW measures differed from manufacture r’s directions but were reliable. Conclusions: The Prairie seating simulator, Jaeger IOS, Respironics NICO, and Hans Rudolph facemasks effectively measured pulmonary mechanics as a function of wheelchair seating parameters in this sample Upper extremity and lateral trunk supports most greatly reduced RAW, maintaining MV. Verbal children tolerated the procedure best.


1 Chapter One: Introduction1 When children with severe cerebral pa lsy (CP) do not develop trunk control, including the ability to sit independently, a series of events unfolds. They experience functional difficulties with eating, speaking, moving about, and executing other activities of daily living. Their immobility is associated with increas ed health risks known as the “hazards of immobility” (Olson, 1967). Over time, spinal deformity, or scoliosis, develops in individuals who lack ability to co-contract their paraspinal and abdominal muscles in a balanced fashion on right and left sides when in upright position. When trunk control is absent, the upper trunk collapses downward, with associated spinal deformity. In the beginning this spinal deformity is somewhat flexible, but over time it becomes “fixed” or immovable (Hale, et al., 1987; Shannon, et al., 1971) Without intervention, lateral and rotational spinal deformity develops in nonambulatory indi viduals who lack trunk control, generally becoming fixed during and after the adoles cent growth spurt. Persons who never ambulate must be positioned extensively in wheelchairs for mob ility, stability, and activities of daily living (F raser, 1990; Cox, 1987). Morbidity and mortality develop chiefly in the respiratory system (Eyman, et al., 1990). As they age and increase in size, nonambulatory persons often move into long term care facilities with associated increases in cost (Braddock, 2001) and loss of daily family interaction. 1 Portions of Chapters One and Two have been published by this author (Barks, 2004). They are updated and reprinted with permission for this dissertation.


2 Recent prevalence data (United States (U.S.), 2003) show 258,000 have a diagnosis of infantile CP, and 2 per 1000 are under 18 years of age. Thirty-nine percent of these persons have had a related ho spitalization annually, and 74% experienced limitation of activity (U.S. National Cent er on Health Statistics, 2003). All had experienced one or more related physician visi ts. For all racial groups, age of death with CP as an underlying cause is significantly young er than in the gene ral population (United CP, 1995). Death from respiratory causes is lis ted as the leading cause of death in this group (42.1%); in the non-disa bled population, respiratory ca uses are not among the top four (United CP). Healthy People 2010 states this goal: “Promo te the health of people with disabilities, prevent secondary conditions and eliminate disparities between people with and without disabilities in the US popul ation” (U.S. Department of Health and Human Services, 2000, p. 7). For such persons, medical treatment alone is insufficient in improving function and respiratory outcom es. Finding a way to augment medical treatment of pulmonary issues through better posi tioning could be worthwhile and could also reduce health disparities. Wheelchair seating for people with CP and sc oliosis is the most prevalent form of upright positioning. Proper wheelchair positioni ng appears simple but can be deceptively complex. It may either be “adaptive,” conforming to the prevailing posture of the disability and gravity’s pull, or it may be st rategic and “therapeutic,” when the main goal is to prevent or reverse development of de formity (Fraser, et al., 1990; Cox, 1987). Each supporting surface of a wheelchair contributes to the resulting seated posture of a person placed in it. In practice, wh en wheelchairs are contoured to accommodate the scoliotic curve, the pelvis may be level and secure but upper extremity support is absent; or there


3 may be upper extremity support, but the pelv is is positioned obliquely with contoured surfaces. In either case, there may be malalignment of the trunk, such that there is reduced respiratory volume exchange and a resi stive, rather than obs tructive, pattern of breathing (Canet, et al., 1998). It is well known that reduced respiratory volume exchange can lead to hypoxemia, pneumonia, and at electasis, which may all be seen in the population of those with CP. The mechanism pr oducing increased resistance of the chest, with resulting reduced volume exchange in persons with scoliosis, has been summarized by Canet, et al. (p. 796). Possi bly total airway resistance (RAW) is increased, related to the spinal deformity, and is also responsible. At the present time, this is unknown, as is the effect on other pulmonary mechanics--tidal volume (VT), minute ventilation (MV), and deadspace to tidal volume ratio (VD/VT)--of wheelchair parameters supporting posture of the trunk. Tidal volume, minute ventilation, a nd deadspace to tidal volume ratio can be used to clinically interpret cha nges in total airway resistance. In fact, many custom wheelchairs are made using contoured surfaces that hold the rib cage statically in its deformity. Such wh eelchair seating parameters as securing and leveling the pelvis, lateral trunk supports, a nd upper extremity supports are not always used in wheelchair seating and may be critic al factors in pulmonary mechanics. Leveling the pelvis and securing it to remain level provides a biomech anical base of support for the rest of the spine; lateral trunk supports may straighten the trunk, and upper extremity supports can lift weight of the upper trunk o ff the abdomen and dependent lung segments, improving volume exchange and decreasing re striction by allowing greater movement. Airway resistance may improve with trunk straightening.


4 In addition to genesis of scoliosis, exte nt of the problem, and possible role of parameters of wheelchair sea ting, there are considerable i ssues in studying the population with CP. Traditionally, spirometric testi ng, including plethysm ography using a “body box,” has been used to conduct pulmonary function tests; however, this form of testing is not appropriate for children with CP for thr ee reasons: (1) children with disabilities have difficulty complying with required spirometric protocols; (2) proce dures for manipulating postural supports would be logistically impossible in a body box, and (3) primitive reflexive movement can prohibit reliable m easurement. The Respironics Non Invasive Cardiac Output monitor (NICO) can be us ed passively by all ages for VT, MV, and VD/VT measurement, but another instrument is needed for the primary outcome variable, RAW. Although one instrument, the Jaeger Impulse Oscillometry System (IOS), has been marketed for testing RAW in children, its factory-approved usage employs a mouthpiece which, although intended for use in children, cannot be used in children with CP, due to its oral intrusiveness. Made of hard silic one rubber and extending far into the mouth over the tongue, this mouthpiece can trigger primitiv e reflexes common in children with CP, such as tonic biting, gagging, and uncoordi nated movements of lips and tongue, making consistent measurement potentially noxious and unreliable. A silicone facemask may be used to ob tain measures with the IOS and NICO, if an adequate seal is obt ained, although obtained RAW measures may not correspond to usual clinical values. RAW measures could be expected to increase because the surface of the face, lips, tongue, and inner surface of the facemask could provide some increased resistance to airflow. This resistance c ould be greater than the resistance of the recommended method, in which the mouthpiece is placed into the mouth at about mid-


5 tongue. Finally, the Waugh protoc ol (2004) contains at least nineteen possible wheelchair determinants of posture: pelvis support; tw o upper extremity supports; three lateral trunk supports; tilt in space support; left and right seat depths; seat wi dth; back rest height; left and right leg rest lengths; left and right seat to leg rest angles; right and left hip abduction/adduction; tilt in space; a nd headrest/neck flexion or extension. Problem Statement The cumulative effect of poor trunk control and the inability to sit independently can lead to spinal deformity and pulmonary compromi se in children with disabilities including severe CP (Canet, et al., 1998). This postura l instability limits mobility, with resultant health risks and costs of long-term car e (Braddock, 2001). For persons lacking trunk control, supporting surfaces, including wheelchair parameters, dictate posture. Although some evidence links posture and pulmonary function, the extent to which wheelchair parameters influence pulmonary mechanics in seating of children with disabilities is largely unknown. Purpose The purpose of this study was to estimate the relative importance of five wheelchair configuration parameters as predictors of pulmonary mechanics improvement, reflected in total airway resistance of school-aged chil dren with CP who cannot sit alone but have flexible spines. These configur ation parameters were: (1) left and right upper extremity supports; (2) left and right lateral trunk supports ; (3) secured, level, and derotated pelvis; (4) tilt in space; and (5) presence of all four parameters (“totally supported”).


6 Research Questions The questions to be answered in this study are: 1. What is the relative cont ribution of each of five wheelchair parameters to improvement in pulmonary mechanics? 2. What are the challenges associated with subject recruitment and retention in a sample of school-aged children with CP? 3. What is the response of children with CP to the data collection protocol? Definition of Terms For purposes of this study, terms are defined and listed in Table 1 on the following page.


7 Table 1 Definition of Terms Term Definition Deadspace to tidal volume ratio (VD/VT) A pulmonary convention denoting the ratio of the volume of gas contained in the trachea, bronchi, and bronchioles (which do not participate in gas exchange), divided by the tidal volume. Pulmonary convention is VD/VT. Planar system A positioning system in which cushioned, planar surfaces passively apply some force to a body part or parts. Generally, force applied is equal to the force of gravity such as in an upper extremity support or armrest. Pulmonary function Capability of the pulmonary system to function, measured by clinical laboratory tests such as forced expiratory volume in 1 second (FEV1), the volume of gas measured in one second of forced exhalation. Generally, pulmonary function measures require the participant’s ability to actively engage in volitional physical responses. Children with CP may or may not be capable of such responses. (Often, the term pulmonary function also includes measures of pulmonary mechanics, but pulmonary function includes at least some volitional response. It does not refer to any variables in this study; it is defined here onl y to distinguish traditional, clinical, pulmonary function measures from pulmonary mechanics measures .) Pulmonary mechanics Pulmonary measures of mechanical forces and volumes and their quotients in ventilation. In this study, these are total airway resistance (RAW), tidal volume (VT), minute ventilation (MV), and deadspace to tidal volume ratio (VD/VT). Participation may be passive; that is, some instruments that measure pulmonary mechanics do not require active, volitional responses.


8Term Definition Minute ventilation (MV) A pulmonary convention denoting the amount of gas in liters exhaled from the airway in one minute, equal to the tidal volume multiplied by the respiratory rate. Pulmonary convention is MV. Tidal volume (VT or VTE) A pulmonary convention denoting the amount of gas in milliliters exhaled from the airway in one exhalation. Pulmonary convention is VT or VTE. Total airway resistance (RAW, or R at 5Hz) A pulmonary convention denoting the total resistance of the airway to gas flow, at 5 Hz frequency of oscillation. Measured in kPascals/L/sec. Pulmonary convention is RAW. Wheelchair parameters Nineteen characteristics of a wheelchai r that combine to shape the posture of a seated human body according to the Waugh protocol (2004). Six “wheelchair seating conditions” varied in this study include: one condition of each of the four wheelchair parameters, one condition of total absence of the four parameters (“unsupporte d”), and one condition of presence of all of the four parameters (“totally supported”). Significance to Nursing There is no unifying nursing mid-range th eory of positioning despite widespread, strategic use of positioning to relieve pressure and protect skin integrity in immobile persons. Based on clinical intervention studi es, nurses now positi on strategically for pneumonia and mechanical ventilation a nd Acute Respiratory Distress Syndrome (ARDS) (Yeaw, 1996; Bridges, 2001). In addi tion to nursing use of positioning in other conditions, positioning could be strategic for persons with disabilities due to the fact that various medical treatments, such as spinal surgery, often have limited success when used


9 alone. The primary goal of surgical treatme nts is to arrest progression of spinal deformity, thereby reducing morbidity and mortality and improving quality of life. Surgical interventions are associated with serious risks, are not always employed, and do not appreciably reverse existing pulmona ry compromise (Cassidy, et al., 1994). Practitioners in the field know that surgically -placed rods can dissect out of the spine. Therefore, there is still a need for upright, supported, seated positioning of individuals who do not have trunk control. It is reasona ble for a descriptive nursing study to inquire into factors that may produce healthy pul monary function, in support of future interventions. Nurses have responsibility for positioning their patients and may do so strategically to oppose the pu ll of gravity, prevent deformity, and assist pulmonary mechanics. The effects of various wheelchai r parameters on pulmonary mechanics have not been definitively studied, and current ly, the wheelchair industry does not show evidence that positioning for adequate brea thing is understood. In fact, many custom wheelchairs are made using contoured surfaces that hold the rib cag e statically in its deformity. Such wheelchair seating parameters as securing and leveli ng the pelvis, lateral trunk supports, and upper extremity supports are not always used in wheelchair seating yet may be critical factors in pulmonary mech anics. Leveling the pe lvis and securing it to remain level provides a biomechanical base of support for the rest of the spine. Lateral trunk supports may straighten the trunk; upper extremity supports can lift the upper trunk’s weight from the abdomen and depe ndent lung segments, thus improving volume exchange and decreasing rest riction by allowing greater m ovement. Trunk straightening may decrease airway resistance. These circum stances point to the need for development


10 of evidence-based nursing recommendations on seating parameters to support healthy pulmonary function. Chapter One describes the genesis of sco liosis in children with CP who cannot sit alone and discusses morbidity and morta lity for this group. Other than medical interventions, such as spinal stabilization surgery, there are few interventions for this problem, and surgical interventions are not al ways effective or rest orative. Although this is a small population, disability is severe and expensive. It dire ctly confers health disparity from the general population, along wi th separation of the individual from the family, as care demands rise with immobility. There are many parameters of wheelchair seating, and children with CP are a comp lex group to study. A within-subjects, descriptive study was proposed to test a new method for determining the extent to which wheelchair parameters may influence pulmonary mechanics.


11 Chapter Two: Review of Literature2 This chapter presents a synthesis of th e research literature in the domain of posture, particularly wheelchair seating, and pulmonary function in persons with CP who lack trunk control. An electronic search wa s conducted of PubMed, Ovid Medline, and the Cumulative Index to Nursing and Allied Health Literature (C INAHL) databases from 1966 through February 2007. Search terms included posture body position wheelchair CP pulmonary function respira and seat. Ancestral searching and the ISI Web of Knowledge were then used to find studi es citing Nwaobi and Smith (1986), who conducted a study of the effects of wheelchai r seating on pulmonary function of children with severe CP. The number of articles and websites review ed was 107; an additional 16 articles or websites were reviewed about the instrume nts, with a total of 65 articles eventually used. In the synthesis, peopl e with spinal cord injuries were excluded because their spasticity may be symmetrical, so that sc oliosis may not develop. Those with spina bifida, including myelomeningocele, were also excluded. In spina bifida, there is some trunk control down to the level of the spinal defect, and the le vel of the defect can vary over most of the length of the spine. Resulti ng posture may differ from posture of persons with CP who lack control throughout th e trunk, possibly affec ting breathing. Other excluded research reports addressed body positio ning, posture, and breathing in a host of 2 Portions of Chapters One and Two have been published by this author (Barks, 2004) and are updated and reprinted with permission for this dissertation.


