THE ROLE OF PATIENTS' EXPECTATION OF APPROPRIATE INITIAL MANUAL THERAPY TREATMENT IN OUTCOMES FOR PATIENTS WITH LOW BACK

PAIN
Megan Donaldson, PT, PhD, a Kenneth Learman, PT, PhD, b Bryan OHalloran, DPT, c Christopher Showalter, PT, d and Chad Cook, PT, PhD e

ABSTRACT
Objectives: The purpose of this study was to compare the outcomes of patients with low back pain who had a matched, unmatched, or were indifferent with their pretreatment expectation/preference to the manual therapy thrust or nonthrust intervention. A secondary purpose was to explore baseline patient characteristics affiliated with a preference of manual therapy treatment type (thrust or nonthrust manipulation). Methods: The study is a secondary analysis of a prospective randomized controlled trial that enrolled 149 patients with low back pain, from 16 distinct outpatient physical therapy practices within the United States. Patient's pretreatment expectations were identified before randomization. The intervention included manual therapy (thrust or nonthrust manipulation) with a standardized exercise program, which was carried out for 2 treatment sessions. Six outcomes captured the constructs of (a) disability, (b) pain perception, (c) care intensity, (d) fear avoidance behaviors, and (e) perception of extent of recovery. Descriptive statistics, analysis of variance (and Tukey, a post hoc test), Fisher exact test, and a multivariate logistic regression analysis were used for analysis. Results: There were no statistical differences in any of the 6 dedicated outcomes measures between the matched, unmatched, and no preference groups. Therapist-determined patient irritability was associated with patient preference of nonthrust treatment, and higher body mass index was associated with patient preference of thrust treatment. Conclusions: This study shows that patients demonstrated no statistical difference in disability or pain outcome measures when matched, unmatched, or indifferent to the intervention. (J Manipulative Physiol Ther 2013;36:276-283) Key Indexing Terms: Low Back Pain; Patient Preference; Musculoskeletal Manipulations

F
a b

or a randomized controlled trial (RCT), a state of equipoise or genuine uncertainty in the comparative merits of the treatments being assessed is necessary for a fair; unbiased; and, in some cases, ethical

Assistant Professor, Walsh University, North Canton, Ohio. Associate Professor, Youngstown State University, Youngstown, Ohio. c Director, Pain Relief and Physical Therapy, Havertown, Pa. d Clinical Director, Maitland-Australian Physiotherapy Seminars, Cutchogue, NY. e Professor, Walsh University, North Canton, Ohio. Submit requests for reprints to: Megan Burrowbridge Donaldson, PT, PhD, Associate Professor, FAAOMPT, 2020 East Maple St NW, North Canton, Ohio 44720 (e-mail: mdonaldson@walsh.edu). Paper submitted September 11, 2011; in revised form December 20, 2012; accepted December 27, 2012. 0161-4754/$36.00 Copyright 2013 by National University of Health Sciences. http://dx.doi.org/10.1016/j.jmpt.2013.05.016

design. 1 Equipoise is categorized in many ways, including whether there is no evidence within the literature that one intervention is genuinely more effective than another (clinical equipoise), whether there is no evidence that the expert clinical practitioners favor one intervention over another (community or personal equipoise), or whether there is no evidence that patients have expectations/ preferences toward one treatment arm or the other. 2 In reality, most patients likely do have preferences/expectations regarding the treatment choices provided, which could potentially influence the within-group changes and between-group differences. 3 Furthermore, it has been hypothesized that when patients are given the opposite intervention to their expectation/preference (unmatched treatment), they have a greater probability of experiencing worse outcomes in comparison with those receiving matched treatment. 4 The definition of patient preference differs only slightly from patient expectations. 5 Of the 2, patient expectation is

