A Six Week Study Of The Pain Relieving Effects Of Celecoxib 200 Mg Twice Daily Compared To Tramadol 50 Mg Four Times Daily In Patients With Chronic Low Back Pain

• Sponsor: Pfizer's Upjohn has merged with Mylan to form Viatris Inc. (Industry)
• Overall Status: Completed
• CT.gov ID: NCT00662558
• Collaborator: (none)
• Enrollment: 802
• Locations: 60
• Arms: 2
• Duration (Months): 8
• Patients Per Site: 13.4
• Patients Per Site Per Month: 1.7

Study Details

Study Description

Brief Summary

To compare the analgesic effectiveness of celecoxib and tramadol in subjects with Chronic Low Back Pain measured by the Numerical Rating Scale (NRS-Pain) at Week 6

Condition or Disease	Intervention/Treatment	Phase
Low Back Pain	Drug: celecoxib; Drug: tramadol HCL	Phase 3

Study Design

• Study Type: Interventional
• Actual Enrollment: 802 participants
• Allocation: Randomized
• Intervention Model: Parallel Assignment
• Masking: Double (Participant, Investigator)
• Primary Purpose: Treatment
• Official Title: A Six Week Double-Blind, Randomized, Multicenter Comparison Study Of The Analgesic Effectiveness Of Celecoxib 200 Mg BID Compared To Tramadol Hydrochloride 50 Mg QID In Subjects With Chronic Low Back Pain
• Study Start Date: Jan 1, 2008
• Actual Primary Completion Date: Sep 1, 2008
• Actual Study Completion Date: Sep 1, 2008

Arms and Interventions

Arm	Intervention/Treatment
Experimental: celecoxib	Drug: celecoxib; 200 mg capsules BID for 6 weeks
Active Comparator: tramadol	Drug: tramadol HCL; 50 mg capsules QID for 6 weeks

Outcome Measures

Primary Outcome Measures

1. Treatment Responders Based on the Numerical Rating Scale-Pain (NRS-Pain) [Week 6 or Early Termination (ET)]

   A subject who met the following criteria was considered as a successful responder at Week 6: completed 6 weeks of treatment with study medication and had a 30% improvement from Baseline to Week 6/ET on the NRS-Pain. NRS-Pain scale assessed the severity of a subject's lower back pain on a scale of 0 (No pain) and 10 (Worst possible pain).

Secondary Outcome Measures

1. Change From Baseline in Severity of Chronic Low Back Pain as Measured by NRS-Pain [Baseline, Week 6/ET]

   NRS-Pain scale assessed the severity of a subject's lower back pain on a scale of 0 (No pain) and 10 (Worst possible pain). NRS-Pain scale: Change = mean score at Week 6/ET minus mean score at Baseline.

2. Change From Baseline in Severity of Low Back Pain as Measured by Visual Analogue Scale (VAS) [Baseline, Week 6/ET]

   VAS was a 100 millimeter (mm) scale that subjects used to assess the severity of their lower back pain. Based on the following question, "During the past day, how much back pain did you have?", the subject was instructed to place a vertical line on the VAS to indicate the magnitude of his/her lower back pain. 0 mm = no pain and 100 mm = worst possible pain. VAS: Change = mean score at Week 6/ET minus mean score at Baseline.

3. Patient's Global Assessment of Disease Activity [Week 6/ET]

   Number of subjects with a graded level of disease activity using the Patient's Global Assessment of Disease Activity 5-point scale (1=very good, 2=good, 3=fair, 4=poor, and 5=very poor). Subjects were classified as "Improved" if their assessment reduced at least 2 grades from baseline or if their assessment changed to Grade 1 (Very Good). Subjects were classified as "Worsened" if their assessment increased at least 2 grades from baseline or if their assessment changed to Grade 5 (Very Poor). Subjects were classified as "No Change" otherwise.

4. Physician's Global Assessment of Disease Activity [Week 6/ET]

   Number of subjects with a physician's grading of disease activity using the Physician's Global Assessment of Disease Activity 5-point scale ((1=very good, 2=good, 3=fair, 4=poor, and 5=very poor). Subjects were classified as "Improved" if their assessment reduced at least 2 grades from baseline or if their assessment changed to Grade 1 (Very Good). Subjects were classified as "Worsened" if their assessment increased at least 2 grades from baseline or if their assessment changed to Grade 5 (Very Poor). Subjects were classified as "No Change" otherwise.

