DPRS consists of an 11-point rating scale ranging from 0 (no pain) to 10 (worst possible pain). A 50% responder at endpoint is a subject who has a 50% or more reduction in mean pain score at endpoint compared to baseline.

DPRS is 11-point rating scale (0=no pain to 10=worst possible pain). Subjects instructed to describe pain (upon awakening) during preceding 24 hrs by choosing appropriate number between 0-10. Mean endpoint pain score obtained from last 7 available DPRS scores of daily pain diary while subject on study medication. Overall Comparison=8-week average.

DAAC is a score used to assess treatment effects averaged over the entire length of the study. DAAC from baseline to weekly mean pain score was derived by calculating the mean of all post-baseline scores minus baseline mean pain score, and then weighing this according to the proportion of the planned study duration the subject actually completed.

DSIS consists of an 11-point rating scale (0 = pain did not interfere with sleep to 10 = completely interfered with sleep). Higher score indicating greater level of sleep disturbance.

DSIS consists of an 11-point rating scale ranging from 0 = pain did not interfere with sleep to 10 = pain completely interfered with sleep. Overall Comparison= 8-week average.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Sleep Disturbance sub-scale scores range from 0 to 100, lower scores indicate less disturbance.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Snoring sub-scale scores range from 0 to 100, lower scores indicate less snoring.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Awaken Short of Breath or with a Headache sub-scale scores range from 0 to 100, lower scores indicate less difficulty.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Sleep Quantity sub-scale scores range from 0 to 24 (number of hours slept).

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Optimal Sleep sub-scale score is a binary outcome derived from Sleep Quantity (SQ): the response is YES (or 1) if SQ = 7 or 8 hours per night.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Sleep Adequacy sub-scale scores range from 0 to 100, higher scores indicate greater sleep adequacy.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Somnolence sub-scale scores range from 0 to 100, lower scores indicate less somnolence.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Overall Sleep Problems Index is a 9-item sub-scale; scores range from 0 to 100, lower scores indicate fewer sleep problems.

EQ-5D, a subject-completed questionnaire, assesses health-related QOL. QOL Health State Profile (HSP) is designed to record subject's level of current health across 5 domains (mobility, self-care, usual activities, pain/discomfort & anxiety/depression); scores are used to calculate EQ-5D Utility Score; range: -0.594 to 1.000 (from worst to best).

EQ-5D is a subject-completed questionnaire to assess health-related QOL (Health State Profile (HSP) & Visual Analog Scale (VAS)). The VAS is designed to rate the subject's current health state on a scale from 0 to 100 where 0 represents the worst imaginable health state and 100 represents the best imaginable health state.

HADS-A consists of 7 items that are assessed by a score of 0 = no anxiety to 3 = severe feeling of anxiety. The anxiety subscale determines a state of generalized anxiety (including anxious mood, restlessness, anxious thoughts, panic attacks). Score range = 0 to 21; higher scores indicate a greater intensity of anxiety

HADS-D consists of 7 items that are assessed by a score of 0 = no depression to 3 = severe feeling of depression. The depression subscale focuses on the state of lost interest and diminished pleasure response ("lowering of hedonic tone"). Score range = 0 to 21; higher scores indicate a greater intensity of depression

PGIC is a subject-rated instrument that measured change in subject's overall status on a 7-point scale ranging from 1 (very much improved) to 7 (very much worse). Improved = minimally improved, much improved, and very much improved; Worse = minimally worse, much worse, and very much worse.

CGIC is a clinician-rated instrument that assesses the subject's overall global improvement on a 7-point scale ranging from 1 (very much improved) to 7 (very much worse). Improved = minimally improved, much improved, and very much improved; Worse = minimally worse, much worse, and very much worse.

DAAC is a score used to assess treatment effects averaged over the entire length of the study. DAAC from baseline to weekly mean pain score was derived by calculating the mean of all post-baseline scores minus baseline mean pain score, and then weighing this according to the proportion of the planned study duration the subject actually completed.

Total of all reporting groups

Subjects who were randomized to placebo remained on placebo throughout the remainder of the study.

DPRS consists of an 11-point rating scale ranging from 0 (no pain) to 10 (worst possible pain). A 50% responder at endpoint is a subject who has a 50% or more reduction in mean pain score at endpoint compared to baseline.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. For subjects who did not complete the study, the last 7 available daily pain scores while on study medication were carried forward (LOCF) to calculate the endpoint mean score.

DPRS is 11-point rating scale from 0 (no pain) to 10 (worst possible pain). Subjects instructed to describe their pain (daily upon awakening) during preceding 24 hrs by choosing the appropriate number between 0-10. Mean endpoint pain score was obtained from the last 7 available DPRS scores of the daily pain diary while subject on study medication.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. For subjects who did not complete the study, the last 7 available daily pain scores while on study medication were carried forward (LOCF) to calculate the endpoint mean score.

DPRS is 11-point rating scale from 0 (no pain) to 10 (worst possible pain). Subjects instructed to describe their pain (daily upon awakening) during preceding 24 hrs by choosing the appropriate number between 0-10. Mean endpoint pain score was obtained from the last 7 available DPRS scores of the daily pain diary while subject on study medication.

Evaluable (EVAL) Population: Subset of ITT subjects with ≥4 daily pain diaries in the 7 days before Visit 2 (randomization) with average score ≥4; ≥14 days of treatment; ≥14 days of DB daily pain diaries; not withdrawn due to "Protocol violation" or "Did not meet entrance criteria"; not previously participated in the study.

