Physical therapy and health outcomes in patients with knee impairments.

Key Words: Knee injuries; Lower Extremity, knee; Outcome and process assessment (health care).

Musculoskeletal impairments account for a large percentage of conditions for which medical care is sought in the United States,[1] and there are significant costs related to the disabilities caused by these impairments. In 1984, Cunningham and Kelsey,[2] using data from the National Health and Nutrition Examination Survey (NHANES) I, estimated that 32.6% of individuals in the United States between the ages of 25 and 74 years were affected by some type of physician-observed musculoskeletal impairment and 29.7% of this population had self-reported musculoskeletal impairments. Impairments related to the spine had the highest prevalence, followed by impairments related to the knee. In the National Medical Care Utilization and Expenditure Survey (NMCUES), about 20% of the 1980 noninstitutionalized population reported having a musculoskeletal problem involving the back or joints that resulted in some type of disability or use of the health care system.[3]

Conservative treatment of musculoskeletal impairments often includes physical therapy, and there may be a trend toward increased use of physical therapy.[4] The NMCUES indicated that 13.3% of total charges for treatment of musculoskeletal conditions was attributed to care given by health care professionals other than physicians, including physical therapists.[3] In a national survey of outpatient physical therapy providers, knee or hip pain was the second most frequently cited reason for physical therapy care.[5] Moderate osteoarthritis of the knee has been shown to be related to reductions in self-reported health, influencing ambulation, recreation, sleep and rest, and emotional behavior.[6] Although physical therapy is usually initiated with the belief that these reductions in health can be minimized or prevented, there is little evidence in the literature that any particular treatment approach produces better health outcomes than another.

The goals of our study were (1) to describe the health outcomes of patients with knee impairments receiving physical therapy and (2) to determine the relationship of these health outcomes to the demographic and clinical characteristics of the patients and the characteristics of the physical therapy episode of care.

Method

Subjects

The data for the study were derived from a database generated by the Focus on Therapeutic Outcomes (FOTO) network.(*) In 1993 and 1994, the (FOTO) network was a privately funded consortium of five leading outpatient rehabilitation companies developed for the purpose of generating an outcome-oriented, standardized information management system for use in outpatient physical therapy settings. The companies were required to maintain an agreed-on data quality standard and record completion rate to remain in the FOTO network. The database for this time period contains information from 1,308 patients admitted for physical therapy, care for knee impairments. Because not all patients completed an episode of care or completed all data collection forms, the sample used for this study, consisted of 426 patients who had a completed episode of care and filled out both initial and discharge health outcomes questionnaires. The care was provided in 63 physical therapy practices by in physical therapists over a 1-year period beginning in July of 1993. The project was approved by the Human Subjects Review Board of New England Research Institutes, which was responsible for the design and administration of the database during that year.

Practices. The practices were located across the United States: 2% from the Middle Atlantic States, 23% from the South Atlantic States, 8% from the South Central States, 57% from the North Central States, and 12% from the Mountain States. The practices ranged in size, with 27% having one full-time physical therapist, 35% having two full-time physical therapists, 23% having three full-time physical therapists, and 15% having more than three full-time physical therapists.

Physical therapists. The average age of the physical therapists was 32.1 years (SD=6.7, range=22-60). Seventy-two percent were women. The highest credential of 84% of the physical therapists was a bachelor's degree. Ten percent of the therapists had an entry-level master's degree, and 6% had an advanced master's degree. The average years of practice was 7.8 (SD=6.1, range=<1.0-33). Eighty-three percent of the therapists worked full time, treating an average of 52, patients per week.

Patients. Table 1 shows the demographic and clinical characteristics of the 426 patients providing data for this study.