5. Change From Baseline in Roland-Morris Disability Questionnaire (RMDQ) Total Score [Baseline, Week 6/ET]

   Each subject assessed his/her own disability due to low back pain using the RMDQ worksheet, which consisted of 24 statements of disability. The RMDQ total score was calculated as the total number of statements that were checked; the RMDQ total scores could have ranged from 0 to 24, with higher scores indicating greater disability. RMDQ: Change = mean score at Week 6/ET minus mean score at Baseline.

6. Change From Baseline in Modified Brief Pain Inventory (m-BPI-sf) [Baseline, Week 6/ET]

   m-BPI-sf scale assessed pain severity (0 = no pain to 10 = worst possible pain), and pain interference of functional activities (0 = does not interfere to 10 = completely interferes) during the 24 hour follow-up period. Subjects indicated: how much pain now; worst pain; average level of pain; how much pain interfered with general activity, mood, walking ability, relations with other people, sleep, normal work (including housework), and enjoyment of life. m-BPI-sf: Change = mean score at Week 6/ET minus mean score at Baseline.

7. Change From Baseline in Medical Outcomes Study (MOS) Sleep Scale [Baseline, Week 6/ET]

   MOS sleep scale included the following attributes: sleep disturbance, snoring, awaken shortness of breath or headache, quantity of sleep, sleep adequacy, somnolence, Sleep Problem Index I, and Sleep Problem Index II. Score ranged from 0-100, with a higher score indicating more of the scale attribute (e.g., more sleep disturbance, etc.). A negative change indicated subject improvement. MOS sleep scale: Change = mean score at Week 6/ET minus mean score at Baseline.

8. Number of Subjects With Change From Baseline in MOS Optimal Sleep Scale Scores [Baseline, Week 6/ET]

   The Optimal Scale is scaled from 0 or 1 with 1 indicating 7 or 8 hours of sleep per night and 0 otherwise. Number of subjects with change of improvement (0 to 1), no change (1 to 1 or 0 to 0), or worsening (1 to 0) from baseline as indicated by the MOS Optimal sleep scale.

9. Change From Baseline in Work Limitations Questionnaire (WLQ) [Baseline, Week 6/ET]

   The WLQ included the following: Time Scale, Physical Scale, Output Scale, Mental-Interpersonal Scale, and Index Scale. The scales ranged from 0 (Limited none of the time) to 100 (Limited all of the time). A negative change indicated subject improvement.

10. Patient's Global Evaluation of Study Medication [Weeks 1, 3, and 6/ET]

    Number of subjects with an overall response to study medication of poor, fair, good, very good, and excellent.

11. Patient's Satisfaction Questionnaire (With Pain Relief Scale) [Week 6/ET]

    Number of subjects at varying levels of pain relief (1 = very dissatisfied to 10 = very satisfied).

12. Patient's Satisfaction Questionnaire (With Walking and Bending Ability Scale) [Week 6/ET]

    Number of subjects at varying levels of pain relief (1 = very dissatisfied to 10 = very satisfied).

13. Chronic Low Back Pain Responders Based on VAS, Patient's Global, and RMDQ [Week 6/ET]

    Subjects were successful responders if they had: > = 30% improvement from baseline to final visit in VAS assessment (as identified by 100 millimeter scale); > = 30% improvement from baseline to final visit in Patient's Global assessment (classified as improved if assessment reduced at least 2 grades from baseline or if assessment changed to Grade 1, worsened if assessment increased at least 2 grades from baseline or if assessment changed to Grade 5, or no change; and < 20% worsening from baseline to final visit in RMDQ assessment (lower scores indicated greater disability).

Eligibility Criteria

Criteria

• Ages Eligible for Study: 18 Years and Older
• Sexes Eligible for Study: All
• Accepts Healthy Volunteers: No

Inclusion Criteria:

• The subject presents with duration of chronic low back pain of > 3 months requiring regular use of analgesics (> 4 days/week), except for acetaminophen which cannot have been the sole analgesic used

Exclusion Criteria:

• The subject has chronic low back pain, which is neurologic in etiology (i.e., radiculopathy, neuropathy, myelopathy)

