Back Care Instructions in Physical Therapy: A Trend Analysis of Individualized Back Care Programs.

Back pain is one of the major health problems in Western industrialized countries.[1] The annual incidence has been reported to be about 5%.[2] Back pain is also one of the most frequent reasons for visiting a general practitioner or a physical therapist.[3-6] In the Netherlands, 22% of the patients referred by general pactitioners for physical therapy have back pain.[7] In the United States, patients with low back pain represent 25% of all outpatient discharges from physical therapy practices.[8] The primary physical therapy intervention in the treatment of people with back pain is often exercise[9,10] aimed at obtaining recovery with a minimal chance of relapse.[11] Another therapeutic method is to educate patients about anatomy, the natural history of disorders of the back, the principles underlying posture, back care during daily activities, and a healthy lifestyle.[12] Consequently, the effects of back care programs may depend on the patient's adherence to the therapist's instructions.[13] Adherence, however, may be difficult to obtain.[13-17] To facilitate treatment adherence, some authors recommend setting attainable goals in cooperation with the patient.[18] A manageable number of instructions, spread out in a logical sequence, seems like a practical strategy for patient education.[19] Thus, back care instructions require a careful planning. Knowledge of the content and sequence of back care instructions is needed to plan an educational program.

Thus, the first objective of our study was to investigate the content and sequence of back care instructions given by physical therapists. Although our study involved physical therapists in the Netherlands, its scope was much wider. We believe that not only do the Netherlands and the United States have comparable clinical guidelines for patients with low back pain[20,21] but therapists from both countries use the same categories of patient education.[22,23]

Our second objective was to explore similarities and differences among physical therapists. Research has shown considerable differences among physical therapists in the amount of information given to patients.[24] Comparable results have been reported for general practitioners[25] and nurses.[26] To enhance the quality of care, efforts have been undertaken by governmental agencies and professional associations to diminish treatment differences among physical therapists. Professional organizations are developing guidelines for practice that contain, for example, the optimal treatment for low back symptoms.[27,28] A clinical guideline regarding acute low back problems in adults[21] emphasizes patient education as one of the few moderately evidence-based interventions for people with low back pain. Connolly[29] commented that the guideline developed by the Agency for Health Care Policy and Research "emphasized telling patients what to do rather than training patients how to do" and that the guideline lacks items regarding exercise prescription and instruction in symptom control.

We believe that education should be tailored to the patient's needs and attuned to his or her individual situation.[14] Therefore, we examined the relative influence of patients and therapists on the number and content of instructions because we assumed that optimal care is reflected in more variability related to patients and in less variability related to therapists. This assumption is not to say that all therapists should act uniformly, but rather it emphasizes the importance of variability (ie, tailoring the educational experience) among different patients with the same therapist, because each patient may have different needs at different times during treatment. An individual tailoring of patient education to these different needs will become manifest in larger patient variability.

Theoretical Framework

In general, physical therapy for back problems, in our opinion, has 3 main objectives: pain relief, recovery of function, and prevention of recurrence. These objectives constitute a basis for palliative, curative, and preventive interventions and instructions.

During the pain management phase, back care instructions may concern applying warmth, taking rest, use of pain medication, and so on. Some authors[30] consider pain management to be preparatory for the primary intervention: exercise[30] Pain management instructions refer to the acute complaints. Patients are rewarded immediately for their adherence, and the correct execution of these instructions is under control of the therapist. Exercise therapy concerns the mobility of the lumbar spine, strengthening important muscle groups, and so on. By means of exercise instructions, patients learn to master the exercises. Again, these instructions are under control of the therapist, as he or she can assess the patient's adherence at the next visit. Patients receive many instructions about taking care of their back (ie, about lifting, posture, locomotion, and so on). During the course of treatment, additional attention may be given to instructions about future management of back problems. Here, the reward for the patient may be less immediate, and it is important to repeat these kinds of instructions throughout the entire series of therapy sessions. To prevent the recurrence of back problems, patients are also often given advice about general fitness and health behaviors.[31] Here, the patient is in charge and responsible for the long-term nonsupervised adherence.[14]

Given these considerations, we postulate the following optimal sequence of back care instructions:

1. Pain management instructions are given most often at the start of treatment and then gradually decrease in number.

2. Exercise instructions are introduced after the start of treatment and then repeated throughout the rest of the treatment.

3. Instructions about taking care of the back during activities of daily living are given throughout the course of treatment.