Table 1.
Demographic and Clinical Characteristics of Patients With Knee
Impairments

Characteristic

Age (y)
  X                        42.7
  SD                       14.5
Gender
  Female                    52%
  Male                      48%
Ethnicity
  White                     90%
  Black                      6%
  Native American           <1%
  Asian                     <1%
  Hispanic                   2%
Employment
  Full-time                 38%
  Light duty                 9%
  Off because of health     31%
  Retired                   12%
  Unemployed                10%
Education
  No high-school diploma     8%
  High-school diploma       59%
  College degree            25%
  Graduate degree            8%
Income
  <$15,000                   8%
  $15,000-$25,000           15%
  $26,000-$35,000           16%
  $36,000-$45,000           13%
  >$45,000                  30%
  Refused to answer         18%
Acuity
  Acute                     17%
  Subacute                  60%
  Chronic                   23%
Surgery
  Yes                       44%
  No                        56%
Depressed
  Yes                       25%
  No                        75%
Comorbidities
  None                      56%
  1 category                28%
  2 categories              13%
  More than 2 categories     3%
Impairment level
  X                         7.4
  SD                        2.4

Treatments. Table 2 shows the percentage of patients receiving various physical therapy treatments over the episode of care.

Table 2.
Characteristics of the Episode of Care for Patients With Knee
Impairments
                           Percentage of Patients
                           Receiving This Type of
Characteristic             Treatment

Mobilization               35%
Flexibility exercises      82%
Strengthening exercises    95%
Endurance exercises        72%
Massage techniques         15%
Heat modalities            37%
Cold modalities            67%
Episode length (D)
  X                       32.8
  SD                      26.2

Procedure

Data obtained from the patients through standardized questionnaires at the initial physical therapy visit included age, gender, ethnicity, education, income, employment status, comorbid conditions, history of depression, duration of the knee problem, and surgical history. The physical therapist provided information about the total duration of the episode of care and type of treatments provided.

Each patient completed the standard form of the Medical Outcome Study 36-Item Short-Form Health Survey (SF-36)[7] and the Lysholm Knee Rating Scale[8] at initiation and completion of treatment. The SF-36 queries the patient concerning health over the past 4 weeks in eight different health dimensions: energy/fatigue, general health perception, mental health, bodily pain, physical functioning, role limitation due to emotional problems, role limitation due to physical problems, and social functioning. The SF-36 provides a profile of health with a score for each of the eight dimensions ranging from 0 to 100; higher scores represent better health. The reliability and validity of the SF-36 have been extensively tested.[9,10]

The Lysholm Knee Rating Scale assesses eight areas related to knee dysfunction: limp, support, locking, instability, pain, swelling, ability to squat, and ability to negotiate stairs. The test-retest reliability is .97, and the scale has been shown to correlate with patient activity levels and other functional measures.[8] The Lysholm Knee Rating Scale provides a comprehensive score from 0 to 100, with 100 representing the best status.

Data Analyses

Health outcomes scores for patients with knee impairments at the initial and discharge physical therapy visits were described by graphically comparing gender- and age-controlled SF-36 scores of these patients with the adult population norms in the United States.[11] The comparisons were made by computing standard scores, which were derived by subtracting the mean score for patients from the mean general population score on each of the scales and dividing by the standard deviation of the general population score.[12].

Changes in health for patients with knee impairments were estimated by computing effect sizes for each of the SF-36 scales and the Lysholm Knee Rating Scale. Effect size presents change in a scale as it relates to the standard deviation of initial scores for that scale, and is computed by subtracting the mean initial score from the mean final score and dividing the result by the standard deviation of the initial score.[13] Demonstrating change in this way allows a person to compare the changes occurring across a variety of scales.

To examine the relationship of health outcomes to demographic and clinical characteristics of the patients and characteristics of the physical therapy episode, nine sets of analyses were conducted, with follow-up scores on the SF-36 and Lysholm Knee Rating Scale as the dependent variables and baseline scores as covariates. The independent variables of interest consisted of treatment characteristics and relevant demographic and clinical characteristics of the patients. Table 3 describes the variables in detail.

