Transcript
Page 1: Clinical outcomes measurement: Classification of mild, moderate and severe pain due to diabetic peripheral neuropathy based on levels of functional disability

(1022) Impact of worker‘s compensation status on pain re-porting in multidisciplinary pain treatment

S. Stanos, T. Houle, T. Remble, R. Harden; Rehabilitation Institute of Chicago,Chicago, ILMuch attention has been paid to the effect of compensation on thepresentation and outcome of treatment in chronic pain. Findings havebeen mixed, but most authors have agreed that compensation statusshould be included as a relevant factor in the research of chronic paintreatment. Recently, Suter (2002) found that litigation and workerscompensation status predicted outcome over time. The present studyexpands on previous research by examining the effect of worker‘s com-pensation (WC) status on pain reporting while patients are actuallyundergoing multidisciplinary treatment. Participants were 618 chronicpain patients (180 WC, 438 not WC) presenting for treatment at a mul-tidisciplinary pain treatment center. Treatment was multidisciplinaryand lasted for 4 weeks. The McGill Pain Questionnaire (MPQ) was ad-ministered at evaluation, during treatment weeks 1- 4, and at discharge.At evaluation, the non-WC group was significantly (p‘s � .05) moreeducated (14.3 vs. 12.9 years), older (45.4 vs. 42.0 years), had a higherproportion of females (49% vs 13%), and had pain for a greater dura-tion (54.0 vs. 32.3 months). The WC group presented with greater levelsof pain, as measured by the VAS of the MPQ, and responded to treat-ment differently than non WC. A 2 6 ANOVA found that reported painin both groups significantly decreased over time (p � .001). However, asignificant Group Time interaction revealed that the WC group re-ported less decreases in pain than non-WC, particularly late in treatment(Overall decrease in pain of 18% vs. 34%). The results support the notionthat the workers compensation system adversely affects pain reporteven during rehabilitation.

(1023) Development and psychometric properties of the Post-operative Activity Scale

M. Good, S. Huang, X. Cong, Y. Seo; Frances Payne Bolton School of Nursing,Case Western Reserve University, Cleveland, OHThe purpose was to report the psychometric properties of the postop-erative activity scale (PAS) to quantify the amount of activity in patientsafter major abdominal surgery. Measuring postoperative activity maycontribute to a theory of activity in relation to recovery (Redeker &Wykpisz, 1999). The PAS measures the intensity, duration and frequencyof six ranked postoperative activities according to distance (.5-5): dan-gling, and walking: to chair, in room, to bathroom, in hall (short) and inhall (long). Patients (N � 238) reported their activities in terms of dis-tance and frequency four times daily. Scoring was by multiplying thedistance by the frequency, and summing at each time point (possible �0 - 46.5). The sample was aged M � 48�11 years, 78% female, 63%white, and 68% with some college. Using repeated measures ANOVA,criteria-related validity of PAS was supported by significant increases inmeans across each day, and from day 1 to day 2, p� .001. Further supportwas found in correlations showing that patients with more activity hadless pain, r � .15 to .26, were younger, r � .15 to .28, and had shortersurgery, r �.20 to .29. Test-retest reliability was supported by significantpositive correlations on both days, r � .48 to .65, that were stronger atadjacent points. Postoperative activity would be expected to increase bytime, thus supporting the reliability of the PAS. The PAS can be used bynursing researchers to study relationships between postoperative activ-ity, pain, sleep, immune factors, and recovery. Supported by the Na-tional Institute of Nursing Research, RO1 NR03933 (2001-2005), M. Good

(1024) A latent growth analysis of a multilevel model of emo-tional factors related to treatment outcome

P. Davis, J. Reeves II, B. Naliboff; Cedars Sinai Medical Center, Los Angeles, CAAnxiety and depression are related to pain adaptation, however amodel that describes their differential effect has not been demon-strated. Six hundred consecutive patient referrals to the Pain Center atCedars-Sinai Medical Center were classified into one the following fourdiagnostic categories: Myofascial Pain Syndrome (n�109); NeuropathicPain (n�51), Neurovascular Pain (n�228), and unclassified (n�2). Themean duration of pain, was 146 months. Ages ranged from 16 to 87 witha mean of 45 years. Prior to receiving any treatment, patients received apacket of instruments and forms requesting demographic data. Theyensured that each participant was alone and did not receive any assis-tance that might bias their responses. The evaluation process involvednested models that are similar to the authors’ previous (Davis et al,2000), in that all sources of error and sample specific characteristics havebeen isolated. In this sample, a Monte Carlo Bootstrapping procedurewas used to give an indication of the stability of model parameters in200 randomly generated groups, a unique strength of AMOS The suc-cessful model confirmed previous findings that pain and chronic illnessdo not directly affect emotional levels, but causation derives from de-pression and anxiety. When emotional variables were placed in a medi-ating position between pain and PI/SO, the model did not converge,confirming that anxiety and depression are exogenous constructs. Painmeasures were linear combinations of VAS-P, VAS-H and VAS-FL, ar-ranged in a “Curve of Factors” design The present study confirmed thepredominance of anxiety in determining outcome to medical proce-dures in an migraine and orofacial pain population. The influence ofdepression was differential, mediated by variables of pain impact andsomatization, though both emotional variables were equally determi-native of initial status on pain measures. The study confirmed that de-pression and anxiety determine other variables in a causal mechanism.

(1025) Classification of mild, moderate and severe pain due todiabetic peripheral neuropathy based on levels offunctional disability

D. Zelman, E. Dukes, N. Brandenburg, A. Bostrom, M. Gore; Avalon HealthSolutions, Inc., Philadelphia, PAThis study identified discrete categories of pain severity in a sample ofpatients with painful diabetic peripheral neuropathy (DPN), throughderivation of cut-points on a 0-10 scale of pain severity (modified BriefPain Inventory-DPN, m-BPI-DPN). Cut-points are important for charac-terizing patients and outcomes and address nonlinearity in patients’interpretation of numeric pain severity scales. We adapted the methodof Serlin and colleagues that established cut-points for cancer pain se-verity based on interference with function. Since neuropathic pain is aunique condition, replication is appropriate. Subjects were participantsin a burden of illness survey (N�255). Average and Worst Pain (m-BPI-DPN) were 5.0(SD�2.5) and 5.6(SD�2.8). We considered all possible cut-points between 4 and 8. Optimal cut-points were those that createdthree pain severity categories producing maximum between-categorydifferences on the seven m-BPI-DPN Interference items, using MANOVA.Cut-points of 4 and 7 optimally classified the sample, creating categoriesof �4, 4 to �7 and 7 and higher, (Hotelling Lawley trace F � 22.95 and16.20 for Worst and Average Pain, p�.0001). Those reporting pain �4characterized their pain over the past month as “mild” or “very mild,”those with pain 4 to �7 generally classified pain as “moderate,” andthose �/� 7 generally classified pain as “severe,” confirming the termsmild, moderate and severe for the cut-point-derived categories. Meanm-BPI-DPN Interference was 2.1(SD�2.1), 4.9(SD�1.9) and 7.4(SD�1.6)for the mild, moderate and severe pain categories. Patients in the threecategories differed on patient-rated outcomes (the SF-12v2, HADS,MOS-Sleep, p�.001), medication satisfaction (p�.01), and on meanDPN-related healthcare visits over the past three months (p�.01):2.26(SD�3.43), 3.38(SD�7.86) and 5.53(SD�7.04). This research showsthat three categories of DPN pain severity can be identified based oninterference with daily function, and these categories influence patientoutcomes and medical utilization. Research was supported by Pfizer, Inc.

114 Abstracts