Contacts and Locations

Locations

	Site	City	State	Country	Postal Code
1	Pfizer Investigational Site	Birmingham	Alabama	United States	35126
2	Pfizer Investigational Site	Birmingham	Alabama	United States	35235
3	Pfizer Investigational Site	Birmingham	Alabama	United States	35242
4	Pfizer Investigational Site	Phoenix	Arizona	United States	85023
5	Pfizer Investigational Site	Little Rock	Arkansas	United States	72205
6	Pfizer Investigational Site	Anaheim	California	United States	92801
7	Pfizer Investigational Site	Long Beach	California	United States	90806
8	Pfizer Investigational Site	Oceanside	California	United States	92056
9	Pfizer Investigational Site	Sacramento	California	United States	95823
10	Pfizer Investigational Site	Sacramento	California	United States	95825
11	Pfizer Investigational Site	Wildomar	California	United States	92595
12	Pfizer Investigational Site	Boulder	Colorado	United States	80304
13	Pfizer Investigational Site	Colorado Springs	Colorado	United States	80904
14	Pfizer Investigational Site	Denver	Colorado	United States	80220
15	Pfizer Investigational Site	Cos Cob	Connecticut	United States	06807
16	Pfizer Investigational Site	Jacksonville	Florida	United States	32216
17	Pfizer Investigational Site	Jacksonville	Florida	United States	32257
18	Pfizer Investigational Site	Pinellas Park	Florida	United States	33781
19	Pfizer Investigational Site	West Palm Beach	Florida	United States	33409
20	Pfizer Investigational Site	Woodstock	Georgia	United States	30189
21	Pfizer Investigational Site	Wichita	Kansas	United States	67206
22	Pfizer Investigational Site	Wichita	Kansas	United States	67214
23	Pfizer Investigational Site	Baton Rouge	Louisiana	United States	70809
24	Pfizer Investigational Site	Baltimore	Maryland	United States	21218
25	Pfizer Investigational Site	Columbia	Maryland	United States	21045
26	Pfizer Investigational Site	Rockville	Maryland	United States	20852
27	Pfizer Investigational Site	Wheaton	Maryland	United States	20902
28	Pfizer Investigational Site	Saint Paul	Minnesota	United States	55108
29	Pfizer Investigational Site	Jackson	Mississippi	United States	39202
30	Pfizer Investigational Site	Saint Louis	Missouri	United States	63141
31	Pfizer Investigational Site	Springfield	Missouri	United States	65807
32	Pfizer Investigational Site	Omaha	Nebraska	United States	68134
33	Pfizer Investigational Site	New Windsor	New York	United States	12553
34	Pfizer Investigational Site	New York	New York	United States	10022-1009
35	Pfizer Investigational Site	Rochester	New York	United States	14618
36	Pfizer Investigational Site	Williamsville	New York	United States	14221
37	Pfizer Investigational Site	Portland	Oregon	United States	97219
38	Pfizer Investigational Site	Bridgeville	Pennsylvania	United States	15017
39	Pfizer Investigational Site	Camp Hill	Pennsylvania	United States	17011
40	Pfizer Investigational Site	Columbia	South Carolina	United States	29204
41	Pfizer Investigational Site	North Charleston	South Carolina	United States	29406
42	Pfizer Investigational Site	Bristol	Tennessee	United States	37620
43	Pfizer Investigational Site	Collierville	Tennessee	United States	38017
44	Pfizer Investigational Site	Johnson City	Tennessee	United States	37601
45	Pfizer Investigational Site	Kingsport	Tennessee	United States	37660
46	Pfizer Investigational Site	New Tazewell	Tennessee	United States	37825
47	Pfizer Investigational Site	Austin	Texas	United States	78705
48	Pfizer Investigational Site	Beaumont	Texas	United States	77701
49	Pfizer Investigational Site	Beaumont	Texas	United States	77706
50	Pfizer Investigational Site	Dallas	Texas	United States	75230
51	Pfizer Investigational Site	Dallas	Texas	United States	75240
52	Pfizer Investigational Site	Grapevine	Texas	United States	76051
53	Pfizer Investigational Site	Houston	Texas	United States	77074
54	Pfizer Investigational Site	Lake Jackson	Texas	United States	77566
55	Pfizer Investigational Site	San Angelo	Texas	United States	76904
56	Pfizer Investigational Site	San Antonio	Texas	United States	78217
57	Pfizer Investigational Site	San Antonio	Texas	United States	78229
58	Pfizer Investigational Site	Salt Lake City	Utah	United States	84107
59	Pfizer Investigational Site	Richmond	Virginia	United States	23294
60	Pfizer Investigational Site	Weber City	Virginia	United States	24290

Sponsors and Collaborators

• Pfizer's Upjohn has merged with Mylan to form Viatris Inc.

Investigators

• Study Director: Pfizer CT.gov Call Center, Pfizer

Study Documents (Full-Text)

None provided.