Table 3.
Independent Variable

Variable                 Type          Measurement

Patient demographic
  Age                    Continuous    Years
  Gender                 Categorical   Male, female
  Income                 Categorical   <$15,000, $15,000-$25,000,
                                         $26,000-$35,000,
                                         $36,000-$45,000, >$45,000,
                                         refused to answer
  Education              Categorical   No high-school diploma,
                                         high-school diploma or
                                         some college, college
                                         diploma or some graduate
                                         work, graduate degree
  Employment             Categorical   Full time, light duty, off
                                         because of health,
                                         retired, unemployed
  Ethnicity              Categorical   White, Black, Hispanic,
                                         Asian, Native American
Patient clinical
  Comorbidity            Categorical   Disease(s) in 0 categories,
                                         1 category, 2 categories,
                                         > 2 categories
  Depression             Categorical   Depressed within past year,
                                         not depressed within past
                                         year
  Surgery                Categorical   Surgery related to current
                                         knee problem, no surgery
  Acuity                 Categorical   < 2 wk duration, 2 wk to 6
                                         mo, > 6 mo
  Impairment             Continuous    Score (0-35) based on part
                                         of Lysholm Knee Rating
                                         Scale (higher is worse)
Treatment
  Mobilization           Categorical   Having mobilization as part
                                         of treatment, not having
                                         mobilization as part of
                                         treatment
  Flexibility exercise   Categorical   Having flexibility exercise
                                         as part of treatment, not
                                         having flexibility
                                         exercise as part of
                                         treatment
  Strength exercise      Categorical   Having strengthening
                                         exercise as part of
                                         treatment, not having
                                         strengthening exercise as
                                         part of treatment
  Endurance exercise     Categorical   Having endurance exercise as
                                         part of treatment, not
                                         having endurance exercise
                                         as part of treatment
  Massage techniques     Categorical   Having massage as part of
                                         treatment, not having
                                         massage as part of
                                         treatment
  Cold modalities        Categorical   Having cold modalities as
                                         part of treatment, not
                                         having cold modalities as
                                         part of treatment
  Heat modalities        Categorical   Having heat modalities as
                                         part of treatment, not
                                         having heat modalities as
                                         part of treatment

Episode Length           Continuous    Total days from admission to
                                         discharge

Initially, univariate analyses were used to determine the relationship between the patient characteristics and outcomes. An alpha level of .01 was used to determine statistical significance. Then, using general linear models and a backward deletion process, treatment and patient variables were determined for models controlling for initial SF-36 or Lysholm Knee Rating Scale scores and those patient variables that had a significant univariate correlation to the outcome score in the initial step. Because we hypothesized a possible interaction of surgical status and treatment with health outcomes, two-way interaction terms were included for all treatment variables. An alpha level of .05 was used as the criterion for a treatment or patient variable to remain in a model. Only data of patients with complete data for the independent variables of interest were included in the analyses Due to missing data for some variables, the number included in each set of analyses varied. All analyses were performed using SAS software.([dagger])

Results

To facilitate discussion of the results, it is helpful to conceptualize the scales of the SF-36 as measuring two basic components of health: physical and psychological.[14] Patients with knee impairments had poorer physical health than the general population. At the initial physical therapy visit, physical function, role limitations due to physical problems, and bodily pain were each nearly 2 standard deviations below the US adult population norms (Fig. 1). The psychological dimension of health was considerably less affected than physical health in patients with knee impairments, with mental health, role limitation due to emotional problems, and energy/fatigue less than 0.30 standard deviation below the general population norms. Social function was nearly 1 standard deviation below the population norm.

Health status in all but one scale improved over the course of physical therapy care, which lasted, on average, 33 days (Fig. 2). The most notable changes occurred in the physical dimension of health, as measured by physical function (effect size=0.81), bodily pain (effect size=0.88), and the Lysholm Knee Rating Scale (effect size=0.93). Moderate changes occurred in role limitation due to physical problems (effect size=0.56) and social function (effect size=0.52).

The multivariate models explaining health outcomes are shown in Table 4. Significant models (P[less than or equal to].05) were generated for all but one scale, health perceptions. The models explained 23% to 37% of the variability in health outcomes, depending on the specific outcome scale. Depression was related to poorer health outcomes in all scales for which significant models were generated. Patients who had knee surgery had better outcomes than patients who had not had knee surgery, primarily in the physical health scale. Greater initial impairment, defined as knee locking, swelling, or instability, was associated with poorer outcomes in physical health. Episode length was not related to any health outcomes. Treatment that included mobilization led to better outcomes in mental health, and treatments that did not include heat were associated with better outcomes in bodily pain and energy/fatigue. The effect of endurance exercise on the Lysholm Knee Rating Scale depended on whether the patient had had surgery. Inclusion of endurance exercise for patients who had received surgery was associated with better outcomes; the inclusion of endurance exercise for patients who had not had surgery was not associated with outcome.