4. Recommendations about general fitness are mainly given toward the end of treatment, so they they gradually increase in number.

Method

Patients

In the Netherlands, the majority of all patients visit small, private outpatient practices.[32] Most of the patients are treated in a series of sessions that last almost a half hour each.:[33,34] Our database contained information on 1,151 therapy sessions for 132 patients with back pain who were treated by 21 physical therapists. The study was conducted in private outpatient practice. The average number of therapy sessions was 8.5 per patient (range=1-22), and the average number of patients was 6 patients per physical therapist (range=2-12). Data on the average age and sex of the patients are displayed in Table 1, which also contains information on the treatment goals.

Table 1.
Description of Patients (N = 132) and Treatments

                             [bar]   SD     Range   Percentage
                               X
Men                                                    41.7

Women                                                  58.3

Age (yr)                      42.8   13.9

Duration at start of
    treatment (wk)            11.3   17.1
No. of treatment goals         2.9    0.9
 Range of motion                                       65.2
 Pain reduction                                        53.0
 Posture improvement                                   30.3
 Muscle strengthening                                  26.5
 Facilitation of activities
   of daily living                                     19.7
 Muscle tone regulation                                15.2
 Education in body
   mechanics                                           15.9

Seriousness of complaints      2.2    0.8   1-4

Likelihood of recurrence       2.5    0.6   1-4

Psychosocial influence         2.1    0.8   1-4

Importance of exercises        3.2    0.8   1-4

Seven different treatment goals were identified. Treatment was most often aimed at increasing lumbar spine range of motion and pain reduction (Tab. 1). The average treatment had 3 goals.

Physical therapists also estimated on 4-point scales the likelihood of recurrence within a year (complaint will certainly not/probably not/probably/certainly recur), the influence of psychosocial aspects (psychosocial aspects have no/some/much/very much influence), and the importance of doing exercises in the prevention of back pain recurrence (exercises are not important/ somewhat important/important/very important to prevent recurrence).

Registration Form

The physical therapists recorded the kind of instructions given to their patients in each session by means of a registration form. In physical therapy, registration forms can be a reliable source of information,[35,36] The form contained 34 topics in 4 areas: instructions about pain management (8 topics), instructions about taking care of the back when performing daily activities (14 topics), instructions about doing exercises (9 topics), and recommendations on general fitness (3 topics) (Tab. 2). The registration form had 10 additional items that could be used freely by the therapist. The list was developed in 2 stages. First, all available information used in practice by physical therapists was explored. All kinds of unofficially published brochures and leaflets were investigated, supplemented by an overview of 70 back school programs by Knibbe et al[37] and by a book edited by Goeken[38] that contains descriptions of back management programs. From these sources, a list was compiled of all kinds of advice given to patients with back pain. We checked this list for completeness and condensed it into major categories. Second, this list was pilot tested by 4 experienced physical therapists to determine whether use of the form could be part of their daily routine.

Data Analysis

We analyzed the data by means of a special form of linear regression analysis: hierarchical linear modeling (HLM).[39,40] We used HLM to analyze the data for 2 reasons: (1) because the number of instructions is measured repeatedly during sessions[41] and (2) because we had a 2-stage sample of patients within physical therapist.[42] Thus, we had information about sessions, patients, and physical therapists. The data, therefore, were not from independent observations, violating a major assumption of traditional linear regression analysis.[43] In HLM, both of these factors are taken into account. Several health services research projects have applied HLM.[44-47] Data analysis was carried out by means of MLN software.[48]

We used HLM in 5 separate analyses to determine which factors were statistically significant predictors of the number of overall instruction statements as well as the number of statements in the areas of pain management, back care during activities of daily living, exercises, and fitness. Several classes of predictors were used. These classes of predictors included variables related to the progression of sessions (overall trend, first session, last session), patient characteristics (age and sex), treatment goals (range of motion, pain reduction, posture improvement, muscle strengthening, facilitation of activities of daily living, muscle tone regulation, and education in body mechanics), and therapist opinion about the case (the likelihood of recurrence within a year, the influence of psychosocial aspects, and the importance of doing exercises in the prevention of back pain recurrence).

Regression coefficients and their standard errors are used to test for statistical significance. If their division is greater than 1.96 or smaller than -1.96, the coefficient is statistically significant (alpha=.05), as in normal linear regression analysis. The proportion of the total variability that could be accounted tot by variance between therapists, patients, and sessions was calculated for each analysis.