More Information

Additional Information:

• To obtain contact information for a study center near you, click here.

Publications

None provided.

• Responsible Party: Pfizer's Upjohn has merged with Mylan to form Viatris Inc.
• ClinicalTrials.gov Identifier: NCT00662558
• Other Study ID Numbers: A3191338
• First Posted: Apr 21, 2008
• Last Update Posted: Feb 21, 2021
• Last Verified: Jan 1, 2021

• Keywords provided by Pfizer's Upjohn has merged with Mylan to form Viatris Inc.: Chronic low back pain
• Additional relevant MeSH terms: Back Pain; Low Back Pain; Celecoxib; Tramadol

Study Results

Participant Flow

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Period Title: Overall Study
STARTED	398	404
Received Treatment	396	396
COMPLETED	342	294
NOT COMPLETED	56	110

Baseline Characteristics

Arm/Group Title	Celecoxib	Tramadol HCL	Total
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks	Total of all reporting groups
Overall Participants	396	396	792
Age, Customized (participants) [Number]
18 to 44 years	164 41.4%	167 42.2%	331 41.8%
45 to 64 years	191 48.2%	182 46%	373 47.1%
> = 65 years	41 10.4%	47 11.9%	88 11.1%
Sex: Female, Male (Count of Participants)
Female	237 59.8%	213 53.8%	450 56.8%
Male	159 40.2%	183 46.2%	342 43.2%

Outcome Measures

1. Primary Outcome

Title	Treatment Responders Based on the Numerical Rating Scale-Pain (NRS-Pain)
Description	A subject who met the following criteria was considered as a successful responder at Week 6: completed 6 weeks of treatment with study medication and had a 30% improvement from Baseline to Week 6/ET on the NRS-Pain. NRS-Pain scale assessed the severity of a subject's lower back pain on a scale of 0 (No pain) and 10 (Worst possible pain).
Time Frame	Week 6 or Early Termination (ET)

Outcome Measure Data

Analysis Population Description

Intent-to-treat = all subjects who were randomized and received at least one dose of study medication. Missing values were imputed using last observation carried forward (LOCF).

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Responders	254 64.1%	218 55.1%
Non-responders	142 35.9%	178 44.9%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments
	Type of Statistical Test	Non-Inferiority or Equivalence (legacy)
	Comments	Differences in treatment proportions and the 95% confidence interval (CI) around the difference were estimated by calculating the risk difference between the treatment arms using a generalized linear model with treatment and center as factors. A lower 95% CI for the risk difference greater than -0.10 would demonstrate that celecoxib 200 mg BID is not inferior to tramadol hydrochloride 50 mg QID.
Statistical Test of Hypothesis	p-Value
	Comments
	Method
	Comments
Method of Estimation	Estimation Parameter	Risk Difference (RD)
	Estimated Value	0.0910
	Confidence Interval	() 95%; 0.0255 to 0.1565
	Parameter Dispersion	Type: Value:
	Estimation Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	If celecoxib 200 mg BID was found to be non-inferior to tramadol hydrochloride 50 mg QID then the second step was to test the superiority of celecoxib 200 mg BID over tramadol hydrochloride 50 mg QID using a two-sided test of proportions. Differences in proportions were tested using the General Association Test of the Cochran-Mantel-Haenszel (CMH) procedure stratified by center.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.008
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

2. Secondary Outcome

Title	Change From Baseline in Severity of Chronic Low Back Pain as Measured by NRS-Pain
Description	NRS-Pain scale assessed the severity of a subject's lower back pain on a scale of 0 (No pain) and 10 (Worst possible pain). NRS-Pain scale: Change = mean score at Week 6/ET minus mean score at Baseline.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Missing values were imputed by LOCF. Number of subjects with NRS-Pain scale scores at Baseline and Week 6/ET: celecoxib n=386, tramadol HCl n=385.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Mean (Standard Error) [scores on a scale]	-3.38 (0.12)	-3.25 (0.13)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	The change from Baseline was compared between the two treatment groups using analysis of covariance (ANCOVA), with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.234
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.20
	Confidence Interval	() 95%; -0.53 to 0.13
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.17
	Estimation Comments	The mean difference reported is the least squares (LS) mean difference.

