Advertisement

Preliminary Results of Patient-Defined Success Criteria for Individuals With Musculoskeletal Pain in Outpatient Physical Therapy Settings

Published:January 27, 2012DOI:https://doi.org/10.1016/j.apmr.2011.10.007

      Abstract

      Zeppieri G Jr, Lentz TA, Atchison JW, Indelicato PA, Moser MW, Vincent KR, George SZ. Preliminary results of patient-defined success criteria for individuals with musculoskeletal pain in outpatient physical therapy settings.

      Objectives

      (1) To investigate patient-defined parameters of treatment success in an outpatient physical therapy setting with musculoskeletal pain, (2) to determine whether patient-defined treatment success was influenced by selected demographic and clinical factors, and (3) to examine whether patient subgroups existed for ratings of importance for each treatment outcome domain.

      Design

      Cross-sectional study.

      Setting

      Outpatient physical therapy clinic.

      Participants

      Consecutive patients (N=110) with complaints of musculoskeletal pain.

      Interventions

      Not applicable.

      Main Outcome Measure

      We reported patient-defined treatment success targets for pain, fatigue, emotional distress, and interference with daily activities using the Patient-Centered Outcomes Questionnaire (PCOQ). We also investigated whether patient subgroups existed based on perceived importance of improvement for these same outcome domains.

      Results

      Patient-defined criteria for treatment success included mean reductions (from baseline scores) in pain of 3.0 points, in fatigue of 2.3 points, in emotional distress of 1.4 points, and in interference with daily activities of 3.4 points. There were no differences in patient-defined criteria for treatment success based on sex, age, postoperative rehabilitation, prior physical therapy, other prior health care interventions, duration of symptoms, and anatomical location of symptoms (P>.01). Cluster analysis of the PCOQ importance ratings indicated a 2-cluster solution. The multifocused subgroup demonstrated higher importance for improvement ratings in each treatment outcome domain when compared with the pain-focused subgroup (P>.05).

      Conclusions

      These data indicate that patient-defined criteria for treatment success required greater reductions in the studied outcome domains to be considered successful. These data suggest the potential existence of patient subgroups that either rate improvement in all outcome domains as important or rate pain relief as the most important outcome.

      Key Words

      List of Abbreviations:

      ANOVA (analysis of variance), MCID (minimal clinically important difference), PCOQ (Patient-Centered Outcomes Questionnaire)
      A STANDARD APPROACH of determining treatment effect is to use statistical inferential testing. In this approach, statistical models are used to determine if a statistical threshold has been satisfied. Determining statistical significance provides the ability to detect a change, but the inherent flaw to this type of analysis is that change scores indicate only that the change could be reliably detected in comparison with associated error.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      • Beaton D.E.
      • Bombardier C.
      • Katz J.N.
      • Wright J.G.
      A taxonomy for responsiveness.
      • Fritz J.M.
      • Hebert J.
      • Koppenhaver S.
      • Parent E.
      Beyond minimally important change: defining a successful outcome of physical therapy for patients with low back pain.
      Therefore, these changes may not be indicative of a clinically meaningful or important change.
      • Beaton D.E.
      • Bombardier C.
      • Katz J.N.
      • Wright J.G.
      A taxonomy for responsiveness.
      Another way to measure treatment effect is by comparison with established minimal clinically important difference (MCID) scores. MCIDs, according to Jaeschke et al,
      • Jaeschke R.
      • Singer J.
      • Guyatt G.H.
      Measurement of health status Ascertaining the minimal clinically important difference.
      are defined as “the smallest difference in a score in a domain of interest that is perceived as beneficial and 1 that would mandate, in the absence of side effects a change in the patient's management.”
      • Jaeschke R.
      • Singer J.
      • Guyatt G.H.
      Measurement of health status Ascertaining the minimal clinically important difference.
      (p407-15) Treatment outcomes that exceed MCID scores imply that clinical significance has occurred owing to a comparison with an external criterion. MCIDs then are not based on comparison with associated error, but instead are related to a minimum amount of change that would be expected to reflect meaningful change in clinically relevant external criteria.
      • Fritz J.M.
      • Hebert J.
      • Koppenhaver S.
      • Parent E.
      Beyond minimally important change: defining a successful outcome of physical therapy for patients with low back pain.
      • Hajiro T.
      • Nishimura K.
      Minimal clinically significant difference in health status: the thorny path of health status measures?.
      • Hays R.D.
      • Woolley J.M.
      The concept of clinically meaningful difference in health-related quality-of-life research How meaningful is it?.
      • Wolfe F.
      • Michaud K.
      • Strand V.
      Expanding the definition of clinical differences: from minimally clinically important differences to really important differences Analysis in 8931 patients with rheumatoid arthritis.
      • Farrar J.
      • Portenoy R.
      • Berlin J.
      • Kinman J.
      • Strom B.
      Defining the clinically important difference in pain outcome measures.
      Despite the utility of statistical-based and criteria-based methods of evaluating change, neither incorporates the individual patient's perspective in defining his/her treatment success.
      • Ostelo R.W.
      • Deyo R.A.
      • Stratford P.
      • et al.
      Interpreting change scores for pain and functional status in low back pain: toward international consensus regarding minimal important change.
      • Gross D.P.
      • Ferrari R.
      • Russell A.S.
      • et al.
      A population survey of back pain beliefs in Canada.
      Consequently, there have been numerous attempts to develop patient-centered outcome measures that consider success, expectation, desire, and importance of therapy-related interventions.
      • Cook C.E.
      The minimally clinically important change score (MCID); a necessary pretense.
      Patient-centered methods are readily available, but success of treatment episodes is more commonly determined by statistical methods or by MCID scores. In light of this, patient-centered views instead of purely arithmetical views of treatment success have been advocated.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      • Beaton D.E.
      • Bombardier C.
      • Katz J.N.
      • Wright J.G.
      A taxonomy for responsiveness.
      • Fritz J.M.
      • Hebert J.
      • Koppenhaver S.
      • Parent E.
      Beyond minimally important change: defining a successful outcome of physical therapy for patients with low back pain.
      • Ostelo R.W.
      • Deyo R.A.
      • Stratford P.
      • et al.
      Interpreting change scores for pain and functional status in low back pain: toward international consensus regarding minimal important change.
      • Gross D.P.
      • Ferrari R.
      • Russell A.S.
      • et al.
      A population survey of back pain beliefs in Canada.
      • Cook C.E.
      The minimally clinically important change score (MCID); a necessary pretense.
      • Casarett D.
      • Karlawish J.
      • Sankar P.
      • Hirschman K.
      • Asch D.A.
      Designing pain research from the patient's perspective: what trial end points are important to patients with chronic pain?.
      • Brown J.L.
      • Edwards P.S.
      • Atchison J.W.
      • Lucey-Lafayette A.
      • Wittmer V.T.
      • Robinson M.E.
      Defining patient-centered multidimensional success criteria for treatment of chronic spine pain.
      • O'Brien E.M.
      • Staud R.M.
      • Hassinger A.D.
      • et al.
      Patient-centered perspective on treatment outcomes in chronic pain.
      • Vowles K.E.
      • Robinson M.E.
      Progressing towards acceptable treatment outcomes.
      • Daniel M.
      • Long C.
      • Murphy W.
      • Kores R.
      • Hutcherson W.
      Therapist's and chronic pain patient's perceptions of treatment outcome.
      • Liang M.H.
      Longitudinal construct validity: establishment of clinical meaning in patient evaluative instruments.
      • Morley S.
      Efficacy and effectiveness of cognitive behaviour therapy for chronic pain: progress and some challenges.
      A patient-centered outcome can be defined operationally as any outcome that represents a treatment success based on the patient's self-determined criterion for success. Patient-centered outcomes are important to consider because outcomes are specific to the individual patient and may be influenced by a wide variety of factors, including the patient's past medical history.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      • Gross D.P.
      • Ferrari R.
      • Russell A.S.
      • et al.
      A population survey of back pain beliefs in Canada.
      • Vowles K.E.
      • Robinson M.E.
      Progressing towards acceptable treatment outcomes.
      • Daniel M.
      • Long C.
      • Murphy W.
      • Kores R.
      • Hutcherson W.
      Therapist's and chronic pain patient's perceptions of treatment outcome.
      In essence, if we look to consider patient outcomes only from statistical or MCID perspectives, we potentially ignore the individual needs of the patient.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      For example, a desired outcome might be defined by 1 patient as a measureable reduction in 1 domain, such as pain, while another individual might desire an improvement in a different domain, such as function.
      • Gross D.P.
      • Ferrari R.
      • Russell A.S.
      • et al.
      A population survey of back pain beliefs in Canada.
      • O'Brien E.M.
      • Staud R.M.
      • Hassinger A.D.
      • et al.
      Patient-centered perspective on treatment outcomes in chronic pain.
      • Vowles K.E.
      • Robinson M.E.
      Progressing towards acceptable treatment outcomes.
      Due to the diversity of various patient rehabilitation needs, it is imperative to consider patient-centered measures of outcome.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      One measure, which has been used in the chronic pain population, is the Patient-Centered Outcomes Questionnaire (PCOQ).
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      The PCOQ obtains the patient perspective of treatment goals across 4 constructs (pain, fatigue, emotional distress, and interference with daily activities).
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      Additionally, the questionnaire assesses how each of these areas has impacted the patient and what the patient definitions of expected, successful, and desired outcomes are after their rehabilitation treatment.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      The PCOQ has been investigated only in chronic pain settings, and the purpose of this article is to investigate whether the PCOQ has an application in outpatient physical therapy settings. The first objective was to identify patient-centered targets for success, expectation, and desired posttreatment levels for pain, fatigue, emotional distress, and interference with daily activities in an outpatient physical therapy population of individuals with musculoskeletal pain. Furthermore, we investigated whether these levels were dependent on duration of symptoms, prior physical therapy interventions, prior health care interventions, sex, anatomical location of symptoms, and if prior interventions were positive or not. The second objective was to investigate subgroups of the PCOQ based on importance of improvement in each of the treatment domains (pain, fatigue, emotional distress, and interference with daily activities). This analysis of potential patient subgroups would provide information on whether different patient subgroups exist based on the perceived importance of treatment domain improvement.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      Overall, this preliminary study has the potential to provide an initial report for definitions of treatment success from the perspective of the patient, as well as indicate what are the most important domains for improvement from the patients' perspective. This study also provides a basis on which to evaluate patient outcomes based on their criteria for success, where data generated from patient-centered outcomes from the PCOQ may provide possible clinical benchmarks and thus improve the quality of clinical practice.
      • Gross D.P.
      • Ferrari R.
      • Russell A.S.
      • et al.
      A population survey of back pain beliefs in Canada.
      • Morley S.
      Efficacy and effectiveness of cognitive behaviour therapy for chronic pain: progress and some challenges.