Results

Back Care Instructions

As outlined in Table 2, a total of 6,078 instructions were given to 132 patients over 1,151 sessions. The mean number of instructions was 5.3 per session and 46 per patient. These were not all different kinds of instructions. Information was often repeated in subsequent sessions. On average, patients received 16 different instructions, so instructions were repeated an average of 3 times. Of the total of 6,078 instructions, 88.5 instructions concerned pain management. Resting to avoid pain (23.3% of the sessions) and doing analgesic exercises (19.4% of the sessions) were most common types of instructions in this area. Taking care of the back while performing daily activities was the focus of 2,379 instructions. Instructions on sitting posture (29.2% of the sessions) and standing posture (2.5.4% of the sessions) were most frequently encountered in this area. Information on how to avoid overloading the back (20.1% of the sessions) and how to alternate body position during daily activities (24.6% of the sessions) also was often given. Home exercises were the focus of 2,234 instructions. Exercises to increase the mobility of the lumbar spine were advocated most often (48.6% of the sessions). Exercises to strengthen the abdominal muscles (27.5% of the sessions) and the back muscles (22.0% of the sessions) also were often discussed. Information about physical fitness was less common, with 444 recommendations related to this topic area.

Table 2. Number and Percentages of Sessions of Information (1,151 Sessions, 132 Patients, 21 Physical Therapists)

                                    Number   Percentage

Pain management
  Resting                            268     23.3
  Doing analgesic exercises          223     19.4
  Using warmth (eg, showers,
    hot-water bottles)               141     12.3
  Using pain medication               50      4.3
  Seeking distraction                 43      3.7
  Not using pain medication           32      2.8
  Thinking consciously of
    something else                    32      2.8
  Ignoring pain                       16      1.4
  Other                               80      7.0
  Subtotal                           885

Back care in activities of daily
    living
  Sitting posture                    336     29.2
  Standing posture                   292     25.4
  Alternating body position          283     24.6
  Recognizing limitations of back    231     20.1
  Slowing down/taking it easy        208     18.1
  Handling stress/strain             197     17.1
  Using correct lifting technique    178     15.5
  Resuming labor                     104      9.0
  Adjusting height of table/desk     100      8.7
  Getting in/out of bed               83      7.2
  Choosing good chair                 77      6.7
  Wearing good shoes                  39      3.4
  Losing weight                       29      2.5
  Wearing support belt                26      2.3
  Other                              106      9.2
  Subtotal                         2,379

Exercises
  Mobility of lumbar spine           559     48.6
  Abdominal muscles                  317     27.5
  Back muscles                       253     22.0
  Leg muscles                        150     13.0
  Relaxation                         139     12.1
  Endurance                          111      9.6
  Stretching illiopsoas muscles       93      8.1
  Stretching hamstring muscles        93      8.1
  Gluteus muscle                     192     16.7
  Other                              327     28.4
  Subtotal                         2,234

General fitness
  Swimming                           124     10.8
  Walking/strolling                  104      9.0
  Cycling                            103      8.9
  Other                              113     10.2
  Subtotal                           444

Other
  Subtotal                           136

Grand total                        6,078

Instructions Throughout Treatment Sessions

Figure 1 displays graphically the number of instructions during the different sessions. The first session of the treatment contained the most items of information. The last session of the treatment is hidden in this figure, as this may have been the 6th session in one treatment and the 18th session in another treatment.

During the subsequent sessions, the number of topics discussed decreased. In the 16th session, for instance, an average of about 3 topics were discussed. Some topics showed a sharper decline than other topics. Pain management, for instance, was discussed relatively often in the initial sessions and relatively rarely in the later sessions. The same was true for back care instructions in daily activities, although the decrease was less sharp and more irregular. Information about home exercises seemed to be more stable throughout the sessions.

Trend Analysis

A trend analysis demonstrated these results more precisely and took the possible confounding variables into consideration. Overall, there was a negative trend related to instructions in therapy (Tab. 3), indicating that the frequency of instructions declined as the sessions progressed. There was, however, a positive coefficient related to both the first session and the last session, indicating that more than typical instructions were given during the first and last sessions. The only other variable that predicted the total number of instructions was the therapist's opinion on the importance of exercise in prevention, indicating that a greater belief in the importance of exercise was reflected in higher numbers of instructions to patients. All of these findings were statistically significant. With respect to the sources of variance within the analysis, 5.5% of the variability was related to differences among sessions, 34% was related to differences among therapists, and only 11% was related to differences among patients.

