Abstract
Background A variety of diagnostic classification systems are used by physical therapists, but little information about how therapists assign diagnostic labels and how the labels are used to direct intervention is available.
Objective The purposes of this study were: (1) to examine the diagnostic labels assigned to patient problems by physical therapists who are board-certified Orthopaedic Clinical Specialists (OCSs) and (2) to determine whether the label influences selection of interventions.
Design A cross-sectional survey was conducted.
Methods Two written cases were developed for patients with low back and shoulder pain. A survey was used to evaluate the diagnostic label assigned and the interventions considered important for each case. The cases and survey were sent to therapists who are board-certified OCSs. Respondents assigned a diagnostic label and rated the importance of intervention categories for each case. Each diagnostic label was coded based on the construct it represented. Percentage responses for each diagnostic label code and intervention category were calculated. Relative importance of intervention category based on diagnostic label was examined.
Results For the low back pain and shoulder pain cases, respectively, “Combination” (48.5%, 34.9%) and “Pathology/Pathophysiology” (32.7%, 57.3%) diagnostic labels were most common. Strengthening (85.9%, 98.1%), stretching (86.8%, 84.9%), neuromuscular re-education (87.6%, 93.4%), functional training (91.4%, 88.6%), and mobilization/manipulation (85.1%, 86.8%) were considered the most important interventions. Relative importance of interventions did not differ based on diagnostic label (χ2=0.050–1.263, P=.261–.824).
Limitations The low response rate may limit the generalizability of the findings. Also, examples provided for labels may have influenced responses, and some of the label codes may have represented overlapping constructs.
Conclusions There is little consistency with which OCS therapists assign diagnostic labels, and the label does not seem to influence selection of interventions.
Diagnosis has been defined in various ways. The traditional medical view of diagnosis is expressed most simply as “the identification of a disease or condition.”1(p515) The identification of a disease involves an investigative process and culminates in the assignment of a label to the condition, which in turn leads to a specific intervention or a defined set of intervention options. This traditional approach to diagnosis often occurs within the context of a well-developed taxonomy for health conditions.2 With the evolution of physical therapy as a doctoring profession, there has been an increased focus on the role of diagnosis in both clinical practice and research. Defining diagnosis for physical therapists has become an important part of the discussion in the literature, especially in light of the perceived limitations of traditional approaches to medical diagnosis.3 The American Physical Therapy Association (APTA) has defined diagnosis as “both a process and a label.”4(p45)
There have been many efforts in physical therapy to develop a consistent system for labeling patient conditions. Several diagnostic classification systems have been developed by physical therapists for the purpose of directing appropriate intervention.5 These systems use named categories to distinguish people with movement system impairments, mechanical dysfunction, or clinical syndromes amenable to physical therapy intervention and are most well developed for low back conditions. Examples of these systems include mechanical diagnosis and therapy,6 movement system impairment,7 and treatment-based classification.8 Physical therapists may incorporate aspects of these named patient categories into the diagnostic labels they assign to an individual patient. However, utilization of different classification systems may lead to inconsistency in the assignment of diagnostic labels to individual patients.
Several groups of physical therapy professionals also have come together with a goal of reaching some consensus points on physical therapy diagnosis and treatment. One group working to develop a “Diagnosis Dialog” recognized that there is a lack of consistency in the “diagnostic descriptors” used to identify the various clinical subtypes of movement system dysfunction, and the group has been working to develop a more standardized way of labeling conditions of the movement system.9,10 The Orthopaedic Section of APTA has been working to develop clinical practice guidelines. These guidelines have moved beyond the triage-based system of diagnosis proposed in several medical low back guidelines11 and have established a diagnostic classification system that integrates impairment-based terminology consistent with the International Classification of Functioning, Disability and Health (ICF).12
Given the many different initiatives to develop diagnostic frameworks in orthopedic physical therapist practice, there is little information about how existing frameworks are being used in clinical practice and the consistency with which therapists are assigning diagnostic labels. Consistency with which diagnostic labels are assigned is important for several reasons. First, diagnostic labels are the most succinct way to communicate what the patient problem is to other health care providers and the health care community. They are important because they serve to identify the condition for which the health care practitioner is directing intervention. Labels also are used to subgroup patients so that intervention can be best targeted to their problems. However, it also is unclear how the diagnostic labels are being used to direct intervention.
