Abstract
Background Bipedal locomotor control requirements may be useful as classifications for walking dysfunction because they go beyond gait analysis to address all issues contributing to walking dysfunction.
Objective The objective of this study was to determine whether locomotor experts could achieve consensus about the requirements for bipedal locomotion.
Design and Methods Locomotor experts from physical therapy and other related professions participated in an electronic mail Delphi survey. Experts recommended additions, deletions, rewording, and merges for 15 proposed locomotor requirements in round 1. In rounds 2 and 3, panelists commented on and rated the validity, mutual exclusiveness, and understandability of each requirement. Consensus was defined a priori as: (1) 75% or more panelists agree or strongly agree that a requirement is valid, mutually exclusive, and understandable in round 3; (2) no difference between round 2 and 3 ratings with kappa coefficients ≥.60; and (3) a reduction in panelists who commented and convergence of comments between rounds 1 and 3. Content analysis and nonparametric statistics were used.
Results Fifty-eight panelists reached full consensus on 5 locomotor requirements (Initiation, Termination, Anticipatory Dynamic Balance, Multi-Task Capacity, and Walking Confidence) and partial consensus for 7 other requirements. There were no significant differences in ratings between rounds 2 and 3, and there was a decrease in the percentage of panelists who commented between rounds 1 and 3.
Limitations The study's 6-month time frame may have contributed to panelist attrition.
Conclusions Locomotor experts achieved consensus on several bipedal locomotor requirements. With validation, these requirements can provide the framework for a clinically feasible and systematic diagnostic tool for physical therapists to categorize locomotor problems and standardize intervention for walking dysfunction.
Physical therapist leaders have long called on the profession to develop and apply classification systems to guide and standardize physical therapist practice, improve communication among colleagues, and categorize patients so that treatment effectiveness can be studied more effectively.1–4 Despite this charge, standardized diagnoses are not a ubiquitous part of current physical therapist practice.5 Coffin-Zadai5 attributed this deficiency to the complexity of the diagnostic process, evolving knowledge, and lack of agreement on diagnostic constructs and language.
Consensus has been reached on some fundamental principles through the Diagnosis Dialog initiative.6 This group of physical therapist leaders, representing the spectrum of physical therapist practice, agreed that the primary focus of physical therapist diagnostic expertise is on diagnosing and identifying syndromes of the human movement system. They described physical therapist diagnosis as both a process of integration and evaluation of examination data in order to guide prognosis and treatment, as well as assigning diagnostic descriptors that identify a condition or syndrome at the level of the system and the whole person. The group suggested that existing models of disablement and enablement should inform, but not constrain, the development of diagnoses. Last, they recommended using standardized anatomical, physiological, or functional terms that concisely describe conditions.
A variety of constructs have been proposed for diagnostic classifications, including disability models,1,7–9 etiology,7,10 body functions and structures (BFS) signs and symptoms,11–15 and response to treatment.16 Nutt et al17 called for new classifications of gait and balance disorders based on function of the locomotor and balance circuits. Accordingly, the Balance Evaluation Systems Test (BESTest) groups clinical tests under 6 subcategories of balance control to help physical therapists identify the underlying postural control systems responsible for impaired balance so that treatments can be directed specifically at the abnormal underlying systems.18 A similar test for walking has been identified as a need by clinicians who work with patients with neurologic dysfunction.19
Patla20 described several locomotor control systems that act on and interact with the musculoskeletal system to influence the expression of skilled locomotor behavior. These systems include: core locomotor pattern, active propulsion, weight support, dynamic equilibrium, steering and accommodation, maintaining structural integrity, minimization of energy expenditure, and cognitive spatial mapping. Patla postulated that it would be more revealing to identify dysfunction associated with these control factors in older adults because deficits in BFS may or may not have a direct relationship to mobility deficits. Similarly, traditional gait analysis may not reveal underlying causes of walking dysfunction in older adults and others with central nervous system involvement. Rather, locomotion, defined as the movement of an organism from one place to another,21 is a broader and potentially more meaningful construct on which to base clinical decision making. We propose, therefore, that requirements of bipedal locomotion could form the initial sorting factor for diagnostic classifications for bipedal locomotor dysfunction and provide the framework for a diagnostic clinical locomotor assessment tool that would enable physical therapists to categorize locomotor problems so that hypotheses about underlying causes and decisions about further examination and treatment can be organized.