276

Journal of Manipulative and Physiological Therapeutics Volume 36, Number 5

Donaldson et al Expectation of Manual Therapy

277

more widely studied. Expectation involves the patient's beliefs regarding the potential benefit of the treatment 6 and the likelihood of an outcome or an expected effect (prognosis). 7 A patient's expectation has been linked to levels of pain and recovery. 5,7,8 By definition, patient preference is derived from patient's expectations about the care that they will receive, 9 and results from deliberation about specific elements, such as anticipated treatments or health outcomes. Although patient preferences for a certain treatment, which are related to, but distinct from, their expectations of the results of that treatment, together both may have an important role in affecting treatment outcome. Recently, in a study by Lurie et al 9 found that the patients' expectations regarding improvement with nonoperative care was the strongest predictor of their treatment preference. 9 In essence, the constructs of the 2 definitions are conditionally related and should be considered in combination when examining an effect during a clinical trial. The determinant that drives a patient to pick an expectation/preference of one intervention over another intervention is less known. The most commonly reported preferences/expectation influence is a patient's past experience with an intervention, a health care provider, or a particular health-related service. 10 Other preferences/expectations involve the ability of the patient to cope with the prescribed intervention, how the intervention may affect lifestyle choices, and verbal and written knowledge of the effectiveness/ineffectiveness of a technique. 11 Preferences/ expectations may also be associated with the perceived nature of some treatments 12: in other words, whether the treatment is complex or simple or whether the actionable component of the treatment choice is complementary to the patient's perception of their underlying problem. Recently, Bishop et al 13 found that patients who received physical therapy for treatment of nonspecific low back pain (LBP) expected active approaches to be more beneficial than passive approaches such as modalities and medications. However, one could argue that a patient who attends a treatment such as physical therapy is looking for an active intervention because this approach is more reflective of what is provided by physical therapists. Those who pursue a massage therapist are likely looking for a passive approach for their care. Yet many of the procedures used in dedicated disciplines are markedly similar in appearance and outcome. What remains unknown is whether patients would have preferences on similar active or passive interventions such as 2 forms of manual therapy or 2 different forms of exercise, if both procedures had similar constructs. At present, the influence of patient expectations/ preferences on outcomes in clinical trials has provided mixed results. 14,15 Differences in findings may be associated with design variations, how the definitions were used within the study for expectations/preferences, and whether

indifference toward an intervention choice was captured. To our knowledge, no studies have looked at the matched, unmatched, and/or indifferent treatment expectations/preferences on recovery in a manual physical therapy oriented clinical trial for mechanical LBP. Therefore, the primary objective of this study is to compare the outcomes of patients with LBP who received care that was (a) matched to expectation/preference, (b) unmatched to expectation/preference or included patients who were indifferent to an expectation/preference. A secondary objective was to explore baseline patient characteristics affiliated with a preference of manual therapy treatment type (thrust or nonthrust manipulation) because the interventions are similar in construct and/or both advocated by treatment guidelines.

METHODS
Design
The secondary database analysis is extracted from an RCT (clinicaltrials.gov no. NCT01438203). The RCT compared thrust and nonthrust manipulation in the treatment of LBP; thus, both groups in the study received a manual therapy approach. The study was approved by the Walsh University Human Ethics Review Board. All subjects gave consent to participate in this study.

Participants
The original RCT study from which this secondary analysis was performed has been described elsewhere and is presented to give context for this current study in Figure 1. 16 Participants were 149 patients with LBP, all of which were treated in 1 of 16 distinct outpatient physical therapy practices within the United States by 1 of 17 skilled outpatient physical therapists. Clinicians had undergone extensive manual therapy training and certification in orthopedic manual therapy or were manual therapy fellows within the American Academy of Orthopaedic Manual Physical Therapists. Study inclusion required an age of 18 years or older with mechanically producible LBP. There were no restrictions on minimum baseline Oswestry Disability Index (ODI) scores or pain scores. Exclusion criteria included the presence of a tumor, metabolic diseases, rheumatoid arthritis, osteoporosis, prolonged history of steroid use, or signs consistent with nerve root compression (any of the following: reproduction of low back or leg pain with straight leg raise b 45, muscle weakness involving a major muscle group of the lower extremity, diminished lower extremity muscle stretch reflex, or diminished or absent sensation to pinprick in any lower extremity dermatome). Participants with a prior surgical history of the lumbar spine and current pregnancy were also excluded.