      Methods

      A consecutive sample was recruited from an outpatient physical therapy clinic that treats over 3700 patients annually and is affiliated with Shands Healthcare and the University of Florida. Patients are scheduled to a therapist by front desk staff by order of therapist availability, which provides therapists the opportunity to see an entire spectrum of potential patients and musculoskeletal pathologies. A convenience sample of 110 subjects (50 men, 60 women) referred to outpatient physical therapy over the course of 1 year with complaints of musculoskeletal pain and treated by a physical therapist (G.Z.) were included in the study. All patients who completed the PCOQ as part of routine intake procedures were eligible for this study. Patients were included in this analysis if they presented with musculoskeletal pathology that was appropriate for physical therapy and were able to read and speak English. No formal screening test was performed for English comprehension, and no patients were excluded based on this criterion.
      At the intake physical therapy session, patients completed the PCOQ prior to their evaluation session. In addition to the standard clinical intake questionnaire, selected demographic questions (appendix 1) were asked of patients during their subject evaluation. All patients treated at the Shands/University of Florida Orthopedic and Sports Medicine Institute sign a written consent approved by the University of Florida Institutional Review Board prior to initiating treatment. This global written consent allows for use of data from questionnaires related to injury that measure pain, functional outcomes, and psychological outcomes in a deidentified manner. Therefore, 15 patients were not required to provide specific informed consent for this study because all data were collected during routine clinical visits to outpatient physical therapy as part of normative care. The collected clinical data were then deidentified and entered into an orthopedic data bank before analysis.
      The PCOQ (appendix 2) comprises 5 sections with 4 domains (pain, fatigue, emotional distress, and interference of daily activities) for usual, desired, successful, and expected levels, as well as how important improvement is for each domain on an 11-point numeric rating scale (0=none, 10=worst imaginable).
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      In each section, patients are asked to rate their level of pain, fatigue, emotional distress, and interference of daily activities on an 11-point numeric rating scale for their usual level, their desired level, and what treatment outcome they would consider successful.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      The patients are then asked to identify what they expect after their physical therapy intervention and how important they consider an improvement in each of these domains.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.