The Delphi survey is a method for a wide range of experts to come to consensus regarding an issue that has not been previously explored.22–24 The Delphi method uses a series of structured surveys to elicit and transform individual opinions into group consensus by feeding back the group results to all individuals in each subsequent survey. This method enables participants to reconsider their previous responses in light of the group's average response while remaining anonymous. The originator of the Delphi survey is able to aggregate knowledge and experience of experts without any one individual dominating the discussion.22–24 Delphi surveys are a good way to initiate discussion on a topic where the relevant group is not easily brought together.25 Guccione8 recommended this method to develop physical therapist diagnostic classifications.
The purpose of this study was for a group of locomotor experts to arrive at consensus on the validity, mutual exclusivity, and understandability of proposed locomotor requirements for bipedal locomotion for future use in a framework of diagnostic classifications of locomotion.
Method
Design Overview
The first author developed a list of 15 bipedal locomotor requirements for consideration by an expert panel utilizing a Delphi method that included a 3-round electronic mail survey.
Development of Proposed Locomotor Requirements
The first author considered the locomotor control systems proposed by Patla20 and determined that they encompassed many neural, biomechanical, behavioral, perceptual, and energetic processes but did not represent an exhaustive list of locomotor requirements. Three critical subtasks of bipedal locomotion—limb clearance,26,27 initiation,28 and termination29—also were deemed important to include. Other additions included self-efficacy or confidence in one's ability to walk safely30–32 along with 2 aspects of executive function: purposefulness (the ability to modulate and produce behavior)33 and dual tasking (the ability to appropriately allocate attention among tasks that are performed simultaneously).33 This process resulted in the development of 15 original locomotor requirements (Tab. 1).
Names and Descriptions of the Original and Modified Requirements for Bipedal Locomotion
Recruitment of Participants
The goal was to create an expert panel that was representative of clinicians as well as academicians and researchers from the range of disciplines that contribute to the body of knowledge of bipedal locomotion. Accordingly, one half of the participants were intended to be academic or research physicians, physical therapists, engineers, biomechanists, and scientists. The other half of the participants would be clinicians, primarily physical therapists or physicians. The target number of participants was 100. An ideal number of participants for Delphi surveys has not been identified and should be based on scope of the problem as well as available resources.34 For this study, a larger cohort of participants was desirable to support the generalizability of the results.
The first author used purposive sampling to recruit participants for the expert panel by e-mailing the center coordinators of clinical education of the physical therapy departments of the top 10 rehabilitation35 and children's36 (neurology and neurosurgery) hospitals as recognized by U.S. News and World Report and asking them to recommend any physical therapists from their department who they considered to be experts in locomotion (Fig. 1). Experts were defined as having a reputation of being an expert in locomotion and, if a physical therapist, having 5 or more years of clinical experience and being an American Board of Physical Therapy Specialties–certified Geriatric, Neurologic, or Pediatric Clinical Specialist. Clinical specialists were recruited because they are recognized by the physical therapy profession as demonstrating advanced clinical knowledge, experience, and skills in a special area of practice.37
Description of expert panel recruitment. CCCE=Center Coordinator of Clinical Education, PTJ=Physical Therapy: Journal of the American Physical Therapy Association.