278

Donaldson et al Expectation of Manual Therapy

Journal of Manipulative and Physiological Therapeutics June 2013

Fig 1. Patient flow diagram. (Color version of figure is available online.)

Intervention
The intervention was a comprehensive rehabilitation approach that included 1 of 2 forms of manual therapy (thrust or nonthrust manipulation) for the first 2 visits only, followed by physical therapistdirected care after the initial 2 visits. Purposely, the trial was both prescriptive (the patients received a dedicated intervention, either thrust or nonthrust manipulation) and pragmatic in that the techniques themselves (thrust and nonthrust manipulation) were

not performed in a prescribed manner but differed depending on what the clinician felt was most appropriate.

Assessment of Patient Expectations/Preferences

After consent and before randomization to thrust or nonthrust intervention and after the details and procedures of the study were explained to the patient, patients were queried regarding their expectation/preferences of which

Journal of Manipulative and Physiological Therapeutics Volume 36, Number 5

Donaldson et al Expectation of Manual Therapy

279

intervention (thrust or nonthrust) would be most effective for their LBP. The patients were asked which of the two techniques do you think would benefit your condition the most? Patients were instructed to choose one or the other but were not pressed if they indicated that they had no preferences of thrust or nonthrust. If needed, patients were shown photos of the 2 techniques, and/or the techniques were further described in detail. In spirit, the question was purposively blended to assess both their expectation and preference for the interventions.

to subside once aggravated (ie, pain persistence). The variable was dichotomously coded, as recommended by Maitland, 23,24 as present or not present, where present qualified as any one or more excessive findings recognized on the 3 identifiers.

Data Analysis
Data analysis included descriptive statistics for each of the 3 groups (matched expectations, unmatched expectations, and no preference). Expectations/preferences were matched if the person selected either thrust or nonthrust and was allocated to thrust or nonthrust, respectively. A 1-way analysis of variance (and Tukey, a post hoc test) was used to determine differences in outcomes among the 3 groups for continuous variables. A Fisher exact test was used to measure differences among the 3 groups for meeting the cut point of the 50% change in the ODI because it is the recommended statistical test for small sample size categorical data. For our secondary objective, a multivariate logistic regression analysis was used to identify characteristics associated with those who expected either thrust or nonthrust manipulation to be most beneficial for their condition (nonpreference was removed from the analysis). As stated previously, we targeted the baseline characteristics of age; BMI; duration of symptoms in weeks; initial (baseline) scores of ODI, NPRS, and FABQ-w; sex; and irritability status as the predictor variables. The dependent variable was a 50% change on the ODI. An value of .05 was selected for significance for all analyses.

Outcomes Measures
There were 6 primary outcomes measures used within the study. The 6 outcomes captured the constructs of (a) disability (ODI in a linear and dichotomous fashion), (b) pain perception, (c) care intensity (patient visits and days of care), (d) fear avoidance behaviors (FABQ), and (e) perception of extent of recovery. The ODI 17 was used to measure disability and is a scale that consists of 10 questions each scored from 0 to 5, with higher scores indicating greater disability. A linear change score and a 50% reduction of the ODI were both used as individual measures. A 50% reduction of the ODI has been suggested as a usual cut point for meaningful change by past authors. 18 The Numerical Pain Rating Scale (NPRS) 19 was used to capture the patient's level of pain. For the NPRS, patients were asked to indicate the intensity of their current back pain, using an 11-point ordinal scale ranging from 0 no pain to 10 worst pain imaginable. The variable, total visits, was used to measure the quantity of treatment, and days in care was used to capture the temporal duration of care. Selfreport of extent of recovery (0%-100%) was used to measure perception of recovery. The question was stated as What percent, 0 percent (meaning not at all) to 100 percent (meaning totally recovered) do you feel that you have recovered at this point? The self-report of extent of recovery is a variant of the Single Alphanumeric Evaluation, 20 which has been used previously with patients with shoulder pain 21 and LBP. 22