12 other conditions and positions such as ba riatrics, surgery, non-seated positions, and pulmonary function testing not useful for a population who cannot follow directions. Of the articles, 21 were primarily literature reviews, and 25 were research reports, which were evaluated using an abbr eviated form of the Research Analysis Tool (RAT) (Moody, 2001). The RAT instrument systematically orde rs 35 characteristics of reported research and can be used to quantify studies. Only one of the articles con cerned nursing theory, process, or interventions (Raven, 1989). The res earch reports fell into three main areas, or constructs: genesis of deformity/scoliosis scoliosis/respiratory dysfunction and seated posture/pulmonary function In addition, a treatment approach identified as postural management for which a review of the evidence ba se has been conducted (Farley, et al., 2003), was also reviewed and a robust li nk found between physiol ogical function and posture. Research Progress Research into the three constructs— genesis of deformity/scoliosis scoliosis/respiratory dysfunction and seated posture/pulmonary function —has shown progress over the review time period from 1966 through early 2007. For each construct, for purposes of this study, the predominating conceptual model was physiologic. In some studies, physical therapists used, implied, a nd sometimes stated a model of “function” (Nwaobi, 1986; Myhr & vonWendt, 1991) In a function model, the treatment goal (and treatment effect) is function in activities of daily living; these studies were excluded because health or physiologic outcomes genera lly were not used as outcome measures. A physiologic model contrasts with a function model in that physiologic, or health,


13 outcomes may be precursors of function su ch as self-feeding, typing, sitting, and dressing. Each model has been the focus of research on wheelchair seating, but only a physiologic model is presented here. In addition to differences between f unction and physiologic models, Fraser, Hensinger, and Phelps (1990) differentiate d the concept of “adaptive” positioning or equipment, from “therapeutic.” The goals of adaptive equipment use and interventions are to accommodate a person’s disability a nd modify the environment; the goal of therapeutic interventions is to change the course of the disa bility. In this study’s model, then, the goal of every nursing intervention can be said to be therapeutic. The concepts of Fraser, et al. are important for nurses in he lping to improve respiratory function through postural support by changing the course of trunk alignment, rather than accommodating abnormal posture. The three constructs support this study in a logical progression: genesis of deformity/scoliosis; scoliosis/respiratory dysfunction, and seated positioning/respiratory function These constructs are related in the follo wing way: Lack of trunk control, which is present in spastic quadrip aretic CP and other neuromuscular disorders, eventually fosters fixed spinal deformity, or scoliosis (Berven & Bradford, 2002; Letts, et al., 1992). In severe CP, scoliosis is associated with a restrictive pattern of respiratory movement and deficient pulmonary mechanics and pulmonary function (Ting & Lyons, 1963; Bjure & Berg, 1970). In the absence of normal pulmonary mechanics, the respiratory system is a known locus of morbidity and mortalit y (Nilsonne & Lundgren, 1968; Nachemsson, 1968); this is particularly true among persons wi th severe CP and scoliosis (Eyman, et al., 1990). If scoliosis and fixed deformity could be prevented, or if at least the individual


14 could be consistently positioned in alignment, much of the respiratory morbidity and mortality might be preventable. Customarily, th is patient group is seated in wheelchairs for long periods daily. Therefore, wheelchair s eating is of interest to nurses engaged in engineering respiratory health in this group, especially if such seating does not provide adequate alignment. Construct One: Genesis of Deformity/Scoliosis For the first construct, genesis of deformity/scoliosis many authors have documented the genesis of spinal deformity in persons with neuromuscular disorders, including severe CP (Berven & Bradford, 2002; Letts, et al., 1992; Makley, et al., 1968). These authors note the collapsing nature of the spinal curves and development of associated deformity over time, all confir med by clinical observation. Conversely, Frankeny, et al. (1982) showed that at least 30 minutes of daily stretc hing is required to result in muscle growth of normal and dys trophic muscle. It is well known that when stretching and movement are ab sent, the soft tissue of the spine becomes rigid. One requisite in prevention of deformity is bala nced stretch of soft tissue on all sides of moveable joints; in order to maintain healt hy muscle tissue, continued stretch is needed, on a daily basis. Cherry (1980) concluded that prolonged maintenance in a desired position to prevent or reverse deformity ma y be more effective than passive manual stretching (Passive Range of Motion), due to very brief periods of execution of the intervention. This conclusion applies to joints and muscles of the spine, as well as to other joints, and may suggest that nursing inte rvention to consistently position individuals lacking trunk control in spinal alignment is needed. While medical interventions have


15 focused on surgical spinal stabilization and orthotic bracing to halt deformity progression, these interventions have not prevented deform ity on a large scale. More importantly, it is well known that scoliosis is a ssociated with respiratory dys function, the second construct for consideration here. Taken together, thes e studies demonstrate a need for strategic positioning in sustained positions for immobile persons with CP to prevent deformity. This is therapeutic positioning; it is a foundation for the second construct. Construct Two: Scoliosis/Respiratory Dysfunction Many studies are concerned with scoliosis/respiratory dysfunction Long ago, Itovici & Lyons (1956) and Bergof sky, et al. (1959) found scolio sis to be associated with decreased vital capacity and maximal breath ing capacity, with re strictive patterns of movement and pulmonary function measures. Ting & Lyons (1964) found the same restrictive pattern with kyphos coliosis as for lateral spinal curvature. In a study to measure static and dynamic lung volumes in subjects with CP, Bjure & Berg (1970) found respiratory dysfunction from respiratory muscle weakness, again with a restrictive pattern and a total lung cap acity 85% of normal volume, with low maximal oxygen uptake. They studied sevento twenty-three ye ar olds with CP (n = 22). Their sample included patients with less severe disabiliti es, because they could voluntarily cooperate with spirometric tests. It also probabl y included, by stated age, a mixed group of postpubertal participants who had completed th eir pubertal growth s purts and already had fixed deformity, along with some who were still children. These studies are important, however, because they show th e relationship between the architecture of the deformity and pulmonary mechanics, primarily restri ction of movement and volume phenomena.


16 Nilsonne and Lundgren (1968) and Nachemsson (1968) noted remarkable cardiopulmonary morbidity and mortality among pa tients with scoliosis. Their studies are important in showing the seriousness of this problem. An inference may also be drawn that studies on scoliosis/respir atory dysfunction show that defi cits culminate in increased need for nursing care, generally in cos tly long term care se ttings (Braddock, 2002). After the 1960s, interventions to straight en the spinal curve(s) and influence respiratory function in the form of pulmona ry mechanics have been the subject of research (Sevastikoglou, et al., 1976; Nwa obi & Smith, 1986; Noble-Jamieson, et al., 1986; Letts, et al., 1992; Cassi dy, et al., 1994; Leopando, et al., 1999; Tangsrud, et al., 2001;). Among these interventions are spinal stabilization surgery, such as Harrington rod insertion, Luque and other procedures, as well as use of positioning with rigid and soft orthoses and seating. Cassi dy, et al. studied 37 persons wi th severe CP and scoliosis, evenly divided into patients with spinal fu sions and patients without in a mixed group of children and adults, ages 11-27 years, most of whom were post-pubertal. Surgery in the fused group did not improve pulmonary func tion after problems had developed, so prevention is of primary importance, particular ly in children whose spines are generally still flexible until the puber tal growth spurt (Samils on, 1972). Even when medical interventions such as surgery or bracing are used, there is s till a need to consistently support the trunk in alignment to reduce latera l and gravitational force on the spine when seated and ensure normal pulmonary mechan ics. Rigid orthoses have been found to decrease tidal volume or vital capacity (Sev astikoglou, et al.; N oble-Jamieson, et al.; Tangsrud, et al.), and although soft orthoses have not reduced lung volumes, they have also produced no improvement (Letts, et al.; Leopando, et al.). This fact increases the


17 importance of noninvasive nursing solutions, underscoring nursing responsibility for strategic positioning to prevent reduced volume exchange, and raising the question of how straightening the spine by positioning mi ght improve pulmonary mechanics if soft orthoses do not. Construct Three: Seated Position/Pulmonary Function Nwaobi and Smith (1986) compared the effects of nonadaptive and adaptive seating on pulmonary function in children with CP, using a within-subjects, quasiexperimental design, spirometry, and both a sling-seat wheelchair and an optimallysupported seating system. A convenience sample was used (n = 8), with sequence of testing randomized. Mean Vital Capacity (V C), mean Fractional Expired Volume at 1 second (FEV1) as a percent of Vital Capacity, and Mean Expiratory Time all increased significantly in “adaptive” seat ing, with differences in pulmonary measures in the two types of seating measured by t tests, with p< .05. Post hoc power analys is showed that this study was sufficiently powered due to very la rge effect sizes (Cohen’s d of 2.07 for VC and 1.53 for 1-second FEV1) although there was a small effect for expiratory time (Cohen’s d = 0.37). The sample included child ren aged 5-12 years with spastic CP. Selection of children uniformly lacki ng trunk control and possessing only CP is important. Postures in other neuromuscular di sorders may be dissimilar, according to the level of spinal defect in spina bifida a nd variations in muscle tone among muscular dystrophy subjects or spinal muscle atrophy subj ects. These variations in posture could affect the mechanics of breathing. Although th is study was well-powered despite small sample size, its limitations are that only participants who could actively assist with


18 voluntary tests of pulmonary function were used, thus excluding the most disabled children with the most pathology who could mo st benefit from therapeutic positioning. This study also did not measure airway resi stance, which may be increased when the weight of the unsupported upper trunk presses on the dependent portions of the lung or whenever the trunk is not straight. The study showed improvement with the aggregate of the pulmonary function variables, but it did not distinguish which of the multiple properties of adaptive wheelchair seating produced improvement. Redstone’s study (2004) is notab le because it inquired in to the effects of seating position on respiratory patterns of preschooler s with CP. However Redstone’s interest was in respiration and positioning the individual for speech production, a function model, rather than physiologic/health outcomes The Nwaobi and Smith study (1986) and the Redstone study are important because they are the first and only studies to link scoliosis, pulmonary function, and CP. In addition to these studies, Druz and Sh arp (1981) suggested that gravity as a stimulus for stretch of human respiratory muscles and increased muscle activity may explain the influence of body position on vent ilation and volume exchange. They asked why chest movement is greater in upright pos tures and abdominal movement is greater in supine positions; they found a “combinati on of increased activation of rib cage inspiratory muscles plus greater activati on of the diaphragm” (p. 1552) in upright positions. Human studies about respiratory func tion and posture from 1960 to the present also included Crosbie and Myles (1985) Dean (1985), Geubelle and Goffin (1962), Hough (1984), Landers, et al (2003), and Moreno and Lyons (1961). Although Hough and Dean published review arti cles, all authors used physiolo gic measures to explain or


19 quantify effects of posture on respiratory function. In pr ospective studies, Moreno and Lyons (1961), Geubelle and Goffin Crosbie a nd Myles, and Landers, et al. measured and compared the effects of seated, prone, and supine postures, among some others, on respiratory volumes. Moreno and Lyons (1961) found ventilation in prone and seated almost identical but found a sta tistically significant reduction in ventilatio n from seated to supine. Geubelle and Goffin al so found prone and seated values virtually equal in children and a significant reduc tion in functional reserve capac ity from seated to supine, in both children and adults Crosbie and Myles found si gnificant differences in ventilation in seated, supine, prone with hips elevated, and slumped-at-45-degrees-seated postures. They noted the sameness of this sl umped posture and sitting upright in bed and that slumping appears to cause decreases of 12% and 15 % respectively in both vital capacity and forced expiratory volume in 1 sec (FEV1). Landers, et al. found statistically significant increases in tidal volume and minute ventilation in seated upright as opposed to slumped postures. The posture of the insufficiently-supported trunk of a person with scoliosis may be equivalent to kyphotic sl umping. This slumping from lack of support may have the same effects on ventilation as any other slumped posture. Conversely, the properties of an effective seated posture produc e better ventilation in seated than slumped postures. A review by Pynt, et al. (2001) eval uated properties of “optimal seated posture” for the lumbar spine of healthy adults, define d as posture associated with intervertebral disc health and absence of pa in. This posture was slightly lordotic, with a slightly forward-tilted pelvis and pe riodic movement about. For immobile persons who do not move themselves about, the hips must be pla ced against the seat back in order to provide


20 a base of support for the spine and to avoi d slumping, and a seatbelt may be needed to maintain seated posture. Presumably, in both the lateral spinal curvature of scoliosis and kyphotic slumping, additional weight of the head, n eck, and upper chest is unsupported by the spine. This constitutes a hypothetical load arm that places additional weight and force on the distal lungs (alveoli), restricting both abdominal and chest wall expansion and probably also increasing both alve olar and total airway resistance. In order to measure the effect of various supports in straightening the posture, then, it is ge rmane to measure total airway resistance (RAW) and the volume of gas exchanged over a unit of time (MV). In clinical research, these measures could be in cluded in a standard protocol of pulmonary function studies that requires the participant to inhale and forcefully exhale. Many people in the population of persons who lack tr unk control, those with great pulmonary pathology, are unable to execute this maneuver, excluding them heretofore from research into these issues. For this reason, a measur ement method that simply yields physical measures of pulmonary mechanics should be use d, rather than one that also includes the functional ability to volitionally produce these measures, pulmonary function tests. Seven Gaps in the Research Based on this systematic literature review, seven gaps in the research were identified: 1) Although Nwaobi and Smith’s study ( 1986) was innovative and sufficiently powered with very large effects in vital capacity change and FEV1 improvement, it aggregated the specific wheelchair seating properties that produced change. Future


21 studies need to characterize changes in pulmonary function according to specified wheelchair seating parameters. 2) Many studies (Makley, et al., 1968; N oble-Jamieson, et al., 1986; Letts, et al., 1992; Sevastikoglou, et al., 1976) us ed a sample of various ages with mixed etiologies for absent trunk control. A ho mogeneous sample of children with CP who have not experienced their pubertal growth spurt and who lack trunk control is needed, however, so that postures are similar before and after correction, and influence on pulmonary mechanics is attributable to the treatment, not to postural differences. These postures are also similar to those of the immobile elderly and may inform later studies. 3) Studies using spirometry (Leopando, et al., 1999; Nwaobi & Smith, 1986) were limited to participants who could cooperate with traditional pulmonary function measures and excluded the most compromised individuals most likely to benefit from intervention. Future studies need to test other measur ement approaches for pulmonary function to capture subjects with more severe disabilities. 4) Except for the Leopando, et al. study (1999), airway resistance has not been widely used to reflect pulmonary mechanics in these studies. Total airway resistance (RAW) is potentially important in reflecting trunk straightness and in allowing sufficient volume exchange. 5) A simple, short procedure is needed, both for positioning and for some measuring of pulmonary mechanics. The Leopando, et al. study used a precise but lengthy method to measure pulmonary mechanics, sometimes requiring up to five hours (G. Rempel, personal communication, Februa ry 12, 2004), thus possibly resulting in fatigue and change in outcome measures. A simpler but precise procedure is needed.


