Individually tailored treatment targeting motor behavior, cognition, and disability: 2 experimental single-case studies of patients with recurrent and persistent musculoskeletal pain in primary health care.

Background and Purpose. This article introduces an individually tailored intervention targeting motor behavior, cognition, and disability in patients managed by physical therapists in primary health care. Effects on self-rated disability, pain intensity, and pain control are described. Subjects. Two women with recurrent or persistent disabling musculoskeletal pain were selected. Methods. Two experimental single-subject [A.sub.1]-B-C-[A.sub.2] studies with multiple baselines across situations were used. Principal outcome data were collected daily with patient-specific continuous measures for 3 weeks before intervention, continuously during intervention, and for 2 weeks during each of the 1-, 4-, 6-, and 12-month follow-up examinations. Results. Disability and pain intensity decreased, and pain control increased in both subjects. The results were maintained at the follow-up examinations. Discussion and Conclusion. Positive outcomes of the intervention were reported from 2 subjects with recurrent and persistent disabling pain. Procedures for systematic tailoring of treatment to behavioral goals and individual patient characteristics are available as a result of the successful application. The results need to be replicated in future clinical controlled group studies. [Asenlof P, Denison E, Lindberg P. Individually tailored treatment targeting motor behavior, cognition, and disability: 2 experimental single-case studies of patients with recurrent and persistent musculoskeletal pain in primary health care. Phys Ther. 2005;85:1061-1077.]

Key Words: Behavioral medicine, Multiple baseline design, Musculoskeletal pain, Primary care, Self-efficacy.

Control over pain, resumed activity in daily life, and maintenance of social relationships are important rehabilitation goals for patients with chronic pain. (1) Such comprehensive treatment goals indicate that the biomedical perspective of pain as a sensory response to tissue damage (2) is too limited. Pain has debilitating consequences that are not only physical. Medication and physical rehabilitation, therefore, are not sufficient modes of treatment.

About 40 years ago, the gate-control theory (3) provided the basis for our understanding of how pain is processed and for development of medical and psychological treatment targeting pain control. Today, it is suggested that the pain experience is a product of the synaptic structure of a neural network in the brain--"the body-self neuromatrix" that "integrates multiple inputs to produce the output pattern that evokes pain." (4(ps121)) The neuromatrix is genetically determined but is continuously influenced and changed by sensory input, including psychological stressors, which also involve neurohormonal mechanisms and the immune system. (4) An interesting assumption for pain management is that multimodal treatment can modify the output of the neuromatrix (2) and thus affect the pain experience.

In a behavioral learning theory perspective, a person's responses to pain are modified by both internal and external factors. For example, pain behaviors can be conditioned to cues in the environmental context, to emotions, or to cognitions (eg, beliefs, expectations, values, interpretations) and are reinforced by their positive and negative consequences. (5) Consequently, interventions based on such behavioral principles usually emphasize the influence of the context in controlling pain behaviors and individuals' self-regulation of behavior. (6)

Turk and colleagues (7) extended the original behavioral model of pain by including cognitive factors, and recently Sharps called attention to a modified cognitivebehavioral model that more directly targets patients' thoughts and interpretations. Treatment would still acknowledge behavioral components, but also target individuals' appraisal of pain, such as their expectations and interpretations of painful events, more directly. The neurobiological, behavioral, and cognitive-behavioral perspectives of the pain experience rather complement than contradict each other, and new pain management interventions presumably benefit from integration.

There is strong evidence that physical modalities, often included in physical therapy interventions, do not contribute to durable pain control, (9-11) whereas physical activity and exercise promote activity and return to work in patients with chronic spinal pain. (12,13) Psychological and psychosocial factors influence the course of development from subacute to chronic pain and play an important role for adjustment to chronic pain. (14) Recent reviews and meta-analyses of randomized controlled trials dealing with chronic pain provide evidence of the effectiveness of cognitive behavior therapy (CBT) and behavior therapy (BT) in changing pain perception, functional disability, and cognitive coping and appraisal, as well as evidence of the reductions of pain behaviors. (15-17) When different experimental conditions are compared, CBT and BT are superior to waiting-list control conditions, but not to other active interventions such as relaxation training and exercise. (16,17) Physical and functional restoration programs that include cognitivebehavioral principles can reduce the number of sicklisted days for patients with subacute and chronic spinal pain. (18) Combined physical therapy and BT/CBT programs are sparse in the literature, (19-23) and the different perspectives could be further integrated to provide more potent intervention for behavior change, self-management, and maintenance of adaptive strategies despite the presence of pain.

Many patients are not responsive to available active treatment methods, (24) which may be due to the fact that they do not receive treatment that specifically addresses their assets and needs (ie, treatment is not individually tailored). (25) Psychological treatment tailored to subgroup characteristics of psychological and psychosocial disability has recently gained support for patients with temporomandibular disorders (26) and early rheumatoid arthritis. (27) Trials involving physical therapy management aimed at patients with benign musculoskeletal pain (MSP) have hitherto not included interventions that are systematically and individually tailored to motor skills in combination with evidence-based psychological and psychosocial risk factors.

We developed a new intervention for people with recurrent or persistent pain managed by physical therapists in primary health care (PHC) based on: (1) a theoretical perspective of behavior change derived from Social Cognitive Theory (SCT), (28,29) (2) empirical support for the relationship between psychological and psychosocial factors for the onset of and adjustment to chronic pain and pain-related disability, (30-36) (3) integration of CBT/BT principles and physical and functional restoration programs, and (4) the need of individually tailored pain management interventions aimed at systematically identified risk factors and motor behaviors.

The clinical purposes of the intervention are attainment of individualized behavioral goals and self-management of pain. Targets for the intervention are physical performance, cognition, and motor behavior related to important, frequent, and distressing activities or situations of everyday life. The use of systematic, multiaxial assessments and functional behavioral analyses (FBA) (37,38) that include specific contextual, psychological, and physical factors controlling pain behaviors and perceived disability renders application of unique combinations of components in the treatment in each individual. (39,40) The treatment program is generally outlined in the "Method" section, whereas a more specific description of the strategies for the individual tailoring is provided in the Appendix.

In the Swedish public health care system, the PHC takes the main societal responsibility for provision of care to patients with MSP. General practitioners have the medical responsibility in cooperation with physical therapists, occupational therapists, and medical social workers, who provide medical services and interventions. Clinical psychologists and multidisciplinary teams are seldom available in PHCI and patients with persistent and disabling pain are usually offered treatment on an outpatient basis by physical therapists only. The patients in PHC differ from pain clinic populations where the majority of existing controlled studies are carried out. (41) Patients managed in PHC settings display lower levels of pain intensity, disability, and psychological stress, (41,42) and the comprehensive goal of treatment is either to prevent recurrent pain from becoming chronic and disabling or to provide proper adjustment and self-management for those with chronic pain. Interventions addressing psychological and psychosocial factors are important but are not routinely available. (43)

We chose the experimental single-case design (44) to study the new intervention and its effects because it is especially useful for the understanding of the impact of an intervention on an individual. (45) The key characteristics of the design (ie, continuous assessment, baseline assessment, endeavors for stability of performance, and the use of different phases) provide rigorous control and enhance internal validity, which permit causal inferences to be drawn in each individual case. (44) Besides, details about a new intervention that may go unnoticed in group studies may be revealed by the experimental single-case design, (46) and it makes refinements of the intervention possible before it is applied in large-scale group designs.

The purpose of this study was to describe and evaluate a recently developed individually tailored treatment program for patients with recurrent or persistent MSP. More specific aims were: (1) to study and describe the effects of the intervention on self-reported disability, pain control, and pain intensity in 2 experimental single-case studies and (2) to illustrate how the intervention was individually tailored to 2 patients with persistent MSP managed by physical therapists in PHC.