RESULTS
The descriptive statistics for this sample are reported in Table 1. There were no significant differences in baseline descriptive criteria among the 3 expectation groups for any of the variables. In total, 52 individuals selected thrust manipulation as the most beneficial for their condition, 89 selected nonthrust manipulation as the most beneficial, and 8 demonstrated no preference/expectation; however, 84 received matched care, When analyzing these data, the assumption of homogeneity of variance was upheld; however, there were several violations of normality of distributions. Bearing in mind that the no-preference group included such a small sample resulting in large differences in group sample sizes, the investigators elected to abandon the use of the 1-way analysis of variance and analyze these data with a KruskalWallis test. There were no statistical differences in the NPRS or ODI change scores, the percent recovery, and the total visits outcomes measures between the matched, unmatched, and no-preference groups. There were, however, significant differences between the FABQ change scores and the total days in care with P = .045 and P = .037, respectively (Table 2).

Predictive VariablesDemographics
To our knowledge, there are no predictors that have been reported in the literature that are reflective of the tendency to exhibit preference/expectations toward thrust or nonthrust manipulation. Consequently, we selected irritability, sex, age, body mass index (BMI), duration of symptoms, the NPRS at baseline, the FABQ-w at baseline, and the ODI at baseline as potential predictor variables. Duration of symptoms was described in weeks. Irritability was a concept espoused by Maitland 23 and includes 3 primary situational identifiers: (1) the vigor of activity required to provoke a patient's symptoms, (2) the severity of those symptoms, and (3) the time it takes for the symptoms

280

Donaldson et al Expectation of Manual Therapy

Journal of Manipulative and Physiological Therapeutics June 2013

Table 1. Descriptive characteristics of the patients divided into matched vs unmatched groups (n = 149)
Variable Age (y) Duration of symptoms (wk) Sex Race Matched subjects, mean (SD)/frequency, n = 84 49.8 (15.4) 43.2 (127.5) 39 = male 45 = female 78 = White 3 = Black 2 = Hispanic 0 = Asian 0 = other 1 = missing 67.4 (3.9) 169.4 (33.0) 26.2 (4.2) 30.5 (15.5) 11.1 (10.2) 5.2 (2.1) 34 = Thrust 50 = Nonthrust 0 = No preference Unmatched subjects, mean (SD)/frequency, n = 57 45.5 (13.8) 22.9 (37.3) 27 = male 30 = female 51 = White 0 = Black 1 = Hispanic 2 = Asian 2 = other 1 = missing 67.2 (3.9) 171.4 (36.9) 26.6 (4.9) 31.6 (16.9) 13.1 (11.7) 5.4 (2.1) 18 = Thrust 39 = Nonthrust 0 = No preference Indifferent about assignment, n = 8 51.2 (16.4) 14.0 (21.1) 4 = male 4 = female 7 = White 0 = Black 0 = Hispanic 1 = Asian Pa .21 .41 .98 .13

Height Weight BMI Baseline ODI Baseline FABQ Baseline numeric analog for pain Patient expectations

67.0 (2.8) 175.1 (51.9) 27.4 (8.6) 24.0 (7.8) 11.6 (7.2) 4.2 (2.2) 0 = Thrust 0 = Nonthrust 8 = No preference

.94 .88 .69 .44 .58 .41 b .01

BMI, body mass index; FABQ, fear avoidance behaviors; ODI, Oswestry Disability Index. a Significance at the 0.05 level.