Based on the results shown in Table 3, an individual trend line was calculated for each patient (Fig. 9, upper panel). Patients given many instructions at the start of treatment showed a sharp decrease in the number of instructions received throughout the sessions. They ended up the same as patients with only a modest number of topics discussed in the first session. The trend lines of all patients with long treatments, no matter how they started, converge into the same small number of instructions. The same procedure for the 21 individual therapists resulted in a slightly different picture (Fig. 2, lower panel). At the level of the physical therapist, the lines are more entangled. One therapist began, on average, with little information and ended up with more information; most of the other therapists did the opposite.

[Figure 2 ILLUSTRATION OMITTED]

Table 3. Regression Coefficients and Standard Errors (in Parentheses) of Trend Analysis (1,151 Sessions, 132 Patients, and 21 Physical Therapists)

Mean                                     6.35      (.45)
Trend                                    -.24      (.05)
First session                             .86      (.28)
Last session                              .94      (.23)
Importance of exercise in prevention      .77      (.18)
Variance among physical therapists       3.17      (1.11)
Variance among patients                   .98      (.23)
Variance among sessions                  5.12      (.31)
Total variance                           9.27

Trends of Different Kinds of Instructions

The trend analysis, with the total number of instructions, is an introduction to the estimation of another series of trends for each of the 4 distinctive kinds of instructions mentioned in the theoretical framework: instructions on pain management, instructions on back care in activities of daily living, instructions on exercise, and recommendations for general fitness. The number of instructions differed greatly among the 4 areas (Tab. 2). To overcome this problem of scale, the dependent variables were standardized to z scores before modeling the independent variables. This procedure allowed us to better compare the results of the trend analyses summarized in Table 4.

Table 4. Regression Coefficients and Standard Errors (in Parentheses) of Trend Analyses for Pain Management, Back Care in Activities of Daily Living, Exercises, and Fitness (1,151 Sessions, 132 Patients, and 21 Physical Therapists)(a)

                                                     Back Care
                                        Pain        in Activities
                                     Management    of Daily Living

Mean(b)                              -.55 (.21)       -.85 (.26)
Trend                                -.06 (.01)       -.06 (.01)
First session                         .98 (.10)        .29 (.10)
Last session                                           .24 (.07)
Age                                   .01 (.002)
Treatment goal
  Pain reduction
  Posture improvement                                  .25 (.10)
  Muscle tone regulation                               .33 (.14)
Psychosocial                          .12 (.05)
Importance of exercise in prevent     .12 (.06)        .15 (.06)
Variance among physical therapist     .07 (.03)        .20 (.07)
Variance among patients(b)            .07 (.02)        .07 (.02)
Variance among sessions(b)            .47 (.03)        .54 (.03)
Total variance                        .61              .81

                                     Exercises           Fitness

Mean(b)                              -.03 (.27)         .30 (.27)
Trend                                                  -.02 (.009)
First session                        -.32 (.09)
Last session                          .21 (.08)
Age
Treatment goal
  Pain reduction                                       -.34 (.11)
  Posture improvement
  Muscle tone regulation
Psychosocial
Importance of exercise in
  prevention
Variance among physical therapist    .16 (.06)          .10 (.04)
Variance among patients(b)           .11 (.03)          .10 (.03)
Variance among sessions(b)           .64 (.03)          .76 (.03)
Total variance                       .91                .96

(a) Apart from the mean, only statistically significant (P <.05) coefficients are shown. The 4 dependent variables are transformed to z scores.

(b) For computational reasons, the trend variable is centered around the sixth session. The mean and variance components estimates, therefore, relate to the sixth session.

The trend coefficient for the number of pain management instructions was -.06 (Tab. 4). There was a downward trend throughout the sessions. The largest coefficient (.98) occurred in the first session for pain management, which reflects the fact that the topic of pain management was discussed very often in the first session. At the patient level, a small effect of age and a larger effect of "psychosocial factors" was estimated. The latter finding means that when patient complaints were related to psychosocial factors, according to the physical therapist, instructions on pain management were discussed more often than would occur with other patients.

The trend effect for instructions on back care in activities of daily living was of the same magnitude as the pain management coefficient (Tab. 4). There was a trend effect for both first and last sessions. At the patient level, we found an effect of 2 treatment goals. In those situations where the therapist was aiming at posture improvement or muscle tone regulation, back care instructions were more common. This finding also applies to patients whose exercises were seen as important in the prevention of relapse.