      Statistical Analysis

      All analyses were performed with SPSS version 17.0a with an alpha equal to .01 due to the number of comparisons. Descriptive statistics were calculated for selected demographic and clinical factors. Normality of PCOQ domains was assessed with 1-sample Kolmogorov-Smirnov tests. Independent t tests, analysis of variance (ANOVA) models, or chi-square tests were used to determine differences in PCOQ domains based on available demographic and selected clinical factors. A hierarchical agglomerative cluster analysis was performed using Ward's clustering method with p squared Euclidean distances as the similarity measure to create distinct cluster profiles. Agglomeration coefficients were inspected; percent change between adjacent cluster solutions and plot characteristics were considered to be the most optimal solution between 2 and 4 clusters. Furthermore, demographic and clinical factors as well as subgroups identified by cluster analysis were examined to identify differences for usual ratings for pain, fatigue, emotional distress, and interference with daily activities using ANOVA models for continuous measures and chi-squared levels for categorical measures.

      Results

      Descriptive statistics for sample demographics are listed in table 1. The mean age ± SD of the 110 patients was 40.1±20.8 years (range, 11–89). Duration of musculoskeletal symptoms ranged from 1 week to 520 weeks (mean ± SD, 103.4±304.5).
      Table 1Demographic, Clinical, and PCOQ Data for the Total Sample and Cluster Subgroups
      VariableTotal Sample (N=110)Cluster 1 (pain-focused group; n=36)Cluster 2 (multifocused group; n=74)P
      Age, mean ± SD (y)40.2±20.741.2±23.739.2±19.3.518
      Sex.504
       Men50 (45.4)18 (50.0)32 (43.2)
       Women60 (54.5)18 (50.0)42 (56.8)
      Race.511
       White91 (82.7)30 (83.3)61 (82.4)
       Hispanic3 (2.7)2 (5.6)1 (1.4)
       Black15 (13.6)4 (11.1)11 (14.9)
       Asian1 (0.9)0 (0.0)1 (1.4)
      Duration, mean ± SD (wk)159.0±429.776.3±218.4.183
       Acute80 (72.7)23 (63.9)57 (77)
       Subacute16 (14.6)6 (16.7)10 (13.5)
       Chronic14 (12.7)7 (19.4)7 (9.5)
      Anatomical location.720
       Lower extremity64 (58.2)19 (52.8)45 (60.8)
       Spine32 (29.1)12 (33.3)20 (27.0)
       Upper extremity14 (12.7)5 (13.9)9 (12.2)
      Postoperative.267
       Yes48 (43.6)13 (36.1)35 (47.3)
       No62 (56.4)23 (63.9)39 (52.7)
      Prior physical therapy.344
       Yes56 (50.1)16 (44.4)40 (54.1)
       No54 (49.1)20 (55.6)34 (45.9)
      Other prior health care.814
       Yes32 (29.0)11 (30.6)21 (28.4)
       No78 (71.0)25 (69.4)53 (71.6)
      PCOQ – usual painNA4.7 (2.2)4.8 (2.6).857
      PCOQ – usual fatigueNA3.1 (2.9)4.3 (2.5).023
      PCOQ – usual emotional distressNA2.0 (2.4)3.0 (2.6).055
      PCOQ – usual interference with daily activitiesNA4.2 (3.2)5.2 (3.1).183
      NOTE. Values are no. (%) or as otherwise indicated. P values represent cluster subgroups results from ANOVA continuous and chi-square categorical variables.
      Abbreviation: NA, not applicable.
      Duration was nonnormally distributed, with skewness of 5.78 (SE=0.23) and kurtosis of 39.6 (SE=0.46). Of the 110 patients, 91 (82.7%) identified themselves as white of non-Hispanic origin, 15 (13.6%) identified themselves as black, 3 (2.7%) identified themselves as Hispanic, and 1 (0.9%) identified himself/herself as of Asian origin; 48 (43.6%) of the 110 patients presented to physical therapy after surgery.
      Additionally, 64 (58.2%) patients presented at initial examination with primary complaints of lower-extremity symptoms, 32 (29.1%) patients reported spine-related symptoms, and 14 (12.7%) patients reported upper-extremity symptoms. Fifty-six (50.9%) patients reported prior physical therapy treatments. These prior treatments were unrelated to current musculoskeletal episodes, with 89% reporting the prior treatment as successful. Thirty-two (29.1%) patients reported other prior health care interventions including massage (most frequent), chiropractic, steroid injections, acupuncture, and combination of these intervention strategies. Of these, 77% were reported as successful.
      Means and medians for all PCOQ domains are listed in table 2, and PCOQ means are displayed graphically in figure 1. All PCOQ domains were positively skewed, so data were reported as mean and median values. Inferential testing did not identify differences when nonparametric tests were used to test central tendency, so only results from the parametric tests are reported in this article. Participant-defined success criteria included mean reductions (calculated by usual rating-successful rating) in pain intensity of 3.0 points, in fatigue of 2.3 points, in emotional distress of 1.4 points, and in interference with activities of daily living of 3.4 points.
