Background and Purpose. Opponents of direct access to physical therapy argue that physical therapists may overlook serious medical conditions. More information is needed to determine the ability of physical therapists to practice safely in direct-access environments. The purpose of this study was to describe the ability of physical therapists to make decisions about the management of patients in a direct-access environment. Subjects. Of a random sample of 1,000 members of the Private Practice Section of the American Physical Therapy Association, 394 participated. Methods. A survey included 12 hypothetical case scenarios. For each case, participants determined whether they would provide intervention without referral, provide intervention and refer, or refer before intervention. The percentage of correct decisions for each group of scenarios was calculated for each participant, and participants were classified as having made correct decisions for 100% of cases or less for each group. Three sets of logistic regressions were completed to determine the characteristics of the participants in relation to the decision category. Results. The average percentages of correct decisions were 87%, 88%, and 79% for musculoskeletal, noncritical medical, and critical medical conditions, respectively. Of all participants, approximately 50% made correct decisions for all cases within each group. The odds of making 100% correct decisions if a physical therapist had an orthopedic specialization were 2.23 (95% confidence interval=1.35–3.71) for musculoskeletal conditions and 1.89 (95% confidence interval=1.14–3.15) for critical medical conditions. Discussion and Conclusion. Physical therapists with an orthopedic specialization were almost twice as likely to make correct decisions for critical medical and musculoskeletal conditions.
It has been said that “No other health profession has patient access to its services so severely limited, and few other health professions are so demonstrably well qualified for practice without referral” than physical therapy.1 Indeed, since 1957, the American Physical Therapy Association (APTA) has lobbied at both the state and the national levels to change laws and practice acts so that physical therapists can obtain direct access, allowing them to manage patients and clients without referral from a physician or other health care practitioner.2 At present, some form of direct access is allowed in 39 states within the United States, with 48 states allowing physical therapists to examine and evaluate patients without a physician referral.3 The APTA recently developed a vision statement supporting direct access to physical therapy services: “By 2020, physical therapy will be provided by physical therapists who are doctors of physical therapy, recognized by consumers and other health care professionals as the practitioners of choice to whom consumers have direct access for the diagnosis of, interventions for, and prevention of impairments, functional limitations, and disabilities related to movement, function, and health.”3
Promoters of direct access argue that it can extend consumers’ choice of health care providers, improve access to services that promote prevention and wellness, and reduce delays in receiving physical therapy services. Additionally, physical therapists have been shown to have adequate knowledge to practice safely in the direct-access environment. Childs et al4 studied the knowledge of experienced physical therapists in the US uniformed health services and physical therapist students in managing musculoskeletal (MS) conditions. Although they did not directly compare physical therapists’ scores with those of physicians, on the basis of data from previous studies of physicians and medical students,5,6 the authors concluded that the knowledge of experienced physical therapists was greater than that of medical students, physician interns and residents, and all physician specialists except for orthopedists. Physical therapists who were board certified in orthopedic or sports physical therapy achieved significantly higher scores and passing rates than non–board-certified professionals. Additionally, in a study of physical therapy provided to over 50,000 patients seen through direct access at 25 military clinical sites, no adverse events were reported over a 40-month period.7 Daker-White et al8 reported that physical therapists and physicians seeing patients with orthopedic problems did not differ in their abilities to make diagnoses or in the amounts of consultation with senior physicians. Additionally, the patients did not differ in their outcomes. The patients, however, were all referred by general practitioners and had been screened for urgent medical problems.
Opponents of direct access argue that physical therapists may overlook serious medical conditions because they may not be able to refer a patient directly for diagnostic testing and are not trained to make medical diagnoses.2 Leerar et al9 found that physical therapists regularly (≥85% of cases) documented the presence or absence of only 9 of 17 “red-flag” items in patients with lumbosacral spine conditions. For example, weight loss was documented in only 5% of records. Similarly, Riddle et al10 found that approximately 25% of physical therapists would not contact physicians when the probability of deep vein thrombosis in patients with MS conditions was high. These findings suggested that more information is needed to determine the performance of physical therapists in a direct-access situation.