Table 4 also shows the results of the trend analysis with the number of exercise instructions as a dependent variable. The trend coefficient was not statistically significant. This finding means that the amount of home exercise information was spread evenly across the sessions, except for the beginning and the end of treatment where the first and last sessions had significant effects. Thus, from the first session to the second session, the mean number of exercise instructions increased (.32) and stayed at that level during subsequent sessions, until the last session where it increased (.21) again. The physical therapist's estimation of the importance of home exercises in preventing the recurrence of back pain was also associated with the amount of information given on home exercise.

Recommendations to promote general fitness were the last of the instructions analyzed (Tab. 4). The trend coefficient was small, but negative, and there was neither a first-session trend effect nor a last-session trend effect. In treatments explicitly targeting pain reduction, fewer recommendations about general fitness were registered.

With respect to the sources of variance within the analysis, most of the variability in the number of instructions was related to differences among sessions, a smaller part was related to differences among therapists, and an even smaller part was related to differences among patients. These findings applied to all 4 areas of instruction.

Discussion and Conclusions

We explored the content and sequence of instructions given by 21 physical therapists in private outpatient practices in the Netherlands. The treatment for back pain constitutes a great challenge tot secondary prevention. Consequently, we expected much information to deal with home exercises and back care instructions. Most of the topics discussed related to these areas. The actual number of instructions, however, was difficult to interpret. Earlier studies have focused on the number of informative statements obtained by analyzing audio-taped sessions. Sluijs[49,50] reported about 20 educational remarks and Gahimer and Domholdt[23,51] mentioned about 12 educational remarks, whereas the number of instructions averaged about 6 in our study. We should point to an important difference between the 2 studies mentioned and our study. In the previous studies,[23,49-51] the authors counted each occurrence, whereas we had an aggregate score for each session where each kind of instruction could count only once. Another difference between the previous studies and ours concerns the difference between observational studies and the use of registration forms. Comparison of these 2 methods invariably shows that registration forms underestimate the amount of information given.[52,53] The different method used in our study makes a comparison with other studies quite difficult.

From a consideration of the course of physical therapy, we arrived at 4 assumptions about the particular sequence (if the instructions in back care programs. Two of the assumptions were confirmed, and 2 assumptions were rejected.

Pain management instructions were given at the start of treatment and then decreased in number, as we anticipated. Evidently, pain prevails in the first phase of treatment of the majority of patients; therefore, pain management is apparently seen as vital in this step. The number of pain management instructions seems to follow the usual course of recovery.

As we expected, exercise instructions were introduced after the start of treatment and were spread evenly across the visits. In an earlier study,[49] a sharp decrease in exercise instructions across the therapy visits was found, with a maximum number of instructions in the second session and only a few instructions at the end. In our opinion, practicing therapists are nowadays probably guided more by patient education principles as far as exercise therapy is concerned.

Instructions about taking care of the back during daily activities followed a different course from our expectations. Back care instructions were not evenly distributed over the sessions. The trend was downward. In the last session, there were relatively more back care instructions, probably because the therapist wanted to recapitulate the main items of self-management. The fact that back care instructions were predominant in the first stage of treatment and received less attention in the second half of treatment (except for the last session) does not meet our recommendation to spread the information equally across the visits. It could be argued, however, that instructions for activities of daily living may be provided with more efficacy in the early stages of treatment, when motivation to change behavior may be high. Currently, no evidence-based recommendation can be given.

Physical therapists apparently can influence a patient's health status, not only by prescribing efficacious treatments for musculoskeletal dysfunctions, but also by making recommendations for regular exercise to promote overall fitness.[54-56] In contrast to our assumptions, the number of recommendations about general fitness decreased over the course of treatment, and information about physical fitness was not widely given when compared with the other areas.

Assumptions about a rational sequence of back care instructions were derived from the usual course of physical therapy. We do not pretend that our supposition about the sequence of instructions is perfect. It would seem that it is important to find out what the optimal sequence is in order to enable the physical therapist to plan the information properly and to increase the likelihood of adherence because a well-planned intervention ensures the best adherence. We recommend that future guidelines for optimal treatment of patients with back pain should contain evidence-based information about the optimal sequence of instructions.