The first author also e-mailed the directors of the geriatric, neurologic, and pediatric residency programs of the American Physical Therapy Association38 and asked them to recommend clinical specialists having a minimum of 5 years of experience who were associated with their residency program and who they considered locomotor experts. Last, the first author also contacted members of the editorial board of the 2 professional gait organizations, as well as the authors, guest editors, and manuscript reviewers for the PTJ Jacquelin Perry special issue “Stepping Forward With Gait Rehabilitation”39 and asked these individuals to recommend colleagues with a variety of professional backgrounds who they deemed to be experts in locomotion. Individuals received 2 or 3 e-mail reminders. Potential participants who currently work with the authors were excluded. This process netted 220 potential participants.
The first author then used a snowball sampling technique to recruit additional expert panelists (Fig. 1). The first author asked 166 of the identified experts to recommend colleagues who they believed were locomotor experts. Specific criteria for clinical experts were not used, assuming that these individuals would identify credible locomotor experts and the clinical specialist designation was not applicable internationally. This process netted 67 additional potential participants, for a total of 287.
Delphi Process
Round 1: expert panel demographics, addition, deletion, rewording, and merge recommendations for original locomotor requirements.
Figure 2 illustrates the Delphi process. To initiate the first round, the first author sent an e-mail via Survey Monkey (Survey Monkey, Palo Alto, California) to each of the 287 identified experts. This e-mail provided an introduction to the purpose, scope, time frame of the study, estimated time commitment, and informed consent and emphasized the importance of committing to the entire process. The e-mail message contained a link to the first survey, where respondents were asked to provide demographic information about their age, sex, profession, years in profession, and academic and professional credentials. Participants were asked to rank how they spend their professional time in terms of clinical, research, teaching and to estimate the percentage of total work time devoted to working with adults over 18 years of age or with children.
Description of the Delphi process.
Participants were presented with the names and descriptions of the 15 bipedal locomotor requirements (Tab. 1) and were asked to respond to the following 4 open-ended questions: “Are there any classifications* that you think should be added (removed, reworded, merged)? If so, please describe and provide your rationale.” Participants also were asked for any overall comments. Participants were given 3 weeks to respond to this and subsequent rounds of the survey. Reminder e-mails were sent out 1 week and 2 days before and 1 week after the deadlines.
The responses to the round 1 open-ended questions were downloaded verbatim and de-identified prior to analysis by the qualitative analysis team (QAT). The QAT consisted of the first author and 2 qualitative analysis experts, one of whom is also a locomotor expert. The team used content analysis40 to code open-ended comments. Each team member coded the comments individually, and then the QAT met to arrive at consensus on the codes. The QAT coded comments about a specific requirement as “remove” or “reword” and coded other comments as “addition,” “merge,” or “global.” All coded comments with the exception of “remove” comments were further coded to more specifically reflect their meaning. The first author calculated code frequencies41 and used the relative code frequency to modify the names or descriptions of the locomotor requirements, to develop additional requirements, and to create an alternative organization of the requirements for the QAT's consideration.
Round 2: Likert ratings of and qualitative responses to original, modified, and additional locomotor requirements.
The round 2 survey contained a summary of the demographics of the round 1 expert panel, overall response rate, and percentage of the experts who responded to each of the round 1 open-ended questions (Fig. 2). This summary was followed by an explanation of the data analysis procedure and a summary of the global comments with some clarifications. Each original locomotor requirement was presented with the codes and their frequency, followed by the modified requirement name and description. Participants were instructed to review this information prior to rating the validity, mutual exclusiveness, and understandability of each original, modified, and additional requirement using a 4-point Likert scale (strongly agree, agree, disagree, strongly disagree) and provide any open-ended comments. Finally, an alternate organization of all the locomotor requirements was presented for participant feedback. Because of the length of the survey, participants were offered access to their personal responses from round 1. Content analysis and calculation of code frequencies were completed for the open-ended comments.
Round 3: repeated Likert ratings of and qualitative responses to original, modified, and additional locomotor requirements.