The purpose of the current study was to examine: (1) the diagnostic labels assigned to patients by Orthopaedic Certified Specialists (OCSs) based on examination data from written patient cases and (2) the influence of the diagnostic label on selection of interventions. With the lack of a standardized taxonomy in physical therapist practice, although general elements of the diagnostic process may be similar among physical therapists, we hypothesized that there will be inconsistency in assigning diagnostic labels to patient conditions and that the labels will not influence intervention decisions.
Method
Patient Cases
Two written patient cases were developed using information from actual patient examinations at an orthopedic outpatient clinic (Appendixes 1 and 2). The first case provides examination data for a patient with low back pain (LBP), and the second case provides examination data for a patient with shoulder pain. Low back pain and shoulder pain were selected because they are 2 of the more prevalent symptom complaints for which a patient seeks care at an outpatient orthopedic clinic.13
Briefly, the case for a patient with LBP (Appendix 1) included subjective information on apparent mechanisms of injury, history of LBP, numeric rating scale of LBP symptoms, symptom behavior with movements and positions, a systems review, and scores on relevant outcome measures (Oswestry Disability Index and Fear-Avoidance Beliefs Questionnaire). Examination findings were included from a postural assessment, single and repeated trunk movement tests, limb movement tests, manual muscle tests, a neurological screen, and relevant orthopedic special tests. The items included in the examination for the patient case were intended to be comprehensive and representative of aspects of the most commonly used diagnostic classification systems in outpatient orthopedic practice,14 including but not limited to the following systems: treatment-based classification,8,15,16 McKenzie mechanical diagnosis and therapy,6 and movement system impairment.7 The patient case for shoulder pain (Appendix 2) included subjective information on the apparent mechanism of injury, history of pain, a systems review, and score on the Disabilities of the Arm, Shoulder and Hand (DASH) outcome measure. Examination findings were included from a postural assessment, shoulder movement tests, manual muscle tests, muscle length tests, and relevant orthopedic special tests. The items included in the examination for the patient case were intended to be comprehensive and representative of a typical orthopedic examination.
Survey
A survey was developed to evaluate what diagnostic labels clinicians would assign and which interventions would be considered important for each patient case (Appendix 3). The first series of survey questions related to respondent demographics, education and training, practice experience, advanced certification, and assessing what clinical classification systems were being utilized in practice. Patient cases then were presented for each respondent to read. After each patient was presented, both open-ended and closed-ended survey questions were posed about the case. Specifically, case-related questions were intended to address the clinical decision-making process associated with diagnosis and intervention. To this end, clinicians were asked what additional tests and measures would be needed and what classification system would be utilized to determine a diagnosis. With an open-ended question, the respondent then was asked to assign a diagnostic label to each case. With a closed-ended question, the respondent was asked to rate each of 9 selected intervention categories on a 4-point (0–3) Likert scale for each case (0=“Would not use,” 1=“Not very important,” 2=“Somewhat important,” 3=“Very important”). Categories were selected based on interventions typically used in physical therapist practice, including: strengthening, stretching/range of motion (ROM), motor control/neuromuscular re-education, functional training, balance training, mobilization/manipulation, soft tissue techniques, and modalities.17
Approach
Patient cases and the accompanying survey questions were sent to board-certified OCS therapists from a public domain database on the APTA website for the Orthopaedic Section. Participants were selected using a stratified systematic sampling procedure. The sampling procedure was utilized to ensure a more representative sample across the country, minimize regional bias, and improve generalizability of the findings. Data reported on the American Board of Physical Therapy Specialties (ABPTS) website were used to determine the number of physical therapists with an OCS in each state.