Considering the information given by the physical therapists within the framework of guidelines,[21] we conclude that physical therapists comply with the recommendation to pay attention to patient education.[21] We found many differences among therapists, however, in the amount of information they provide. The largest difference was found for the instructions on taking care of the back in daily activities. Some therapists offered a lot of advice, whereas other therapists did not. This finding is in accordance with the findings of one of our earlier studies.[22] Evidently, physical therapists have considerable flexibility when instructing their patients. It is unclear what degree of flexibility is desirable. Guideline developments tend to restrict the caregiver's degree of freedom. Within each guideline, however, some degree of freedom appears to be necessary to tailor the treatment to the individual patient's needs and circumstances. We expect that finding the balance between uniformity and flexibility remains a point of careful consideration in health care and, thus, also in physical therapy.

Given the diversity of the instructions given to patients, we would like to think about the intervention as an individualized back care program, that is, a program tailored to each patient's needs and attuned to the patient's individual situation. We assumed that individualized back care programs would lead to differences among patients in the information given by the physical therapist. The magnitude of the variance components shows us that this assumption is not true. The results indicate that the number of instructions in back care programs are determined to a high degree by individual therapists and not by the patient. The differences among patients were rather small. In our opinion, this finding indicates that the information and instructions were not fully tailored to each patient's situation, which is one of the strategies for enhancing patients' adherence to treatment. To enhance adherence, some authors[57-59] recommend tailoring exercises and advice as much as possible, considering the patient's particular situation and routines. We believe that adherence is even better when these instructions are also integrated into each patient's daily activities to prevent adherence problems.[60] We recommend, therefore, that tailoring information and instructions to the individual patient should be considered by therapists, but we cannot present data to show that it would lead to better patient outcomes.

References

[1] Frymoyer JW. Back pain and sciatica. N Engl J Med. 1988;318: 291-300.

[2] Frymoyer JW, Cats-Baril WL. An overview of the incidences and costs of low back pain. Orthop Clin North Am. 1991;22:263-271.

[3] van Tulder M. Diagnostics and Treatment of Chronic Low Back Pain in Primary Care. Amsterdam, the Netherlands: Thesis Publishers; 1996.

[4] van der Velden J, de Bakker DH, Claessens AAMC, Schellevis FG. Morbidity in General Practice: Dutch National Survey of General Practice. Utrecht, the Netherlands: Netherlands Institute for Primary Health Care; 1992.

[5] Dekker J, van Baar ME, eds. Beleidsgericht Evaluatie: En Effectonderzoek Extramurale Fysiotherapie. Utrecht, the Netherlands: Nederlands Instituut Voor Onderzoek van de Gezondheidszorg; 1995.

[6] Lamberts H. In Her Huis van de Huisarts: Verslag van een Transitie-project. Lelystad, the Netherlands: Meditekst; 1991.

[7] Kerssens JJ, Groenewegen PP. Referrals to physiotherapy: the relation between the number of referrals, the indication for referral, and the inclination to refer. Soc Sci Med. 1990;30:797-804.

[8] Jette AM, Smith K, Haley SM, Davis KD. Physical therapy episodes of care tot patients with low back pain. Phys Ther. 1994;74:101-110.

[9] Wells P, Lessard E. Movement education and limitation of movement. In: Wall PD, Melzack R, eds. Textbook of Pain. 2nd ed. Edinburgh, Scotland: Churchill Livingstone; 1989:952-963.

[10] Koes BW, Bouter LM, Beokerman H, et al. Physiotherapy exercises and back pain: a blinded review. BMJ. 1991;302:1572-1576.

[11] Nachemson AL. Exercise, fitness, and back pain. In: Bouchard C, Shepard RJ, Stephens T, et al, eds. Exercise, Fitness, and Health: A Consensus of Current Knowledge. Champaign, Ill: Human Kinetics Inc; 1990.

[12] Yeh C, Gonyea M, Lemke J, Volpe M. Physical therapy: evaluation and treatment of chronic pain. In: Aronoff GM, ed. Evaluation and Treatment of Chronic Pain. Baltimore, Md: Urban & Schwarzanberg; 1985.

[13] Reilly K, Lovejoy B, Williams R, Roth H. Differences between a supervised and independent strength and conditioning program with chronic low back syndromes. J Occup Med. 1989;31:547-550.

[14] Sluijs EM, Knibbe JJ. Patient compliance with exercises: different theoretical approaches to short-term and long-term compliance. Patient Educ Couns. 1991; 17:191-204.