Method

Design

Two single-subject [A.sub.1]-B-C-[A.sub.2] designs with multiple baselines across situations were used. (47) Each baseline represented an important activity of everyday life, specific to a certain situation. The subjects were not exposed to any part of the treatment program during the initial [A.sub.1] phase. Pretreatment baseline scores were collected during this phase. During the B phase, the initial steps of the intervention were introduced (ie, behavioral goal identification, self-monitoring, functional analyses, and acquisition of basic skills) (Tab. 1). The C phase covered application of skills in situations of everyday life and was initiated at different time points for each baseline. Finally, treatment was withdrawn in the [A.sub.2] phase except for planned booster sessions. Valid conclusions from a multiple baseline design can be drawn if changes occur when the specific behavior is targeted (47) (ie, during the C phase in the present design).

Setting and Participants

The study was conducted at a physical therapy department within a typical PHC setting in Sweden. The particular PHC department employed general practitioners, nurses, midwives, physical therapists, and administrative staff. Neither multidisciplinary teams nor clinical psychologists were available for the regular practice. The intervention was provided by an experienced PHC physical therapist with postgraduate education in behavioral medicine (PA). Supervision was provided by a psychologist (PL).

In using single-subject designs, researchers strive to include typical clinical cases. (47) Consequently, subjects were recruited among ordinary patients seeking care at the physical therapy department. Criteria for inclusion were stated in advance. Participants should be between 18 and 65 years of age, have a duration of MSP of longer than 4 weeks, and experience activity limitation in daily life due to pain. Specifically, we wanted to include 1 subject with chronic disabling pain and 1 subject with lower levels of disability but experiencing recurrences and, therefore, at risk. Patients with osteoporosis, fibromyalgia, signs of trauma, systemic, infectious, or malignant diseases were not considered for the study.

Participant 1. The first participant (P1) was a 49-year-old woman with a history of low back pain and unilateral sciatic pain in the left leg and foot due to a suspected disk herniation 3 years previously. The pain had gradually spread to the upper extremities. An orthopedic surgeon did not find any indication for surgery, and a conservative treatment plan with sick leave and physical therapy in a PHC setting was decided on. The physical therapy consisted of 36 therapist-led sessions, including a regimen of rest, assisted movements, spinal extension exercises, acupuncture, transcutaneous electrical nerve stimulation, massage, hot packs, manual traction, cervical nerve mobilization, stretching, and physical exercise. The sciatic pain decreased over the treatment period, but 6 months after discharge, she reported a worsening of the diffuse and generalized pain. After 1 year of sick leave from her employment as an assistant nurse, she was transferred from night duty to day shifts. Initially, she worked half-time, but then she resumed sick leave involving considerable inactivity, dissatisfaction, and an incipient social withdrawal. Medication did not help her to sufficiently control pain any longer, which was the main reason for her general practitioner's second referral for physical therapy.

Participant 2. The second participant (P2) was a 49-year-old woman. Four years prior to this study, she slipped and fell. She claimed MSP as a consequence of damaging her left arm and shoulder. She recovered gradually but reported recurrent shoulder pain several times in connection with increased demands of arm and shoulder functions at work (employed as a personal assistant to people with multiple handicaps) or when engaged in her hobby (pony-harness racing). Six months before inclusion into this study, she reported increased pain in the low back, neck, and left shoulder. An x-ray examination did not show any signs of arthritis in the shoulder joints. Consequently, her general practitioner referred her for physical therapy. She worked full-time by the time of inclusion into this study.

Intervention

Social Cognitive Theory (29) and the cognitive-behavioral perspective (48) were used as organizing strategies for the intervention. The clinical purpose of the treatment program was to attain individual behavioral goals by acquisition of physical, cognitive, and behavioral skills related to important, frequent activities of everyday life and situations where patients reported hindrance due to pain. Our assumption for the program was that an individual who manages to control pain and perform one specific, important activity in one cumbersome situation gradually will control more and more activities and situations despite the presence of pain. Physical activity and motor behaviors related to the specified behavioral goals were the main targets for the intervention. However, related cognitions (eg, expectancies, fears) also were addressed, as were contextual factors and the organization of tasks and activities.

The program comprised 7 universal components or phases (Tab. 1), but the tactics within each phase differed according to each individual's goals, assets, and needs. Treatment was scheduled over a 4- to 6-month period, and the patients met the therapist in 45- to 60-minute sessions, initially once a week and later every second or third week. A central feature of the program was that the intervention was managed by the patients themselves according to homework assignments. The therapist-led sessions were used primarily to progress and tailor the intervention through its phases. The therapist introduced and monitored homework assignments, reinforced patients' progress, and helped to solve problems. Each patient had a personal file for the documentation (ie, diaries, exercises, educational material, and personal maintenance and relapse prevention and management programs).

Initially, the purpose of the intervention was explained and discussed thoroughly (ie, the emphasis on performance of important activities in everyday life, rather than on elimination of the pain). The requirement of active patient participation that included adherence to homework assignments also was emphasized.

The 7 components of the program were:

1. Behavioral goal identification and assessment. The first step aimed to create a priority list of each patient's behavioral goals for the coming treatment. The patients were asked to list important and specific activities and situations that they did not manage as well as they wanted due to the pain and that they expected to be affected by the treatment. The goals then were discussed by the patient and the physical therapist with reference to: (1) importance (ie, the patient's ranking or priority of the goals), (2) frequency (ie, how often they occur in the patient's daily life), and (3) magnitude of efficacy expectations (ie, the ordering of the goals by the patient's perception of level of difficulty). (28) As a result of these 3 rankings, a single behavioral goal was selected as the first target for analysis and treatment. This goal had to be highly ranked by the patient and had to occur frequently in daily life. Furthermore, the patient should have ranked the goal as one of the most easy to attain.

During this phase, further information also was obtained by interviewing the patient about the pain history and relevant medical, physical, behavioral, and psychosocial information, that is, information related to reported disability and prioritized goals. Finally, an individualized physical examination was done.

2. Self-monitoring. A diary was adapted, or tailored, to monitor activity performance related to the initially selected behavioral goal. Examples of factors included in the self-monitoring were situation-specific conditions, expectancies (eg, self-efficacy), fears, negative thoughts, motor behaviors, evaluation on actual performance, and immediate consequences of behavior. Each patient was asked to monitor and record activity and related cognitions and behaviors in specific everyday-life situations.

3. Individual functional behavioral analysis. Data that were collected in the initial phases of the program provided information for the FBA. The functional analysis is the identification of important, controllable, causal functional relationships applicable to specified behaviors for an individual. (38) Thus, according to the S-I-R-C format, antecedents or situations (S), individual capabilities (I), behavioral responses (R), and short- and long-term consequences (C) (49) related to the behavioral goal were identified. Preliminary hypotheses of causal relationships among these variables then were introduced and discussed with the patient. The initial FBA was focused on the selected initial goal, and the variables included must have a potential for change. Each hypothesis then resulted in further specifications of the treatment goal and justified the treatment tactics. The patient's progress was continuously evaluated in relation to the hypothesized relationships between the functional variables and the goals. This feedback was supposed to contribute to the patient's gain of knowledge and sense of control over the situation.

The FBA is preliminary and would be open to modification during treatment according to changes in the hypothesized relationships and to the different treatment goals. In the end, attainment of initially prioritized goals would determine when to discharge the patient, which makes the goal-setting procedure crucial. The patients are supposed to fully take over the responsibility for additional goal attainment and management of relapses.

4. Basic skills acquisition. This component refers to acquisition of skills deemed necessary for goal attainment. Skills were always related to the behavioral goal and specified by the previous FBA. Voluntary activation of muscles, coordination of movement patterns, strength (force-generating capacity of muscles), endurance, and mobility are examples of possible motor skills to acquire. Cognitive skills (eg, recognition of negative interpretations and cognitions) also were practiced. Self-efficacy related to behavior in specified activities and situations was reinforced. Subsequently, organization of the surrounding environment was discussed with each patient, as was organization of events with respect to fluctuations in pain and realistic efforts needed to manage the activities. General physical activity was encouraged during this phase. Finally, homework assignments were agreed on, and the training and completion of tasks were recorded by the patient in a diary.