Table 2. Change scores of matched vs unmatched subjects to their outcome expectation or preferred treatment (n = 141)
Outcome variable ODI (change score) Numeric analog scale for pain (change score) Fear avoidance beliefs characteristics (work subscale) Self-report, percent recovered ODI 50% improvement Total physical therapy visits Total days from initiation to discharge ODI, Oswestry Disability Index. a Significance at the 0.05 level. Matched subjects, mean (SD)/frequency, n = 84 14.7 (11.6) 3.2 (2.5) 1.1 (5.9) 78.7 (22.2) 47 = Yes 37 = No 6.9 (4.7) 33.9 (30.6) Unmatched subjects, mean (SD)/frequency, n = 57 15.7 (13.9) 3.6 (2.2) 1.7 (5.2) 80.9 (21.2) 35 = Yes 22 = No 7.3 (4.6) 39.9 (29.6) Indifferent about assignment, n = 8 7.7 (11.2) 1.8 (2.6) 6.0 (4.8) 70.7 (32.1) 3 = Yes 5 = No 5.7 (2.6) 20.8 (11.5) Pa .25 .22 .05 .70 .47 .45 .04

There were 2 variables that were significantly associated with the likelihood of selecting either thrust or nonthrust manipulation as an expectation for benefiting their condition the most. Individuals with higher BMI were 1.1 times as likely (95% confidence interval [CI], 1.04-1.3) to demonstrate preferences/expectations toward thrust manipulation than those with a lower BMI. Individuals who were identified as irritable by the physical therapists were 4.5 times less likely (odds ratio [OR], 0.22; 95% CI, 0.08-0.63) to identify thrust manipulation as those without irritability as the most potentially beneficial treatment. Table 3 outlines the ORs, significance, and 95% CIs of the predictors.

DISCUSSION
This study's primary objective was to compare the outcomes of LBP patients who had a matched, unmatched, or were indifferent with their pretreatment expectations/ preferences to the manual therapy thrust or nonthrust manipulation interventions, and there were no differences

among any of the targeted outcomes measures. With respect to our second objective, BMI and irritability were associated with the selection of a specific treatment intervention: those with higher BMI more likely to select thrust manipulation and those who were identified as irritable by the physical therapist at baseline were more likely to identify nonthrust manipulation. To our knowledge, we are the first that have attempted to identify preferences/expectations for similar physical therapy techniques such as thrust and nonthrust manipulations, 2 techniques that have very similar constructs. Our finding that matching one's expectation/preference to their randomized treatment did not lead to an improved outcome is similar to the outcomes reported previously. George and Robinson 14 examined behavioral interventions for acute and subacute LBP and, in a well-designed study, separated levels of patient expectation, levels of patient preference, and patient satisfaction. They reported similar 4-week and 6-month improvements for pain intensity and disability for subjects receiving matched and unmatched

Journal of Manipulative and Physiological Therapeutics Volume 36, Number 5

Donaldson et al Expectation of Manual Therapy

281

Table 3. Logistic regression analysis of predictors for thrust or nonthrust manipulation for expectations of beneficial treatment
Variable Age BMI Irritability Duration NPRS at baseline Fear avoidance behaviors beliefs work subscale at baseline ODI at baseline Sex Odds Ratio (95% CI) 0.98 (0.98-1.01) 1.14 (1.04-1.25) 0.22 (0.08-0.63) 1.00 (0.99-1.01) 1.11 (0.93-1.34) 1.03 (0.96-1.10) 1.03 (0.99-1.06) 0.85 (0.40-1.8) Pa .36 b .01 b .01 .29 .25 .33 .12 .68

BMI, body mass index; NPRS, Numerical Pain Rating Scale; ODI, Oswestry Disability Index. a Significance at the 0.05 level.

