| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
SK Carter, PT, PhD, OCS, is Assistant Professor, Department of Physical Therapy, College of Health Science and Human Service, University of Toledo–Health Science Campus, Mail Stop 1027, 3000 Arlington Ave, Toledo, OH 43614-2598 (USA)
JA Rizzo, MD, is Professor, Department of Preventive Medicine and Economics, and Director, Center for Health Services and Outcomes Research, Stony Brook State University of New York, Stony Brook, NY
Address all correspondence to Dr Carter at: stephanie.carter2{at}utoledo.edu
Submitted July 14, 2005;
Accepted January 8, 2007
 | Abstract |
|---|
Subjects: The subjects were adult noninstitutionalized civilians who participated in the Medical Expenditure Panel Survey from 1996 to 2000 and who had at least one musculoskeletal condition (N=18,546).
Methods: Logistic regression was used to identify predisposing, need, and enabling variables associated with receiving outpatient physical therapy services.
Results: Factors that were positively associated with receiving physical therapy services were having more than one musculoskeletal condition, having some limitation in function, having 7 or more ICD-9 (International Classification of Diseases, 9th Revision) codes, having a college or advanced degree, and residing in an urban area. Factors that were negatively associated with receiving physical therapy services were being older than 65 years of age, having no high school degree, Hispanic ethnicity, African-American race, having public insurance or no insurance, and living in any US census region besides the Northeast.
Discussion and Conclusion: The results of this study indicate that health- and non–health-related factors influence outpatient physical therapy use.
| Introduction |
|---|
 TopAbstract Introduction MethodResultsDiscussionConclusionAppendix.References |
|---|
Musculoskeletal conditions affect both physical and mental aspects of health-related quality of life. Forty-two percent of people with a musculoskeletal condition report some activity limitation primarily attributable to that condition.3 The decline in physical function attributable to a musculoskeletal condition may begin in midlife and contributes to a significant portion of all disability in later years.4 The decline in physical function increases the risk for depression and poor mental health.5 In some people, the physical and mental limitations are severe enough to affect their ability to work and can lead to chronic disability.3
Physical therapy interventions are directed toward eliminating or reducing pain, restoring function, and educating patients about the appropriate management of their conditions.6 Therefore, physical therapy services may be an appropriate health care resource for patients with musculoskeletal conditions.6 Many studies7–11 have shown physical therapy intervention to be effective for various musculoskeletal conditions. The results suggest that physical therapy services may be essential for providing cost-effective, high-quality care to people with musculoskeletal conditions.7–11
Although people with musculoskeletal conditions account for the majority of outpatient visits to physical therapists, many do not receive physical therapy services.12–15 It is not known how many people with musculoskeletal conditions actually need physical therapy services, and no criteria exist for making this determination. Although health-related factors, such as severity of condition, number of symptoms, and functional limitations, are associated with receiving physical therapy services, other non–health-related factors influence the use of outpatient physical therapy (OPT) services. A few studies have investigated non–health-related determinants of the use of physical therapy services for patients with musculoskeletal conditions.
The most common condition used for investigating determinants of physical therapy intervention has been spine dysfunction. Mielenz et al13 reported that patients with acute and subacute low back pain were more likely to receive physical therapy services if they were educated beyond high school and covered by workers’ compensation insurance. In a more recent study investigating patients with back and neck pain,14 the authors reported that receiving physical therapy services was positively associated with postsecondary education, workers’ compensation insurance, and having taken legal action. For these patients, the use of physical therapy services was negatively associated with being older than 50 years of age and living in the South and Midwest regions of the United States rather than the Northeast.14
Besides spine disorders, other musculoskeletal conditions have been studied with regard to use of physical therapy services. Patients with lower-extremity trauma were more likely to receive physical therapy intervention if they had some college education, higher income, and private insurance and were female.16 Community-dwelling adults who were 65 years of age or older were more likely to receive physical therapy intervention in any treatment setting if they had an income greater than $30,000, had supplemental private or managed-care insurance, and lived in the Pacific region of the United States rather than the South Atlantic.15 Years of education and physical therapist supply also were positively associated with physical therapy use. These older adults were less likely to receive physical therapy services if they were accompanied on physician visits, lived in the East South Central region of the United States rather than the South Atlantic, and were 80 years of age or older.15
In many situations, OPT services are reimbursed only when a patient is referred by a physician. Therefore, physician referral and type of physician seen affect the use of physical therapy services. Patients with musculoskeletal disorders were more likely to be referred if they were seen by an orthopedic surgeon or an osteopathic physician than if they were seen by a primary care physician or an allopathic physician, respectively.17 Having workers’ compensation insurance and greater physical therapist supply also were positively associated with a referral to a physical therapist.17 Patients with spine disorders and being seen in US spine centers were more likely to be referred to a physical therapist if they were college educated and had a high expectation of improving function.18 Patients with spine disorders were less likely to be referred to a physical therapist if they were receiving disability insurance payments, 50 years of age or older, and male.18