[15] Mazucca SA. Does patient education in chronic disease have therapeutic value? Journal of Chronic Diseases. 1982;35:521-529.

[16] Turk DC, Salovey P, Litt MD. Adherence: a cognitive-behavioral perspective. In: Gerber KE, Nehemkis AM, eds. Compliance: The Dilemma of the Chronically Ill. New York, NY: Springer Publishing Co; 1986:44-72.

[17] Green CA. What can patient health coordinators learn from ten years of compliance research? Patient Educ Courts. 1987;10:167-174.

[18] Lorig K. Patient Education: A Practical Approach. 2nd ed. Newbury Park, Calif: Sage Publications Inc; 1996.

[19] Sluijs EM, Kok GJ, van der Zee J. Correlates of exercise compliance in physical therapy. Phys Ther. 1993;73:771-781.

[20] Oostendorp RAB, Warns HWA, Hendriks HJM. Fysiotherapie en lagerugpijn: een nieuw paradigma. Nederlands Tydschrift Voor Fysiotherapie. 1997;107:102-110.

[21] Rothstein JM, Delitto A, Scalzitti DA. Understanding AHCPR Clinical Practice Guideline No. 14: Acute Low Back Problems in Adults. PT--Magazine of Physical Therapy. 1995;3 (9) :65-88.

[22] Sluijs EM, van der Zee J, Kok GJ. Differences between physical therapists in attention paid to patient education. Physical Theory & Practice. 1993;9:103-107.

[23] Gahimer JE, Domholdt E. Amount of patient education in physical therapy practice and perceived effects. Phys Ther. 1996;76:1089-1096.

[24] Kerssens JJ, Curfs EC. Extramurale Fysiotherapie [Outpatient Physical Therapy]. Utrecht, the Netherlands: Netherlands Institute of Primary Health Care; 1993.

[25] Verhaak PFM, van Busschbach JT. Patient education in general practice. Patient Educ Courts. 1988;11:119-129.

[26] Minnick A. Patient Teaching by Registered Nurses. Ann Arbor, Mich: UMI Research Press; 1982.

[27] Edelman B. Federal agency to draft low back pain guidelines. Orthopedics Today. 1992;12:1, 10.

[28] Bigos S, Bowyer O, Braen G, et al. Clinical Practice Guideline No. 14: Acute Low Back Problems in Adults. Rockville, Md: Agency for Health Care Policy and Research, Public Health Service, US Department of Health and Human Services; 1994. AHCPR publication 95-0642.

[29] Connolly J. AHCPR Guideline No. 14, "Acute Low Back Problems in Adults": a commentary. PT--Magazine of Physical Therapy. 1995;3(9): 89-97.

[30] DeRosa CP, Porterfield JA. A physical therapy model for the treatment of low back pain. Phys Ther. 1992;72:261-269.

[31] Philips HC, Grant L. The evolution of chronic back pain problems: a longitudinal study. Behav Res Ther. 1991;29:435-441.

[32] Hingstman L, Boon H. Regional dispersion of independent professionals in primary health care in The Netherlands. Soc Sci Med. 1989;28:121-127.

[33] Dekker J, van Baar ME, Curfs EC, Kerssens JJ. Diagnosis and treatment in physical therapy: an investigation of their relationship. Phys Ther. 1993;73:568-577.

[34] Valk RWA, Dekker J, van Baar ME. Physical therapy for patients with back pain: a description. Physiotherapy. 1995;81:345-354.

[35] van Triet EF, Dekker J, Kerssens JJ, Curfs EC. Reliability of the assessment of impairments and disabilities in survey research in the field of physical therapy. Int Disabil Stud. 1990;12:61-65.

[36] Hendriks EJM, Brandsma JW, Heerkens YF, et al. Intraobserver and interobserver reliability of assessments of impairments and disabilities. Phys Ther. 1997;77:1097-1106.

[37] Knibbe JJ, Knibbe NE, Elvers JWH, et al. Inventarisatie van Rugscholen in Nederland [Overview of Back Schools in the Netherlands]. Amersfoort, the Netherlands: Locomotion/SWSF; 1992.

[38] Goeken LNH, ed. Rugscholen in Nederland [Back Schools in the Netherlands]. Utrecht, the Netherlands: ISPO/LEMMA; 1995.

[39] Bryk AS, Raudenbusch SW. Hierarchical Linear Models: Applications and Data Management Methods. Newbury Park, Calif: Sage Publications Inc; 1992.