5. Applied skills acquisition. In the fifth phase of the treatment program, the acquired basic skills are merged to create more complex behaviors (ie, motor behaviors, including necessary cognitive skills to manage the target activity). Initially, behaviors were practiced in contrived environments at the clinic (ie, the therapist created a situation much similar to the patient's everyday-life situation). The performance of the activity then was practiced with guidance from the therapist. Strong emphasis was put on the patient's ability to combine basic skills to create an appropriate motor behavior. Homework assignments including applied skills were agreed on, and the training and completion of tasks were recorded by the patient in a diary.

6. Generalization. When the initially selected goal was attained, the next goals on the priority list were targeted, and the procedure was repeated (ie, the FBA was complemented and additional basic skills were practiced and subsequently applied in contrived situations as well as in situations of everyday life).

7. Maintenance and relapse prevention. Patients were prepared for relapses that are commonly present in the process of behavior change. (50) High-risk situations were identified, and problem-solving strategies were discussed and expanded to equip the patient with an arsenal of strategies to use in different situations. This preparation resulted in a personal, written relapse prevention and management document. Finally, strategies for maintaining performance in relation to attained goals were discussed and reinforced.

Measures

Daily ratings of self-reported disability and pain. The principal outcome data were collected daily with continuous patient-specific measures for 3 weeks before the introduction of treatment (phase [A.sub.1]), throughout the intervention (phases B and C), and for 2 weeks during each of the follow-up examinations at 1, 4, 6, and 12 months after treatment (phase [A.sub.2]).

Self-reported disability related to each patient's prioritized goals for treatment (51) was scored on 11-point (0-10) numerical rating scales (NRS), where low scores indicated low disability. One scale was adopted for each goal (for more details about the goals, see the "Individual Tailoring of the Treatment Program" section of the Appendix). A pilot study including a Swedish PHC sample of 37 patients waiting for physical therapy consultation for persistent musculoskeletal pain showed that data on daily self-reported disability related to prioritized behavioral goals were stable across a 3-week period before treatment (Dagson and Peterson, unpublished master's thesis).

Pain control was scored on a 7-point NRS extracted from the Swedish version of the Coping Strategies Questionnaire (CSQ), (52) with low scores indicating low pain control. The Swedish version of the CSQ has shown high internal consistency (Cronbacb alpha = .70 - .80). The pain control item usually is analyzed separately from the other subscales included in the questionnaire. The moderate test-retest reliability (r = .63) of data for the single item of pain control has been found by other authors. (52)

Pain intensity was scored on an 11-point (0-10) NRS, with low scores indicating low pain intensity. The validity of NRS data for pain intensity ratings is well documented, and findings include positive, significant correlations with other measures of pain intensity. Numerical rating scales also have demonstrated sensitivity to treatments that are expected to produce pain intensity changes. (53) Data from the pilot study mentioned previously did not show any reactivity in the pain intensity measure using daily self-reports 3 weeks before the start of physical therapy (Dagson and Peterson, unpublished master's thesis).

Standardized measure of pain-related disability. Pain-related disability was measured on 6 occasions: before treatment, immediately after treatment, and at the 1-, 4-, 6-, and 12-month follow-up examinations. For this purpose, the Pain Disability Index (PDI) (54,55) was used. The PDI was chosen due to its behavioral medicine origin, its explicit purpose of collecting data that complement behavioral assessments and assessments of physical impairments, and its sound psychometric properties. Additionally, it complemented the main outcome measure of pain-related disability because it is standardized and generic rather than patient-specific and condition-specific (eg, to low back pain, neck pain, shoulder pain). Thus, the PDI represents a brief, generic 7-item inventory, with validity and reliability, for the measurement of pain-related disability or pain interference in the performance of activities related to the following areas: family and home responsibilities, recreation, social activity, occupation and education, sexual behavior, self-care, and life-support activity. (54,55) The degree of interference was rated on an 11-point NRS, ranging from 0 ("no interference") to 10 ("total interference"). A general disability score ranging from 0 to 70 then was calculated by summing scores of the 7 items. (55) A Swedish version of the PDI (33) was used in this study. The internal consistency was found to be high in a Swedish PHC sample of patients with persistent pain (Cronbach alpha = .85). (33) The item total correlation coefficients ranged from .41 to .71.

Clinical assessment. The data for the FBA were collected using a clinical version of a patient-specific tool, the Patient Goal Priority Questionnaire (PGPQ). (51) The PGPQ originally was designed to collect data concerning patients' priorities of behavioral goals. The patients listed activities that they: (1) were unable to perform or had difficulty performing due to pain and (2) expected to improve as a result of the treatment. They also ranked the relative importance of the activities. Subsequently, we investigated patients' perceptions of: (1) current level of behavioral performance, (2) frequency of behavioral performance, (3) satisfaction with current level of behavioral performance, (4) self-efficacy or confidence in behavioral performance, (5) fear of behavioral performance, (6) expectations of future level of behavioral performance, and (7) readiness to adopt new behaviors to achieve prioritized goals. This information was collected for each of the listed goals. Observations of motor behavior and activity performance also were done.

Standardized physical tests were included as a complement to the regular physical assessment. Active range of motion (ROM) in the affected joints was measured. Active ROM of the shoulder (P2) was measured with a goniometer with the patient in a supine position. Active ROM of the neck (P1, P2) was measured with the patient in a sitting position without back support using a Myrin meter * (ie, a compass with an inclination needle). Thoracic and lumbar active ROM (P1) was measured with a kyphometer as described by Ohlen. (56) Isometric muscle endurance tests of the neck extensors and flexors and the back extensors were performed as described by Ljungquist et al. (57) The total number of sit-ups, performed with supported feet, knees bent to 90 degrees, and hands striving to reach the bases of the patellae, was counted, as was the total numbers of push-ups performed with the knees and lower part of the legs supported by the floor. (58) Functional lifting ability was assessed with a cervical lifting test, the Progressive Isoinertial Lifting Evaluation. (59) The patients were asked to lift a plastic box containing weights from waist to shoulder (0.72-1.34 m). The initial weight was 2 kg. Four lifts were performed during 20 seconds; the weight then was increased by 2 kg every fourth lift. The weight managed during the final trial was used as the test result. Data from the physical tests are presented in Table 2. Additional data from the clinical assessment will not be reported separately but are integrated in the case descriptions of the individual tailoring of the treatment program (Appendix).

Data Management and Data Analyses

Continuous data for each baseline measurement were initially checked for serial dependency by calculation of autocorrelation coefficients. (60) Due to the large amount of data points (one per day), weekly average values were calculated and displayed graphically for visual inspection of means, trends, levels, and variability between phases. (44) To illustrate changes in variability between the phases, box plots for each phase were constructed. To conclude whether the interventions were efficacious and clinically significant, criteria for improvement were formulated in advance. The treatment goals for self-reported disability were considered attained with scores between 0 and 2 on the NRS. Percentages of nonoverlapping disability scores in the [A.sub.2] phase compared with baseline scores in the [A.sub.1] phase also were calculated. Higher percentage scores are proposed to reflect a more efficacious intervention. (60) Finally, the 2-standard deviation band method (61) was used to decide the significance of the effects. This is a semistatistical procedure based on the mean value and standard deviation of the baseline scores. A band was drawn at the distance of 2 standard deviations from the mean value of the baseline scores (phase [A.sub.1]). Data in the [A.sub.2] phase then were checked. It is proposed that, if at least 2 successive data points fall outside the 2-standard deviation band, a statistically significant change has occurred. This assumption is based on the likelihood of such change occurring considering the baseline score is less than 5 out of 100. (61) Data from the generic measure of pain-related disability (PDI) were presented as raw score changes. Based on previous findings in a Swedish PHC sample (Denison et al, unpublished data, 2004), a 4-point change was considered to be a clinically significant effect. Mean values for patients' daily ratings of pain control and pain intensity were calculated for 3 weeks before treatment, for 2 weeks after treatment, and for 2 weeks during each of the 1-, 4-, 6-, and 12-month follow-up examinations. For pain intensity, we adhered to the criteria presented by Farrar and colleagues (62) suggesting that a 2-point change on the NRS, or a 30% decrease, is a clinically meaningful pain reduction.