Research also has been done on this same topic in other countries. In the Netherlands, patients with hip and knee pain, 55 to 74 years of age, diagnosed with arthritis, and referred to a specialist were more likely to receive physical therapy intervention than patients who were not referred to a specialist.19 In a study of Canadian patients with osteoarthritis of the hip or knee, patients who had a total joint replacement were more likely to be referred to a physical therapist if they had a more severe presentation of illness.20 Patients who did not have a total joint replacement were more likely to be referred to a physical therapist if they had more days in which their level of activity was limited, were older, and had few comorbidity diagnoses.20
As evidenced by the research reviewed, other factors besides functional limitations, such as need, affect the use of physical therapy services. However, these studies have limitations. First, a few articles focused specifically on patients with spinal dysfunction.13,14,18 Although this patient group makes up a significant percentage of patients seen in outpatient centers, it is also important to investigate the use of physical therapy services by people with other musculoskeletal conditions. Second, a focus on people 65 years of age or older has implications for Medicare therapy policies and reimbursement, but use of physical therapy services by people under age 65 years has broad implications for many private and public third-party payers.15
Third, research in other industrialized countries shed light on the use of physical therapy services and factors that may affect it; however, the health care systems in these countries are quite different, and the results may not be generalizable to health care in the United States.19,20 Fourth, investigation of patterns of physician referral to a physical therapist may not relate to the actual use of outpatient physical therapy services.17 Finally, these studies addressed physical therapy in many different treatment settings, and the results may not relate specifically to the use of outpatient physical therapy services.15,16 Outpatient therapy is quite different from therapy received in an inpatient setting. Because people who use OPT services live in the community and may continue to work, attendance may be hindered by other factors. It is important to study the OPT setting independently from other therapy settings because Medicare and many private insurers have separate outpatient benefits with specific limitations.
Because of the limitations of previous research, further investigation into the determinants of outpatient physical therapy care was warranted. In this study, we attempted to overcome some of these limitations by using a nationally representative sample of people with a variety of musculoskeletal conditions. The purpose of this study was to investigate the health-related and non–health-related determinants of the use of OPT services by people with musculoskeletal conditions. The results of this study will help stakeholders, such as physicians, patients, third-party payers, and health care policy makers, understand whether OPT services were used appropriately or inappropriately by people with musculoskeletal conditions.
Although illness and health-related factors affecting the use of physical therapy services reflect differences in need, non–health-related factors suggest either underuse or overuse of services.14,15 If underuse is occurring, physical therapists may want to develop policies or programs that specifically target certain patient groups. If overuse is occurring, physician and insurance stakeholders may want to develop criteria that outline the need for OPT services in order to make appropriate referral and reimbursement decisions. This research also is important with regard to planning for future needs and costs of OPT services for people with musculoskeletal conditions.
| Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionAppendix.References |
|---|
The survey consisted of an overlapping, unbalanced panel design in which people were interviewed in 5 rounds every 5 to 6 months over 24 months to provide 2 years of longitudinal data (Figure).22 The interview was administered as a computer-assisted personal interview of the head of the household. The head of the household provided proxy responses regarding health care information for each person living in the home.23
|
The person-level file was combined with the condition-level file by use of a unique identifier. Each record in the condition-level file represented a unique medical condition reported for each person. Condition-level information was collected at each interview and recorded verbatim by the interviewer. The conditions then were coded into ICD-9 (International Classification of Diseases, 9th Revision) codes by professional coders.23 At each interview, updates were made with regard to new conditions and health care use for both new and existing conditions. For this study, an attempt was made to link condition with receiving OPT services. However, not all physical therapy visits had conditions linked to them, and such links would have resulted in a much smaller sample size. The combined data set contained information regarding demographics, medical procedures, functional status, and medical conditions collected for each person during the year.
To produce estimates of health care use for a calendar year, we pooled the data across the 2 distinct nationally representative MEPS panels. For example, data for the year 1997 include the second year of data from the 1996–1997 MEPS panel and the first year of data from the 1997–1998 MEPS panel (Figure). Therefore, one person may be represented in 2 different years of data.26 Because each year of the MEPS sample is nationally representative, it was acceptable to use data from multiple years in this manner.