22 6) Except for that of Nwaobi and Smith ( 1986), studies did not determine the role of multiple wheelchair seati ng parameters in multiple planes in producing pulmonary outcomes. The difficulty is to meaningfully evaluate one parameter at a time because multiple parameters probably produce respiratory movement in multiple planes. 7) In the literature, medical research ha s attempted various interventions to find adequate treatment. To inform interventi on studies with this population, a preliminary study is needed to test a new method of m easuring pulmonary mechanics in persons who cannot comply with the testing and to dete rmine the specific positioning variables that may predict improvement in pulmonary mechanics. Conceptual Framework For each of the three constructs, the predominating conceptual model is a physiologic model. Intentionally positioning the individual to maintain passive stretch and gain range in joints that have no ac tive motion (Samilson, et al., 1972; Frankeny, et al., 1983; Cherry, 1980) may decrease internal physiologic deficits in respiratory and other systems by allowing movement and intern al space. This approach is congruent with Orem (2001) in supplying trunk s upport to address the Self-Car e Deficit of lack of trunk control and meet the Universal Self-Care Requisite of Air. There is a need for sufficiently-powered research with children with CP that can fill gaps in understanding of respiratory function for this group. This study is intended to provide the next level of evidence in the research, with particular attention to the se ating properties that may be associated with any changes in pulmona ry mechanics. In the future, greater understanding of the physiologic outcomes of specific properties of wheelchair


23 positioning may allow research of more effective interventions that reduce associated morbidity and mortality, improve health and quality of life, and decrease costs. A logic model is presented in Figure 1. Note: Shading denotes variables used in th is study. All other va riables were held constant. Figure 1. Wheelchair Seating/Pulmonary Mechanics Logic Model. Patient Characteristics Age Gender Medications Spasticity Ethnicity Neurological Deficits Pulmonary Mechanics Total Airway Resistance (RAW) (Improved: decreased from usual) Tidal Volume (VT) (Improved: increased from usual) Minute Ventilation (MV) (Improved: increased from usual) Deadspace to Tidal Volume Ratio (VD/VT) (Improved: Decreased from usual) Environment/Wheelchair Seating Parameters Seat Depth Seat Width Seat to Back Angle Back Rest Height Pelvis Level Pelvis Secured Pelvis Derotated R Arm Rest height L Arm Rest Height L Lower Lateral Trunk Support Ht. R Lateral Trunk Support Ht. L Upper Lateral Trunk Support Ht. Leg Rest Length Seat to Leg Rest Angle R Hip ab-, adduction L Hip ab-, adduction Tilt in Space Angle Health Outcomes: Physiologic Function Functional Status Morbidity Mortality


24 Patient characteristics such as age, ge nder, ethnicity, medi cations influencing muscle tone, degree of spastic ity, and neurological deficits may also be predictors of pulmonary mechanics. Although effects of wheelchair seating pr operties are unknown, Figure 1 depicts their hypothesized relationshi p as predictors of pulmonary mechanics measured in RAW, VT, MV, and VD/VT. Under these circumstances, pulmonary mechanics may be observed as usual for the individual in an unsupported position or improved from the unsupported position as the trunk is straightened. Pulmonary mechanics may be predictors of the health outcomes of physiologic function, functional status, morbidity, and mortality, but these ou tcomes will not be studied in the proposed study. This is a physiologic model. Currently lacking in the scien tific literature, an intervention study is anticipated, building on this description of relationship of wheelchair seating properties to pulmonary m echanics. The independent variables were the unsupported condition (bas eline) and wheelchair seati ng parameters; the patient characteristics may be treated as covariates in analysis. The de pendent variable was pulmonary mechanics, reflected in RAW, and minute ventilation (MV) will clinically explain some changes in RAW. In the future study, tidal volume (VT) and deadspace to tidal volume ratio (VD/VT) will also be used to explain changes in RAW.


25 Chapter Three: Method Research Design The purpose of this descriptiv e, within-subjects study was to estimate the relative importance of five wheelchair configuration parameters as predictors of pulmonary mechanics, reflected in total airway resi stance of school-aged ch ildren with CP who cannot sit alone but have flexible spines During a single data collection session, participants were placed in a wheelchair simu lator in which five seating parameters were manipulated; pulmonary mechanics served as the dependent variable using the Viasys Jaeger Impulse Oscillometry System, Resp ironics NICO, and Hans Rudolph facemasks (Appendix D). Spasticity was measured using the Modified Ashworth Scale. Patient characteristics and medications were reco rded. A process log was used to capture challenges associated with subject recruitm ent and retention as well as the subjects’ responses to the data collection protocol. A wheelchair seating simulator provide d six conditions of support (four experimental wheelchair parame ters and one condition with all four parameters present, plus one control condition with none), empl oyed independently during configuration of the simulator for each participant. The orde r of introduction of the four experimental wheelchair parameters—two upper extremity s upports; two lateral tr unk supports; level, de-rotated pelvis secured with seatbelt; and tilt in space—was randomized. These parameters were intended to serve as predic tors of change in the dependent variable,


26 pulmonary mechanics, measured in four components, RAW, VT, MV, and VD/VT. However, VT, MV, and VD/VT were included only to provide further clinical explanation of potential differences in RAW. Objectives The study had three objectives: 1. determine the relative contribution of each of five wheelchair configuration parameters to im provement in pulmonary mechanics (total airway resistance (RAW), tidal volume (VT), minute vent ilation (MV), and deadspace to tidal volume ratio (VD/VT)); 2. describe the challenges associated with subject recruitment and retention; 3. discuss the re sponse of children with CP to the data collection protocol. Purpose The purpose of this within-subjects, descrip tive study was to estimate relative importance of five wheelchair configurat ion parameters as predictors of pulmonary mechanics, reflected in total airway resistance of school -aged children with CP who cannot sit alone but have flexible spines. Research Questions The questions to be answered in this study are: 1. What is the relative contri bution of each wheelchair parameter to improvement in pulmonary mechanics? 2. What are the challenges associated with subject recruitment and retention?


27 3. What is the response of children with CP to the data collection protocol? Variables Table 2, below, presents a list of study variables in cluding the domain, operational definitions, data source, and type of variable. Table 2 Variables and Measurement Plan Domain Variable Operational Definition Data Source/Recording Type, Treatment of Variable Age Age in years Continuous Gender Male/female Dichotomous Ethnicity White, non-Hispanic; African American; Hispanic; Other Nominal Medications Names of currently used medications that can reduce muscle tone Interview with parent Nominal Patient Characteris tics Spasticity Greatest am ount of spasticity present in body, as measured by Ashworth scale. Direct observation Ordinal Environment/ Wheelchair Seating Parameters Unsupporteda Seated in wheelchair seating simulator with only seat surface for support and hands of investigator to hold onto surface; without back support, but with le g support. This is the control condition Direct observation; Prairie Reflections Seating Simulator Dichotomous (support present or absent); Predictor variable


28Domain Variable Operational Definition Data Source/Recording Type, Treatment of Variable Partially Supported (Wheelchair Seating Parameters 1 – 4)a Pelvis Support Upper Extremity Supports (2) Lateral Trunk Supports (2) Tilt in Space Support No conditions of support present except head, seat, back, leg and foot support; one condition of partial supportb introduced at a time, then removed for introduction of next condition. Direct Observation Other wheelchair seating parameters held constant in this study. Dichotomous, (support present or absent); predictor variable Seat depth Depth of seat; equal to length of longest of both femurs at hip flexion of 100 degrees Seat width Width of seat, equal to width of buttocks in hip flexion, no wider Back rest height Height of back rest at the participant’s shoulder level Environment/ Wheelchair Seating Parameters Continued Leg rest length (Left and Right leg rests) Length of foot support from the seat surface, equal to length of the lower leg from lower surface of thigh to bottom of foot, for each leg, ankle dorsiflexed Direct observation & recording of presence of these conditions within subjects, as set into seating simulator, using Waugh Protocol (2004). (These conditions will vary between subjects, but will be constant within subjects.) Dichotomous; presence or absence of equal measures; Constant within subjects, across all conditions of support


29Domain Variable Operational Definition Data Source/Recording Type, Treatment of Variable Seat to leg rest angle Angle of leg rest to seat, equal to angle of tibias to femurs of both legs, seated. (90 degrees for all). Right and left hip abduction/adduction Positioning of thighs parallel to seat sides, maintaining seat width as stated above. Tilt in space Seat and back will be tilted as a unit, 30 degrees from vertical for each participant Same as above Same as above Environment/ Wheelchair Seating Parameters Continued Headrest position Headrest will be present throughout all conditions, placed at a height equal to the child’s neck length, occiput to top of shoulders, in same plane as seat back. Same as above Same as above Totally supported (Wheelchair Seating Parameter 5, manipulated in this study)a All conditions of support pres ent Direct observation of presence of all conditions of support Dichotomous (support present or absent); predictor variable


30Domain Variable Operational Definition Data Source/Recording Type, Treatment of Variable Total airway resistance (RAW) Tidal volume (VT) Minute ventilation (MV) Pulmonary Mechanics Dead space to tidal volume ratio (VD/VT) As measured with Jaeger IOS, Respironics NICO, and reusable Hans Rudolph facemask Observation & recording of IOS & NICO measures, in each positioning condition, by instrument instructions. Continuous; Criterion variable Subject recruitment and retention Challenges associated with study recruitment and retention of children and parents to the study. Subject Responses Data collection protocol Child ren’s response to equipment (facemask, simulator, IOS) and data collection procedures. Direct observations recorded in Process Log Qualitative data Notes. a One seating condition (unsupported) an d five wheelchair parameters were manipulated in this study. b With: pelvis secured with seat belt at 60 degrees to horizontal, snug (admits 2 fingers only under belt); anterior superior iliac spines (ASIS’S) level, with a line through both ASIS’s congrue nt with frontal plan e & parallel to level seat surface; armrests placed directly under elbows/under shoulders, at a distance from the seat equal to the height of the elbow from seat when the pelvis is at neutral or in anterior tilt; 2 trunk supports in place directly at alternati ng sides of the trunk, with first one at apex of scoliotic curve if present; the seat and back t ilted as a unit, backward in space, 30 degrees from vertical; and the headrest present, with support height adjusted to the length of the par ticipant’s neck to s houlder, at the occiput, and no recline from vertical.


31 Instruments and Reliability Prairie Reflections Seating Simulator The fully adjustable Prairie Reflections Seating Simulator (Prairie Seating Corporat ion, 2007), a clinical tool, provides multiple surfaces as described by the list of seating pa rameters and variables; all measurements were made manually with a goniometer or metal tape measure. The simulator also has a gauge at the seat to backrest angle vertex gi ving a measure in degrees of tilt in space used to measure 30 degrees backward tilt from verti cal. It was zeroed daily. The investigator was trained and certified to measure and appl y the seating parameters uniformly by two experts in the Waugh protocol (Waugh, 2004; Shaw, 2006) to + or – 10% of the expert’s measurements. The simulator was decorate d with a large span gled hat and ladybug, flower, and bird hand puppets to appear softened and more child-friendly.