3. Secondary Outcome

Title	Change From Baseline in Severity of Low Back Pain as Measured by Visual Analogue Scale (VAS)
Description	VAS was a 100 millimeter (mm) scale that subjects used to assess the severity of their lower back pain. Based on the following question, "During the past day, how much back pain did you have?", the subject was instructed to place a vertical line on the VAS to indicate the magnitude of his/her lower back pain. 0 mm = no pain and 100 mm = worst possible pain. VAS: Change = mean score at Week 6/ET minus mean score at Baseline.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Missing values were imputed by LOCF. Number of subjects with VAS scores at Baseline and Week 6/ET: celecoxib n=386, tramadol HCl n=385.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Mean (Standard Error) [mm]	-34.78 (1.33)	-34.27 (1.42)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.595
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.94
	Confidence Interval	() 95%; -4.39 to 2.52
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.76
	Estimation Comments	The mean difference reported is the LS mean difference.

4. Secondary Outcome

Title	Patient's Global Assessment of Disease Activity
Description	Number of subjects with a graded level of disease activity using the Patient's Global Assessment of Disease Activity 5-point scale (1=very good, 2=good, 3=fair, 4=poor, and 5=very poor). Subjects were classified as "Improved" if their assessment reduced at least 2 grades from baseline or if their assessment changed to Grade 1 (Very Good). Subjects were classified as "Worsened" if their assessment increased at least 2 grades from baseline or if their assessment changed to Grade 5 (Very Poor). Subjects were classified as "No Change" otherwise.
Time Frame	Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with Patient's Global Assessment of Disease Activity scores at Week 6/ET: celecoxib n=375, tramadol HCl n=367.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Improved	108 27.3%	107 27%
No Change	264 66.7%	259 65.4%
Worsened	3 0.8%	1 0.3%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	CMH tests using row mean score and adjusted for center was used to compare the two treatment groups.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.829
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

5. Secondary Outcome

Title	Physician's Global Assessment of Disease Activity
Description	Number of subjects with a physician's grading of disease activity using the Physician's Global Assessment of Disease Activity 5-point scale ((1=very good, 2=good, 3=fair, 4=poor, and 5=very poor). Subjects were classified as "Improved" if their assessment reduced at least 2 grades from baseline or if their assessment changed to Grade 1 (Very Good). Subjects were classified as "Worsened" if their assessment increased at least 2 grades from baseline or if their assessment changed to Grade 5 (Very Poor). Subjects were classified as "No Change" otherwise.
Time Frame	Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with Physician's Global Assessment of Disease Activity scores at Week 6/ET: celecoxib n=368, tramadol HCl n=361.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Improved	113 28.5%	102 25.8%
No Change	253 63.9%	258 65.2%
Worsened	2 0.5%	1 0.3%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	CMH tests using row mean score and adjusted for center was used to compare the two treatment groups.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.470
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

6. Secondary Outcome

Title	Change From Baseline in Roland-Morris Disability Questionnaire (RMDQ) Total Score
Description	Each subject assessed his/her own disability due to low back pain using the RMDQ worksheet, which consisted of 24 statements of disability. The RMDQ total score was calculated as the total number of statements that were checked; the RMDQ total scores could have ranged from 0 to 24, with higher scores indicating greater disability. RMDQ: Change = mean score at Week 6/ET minus mean score at Baseline.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Missing values were imputed by LOCF. Number of subjects with Roland-Morris Disability total scores at Baseline and Week 6/ET= celecoxib n=391, tramadol n=389.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Mean (Standard Error) [scores on a scale]	-4.90 (0.26)	-4.45 (0.25)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.339
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.31
	Confidence Interval	() 95%; -0.95 to 0.33
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.32
	Estimation Comments	The mean difference reported is the LS mean difference.