The purpose of this study was to describe the ability of physical therapists to make decisions about whether management by a physical therapist or referral to a medical professional was appropriate for hypothetical patients seeking care without a referral. We also examined the factors associated with making appropriate decisions.
A random sample of 1,000 members of the Private Practice Section (PPS) was obtained from APTA. Individuals were sent an initial mailing of a paper survey questionnaire with the option to complete and return it via US mail or to complete it online through a Web site link. Nonrespondents were sent a reminder e-mail approximately 2 weeks after the first mailing with a link to the survey Web site. A second paper mailing was sent to those who did not reply to either the first mailing or the e-mail. Data collection occurred May through August 2005. We also secured information from APTA regarding the demographic characteristics of all members of the PPS to compare with our sample.
Five survey questionnaires were returned through the mail as undeliverable. Of the remaining questionnaires, 39.9% (n=399) were returned. Five individuals indicated that they did not want to complete the survey, leaving 394 usable questionnaires. Eighty-one percent of the participants practiced in direct-access states. Ninety percent of the participants practiced in settings in which more than 50% of their patients had orthopedic conditions. Twenty-nine percent of the participants had a master’s degree, and 3% had a doctorate of physical therapy (DPT) at entry level; 30% practiced with a degree beyond entry level. Twenty-five percent of the participants had an orthopedic specialization (OS), including American Board of Physical Therapy Orthopaedic Certified Specialist (OCS), Fellow of the American Academy of Orthopaedic and Manual Therapy (FAAOMPT), or any other indication of formal orthopedic or manual therapy specialization indicated by the participant. Nearly 79% of the participants had been practicing for more than 10 years (Tab. 1). The regional distribution of the participants was similar (P<.05) to the regional distribution of all members of the PPS, with most of the participants residing in the Pacific and South Atlantic regions of the United States. Participants differed from members of the PPS in terms of entry-level (professional) degree and years of experience, with more of the sample having baccalaureate entry-level degrees (P<.01) and more than 10 years of experience (P<.01). Data were not available on the number of PPS members with an OS as defined in this study.
We created a survey instrument that included 12 case scenarios that represented conditions for which we believed patients might come to physical therapists’ offices without a referral (Tab. 2).11–35 Conditions were either medical conditions outside the realm of physical therapist practice alone or MS conditions that would be appropriate for physical therapists to manage without referring the patient for medical management. The medical conditions were considered either noncritical or critical on the basis of our perception of the urgency of the need for medical attention. The case scenarios provided short initial patient histories with a description of the patient’s signs and symptoms. On the basis of the scenarios, participants were asked to determine a course of management. Management decisions included 3 choices: providing physical therapy intervention to the patient without medical referral, providing intervention and then referring the patient to a medical professional, or referring the patient to a medical professional before any physical therapy intervention.
Case Content Validity
Case scenarios were based on information from literature that included information on symptoms, risk factors, and differential diagnosis for MS problems and problems that might mimic MS problems. Cases were developed to include common symptoms or risk factors for MS or non-MS conditions or the presence or absence of “red flag” symptoms, that is, symptoms that are associated with an increased likelihood of serious conditions.11–14 For example, Boissonnault and Bass15 noted that mechanical MS dysfunction is classically aggravated or alleviated by changes in position or movement. Saunders16 listed restricted and painful movements and localized loss of active mobility and passive mobility as examination findings for many MS spinal conditions. Pain with palpation and pain relieved by rest are also symptoms related to mechanical MS dysfunction.15,16 Patients’ reports of insidious symptoms not precipitated by an incident or accident, not related to movement or posture changes, and worsening with rest or interfering with sleep should arouse caution and suspicion.15 Additionally, risk factors for certain medical conditions should be considered when determining the likelihood of serious conditions that might be manifested as back pain.15,17 Focal neurological deficits or a confusing clinical picture also should arouse suspicion of serious conditions.17 Table 2 includes the cases, brief explanations, and relevant citations to support their content validity.