DPRS is 11-point rating scale (0=no pain to 10=worst possible pain). Subjects instructed to describe pain (upon awakening) during preceding 24 hrs by choosing appropriate number between 0-10. Mean endpoint pain score obtained from last 7 available DPRS scores of daily pain diary while subject on study medication. Overall Comparison=8-week average.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. Missing Weekly mean scores were not imputed. Number of subjects analyzed differed by week; noted as n= (pregabalin, placebo).

DAAC is a score used to assess treatment effects averaged over the entire length of the study. DAAC from baseline to weekly mean pain score was derived by calculating the mean of all post-baseline scores minus baseline mean pain score, and then weighing this according to the proportion of the planned study duration the subject actually completed.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. Subjects without any post-baseline daily pain scores would have no DAAC; missing DAACs were not imputed.

DSIS consists of an 11-point rating scale (0 = pain did not interfere with sleep to 10 = completely interfered with sleep). Higher score indicating greater level of sleep disturbance.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. For subjects who did not complete the study, the last 7 available daily sleep interference scores while on study medication were carried forward (LOCF) to calculate the endpoint mean score.

DSIS consists of an 11-point rating scale ranging from 0 = pain did not interfere with sleep to 10 = pain completely interfered with sleep. Overall Comparison= 8-week average.

ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. Missing Weekly mean scores were not imputed. Number of subjects analyzed differed by week; noted as n = (pregabalin, placebo).

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Sleep Disturbance sub-scale scores range from 0 to 100, lower scores indicate less disturbance.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Snoring sub-scale scores range from 0 to 100, lower scores indicate less snoring.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Awaken Short of Breath or with a Headache sub-scale scores range from 0 to 100, lower scores indicate less difficulty.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from the analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Sleep Quantity sub-scale scores range from 0 to 24 (number of hours slept).

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from the analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Optimal Sleep sub-scale score is a binary outcome derived from Sleep Quantity (SQ): the response is YES (or 1) if SQ = 7 or 8 hours per night.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from the analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Sleep Adequacy sub-scale scores range from 0 to 100, higher scores indicate greater sleep adequacy.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from the analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Somnolence sub-scale scores range from 0 to 100, lower scores indicate less somnolence.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from the analysis.

MOS Sleep Scale is a subject-rated questionnaire consisting of 12 items that assess the key constructs of sleep; assesses sleep for the 4 weeks prior to evaluation. The MOS Overall Sleep Problems Index is a 9-item sub-scale; scores range from 0 to 100, lower scores indicate fewer sleep problems.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. MOS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from the analysis.

EQ-5D, a subject-completed questionnaire, assesses health-related QOL. QOL Health State Profile (HSP) is designed to record subject's level of current health across 5 domains (mobility, self-care, usual activities, pain/discomfort & anxiety/depression); scores are used to calculate EQ-5D Utility Score; range: -0.594 to 1.000 (from worst to best).

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. EQ-5D was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

EQ-5D is a subject-completed questionnaire to assess health-related QOL (Health State Profile (HSP) & Visual Analog Scale (VAS)). The VAS is designed to rate the subject's current health state on a scale from 0 to 100 where 0 represents the worst imaginable health state and 100 represents the best imaginable health state.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. EQ-5D was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

HADS-A consists of 7 items that are assessed by a score of 0 = no anxiety to 3 = severe feeling of anxiety. The anxiety subscale determines a state of generalized anxiety (including anxious mood, restlessness, anxious thoughts, panic attacks). Score range = 0 to 21; higher scores indicate a greater intensity of anxiety

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. HADS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

HADS-D consists of 7 items that are assessed by a score of 0 = no depression to 3 = severe feeling of depression. The depression subscale focuses on the state of lost interest and diminished pleasure response ("lowering of hedonic tone"). Score range = 0 to 21; higher scores indicate a greater intensity of depression

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. HADS was measured at baseline and Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

PGIC is a subject-rated instrument that measured change in subject's overall status on a 7-point scale ranging from 1 (very much improved) to 7 (very much worse). Improved = minimally improved, much improved, and very much improved; Worse = minimally worse, much worse, and very much worse.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. PGIC was measured at Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

CGIC is a clinician-rated instrument that assesses the subject's overall global improvement on a 7-point scale ranging from 1 (very much improved) to 7 (very much worse). Improved = minimally improved, much improved, and very much improved; Worse = minimally worse, much worse, and very much worse.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. CGIC was measured at Week 8. For subjects not completing the study, their final data were considered Week 8 data; those without final data were excluded from analysis.

DAAC is a score used to assess treatment effects averaged over the entire length of the study. DAAC from baseline to weekly mean pain score was derived by calculating the mean of all post-baseline scores minus baseline mean pain score, and then weighing this according to the proportion of the planned study duration the subject actually completed.

The ITT population consisted of subjects who received ≥1 dose of study medication and had contributed to the analysis of any efficacy parameter. For subjects who did not complete the study, the last 7 observations while on study medication were carried forward (LOCF).

Subjects who were randomized to placebo remained on placebo throughout the remainder of the study.

This was a supportive analysis of the primary endpoint to determine if the main results were generalizable across centers.

In supportive model 1, a "treatment by center" independent variable added to the main model.

Interaction p-value based on adding interaction term to the main model.

This was a supportive analysis of the primary endpoint to determine if the main results were generalizable across different baseline values.

In supportive model 2, a "treatment by baseline mean pain" independent variable added to the main model.

Interaction p-value based on adding interaction term to the main model.

Week 1 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 2 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 3 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 4 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 5 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 6 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 7 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 8 Mean Pain Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 1 Sleep Interference Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 2 Sleep Interference Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 3 Sleep Interference Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.

Week 4 Sleep Interference Score

Pregabalin minus Placebo

Estimated from a linear mixed model with treatment, week, center and treatment by week interaction as factors and baseline value as a covariate.

All statistical testing was 2-sided and was conducted at the 5% level of significance.