      Table 2PCOQ Domain Distributions for Total Data Sample
      PCOQ Domain VariablesMean ± SDMedian (interquartile range)
      Usual levels
       Pain4.7±2.45.0 (3.0–7.0)
       Fatigue3.9±2.74.0 (2.0–6.0)
       Emotional distress2.7±2.62.0 (0.0–4.0)
       Interference with daily activities4.9±3.25.0 (2.0–8.0)
      Desired levels
       Pain.93±1.70.0 (0.0–2.0)
       Fatigue.83±1.40.0 (0.0–1.25)
       Emotional distress.56±1.30.0 (0.0–1.0)
       Interference with daily activities.55±1.40.0 (0.0–0.0)
      Successful levels
       Pain1.7±1.61.5 (0.75–2.0)
       Fatigue1.6±1.41.0 (0.0–2.0)
       Emotional distress1.3±1.61.0 (0.0–2.0)
       Interference with daily activities1.5±1.71.0 (0.0–2.0)
      Expected levels
       Pain1.5±2.01.0 (0.0–2.0)
       Fatigue1.4±1.71.0 (0.0–2.25)
       Emotional distress1.1±1.50.0 (0.0–1.0)
       Interference with daily activities6.4±3.60.0 (0.0–2.0)
      Important levels
       Pain8.4±2.610 (8.0–10.0)
       Fatigue6.4±3.67.0 (4.0–10.0)
       Emotional distress5.7±4.06.0 (2.0–10.0)
       Interference with daily activities7.7±3.410 (6.75–10.0)
      Figure thumbnail gr1
      Fig 1PCOQ domain means for patients seeking outpatient physical therapy.
      Differences in PCOQ domain scores were assessed for sex, postoperative rehabilitation, prior physical therapy intervention, and other prior health care. All tests revealed a lack of statistical significance across all PCOQ domains (P>.01), indicating that desired and successful treatment outcomes, treatment expectations, and overall importance of treatment improvement were similar among sex, postoperative rehabilitation, prior physical therapy intervention, and other health care intervention.
      Differences in PCOQ domain scores for age, duration of symptoms (initial onset), and anatomical location of symptoms (medical diagnosis) were investigated with ANOVA. Results indicated no differences (P>.01) among age, duration of symptoms, and anatomical location of symptoms for PCOQ domains, further suggesting that desired and successful treatment outcomes, treatment expectations, and overall importance of treatment improvement were not influenced by these factors.
      The hierarchical agglomerative cluster analysis (Ward's method, squared Euclidian distance) of importance ratings across the 4 treatment domains indicated potential cluster solutions for 2, 3, and 4 subgroups. Inspection of agglomeration coefficients revealed that the percent change was small between the 3-cluster and 2-cluster, suggesting that a 2-cluster solution is appropriate, which was further confirmed by visual inspection of plotted agglomeration coefficients. Therefore, only the data for the 2 subgroups solution are presented in this article. The first cluster was identified as the multifocused subgroup. This subgroup rated all domains with high importance ratings (mean range, 7.4–9.2) that did not differ statistically (P>.01) within the cluster. The second cluster was identified as the pain-focused subgroup. This subgroup rated the importance of pain relief (mean. 6.9) higher (P<.01) than the other outcome domains (mean range, 1.4–4.4) within the cluster. Further investigation of the clusters indicated that the multifocused subgroup had higher importance ratings for all domains in comparison with the pain-focused subgroup. Figure 2 provides a graphical depiction of importance ratings for the 2 subgroups identified in this study.
      Figure thumbnail gr2
      Fig 2Cluster subgroups based on importance of PCOQ domains. Abbreviation: ADLs, activities of daily living.
      The cluster subgroups were then investigated by ANOVA (continuous data) and chi-square (categorical data) for differences in demographic and clinical variables, as well as PCOQ ratings for usual pain, fatigue, and emotional distress; there were no differences in the subgroups for any of the demographic and clinical variables (P>.01). The PCOQ ratings showed potential for differences in the fatigue and emotional distress domains. The multifocused subgroup had slightly higher ratings in these domains (P<.05), but not at the type 1 error rate used for this study. The descriptive data for the cluster subgroup comparisons are reported in table 1.