Results

Patient-specific data illustrating the treatment process and the intervention effects are presented. Because the treatment program has not been previously described, the systematic strategies for the individual analysis and the tailoring of the intervention are detailed for each of the 2 cases in the Appendix.

Self-reported Disability Related to Prioritized Treatment Goals (P1)

Visual inspection of raw data suggested that there were no trends in data during the pretreatment baseline period. This assumption was supported by the calculation of serial dependency, which showed no significant autocorrelation coefficient in any of the 4 baseline measurements. The principal outcomes are illustrated in Figure 1A, which shows P1's self-rated disability in the 4 most important everyday-life situations. Weekly average values are displayed for the pretreatment baseline measurements (phase [A.sub.1]), for the intervention phases (B, C), and for 2 weeks at each follow-up examination (phase [A.sub.2]). The variability in reported disability during the baseline period ranged from 6 to 1 (low scores indicate low disability) (Fig. 2A). Reductions of disability were gradual during treatment (Fig. 1A), but shifts in levels are obvious in the comparison of data from phases [A.sub.1], B, and [A.sub.2]. Furthermore, the variability in reported disability gradually decreased, and stable data patterns were evident in the [A.sub.2] phase (Fig. 2A). The percentages of nonoverlapping data in the [A.sub.2] phase compared with the [A.sub.1] phase were 100% (household chores by the sink), 100% (driving), 100% (make the beds), and 83% (buy food for the household). Statistical evaluation of changes was performed using the 2-standard deviation band method, (61) and the results suggested that differences of mean levels between the [A.sub.1] and [A.sub.2] phases did not occur by chance. The disability scores were zero or close to zero in all activities after treatment and at the follow-up examinations, confirming the clinical significance of the results. This finding also was validated by the generic measure of disability. Table 3 displays raw scores and percentage changes in pain-related disability.

[FIGURE 1 OMITTED]

Self-reported Disability Related to Prioritized Goals (P2)

P2's reports of disability in her 3 most important everyday-life situations are illustrated in Figure 1B. No trends in raw data were seen during the [A.sub.1] phase, which were confirmed by nonsignificant autocorrelation coefficients. The variability in reported disability during the baseline period ranged from 2 to 4 (low scores indicate low disability) (Fig. 2B). Shifts in levels were present between the [A.sub.1], B, and [A.sub.2] phases. The variability in reported disability decreased during treatment, and stable data patterns were seen for all activities in the [A.sub.2] phase (Fig. 2B). The percentages of nonoverlapping data in the [A.sub.2] phase compared with the [A.sub.1] phase were 100% (driving), 100% (pony-harness racing), and 100% (handling of wheelchairs). Calculations according to the 2-standard deviation band method (61) suggested that the changes in mean levels between the [A.sub.1] and [A.sub.2] phases were statistically significant for all activities. The post-treatment disability scores ranged from 0 to 1, pointing to clinically significant changes that were validated in the generic measure of pain-related disability (Tab. 3).

[FIGURE 2 OMITTED]

Pain Control and Pain Intensity (P1, P2)

The patients' perception of pain control gradually increased during the course of the intervention. Furthermore, self-reported pain intensity decreased in both patients (Tab. 3). Pain-intensity reductions of 65% (P1) and 97% (P2) indicate that these changes were clinically significant.

Discussion

The primary aim of this article was to introduce, describe, and evaluate an individually tailored treatment program for use by physical therapists in PHC. Behavioral goal assessment and systematic individualization of the treatment were especially emphasized. The main clinical philosophy behind the program is to guide patients toward resumed activity. Through acquisition of necessary physical, cognitive, and behavioral skills that were related to important and frequent instances of everyday life, both patients in this study attained preset individual behavioral goals.

Components of the Treatment Program and the Individual Tailoring

In this article, a detailed treatment description is provided because of its novelty and because of the lack of comprehensive treatment manuals in the present literature. (63) Such descriptions are necessary to develop, implement, and evaluate individually tailored treatment programs. The first step in the program was to ensure the clinical relevance of the treatment goals for each individual and provide a basis for tailored analyses and treatment strategies. The goals on each patient's priority list were ranked according to importance, frequency, and magnitude of self-efficacy expectations. A mutual agreement between the patient and the physical therapist, considering the ranking of the goals according to these 3 hierarchies, decided which goal to target initially in treatment.

The treatment goals in our program are behavioral because these are directly controlled by the individual and more strongly relate to efforts, endurance, and concentration of goal attainment than pathophysiological and physical impairment goals. (64) The goal assessment procedure is central because, if it is done systematically, it is a potential intervention component by itself. Goals can instill motivation for goal-directed performance provided that they are important and not in conflict with other goals. (65) Consequently, we asked for the individuals' most important treatment goals, and their rankings were acknowledged in the treatment planning. To illustrate, P1 was on sick leave, but prioritized goals related to her responsibilities at home. She hesitated about the goal of returning to work because she considered herself unable to manage both work and household responsibilities, the latter being the most important to her. Consequently, the home- and family-related goals were targeted first. A gradual acquisition of skills and self-efficacy for performance of these activities could subsequently be generalized to social and work-related activities.

The rationale for ordering of goals by their difficulty level, or magnitude of self-efficacy expectations, was that high self-efficacy more likely promotes effective problem-solving strategies and engagement in behavior than low self-efficacy. (66) Because self-efficacy expectations are contextually and behaviorally specific but also possible to generalize, we adhered to Bandura's (28) proposal and analyzed the magnitude level of self-efficacy for each goal on the patients' priority list in relation to the other goals and the strength of the efficacy expectations for each goal separately. The magnitude level was used to select the first target in treatment, and the ratings of self-efficacy strength were used for the specific FBA. For example, P2's highest priority was to manage her job, although she reported a lower magnitude level for pushing wheelchairs than for driving her car. Consequently, to drive her car became the first target, and strengthened by the success in this task, a more difficult goal could be approached.

The FBA was built upon information collected in the prospective, systematic self-monitoring of specified behaviors related to the target goal. (38,67) Functional behavioral analyses differ from diagnostic approaches (68) and are used to specify variables and relationships between variables that account for variance in behavioral goals and problems. The intervention in this study was designed to target the functions between individual factors and the target goal. The FBAs were consequently used to individually tailor the treatment. Thus, treatment plans were not based on the patients' diagnoses. Pain-related disability is controlled by the individual patient's behavioral responses (including motor and cognitive responses) in interaction with the context. (68,69) The content within each phase of the treatment program, therefore, differs across patients and hypothetically will do so even if patients have similar diagnoses.

The systematic way of collecting data, the analysis, and the coherent treatment plan motivate patients to participate and to take responsibility for their own rehabilitation. Furthermore, the reactive effects of self-monitoring are well-known, (67) making it a potential therapeutic agent in itself (ie, patients are provided with a sense of control over their situation, which, in turn, contributed to problem solving and acquisition of necessary skills to attain their goals). P1 used her self-monitoring diary throughout the applied acquisition phase because it reminded her to apply necessary skills and made it possible for her to check her progress. This example illustrates how the self-monitoring and the introduction of the FBA to the patient are double-acting tools for analyses and treatment.