The cases were reviewed and revised by 4 expert physical therapists on 2 separate occasions. All of the expert physical therapists had practiced for more than 6 years, and 2 had practiced for more than 20 years. Two experts had spent more than 10 years in private practice, and 1 had spent over 20 years in private practice. Two had advanced master’s degrees, 3 had transition DPT degrees, and 3 had an OCS. Modifications to the case scenarios were made on the basis of their input.
Participants were told in a cover letter included with the survey questionnaire that the purpose of the study was to describe the ability of physical therapists to determine the appropriateness of treatment or referral for a patient being evaluated through direct access. The letter acknowledged the fact that in all cases described in the survey instrument, a physical therapist likely would carry out more tests before making a final decision, and explained that scenarios were brief to keep the survey from being burdensome for busy people.
Descriptive statistics were used to characterize the participants and their practice settings as well as the percentages of participants who were able to correctly determine the correct management decision in each case. Decisions about the management of each patient in the case scenarios were classified as correct or incorrect. We defined a correct decision for the 5 MS conditions as physical therapy intervention without medical referral or physical therapy intervention with follow-up medical referral. An incorrect decision was medical referral prior to physical therapy intervention. We defined a correct decision for the noncritical medical conditions as physical therapy intervention with follow-up medical referral or medical referral prior to intervention. For the critical medical conditions, we defined a correct decision as medical referral prior to any physical therapy intervention.
The mean percentage of correct management decisions was calculated for each group of case scenarios (MS, noncritical medical, and critical medical). Participants also were classified as having made correct management decisions for 100% of the items or for fewer than 100% of the items for each group of case scenarios. We then completed 3 sets of logistic regressions, 1 set for each group of case scenarios, to create models to explain the factors associated with making correct management decisions. In each case, the dependent variable was the decision classification (100% of items in the group correct or not), and the independent variables were years of experience (≤10 years or 10 years), direct-access state (yes or no), OS (yes or no), highest degree earned (entry level or beyond entry level), entry-level degree (baccalaureate or postbaccalaureate), and percentage of patients with orthopedic conditions (≤50% or 50%). To determine a parsimonious model to explain decisions for each group of case scenarios, we first examined bivariate relationships and then added variables in a forward selection process, stopping when the addition of a variable did not yield a significant partial F value.36
The ages of the patients seen at the practices of the participants varied widely. The largest percentage of patients was adults 41 to 60 years of age (∼33%). Ninety-eight percent of the participants reported collecting a routine medical history for all patients, and 94% reported collecting information about medications for more than 75% of their patients. Only about 7% of the participants reported checking vital signs for more than 75% of their patients.
Across all cases related to MS conditions, participants made a correct management decision for 87.3% (SD=15.8%) of them. For each case individually, 61.7% to 98.4% made a correct management decision (Tab. 3). Fifty percent of the participants made a correct management decision for all MS cases (Tab. 4). Participants with an OS were more likely to make a correct management decision for MS cases (odds ratio [OR]=2.23, 95% confidence interval [CI]=1.35–3.71), and participants whose patient caseload was more than 50% patients with orthopedic conditions were more likely to make a correct management decision (OR=2.23, 95% CI=1.05–4.74) (Figure). When both variables were entered together in a logistic regression model, the variable representing caseload was not significant (Tab. 4).
Noncritical Medical Conditions
Across all cases related to noncritical medical conditions, participants made a correct management decision for 87.8% (SD=14.7%) of them. For each case individually, 59.4% to 99.7% made a correct management decision (Tab. 3). Fifty-six percent of the participants made a correct management decision for all noncritical medical cases (Tab. 4).