32 Figure 2. Prairie Reflections Seating Simulator. The decorated simulator used in this st udy is shown on the left, and a proprietary photograph (Prairie Seating Co rporation, 2007) on the right. Respironics Non-Invasive Cardiac Output (NICO) Monitor The NICO measures flow rate and pressure of respired gas flow ing through a fixed orifice differential pressure pneumotachometer. A capnostat sensor passes an infrared beam through a tubing sample of respired air, w ith carbon dioxide (CO2) sensitively absorbing the infrared light energy, in direct proportion to CO2 concentration. A photodetector measures and subtracts the remaining light energy with re sultant difference equaling the amount of carbon dioxide in the tubing sample. NICO software compensations in the monitor allow accurate computation of volume and flow of respired gas (Respironics, 2007) This technology is distinguished by its infrared beam and effectiv e isolation of the proprietary sensors from


33 respiratory secretions, along w ith its accuracy and point of measurement at the immediate external airway. Accuracy is ensured by several mechanisms. First, the electronic capnostat signal of CO2 concentration is compared to a signal of a known CO2 concentration stored at the factory in th e monitor’s memory, requiring no calibration by the user. Second, single-use flow and pressure sensors are also correlated to factoryinstalled flow measurements stored inside the monitor, and us er calibration is not required, due to ability of the plastic injec tion mold to produce precision sensors. The pressure transducer automatically “zeroes” to correct for changes in ambient temperature and electronics, with software compensati ons. This was done every morning at startup. Third, carbon dioxide elimination is calcula ted based on a mathematical integration of flow and CO2; both measures are obtained from sensors at the same point on the external airway itself, rather than some other point in the tubing (Respironi cs, 2007c; Respironics Novametrix 2003). The sensor and short tubing are disposable. Pediatric size for persons between 15 and 45 pounds (Reference number 97 66) and adult size for persons over 45 pounds (Reference number 9767) were used in this study, after each participant was measured for weight and height. P ublished proprietary data state CO2 sensor accuracy is + or – 2mm Hg for 0-40 mmHg, + or – 5% of reading for 41-70 mmHg, and + or – 8% of reading for 71-150mm Hg. Accuracy of the flow [volume] sensor is + or – 3% of the reading or .5 L/min (Respironics, 2003a). Other NICO measures (MV, VT, and VD/VT) are calculated by the instrument, based on end tidal CO2 (ETCO2) and volume measurement. Minute ventilation (MV) is measured in liters per minute, tidal volume (VT) in milliliters or liters per one breath, and deadspace to tidal volume ratio (VD/VT), or physiologic deadspace, as a numerical ratio,


34 calculated by the NICO to two decimal places. In order to obtain VD/VT, an adjustment value of 5 mmHg for less healthy in dividuals is input to each ETCO2 measure, and this was done for each position, for each partic ipant, according to the manufacturer’s instructions (Respironics, 2007b; G. W illiams, personal communication, October 27, 2004). (Only values for MV we re used in addition to RAW values in the analysis for this study as values for VT and VD/VT were desi gned to assist with further clinical interpretation of the RAW measures.) Independent eval uation of NICO’s precursor, VenTrak1550, found accuracy within clinically acceptable limits (Bak, et al., 2002). The manufacturer’s representative, a biomedical engineer, conduct ed training and certified the investigator in use of the NICO System (Williams, 2006). Both NICO and IOS displayed measurements electronically until recorded and purged.


35 Figure 3. Respironics Non-Invasive Cardiac Output Monitor.


36 The NICO monitor, shown in Figure 3 with cartoon sticker, includes a transcutaneous capillary oxygen concentration sensor (“ pulseox”), which was used to monitor participants’ heart rate a nd peripheral oxygen saturation dur ing the study procedure. Oxygen capillary saturation of 92% or less wa s used to identify desaturation and excuse the participant. This criterion was provid ed by a pediatric pulmonary researcher (P. Kuster, personal communication, February 17, 2005). No child’s oxygen saturation dropped below 95% at any time. Pulmonary Measurement—NICO. The recommended locus for sampling in NICO measurement of VT, MV, and VD/VT is immediately proximal to the airway, and a facemask is acceptable. No difficulty was expe rienced with the NICO in collecting data at 30 degrees tilt in space, because the tubing and sensor do not need to be presented solely in the frontal plane and were quite m obile. Only the flow sensor was used; it was not necessary to use the Fick rebreathing method (with a rebreathing loop) to obtain reliable data for VT, MV, and VD/VT; the rebreathing loop is used primarily to obtain a CO2 baseline for cardiac output measures (R espironics, 2007b). The only requirement was to allow approximately 30 seconds, or 8 breaths, for end tidal CO2 (ETCO2) determination, in order for the VD/VT measurement to be made and display on the screen. If insufficient breaths were exchanged for a determination, there simply was no display until a reliable ETCO2 value had been obtained and input. The NICO manufacturer (Respironics, Wallingford, CN ) states that NICO ETCO2 is 3-5 torr, or mmHg below arterial ETCO2, so to obtain the VD/VT value, ETCO2 must be obtained and 3 to 5 mmHg added to it, with 5 added for individuals who are chronically challenged, and 3 for well individuals. A value of 5 was systematically added to ETCO2 for all VD/VT measures, for


37 all participants. VT, MV, and VD/VT values were recorded as soon as displayed after 30 seconds or 8 breaths to get a representative measure of the parameter efficiently and limit participant burden due to dur ation of the study. For this r eason, no additional time for “settling” was allowed, and the measurement wa s taken immediately as positioned by the investigator. The Viasys Jaeger Impulse Oscillometry System (IOS). The Viasys Jaeger IOS uses a pseudorandom noise signal in a normal tid al breathing column (“forced oscillation technique”) with advanced di gital signal analysis. An attached computer uses Fast Fourier Transforms to fit the forced oscillat ion technique data to algorithms and deliver measurements of airway status during tidal breathing previously available only with more invasive procedures. The IOS allows dete rmination of total airway resistance (RAW) in kiloPascals/Liter/second, with virtually no active patient participation, only relaxed breathing. R5, or resistance at 5 Hz, was used in this study as the measure of total airway resistance. Accuracy of the IOS pneumotach is 0.2-12 Liters/second, + or –2%, with accuracy of the mouth pressure transducer [e rror rate] of + or –2%. Disposable Viasys Microgard microbial filters, reference number 769344G, were used with the IOS. The IOS has been found to offer reliable measures in subjects who cannot comply (Horan, et al., 2001). An expert skilled in use of the IOS trained and certified th e investigator in use of the Viasys Jaeger system at the Shands/University of Florida pulmonary neurophysiology laboratory (P. Davenport, pe rsonal communication, January 26, 2006). An IOS volume calibration check was completed monthly, as recommended by the manufacturer in proprietar y information (Viasys, 2000b).


38 Figure 4. Viasys Jaeger Impulse Oscillometry System. The system is shown here with a filter/mouthpiece, arm, and computer. A facemask/airway is recommended by the manufacturer for IOS use with children; however, it proved unacceptable to the particip ants in this study by stimulating gagging in some. This study’s participants, who have CP and oral incoor dination, were also unable to close their lips around it, to pr ovide a seal, so a noninvasive replacement facemask by Hans Rudolph, Inc., providing an adequate seal, was found. This facemask design produced R at 5Hz (RAW) measures that were diffe rent from the software predicted values, but it was consistently applie d to all participants, who served as their own controls, and it was not used for clinical testing, so the values obtained were deemed


39 reliable. In addition, the pr edicted software values were means for normally-abled individuals, for height, weight age, and gender. No such m eans were offered as predicted values for individuals with disabilities. In order to reduce the s hunt impedance of the cheeks, measurement protocol from the manuf acturer (Viasys, 2000b) requires the patient to hold his hands against his cheeks; however the study participants were unable to do so, and the HRI facemasks passively a pplied this pressure in the study. Figure 5. Child with Viasys Jaeger Impulse Oscillometry System Attached to Facemask. Pulmonary Measurement—IOS. The manufacturer states (Viasys, 2000a, p. 10) that it is possible for small leakages to occur at the corners of the mouth around the mouthpiece and lower the mouth pressure and recorded respiratory resistance measurement; however, the Hans Rudolph oro-na sal facemasks were used instead of the


40 Jaeger mouthpiece and a seal was scrupul ously secured and monitored before measurement took place. With loss of the seal, the flow and pressure trend in the lower half of the screen display is lost, so that a clear display of red pre ssure tracing and green overlay of tidal breathing is not visualized, only an irregular, red, in distinct pattern. When this occurred, the seal was restored and a cl ear tracing visualized; if 30 seconds of data were obtained, the value was displayed onscreen and data were kept. This occurred for 3 participants retained in the completing sample The same was true of obstruction of the airway with the tongue, which was observed on several occasions as an irregular red tracing without its green overlay and with markedly elevated RAW values on the screen display; this occurred for 2 pa rticipants who were excused. In either case, the participant was excused if 30 seconds of valid tracing coul d not be obtained with an observable seal for each parameter. The IOS pneumotach and elbow piece are desi gned to be presented in the frontal plane, with the participant si tting upright. In the 30 degrees tilt in space condition, the facemask of some of the participants lost a seal at the na sal bridge. In these situations, the guardian was recruited to lightly hold the mask down with one finger over the bony prominence of the nasal bridge. The diffe rence in the IOS tracing was immediately restored, and additional pressure was not placed on the cheeks or the airway, so no increase in RAW is expected to have been reflecte d. The IOS also will not display a RAW value if there are insufficien t tidal breathing cycles before attempting to capture data. Then there must be at least 30 seconds for a reliable measurement, which was secured for each participant with the tilt in space parameter.


41 Hans Rudolph Oro-Nasal NIV Facemasks, Headgear, and Custom Connectors. Soft silicone rubber facemasks in various sizes (Hans Rudolph, Inc., 2006) were used to secure tubing to each child’s head, by mean s of an attaching, quick-release, porous mesh headset. Two custom connectors were fabricat ed by the manufacturer and affixed to the NICO tubing and IOS plastic pneumotach and filt er, respectively; to al ternately attach the NICO and IOS to the facemask with a competent seal (Figure 6, below). Figure 6. Four HRI Facemasks. This figure shows one facemask attached to a mesh headset with clips, and two connectors for attaching a facemask to IOS and NICO. Modified Ashworth Scale. Spasticity is defined here as excessive muscle tone. Presence of spasticity was measured with th e Modified Ashworth Scale (MAS) (Haley &


42 Inacio, 1990), a single scale w ith six possible choices (0, 1, 1+, 2, 3, 4; Appendix A) for grading the extent of muscle tone in joints of the extremities. Since trunk control was judged to be a very important contributor to posture and breathing, sc ores on muscle tone at the hip and shoulder joints (proximal to th e trunk, in flexion /extension) were averaged to obtain an overall score for each participant. This was done by the investigator during measurement of height, wei ght, and body dimensions for seating in the simulator. Gregson, et al. (1999) found reli ability very good (kappa = .84 for interrater and .83 for intrarater comparisons). The investigator 's Modified Ashworth Scale scoring was certified by an occupational therap ist skilled in its use (Shaw, 2006). Power A priori power analysis was conducted, assumi ng an expected small effect size (Cohen’s d of .20 for small effect size = 0.1 to 0.3). Therefore, assuming small effect size and controlling for expected power of 0.8 (the condition in which 80% of variance in the dependent variable is explained by the phenomenon(a) of interest), the a priori estimation of sample size was 42. Over a 2month period, 5 participants were recruited for a pilot test of the protocol. This sample size was far less than the a priori sample size expectation of 42. Given that alpha remains c onstant, but with sma ller sample size than originally estimated, power may still be sufficien t if effect size is large. Therefore it is important to reassess power and sample size af ter the pilot study has been completed to determine an effect size which may ultimately be used to estimate the number of participants actually needed in a future study. In or der to know how many more participants would be needed to achieve sufficient power, and with alpha set at 0.05, a


43 post hoc analysis to estimate effect size and pow er was conducted for the pilot. Cohen’s d was used as the measure of effect size, s ubtracting the means of totally supported and unsupported conditions, divided by the pooled standard deviation. This revealed Cohen’s d of 0.69, indicating a medium effect size fo r n = 5; the pilot st udy was expanded with efforts to recruit participants continuing, based on Cohen’s table (1992, p. 158). The table indicates 35 participants are needed for a medium effect size with ANOVA of 6 wheelchair parameters (groups) and alpha set at 0.05. There are different measure of effect size, according to the type of test. Cohen’s d for a simple (two group) t-test and Cohen’ s f for an ANOVA (comparing more than two means) are two examples of effect size measur e that can be used to estimate the power of completed studies. A priori estimates of power are used to plan the number of subjects needed for adequate power in a study and effect size is estimated; post hoc power estimates are used to estimate the power of a completed study in which actual effect size has been discerned for a given sample. For one-way analysis of variance using Cohen’s f, Cohen established small, medium, and large effect sizes as 0.10, 0.25, and 0.40, respectively (Cohen, 1992, p. 157). Inclusion/Exclusion Criteria Inclusion criteria were the following: English-speaking, pre-pubertal children aged 5 to 10 years with CP who do not sit alone, in equal nu mbers of males and females, who have not had a full meal in the previous 2 hours. Exclusion criteria were: children with spinal cord injury, spina bifida/myel omeningocele, or degenerative neurologic diseases, due to potential variety in postur es; children with existing cardiovascular,


44 hematologic, or respiratory path ology; or history of either respiratory illness in the past 3 weeks or of apnea, based on a nursing hist ory and respiratory assessment performed by the investigator onsite; children with a hi story of pain upon body movement; and those with stated, fixed spinal deformity or hi story of spinal surgery, on physical therapy assessment at the clinic visit. Setting The sample for this study was recruited fr om two sites, the outpatient pediatric neurology clinic for the Shands Children’s Hospital in Gainesvill e, Florida, and the outpatient pediatric orthopedic clinics at the Shiners’ Hospitals for Children, Tampa, Florida. The outpatient pediatric neurology clin ic at Shands in Ga inesville is in the pediatric outpatient department of a freesta nding outpatient building on the University of Florida campus, adjacent to a tertiary care teaching hospital in north central Florida. Patients receive either Medicaid or private insurance be nefits. The Shriners Hospital for Children, Tampa, is on the University of Sout h Florida campus and serves all patients at no cost to the family. Patients come fro m all of Florida, south Georgia, and internationally, especially for spinal, ha nd, and foot surgery. At least 50% of the participants reside within 30 minutes’ drive of the clinics. Procedures Approvals Approval was obtained from Institutional Review Boards at both the University of South Florida and the Univers ity of Florida (Appendix B), as well as from