      Discussion

      This study was a preliminary investigation of whether the PCOQ, a measure developed for use with chronic pain populations, has a potential application in outpatient physical therapy settings for patients with musculoskeletal pain. The ability to determine patient-centered successful outcomes may facilitate designing appropriate treatment interventions and be used as a method for defining thresholds of successful treatment outcomes. This study provided a preliminary indication of what successful outcomes may be for patients seeking outpatient physical therapy, which did not vary based on demographic or clinical factors. This study also provided information on the importance of improvement in each domain, indicating there are 2 potential subgroups of patients with musculoskeletal pain: those that are focused on the importance of pain relief, and those that desire relief in all treatment domains. These data may be helpful in providing an insight into the patient's perspective so that physical therapy interventions can be used to guide rehabilitation in the attainment of health care goals.
      Our results indicated that prior physical therapy interventions, other health care interactions (whether positive or negative), history of surgical interventions, and anatomical location of symptoms were not associated with perceived successful outcomes. Particularly interesting was that previous treatment experiences were not associated with the patient's expectation of a successful outcome. This result seems counterintuitive because one would speculate that a negative experience with a prior physical therapy encounter would tend to bias expectations and perceptions of success. However, the results implied that current treatment expectations were not influenced by past experiences, although this may be an artifact of selection bias, because patients who were unsatisfied with previous negative physical therapy experiences may be less likely to return for physical therapy.
      Additionally, expectations were not associated with anatomical distribution of symptoms. The clinical assumption that patient expectations differ among patients with spine, lower-extremity, or upper-extremity pathology was not supported by our data. This is an interesting finding because there are population data to suggest that negative beliefs about specific anatomical pain are common; however our sample size may not have been adequately powered to represent this trend. Additionally, our outpatient population may not have presented with the same levels of emotional distress that are commonly reported in a chronic pain setting.
      Therefore, the PCOQ may not be sensitive enough to detect changes in emotional distress because of the overall low levels presented in this sample, despite overall reductions reported by this sample in this domain.
      This study also corroborates Robinson et al's
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      work in chronic pain settings, identifying the importance of patient expectations across multiple success domains.
      Clinically, patients describe physical health, emotional well-being, societal acceptance, and return to work as important outcomes when identifying treatment success. Cluster analyses indicated that patient subgroups may exist in physical therapy settings that are very similar to the pain-focused or multifocused groups previously described.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      Interestingly, the multifocused subgroup had higher pain ratings for all domains, suggesting that this subgroup not only is multifocused, but rates all domains with high importance. Our findings suggest that physical therapists may consider modification of treatment plans specific to these patient subgroups.
      • Robinson M.E.
      • Brown J.L.
      • George S.Z.
      • et al.
      Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
      For example, patients wanting improvements in domains not explicitly addressed by physical therapy (eg, fatigue, emotional distress) may be identified early and considered for comanagement with other health care providers. There may also be benefit in identifying patients that have pain-focused beliefs, especially if pain relief is not a primary goal of the ongoing therapy (eg. focused on reducing other impairments or increasing activities).
      Future research should include longitudinal studies to determine the extent of changes in pain, fatigue, emotional distress, and interference of daily activities throughout the course of physical therapy in relation to the patient-determined thresholds for success. The results of those findings may indicate that patients undergo an alteration of expectations during the intervention process, similar to what has been reported in the chronic pain setting.
      • Brown J.L.
      • Edwards P.S.
      • Atchison J.W.
      • Lucey-Lafayette A.
      • Wittmer V.T.
      • Robinson M.E.
      Defining patient-centered multidimensional success criteria for treatment of chronic spine pain.
      Moreover, future investigation into the effects of matching physical therapy treatments based on the patient's reported outcome importance ratings may yield new insight into decreasing treatment morbidity and increasing the rate of successful treatment outcomes. Additionally, future research might also include validity and reliability of the PCOQ in outpatient physical therapy, including examination of the composition of the PCOQ (ie, domains) and its measurement properties.

      Study Limitations

      This study has limitations, which should be considered when interpreting the results. One limitation is that the PCOQ psychometric properties have only been validated in a chronic pain setting and not in an outpatient physical therapy setting with individuals with musculoskeletal pain; therefore, its exact measurement properties are unknown in this population. Another limitation to this study includes the potential influence of 1 domain by another. For example, the subject may subconsciously rate the expectation of a decrease in 1 domain high because there may be a carryover in his/her perception of a successful treatment outcome in his/her actual target domain. This may lead to an allocation of resources to decrease a domain of diminished importance. Moreover, the domains of the PCOQ are predefined and are not comprehensive. The patient-centered part of the PCOQ is the rating of expected and desired outcomes, as well as the importance of the levels of improvement, and not the selection of the domains. Furthermore, due to the cross-sectional nature of this study, we could not address how patients' perceptions changed over time. The PCOQ was completed prior to any physical therapy intervention, and therefore we do not know if patients' expectations or success criteria changed in response to therapy. In addition, this study should not be interpreted as a validity study owing to the lack of an external criterion; however, it should be considered as a preliminary description of the PCOQ applicability in the outpatient physical therapy setting. Additionally, the data gathered for this convenience sample of patients are not representative of individuals seeking physical therapy for musculoskeletal pain from all outpatient clinics. Additionally the definition of desired, expected, or successful levels of PCOQ domains did not address time to attainment.