Subjects
Subjects from the linked MEPS databases from 1996 to 2000 were selected for analysis when at least one Clinical Classifications Software (CCS) code was related to the diagnosis of a musculoskeletal condition. Clinical Classification Software codes are clinically meaningful groups of ICD-9 codes that were developed by the AHRQ.27 The CCS codes and related ICD-9 codes used for this study are listed in the Appendix. Musculoskeletal conditions for this sample were typical of patients using OPT services and were consistent with those in other studies of health care use (Tab. 1).6,27,28 The sample included subjects 16 years of age or older and having at least one musculoskeletal condition during the years 1996 to 2000, resulting in 18,796 subjects. The age of 16 years or older was used because many variables (such as marital status and education attainment) in the MEPS database were constructed with this cutoff. Missing values in some of the response items caused 250 observations to be lost, leaving a sample of 18,546 subjects available for the multivariable analysis.
|
The independent variables selected for investigation were based on previous research with the Andersen-Newman model, the availability of data in the MEPS database, and the knowledge and experience of the authors.29 In accordance with previous studies,14,15,17,29 the determinants were categorized as predisposing, need, and enabling.
Individual characteristics present before illness may affect the propensity to use health care services. These predisposing factors include demographic, social structure, and attitudinal-belief variables. For the purpose of this study, the following variables were available for use as predisposing variables in the MEPS database: age, sex, race, educational attainment, and marital status. People of different age groups are affected by different types of diseases and illnesses. Age was determined by using a person's birth date to calculate age at the earliest interview of the year. Given the chronic nature of some musculoskeletal conditions, such as arthritis, age may affect the use of OPT. Sex was investigated because men and women may have different beliefs and behaviors about the use of health care services that may affect the use of OPT services.
Level of education has been used in other studies as a proxy for socioeconomic status and has been shown to affect the use of physical therapy services.13,14,16 Disparities in health care use based on race or ethnicity have been found and therefore were investigated to determine whether these differences exist in use of OPT services.30–33 Race and ethnicity were collected as separate data points during the MEPS interview; however, during editing and management of the MEPS data, an imputed variable combining race information and ethnicity information was created (RACETHNX) and used instead of 2 separate variables (Tab. 2). Marital status at the first interview of the year was investigated because the support of another person may affect the use of care. Never being married, being, widowed, or being divorced was considered to be "not married" (Tab. 2).
|
Need variables were transformed into either dichotomous or categorical variables in order to interpret the logistic regression results. The number of musculoskeletal conditions was dichotomized into 1 and greater than 1. For the number of ICD-9 codes, a frequency distribution was performed, and the variable was divided into 4 categories based on the resultant quartiles of the distribution (1 or 2, 3, 4–6, and 
7 ICD-9 codes) (Tab. 2). For limitations of physical function, respondents were asked to rate their performance on 8 functional activities as follows: no difficulty (1), some difficulty (2), a lot of difficulty (3), or unable to do (4). The functional activities surveyed were lifting 4.5 kg (10 lb), walking up 10 steps, walking 3 blocks, walking 1.6 km (1 mile), standing for 20 minutes, bending or stooping, reaching overhead, and grasping with fingers. All 8 of the physical function scales were summed to create a physical difficulty score with a range of 8 to 32. The physical function variable was created from this score and was dichotomized, with 8 indicating no difficulty and greater than 8 indicating some difficulty with at least 1 functional activity (Tab. 2).
The enabling factors included the availability of resources and access to care.29 The enabling variables investigated were rural versus urban residency, geographic region of residence, barriers to care, and insurance status. The availability of resources was measured by whether the person lived in a rural or an urban area and the geographic region of residence (Tab. 2). Living in an urban area was expected to have a positive effect, as the density of OPT services may be greater there. The geographic region of residence was investigated because physical therapy resources may not be equitably distributed throughout the country. Binary variables indicating the year in which an observation was collected were included to investigate intertemporal variations in the use of OPT services. In addition, previous research into area variations suggested differences in use based on geography.14,34
Access to care was measured by the person's (or proxy's) perception of barriers to care. During the interview, respondents were asked whether any of the following limited access to care: insurance, transportation, communication, physical access, work schedule, child care, time, did not know where to obtain care, or refused care. Affirmative responses were summed, and a dichotomous variable (no=0 affirmative responses; yes=1 or more affirmative responses) was created (Tab. 2). Having a barrier to care was expected to exert a negative effect on receiving physical therapy intervention. Respondents also were asked about payment for medical care at any time during the year (ie, insurance status). Three payment categories were developed from a more detailed list used during the interview: any private insurance (including Champus/VA), public insurance only, or no insurance (Tab. 2).