[TABULAR DATA OMITTED]

Discussion

The results of this study provide new knowledge concerning the health outcomes of patients with knee impairments receiving physical therapy care. Physical health is substantially diminished in the presence of knee impairment, and changes in health outcomes for patients with knee impairments occur most notably in this domain. This study is also the first, to our knowledge, to explore the factors related to health outcomes in patients with knee impairments seen in outpatient physical therapy practice. Patients who were older, had no surgery, were out of work because of their health, and were depressed consistently had poorer outcomes following physical therapy intervention compared with patients without these characteristics.

Katz and colleagues[15] have also used the SF-36 to describe the physical function of patients undergoing arthroscopic meniscectomy. They reported a mean preoperative physical function slightly higher than that of our patients. With a mean follow-up time of 18.5 months, the effect size was 1.26 somewhat greater than the effect size (0.81) for patients undergoing physical therapy in our study. Cohen[16] suggested the following interpretation of effect size: 0.2 to 0.4 is small, 0.5 to 0.7 is moderate, and 0.8 or greater is large. Although Cohen has qualified effect size to facilitate interpretation, the numerical value does not directly assess the clinical importance of a change. For most measures, the clinical importance of any change is determined by observation of many cases and the functional implications of variations in the value. The clinical relevance of the change in health in our sample may be most evident in Figure 1. This figure demonstrates the degree to which a change in health brings a score closer to the norms at discharge from physical therapy.

Approximately 56% of the patients in our study had had surgery, and similarities can be found between our findings and those of Katz and colleagues.[15] The poorer outcomes associated with being off work because of poor health may well reflect the same effect noted by Katz et al[15] for patients receiving Workers' Compensation. Although the working status of their subjects was not reported, one might expect that those receiving workers' compensation were not working because of their poor health. Our finding that poorer outcomes were associated with greater impairment could also be considered similar to the finding of Katz et al[15] that the degree of cartilage damage was related to outcome. Summers et al[17] also reported a moderate univariate relationship of disease severity with measures of physical health in patients with osteoarthritis.

Previous reports have shown that patients with knee impairments have problems in the psychological health domain.[6-18] Although the mental health score on the SF-36 was near the general population norm for the patients in our study, 25% of the patients reported that they had experienced depression in the past year. The relationship of depression to health outcomes over an episode of outpatient physical therapy has not been previously described for patients with knee impairment. Anxiety and depression have been shown to be related to increased levels of disability in patients with osteoarthritis of the hip or knee.[17-19] Summers et al[17] found that more of the variability in physical health was accounted for by anxiety and depression than by disease severity. Reasons for this effect can only be surmised, but some evidence suggests that the more effective patients' coping strategies, the more likely they are to adapt to chronic disease with low levels of disability.[20]

In contrast to our previous working involving spinal impairments, we found that type of physical therapy intervention did not consistently help explain health outcomes for patients with knee impairments once patient demographic and clinical variables were considered. The application of heat was associated with poorer outcomes for pain. This effect persisted even though the acuity of the problem was considered in the analysis. The effect of heat on pain outcomes man, well reflect the tendency for heat to increase tissue inflammation and, therefore, pain. The effect of treatment with mobilization techniques on mental health outcomes cannot be readily explained. One possibility is that treatment with mobilization techniques requires more direct hands-on, one-on-one care than the other types of treatments provided, thus allowing greater communication with patients and subsequently improved mood and effect.

We did not find a consistent relationship between health outcomes and type of exercise provided, although patients who had had surgery, had better health outcomes as measured by the Lysholm Knee Rating Scale when endurance exercise was part of the physical therapy regimen. Similarly, in a randomized study of the effects of a supervised physical therapy program that included fitness walking for patients with osteoarthritis of the knee, Knee et al[22] found greater improvements in pain and physical activity in the walking group. Whitelaw et al[23] reported that patients with knee pain who participated in a recommended home exercise program at least twice per week were able to increase activity intensity with less pain than those who did not follow through with home exercise. In a study on spinal impairments,[21] we also found an effect of endurance exercise on outcome.