7. Secondary Outcome

Title	Change From Baseline in Modified Brief Pain Inventory (m-BPI-sf)
Description	m-BPI-sf scale assessed pain severity (0 = no pain to 10 = worst possible pain), and pain interference of functional activities (0 = does not interfere to 10 = completely interferes) during the 24 hour follow-up period. Subjects indicated: how much pain now; worst pain; average level of pain; how much pain interfered with general activity, mood, walking ability, relations with other people, sleep, normal work (including housework), and enjoyment of life. m-BPI-sf: Change = mean score at Week 6/ET minus mean score at Baseline.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Missing values were imputed by LOCF. Number of subjects with m-BPI-sf scores at Baseline and Week 6/ET: celecoxib n=373, tramadol HCl n=367.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
How Much Pain Now	-2.87 (0.12)	-2.88 (0.13)
Worst Pain in Past 24 Hours	-2.84 (0.13)	-2.87 (0.13)
Average Pain in Past 24 Hours	-2.64 (0.11)	-2.59 (0.13)
Pain Interfered With General Activity	-2.62 (0.12)	-2.57 (0.13)
Pain Interfered With Mood	-2.38 (0.14)	-2.25 (0.15)
Pain Interfered With Walking Activity	-2.26 (0.13)	-2.36 (0.13)
Pain Interfered With Relations With Others	-1.79 (0.13)	-1.74 (0.14)
Pain Interfered With Sleep	-2.52 (0.13)	-2.51 (0.15)
Pain Interfered With Normal Work	-2.57 (0.13)	-2.46 (0.15)
Pain Interfered With Enjoyment of Life	-2.43 (0.14)	-2.43 (0.15)
Pain Interference Subscale	-2.37 (0.11)	-2.33 (0.12)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	How Much Pain Now. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.741
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.05
	Confidence Interval	() 95%; -0.36 to 0.25
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.16
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Worst Pain in Past 24 Hours. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.796
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.04
	Confidence Interval	() 95%; -0.38 to 0.29
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.17
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Average Pain in Past 24 Hours. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.492
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.11
	Confidence Interval	() 95%; -0.41 to 0.20
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.16
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With General Activity. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.893
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.02
	Confidence Interval	() 95%; -0.28 to 0.33
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.16
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With Mood. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.618
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.08
	Confidence Interval	() 95%; -0.40 to 0.24
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.16
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With Walking Activity. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.440
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.12
	Confidence Interval	() 95%; -0.19 to 0.43
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.16
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With Relations With Others. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.806
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.04
	Confidence Interval	() 95%; -0.25 to 0.32
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.14
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With Sleep. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.739
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.05
	Confidence Interval	() 95%; -0.38 to 0.27
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.17
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With Normal Work. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.992
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.00
	Confidence Interval	() 95%; -0.32 to 0.33
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.17
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interfered With Enjoyment of Life. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.793
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.04
	Confidence Interval	() 95%; -0.28 to 0.36
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.16
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 11

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Pain Interference Subscale. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.973
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.00
	Confidence Interval	() 95%; -0.27 to 0.28
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.14
	Estimation Comments	The mean difference reported is the LS mean difference.

8. Secondary Outcome

Title	Change From Baseline in Medical Outcomes Study (MOS) Sleep Scale
Description	MOS sleep scale included the following attributes: sleep disturbance, snoring, awaken shortness of breath or headache, quantity of sleep, sleep adequacy, somnolence, Sleep Problem Index I, and Sleep Problem Index II. Score ranged from 0-100, with a higher score indicating more of the scale attribute (e.g., more sleep disturbance, etc.). A negative change indicated subject improvement. MOS sleep scale: Change = mean score at Week 6/ET minus mean score at Baseline.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with MOS scores at Baseline and Week 6/ET: n=celecoxib, tramadol HCl.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Sleep Disturbance (n=374, 367)	-17.91 (1.23)	-16.34 (1.17)
Snoring (n=372, 365)	-7.53 (1.31)	-5.10 (1.28)
Awaken Shortness of Breath, Headache (n=374, 367)	-8.61 (1.27)	-6.21 (1.34)
Quantity of Sleep (n=373, 365)	0.00 (0.23)	0.58 (0.28)
Sleep Adequacy (n=373, 367)	7.32 (1.48)	9.67 (1.34)
Somnolence (n=374, 366)	-9.63 (1.05)	-7.01 (1.16)
Sleep Problem Index I (n=373, 367)	-11.59 (0.96)	-10.54 (0.92)
Sleep Problem Index II (n=373, 367)	-12.97 (0.94)	-11.82 (0.89)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Sleep Disturbance. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.392
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-1.29
	Confidence Interval	() 95%; -4.26 to 1.67
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.51
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Snoring. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.691
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.66
	Confidence Interval	() 95%; -3.94 to 2.61
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.67
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Awaken Shortness of Breath or Headache. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.549
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.87
	Confidence Interval	() 95%; -3.70 to 1.97
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.45
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Quantity of Sleep. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.336
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.22
	Confidence Interval	() 95%; -0.67 to 0.23
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.23
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Sleep Adequacy. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.370
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-1.51
	Confidence Interval	() 95%; -4.83 to 1.80
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.69
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Somnolence. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.341
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-1.27
	Confidence Interval	() 95%; -3.90 to 1.35
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.34
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Sleep Problem Index I. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.604
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.59
	Confidence Interval	() 95%; -2.85 to 1.66
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.15
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Sleep Problem Index II. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.547
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.68
	Confidence Interval	() 95%; -2.92 to 1.55
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.14
	Estimation Comments	The mean difference reported is the LS mean difference.