Data Analysis
Multivariable logistic regression was used to predict the odds that a person would receive OPT services and to identify the impact of the predisposing, need, and enabling factors mentioned above on the likelihood of receiving OPT services. Multivariable logistic regression is a standard approach for modeling the likelihood of an event occurring (in this case, OPT) as a function of multiple factors that potentially explain the event. These factors are entered into the model simultaneously. Although multicollinearity can be a concern with multiple independent or explanatory variables, correlations among the variables were quite low. As an additional test for multicollinearity, tolerance and variance inflation factors were examined for each variable. In logistic regression, variance inflation factor values above 2.5 may be cause for concern.35 The results of these tests were all below 2.5 and indicated that multicollinearity was not a problem in this analysis (Tab. 3).
|
.05. Wald 95% confidence intervals (CIs) were reviewed to examine the magnitude of the effect. Continuous explanatory variables were transformed into categorical variables to facilitate the interpretation of odds ratios (ORs) obtained from the logistic regression analysis. For ordinal variables, 2 or more dummy variables were created on the basis of a logical grouping of data, essentially making them categorical variables. For example, the number of ICD-9 codes was grouped according to the quartile distribution. Dummy variables also were created when categorical variables had more than 1 level. In the analysis, one dummy variable was used as the reference so that the OR estimates for other levels in that category would have a basis for comparison. For example, high school degree was assigned as the reference variable for education level. People with other education levels then were interpreted as being more or less likely to receive OPT services than those with a high school degree. The estimation for the reference variable will always equal 1. An OR of greater than 1 implies that the use of OPT services was more likely for people who had the characteristic than for those who did not. An OR of less than 1 implies that the use of OPT services was less likely for people who had the characteristic than for those who did not. Estimation was performed with SAS version 8 software.* | Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionAppendix.References |
|---|
The likelihood ratio test for the goodness of fit of the model was significant (
2=516.42, P<.0001). Interactions among variables such as age and physical difficulty were assessed, but none were found to be significant. The ORs obtained from the multivariable logistic regression analysis are shown in Table 4. Significant need factors with an OR of greater than 1 were the presence of more than 1 musculoskeletal condition (OR=2.10, 95% CI=1.84–2.39), at least some difficulty with physical function (OR=1.52, 95% CI=1.31–1.76), and having more than 7 ICD-9 codes (OR=1.33, 95% CI=1.11–1.49). People with more than 1 musculoskeletal condition were more than 2 times as likely to receive OPT services as people with only one musculoskeletal condition. People with a combined physical difficulty score of 9 or more (indicating some difficulty with at least one functional task) were 1.52 times as likely to receive OPT services as people who had no to little limitation of physical function. People with more than 7 ICD-9 codes were 1.33 times as likely to receive OPT services as people who had 1 to 6 ICD-9 codes. In contrast, self-perceived health status was not associated with the odds of receiving OPT services.
|
With regard to the enabling factors, geographic differences existed in the odds of receiving OPT services, with people in the Midwest (OR=0.71, 95% CI=0.60–0.84), South (OR=0.55, 95% CI=0.46–0.65), and West (OR=0.76, 95% CI=0.64–0.90) census regions being less likely to receive OPT services than people in the Northeast census region. Urban residency had a positive influence on the use of OPT services (OR=1.28, 95% CI=1.10–1.48). Insurance status also affected the use of OPT services. People with public insurance (OR=0.66, 95% CI=0.55–0.80) or no insurance (OR=0.57, 95% CI=0.43–0.74) during the year were less likely to receive OPT services than those with private insurance at any time during the year. Having a perceived barrier to care did not affect the likelihood of receiving OPT services. The year indicators also did not reveal any significant intertemporal trend in the use of OPT services.
| Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionAppendix.References |
|---|
The presence of more than one musculoskeletal condition and other comorbidities has been shown to negatively influence a person's functional status and may affect the severity of the condition.41 For these reasons, both of these factors were positively associated with the use of OPT services; this finding may indicate that patients receiving OPT services for musculoskeletal conditions have complex conditions that affect their care. These complex conditions need to be addressed through the screening of comorbidities during the examination, modification of interventions to account for common diagnoses in other systems, and adjustment of the expected outcomes because of the effects of these conditions on healing and prognosis.42,43
In the present study, being 65 years of age or older was negatively associated with the use of OPT services. This result is supported by the findings of Freburger and Holmes15 in a study of community-dwelling older people; they found that people who were 80 years of age or older were less likely to receive physical therapy intervention in any treatment setting. These results are perplexing because function decreases with advancing age, and this decrease in function would seem to suggest a greater need for physical therapy services. Because people in this age group receive Medicare, policies that regulate OPT services may limit access to or coverage of these services or both.