Our finding that the effect of endurance exercise on the Lysholm Knee Rating Scale scores depended on whether the patient had had surgery led us to further explore the relationship of surgical status to treatment provided. We found that patients who had had surgery tended to receive a different pattern of care. They received more of each type of exercise during the episode of care and were more likely to receive combinations of exercise than patients who had not had surgery. For example, although only 2% of patients received no exercise, these were all patients who had not had surgery.

The lack of a consistent relationship between health outcomes and physical therapy treatment might be explained by, the high degree of commonality in treatment type among patients. In an unpublished study by Jette and Delitto (Alan M Jette, Anthony, Delitto; unpublished research), the majority of episodes of care for patients with knee impairment included both exercise and passive forms of treatment, regardless of the phase of the episode. In further analysis, we found that 60% of the episodes of care included strength, endurance, and flexibility, exercises; 31% of the episodes of care included two of the three types of exercise. Nearly all episodes of care included strengthening exercises. Inclusion of specific exercises in a treatment regimen was not related to the degree of knee impairment (locking, instability, and swelling) or acuity of the problem. Although the common perception is that physical therapists focus on impairments and consider the acuity or chronicity of a problem in designing their treatments, we have not found evidence for this belief. With little variability in treatment episodes, health outcomes can be expected to be associated with clinical and demographic variables rather than treatment characteristics.

Limitations

Our study was observational in nature, which precludes conclusion about the effectiveness of physical therapy treatments. Patients were not randomly assigned to receive treatments, and unknown factors may have influenced treatment choices made the physical therapists. In addition, individual physical therapists indicated whether they had used particular treatment approaches. These approaches (eg, mobilization) were broadly defined, and no attempt was made to assess the reliability, of therapists, entries with the database. We also did not account for the subtle changes in patients' conditions within and between individual treatment sessions and the treatment adjustments that occur in response to these changes. Patients were simply classified as to whether they had at any time received a particular treatment during the episode of care.

No attempt was made to classify patients according to diagnoses. The database used for this study also lacked data that physical therapists might consider important in caring for patients with knee impairments, for example, range of motion and muscle force. In addition, the standard form of the SF-36, in use at the time of the study, queried the patient concerning his or her health over the past 4 weeks. This time period may, have been too long to detect changes in health, especially if the episode of care was brief Such a limitation, however, would lead to an underestimation of the changes that occur over an episode of care.

The findings reported here are based on analysis of a clinical database. These data were generated for the major purposes of quality assurance and business decision making, and using them for predicting outcomes presents limitations. Because of these limitations, this study should not be viewed as testing hypotheses. Rather, the study has generated hypotheses that must be further tested concerning the relationships between health outcomes and factors such as patients' depression and work status and application of heat modalities, and whether specific physical therapy interventions affect health outcomes.

The limitations related to the use of clinical databases for research have been clearly outlined by Pryor and Lee[24] and include the use of data to answer questions not determined a priori, missing observations, and other biases that limit validity. These biases include selection bias and referral bias. In addition, this database included large numbers of variables that needed to be reduced in some way to facilitate interpretation of the predictive models. Such reduction results in the loss of some information and the increased chance of misclassification. In spite of reducing variables to a manageable number and selecting variables that measured unique attributes, many predictor variables were examined. This large number of variables, defined in unique ways by the investigators, is likely to result in idiosyncratic models. Certainly, further studies are needed to test the validity of the predictions.

In spite of these limitations, this study provides information not previously reported concerning the correlates of health-related outcomes in patients with knee impairments. Although observational studies using clinical databases cannot provide the same rigorous results as clinical trials concerning the effectiveness of treatments, they can provide variable information for conceptualizing efficient and effective future trials.

Conclusion

Individuals with knee impairments demonstrate decrements in the physical domain of health. Over a course of physical therapy care, measures of health outcomes show improvement. Patients at risk of poorer health outcomes are those who are older, depressed, have not had surgery, or are out of work because of their health. Type of physical therapy does not seem to have a consistent effect on health outcomes once demographic and clinically related variables are considered. Because of the design of this study, it cannot provide evidence for the effectiveness of physical therapy in patients with knee impairments, and future studies are needed to validate the models reported here. With these caveats in mind, the findings suggest that physical therapists consider adjustments in their examinations to identify those patients who are potentially at risk for poor health outcomes. With additional information about patients at risk, referrals might be sought for psychological or vocational counseling or treatment approaches might be considered that focus on the problem areas.