With a goal of e-mailing 1,000 surveys, we used the percentage of OCS therapists in each state to determine the number of surveys sent to that state. Within each stratum, or state, we used a sampling interval to select participants. The sampling interval was selected based on the total number of therapists in the state and the number of therapists targeted for the survey. The survey was successfully mailed electronically through Survey Monkey (SurveyMonkey, Palo Alto, California) in July 2010 to 877 physical therapists with an OCS certification. The actual mailing (n=877) was less than the target mailing (N=1,000) due to problems associated with e-mail delivery errors. Respondents who received the electronic mailing were asked to review each case and respond to the survey questions following each case. Anonymous responses were collected electronically through SurveyMonkey over a 6-week period through August 31, 2010, at which time they were exported for data analysis. To provide a basis for comparison with the current sample, demographics such as sex, age, and practice setting for OCS therapists were obtained from the Office of Postprofessional Certification and Credentialing at APTA (Derek D. Stepp, Director, Postprofessional Certification & Credentialing, APTA; e-mail communication; October 2013).
Data Analysis
Survey data were analyzed using SPSS statistical software (version 19.0.0, IBM Corp, Armonk, New York). The response rate for the survey and summary statistics on demographics and other respondent characteristics were calculated. To determine whether the current sample was representative of the population, geographic representation and demographics were compared between the sample and population of OCS therapists. For the closed-ended survey question to determine which classification system would be used for each patient case, we calculated the percentage of respondents who indicated they would use each system.
Five separate investigators examined responses to the open-ended question in which respondents were asked to assign a diagnostic label to each case. After examining all responses, investigators identified common constructs that represented existing approaches to diagnostic classification in orthopedic physical therapist practice.6–8 For responses that included terminology not consistent with the constructs found in any of the existing clinical classification systems, but more consistent with disablement models,12 we developed a diagnostic label code to best represent the identified construct. Examples of diagnostic label codes included: “Impairment,” “Pain Location,” “Pathology/Pathophysiology,” “Practice Patterns,” “McKenzie,” “Treatment-Based Classification,” and “Movement System Impairment.” An operational definition for each diagnostic label code was established and is included in Appendix 4. Each investigator then examined the responses a second time and coded each response for the diagnostic label based upon the identified construct or constructs represented by the label and using the established diagnostic label codes. Diagnostic labels that represented multiple constructs in a single label were coded as “Combination.” The investigators then met, discussed discrepancies, and resolved any discrepancies via consensus. Frequency and percentage of responses for each diagnostic label code were calculated separately for each case.
Investigators then examined responses to the question related to selection of interventions. Upon inspection, few respondents rated specific intervention categories as 0 (“Would not use”) or 3 (“Very important”). Furthermore, the investigators felt that the response options 0 (“Would not use”) and 1 (“Not very important”) represented a similar construct, and the intervention was considered not important. Similarly, response options 2 (“Somewhat important”) and 3 (“Very important”) represented a similar construct, and the intervention was considered important. Therefore, response options 0 (“Would not use”) and 1 (“Not very important”) were collapsed and labeled “Not important,” and response options 2 (“Somewhat important”) and 3 (“Very important”) were collapsed and labeled “Important.” Frequency and percentage of responses for each intervention category were calculated separately for each case.
To examine the influence of the diagnostic label on selection of intervention, we calculated the percentage of respondents who considered each intervention category as important for the 2 most prevalent diagnostic label codes for each case. Then, we conducted separate chi-square analyses for each intervention category to examine the relative importance of the intervention category (important versus not important) based on the diagnostic label code used (“label code 1” versus “label code 2”).
Results
Demographics
Of the 877 surveys that were mailed electronically, 135 responses were received. Of the 135 responses, 107 surveys were completed adequately to use for data analysis, for a response rate of 12.2%. Of the 107 completed surveys, all were completed for the questions following the LBP case, and 103 were completed for the questions following the shoulder pain case. Respondents represented every geographic region in United States, and 55% were male. More than 90% of respondents were between the ages of 36 and 55 years, and 72% had been in practice for more than 15 years. Additional respondent characteristics are reported in Table 1. Geographic representation in the current sample very closely represents that of the population of OCS therapists. Although there were differences in how the age data were captured between our study and APTA, the current sample also fairly closely represents the population of physical therapists with an OCS certification based on sex, age, and practice setting.