treatments. Ironically, patients without a treatment preference had the largest, statistically significant (P b .05), 6month improvements for pain intensity and disability scores. Furthermore, George and Robinson 14 found that patient expectation for recovery was more predictive of actual perceived recovery. This may suggest that the intervention is or is not effective based on its own capacity, vs purely upon the patient's expectations/preferences. An example is provided by Bishop et al, 13 who indicated that when thrust manipulation was matched to groups who met a clinical prediction rule for spinal thrust manipulation, their outcomes were better after receiving the intervention, regardless of whether they felt thrust manipulation would be useful in their condition. 13 In a study of 323 subjects with major depression, patient preference allocation had no impact over random allocation for the treatment of moderate depression. 25 These findings are in contrast to that of Lin et al 26 who found that allocation by preference provided superior results to randomization for treatment of depression in the short term with differences becoming nonsignificant at a 9-month follow-up. A portion of these differences may be attributed to study design particularly concerning the operational definition of preference matching. As stated, other studies have shown that patient's expectation has been linked to levels of pain and recovery. 5,7,8,14,2730 This seeming contradiction between matching patient's expectations in general being predictive of outcome and patient's expectation for a given treatment being predictive of outcome may be influenced by an individual's lack of experience with a given treatment, which may attenuate the potential effect of expected outcome. Several studies in oncology support the notion that patients routinely make treatment decision on misconception, fear, and perceived survival potential 31 and that initial treatment preference may be altered using extensive education from physicians. 32 A second possibility is that patients who are optimistic about recovery possess this belief independent of specific interventionsthey are optimistic and expect a better

outcome regardless of treatment. This concept is partially supported by George and Robinson, 14 who found that subjects who had higher expectation had higher perceived effectiveness independent of treatment type. In addition, Linde et al 3 found that subject's beliefs regarding the effectiveness of care (in this case, acupuncture) did not necessarily predict perceived expectation of treatment outcome in subjects with chronic LBP, knee osteoarthritis, chronic tension headache, and migraine headaches. Carefully designed studies that may account for a patient's knowledge and experience with given techniques before asking about preferences may allow further investigation of this phenomenon. An interesting finding in the study was the association of BMI and a status of irritability, toward the likelihood of expecting/preferring one manual therapy intervention over another. We have found no explanatory evidence of why patients with higher BMI had expectations/preferences that were greater for thrust than nonthrust manipulation. Application-wise, a thrust manipulation is a more challenging procedure to perform on larger patients; however, this does not explain the patient's preferences. Both interventions are passive, and both are performed exclusively by the clinician. In addition, those patients who were identified at baseline as irritable by the physical therapist were over 4 times as likely to identify nonthrust manipulation more beneficial to their condition than thrust manipulation. Patients who are identified as irritable often exhibit a complex clinical picture that suggests that their condition is easy to irritate (worsen) and difficult to calm down. Although there is limited research on the concept of irritability, a study by Barakatt et al 23 found that pain persistence is a pain behavior that is required to identify a patient that is deemed to have a condition that is irritable. This suggests that there may be a reasonable relationship between what a patient perceives about their own condition and that which is identified by the physical therapist during the examination. In addition, Barakatt et al 24 found reasonable reliability among clinicians assessing irritability ( = .44). As a result of our study, the patient with an irritable condition may exhibit pain behaviors and concerns that are different from those who do not have an irritable condition. Therefore, the description or demonstration of the thrust technique during the information session may observe thrust manipulation as being different in some negative manner from the nonthrust technique and altered his/her treatment expectation/ preference toward a nonthrust treatment choice.

LIMITATIONS
There is no formal measurement tool used in physical therapy to assess patient expectation, other than to quantify expectation for overall improvement in a condition. We used a blended preference/outcome expectation question to

282

Donaldson et al Expectation of Manual Therapy

Journal of Manipulative and Physiological Therapeutics June 2013

identify which treatment (thrust or nonthrust) the patient thought would give him/her the most benefit for his/her condition. The construct of patient outcome expectation itself has a multitude of dimensions, which may influence the patient including self-efficacy, social influence, media influence, patient satisfaction, and demographic factors. 33 Our study gathered demographic information; however, the aforementioned dimensions were not accounted for in this trial. In addition, we did not capture the strength of the patient's preference or outcome expectation during this trial and that may have some effect that could not be captured in this study. Other studies have examined this on a Likert scale to identify strength of preference or expectation, and this could have altered our findings. Lastly, we were unable to capture the calculated effect of indifference or no preference/expectation toward the interventions due to the significantly small sample size in this group, which would over underrepresent this population statistically.