A central feature of the program is that acquired basic skills are applied in patients' everyday environment. Patients are not expected to transfer acquired basic skills (eg, balance, coordination of movement, relaxation, constructive thinking, optimization of the use of technical aids) to situations of everyday life on their own. Instead, skills are merged and practiced in contrived as well as natural environments as a part of the program. A look at the documented changes in self-rated disability in relation to the different treatment components in our study reveals that changes occurred when skills were applied in situations of everyday life, especially in the first targeted goal. This component, therefore, seems crucial for goal attainment.

Generalization of behavior change includes generalization across time, settings, and behaviors. (70) The generalization across prioritized goals or situations is built into the program and subsequently is followed up in booster sessions. More specifically, we used a patient's perception of mastery of a specific activity in a specific situation to boost the individual for further success in activity performance. (28) Thus, a successful initial goal attainment facilitates generalization and amplifies skills needed to reach the other goals on the priority list. The initial goal took longer to attain than the other goals, and more supervision from the therapist was needed. Possibly, many patients can generalize skills on their own, but individual differences should be acknowledged in the generalization phase. In our study, P1 did most of her generalization tasks guided by the therapist during the therapist-led sessions. P2 had higher initial self-efficacy expectations and therefore was encouraged to generalize skills to additional tasks and situations mainly on her own after the end of the therapist-led sessions.

The maintenance and relapse prevention programs were based on the patients' own prediction of risk situations. Adaptive situational strategies were reinforced, possibly contributing to the patients' self-management of pain in the future. The intervention effects in this study were still maintained 1 year after the end of the therapist-led sessions, indicating that this strategy was useful.

Intervention Effects and Internal Validity

Clinically significant changes in self-rated disability were seen in both patients during the course of treatment, but could these changes be causally related to the intervention? The patients' disability levels were moderate during the baseline period, but they were comparable to levels in a representative Swedish PHC sample of patients with recurrent and persistent MSP (51) that differs from strictly selected samples treated at pain clinics. Changes in levels for weekly mean values of self-rated disability were seen during treatment. These changes hypothetically could be referred to a pre-existing downward trend and not to the intervention. (60) Another explanation for the effects could be maturation (47) (ie, the patients were under spontaneous physical recovery, which contributed to decreased disability and pain intensity and to recovery in the long-term). However, both patients' histories of recurrent pain and pain duration for several years and earlier treatment without sustained effects minimize but do not rule out such explanations. Furthermore, no threats in terms of changes in history (eg, changes in family conditions or work conditions, additional treatment or diseases) during treatment were reported. The absence of any trends in pretreatment baseline data, controlled for by analyzing the autocorrelation coefficients, also indicates that no such trends existed for either patient. The multiple baseline design across situations permits causal inferences, provided that changes occur only when the targeted intervention is introduced. (47) Our results brought some ambiguities into the interpretation because changes occurred in some baseline situations when the patients were exposed to a preceding situation. This may be due to the fact that the baseline situations were not completely independent of each other and that patients spontaneously generalized acquired skills to new situations. This is problematic from a methodological point of view and reduces the possibility to draw firm causal inferences. However, from a clinical perspective, it is valuable if acquired skills and exposure to one specific situation are generalized to other situations of everyday life by the patient. The prerequisites for such generalizations should be a target for further study.

The changes in level from moderate disability to no reported disability were replicated in both patients and in all specific situations. The patients' earlier resistance to treatment increased the likelihood of the intervention being responsible for the improvements. Thus, the effects would be related to the increased control of pain and to the changes in cognitions, behaviors, and context that were reported by both patients in their self-monitoring diaries. We also suggest that the changes were reflected in the data patterns. The variability in disability levels decreased during the course of treatment and reached stable, low levels that were maintained at all follow-up examinations in the [A.sub.2] phase. These data patterns also acknowledge the rationale behind the treatment program that an individual who manages to control and perform one specific, important activity in one cumbersome situation gradually will manage to control more and more activities and situations. The results were validated by the generic measure of pain-related disability, which showed raw score changes between 90% (P2) and 95% (P1). Furthermore, decreases in average pain intensity of 65% (P1) and 97% (P2) were reported. Thus, our preset criteria for an effective intervention were fulfilled in all available outcome measures.

Considering the novelty and the patient-specific characteristics of the principal outcome measures, reliability and validity have not been extensively investigated. Nevertheless, the pilot study did not show any reactivity in the continuous measure, and the actual changes in the outcome measures during the course of treatment in this study indicate their sensitivity to change. The complement with generic measures whose results were coherent with the continuous measures also validated the findings. However, the clinically significant findings for the 2 individuals in this study need to be replicated in controlled group study designs before it is possible to discuss the generality of the outcomes. (61)

Conclusions

A systematic, individualized approach to physical therapy management developed from a behavioral medicine perspective was described in this article. Individualized procedures for goal identification, prospective self-monitoring, tailored behavioral functional analysis, and applied activities training were especially emphasized. Positive outcomes of the intervention were reported from 2 patients with recurrent and persistent disabling pain. Their goals for self-rated disability related to important behavioral goals were successfully attained. The utility of the experimental single-subject design in the analysis of effects in individual cases and in the development of an intervention research program also was shown.

As a result of this study, the 7 main components of the treatment program have been standardized for further evaluation of the individually tailored treatment program in a controlled between-group study now being conducted by physical therapists in a PHC setting. Provided that the results are replicated, the future clinical application will require provision of continuing education of physical therapists using this behavioral medicine approach to affect motor behaviors, cognitions, and activity in patients along the complex course of pain and disability.

Appendix.

Algorithm for the Individually Tailored Intervention

Phases and Components

1.

Behavioral goal identification and assessment Create a goal priority list including: goal identification, goal ranking (importance-frequency-level of difficulty), self-reports of activity, satisfaction, self-efficacy, fear avoidance, readiness to adopt new behaviors, treatment expectations Take a pain history Conduct a physical examination

2.

Self-monitoring Select initial behavioral treatment goal Tailor a self-monitoring diary for prospective recording (eg, activity, situation, pain, cognition, motor behavior, immediate consequences of behavior)

3.

Individual functional behavioral analysis

Introduce and discuss the hypothesized physical, cognitive, behavioral, and psychosocial determinants of disability Specify goals and intermediate objectives

4.

Basic skills acquisition

Introduce physical, psychological, and planning skills constituent to the initial treatment goal Exercise and record (quota, success, and failures)

5.

Applied skills acquisition

Merge basic skills and shape adequate behaviors Guide behavioral performance in contrived situations at the physical therapy clinic Apply activities training in everyday life and record (quota and experiences)

6.

Generalization of skills

Target subsequent goals on the priority list, extend the functional behavioral analysis Acquire additional skills and apply in everyday life, record (quota and experiences)

7.

Maintenance and relapse prevention

Discuss strategies for maintenance of skills Identify future risk situations, discuss and document management strategies Conduct booster sessions, including rehearsal and problem solving

The individual tailoring of the treatment program

Patient 1 (P1)

In the first phase of the program, P1 identified and ranked the following behavioral goals (1 ="most important to manage"): (1) to drive her car, (2) to make the beds, (3) to buy food for the household, (4) to manage household chores by the kitchen sink, (5) to vacuum clean the house, and (6) to hang up the laundry. It was decided to target household chores by the kitchen sink initially. P1 considered this activity important, and it occurred daily. She also reported the highest self-efficacy magnitude of attaining this goal in relation to the other goals on the priority list, and it included a well-defined motor behavior.

A self-monitoring diary (phase 2) was designed for daily records of activity performance by the kitchen sink. She recorded contextual contingencies, self-efficacy beliefs, negative cognitions, motor behavior, pain intensity, activity, and self-satisfaction each time she performed the activity.