45 the Medical Research Department at the Shin ers’ Hospitals for Children, Tampa, and the supervising clinic staff. Human Subjects Protections The study participants came from a vulnerable population, defined as nonambul atory minor children with relatively high respiratory morbidity and mortality. Some of the children also had cognitive delays and communication disabilities. Prepubertal childre n were needed, because it is possible to use a within-subjects de sign if subjects have flexible sp ines and can assume both control and treatment positions in the wheelchairs, thereby requiring a smaller number of subjects overall. Legal guardians who received a flyer in clinic and indicated interest were visited by the investigator in the clinic patient room, and questions about the study were answered. Guardians supplied informed consent, and assent was also sought from the children at this time, if able to communicate. A copy of the signed, combined informed consent/assent form was given to guardians (Appendix C). Participant Recruitment Enrollment took place over a four-month period. The sampling goal was to reflect the ethnic and gender composition of Florida, 51.2% female; 48.8% male; 65.4% white, non-Hispanic; 16.8 %Hispanic; 14.6% Black or African American; 3.2% other groups (U.S. Census Bureau, 2000). Parents and children were approached through the clinic staff. First, a flyer was offered to the guardian of all nonambulatory, school-aged children by a nurse in the private patient rooms of the outpatient clinic. If the guardian indicated interest, the investig ator went into the room to answer questions. Guardians were consented and assent sought from children. Then the children were screened for study criteria. The study took place by appointment, in the pulmonary neurophysiology research laboratory in the University of Florida Department


46 of Physiological Sciences and in a wheelchai r fitting room in the seating department at Shriners. All guardians but two chose to ha ve the study procedure immediately following the medical visit on the clinic day. Data Collection For each participant, data were collected in one session after recruitment, over approximately 1 to 1.5 hours, with the trunk in each of the six seating conditions. Children were tested no less than 2 hours after eating a full meal. Some had had a small snack in order to get them through the procedure without pronounced hunger, but parents stated that in all cases they “were not full,” which otherwise could alter pulmonary mechanics values. The order in which the six seating conditions were introduced independently was determined by drawing number sequences from a hat. The pulmonary measures of tidal volume (VT), minute ventilation (MV), and deadspace to tidal volume ratio (VD/VT) were collected solely for clini cal interpretation of changes in total airway resistance (RAW). Qualitative data such as ve rbal and cognitive ability were recorded in a study log following the meas urement session and later analyzed. Use of only one data collector the investigator, reduced inter-rater variability of all measurements. Data collection steps are listed in Table 3 below. After admission to the study, the investigator screen ed each participant, measured vital signs, height, weight, and body dimensions for seati ng; collected demographic in formation; inspected and auscultated the child’s ches t; and made a Modified Ashw orth Scale determination.


47 Table 3 Data Collection Protocol Step Details 1. Informed consent obtained, with assent solicited. 2. Screening of participant Respirator y history for recent past also taken, as well as presence of inclusion and exclusion criteria 3. Overall spasticity estimated and reco rded Using Modified Ashworth Scale 4. Patient characteristics, recent health history, and vital signs recorded 5. Mat examination to measure: elbow height above seat; hip, knee and ankle flexion angles; trunk length; thigh length; and lower leg length According to Waugh (2004) protocol purpose: to formulate seating simulator dimensions for seating conditions; angles measured with a manual goniometer; height and lengths measured from joint to joint with a metal tape measure, according to published protocol 6. Facemask explained to the child; child allowed to handle it, and mask placed 30 seconds allowed to become accustomed to mask


48Term Definition 7. Partially-supported conditions applied in random order, independently. Using the NICO and the Jaeger IOS, pulmonary mechanics (4) measured in each partially-supported condition. Position condition noted for each. Each instrument digitally displays measurements. Investigator records measurements onto data flow sheet, along with PaO2. For NICO, 8 breaths must elapse, approximately 30 seconds, before ETCO2 analysis allows VD/VT display. For IOS, 3 full breaths precede data collection of RAW (R @ 5 Hz), which takes at least 30 seconds of clear flow cycle tracing, according to manufacturer’s instructions. If tracing shows obstructed pattern or lack of seal, data cannot be collected until at least 30 seconds of clear tracing are obtained. 8. Fully-supported and unsupported seating conditions are applied in simulator, with pulmonary mechanics measures recorded Measurements taken immediately, upon positioning The following should be noted: The f acemask was positioned snugly with intrinsic Velcro straps and plastic clips. Recording of each pulmonary mechanics measure with the NICO and IOS occurred im mediately following posit ioning in each of the six conditions of support for each child. As soon as the position condition (wheelchair parameter) was introduced, the 30-second period began for the NICO or 3 breaths for the IOS. The same instrument as for the last measure of the last position began the next


49 condition, to minimize skin breakdown due to friction from excessive switching of the tubing on each silicone rubber mask. That is, the starting instrument of the two (NICO or IOS) was alternated, in each condition. In the “unsupported” condition, the ch ild received no extrinsic trunk support except from the seat surface of the simulator and leg support, plus the guardian’s hands holding the thighs on the seat surface to pr event falling. Partially and totally-supported conditions included optimal positioning in th e Prairie Reflections Seating Simulator according to the Waugh protocol (2004). For all conditions of “partially-secured”, one condition of support at a time was imposed. Ob servational monitoring for discomfort and capillary/pulse oximetry detection of any de saturation occurred thr oughout the procedure. The investigator was assisted in positioning by the guardian, who held the child on the seat in the unsecured position and in some cas es, held the top of the facemask lightly on the nose in the tilt in space position, to maintain a seal. The seal was validated by the flow tracing display observed by the investigator. (Lack of se al is displayed as jagged, red, irregular tracing.) Angles of seating parameters (seat to back angle, tilt in space) were measured by manual goniometer, in degr ees, as trained by Waugh: participant ankle, knee, and hip flexion/extension ra nge of motion were ascertained with body measurements at the start of the session, and joint angles were held constant for these settings, for all participan ts. These were 90 degrees ankle and knee flexion and 100 degrees hip flexion. Linear meas urement was by tape measure in inches (back rest height, headrest height, seat depth and width), matc hed to body dimensions, also as trained by Waugh. The hips were positioned against the back rest, with slight anterior pelvic tilt for secured, level pelvis. Participants not allo wing measurement due to invasiveness of the


50 facemask were noted. Any two episodes of desaturation comprised stopping criteria for the study, but none occurred. A “safe” environment was provided in the seating department at Shriners, a signal to these children that they were in a place friendly to children. Stuffed animals were placed on the mat table in the room, cartoons were visible on a television overhead, 4 hand puppets were placed on the simulato r, and a large floppy, sequined hat was available to wear. Choice of a new toy was offered to each child engaging in the procedure, provided by the hospi tal child life specialist. Data Analysis Data are presented descriptiv ely for categorical variables as percentages and as means and standard devia tions for continuous variables. A content analysis of the study log was conducted to answer study questions 2 and 3 using observation for themes and coding. For study que stion 1, the Statistical Package for the Social Sciences (SPSS version 14.0) was used for data management and analysis. The data were examined for missing values, outlier s, and inconsistent data as recommended by Tabachnick & Fidell (2001, pp. 56-9), with a scatterplot executed to identify subjects. To determine the relationship between all six wheelchair seating parameters and total airway resistance, a within-subjects, one-way repeated measures analysis of variance was conducted, with six conditions of wheelchair parameter configurati on and total airway resistance serving as independe nt and dependent variables, re spectively. Type 1 error, or alpha, was set at 0.05. Type 1 error is the probability of rejecting a true null hypothesis (Ho), in effect saying a difference between groups (or treatmen ts) exists when, in fact, one does not. In statistical analysis, the investigator sets al pha, and it is the chance of making a Type 1


51 error, expressed customarily as a percentage such as 0.01 (1%). Another way to say this is that Type 1 error is called alpha and can be decreased by al tering the significance level. For example, if alpha is set at 0.01, there is a 1% chance that the result termed significant would occur by chance alone; the greater the pe rcentage, the greater the possibility that finding a difference is due to chance, rather than to the treatment. Type 2 error, or beta, is the opposite or the probability of accepting the null hypothesis, saying no difference exists between groups when, in fact, one does. Since only one of these errors can occur in a study (Stevens, p. 122), there is an inverse relationship betw een the two. As we control Type 1 error (reduce it), chance of making a Type 2 error increases, and the reverse: when the chance of a Type 1 error increases, the chance of a Type 2 error decreases. The percentage of participants who ha d difficulty with the facemask was noted. Optimal time intervals for introduction of seating parameters and measurement of pulmonary mechanics were determined by noti ng total time elapsed in the procedure, due to the need for measurement immediately upon application of the pa rameter. Participant recruitment and retention rates were calcula ted by computing: the ratio of number approached to number recruite d, ratio of number recruite d to number completing study procedure, and ratio of number completing st udy procedure to total study time elapsed (4 months). Necessary sample size for power of 0.8 was 42.


52 Chapter Four: Results The study results are presented in Chapte r Four. These include a profile of the sample, results of the data analysis, and fi ndings, by study question. In order to evaluate the method, it is useful to also consider th e characteristics and obser ved experience of the excused participants, so this chapter also includes a summary of the profile of the excused participants; all participants were consented before data collection. Sample Characteristics A sample of 8 children was recruited; th ese were prepubertal male and female children (ages 5-10 years) who lack trunk contro l and have a medical diagnosis of CP, as stated in a current physical therapy asse ssment and confirmed by the guardian. There were equal numbers of males and females. Fi ve of the eight participants were 10 years old (the mode); the median age was 8 year s; the mean was 8.9 years. There were 6 Hispanic participants, 1 Caucasian, and 1 Af rican American. Only two participants of eight were receiving any medication, oral Va lium, Baclofen, and Trileptal; this was far fewer than expected for special-needs ch ildren. Extent of spasticity was evenlydistributed, with 2 participants with considerable tone (sco re of 3), 3 with moderately increased tone (score of 2), and 3 with only a sl ight increase in tone (s core of 1). (A score of zero on the Modified Ashworth Scale de notes no increased muscle tone, so each


53 participant had some abnormally increased tone.) No data were missing. A summary of the sample composition is located in Table 4. Table 4 Sample Composition Characteristics Frequency Completing Excused Total n Age in years 10 9 8 7 6 5 Gender Female Male 8 5 0 1 1 1 0 4 4 8 0 1 4 1 0 2 3 5


54 Characteristics Frequency Completing Excused Ethnicity Hispanic 6 3 Caucasian African American Other Medication Use Modified Ashworth Scale score (spasticity) Slightly increased tone (score 1) Moderately increased (score 2) Considerably increased (scores 3 and 4) 1 1 0 2 3 3 2 3 1 1 6 1 2 5 The excused participants were similar in ge nder, age, and ethnicity although there were fewer Hispanic participants. These partic ipants were notably different, however, on medications, spasticity, and ve rbal ability, with 6 of 8 excused part icipants receiving medication on a regular basis; 5 of the 8 with high levels of spasticity; and 7 nonverbal


55 with only 1 verbal. Data ha s been included only for cons ented participants, and all excused participants were consented. Verbal Ability and Motor Control The cognitive ability of these children was notable. Although cognitive testing data were unavailable, 75%, or 6 of 8 part icipants who completed the study were verbal communicators. Two participan ts appeared to have low cognitive function and were nonverbal but completed the procedure; one nonverbal child’s guardian showed remarkable rapport and ability to communicate with her. Nearly the opposite was true for excused participants. Of these, 6 appeared to have very low cognitive function and 88%, or 7 of the 8, were nonverbal. Only two excuse d participants, or 12.5% of the 16 enrolled, appeared to have moderate cognitive function, with some receptive language and ability to cooperate with the procedure but very little motor control to allow stable positioning or pulmonary measurement. Verbal ability did not necessarily corres pond to greater age; 2 of the younger participants were quite verbal and enthusiast ic, and two 10-year-olds were noted to have very limited communication skills. Table 5 shows study completion, age, and verbal ability.


56 Table 5 Study Completion, Age, and Verbal Ability Age in Years Verbal Ability* Completing 10 10 10 10 6 7 10 10 V V V V V V NV NV Excused 5 7 8 8 9 5 8 8 NV NV NV NV NV NV NV V Note. *V = verbal; NV = nonverbal Findings Due to limited time and resources and the challenges associated with recruiting the sample of 42 needed to sufficiently pow er the study, the analysis was limited to


57 descriptive statistics and quali tative data from the proce ss log. The research questions were: 1. What is the relative contribution of each wheelchair parameter to pulmonary mechanics? 2. What are challenges associated with subject recruitment and retention? 3. What is the response of children with CP to the data collection protocol? Research Question One: Contribution of Wheelchair Parameters to Total Airway Resistance and Minute Ventilation The within-subjects, one-w ay analysis of variance revealed that for this sample, total airway resistance varied by seating parameter. Although these data are dependent, research has shown that using Cohen’s d and f is appropriate to reflect effect size in this situation (Becker, 1999). Cohen’s f for effect of all six conditions of wheelchair parameters on RAW, considered simultaneously and using eta squared was found to be 0.27, a medium effect. Power was low, 0.46. Using Cohen’s table (1992, p. 158), this post hoc analysis of the pilot sample size confirms the need for 35 participants to ensure power of 0.8. Power is the probability of detecting a difference if such exists; anything that decreases the probability of a Type 2 error increases pow er and vice versa. More power means that one is more likely to reject the null when it is false. If beta is subtracted from perfect probability, 1.0, the resu lting number is the probabi lity of rej ecting the null hypothesis when it is false, that is, of making a correct decision. This is 1 minus beta, or power, also called the probability of correc tly detecting a relationship between the two variables if a relationship is present. Effect is the strength of this relationship. Power is dependent on the significance (alpha level) set, sample size, and effect size. (The


58 statistical test and research design can also in fluence power.) When the sample is small, it is easier to decide erroneously that a difference exis ts between two groups (reject HO), that is, to have insufficient power. As the size of the effect increases, it becomes easier to detect accurately in a smaller sample; thus, pow er is increased. Effect size is the degree to which the null is false, the magnitude of the effect of an independent variable on the dependent variable, or strength of relationship between two variables. Total airway resistance varied by seating parameter. Mean RAW and MV for each seating parameter are depicted in Figures 7 and 8 below. The MV means show that participants maintained minute ventilati on essentially uniformly, for all seating conditions, to within 1 standard deviation, although RAW varied by seating condition. (VT and VD/VT are not reported here, as they we re only relevant to explain differences clinically, if sufficient power had been achieved.) The seatin g parameters responsible for the lowest (best) total airway resistance we re upper extremity suppor ts and lateral trunk supports.