The third-round survey provided the participants with a summary of the round 2 panel demographics and code frequency of the global comments. Each original and modified requirement was presented along with a summary of the round 2 Likert ratings and code frequency of the open-ended comments. For round 3, participants were asked to re-rate the same requirements after reviewing the group response from round 2. Individual responses from round 2 were sent to each panelist in a PDF electronic file. Following content analysis and calculation of code frequencies, the first author developed a content analytic summary table42 to reveal trends in the open-ended comments across the requirements and rounds.
Data Analysis
The demographic and Likert responses were downloaded and summarized descriptively. The Wilcoxon signed ranks test was used to test the difference in participant Likert responses between rounds 2 and 3. A linear weighted kappa coefficient was calculated to assess intrarater agreement for Likert responses from rounds 2 and 3.43 A significance level of P<.05 was used for all statistical tests. Consensus about the proposed locomotor requirements was defined a priori as:
In round 3, ≥75% of participants in the Delphi survey agree or strongly agree that each requirement represents a valid, mutually exclusive, and understandable fundamental requirement for bipedal locomotion.34,44–46
There will be no statistically significant difference between participant responses in rounds 2 and 3 (Wilcoxon signed ranks test, P≤.05).47
Weighted kappa coefficient will be ≥.60 when comparing participant responses with close-ended questions from rounds 2 and 3.43,48
There will be no statistically significant difference in participant responses in round 2 between panelists who dropped out following round 2 versus those who continued to participate through round 3.
There will be a reduction in the percentage of panelists who comment and a convergence of issues between rounds 1 and 3.43
Results
Expert Panel Recruitment and Demographics
Purposive and snowball recruitment resulted in 287 experts being initially invited to participate in the Delphi survey (Fig. 1). The response rates for rounds 1, 2, and 3 were 115/287 (40.1%), 78/115 (67.8%), and 58/78 (74.4%), respectively. A total of 58 expert panelists participated in all 3 rounds (Fig. 2).
The demographics of the expert panel remained relatively stable over the 3 rounds (Tab. 2). The majority (79.3%) of the 58 panelists who participated in all 3 rounds (round 3 panelists) were physical therapists along with 5 engineers and 9 others from a variety of other disciplines. Nearly 60% (56.9%) of the experts had earned a nonclinical doctoral degree, and 22% of the physical therapists had earned a clinical doctoral degree. Just more half of the panelists (51.7%) identified themselves as primarily researchers, and 34.5% self-identified primarily as clinicians. The majority of the expert panelists (74.1%) indicated that they worked mostly with adults, and 20.7% stated that they worked primarily with children. The mean number of years as professionals was 18.4. The majority (78.9%) of clinicians who were physical therapists were certified as clinical specialists.
Demographics of the Expert Panel
Round 1
All requirements were retained by the QAT. Although 28% of the expert panelists recommended removing one or more of the requirements, fewer than 8% of the panelists recommended removal of any one requirement. There were no recommendations to remove 5 of the locomotor requirements.
The QAT revised all but Initiation and Termination names and the description for Balance Confidence based on recommendations from 75% of panelists (Tab. 1). The locomotor requirements fell into 3 groups according to the scope of the recommendations. Relatively straightforward changes in terminology, descriptions, and definitions were recommended for Initiation, Termination, Proactive Dynamic Equilibrium, Reactive Dynamic Equilibrium, Steering and Accommodation, Dual-Task Capacity, Balance Confidence, and Energy Cost. In contrast, lack of clarity was the primary issue for Spatial Mapping, Purposefulness, and Long-Term Viability, necessitating more substantial modifications. The QAT faced the most difficult task in modifying the locomotor requirements Rhythmical Limb Movement, Antigravity Support, Active Propulsion, and Limb Clearance because of panelists' concerns about clarity and constructs underlying the requirements.
The QAT developed 6 additional locomotor requirements for rating in round 2 (Tab. 3A) from the “addition” recommendations from 70 (60.8%) of the panelists. The QAT developed the alternative organization of the locomotor requirements (Tab. 3B) based on recommendations by 25 (21.7%) of the panelists.