Respondent Demographics, Training, and Practice Characteristicsa
LBP Case
In response to the closed-ended question related to classification system (question 13), most respondents (46.7%) indicated that the McKenzie system would be used to treat the patient described in the LBP case (Tab. 2). For responses to the open-ended question related to diagnostic label for the LBP case, the distribution of diagnostic label codes is illustrated in Figure 1. The most common types of diagnostic labels used were “Pathology/Pathophysiology” (32.7%) and “Combination” (of multiple constructs within a single label) (48.6%). One example of a diagnostic label that was pathology-based was “Herniated Nucleus Pulposus With Radiculopathy.” An example of a diagnostic label that included multiple constructs was “Lumbar Radiculopathy With Left Lower Extremity Pain and Impaired Muscle Performance.” This label was considered to include a combination of the following constructs: pathology/pathophysiology, pain location, and impairment. The least common types of diagnostic labels used were “Pain Location” (1.9%), “Movement System Impairment” (1.9%), and “Practice Patterns” (0.9%).
Percentage of Respondents Who Indicated Use of Each Classification System to Manage the Patient Described in the Low Back Pain and Shoulder Pain Cases
Percentage of respondents for each diagnostic label code for the low back pain case.
The percentages of respondents who considered each intervention category as important for the LBP case are included in the eTable. Greater than 80% of the respondents considered mobilization/manipulation, functional training, neuromuscular re-education, stretching/ROM, and strengthening important interventions. The greatest percentage of respondents ranked functional training as important (91.4%). The fewest percentage of respondents ranked modalities as important (18.3%). To examine the relative importance of interventions based on diagnostic label, subsequent analyses were performed only for respondents who used a pathology-based or combined diagnostic label (81.3% of respondents, n=87). No relationship was noted between diagnostic label (pathology-based versus combination) and relative importance (important versus not important) of any of the interventions for the LBP case (χ2=0.001–1.015, P=.31–.97). Figure 2 illustrates the lack of difference in interventions that were considered important between respondents who used a pathology-based diagnostic label compared with a combination diagnostic label.
Percentage of respondents who used a pathology-based or combined diagnostic label and considered each interventin category “important” for the low back pain case.
Shoulder Pain Case
In response to the closed-ended question related to classification system (question 2), the majority of respondents (57.9%) indicated that a pathoanatomic system would be used to treat the patient described in the shoulder pain case (Tab. 2). For responses to the open-ended question related to diagnostic label for the shoulder pain case, the distribution of diagnostic label codes is illustrated in Figure 3. The most common types of diagnostic labels used for the shoulder pain case were “Pathology/Pathophysiology” (57.3%) and “Combination” (of multiple constructs within a single label) (35.0%). An example of a combined diagnostic label provided for the shoulder pain case is “Right Shoulder Impingement and Supraspinatus Tendinopathy With Scapular Dysfunction.” This label included constructs related to “Pathology/Pathophysiology” and “Impairment.” Two examples of purely pathology-based diagnostic labels included are “Supraspinatus Tendinitis” and “Subacromial Bursitis.” The least common types of diagnostic labels used were impairment (1.0%) and pain location (1.0%).
Percentage of respondents for each diagnostic label code for the shoulder pain case.
The percentages of respondents who considered each intervention category as important for the shoulder pain case are shown in the eTable. Greater than 80% of the respondents considered mobilization/manipulation, functional training, neuromuscular re-education, stretching/ROM, and strengthening as important interventions. The greatest percentage of respondents ranked strengthening as important (98.1%), whereas the fewest percentage of respondents ranked balance training as an important intervention (13.0%). Modalities were considered important by 30.8% of respondents. To examine the relative importance of interventions based on diagnostic label, subsequent analyses were performed only for respondents who used a pathology-based or combined diagnostic label for the shoulder pain case (92.2% of respondents, n=95). No relationship was noted between diagnostic label (“Pathology-Based” versus “Combination”) and relative importance (important versus not important) of any of the interventions for the shoulder pain case (χ2=0.050–1.263, P=.261–.824). Figure 4 illustrates the lack of difference in interventions that were considered important between respondents who used a pathology-based diagnostic label compared with a combination diagnostic label.
Percentage of respondents who used a pathology-based or combined diagnostic label and considered each intervention category as important for the shoulder pain case.