59 Figure 7. Total Airway Resistance Means by Seati ng Parameter (+ or SD), (n = 8). Seating parameters are: 1 = unsupported; 2 = totally supported ; 3 = upper extremity supports; 4 = lateral trunk supports; 5 = secu red, level pelvis; 6 = tilt in space. When airway resistance was highest (s eating parameters 1, 2, 5, and 6), minute ventilation was maintained. TOTAL AIRWAY RESISTANCE BY SEATING PARAMETER0 2 4 6 8 10 12 14 16 18 NoneAllArmrestsLat trunkSeatbeltTiltSeating ParameterTotal Airway Resistance (RAW in kPa/L/s + 1 SD)


60 MINUTE VENTILATION BY SEATING PARAMETER2 3 4 5 6 7 8NoneAllArmrestsLat trunkSeatbeltTiltSeating ParameterMinute Ventilation (MV in Liters/Min+1SD) Figure 8. Mean Minute Ventilation by Seating Parameter (+ or -SD), (n=8). Seating parameters are: 1 = unsupported; 2 = totally supported ; 3 = upper extremity supports; 4 = lateral trunk supports; 5 = s ecured, level pelvis; 6 = tilt in space. A scatterplot of all measures of RAW was constructed to identify subjects and inspect distribution of data points, depict ed in Figure 9, below. The parameters responsible for the highest RAW measures were “unsupported, ” “totally supported,” and “secured, level pelvis.”


61 MINUTE VENTILATION BY SEATING PARAMETER0 2 4 6 8 10 12 14 16 18 20 22 24 26 0123456 Seating ParameterRAW Figure 9. Scatterplot of RAW by Seating Parameter (n = 8). Seating parameters are: 1 = unsupported; 2 = totally supported ; 3 = upper extremity supports; 4 = lateral trunk supports; 5 = s ecured, level pelvis; 6 = tilt in space. Using the means and standard deviations in Table 6 below, the strength of the relationship between total ai rway resistance and seati ng conditions, represented by computed effect size as Cohen’s f, for this sample was found to be 0.27, a medium effect with low power (0.46) when Type I error rate was set at 0.05 and a within-subjects ANOVA was used.


62 Table 6 Effect of Seating Condition on RAW Means (SDs) Position/Condition RAW Mean (SD) 1. Unsupported 9.84 (7.06) 2. Totally secured 8.61 (5.18) 3. UE supports 6.20 (2.25) 4. Lateral trunk 6.94 (3.15) 5. Secured, level pelvis 10.27 (5.83) 6. Tilt in space 10.79 (3.20) F 1.51 df num 5 dfdenom 42 p-value 0.25 Qualitative Data from the Process Log The remainder of this chapter addresses qualitative data and findings from the study pr ocess log that answer Research Questions Two and Three. The process log data were coded and summarized according to emerging themes which are summarized in Table 7 on the following pages.


63 Table 7. Process Log Themes and Data Theme Data Summary, ( n = 8) Challenges of Participant Recruitment and Retention Postural instability The first two participants had severely a lternating muscle tone, were placed in the unsupported position first, and may have had a decreased ability to tolerate the procedure, in a strange pl ace, if unable to feel secure in the upright position without any postural su pport except the seat. It is known that many children with CP have difficulty orienting in space and holding themselves upright. In order to minimize participant burden, the totally supported and unsupported pos itions were moved to next-to-last and last, respectively, for each participant. This change eliminated identified postural instability, although it also eliminated randomization of these two parameters. Verbal ability As described in the sample prof ile, participants with greater verbal ability tolerated the protocol more often than others. Reflexive activity Four participants demonstrated primitiv e reflex activity that is common in many persons with CP. One exhibited a startle each time the IOS was introduced, but was able to complete the protocol after acclimating a few seconds. One participant experienced pers istent gagging and one participant had repeated sneezing, both likely due to facial stimulation of the facemask. One of these two was excused; the other was able to complete the protocol. One participant experienced persistent t ongue thrusting; this interfered with a clear RAW tracing at the end of the protocol, so he was withdrawn by the investigator.


64Theme Data Summary, ( n = 8) Clinic recruitment Due to the fact that these participants had come to clinic intending a medical visit, the investigator was obliged to wait until the morning or afternoon clinic was completed by the family; this could take up to 4 hours. At one point, a clinic staff member stated that she ha d not alerted the investigator to one potential participant out of courtesy because the family had traveled a distance. It is unknown whether this affected recruitment for others. In addition, one spine surgeon left the medi cal team; although clinic staff stated there was no decline in clinic appointments, it is unknown whether this affected recruitment. Airway obstruction Only one participant was excused for identifiable airway obstruction that interfered with IOS RAW reading. Obstruction can be identified by the fairly straight red line of the IOS oscillations in the respiratory cycle, with no green line. This participant experienced repeated gagging and may have also retracted his tongue, resulting in distinguishable upper airway obstruction on the tracing. Intolerance of facemask, simulator, or IOS. Four participants of 16 would not accep t the facemask; two were unable to tolerate the simulator sufficiently to sit on the surface of the seat, and two were unable to accept the IOS. All of these were excused.


65 Data Collection Protocol (Response of Participants to) Situational characteristics Three participants had some difficulty ma intaining the facemask seal at the nasal bridge in the tilt in space position, or 30 degrees from vertical, for which the IOS was designed. This was handled by having the guardian place slight finger pressure at the nasal bridge, a bony prominence, to contact the nose and seal the facemask to the nose and face. Because the finger pressure was slight and was applied over a bony prominence, it was not believed to increase airway resistance. One participant also had extreme lack of tr unk control, so that his unsupported position had him facing the floor; in this position, IOS measurement was impossible, so the guardian held his shoulders up for this condition, without any other support. The resulting position was equivalent to th e slumped position of each of the other participants. There was also a need fo r participants to have as pleasant an experience as possible, while not interfe ring with measurement, so scented lip balm (chocolate cherry, mint, or berry) was used on the facemask, as well as decoration of the simulator for children w ho could see. The Disney channel also played on a television above the instruments. Over the 1to 1.5-hour period, 4 participants developed redness over the nasal bridge from the facemask; some of the lip balm was applied, and no one deve loped any skin breakdown. When upright at 100 degrees hip flexion, for 2 participants without head control, the Whitmyer small headrest did not fully support the head. Guardians held these participants’ occiputs in the headrest, making this positi on equivalent to the other participants without placing stress on the anterior neck and airway. Three participants also were unable to balance their arms on the upper extremity supports alone; guardians held their arms on the armrests to duplicate the posture of others.


66Data Collection Protocol (Response of Participants to) Fatigue Six of eight participants either stated or a guardian stated they were tired at the end of data collection. Increased respiratory secretions One participant experienced increased secr etions; this girl had a history of gastroesophageal reflux. Her peripheral oxyg en saturation level dropped briefly to 95% when her neck was in extension, but then rose to an acceptable level for protocol completion. Research Question Two: Challenges Associ ated with Participant Recruitment and Retention Twenty guardians received study information; 2 of these declined to participate; 2 were excluded due to not meeting study criteria; 16 enrolled; 8 were retained. Participant recruitment and reten tion themes from the Process Log follow. Postural instability was a factor in retention: when the first 2 participants became agitated when their posture was destabilized by rem oving support at the outset of the protocol, they were excused. This change eliminat ed the opportunity to randomize order of presentation of totally supported and uns upported parameters. The second theme was verbal ability; participants w ith greater verbal ability tolerated the protocol more often than others. Third, 4 participants demonstrated primitive reflex activity: startle, persistent gagging, repeated sneezing, and persistent tongu e thrusting. This reflex activity stopped the protocol for half of these participan ts, who had to be excused. Fourth, clinic recruitment was a factor; participants were not available until the end of the clinic duration, which averaged 3.5 hours. It is unknown whether length of time in clinic also influenced staff not to approach potential pa rticipants, or whether loss of one surgeon decreased the number of potential participants in clinic. The fifth challenge to participant


67 recruitment and retention found in the Pr ocess Log was airway obstruction, which occurred for only 1 participant who experienced repeated gagging that interfered with IOS measurement. He may have retracted hi s tongue and involuntarily, briefly obstructed his upper airway, resulting in his being excu sed. Finally, intoleran ce of the instruments was a factor. Of 16 participants, 4 did not to lerate the facemask; 2 did not tolerate the seating simulator, and 2 did not ac cept the IOS; all were excused. Research Question Three: Response of Ch ildren with CP to the Data Collection Protocol The challenge posed to recruitment and retention by in tolerance of the facemask, simulator, or IOS, also was a res ponse of participants to the data collection protocol. Other themes were situational char acteristics such as difficulty maintaining the facemask seal for IOS measurement at 30 degr ees tilt in space, ex treme lack of trunk control of one participant, skin redness at the nasal bridge from the facemask, and poor support of the headrest for part icipants with extreme lack of head control. In addition, three participants were unable to balance their arms on the armrest without having them held. Of 8 participants, 6 stated, or a guardian st ated, that they were tired at the end of the protocol. Finally, 1 participan t experienced increased resp iratory secretions, and her peripheral oxygen concentration dropped brie fly to 95%. She was, however, able to complete the protocol. Seating Simulator and Facemask Acceptability Of the 16 partic ipants enrolled, 2 (12.5%) were unable to tolerate the simulator and assumed a total ex tension posture, with high spasticity and inability to maintain the position for measurement, so they were excused. Of the total 16 enrolled, 4 participan ts (25%) either cried or responded to the mask with reflexive activity such as ga gging, tongue thrusting, or spastic movement


68 preventing measurement; they were also excuse d. Conversely, 12 (75%) of those enrolled found the facemask acceptable. Other Process Log Themes and Data In addition to data about optimal measurement intervals; recruitment and retention; simulator, IOS, and facemask acceptability; and postural instability, othe r themes were also found: situational characteristics of the method; fatigue; verbal ability; reflexive activity related to CP; airway obstruction; and increased airway secr etions. There were va rying amounts of head and trunk control. The facemask seal was difficult to maintain in 30 degrees tilt in space. Uniform application of seati ng parameters was difficult when participants had extremely low muscle tone, complete lack of head a nd trunk control, and movement of arms off upper extremity supports; these were situational characteristics of the protocol when used with persons with CP. Of the participants, 75% experienced fatigue at the end of the procedure. Participants with greater verbal ability tolera ted the protocol better than others. Reflexive activity such as startle and sneezing, occurred but di d not interfere with data collection. Gagging and tongue thrusti ng, however, prevented data collection. Airway obstruction occurred briefly with gaggi ng, with which tongue retraction may also have been associated; this made IOS measur ement impossible for 1 participant. Increased respiratory secretions occurred with only 1 pa rticipant and did not interfere with oxygen saturation level. Discussion of Findings These findings will be discussed in the next chapter, which includes contribution of each wheelchair parameter to pulmonary mech anics; effects of data collection protocol


69 on subject recruitment and retention; s eating simulator and facemask acceptability; reliability of pulmonary measurement; optim al measurement intervals; participant recruitment and retention rates; differences in total airway resistance (RAW) by wheelchair parameter; study limitations; r ecommendations for future research; and conclusions.


70 Chapter Five: Discussion The purpose of this within-subjects, desc riptive study was to estimate the relative importance of four wheelchair configurati on parameters as predictors of pulmonary mechanics, reflected in total airway resi stance of school-aged ch ildren with CP who cannot sit alone but have flexible spines. Th is chapter will discu ss the contribution of each wheelchair parameter to pulmonary mechanic s, effects of data collection protocol on subject recruitment and retention, seati ng simulator and facemask acceptability, reliability of pulmonary measurement, optimal measurement intervals, participant recruitment and retention rates, differences in total airway resistance (RAW) by wheelchair parameter, study limitations, recommendations for future research, and conclusions. Contribution of Each Parame ter to Pulmonary Mechanics For this sample, the lowest RAW means were seen with i ndependent application of the upper extremity supports a nd lateral trunk supports. It is possible that upper extremity and lateral trunk supports contributed most to trunk straightening and support of the upper trunk, lifting weight off dependent l obes and reducing resistance to flow throughout the entire tracheobronc hial tree. Straightening the spine could also remove the load arm on the spine placing force on the lower lung fields. Highest RAW means were seen with secured, level pelvis and tilt in space. It is possi ble that derota ting the pelvis and simply securing with a seatbelt, while fo rming a base of support for the spine, does


71 not remove the load arm of the upper trunk suffi ciently to straighten the upper spine and reduce force on the dependent portions of the lu ng, alveolar resistance, and total airway resistance. It is also possible th at in tilt in space, the tongue s of these participants fell to the back of the throat, increasing upper and to tal airway resistance. It was impossible to ascertain tongue position from outside, but the one person who s howed full obstruction was in the tilt in space position. The two next lowest RAW measures were in the totally supported position and the unsupported position. Minute ventilation was maintained even in the presence of the higher airway resi stance measures, possibly indicating relative relaxation and low fear. In the presence of high RAW, minute ventilation will fall, even with increased rate of breathing. These findings show trends only for this small sample. With sufficient power, ot her findings are possible. Effects on Subject Recruitmen t, Retention, & Responses to Data Collection Protocol Seating Simulator and Facemask Acceptability Characteristics of participants who accepted the seating simula tor and facemask--those completing the protocol in this sample--were higher verbal ability, lower levels of spasticity, and freedom from medication use. Participants with less disabi lity handled the challenge of the procedure more easily, but the intent was to study a very disabled population who could benefit most from future interventi on. For the excused group, with gr eater disability, the seating simulator and facemask were poorly accepted. It may be possible, however, to conduct a similar study with participants lacking full tr unk control, but having greater verbal ability, such as frail elderly persons with spinal cord injury, muscular dystrophy, or multiple sclerosis, to estimate effects of wheelc hair parameters on pulmonary mechanics.