9. Secondary Outcome

Title	Number of Subjects With Change From Baseline in MOS Optimal Sleep Scale Scores
Description	The Optimal Scale is scaled from 0 or 1 with 1 indicating 7 or 8 hours of sleep per night and 0 otherwise. Number of subjects with change of improvement (0 to 1), no change (1 to 1 or 0 to 0), or worsening (1 to 0) from baseline as indicated by the MOS Optimal sleep scale.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with MOS Optimal sleep scale scores at Week 6/ET: celecoxib n=373, tramadol HCl n=365.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
0 to 1 (Improvement)	82 20.7%	74 18.7%
1 to 1 (No change)	94 23.7%	70 17.7%
0 to 0 (No change)	171 43.2%	192 48.5%
1 to 0 (worsening)	26 6.6%	29 7.3%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Optimal sleep was analyzed using CMH general association test.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.196
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

10. Secondary Outcome

Title	Change From Baseline in Work Limitations Questionnaire (WLQ)
Description	The WLQ included the following: Time Scale, Physical Scale, Output Scale, Mental-Interpersonal Scale, and Index Scale. The scales ranged from 0 (Limited none of the time) to 100 (Limited all of the time). A negative change indicated subject improvement.
Time Frame	Baseline, Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with WLQ scores at Baseline and Week 6/ET: n=celecoxib, tramadol HCl.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Time Scale (n=241, 239)	-11.55 (1.65)	-13.58 (1.52)
Physical Scale (n=270, 254)	-13.68 (1.52)	-12.65 (1.45)
Output Scale (n=256, 246)	-11.44 (1.54)	-11.69 (1.70)
Mental-Interpersonal Scale (n=266, 252)	-9.06 (1.35)	-9.54 (1.51)
Index Scale (n=223, 220)	-3.10 (0.36)	-3.14 (0.37)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Time Scale. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.252
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	2.19
	Confidence Interval	() 95%; -1.56 to 5.94
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.91
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Physical Scale. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.798
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.48
	Confidence Interval	() 95%; -3.21 to 4.17
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.88
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Output Scale. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.700
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.74
	Confidence Interval	() 95%; -3.03 to 4.51
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.92
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Mental-Interpersonal Scale. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.902
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.21
	Confidence Interval	() 95%; -3.16 to 3.59
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 1.72
	Estimation Comments	The mean difference reported is the LS mean difference.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	Index Scale. The change from Baseline was compared between the two treatment groups using ANCOVA, with treatment and center as factors, and Baseline value as a covariate.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.581
	Comments
	Method	ANCOVA
	Comments
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.25
	Confidence Interval	() 95%; -0.65 to 1.16
	Parameter Dispersion	Type: Standard Error of the Mean; Value: 0.46
	Estimation Comments	The mean difference reported is the LS mean difference.

11. Secondary Outcome

Title	Patient's Global Evaluation of Study Medication
Description	Number of subjects with an overall response to study medication of poor, fair, good, very good, and excellent.
Time Frame	Weeks 1, 3, and 6/ET

Outcome Measure Data

Analysis Population Description

ITT.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Week 1, Excellent	46 11.6%	38 9.6%
Week 1, Very Good	97 24.5%	83 21%
Week 1, Good	126 31.8%	137 34.6%
Week 1, Fair	78 19.7%	57 14.4%
Week 1, Poor	23 5.8%	18 4.5%
Week 3, Excellent	52 13.1%	41 10.4%
Week 3, Very Good	104 26.3%	97 24.5%
Week 3, Good	122 30.8%	105 26.5%
Week 3, Fair	53 13.4%	51 12.9%
Week 3, Poor	23 5.8%	14 3.5%
Week 6/ET, Excellent	91 23%	68 17.2%
Week 6/ET, Very Good	123 31.1%	112 28.3%
Week 6/ET, Good	76 19.2%	90 22.7%
Week 6/ET, Fair	51 12.9%	51 12.9%
Week 6/ET, Poor	32 8.1%	44 11.1%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	P-value reported is the overall p-value for Week 1.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.614
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	P-value reported is the overall p-value for Week 3.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.786
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	P-value reported is the overall p-value for Week 6/ET.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.044
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