(*) FOTO, PO Box 11444, Knoxville, TN 37939. ([dagger]) SAS Institute Inc. SAS Campus Dr, Cary, NC 27513.

References

[1] Kelsey JL, White AA, Pastides H, Bisbee GE. The impact of musculoskeletal disorders on the population of the United States. J Bone Joint Surg [Am]. 1979; 61:959-964. [2] Cunningham LS, Kelsey JL Epidemiology of musculoskeletal impairments and associated disability. Am J Public Health. 1984; 74:547-579. [3] Murt H, Parsons PE, Harlan WR, et al. Disability, utilization, and costs associated with musculoskeletal conditions. Natl Med Care Util Expend Surv C. 1986; 5:1-64. [4] Frymoyer JW, Cats-Baril WL. An overview of the incidences and costs of low back pain. Orthop Clin North Am. 1991; 22:263-271. [5] Jette AM, Smith K, Haley SM, Davis KD. Physical therapy episodes of care for patients with low back pain. Phys Ther. 1994; 74:101-115. [6] Mattsson E, Brostrom L. The physical and psychosocial effect of moderate osteoarthrosis of the knee. Scand J Rehabil Med. 1991; 23:215-218. [7] Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36), 1: conceptual framework and item selection. Med Care. 1992; 30:473-483. [8] Tegner Y, Lysholm J. Rating systems in the evaluation of knee ligament injuries. Clin Orthop. 1985; 190:43-49. [9] Stewart AL, Hays RD. Ware JE. The MOS short-form general health survey (SF-36). Med Care. 1988; 26:724-732. [10] McHorney CA, Ware JE, Lu R, Sherbourne CD. The MOS 36-item short-form health survey (SF-36), III: tests of data quality, scaling assumptions, and reliability, across diverse groups. Med Care. 1994; 32:40-66. [11] Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Survey Manual and Interpretation Guide. Boston Mass: The Health Institute, New England Medical Center; 1993. [12] Garratt AM, Ruta Whitelaw DA, Abdalla MI, et al. The SF-36 health survey questionnaire: an outcome measure suitable for routine use within the NHS? BMJ. 1993; 306:1440-1444. [13] Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in the health status. Med Care. 1989; 27:S178-S189. [14] McHorney CA, Ware JE. Raczek AE. The MOS 36-item short-form health survey (SF-36), II: psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care. 1993; 31:247-263. [15] Katz JN, Harris TM, Larson MG, et al. Predictors of functional outcomes after arthroscopic partial meniscectomy. J Rheumatol. 1992; 19:1938-1942. [16] Cohen J. Statistical Power Analysis for the Behavior Sciences. New York, NY: Academic Press Inc: 1977. [17] Summers MN, Haley WE, Reveille JD, Alarcon GS. Radiographic assessment and psychological variables as predictors of pain and functional impairment in osteoarthritis of the knee or hip. Arthritis Rheum. 1988; 31:204-209. [18] Drewett RF, Minns RJ. Sibly TF. Measuring outcomes of total knee replacement using quality of life indices. Ann Royal Coll Surg Engl. 1992; 274:286-290. [19] Salaffi F. Cavalieri F. Nolli M. Ferraccioli G. Analysis of disability in knee osteoarthritis: relationship with age and psychological variables but not with radiographic score. J Rheumatol. 1991; 18:1581-1586. [20] Keefe FJ, Caldwell DS, Queen K. et al. Osteoarthritic knee pain: a behaviors analysis. Pain. 1987; 28: 309-321. [21] Jette DU, Jette AM. Physical therapy and health outcomes in patients with spinal impairments. Phys. Ther. 1996; 76:930-945. [22] Kovar PA, Allegrant JP, MacKenzie R, et al. Supervised fitness walking in patients with osteoarthritis of the knee. Ann Intern Med. 1992; 116:529-534. [23] Whitelaw GP, Rullo DJ, Markowtiz HD, et al. A conservative approach to anterior knee pain. Clin Orthop. 1989; 246: 234-237. [24] Pryor DB, Lee KI, Methods for the analysis and assessment of clinical databases: the clinician's perspective. Stat Med. 1991; 10:617-628.