72 Postural Instability and Unsupported Position The earliest 2 participants had extremely athetoid movement, were in consta nt motion, and were initially placed into the unsupported position (baseline) first, which esse ntially destabilized the participant. These participants responded with incr eased movement in the simulator and some agitation that prevented pulmonary measurement of tidal breathing, so they were excused. The remaining 14 participants who were recruite d received random presentation of the four partially-supported conditions, with the fully supported and unsupported conditions applied last, and this problem was resolved. Of these 14 participants, 6 were excused for other reasons as re ported previously. Usefulness of Pulmonary Measurement Measurement of RAW with the IOS was difficult, but obtainable. First, the 30-deg ree tilt in space seating condition did not accommodate the IOS connection well; it was n ecessary to hold the connection together manually and for some, to hold the facemask to the nasal bridge to maintain the seal. The nasal bridge is bony, so it would not be po ssible to exert additi onal pressure on the airway and result in erroneously higher RAW measurement. This problem could be addressed by tilting only 15 degrees, rather th an 30; this would still provide substantial tilt from vertical in seating to help the individual stay in th e seat and isolate this condition for measurement. Second, difficulties with obstruction were overcome by astutely watching the tracing and only including 30 seconds of good data. If 30 seconds of good data could not be obtained, the participan t was excused, but excusing participants contributed to low power. Third, several part icipants experienced difficulty directly related to their disabilities, such as atheto id or reflexive movement; they were excused, and this also contributed to low power. This issue could be addressed in a subsequent


73 study by employing a sample of persons lacking trunk control but without CP, so that the reflexive activity associated with CP woul d not be seen. Finally, the most serious challenge was to get a reasonable RAW measurement using the facemask instead of a mouthpiece: Measurements were not exactly th e same as those predicted by the software, but they were generally within 1 to 2 kPa/ L/sec. Since this st udy was a within-subjects design, participants were compared to themselv es rather than to a clinical standard. The NICO had no similar issues and was us ed easily; however, it cannot measure RAW. Although research exists on use of the Jaeger recommended f acemask with oral tube, this equipment is not useable in persons with CP. In addition, there is no research to date in normally-abled children on use of the Hans Rudolph facemasks and headsets, which compressed the cheeks and provided an excellent seal with the IOS. Using the facemask in another population with le ss disability could allow hypothe sis testing of effects of wheelchair parameters on RAW, as well as establish confidence intervals. This study’s procedure was sufficiently straightforward to allow MV to continue virtually unchanged, even when airway resistance increased, indicating anxiety wa s not a factor for participants who were able to complete the protocol. The Prairie wheelchair seating simulator wa s sufficiently operable to consistently apply the seating conditions for pulmonary measurement. The participant who had to have his arms held on the upper extremity supports still received support for his upper trunk from the armrests equivalent to other participants. The condition was consistently applied although his arms were kept from sli pping off. The participant who had such low trunk control that he had to have his shoulders held up was actually placed in a posture that other participants in the sample intr insically possessed, so his seating condition was


74 consistently applied, as for other participan ts. All participants tolerated the unsupported seating condition when it was applied last. Optimal Measurement Intervals Preserving skin integrity was a potential issue with the facemasks at the nasal bridge, with redness noted at the e nd of the procedure for 4 participants. This resolved in a few minut es, but Chap-Stick was applied to lubricate and reduce friction; no participant developed a skin break. In the interests of skin integrity, the procedure time was kept as short as could be accomplished. However, some participants still showed signs of fatigue, such as verbalizing being tired and dark circles under the eyes. Due to priorities of the patient, the clinic schedule was honored before the participant c ould enter the study procedure, so 5 of the sample did not begin the study procedure until the end of clinic; th e average length of time the child had been in clinic was 3.5 hour s. The other 3 had made appointments to return specifically for the study procedure at a preferred time. It would be difficult to compare RAW data for these participants to participan ts without disabilities due to lack of reference values for school-aged children using facemasks. One-half of the sample experienced fati gue, but it was impossible to shorten the duration of the protocol, which was not unduly burdensome, considering the usual duration of an outpatient clin ic visit, the average durati on of conventional pulmonary function testing (up to one hour), and the study protocol’s lack of invasiveness (wearing a face mask, sitting in a chair, and breathing) The measurement intervals for this sample were no longer than immediately following pr esentation of the parameter, due to the length of the study protocol ( one presentation after another), fatigue, the possibility of movement out of the position, and the fact that pulmonary mechanics measures are


75 available immediately. The protocol took no less than one hour; the longest duration was 1.5 hours. The investigator was familiar with all the instruments and prompt in manipulating them. A better arra ngement may be to conduct th e study at the beginning of clinic, rather than at the end, or to work with a less disabled population who could still offer lack of trunk control, to measure e ffects of wheelchair seating parameters on pulmonary mechanics. Optimal measurement in tervals would be less than those allowed by the 1-1.5 hours session durati on. That is, if protocol duration could be shortened, optimal measurement intervals would be sm aller; however, shortening these was not possible in this protocol. Participant Recruitment and Retention Rates In twelve, 8-hour clinic days at Shands over 2 months, only 2 participants were recruited; both were excused. In 35 days of pediatric orthope dic CP and “spinal” clinics at Shiners, over 4 months, 1594 outpatient visits were registered for the CP and “spinal” clinics, but the overwhelming number of th ese were ambulatory. Before and during the study, no breakdown of patients by ambulatory stat us was available, but the consensus of clinical staff was that the spin al pediatric orthopedic clinic would be a rich resource for study participants. During the study, nursing staff approached 23 families and reported that only 3 guardians of nonambulatory childr en with CP aged 5 to 10 years refused to hear more about the study from the investigat or. Each family was told that refusal to participate would not result in loss or reduction of any health care benefits. Of the 20 who received study information and were consente d, 16 enrolled, 2 were excluded due to not meeting study criteria; the schedule of 1 enrolle e conflicted with that of another enrollee,


76 who had arrived at the only time the former enrollee could participate. One remaining family declined to participate, stating th eir child had experienced many surgeries and would not cooperate with the mask. Sixteen ch ildren with a medical diagnosis of CP, who cannot sit alone, were enrolled to participate in th is study. The ratio of those enrolled to those approached is 16 to 23, or 70%. Eight ch ildren were excused, leaving a sample of 8 children who completed the study: the ratio of the number retained to number recruited is 8 to 16, a 50 % retention rate. Eight part icipants completed the study over a 4-month period, or 8 participants/4 months. There was a preponderance of ambulatory children who did not meet study criteria at both site s, with a correspondingly small number of children meeting study criteria. The small sample was associated with stringent study crite ria, which were intended to control confounding variables in measurement of pulmonary mechanics and posture as well as excusal of participants unabl e to tolerate or comply with the protocol. A better solution may be to study a less disabl ed group such as older adults, who still present lack of trunk control, or to iden tify a larger outpatient clinic base of nonambulatory children with CP. Differences in Total Airway Resistance (RAW) By Wheelchair Parameter Due to low power associated with sm all sample size, it was not possible to conduct hypothesis testing regarding RAW and wheelchair parameters. This should be done when power is achievable.


77 Study Limitations Sample In order to achieve power of 0.8.t o detect a small effect, sample size would need to equal 42. Low enrollment and a 50 % attrition rate contributed to low power. There is no known reason for low enroll ment except for the stringent inclusion criteria. The investigator was present in th e clinic area and observ ed no patients who met inclusion criteria who were not approached by clinic nurses and offered a flyer. Explaining the preponderance of Hispanic pe rsons is difficult, especially since no available breakdown of the Shriners outpatien t population by ethnic group is available, but ethnicity is not a known factor in CP th at would bias pulmonary mechanics findings. Finally, the sample was limited by the particip ants’ low verbal ability. Eligible nonverbal participants tended not to complete the prot ocol, thus reducing the study’s retention and power. Method Because selection was not randomized, there may have been some bias in recruitment of participants. Some desirabl e subjects were selected out by their level of ability to comply with the st udy protocol or possibly how long they had been in clinic and how far they had traveled. Randomizing th e full complement of wheelchair seating parameters was not possible, only the order of presentation of the f our partial supports, due to destabilizing some children when th ey first began the protocol with the unsupported parameter, which was burdensome. This may have introduced some bias. Characteristics of persons with CP were the very attributes that complicated this method: reflexive activity, severely absent head and tr unk control, and lack of verbal ability. In addition, despite every effort to observe obj ectively, the complexity of the procedure made detailed observation challenging. A s econd observer/recorder would have been


78 useful. Further, the method was unwieldy and ma y be difficult to replicate. With only one data collector, there was no estimate of measur er-to-measurer variability, so information about interrater reliability is not available, and co llection of descriptiv e data was limited. Finally, effect of residing locally is unknown. Having traveled a distance to clinic, staying in unfamiliar surroundings, and undergoing unfamiliar routines may have contributed to fatigue or to higher stress on participants, thereby perhaps resulting in intolerance of the protocol. Possible relation between residi ng locally and completion of study protocol were not analy zed, and sample size was limited. Recommendations for Future Research Nurses should continue to investig ate wheelchair parameters, posture, and pulmonary mechanics. A different group within the population of persons lacking full trunk control should be studied, to remove the co nditions of the disability that complicate study of persons with CP. Another possibility w ould be to recruit a larger, verbal sample of children in a clinic with known numbers of nonambulatory patients and to conduct the study at the beginning of clinics rather than at the end. Although the IOS is marketed as a device that can be used for individuals w ho cannot comply with traditional pulmonary testing, the recommended mouthpiece did not prove useable with this sample. Other methods need to be explored. Finally, reference values for Hans Rudolph facemask and headset use should be developed for normally -abled children, to compare the facemask and mouthpiece protocols.


79 Conclusions By study question, the following conclusi ons were drawn for this sample: 1. What is the relative contribution of each wheelchair parameter to improvement in pulmonary mechanics? a. For this sample, RAW was lowest in the presence of upper extremity supports and lateral trunk supports; MV was maintained, even when RAW was highest. 2. Using the IOS, NICO, Prairie seat ing simulator, and Hans Rudolph facemasks, what is the effect on s ubject recruitment, retention, and response to the data collection protocol for RAW, VT, MV, and VD/VT, in this population? The challenges associated with subj ect recruitment and retention were: a. The level of disability of the participants. b. The data collection protocol, particul arly the seating simulator, IOS, and facemasks used, for nonverbal children and those with primitive reflex activity, and presentation of the unsupported position first. c. The small number of eligible participants. d. The logistics of travel and time in clinic. 3. What is the response of children with CP to the data collection protocol? a. School-aged children with CP, lack ing trunk control, participated successfully in measurement of total airway resistance with Hans


80 Rudolph facemasks and headsets, in conjunction with the Viasys Jaeger IOS and Respironics NICO. b. Although the IOS is marketed as a device that can be used for individuals who cannot comply with conventional pulmonary testing, the recommended mouthpiece did not prove useable with this sample. c. For this sample, optimal measurement intervals began immediately upon presentation of each seating parameter and ended when the pulmonary mechanics measure was obtained. d. Seating simulator and facemask accep tability were increased, in the presence of greater verbal ability. e. Wearing an HRI facemask for one hour in this protocol posed a small risk of skin breakdown at the nasal bridge. f. This method may contribute to fatigu e; fatigue may also be related to scheduling of data collection at the end of the clinic schedule. g. Other methods of measuring wheelch air configuration parameters and pulmonary mechanics need to be ex plored, including establishment of RAW reference values using the Jaeg er IOS in normally-abled children for HRI facemasks. h. No conclusions for nursing practice can be drawn as wheelchair influences on breathing still need to be estimated in a larger study.


81 References Bak, E., Johnson, W., Kinninger, W. K., & Burns, D. (2004). Accuracy of the Respiratory Compliance and Resistance Meas urements Provided by the Ven Trak 1550 Respiratory Mechan ics Monitoring System Retrieved from owview.asp?abs=9390&Search= Barks, L. (2004). Therapeutic positioning, wh eelchair seating, and pul monary function of children with cerebral palsy: A research synthesis. Rehabilitation Nursing 29 (5): 146-53. Becker. L. A. (1999). Effect size calculators Retrieved July 11, 2007, from University of Colorado at Colorado Springs Web site: Bergofsky, E., Turino, G., & Fishman, A. (1959). Cardiorespiratory failure in kyphoscoliosis. Medicine 38 : 263. Berven, S., & Bradford, D. S. (2002). Neuromus cular scoliosis: Causes of deformity and principles for evaluation and management. Seminal Neurology 22 (2): 167-78. Bjure, J., & Berg, K. (1970). Dynamic and st atic lung volumes of school children with cerebral palsy. Acta Orthopedica Sca ndinavica Supplement. 204 : 35. Braddock, D. L. (2002). Public financial suppor t for disability at the dawn of the 21st century. American Journal of Mental Retardation 107 (6): 478-89. Bridges, E. (2001). Turning patie nts on mechanical ventilation. Critical Care Nursing 21 (6): 66-8. Canet, E., Praud, J., & Bureau, M. (1998) Chest wall diseases and dysfunction in children. In T. Boat & V. Chernick (Eds.), Kendig's disorders of the respiratory tract in children (pp.794-799). Philadelphia: Saunders. Cassidy, C., Craig, C. L., Perry, A., Karlin, L., & Goldberg, M. (1994). A reassessment of spinal stabilization in severe cerebral palsy. Journal of Pediatric Orthopedics 14 (6): 731-9. Cherry, D. B. (1980). Review of physical therapy alternatives for reducing muscle contracture. Physical Therapy 60 (7): 877-81.