Additional Requirements for Bipedal Locomotion and Alternative Organization of the Requirements for Bipedal Locomotion Created by the Qualitative Analysis Team Based on Recommendations From the Expert Panel in Round 1
Fifty-eight panelists (50.4%) offered global comments. Most comments reflected positive impressions of the requirements and more specifically the perception that the proposed requirements were clinically relevant and comprehensive because they considered dimensions beyond traditional gait analysis. Several panelists questioned the intended use of the requirements and whether they were meant to replace traditional gait analysis. Some panelists did not understand the need for requirements beyond traditional gait analysis. Some panelists expressed concerns about the length and complexity of the requirements as well as therapists' ability to understand and willingness to adopt the terminology. Based on these and other comments, the QAT provided a clarification of the need for diagnostic classifications, the definition of diagnosis in this context, and defined the distinction between locomotion and gait at the beginning of the round 2 survey. The introduction to round 2 also reiterated the purpose and scope of the study.
Round 2
Seventy-five percent or more of the panelists strongly agreed or agreed about the validity, mutual exclusivity, and understandability of the original Initiation, Termination, and Foot Clearance requirements. The same level of agreement was reached for the validity and mutual exclusivity of Energy Cost and for the validity of Reactive Dynamic Equilibrium and Dual-Task Capacity (Tab. 4). For the modified requirements, greater than 75% of the panelists strongly agreed or agreed about the validity, mutual exclusivity, and understandability of Initiation, Termination, Stance Stability, Anticipatory Dynamic Equilibrium, Reactive Dynamic Equilibrium, Multi-Task Capacity, Walking Confidence, and Metabolic Energy Expenditure and the validity of all the other modified requirements with the exception of Navigation to Unseen Locations and Long-Term Musculoskeletal Integrity (Tab. 4).
Round 2 and 3 Likert Ratings for Original, Modified, and Additional Requirements for Bipedal Locomotiona
The open-ended responses generally echoed the Likert ratings for the requirements. Panelist comments reflected a preference for the modified version of the majority of the requirements that reached full consensus. Clarity and construct concerns reflected the lower percentage of panelist agreement on understandability of some of the requirements. Global concerns from round 1 persisted, and the issue of overlap between requirements also was raised in congruence with lack of agreement in Likert ratings about the mutual exclusivity of the majority of requirements.
The validity of 3 of the 6 additional requirements was agreed upon by greater than 75% of the panelists (Tab. 4). Open-ended comments about the additional requirements were minimal. Thirty-three of the 39 panelists who commented on the alternative organization of the locomotor requirements commented that it was a positive change, specifically more understandable and useful.
Round 3
In round 3, the expert panel reached consensus on the validity, mutual exclusivity, and understandability of the original Initiation and Termination requirements and the validity and mutual exclusivity of Energy Cost, as well as the validity of Foot Clearance and Dual-Task Capacity (Tab. 4). The panelists reached consensus on the validity, mutual exclusivity, and understandability of 5 of the modified requirements: Initiation, Termination, Anticipatory Dynamic Balance, Multi-Task Capacity, and Walking Confidence (Tab. 4). These are the same requirements in which the primary theme in round 2 was “prefer modified version.” The panelists reached consensus on the validity and mutual exclusivity of Coordination of Rhythmical Stepping and Arm Swing and Metabolic Energy Expenditure and the validity of all other modified requirements except Navigation to Unseen Locations, Purposefulness, and Long-Term Musculoskeletal Integrity. Concerns about clarity, construct, and overlap characterized the classifications that did not reach full consensus. No consensus was reached for the validity, mutual exclusivity, and understandability of any of the additional locomotor requirements (Tab. 4). No new concerns arose in the round 3 open-ended comments about any of the requirements or the alternative organization of the requirements.
Comparisons Among Rounds
There was no statistically significant difference between participant Likert responses in round 2 versus round 3. The only linear weighted kappa value that exceeded .6 was for validity of the modified Termination classification (κ=.67).