Critical Medical Conditions
Across all cases related to critical medical conditions, participants made a correct management decision for 79.0% (SD=24.1%) of them. For each case individually, 67.6% to 93.4% made a correct management decision (Tab. 3). Nearly 50% of the participants made a correct management decision for all critical medical cases. Participants with an OS were more likely to make a correct management decision for critical medical cases (OR=1.89, 95% CI=1.14–3.15) (Tab. 4, Figure).
The findings of the present study support those of Leerar et al,9 who suggested that there may not always be adequate attention to known “red flag” signs or symptoms. These findings also are somewhat consistent with those of Riddle et al,10 who estimated that 25% of physical therapists would not refer patients with a high probability of deep vein thrombosis to physicians. Our results show that, in general, physical therapists make correct decisions regarding the management of hypothetical patients when the problems are MS in nature and can be managed by a physical therapist. They are less often correct in making decisions about medical conditions that require referral to a medical practitioner. For 4 of 5 MS cases, more than 90% of the participants made the correct management decision. For 2 of the 3 critical medical cases, however, fewer than 80% of the respondents made the correct decision. Exploring our data further, we determined that for the 3 critical medical cases (cases 5, 9, and 12), 3%, 7%, and 1% of the participants would not have made a medical referral at all. These data are of concern when one considers the consequences of lack of referral in some critical situations and suggest that physical therapists may require more education related to screening for medical conditions and differential diagnosis.
Other authors have provided evidence that physical therapists are at least as well equipped as physicians to provide care in a direct-access environment; however, these studies included samples with high proportions of physical therapists in the uniformed health services.4,7,37 Physical therapists in these services routinely function in a primary care role. For example, Childs et al4 found that physical therapists with at least 1 year of experience in the uniformed health services in the United States achieved an average score of nearly 76% on an examination requiring participants to respond to questions related to orthopedic diagnoses. Questions were related to both common problems encountered in primary care environments and emergency situations that would require immediate referral. Orthopedic medicine residents taking the same examination scored, on average, 74%.5 In a study comparing the diagnostic accuracies of physical therapists, orthopedic surgeons, and non–orthopedic medicine providers, Moore et al38 found that the accuracies of physical therapists and orthopedic surgeons were greater than that of non–orthopedic medicine providers, with no differences between physical therapists and surgeons for patients with MS injuries. In addition, it has been shown that physical therapists in primary care settings, whether military or civilian, are more likely to include identification of non-MS signs and symptoms in their patient management than physical therapists in non–primary care settings and are more likely to view such identification as an important professional responsibility.37
Childs et al4 also found that physical therapists with an OS scored better on the examination related to orthopedic diagnoses than those without an OS; this finding is similar to the findings of the present study. It is possible that physical therapists who are orthopedic specialists or who work in primary care settings have recognized the need for strong skills in differential diagnosis and have sought continuing education in advanced orthopedic clinical courses, contributing to greater knowledge and better decision making. Additional knowledge also may result from specialized and focused practice and preparation for the specialization examination. On the other hand, Riddle et al10 reported that board certification status did not affect therapists’ decisions to refer hypothetical patients to physicians. In some settings, physical therapists who are involved in primary care are required to demonstrate competencies that include screening for medical conditions such as diabetes, hypertension, and cardiac disease and differential diagnosis of MS versus non-MS conditions.39 The findings of the present study provide support for such a requirement given the effect of having an OS on management decisions.
Nearly 40% of the participants chose conservative management (referral before physical therapy intervention) for the scenario representing a MS problem for which chest pain was a symptom (case 10). This finding suggests that physical therapists understand that the symptom of chest pain could be cardiac in origin and, therefore, would refer the patient for medical care prior to any physical therapy intervention. Pain in the chest has been described as a “red flag” symptom.40 However, pain of MS, rather than cardiac, origin is increased by palpation and chest wall movements.15,16 For case 9 in the present study, only 77% of the participants made a correct management decision. In this case, we included information that suggested pain that was not likely of MS origin: pain that was unrelated to movement or posture, constant, and progressively worsening. Constant pain anywhere in the body has been described as a “red flag” symptom.40 The relatively low percentages of participants making correct management decisions for the problems represented in cases 10 and 9 may indicate that physical therapists need more emphasis on “red flag” symptoms and differential diagnosis for conditions that include thoracic pain as a symptom. On the other hand, approximately 93% of the participants identified the “red flag” symptoms for case 12. It is possible that the problem represented in case 12 is more commonly encountered in practice or more frequently emphasized in educational settings, references, or both.