      Conclusions

      This study was the first we are aware of to investigate the PCOQ in an outpatient physical therapy setting. These data provided patient-determined levels of treatment success across 4 clinically relevant outcome domains for musculoskeletal pain. Patient-defined success levels were not associated with the demographic or clinical variables used in this study. Our findings provide a preliminary indication that patients may require a greater reduction in perceived outcome expectations than what has been previously reported as clinically meaningful in order for them to consider their physical therapy outcomes successful. Lastly, investigation of patient-defined importance of improvement reveals the potential for 2 subgroups in this area. One subgroup rated high importance for all treatment domains, while the other rated lower importance ratings, with the highest rating for pain relief.
      • a
        SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

      APPENDIX 1. Supplemental Therapist Demographic Question List

      Today's date __________
      Age_____
      Sex _____
      Race _____
      Date of initial onset of symptoms (current problem) ________
      Current problem (medical diagnosis) _________
      Have you had prior physical therapy for your current condition? Yes No
      Have you had physical therapy for any other past condition? Yes No
      If so, please state the condition? __________
      Describe the result of the prior physical therapy intervention? Positive Negative
      Have you had any other prior health care intervention for your current condition (ie, chiropractor, acupuncture, massage, steroidal injection, or any multiple combination)? Yes No
      If so, please state the intervention? __________
      Describe the result of the prior health care intervention? Positive Negative
      Please state your referring physician? _________

      APPENDIX 2. Patient-Centered Outcomes Questionnaire

      MANY PEOPLE EXPERIENCE PAIN, FATIGUE (IE, FEELING TIRED), EMOTIONAL DISTRESS (EG, WORRIES, FEELING SAD), AND INTERFERENCE WITH DAILY ACTIVITIES (EG, NOT BEING ABLE TO WORK OR DO HOUSEHOLD CHORES) AS A RESULT OF THEIR MEDICAL CONDITION. WE WOULD LIKE TO UNDERSTAND HOW YOU HAVE BEEN IMPACTED IN EACH OF THESE AREAS. WE WOULD ALSO LIKE TO LEARN MORE ABOUT WHAT YOU WANT YOUR TREATMENT TO DO FOR YOU.
      _______________________________________________
      FIRST, WE WOULD LIKE TO KNOW YOUR USUAL LEVELS OF PAIN, FATIGUE, EMOTIONAL DISTRESS, AND INTERFERENCE.
      On a scale of 0 (none) to 10 (worst imaginable), please indicate your usual level (during the past week) of …
      • Pain ______
      • Fatigue (or tiredness) ______
      • Emotional distress ______
      • Interference with daily activities __________
      • ______________________________________________
      NOW, WE WOULD LIKE TO LEARN ABOUT YOUR DESIRED LEVELS OF PAIN, FATIGUE, EMOTIONAL DISTRESS, AND INTERFERENCE. IN OTHER WORDS, WE WOULD LIKE TO UNDERSTAND WHAT YOUR IDEAL TREATMENT OUTCOME WOULD BE.
      On a scale of 0 (none) to 10 (worst imaginable), please indicate your desired level of …
      • Pain ______
      • Fatigue (or tiredness) ______
      • Emotional distress ______
      • Interference with daily activities __________
      • ______________________________________________
      PATIENTS UNDERSTANDABLY WANT THEIR TREATMENT TO RESULT IN DESIRED OR IDEAL OUTCOMES LIKE YOU INDICATED ABOVE. UNFORTUNATELY, AVAILABLE TREATMENTS DO NOT ALWAYS PRODUCE DESIRED OUTCOMES. THEREFORE, IT IS IMPORTANT FOR US TO UNDERSTAND WHAT TREATMENT OUTCOMES YOU WOULD CONSIDER SUCCESSFUL.
      On a scale of 0 (none) to 10 (worst imaginable), please indicate the level each of these areas would have to be at for you to consider treatment successful.
      • Pain ______
      • Fatigue (or tiredness) ______
      • Emotional distress ______
      • Interference with daily activities _________
      • ______________________________________________
      NOW, WE WOULD LIKE TO KNOW WHAT YOU EXPECT YOUR TREATMENT TO DO FOR YOU.
      On a scale of 0 (none) to 10 (worst imaginable), please indicate the levels you expect after treatment.
      • Pain ______
      • Fatigue (or tiredness) ______
      • Emotional distress ______
      • Interference with daily activities _________
      • ______________________________________________
      FINALLY, WE WOULD LIKE TO UNDERSTAND HOW IMPORTANT IT IS FOR YOU TO SEE IMPROVEMENT IN YOUR PAIN, FATIGUE, EMOTIONAL DISTRESS, AND INTERFERENCE AFTER TREATMENT.
      On a scale of 0 (not at all important) to 10 (most important), please indicate how important it is for you to see improvement in your …
      • Pain ______
      • Fatigue (or tiredness) ______
      • Emotional distress ______
      • Interference with daily activities ______