The functional behavioral analyses (FBA) (phase 3) were based on data from the behavioral goal identification, the initial interview, the physical assessment, and the analysis of the self-monitoring diary. Four individual factors contributed to her disability: her appraisal of pain as a threat during activity performance, her anticipation of increased pain during and after activity performance, her low level in self-efficacy expectations in relation to task performance, and her physical impairments of low muscular endurance and strength (force-generating capacity of muscles). Her behavioral responses were further debilitating (ie, she used an inappropriate motor behavior and showed no ability to adjust her behavior to the different task demands). She also had considerable negative thoughts during activity performance. Three situational conditions increased her disability: the time of day (worse performance during the afternoons and in the evenings), extended duration of activity performance, and number of family members for whom to cook and clean up. Role expectations about her to be the family-member best suited for these tasks also were significant. The short-term consequences of her actual behaviors were tiredness, increased pain, bad mood, and feelings of hopelessness. She usually had to rest for hours after completion of the task. In the long-term, she was at risk for further physical deconditioning, extended sick leave, further social withdrawal, and isolation.

A primary goal for disability reduction was set at <2 on the 11-point numerical rating scale (0="no limitations"), and the goal was specified to include all kinds of chores by the sink with a duration up to 45 minutes and a frequency of at least 3 times a day. A further goal for return to work was set with quotas going from 0% to 25% of full-time work, then to 50%, and finally to 75% of full-time work, and a time frame for this was planned.

In the basic skills acquisition phase (phase 4), she chose "Nordic walking" with a daily quota of at least 30 minutes as her general physical activity. Relaxation techniques were taught and practiced. Exercises for shoulder flexion, rotation, and extension with a rubber band were introduced that combined coordinated humeroscapular movement patterns and muscle endurance training. She also was given exercises for enhancing trunk muscle strength. Personal quotas were discussed with special regard to her self-efficacy expectations of performing the quotas at home. A diary was used for documentation. Her appraisals of pain as threatening and hindering activity performance were discussed and challenged. She practiced to recognize her own negative cognitions and was encouraged to record her thoughts and behaviors in the self-monitoring diary.

The applied skills acquisition phase (phase 5) was initiated as soon as she recognized her negative cognitions, and she coordinated humeroscapular movements appropriately. She practiced to merge the basic skills of initial relaxation, the recognition of negative cognitions, an appropriate movement start and movement pattern during shoulder-arm movements, and body positioning in a contrived situation at the clinic. Negative cognitions were used as a cue to initiate positive self-talk of how to perform the task with an appropriate motor behavior and to instill self-efficacy to manage the task during present circumstances. Adequate performance was reinforced by the therapist, and a quota for applied exercise at home was discussed and subsequently performed and recorded in a diary.

At this time, she took up 25% of full-time work, which increased tiredness and reduced activity during spare time. To support her to stay active and to find a balance between activity and rest, the therapist helped her to schedule the working days.

When the initial goal was attained, skills should be generalized (phase 6) to subsequent goals on the priority list. The remaining goals were discussed and specified, and P1 was given the opportunity to complement the list. She practiced additional basic skills and applied them in their goal-specific context.

The therapist-led treatment period resulted in a personal maintenance and relapse prevention program (phase 7). P1 identified anticipated future risk situations (ie, vacuum cleaning of the house, wiping of floors, and gardening). Strategies for management were discussed and documented, and appropriate motor behaviors were practiced. She decided to continue her daily Nordic walking but had low self-efficacy expectations of performing the recommended physical exercise program for endurance and strength. At this time, she took up 50% of full-time work. To summarize, P1 participated in 18 therapist-led sessions during a 6-month period. Four booster sessions were carried out during the following 8 months. Her long-term goal of working 75% of full time was adjusted to maintenance of the present level (ie, 50% of full-time work).

Patient 2 (P2)

P2 identified and ranked the following 4 behavioral goals (phase 1): (1) to push wheelchairs at work, (2) to participate in pony-harness racing in her spare time, (3) to drive her car, and (4) to ride a bicycle beside the pony when her children were driving the sulky. All activities, except bicycling, were performed 5 to 7 days a week. Because the bicycling was infrequent, and thus not suitable as an intervention target, it was excluded from the list. She estimated driving her car to be the most easy to attain, and consequently it was chosen as the target for the initial part of the intervention.

She monitored (phase 2) her ability to drive her car back and forth to work 5 days a week in a self-report diary.

In the FBA (phase 3), it was hypothesized that the heavy steering wheel of her car contributed to her difficulties in driving. Three physical factors worsened her activity performance: impingement in the left shoulder, high muscle tension, and low muscle endurance in the neck and shoulder muscles bilaterally. She used an inappropriate motor behavior during driving, with particular difficulties in retracting and relaxing her shoulders, and in adaptation of her grip force. She reported high self-efficacy expectations related to task performance and reported no fear, which gave an advantageous starting point for the intervention. An immediate consequence of the morning drive was increased pain, giving a poor start for the even more demanding tasks at work. The long-term consequences could be described as a vicious circle of increased muscle tension, increased pain intensity, poor activity performance, a cognitive drive to challenge her and manage the task, an even higher muscle tension, more pain, and disability.

A primary goal for disability reduction was set at <2 on the 11-point numerical rating scale (0="no limitations") and the goal was specified to driving at least 4 times a day, 10 to 45 minutes in daylight as well in the dark, and in the winter season on dry as well as on snowy roads. The remaining goals on the behavior goal priority list were planned to be addressed after the first goal was attained.

The basic skills acquisition phase (phase 4) was focused on physical exercises to increase local circulation in soft tissues, to increase the subacromial space in the left shoulder, to improve humeroscapular coordination, and to adjust the grip force to the task at hand. Personal quotas for the exercises were discussed with special regard to her busy days and social role expectations of being available at home. A diary was used for documentation.

In the applied skills acquisition phase (phase 5), she was instructed to shape an appropriate motor behavior by merging basic skills of relaxation, positioning of arms and shoulders, and appropriate grip force. The skills were practiced every time she drove her car. Treatment was progressed toward the next 2 goals on the priority list when she scored car driving between 0 and 2 on the numerical rating scale.

The generalization phase (phase 6) focused on the next 2 goals on the priority list. New FBAs were discussed and goals were specified, and she practiced additional basic skills when needed. The skills then were applied in situations of everyday life.

She identified the following future risk situations to include in her personal maintenance and relapse prevention program (phase 7): to push a wheelchair for extended periods of time, to carry water buckets, and to handle restive horses. Because these situations frequently occurred, she was asked to try to use her skills in these activities on her own after the therapist-led sessions were completed.

To summarize, P2 participated in 11 therapist-led sessions during a 4-month period. Two booster sessions were carried out within the following 4 months.

* Lic Rehab Svetsary, Solna, Sweden.

References

(1) Ashburn MA, Staats PS. Management of chronic pain. Lancet. 1999; 353(9167):1865-1869.

(2) Loeser JD, Melzack R. Pain: an overview. Lancet. 1999;353(9164): 1607-1609.

(3) Melzack R, Wall P. Pain mechanisms: a new theory. Science. 1965;150: 971-979.

(4) Melzack R. From the gate to the neuromatrix. Pain. August 1999(suppl 6):S121-S126.

(5) Linton SJ, Melin L, Gotestam KG. Behavioral analysis of chronic pain and its management. Progress in Behavior Modification. 1984; 18:1-43.

(6) Fordyce WE. Behavioral Methods for Chronic Pain and Illness. St Louis, Mo: Mosby; 1976.

(7) Turk DC, Meichenbaum D, Genest M. Pain and Behavioral Medicine: A Cognitive-Behavioral Perspective. New York, NY: Guilford Press; 1983.

(8) Sharp TJ. Chronic pain: a reformulation of the cognitive-behavioural model. Behav Res Ther. 2001;39:787-800.

(9) Feine JS, Lund JP. An assessment of the efficacy of physical therapy and physical modalities for the control of chronic musculoskeletal pain. Pain. 1997;71:5-23.

(10) Hanada EY. Efficacy of rehabilitative therapy in regional musculoskeletal conditions. Best Pract Res Clin Rheumatol. 2003;17:151-166.