82 Cohen, J. (1992). A power primer. Psychological Bulletin 112 (1): 155-159. Cox, E. (1987). Dynamic positioning treatment: A new approach to customized therapeutic equipment for th e developmentally disabled Tulsa, OK: Christian Publishing Services. Crosbie, W. J., & Myles, S. (1985). An i nvestigation into the effect of postural modification on some aspects of normal pulmonary function. Physiotherapy 71 (7): 311-314. Dean, E. (1985). Effect of body position on pulmonary function. Physical Therapy 65 (5): 613-8. Druz, W. S., & Sharp, J. T. (1981). Activity of respiratory muscles in upright and recumbent humans. Journal of Applied Physiology 51 (6): 1552-61. Eyman, R. K., Grossman, H. J., Chaney, R., & Call, T. (1990). The life expectancy of profoundly handicapped people w ith mental retardation. New England Journal of Medicine 323 (9): 584-9. Farley, R. C., Davidson, J. C., Evans, G., M acLennan, K., Michael, S., Morrow, M., et al. (2003). What is the evidence for the e ffectiveness of postural management? International Journal of Therapy and Rehabilitation 10 (10): 449-455. Frankeny, J. R., Holly, R. G., & Ashmore, C. (1983). Effects of graded duration of stretch on normal and dystrophic skeletal muscle. Muscle & Nerve 6 (4): 269-77. Fraser, B., Hensinger, R., & Phelps, J. (1990). Physical management of multiple handicaps Baltimore: Paul H. Brookes. Geubelle, F., & Goffin, C. (1962) Respiratory studies in ch ildren. IV: Lung volumes and body positions in healthy children. Acta Orthopedica Scandinavica 51 : 255-60. Gregson, J., Leathley, M., Moore, A., Shar ma, A. Smith, T., & Watkins, C. (1999). Reliability of the Tone Assessment Scal e and the Modified Ashworth Scale as clinical tools for assess ing poststroke spasticity. Archives of Physical Medicine and Rehabilitation 80 (9) 1013-1016 Hale, K., & Rasp, F., (Eds.). (1987). Pulmonary function testing In Moe's textbook of scoliosis and other spinal deformities (2nd ed., pp. 482-84). Philadelphia: W.B. Saunders.


83 Haley, S., & Inacio, C. (1990). Evaluation of sp asticity and its effect on motor function. In M. Glenn & J. Whyte, The practical management of spasticity in children and adults (pp. 70-96). Philadelphia: Lea & Febiger. Hans Rudolph, Incorporated. (2007). Oro-Nasal NIV Face Masks and Headgear [Product manual]. Kansas City, MO: Author. Hans Rudolph, Incorporated. (n.d.). Homepage Retrieved July 8, 2007, from Horan, T., Mateus, S., Beraldo, P., Araujo, L ., Urschel, J., Urmenyi, E., et al. (2001). Forced oscillation technique to evaluate tracheostenosis in patients with neurologic injury. Chest 120 (1): 69-73. Hough, A. (1984). The effect of posture on lung function. Physiotherapy 70 (3): 101-104. Itovici, H., & Lyons, H. (1956). Ventilatory and lung volume de terminations in patients with chest deformities. American Journal of the Medical Sciences 232 : 265-75. Keppel, G. (1991). Design and analysis: A researcher’s handbook Upper Saddle River, NJ: Prentice-Hall. Landers, M., Barker, G. Wallentine, S., McW horter, J., & Peel, C. (2003). A comparison of tidal volume, breathing frequency, a nd minute ventilation between two sitting postures in healthy adults. Physiotherapy Theory & Practice 19 (2): 109-119. Leopando, M. T., Moussavi, Z., Holbrow, J., Chernick, V., Pasterkamp, H., & Rempel, G. (1999). Effect of a Soft Boston Ort hosis on pulmonary mechanics in severe cerebral palsy. Pediatric Pulmonology 28 (1): 53-8. Letts, M., Rathbone, D., Yamashita, T., Nic hol, B., & Keeler, A. (1992). Soft Boston Orthosis in management of neuromuscu lar scoliosis: A preliminary report. Journal of Pediatric Orthopedics 12 (4): 470-4. Makley, J., Herndon, C., Inkle y, S., Doershuk, C., Matthews, L., Post, R., et al. (1968). Pulmonary function in paralytic and nonparalytic scoliosis before and after treatment. Journal of Bone and Joint Surgery 50-A (7): 1379-90. Moody, L. (2001). Research Analysis Tool Unpublished Manuscript. Moreno, F., & Lyons, H.A. (1961). Effect of body posture on lung volumes. Journal of Applied Physiology 16 : 27-9.


84 Myhr, U. & von Wendt, L. (1991). Improveme nt of functional sitting position for children with cerebral palsy. Developmental Medicine and Child Neurology 33 (3): 246-56. Nachemsson, A. (1968). A long term fo llow-up study of non-treated scoliosis. Acta Orthopedica Scandinavica 39 : 466-476. Nilsonne, V. L., & Lundgren, K-O. (1968). Longterm prognosis in idiopathic scoliosis. Acta Orthopedica Scandinavica 39 : 456-465. Noble-Jamieson, C. M., Heckmatt, J. Z., D ubowitz, V., & Silverman, M. (1986). Effects of posture and spinal bracing on respirat ory function in neuromuscular disease. Archives of Disability in Children 61 (2): 178-81. Nwaobi, O. M. (1986). Effects of body orientation in space on tonic muscle activity of patients with cerebral palsy. Developmental Medicine and Child Neurology 28 (1): 41-4. Nwaobi, O. M., & Smith, P. D. (1986). Effect of adaptive seating on pulmonary function of children with cerebral palsy. Developmental Medicine and Child Neurology 28 (3): 351-4. Olson, E. V., Johnson, B. J., & Thompson, L. (1990). The hazards of immobility. 1967. American Journal of Nursing 90 (3): 43-8. Orem, D., Taylor, S., & Renpenning, K. (2001). Nursing: Concepts of practice St. Louis: Mosby. Prairie Seating Corporation. (2007). Homepage Retrieved March 19, 2007, from Pynt, J., Higgs, J., & Mackey, M. (2001). S eeking the optimal posture of the seated lumbar spine. Physiotherapy Theory and Practice 17 (1): 5-21. Raven, M. (1988). Application of Orem's self-care model to nursing practice in developmental disability. Australian Journal of Advanced Nursing 6 (2): 16-23. Redstone, F. (2004). The effects of seati ng position on the respir atory patterns of preschoolers with cerebral palsy. International Journal of Rehabilitation Research 27 (4): 283-288. Respironics. (2007a). Products: Non-Invasive Cardiac Output Monitor Retrieved March 19, 2007, from Respironics. (2007b). Products: NICO Cardiopulmonary Management System Retrieved March 8, 2007, from


85 Respironics. (2007c). NICO2 users manual, Model 7600, revision C (pp. 113-5). Wallingford, CT: Author. Respironics Novametrix. (2003). Parameter calculations in the CO2SMO Plus Respiratory Profile Monitor and NICO Cardiopulmonary Monitor [White Paper]. Wallingford, CT: Author. Samilson, R. L., Tsou, P., & Aamoth, G ., & Green, W. (1972). Dislocation and subluxation of the hip in cerebral palsy: Pathogenesi s, natural history and management. American Journal of Bone and Joint Surgery 54 (4): 863-73. Sevastikoglou, J. A., Linderholm, H., & Li ndgren, U. (1976). Effect of the Milwaukee brace on vital and ventilatory capacity of scoliotic patients. Acta Orthopedica Scandinavica 47 (5): 540-5. Shannon, D. C., Riseborough, E. J., & Kazem i, H. (1971). Ventilation perfusion relationships following co rrection of kyphoscoliosis. Journal of the American Medical Association 217 (5): 579-84. Shaw, P. (2006, February). Wheelchair seating and Modified Ashworth Scale. Training presented at Milestones Thera py Center, St. Petersburg, FL. Stevens, J. (1999). Intermediate statisti cs: A modern approach (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Tabachnick, B., & Fidell, L. (2001). Using multivariate statistics Boston: Allyn & Bacon. Tangsrud, S. E., Carlsen, K. C., Lund-Peter sen, I., & Carlsen, K.H. (2001). Lung function measurements in young children with spin al muscle atrophy: A cross sectional survey on the effect of position and bracing. Archives of Disease in Childhood 84 (6): 521-4. Ting, E., & Lyons, H. (1964). The relation of pre ssure and volume of the total respiratory system and its components in kyphoscoliosis. American Reviews in Respiratory Disease 89 : 379-386. United Cerebral Palsy Re search Foundation. (1995). Fact Sheets and General Information Retrieved September 15, 2002, from U.S. Census Bureau. State and county quick facts (2000). Retrieved September 15, 2002, from


86 U.S. Department of Health and Human Services (2000). Healthy people 2010 Washington, D.C.: U.S. Government Printing Office. U.S. National Center on Health Statistics (2003). Vital and health statistics: National health information survey Series 10(94), table 22, p. 38. Retrieved September 29, 2002, from ata/series/sr_10/sr10_217 Viasys Healthcare, Inc. (2007). MasterScreen IOS specifications Retrieved July 8, 2007, from erv/downloads/014_MasterScreen_IOS_ Specs.pdf Viasys Healthcare, Inc. (2000a). Impulse Oscillometry System Measurement Program: 10 [Product Manual]. Retrieved October 12, 2002 from Viasys Healthcare, Inc. (2000b). IOS basic 1e [Product Manual]. Retrieved September 14, 2006, from Waugh, K. (2005, April). Wheelchair seating for postural control and function Training presented at Shriners Hospital for Children, Tampa, FL Waugh, K. (2004, March). Wheelchair seating training Training presented at Children’s Hospital of Denver, Denver, CO. Yeaw, E. M. (1996). The effect of body positioning upon maximal oxygenation of patients with unilateral lung pathology. Journal of Advanced Nursing 23 (1): 5561.


87 Appendix A: Modified Ashworth Scale


88MODIFIED ASHWORTH SCALE (Haley & Inacio, 1990) SCORE CRITERIA 0 No increase in tone. 1 Slight increase in muscle tone, manifested by a catch and release, or by minimal resistance at the end of Range Of Motion (ROM) when the affected part(s) is/are moved into flexion or extension. 1+ Slight increase in muscle tone, manifested by a catch, followed by minimal resistance throughout the remainder (less than half) of the ROM. 2 More marked increase in muscle tone through most of the ROM, but affected part(s) is/are easily moved. 3 Considerable increase in muscle tone, passive movement is difficult. 4 Affected part(s) is/are rigid in flexion or extension.


89 Appendix B: Consent to Participate



PAGE 100


PAGE 101


PAGE 102


PAGE 103


PAGE 104


PAGE 105


PAGE 106


PAGE 107


PAGE 108


PAGE 109

100 Appendix C: Invitation to Participate

PAGE 110


PAGE 111

102 Appendix D: Instrument Specifications

PAGE 112


PAGE 113


PAGE 114

105 Respironics. (2007). NICO Users Manual, Model 7600, pp 113-115. Wallingford, CT: Author.

PAGE 115


PAGE 116

107 Viasys. (2007) http://www.viasyshe /014_MasterScreen_IOS_Specs.pdf

PAGE 117


PAGE 118

109 Hans Rudolph, Inc. Homepage Retrieved July 8, 2007, from /pdf/691148%201195%20A.pdf. Used by permission.

PAGE 119

110 Creating Possibilities "plane and simple" Reflection PSS 98 PLANAR SIMULATOR Specifications Standard equipment Mounted on a reinforced Luv-Base 3 pair lateral pads and mounting brackets 1 pair abductors and mounting brackets 1 pair arm supports and mounting brackets 1 head support mounting bracket to fit Whitmyer & Otto Bock type head supports 1 pair leg supports 1 pair medium Ankle Huggers ™ from BODYPOINT DESIGNS 1 pelvic strap Seat depth 6" to 20" Back height 10" to 24" Seat & back width to 20" Seat to back angle 65 de grees through 180 degrees Tilt 10 degrees anterior to 45 degrees posterior Arm supports fully adjustable, mounted to back rails to recline with back Leg supports fully adjustable Client weight limit 200 lbs Biangular Back 12" x 20" overall dim. hinged pad Weight of Planar Simulator Including all asseccories 120 lbs Specifications subject to change without notice Prairie Seating Corp. 7515 Linder Avenue Skokie, IL 60077 847-568-0001 Fax 847-568-0002 http://www.prairieseatin

PAGE 120

111 Appendix E: Institutional Review Board Approvals

PAGE 121


PAGE 122


PAGE 123


PAGE 124


PAGE 125

About the Author Lee Barks received a Bachelor of Sc ience in Nursing degree from the University of Virginia and a Master of Nursing degree from the University of Florida. She has held positions as staff nurse, clinical nurse special ist, and nurse practitioner, primarily in pediatrics. For the past 23 years, she has consulted on federal litig ation regarding long term care of persons with disabilities of all ag es in fifteen states, and has been admitted as an expert by the US Department of Jus tice. She has extensiv e experience with interdisciplinary teams and has most recently held the position of health services specialist at the Patient Safety Center of Inquiry, Tampa Veterans Administration. Ms. Barks is a 2007 inaugural Inte rdisciplinary Fellow in Patient Safety for the Veterans Administration and has authored publications in Rehabilitation Nursing numerous training programs, and chapte rs in several textbooks.