12. Secondary Outcome

Title	Patient's Satisfaction Questionnaire (With Pain Relief Scale)
Description	Number of subjects at varying levels of pain relief (1 = very dissatisfied to 10 = very satisfied).
Time Frame	Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with Patient's Satisfaction Questionnaire (with pain relief scale) scores at Week 6/ET: celecoxib n=373, tramadol HCl n=364.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
1	23 5.8%	22 5.6%
2	15 3.8%	18 4.5%
3	22 5.6%	19 4.8%
4	22 5.6%	20 5.1%
5	44 11.1%	30 7.6%
6	23 5.8%	29 7.3%
7	21 5.3%	39 9.8%
8	63 15.9%	56 14.1%
9	53 13.4%	62 15.7%
10	87 22%	69 17.4%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	CMH tests using row mean score and adjusted for center was used to compare the two treatment groups.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.870
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

13. Secondary Outcome

Title	Patient's Satisfaction Questionnaire (With Walking and Bending Ability Scale)
Description	Number of subjects at varying levels of pain relief (1 = very dissatisfied to 10 = very satisfied).
Time Frame	Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT. Number of subjects with Patient's Satisfaction Questionnaire (with walking and bending ability scale) scores at Week 6/ET: celecoxib n=373, tramadol HCl n=364.

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
1	24 6.1%	15 3.8%
2	16 4%	22 5.6%
3	19 4.8%	20 5.1%
4	31 7.8%	33 8.3%
5	35 8.8%	40 10.1%
6	32 8.1%	30 7.6%
7	35 8.8%	37 9.3%
8	55 13.9%	46 11.6%
9	56 14.1%	77 19.4%
10	70 17.7%	44 11.1%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	CMH tests using row mean score and adjusted for center was used to compare the two treatment groups.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.545
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

14. Secondary Outcome

Title	Chronic Low Back Pain Responders Based on VAS, Patient's Global, and RMDQ
Description	Subjects were successful responders if they had: > = 30% improvement from baseline to final visit in VAS assessment (as identified by 100 millimeter scale); > = 30% improvement from baseline to final visit in Patient's Global assessment (classified as improved if assessment reduced at least 2 grades from baseline or if assessment changed to Grade 1, worsened if assessment increased at least 2 grades from baseline or if assessment changed to Grade 5, or no change; and < 20% worsening from baseline to final visit in RMDQ assessment (lower scores indicated greater disability).
Time Frame	Week 6/ET

Outcome Measure Data

Analysis Population Description

ITT

Arm/Group Title	Celecoxib	Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks	50 mg capsules four times a day (QID) for 6 weeks
Measure Participants	396	396
Responders	213 53.8%	196 49.5%
Non-responders	183 46.2%	200 50.5%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Celecoxib, Tramadol HCL
	Comments	CMH test adjusted for center was used to compare the two treatment groups.
	Type of Statistical Test	Superiority or Other (legacy)
	Comments
Statistical Test of Hypothesis	p-Value	0.218
	Comments
	Method	Cochran-Mantel-Haenszel
	Comments

Adverse Events

Arm/Group Title	Celecoxib		Tramadol HCL
Arm/Group Description	200 mg capsules two times a day (BID) for 6 weeks		50 mg capsules four times a day (QID) for 6 weeks
All Cause Mortality
	Celecoxib		Tramadol HCL
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	/ (NaN)		/ (NaN)
Serious Adverse Events
	Celecoxib		Tramadol HCL
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	1/ (NaN)		0/ (NaN)
Surgical and medical procedures
Coronary artery bypass	1/396 (0.3%)		0/396 (0%)
Other (Not Including Serious) Adverse Events
	Celecoxib		Tramadol HCL
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	89/ (NaN)		171/ (NaN)
Gastrointestinal disorders
Constipation	8/396 (2%)		25/396 (6.3%)
Dry mouth	6/396 (1.5%)		20/396 (5.1%)
Nausea	31/396 (7.8%)		70/396 (17.7%)
Vomiting	5/396 (1.3%)		28/396 (7.1%)
Nervous system disorders
Dizziness	19/396 (4.8%)		53/396 (13.4%)
Headache	41/396 (10.4%)		54/396 (13.6%)
Somnolence	20/396 (5.1%)		44/396 (11.1%)
Skin and subcutaneous tissue disorders
Pruritus	2/396 (0.5%)		23/396 (5.8%)

Limitations/Caveats

[Not Specified]

More Information

Certain Agreements

Principal Investigators are NOT employed by the organization sponsoring the study.

Pfizer has the right to review disclosures, requesting a delay of < 60 days. Investigator will postpone single center publications until after disclosure of pooled data (all sites), < 12 months from study completion/termination at all participating sites. Investigator may not disclose previously undisclosed confidential information other than study results.