(11) McQuay HJ, Moore RA, Eccleston C, et al. Systematic review of outpatient services for chronic pain control. Health Technol Assess. 1999;1 (6):i-iv, 1-135.

(12) Ont i Ryggen, Ont i Nacken: En Evidensbaserad Kunskapssammanstallning [Back Pain and Neck Pain: An Evidence-Based Review]. Stockholm, Sweden; The Swedish Council on Technology and Assessment in Health Care; 2000:2.

(13) van Tulder M, Malmivaara A, Esmail R, Koes B. Exercise Therapy for Low Back Pain (Cochrane Review) [Update Software]. Oxford, United Kingdom: The Cochrane Library; 2003:1.

(14) Linton S. Why does chronic pain develop? A behavioral approach. In: Linton S, ed. New Avenues for the Prevention of Chronic Musculoskeletal Pain and Disability: Pain Research and Clinical Management. Vol 12. Amsterdam, the Netherlands: Elsevier Science BV; 2002:67-80.

(15) Morley S, Eccleston C, Williams A. Systematic review and metaanalysis of randomized controlled trials of cognitive behaviour therapy and behaviour therapy for chronic pain in adults, excluding headache. Pain. 1999;80:1-13.

(16) Turner JA. Educational and behavioral interventions for back pain in primary care. Spine. 1996;21:2851-2857.

(17) van Tulder M, Ostelo R, Vlaeyen J, et al. Behavioural Treatment for Chronic Low Back Pain (Cochrane Review) [Update Software]. Oxford, United Kingdom: The Cochrane Library; 2003:1.

(18) Schonstein E, Kenny D, Koes B. Work Conditioning, Work Hardening and Functional Restoration far Workers With Back and Neck Pain (Cochrane Review) [Update Software]. Oxford, United Kingdom: The Cochrane Library; 2003:1.

(19) Nicholas MK, Wilson PH, Goyen J. Operant-behavioural and cognitive-behavioural treatment for chronic low back pain. Behav Res Ther. 1991;29:225-238.

(20) Lindstrom I, Ohlund C, Eek C, et al. The effect of graded activity on patients with subacute low back pain: a randomized prospective clinical study with an operant-conditioning behavioral approach. Phys Ther. 1992;72:279-292.

(21) Harding V, Williams A. Activities training: integrating behavioral and cognitive methods with physiotherapy in pain management. J Occup Rehab. 1998;8:47-60.

(22) Johansson E, Lindberg P. Clinical application of physiotherapy with cognitive-behavioural approach in low back pain. Adv Physiother. 2001; 3:3-16.

(23) Soderlund A, Lindberg P. An integrated physiotherapy/cognitive-behavioural approach to the analysis and treatment of chronic Whiplash Associated Disorders, WAD. Disabil Rehabil. 2001;23:436-447

(24) Turk DC. Customizing treatment for chronic pain patients: who, what, and why. Clin J Pain. 1990;6:255-270.

(25) Kreuter M, Skinner C. Tailoring: what's in a name [editorial]? Health Educ Res Theor Pract. 2000; 15 (4) : 1-4.

(26) Turk DC, Rudy T, Kubinski J, et al. Dysfunctional patients with temporomandibular disorders: evaluating the efficacy of a tailored treatment protocol. J Consult Clin Psychol. 1996;64:139-146.

(27) Evers A, Kraaimaat F, van Riel P, de Jong A. Tailored cognitivebehavioral therapy in early rheumatoid arthritis for patients at risk: a randomized controlled trial. Pain. 2002;100:141-153.

(28) Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev. 1977;84:191-215.

(29) Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall; 1986.

(30) Arnstein P, Caudill M, Mandle C, et al. Self-efficacy as a mediator of the relationship between pain intensity, disability and depression in chronic pain patients. Pain. 1999;80:483-491.

(31) Arnstein P. The mediation of disability by self-efficacy in different samples of chronic pain patients. Disabil Rehabil. 2000;22:794-801.

(32) Ayre M, Tyson G. The role of self-efficacy and fear-avoidance beliefs in the prediction Of disability. Australian Psychologist. 2001;36:250-253.

(33) Denison E, Asenlof P, Lindberg P. Self-efficacy, fear avoidance, and pain intensity as predictors of disability in subacute and chronic musculoskeletal pain patients in primary health care. Pain. 2004;111: 245-252.

(34) Crombez G, Vlaeyen J, Heuts P, Lysens R. Pain-related fear is more disabling than pain itself: evidence on the role of pain-related fear in chronic back pain disability. Pain. 1999;80:329-339.

(35) Weiser S, Cedraschi C. Psychosocial issues in the prevention of chronic low back pain: a literature review. Baillieres Best Pract Res Clin Rheumatol. 1992;6:657-684.

(36) Linton S. A review of psychological risk factors in back and neck pain. Spine. 2000;25:1148-1156.

(37) Sturmey P. Functional Analysis in Clinical Psychology. Chichester, United Kingdom: John Wiley & Sons Ltd; 1996.

(38) Haynes SN, Leisen MB, Blaine DD. Design of individualized behavioral treatment programs using functional analytic clinical case models. Psychol Assess. 1997;9:334-348.

(39) Haynes SN. Behavioral assessment of adults. In: Goldstein G, Hersen M, eds. Handbook of Psychological Assessment. 3rd ed. Amsterdam, the Netherlands: Pergamon; 2000:471-502.

(40) Haynes SN, O'Brien WH. Functional analysis in behavior therapy. Clin Psychol Rev. 1990;10:649-668.

(41) Turk DC, Rudy TE. Neglected factors in chronic pain treatment outcome studies: referral patterns, failure to enter treatment, and attrition. Pain. 1990;43:7-25.

(42) Crook J, Weir R, Tunks E. An epidemiological follow-up survey of persistent pain sufferers in a group of family practice and specialty pain clinic. Pain. 1989;36:49-61.

(43) Deyo R, Phillips M. Low back pain: a primary care challenge. Spine. 1996;21:2826-2832.

(44) Kazdin AE. Research Design in Clinical Psychology. 4th ed. Boston, Mass: Allyn & Bacon; 2003: chap 10.

(45) Ottenbacher KJ. Clinically relevant designs for rehabilitation research: the idiographic model. Am J Phys Med Rehabil. 1990;69: 286-292.

(46) Boersma K, Linton S, Overmeer T, et al. Lowering fear-avoidance and enhancing function through exposure in vivo: a multiple baseline study across six patients with back pain. Pain. 2004;108:8-16.

(47) Kazdin A. Single-Case Experimental Designs: Methods for Clinical and Applied Settings. Oxford, United Kingdom: Oxford University Press; 1982.

(48) Turk DC, Okifuji A. A cognitive-behavioural approach to pain management. In: Wall P, Melzack R, eds. Textbook of Pain. 3rd ed. New York, NY: Guilford Press; 1993: chap 63.

(49) Kirk J. Cognitive-behavioural assessment. In: Hawton K, Salkovski P, Kirk J, et al, eds. Cognitive Behaviour Therapy for Psychiatric Problem: A Practical Guide. 7th ed. New York, NY: Oxford University Press; 1995: 13-51.

(50) Prochaska JO, DiClemente CC, Norcross JC. In search of how people change: applications to addictive behaviors. Am Psychol. 1992; 47:1102-1114.

(51) Asenlof P, Denison E, Lindberg P. Behavioral goal assessment in patients with persistent musculoskeletal pain. Phys Theor Pract. 2004;20: 243-254.

(52) Jensen IB, Linton SJ. Coping Strategies Questionnaire (CSQ): reliability of the Swedish version of the CSQ. Scand J Behav Ther. 1993;22:139-145.

(53) Jensen MP, Karoly P. Self-report scales and procedures for assessing pain in adults. In: Turk DC, Melzack R, eds. Handbook of Pain Assessment. 2nd ed. New York, NY: Guilford Press; 2001: chap 2.