Discussion
For the LBP case, the largest percentage of survey respondents assigned a diagnostic label that represented multiple constructs (48.5%). A label that includes multiple constructs may reflect the general view that LBP is multifactorial in nature.18,19 It also is consistent with the general consensus in the physical therapy literature that diagnostic labels based primarily on pathology or pathophysiology, such as those of the International Classification of Diseases (ICD), do not fully capture the clinical problems encountered in physical therapist practice.3,20–22 The single construct that was most prevalent in low back diagnostic labels was consistent with terminology related to pathology or pathophysiology. The single label most commonly applied to the LBP case was “Lumbar Radiculopathy.” This label represents a clinical phenomenon consistent with a cluster of signs and symptoms that implicates tissue sources of pain. However, it does not provide a direct link with the pathology or provide information on which to base intervention.
For the shoulder pain case, the majority of respondents (57.9%) assigned pathology- or pathophysiology-based labels. The fact that multiple constructs were used to label the LBP case and a single construct was used to label the shoulder pain case may provide evidence to support the theory that there are differences in the general approach to diagnosing low back and extremity problems. These data are consistent with the findings of a previous study that showed clinicians were more likely to consider pathology and tissue source of pain important in the diagnostic process for extremity conditions.14 It is noteworthy that our findings are based on the presentations of one patient case for LBP and one patient case for shoulder pain. Had we included multiple case studies, our results might have been different. However, the depth of information presented in the case studies would have been sacrificed for a potentially greater breadth of patient cases.
With regard to relative importance of intervention categories, findings were similar for both the LBP case and the shoulder pain case. Overall, the majority of respondents considered active interventions important for both cases, whereas fewer respondents considered passive interventions important. Greater than 80% of the respondents reported that they would use functional training and various forms of therapeutic exercise to treat both patients. A smaller percentage of respondents reported they would utilize physical and soft tissue modalities. Selection of relatively active therapies over passive therapies is consistent with current clinical guidelines and best practice recommendations for cost-effective care, particularly for managing LBP.19,23
Considering the 2 most prevalent types of diagnostic labels for each case, we found no relationship between the labels assigned to the patient case and the relative importance of different interventions. Thus, there appears to be no relationship between the labels that the therapists with an OCS certification used for the patient problems provided in the cases and the type of intervention they considered important to manage the patient's condition. Given that diagnostic labels categorize patients' problems, the labels can be viewed as a form of classification.20 However, an important function of diagnostic classification is to direct treatment. Our results do not support that the classification for the patient cases provided is directing treatment selection. These findings contrast with those of Dekker et al,24 which supported the relationship between diagnosis and intervention in physical therapist practice. However, in Dekker and colleagues' survey study conducted in the Netherlands, respondents were provided a very specific framework for assigning impairment-based diagnostic labels that were inherently more directly linked to treatment. Although impairment-based labels can be more directly linked to treatment, they do not encompass many other aspects of the condition. These conflicting results suggest that a single diagnostic label may not be able to serve the multiple purposes of diagnostic classification. Perhaps a single label cannot do everything.
Inconsistency in labeling patients' conditions as observed in the current study is not surprising given the lack of a standard taxonomy in physical therapy. The scope of physical therapist practice lies within the movement system, and because the movement system encompasses several hierarchical levels of body structures and functions, establishing a standard classification system is challenging. The classification systems currently used in practice are pragmatic systems primarily designed to direct treatment. There is considerable variance in both the manner in which patients' conditions are categorized with these systems and the way in which these conditions are labeled. Recognizing clinical variance in diagnosis is important because diagnosis is a central element of patient management; it is the focal point around which we interpret patient data and render clinical decisions related to intervention.
There were several limitations of the current study. The response rate was low, which can be attributed in part to the time demand on respondents to read and interpret 2 case presentations, in addition to answering both open-ended and closed-ended questions. The low response rate can negatively affect the external validity of this study and limit its generalizability to all OCS therapists and more broadly to non-OCS therapists. However, when considering the characteristics of all physical therapists in the country with an OCS certification, the current sample fairly closely represents this population with regard to geographic distribution and demographic characteristics such as age, sex, and practice area. Future research should continue to explore the consistency with which OCS therapists make clinical decisions related to diagnosis and selection of interventions, and perhaps look more broadly at diagnosis among non-OCS practitioners. Another factor to consider with the current study is the number of years of practice of the respondents. Ninety-nine percent had been in practice for at least 11 years, and 72% had been practicing for 16 years or longer. These findings could have led to a generational effect on the survey results. More experienced clinicians may be less likely to have had formal education on diagnostic classification or have not been exposed to classification approaches that have been more recently described.