There was a significant reduction in the percentage of panelists who provided global comments in round 3 (22.4%) compared with round 1 (50.4%); however, the codes developed in round 1 generally persisted through round 3, so convergence was not demonstrated. In contrast, there was not a reduction in the percentage of panelists who commented on most requirements between rounds 1 and 3, and the discussion pertained to all of the requirements. Panelist discussion involved the requirements where both clarity and construct were issues, as well as those where there was a clear preference for the modified version. Interestingly, there was a convergence of comments between rounds 1 and 3 for the requirements that reached full consensus. Comments about clarity and construct remained similar between rounds 1 and 3 for those requirements that did not reach full consensus. A comparison of the round 2 ratings between panelists who dropped out after round 2 versus those who continued to participate through round 3 revealed statistically significant differences only for mutual exclusivity of Coordination of Rhythmical Stepping and Arm Swing (P=.032) and understandability of multi-tasking (P=.047).
Discussion
This is the first systematic attempt to achieve consensus amongst a cross-disciplinary group of experts in locomotion about fundamental requirements of bipedal locomotion as a precursor to developing a diagnostic classification framework. The key finding of this study was that the expert panel reached full consensus in the round 3 Likert ratings on 5 of the modified locomotor requirements: Initiation, Termination, Anticipatory Dynamic Balance, Multi-Task Capacity, and Walking Confidence. This consensus provides initial face49,50 and content25 validation for these bipedal locomotor requirements that are related to walking but are not included in traditional gait analysis. Many panelists noted the potential clinical benefits of using the requirements as diagnostic classifications because aspects of walking that extend beyond traditional gait analysis were included. A clinical framework based on locomotor requirements representing the spectrum of walking dysfunction could help physical therapists identify issues that may not be easily derived from traditional gait analysis. For example, people with dementia and posttraumatic brain injury may each have impaired navigation to unseen locations even though they exhibit disparate gait kinematics. Alternatively, an older person may exhibit minimal gait kinematic alterations but have underlying impairments of walking confidence and dynamic equilibrium that put them at risk for falls.
The locomotor requirements are intended to be mutually exclusive but are not meant to represent unique diagnostic classifications. Rather, an individual may likely exhibit limitations in multiple requirements, and often they will interact with each other. For example, a person with cerebellar dysfunction may demonstrate problems with the locomotor requirements Progression During Stance and Swing Limb Advancement that are exacerbated by Anticipatory Dynamic Balance impairments.
A framework based on the locomotor requirements, along with associated clinical measures and interventions, could guide clinical decision making, a critical attribute of the physical therapist diagnostic process.6 It could become a tool to organize the physical therapist examination for walking dysfunction and reveal the patterns of impaired locomotor requirements across a variety of clinical populations. Clusters of impaired requirements may eventually represent unique diagnostic groups for locomotion dysfunction. Eventually, interventions identified for each requirement could help to standardize physical therapy care for walking dysfunction. Last, this framework could inform third-party payers and other health professionals about the scope of physical therapy intervention for walking dysfunction.
The panelists did not reach consensus about the validity of 3 requirements: Navigation to Unseen Locations, Purposefulness, and Long-Term Musculoskeletal Integrity. Qualitative comments revealed the panelists' concerns about the clarity of the constructs underlying these requirements as well as questions about their relevance to the analysis of walking. Panelists did reach consensus about the validity of the locomotor requirements that were closest to traditional gait analysis phases such as Stance Stability, Progression During Stance, and Swing Limb Advancement. They did not agree on their mutual exclusiveness and understandability; perhaps because panelists had the most expertise in these areas, they examined these locomotor requirements more critically or were less willing to modify them. This scrutiny also may have contributed to the lack of consensus on the language used to describe the locomotor requirement Metabolic Energy Expenditure.