However, other intrinsic factors of the older population may account for the negative association with the use of OPT services. For example, a person's living situation may affect the use of health care. Nelson et al44 found that people who were 65 years of age or older and who were living alone were less likely to receive health care services. This finding may be attributable to an inability to access care outside the home. Older people may have received physical therapy intervention at home, and these services were not accounted for in this analysis. Another explanation for the decreased use of health care services may be people's beliefs and attitudes. For example, older people may believe that physical decline is part of the aging process, a belief that may keep them from seeking care. Wittink et al45 found that although older people had lower levels of function than younger people, they had similar levels of pain and health status. The older people also reported having higher levels of control over their lives, lower levels of avoidance because of fear, and lower levels of passive coping than the younger people.45 These beliefs may cause them to not seek the care of a physical therapist.
Because the age of 65 years is consistent with enrollment in public insurance, that is, Medicare, the results of the present study are also consistent with people having public insurance being less likely to receive physical therapy services than those having private insurance. These results may be related to policies that affect the use of OPT services within the Medicare system. Although the Centers for Medicare and Medicaid Services and the Medicare Payment Advisory Commission reported a 2-fold increase in spending for outpatient therapy services between 2000 and 2004 (of which physical therapy is 
75%),46 the results of the present study revealed that older people with musculoskeletal conditions may not be getting the physical therapy care that they need. Physical therapists need to continue to work closely with the Centers for Medicare and Medicaid Services in order to ensure that changes in outpatient therapy policies do not adversely affect the functional outcomes of older people. Physical therapists also need to challenge the assertion that the increased use seen is overuse. Rather, this care may be necessary because of other factors within the health care system, such as early discharge from hospitals and skilled nursing facilities.
Level of education, as a proxy measure for socioeconomic status, was positively associated with the use of OPT services. This finding is consistent with those of several other studies. For community-dwelling older adults, the number of years of education was positively associated with receiving OPT services.15 People with spine dysfunction and lower-extremity trauma were more likely to receive physical therapy intervention if they had some postsecondary education.13,16 In contrast, people with spine dysfunction were less likely to receive a referral to a physical therapist if they had a high school education or less.18 These results may indicate that people of lower socioeconomic status may have reduced resources and access to medical care, including physical therapy services.
In the present study, African American race and Hispanic ethnicity negatively influenced the use of OPT services. These results are consistent with those of research in other areas of health care use by people with musculoskeletal conditions.30–32 However, the present study is the first study specific to OPT services to find a disparity in use based on race or ethnicity. It is not known whether the disparity in care is attributable to the members of minority groups not seeking care or not perceiving a need for care.47 The differences in use may be explained by social and cultural norms that relate to expectations and that cause members of minority groups to not seek health care services.47 Another reason for the differences in use may be a language barrier. Fiscella48 found that Spanish-speaking adults were less likely to see a physician, and this situation also may be the case for seeking OPT services.
The use of medical services has been shown to vary across urban and rural locations.34,49 The present study suggests that area variations in the use of OPT services may follow a similar pattern, with positive associations from residing in an urban area or residing in the Northeast census region.14 Area variations may be indicative of physical therapist supply; this notion is consistent with another study in which physical therapist supply was positively associated with use of physical therapy services by older adults in any setting.15 Other factors may influence the differences in care between rural and urban areas. Area variations in Canada have been shown to be affected by patient willingness to have elective surgery, the percentage of hospital beds in an academic medical center, the number of male physicians, the number of physicians trained in North America, the increased propensity for orthopedic surgeons to perform surgery, and the prediction of outcomes by orthopedic surgeons.50
Insurance status was shown to affect the use of physical therapy services in several other studies.13–18,51 The results of the present study are consistent with those of previous research in that people were more likely to receive physical therapy services if they had some type of private insurance, whether it was private commercial insurance, Medicare supplemental insurance, or workers’ compensation. It was understandable that people who had no insurance would be less likely to use any medical services, including OPT services, but it was a bit disconcerting that people with public insurance were less likely to use OPT services. Physical therapists should be concerned that public insurance policies restrict or limit access to physical therapy services.