(54) Pollard CA. Preliminary validity study of Pain Disability Index. Percept Mot Skills. 1984;59:974.

(55) Tait C, Pollard CA, Margolis RB, Duckro PN. The Pain Disability Index: psychometric and validity data. Arch Phys Med Rehabil. 1987;68: 438-441.

(56) Ohlen G. Spinal Sagittal Configuration and Mobilit. [dissertation]. Stockholm, Sweden: Department of Orthopaedic Surgery, Huddinge Hospital, Karolinska Institute; 1989.

(57) Ljungquist T, Fransson F, Harms-Ringdahl K, et al. A physiotherapy test package for assessing back and neck dysfunction: discriminative ability for patients versus healthy control subjects. Physiother Res Int. 1999;4:123-140.

(58) Johnson EP, ed. A CSM's Guidelines for Exercise Testing and Prescription. 6th ed. Baltimore, Md: Lippincott-Williams & Wilkins; 2000.

(59) Mayer TG, Barned D, Kishino ND, et al. Progressive isoinertial lifting evaluation, I: a standardized protocol and normative database. Spine. 1988;13:993-997.

(60) Franklin RD, Allison DB, Gorman BS. Design and Analysis of SingleCase Research. Mahwah, NJ: Lawrence Erlbaum Associates Inc; 1997.

(61) Ottenbacher KJ. Evaluating Clinical Change: Strategies for Occupational and Physical Therapists. Baltimore, Md: Williams & Wilkins; 1986.

(62) Farrar JT, Young JP Jr, LaMoreaux L, et al. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain. 2001;94:149-158.

(63) Linton S. Psykologiska behandlingsmetoder. In: Ont i Ryggen, Ont i Nacken: En Evidensbaserad Kunskapssammanstallning [Back Pain and Neck Pain: An Evidence-Based Review]. Stockholm, Sweden; The Swedish Council on Technology and Assessment in Health Care; 2000;2:205-233.

(64) Strecher V, Seijts G, Kok G, et al. Goal setting as a strategy for health behavior change. Health Education Quarterly. 1995;22:190-200.

(65) Locke EA, Latham GP. A Theory of Goal Setting and Task Performance. Englewood Cliffs, NJ: Prentice Hall; 1990.

(66) Bandura A; Hastings C, ed. Self-efficacy. New York, NY: WH Freeman & Co; 1997.

(67) Asmundson G. Behavioural assessment--pain assessment: state-ofthe-art applications from the cognitive-behavioral perspective. Behav Res Ther. 2002;40:547-550.

(68) Owens RG, Aschcroft JB. Functional analysis in applied psychology. Br J Clin Psychol. 1982;21:181-189.

(69) Cone JD. Issues in functional analysis in behavioral assessment. Behav Res Ther. 1997;35:259-275.

(70) Drabman R, Hammer D, Rosenbaum M. Assessing generalization in behavior modification with children: the generalization map. Behavioral Assessment. 1979;1:203-219.

P Asenlof, PhD, is Registered Physical Therapist and Senior Lecturer, Department of Public Health and Caring Sciences/Section of Caring Sciences, Uppsala University, Uppsala Science Park, S-751 85 Uppsala, Sweden ([e-mail]). Address all correspondence to Dr Asenlof.

E Denison, PhD, is Registered Physical Therapist and Senior Lecturer, Department of Public Health and Caring Sciences/Section of Caring Sciences, Uppsala University.

P Lindberg, PhD, is Associate Professor in Clinical Psychology, Department of Public Health and Caring Sciences/Section of Caring Sciences, Uppsala University.

All authors provided concept/idea/research design, writing, and data analysis. Dr Asenlof provided data collection. Dr Asenlof and Dr Lindberg provided project management and fund procurement. Dr Lindberg provided facilities/equipment and institutional liaisons.

This study was approved by the ethics committee of the faculty of medicine of Uppsala University.

This research was presented as a part of Dr Asenlof's PhD thesis (ISSN 1651-6206, ISBN 91-554-6240-5, urn:nbn:se:uu:diva-5781).

The study was financially supported by the Swedish Research Council and the Section for the Swedish Council on Technology Assessment in Health Care and Research, Uppsala County Council.

Table 1.
The 7 Components of the Intervention and the Corresponding Phases
in the Multiple Baseline Design (a)

Phases and Components                              Design Phases

1. Behavioral goal identification and assessment   B
     Priority list
     Frequency of behavior
     Level of difficulty
2. Self-monitoring                                 B
     Motor behavior, cognition, and activity in
       situations of everyday life
3. Individual functional behavioral analysis       B
     Goal specification
4. Basic skills acquisition                        B
5. Applied skills acquisition                      C
     Contrived environments
     Situations of everyday life
6. Generalization                                  C
7. Maintenance and relapse prevention              [A.sub.2]

(a) No treatment component is introduced during the [A.sub.1] phase of
the design.

Table 2.
Descriptive Data of the Standardized Physical Tests at Admission and
Discharge for Patients 1 and 2 (a)

                                     Patient 1

                                 Admission   Discharge

Cervical ROM ([degrees])
  Forward flexion                       66          66
  Backward extension                    44          40
  Total rotation                        90         126
  Total lateral bending                 36          52
Shoulder ROM ([degrees])
Thoracic ROM ([degrees])
  Total forward flexion and             20           5
    backward extension
Lumbar ROM ([degrees])
  Total forward flexion and             49          48
    backward extension
Isometric muscle endurance (s)
  Neck extensors                        19          26
  Neck flexors                           4           7
  Neck extensors3                       17          13
Sit-ups                                  8          15
Push-ups
Cervical lifting test (kg)               8           8

                                      Patient 2

                                 Admission   Discharge

Cervical ROM ([degrees])
  Forward flexion                       54          62
  Backward extension                    70          70
  Total rotation                       126         144
  Total lateral bending                 60          66
Shoulder ROM ([degrees])              N/A         N/A
Thoracic ROM ([degrees])
  Total forward flexion and
    backward extension
Lumbar ROM ([degrees])
  Total forward flexion and
    backward extension
Isometric muscle endurance (s)
  Neck extensors                       360         360
  Neck flexors                          35          64
  Neck extensors3                      174         178
Sit-ups                                 17          24
Push-ups                                 3           7
Cervical lifting test (kg)               8           8

(a) ROM = range of motion, N/A = not affected.

Table 3.
Raw Scores and Percentages of Change in General Pain-Related Disability
and Mean Values and Standard Deviations (in Parentheses) for Pain
Intensity and Pain Control Before Treatment (Pre), After Treatment
(Post), and at 1-, 4-, 6-, and 12-Month Follow-up Examinations
(F1-F4) (a)

      Patient 1

       Pre    Post    %     F1      F2      F3      F4     %

PDI    42       7     83     5       5       6       2     95
PI      4       2     50    2.1     1.9     2.5     1.4    65
      (1.6)   (0.7)        (0.7)   (0.8)   (0.5)   (0.8)
PC     3.3      5     51    4.6     5.5     5.7     5.8    76
      (0.8)   (0.4)        (1.3)   (0.9)   (0.7)   (0.4)

      Patient 2

       Pre    Post    %     F1      F2      F3      F4     %

PDI    10       3     70     3       2       3       1     90
PI     3.3     1.1    67    0.7     0.5     0.1     0.1    97
      (0.8)   (0.5)        (0.5)   (0.5)   (0.3)   (0.4)
PC     3.9      5     28     5      5.8      6      5.9    51
      (0.8)   (0.0)        (0.0)   (0.4)   (0.0)   (0.3)

(a) PDI = Pain Disability Index (low scores indicate low disability,
70 = maximum score), PI = pain intensity (0-10 numerical rating scale,
low scores indicate low intensity), PC = pain control (0-6 numerical
rating scale, low scores indicate low control, % = percentage of
change.