REFERENCES
1. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317:141-5. 2. Elwyn G, Edwards A, Kinnersley P, Grol R. Shared decision making and the concept of equipoise: the competences of involving patients in healthcare choices. Br J Gen Pract 2000; 50:892-9. 3. Linde K, Witt CM, Streng A, et al. The impact of patient expectations on outcomes in four randomized controlled trails of acupuncture in patients with chronic pain. Pain 2007;128: 264-71. 4. Torgerson DJ, Klaber-Moffett J, Russell IT. Patient preferences in randomised trials: threat or opportunity? J Health Serv Res Policy 1996;1:94-7. 5. Mondloch MV, Cole DC, Frank JW. Does how you do depend on how you think you'll do? A systematic review of the evidence for a relation between patients' recovery expectations and health outcomes. CMAJ 2001;165:174-9. 6. Barron CJ, Moffett JA, Potter M. Patient expectations of physiotherapy: definitions, concepts, and theories. Physiother Theory Pract 2007;23:37-46. 7. Preference Collaborative Review Group. Patients' preferences within randomized trials: systematic review and patient level meta-analysis. BMJ, 337; 2008. p. 1864. 8. Price DD, Barrell JJ. An analysis of factors that contribute to the magnitude of placebo analgesia in an experimental paradigm. Pain 1999;83:14-56. 9. Lurie JD, Berven SH, Gibson-Chambers J, et al. Patient preferences and expectations for care: determinants in patients with lumbar intervertebral disc herniation. Spine 2008;33: 2663-8. 10. Bowling A, Rowe G, Lambert N, et al. The measurement of patients' expectations for health care: a review and psychometric testing of a measure of patients' expectations. Health Technol Assess 2012;16:i-xii, 1509. [Epub 2012/07/04]. 11. Chanouzas D, Ng KP, Fallouh B, Baharani J. What influences patient choice of treatment modality at the predialysis stage? Nephrol Dial Transplant 2012;27:1542-7 [Epub 2011/08/26]. 12. Umar N, Litaker D, Schaarschmidt ML, Peitsch WK, Schmieder A, Terris DD. Outcomes associated with matching patients' treatment preferences to physicians' recommendations: study methodology. BMC Health Serv Res 2012;12:1 [Epub 2012/01/05]. 13. Bishop MD, Bialosky JA, Cleland JE. Patient expectations of benefit from common interventions for low back pain and effects on outcome: secondary analysis of a clinical trial of manual therapy interventions. J Man Manipulative Ther 2011; 19:20-5. 14. George SZ, Robinson ME. Preference, expectation, and satisfaction in a clinical trial of behavioral interventions for acute and sub-acute low back pain. J Pain 2010;11: 1074-82. 15. Bialosky JE, Bishop MD, Cleland JA. Individual Expectation: an overlooked, but pertinent, factor in the treatment of individuals experience musculoskeletal pain. Phys Ther 2010; 90:1345-55. 16. Cook CE, Showalter C, Kabbaz V, O'Halloran B. Can a within/between-session change in pain during reassessment predict outcome using a manual therapy intervention in patients with mechanical low back pain? Man Ther 2012;17: 325-9. 17. Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine 2000;25:2940-52.

CONCLUSION
This study showed that patients enrolled in a randomized patient preference/outcome expectation trial demonstrated no statistical difference in disability or pain outcome measures when matched or unmatched to the manual therapy intervention. The characteristics of those who selected the type of intervention appear novel, as patients deemed irritable by the clinician were 4 times more likely to select nonthrust interventions and patients with a higher BMI were more likely to select thrust interventions. Based on our findings, the construct of irritability, although defined by Maitland, may need further exploration in LBP patient expectation/preference trials.