[40] Goldstein H. Multilevel Statistical Models. 2nd ed. New York, NY: Halsted Press; 1995.

[41] Yang M, Goldstein H. Multilevel models for longitudinal data. In: Engel U, Reinecke J, eds. Analysis of Change: Advanced Techniques in Panel Data Analysis. Hawthorne, NY: Walter de Gruyter Inc; 1996: 191-220.

[42] Paterson L, Goldstein H. New statistical methods for analysing social structures: an introduction to multilevel models. British Educational Research Journal. 1991;17:387-393.

[43] Draper N, Smith H. Applied Regression Analysis. 2nd ed. New York, NY: John Wiley & Sons Inc; 1981.

[44] Jones K, Moon G. Multilevel assessment of immunisation uptake as a performance measure in general practice. BMJ. 1991;303:28-31.

[45] Bensing JM, Kerssens JJ, van der Pasch M. Patient-directed gaze as a tool for discovering and handling psychosocial problems in general practice. Journal of Nonverbal Behavior. 1995;19:223-242.

[46] Kerssens JJ, van Yperen EM. Patient's evaluation of dietetic care: testing a cognitive-attitude approach. Patient Educ Couns. 1996;27: 217-226.

[47] Duncan C, Jones K, Moon G. Health-related behaviour in context: a multilevel modeling approach. Soc Sci Med. 1996;42:817-830.

[48] Rasbash J, Woodhouse G. MLN Command Reference. London, England: Institute of Education, University of London; 1995.

[49] Sluijs EM. Patient education in physical therapy: towards a planned approach. Physiotherapy. 1991;77:503-508.

[50] Sluijs EM. A checklist to assess patient education in physical therapy practice: development and reliability. Phys Ther. 1991;71:561-569.

[51] Gahimer JE. Prevalence and Effectiveness of Patient Education in Physical Therapy Practice. Bloomington, Ind: Indiana University; 1995. Dissertation.

[52] Sluijs EM. Patientenvoorlichting Door Fysiotherapeuten: Ontwikkeling van het Observatieprotocol [Patient Education by Physical Therapists: Development of Observation Protocol]. Utrecht, the Netherlands: Netherlands Institute of Primary Health Care; 1988.

[53] Rethans JJE, van Boven CPA. Simulated patients in general practice: a different look at the consultation. BMJ. 1987;294:809-812.

[54] Lyne PA. The professions allied to medicine: their potential contribution to health education. Physiotherapy. 1986;72:10-12.

[55] Bouchard C, Shepard RL, Stephens T, et al, eds. Exercise, Fitness, and Health: A Consensus of Current Knowledge. Champaign, Ill: Human Kinetics Inc; 1988.

[56] Lorish C, Francis K, Jensen G, Sluijs EM. Enhancing the health status of patients by increasing adherence to therapeutic and voluntary exercise. In: Physical Therapy Course Material: A Compendium of Conference Handouts. Alexandria, Va: American Physical Therapy Association; 1996:21-25.

[57] Fisher AC. Adherence to sports injury rehabilitation programmes. Sports Med. 1990;73:151-158.

[58] Meichenbaum D, Turk DC. Facilitating Treatment Adherence: A Practitioner's Guidebook. New York, NY: Plenum Press; 1987.

[59] Bartlett EE. Behavioural diagnosis: a practical approach to patient education. Patient Educ Couns. 1982;4:29-35.

[60] Sluijs EM, Kerssens JJ, van der Zee J. Adherence to physiotherapy. In: Myers L, Midence K, eds. Adherence to Medical Treatment. Reading, England: Harwood Academic Publishers; 1998.

JJ Kerssens, PhD, is Research Associate, Netherlands Institute of Primary Health Care, Utrecht, the Netherlands. Address all correspondence to Dr Kerssens at NIVEL, PO Box 1568, 3500 BN Utrecht, the Netherlands ([e-mail]).

EM Sluijs, PhD, is Research Associate, Netherlands Institute of Primary Health Care.

PFM Verhaak, PhD, is Senior Research Associate, Netherlands Institute of Primary Health Care.

HJJ Knibbe, PT, is Research Associate, Netherlands Institute of Primary Health Care.

IMJ Hermans is Research Assistant, Netherlands Institute of Primary Health Care.

This work was supported by Grant No. 002822970 from the PREVENTIEFONDS (Prevention Fund).

This article was submitted December 9, 1998, and was accepted November 4, 1998.