Limitations
The present study is the first to use a nationally representative sample of adults with musculoskeletal conditions to investigate the use of OPT services. However, the study does have several limitations. The selection of subjects was based on the presence of a musculoskeletal condition. However, the chronicity and severity of the condition were not known; therefore, these variables were not controlled for in the present study. These variables may have affected the use of OPT services, as some research suggested that people with musculoskeletal disorders of a longer duration, with more severe symptoms, or both were more likely to receive physical therapy services.52 We attempted to use other measures indicating illness severity, complexity, or both (ie, comorbidity and function), but these attempts were limited by the data available in the MEPS.
The fact that data were reported by the head of the household also was a limitation. Self-reported data or data reported by a proxy may not be as reliable as data collected at the point of service.53,54 However, the results of the present study, including proxy responses, were similar to those obtained in another study in which interviews and Medicare claims data were used.15 The similarity of results regarding the use of physical therapy services from different secondary databases suggests the validity of the methods used in the MEPS.
The timing of data collection may also be a limitation. Functional status in the MEPS was collected at an interview during the first 6 months of the year (ie, either round 1 or round 3 of consecutive panels).23 The functional status of people with musculoskeletal conditions may change throughout the year given the nature of these conditions. Outpatient physical therapy services could have occurred at any time during the year to count as use of these services. Thus, it is not known how functional status corresponded to the timing of an OPT intervention.
The results are limited to an adult noninstitutionalized civilian US population. Therefore, it is not known whether the results can be generalized to people in the military or those living in long-term care or assisted-living facilities.
Future Research
Even given the limitations of the present study, the MEPS database has much potential for future research with regard to the use of physical therapy services. A next logical project would be to investigate determinants of use of physical therapy services for people with diagnoses other than those of musculoskeletal origin. For example, people with neurological diagnoses have different functional limitations and needs for physical therapy services. Although this patient population is not as large as the population of patients with musculoskeletal conditions, this group typically generates more visits and expenditures for physical therapy care.55 Understanding the determinants of physical therapy use also would help in planning for future needs and costs of care for these patients.
The MEPS database is extremely large and complex, with many more details regarding demographic, use, and expenditure variables than were used for the purpose of the present study. Therefore, further research could be performed with these details. For example, areas not explored in the present study were the amount of physical therapy services (ie, number of visits) and expenditures. This information could provide nationally representative benchmarks for the care of people with specific diagnoses. Currently, the Guide to Physical Therapist Practice6 provides a wide range of expected visits for each practice pattern. If the CCS code categories correspond to the practice patterns, then a range of expected visits based on evidence could be developed.
In order to better understand whether the pattern of use of OPT services is appropriate, functional outcomes before and after care should be examined. Because physical therapy intervention is directed toward the restoration of function, a change in function would be expected during the time frame of this service. This change in function could be compared to function in people who did not receive OPT services over the same period of time. Such a comparison may be possible with MEPS data by using a much smaller sample size and by linking use of OPT services to the collection of functional outcomes data. Besides the functional outcomes data used for the present study, Short-Form Health Survey (SF-12) data also are included in recent MEPS panels.
One interesting finding of the investigation of the use of physical therapy services was that respondents reported that other health care practitioners provided interventions used by physical therapists. In the present study, only physical therapy visits (ie, services provided by a physical therapist) were used, but 722 people reported that they received OPT services from a chiropractor. Although this finding may have been attributable to a misunderstanding by respondents regarding the differences in care provided by these practitioners, it may provide an opportunity to compare use (ie, number of visits) and expenditures between physical therapists and chiropractors for the treatment of people with musculoskeletal conditions.
| Conclusion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionAppendix.References |
|---|
| Appendix. |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionAppendix.References |
|---|
|
| Footnotes |
|---|
A platform presentation of this research was given at the Combined Sections Meeting of the American Physical Therapy Association; February 23–27, 2005; New Orleans, La.
* SAS Institute Inc, PO Box 8000, Cary, NC 27511. 
| References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionAppendix.References |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Deutscher, D. L Hart, R. Dickstein, S. D Horn, and M. Gutvirtz Implementing an Integrated Electronic Outcomes and Electronic Health Record Process to Create a Foundation for Clinical Practice Improvement Physical Therapy, February 1, 2008; 88(2): 270 - 285. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||