Practical Applications
This study shows that patients enrolled in a fully randomized preference or outcome expectation trial demonstrated no statistical difference in disability or pain outcome measures when matched, unmatched, or indifferent to the intervention. Therapist determined patient irritability was associated with patient preference of nonthrust treatment, and higher BMI was associated with patient preference of thrust treatment.

FUNDING SOURCES AND POTENTIAL CONFLICTS OF INTEREST

No funding sources or conflicts of interest were reported for this study.

Journal of Manipulative and Physiological Therapeutics Volume 36, Number 5

Donaldson et al Expectation of Manual Therapy

283

18. Fritz JM, Hebert J, Koppenhaver S, Parent E. Beyond minimally important change: defining a successful outcome of physical therapy for patients with low back pain. Spine 2009;34:2803-9. 19. Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in patients with low back pain. Spine 2005; 30:1331-4. 20. Kamper S, Maher C, Mackay G. Global rating of change scales: a review of strengths and weaknesses and considerations for design. J Man Manipulative Ther 2009;17:163-70. 21. Williams G, Gangel T, Arciero R, Uhorchak J, Taylor D. Comparison of the single assessment numeric evaluation method and two shoulder rating scales. Am J Sports Med 1999;27:214-22. 22. Van Kleef M, Barendse GA, Kessels A, Voets HM, Weber WE, de Lange S. Randomized trial of radiofrequency lumbar facet denervation for chronic low back pain. Spine 1999;24: 1937-42. 23. Barakatt ET, Romano PS, Riddle DL, Beckett LA, Kravitz R. An exploration of Maitland's concept of pain irritability in patients with low back pain. J Man Manipulative Ther 2009;17:196-205. 24. Barakatt ET, Romano PS, Riddle DL, Beckett LA. The reliability of Maitland's irritability judgments in patients with low back pain. J Man Manipulative Ther 2009;17:135-40 [Epub 2010/01/05]. 25. Bedi N, Chilvers C, Churchill R, et al. Assessing effectiveness of treatment of depression in primary care. Partially randomised preference trial. Br J Psychiatry 2000;177:312-8 [Epub 2000/12/16].

26. Lin P, Campbell DG, Chaney EF, et al. The influence of patient preference on depression treatment in primary care. Ann Behav Med 2005;30:164-73 [Epub 2005/09/22]. 27. Myers SS, Phillips RS, Davis RB, et al. Patient expectations as predictors of outcome in patients with acute low back pain. J Gen Intern Med 2008;23:148-53. 28. Carroll LJ, Holm LW, Ferrari R, Ozegovic D, Cassidy JD. Recovery in whiplash-associated disorders: do you get what you expect? J Rheumatol 2009;36:1063-70 [Epub 2009/02/21]. 29. Schneider MJ, Brach J, Irrgang JJ, Abbott KV, Wisniewski SR, Delitto A. Mechanical vs. manual manipulation for low back pain: an observational cohort study. J Manipulative Physiol Ther 2010;33:193-200. 30. Smeets RJ, Beelen S, Goossens ME, Schouten EG, Knottnerus A, Vlaeyen JW. Treatment expectancy and credibility are associated with the outcome of both physical and cognitive behavioral treatment in chronic low back pain. Clin J Pain 2008;24:305-15. 31. Denberg TD, Melhado TV, Steiner JF. Patient treatment preferences in localized prostate carcinoma: the influence of emotion, misconception, and anecdote. Cancer 2006;107: 620-30 [Epub 2006/06/28]. 32. Mazur DJ, Hickam DH. The effect of physician's explanations on patients' treatment preferences: five-year survival data. Med Decis Making 1994;14:255-8 [Epub 1994/07/01]. 33. Weiner B. Matching patient specific expectations with physician-specific outcomes: toward transparency in treatment decision-making and consent. Med Hypotheses 2007; 68:1287-91.