      References

        • Robinson M.E.
        • Brown J.L.
        • George S.Z.
        • et al.
        Multidimensional success criteria and expectation for treatment of chronic pain: the patient's perspective.
        Pain Med. 2005; 6: 336-345
        • Beaton D.E.
        • Bombardier C.
        • Katz J.N.
        • Wright J.G.
        A taxonomy for responsiveness.
        J Clin Epidemiol. 2001; 54: 1204-1217
        • Fritz J.M.
        • 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: 2803-2809
        • Jaeschke R.
        • Singer J.
        • Guyatt G.H.
        Measurement of health status.
        Control Clin Trials. 1989; 10: 407-415
        • Hajiro T.
        • Nishimura K.
        Minimal clinically significant difference in health status: the thorny path of health status measures?.
        Eur Respir J. 2002; 19: 390-391
        • Hays R.D.
        • Woolley J.M.
        The concept of clinically meaningful difference in health-related quality-of-life research.
        Pharmacoeconomics. 2000; 18: 419-423
        • Wolfe F.
        • Michaud K.
        • Strand V.
        Expanding the definition of clinical differences: from minimally clinically important differences to really important differences.
        J Rheum. 2005; 32: 5835-5889
        • Farrar J.
        • Portenoy R.
        • Berlin J.
        • Kinman J.
        • Strom B.
        Defining the clinically important difference in pain outcome measures.
        Pain. 2000; 88: 287-294
        • Ostelo R.W.
        • Deyo R.A.
        • Stratford P.
        • et al.
        Interpreting change scores for pain and functional status in low back pain: toward international consensus regarding minimal important change.
        Spine. 2008; 33: 90-94
        • Gross D.P.
        • Ferrari R.
        • Russell A.S.
        • et al.
        A population survey of back pain beliefs in Canada.
        Spine. 2006; 31: 2142-2145
        • Cook C.E.
        The minimally clinically important change score (MCID); a necessary pretense.
        J Man Manip Ther. 2008; 16: E82-E83
        • Casarett D.
        • Karlawish J.
        • Sankar P.
        • Hirschman K.
        • Asch D.A.
        Designing pain research from the patient's perspective: what trial end points are important to patients with chronic pain?.
        Pain Med. 2001; 2: 309-316
        • Brown J.L.
        • Edwards P.S.
        • Atchison J.W.
        • Lucey-Lafayette A.
        • Wittmer V.T.
        • Robinson M.E.
        Defining patient-centered multidimensional success criteria for treatment of chronic spine pain.
        Pain Med. 2008; 9: 851-862
        • O'Brien E.M.
        • Staud R.M.
        • Hassinger A.D.
        • et al.
        Patient-centered perspective on treatment outcomes in chronic pain.
        Pain Med. 2010; 11: 6-15
        • Vowles K.E.
        • Robinson M.E.
        Progressing towards acceptable treatment outcomes.
        Pain. 2009; 144: 228-229
        • Daniel M.
        • Long C.
        • Murphy W.
        • Kores R.
        • Hutcherson W.
        Therapist's and chronic pain patient's perceptions of treatment outcome.
        J Nerv Mont Dis. 1983; 171: 729-733
        • Liang M.H.
        Longitudinal construct validity: establishment of clinical meaning in patient evaluative instruments.
        Med Care. 2000; 38: II84-II90
        • Morley S.
        Efficacy and effectiveness of cognitive behaviour therapy for chronic pain: progress and some challenges.
        Pain. 2011; 152: S99-S106