The implications of the present study are constrained by its limitations. The instrument used to collect information on decision making was researcher designed. The case scenarios provided to the participants were short, and the limited information likely contributed to the difficulty in identifying the correct management decision. It is probable that physical therapists would collect additional data from the physical examination and history before making a final decision regarding whether or not to make a referral. Additional data from the examination likely would improve decision-making accuracy. We believe, however, that were the cases longer, busy practitioners would have been less likely to participate in our study. Additionally, the cases were hypothetical and included descriptions of symptoms proposed in the literature to be suggestive of MS conditions or medical conditions. To improve the authenticity of the cases, they were reviewed by a panel of expert physical therapists and edited on the basis of their feedback. The panel was small, however, and there was no attempt to match the panel’s credentials to those of the population being studied. The panel contained only physical therapists and not physicians. Therefore, there is no corroboration of the medical diagnoses represented in the cases except through our interpretation of the literature.
We have no information on how often patients with such symptoms may come to the average physical therapist’s office, although the literature suggests that the most severe conditions are rare. For example, Deyo et al12 cite a prevalence of 0.7% for patients with spinal malignant neoplasms presenting in primary care settings. We also do not know whether physical therapists would make similar decisions with actual patients presenting with the same scenarios.
The response rate of approximately 40% was relatively low, and there may have been some response bias, in that those who found the case scenarios difficult may not have participated at as high a rate as those who were comfortable with the material. Such a bias is suggested by the difference in years of experience and board certification status between our sample and the population of PPS members. Additionally, the reliability of the survey instrument was not sufficiently tested.
Interpretation of the percentage of participants correctly determining management decisions must be made in light of the fact that there was a 67% chance of being correct for the MS and noncritical medical cases and a 33% chance of being correct for the critical medical cases. In other words, for 1 of the MS cases (case 10) and for 1 of the noncritical medical cases (case 2), physical therapists did no better than chance in determining correct management.
Finally, our sample included only physical therapists who were members of the PPS of APTA. These physical therapists may be a relatively select group who, through their membership in APTA and the section, demonstrate their interest in professionalism, including continued competency and high standards for patient care. We do not have information on how this group compares to the general membership of APTA or to physical therapists who are not members of APTA. The sample was similar to members of the PPS in terms of geographic region but had more experience. Their experience may have resulted in better decision-making skills than might be expected, on average, among all members of the PPS, thus providing results that overestimate the ability of physical therapists to make good decisions.
On the basis of the limited information in the case scenarios, more than 90% of the participants in the present study made correct management decisions for 8 of 12 hypothetical patients with a variety of medical and MS conditions. For 2 of 3 critical medical conditions, however, less than 80% of the physical therapists recognized the need for immediate referral. Physical therapists with OS credentials were more likely to make correct decisions for patients with MS conditions and critical medical conditions. The results suggest the need for further emphasis on education in medical screening, identification of “red flag” symptoms, and differential diagnosis.
All authors provided concept/idea/research design, writing, and data collection. Dr Jette provided data analysis and project management.
This study was approved by the Institutional Review Board of Simmons College.
The study was partially funded by the Simmons College Faculty Fund for Research.
This research, in part, was presented as an abstract and poster presentation at PT 2006: Annual Conference and Exposition of the American Physical Therapy Association; June 21–24, 2006; Orlando, Fla.
- Received December 19, 2005.
- Accepted July 24, 2006.
- Physical Therapy