Another important limitation of the current study is the potential for overlap with some of the constructs used within the diagnostic labels. For example, there are similarities in terminology associated with the concept of directional preference as it pertains to active movement test findings with the McKenzie and treatment-based classification systems.6,8 There also is overlap in terminology when using the practice patterns and impairment-based labels. This overlap could negatively affect the validity of the coding scheme used in analyzing the diagnostic labels. However, the diagnostic label codes were operationally defined and based on a consensus of 5 investigators. Another limitation is that the survey respondents were provided examples of diagnostic labels prior to being asked to label each case. Although the intent of providing examples of labels was to clarify the survey question, this approach may have had the unintended consequence of imparting bias into some of the responses such that respondents may have included diagnostic labels that were similar to the examples provided. Furthermore, the case presentation itself may have introduced bias and been suggestive of a pathology-based diagnostic label. However, the cases describe very common clinical presentations for people with LBP and shoulder pain. Last, although the intervention choices presented to respondents reflect standard treatment categories found in the literature, these choices nonetheless may not have provided enough specificity to discern a relationship between diagnosis and treatment.
In conclusion, it appears that our respondents, who were clinical specialists in orthopedic physical therapist practice, did not label a patient problem based on a clinical case in a consistent manner. Not only is there variation in the terminology used to describe each clinical condition, but assigned labels actually represent different constructs. These limitations likely reflect, in part, the lack of a universally accepted taxonomy in physical therapy to classify the clinical phenomena. They also may reflect the inherent limitation of using labels to describe clinically complex conditions and to serve multiple purposes. The inconsistency in naming clinical conditions has practical implications for communications with other health care professionals and third-party payers, variance in clinical practice, and research.
Appendix 1.
Low Back Pain Case
a ROM=range of motion, N/A=not applicable, R=right, L=left, MR=medial rotation, LR=lateral rotation, LE=lower extremity, WNL=within normal limits, SLR=straight leg raising, FABQ=Fear-Avoidance Beliefs Questionnaire, FABQPA=Fear-Avoidance Beliefs Questionnaire–physical activity subscale, FABQW=Fear-Avoidance Beliefs Questionnaire–work subscale.
Appendix 2.
Shoulder Pain Case
a IR=internal rotation, ER=external rotation, AROM=active range of motion, PROM=positive range of motion, MMT=manual muscle testing.
Appendix 3.
Survey Questionsa
a OCS=Orthopaedic Certified Specialist, CCS=Cardiopulmonary Certified Specialist, ECS=Clinical Electrophysiologic Certified Specialist, GCS=Geriatric Certified Specialist, NCS=Neurologic Certified Specialist, SCS=Sports Certified Specialist, ROM=range of motion.
Appendix 4.
Operational Definitions for Diagnostic Label Codesa
a ICD=International Classification of Diseases, APTA=American Physical Therapy Association, MDT=mechanical diagnosis and therapy.
Footnotes
Both authors provided concept/idea/research design, writing, data analysis, and consultation (including review of manuscript before submission). Dr Miller-Spoto provided data collection, project management, and study participants.
The authors thank the following graduates of the Nazareth College Physical Therapy Program for their assistance in the development of the survey and collecting and analyzing the data as part of their student/faculty research project: Amanda Gatch, PT, DPT, Megan Goetz, PT, DPT, Nicole Nolan, PT, DPT, and Brooke Redding, PT, DPT.
A poster presentation of this research was given at the Combined Sections Meeting of the American Physical Therapy Association; February 8–11, 2012; Chicago, Illinois.
- Received June 10, 2013.
- Accepted February 14, 2014.
- © 2014 American Physical Therapy Association
References
Issue highlights