A major strength of this study was the multiple measures of consensus that were used to help increase confidence that agreement was reached on the bipedal locomotor requirements.41 Along with the Likert ratings, expert panel consensus was demonstrated by the overall lack of statistically significant difference between the Likert ratings in rounds 2 and 3 and a decrease in the percentage of panelists who made global comments between rounds 1 and 3, even though the issues that emerged from their comments did not converge appreciably. Additionally, the qualitative comments and the quantitative ratings of the requirements often corresponded, indicating agreement between the 2 types of panelist responses.
Another strength of the study, intended to be a first step in the development of locomotor diagnostic classifications, is that it sought input from a broad representation of experts, including both physical therapists and non–physical therapists, clinicians, and researchers, as well as those who work with adults and children. Diagnostic classifications must be understood by multiple audiences while remaining meaningful to physical therapist practice. Additionally, the expert panel was highly educated and experienced, and many panelists held clinical specialist certification. Panelists seemed very engaged in the process, as evidenced by the percentage who provided open-ended responses and the substantive nature of many comments. The numbers of comments about individual requirements did not decrease and in many cases increased in the later rounds and included references to comments from previous rounds. The response rates for rounds 2 and 3 were high, and the lack of statistically significant difference in the Likert ratings of the panelists who dropped out after round 2 versus those who continued to participate through round 3 did not support the possibility that individuals dropped out because they perceived their perspective to be significantly different from the mainstream. The qualitative analysis and locomotor expertise of the QAT members also strengthened the study. Finally, the QAT was responsive to questions that the experts expressed in round 1 and provided additional information for them.
The study's 6-month time frame was lengthy for maintaining panelists' focus and motivation and may have contributed to the dropout rate. Surveys 1 and 2 each contained 110 questions that generated a significant amount of qualitative data to process, contributing to the gaps between rounds. The length of the surveys also may have contributed to panelist attrition. The decision to provide panelists with their individual responses from previous rounds may have contributed to individuals simply repeating their responses rather than using them to compare their responses with the group responses. Survey Monkey did not support optimal presentation of data in tables, so panelists had to scroll through text to find information. Finally, the survey was concluded at the end of round 3 even though full consensus was not achieved for all of the requirements. An indication that panelists were changing their ratings between rounds 2 and 3 was the fair to moderate linear weighted kappa values. Although it is possible that greater consensus would have been reached with additional rounds, the decision was made to end the study due to concerns about panel fatigue and associated low response rates.
The method of this study provides a model for the initial development of physical therapy diagnostic classification systems. Further work is needed to establish the validity of the locomotor requirements. A version of the modified requirements with explicit operational definitions for all terms needs to be evaluated by focus groups consisting of experts from multiple disciplines as well as subgroups of physical therapists. Consensus on the underlying constructs and language should be prioritized before identification and testing of clinical measures and interventions for each requirement are pursued. The eventual goal is to develop a clinically feasible and systematic diagnostic tool for physical therapists to categorize locomotor problems and standardize interventions for walking dysfunction.
Footnotes
Dr Hedman, Dr Morris, Dr Graham, Dr Brown, Dr Ford, and Dr Ingram provided concept/idea/research design. Dr Hedman, Dr Morris, Dr Graham, Dr Brown, and Dr Ford provided writing. Dr Hedman provided data collection and project management. Dr Hedman, Dr Graham, Dr Hilliard, and Dr Salzman provided data analysis. Dr Morris, Dr Graham, Dr Brown, Dr Ford, Dr Ingram, Dr Hilliard, and Dr Salzman provided consultation (including review of the manuscript before submission). The authors thank the expert panelists, whose participation made this study possible.
This project was completed in partial fulfillment of requirements for Dr Hedman's doctor of science in physical therapy degree at the University of Alabama at Birmingham.
The institutional review boards of Northwestern University Feinberg School of Medicine and the University of Alabama at Birmingham approved the study.
↵* Within the surveys, locomotor requirements were called “diagnostic classifications” and were defined as fundamental requirements of bipedal locomotion toward which examination and intervention can be directed.
- Received December 21, 2012.
- Accepted August 2, 2013.
- © 2014 American Physical Therapy Association
References
Issue highlights
