Tender joint count (TJC) was to be determined by assessing 68 joints (34 right side, 34 left side) for pain using the following grading system: 0=No pain, 1=Patient states that there is pain, 2=Patient states that there is pain and winces, or 3=Patient states that there is pain, winces, and withdraws. The total number of joints graded 1, 2, or 3 were then to be counted to yield the TJC. TJC ranges from 1-68, with increasing score indicating greater number of tender joints. TJC was to be averaged over weeks 8, 10, and 12 to yield a Treatment Period Mean. Change from Baseline = Treatment Period Mean TJC - Baseline TJC.

At each clinic visit, participants were to assess disease activity using a 100 mm visual analog scale (VAS) in reponse to the question: "Considering all the ways your arthritis affects you, mark an (X) through the line for how well you are doing." The VAS ranges from "Very Well" (0) to "Very Poor" (100). The mean score at Treatment Weeks 8, 10 and 12 was calculated. A lower score indicates a better disease activity.

At each clinic visit, the Investigator was to make a global assessment of participant disease activity on a 5-point Likert scale with grading as follows: 1=Very well, 2=Well, 3=Fair, 4=Poor, or 5=Very poor (scale range: 1-5). A lower score indicates a more positive assessment of participant disease activity.

Participants were to rate their overall response to the study drug on a 5-point Likert scale with grading as follows: 0=None, 1=Poor, 2=Fair, 3=Good, or 4=Excellent (scale range: 0-4). A higher score indicates a more positive response to study drug.

The Stanford Health Assessment Questionnaire Disability Index assesses participant functional ability based on 20 questions in 8 categories of functioning: dressing, rising, eating, walking, hygiene, reach, grip, and usual activities. Responses range from 0=No disability to 3=Completely disabled. The score for each category subscale is the single response within the category with the highest score (greatest difficulty). The overall score for the Disability Index is the mean of the 8 category scores and also ranges from 0-3, with a lower score indicating less disability.

At each clinic visit, participants were to assess their amount of pain due to arthritis during the previous 48 hours on a 100 mm visual analog scale (VAS) that ranged from "No pain" (0) to "Extreme pain" (100). A lower score indicates less pain.

C-reactive protein levels rise in response to inflammation in the body. The ratio of On-treatment serum C-reative protein:Baseline serum C-reactive protein was calculated to determine a treatment effect. On-treatment C-reactive protein = the mean of serum C-reactive protein levels for Treatment Weeks 8, 10 and 12. A ratio of less than 1.0 is consistent with lower inflammation and was to be considered an improvement.

Total of all reporting groups

Participants receive matching placebo to MK-0873 twice daily for 12 weeks

Participants were categorized as meeting ACR20 criteria when they had at least 20% improvement from Baseline in tender and swollen joint counts, and improvement from Baseline in at least 3 of 5 of the following domains: Pain Visual Analog Scale (VAS), Patient Global Assessement, Physician Global Assessment, Patient Physical Function (Disability) Score and acute-phase reactant (Erythrocyte Sedimentation Rate [ESR] or C-Reactive Protein [CRP]). The average percentage of participants that met the ACR20 responder criteria over Treatment Weeks 8, 10 and 12 was calculated.

The APT population consisted of all participants with a baseline and at least one postbaseline observation.

Tender joint count (TJC) was to be determined by assessing 68 joints (34 right side, 34 left side) for pain using the following grading system: 0=No pain, 1=Patient states that there is pain, 2=Patient states that there is pain and winces, or 3=Patient states that there is pain, winces, and withdraws. The total number of joints graded 1, 2, or 3 were then to be counted to yield the TJC. TJC ranges from 1-68, with increasing score indicating greater number of tender joints. TJC was to be averaged over weeks 8, 10, and 12 to yield a Treatment Period Mean. Change from Baseline = Treatment Period Mean TJC - Baseline TJC.

No additional analyses were performed if the primary (Swollen Joint Count) and major secondary (ACR20) outcome measures resulted in a p-value of >0.05.

At each clinic visit, participants were to assess disease activity using a 100 mm visual analog scale (VAS) in reponse to the question: "Considering all the ways your arthritis affects you, mark an (X) through the line for how well you are doing." The VAS ranges from "Very Well" (0) to "Very Poor" (100). The mean score at Treatment Weeks 8, 10 and 12 was calculated. A lower score indicates a better disease activity.

No additional analyses were performed if the primary (Swollen Joint Count) and major secondary (ACR20) outcome measures resulted in a p-value of >0.05.

At each clinic visit, the Investigator was to make a global assessment of participant disease activity on a 5-point Likert scale with grading as follows: 1=Very well, 2=Well, 3=Fair, 4=Poor, or 5=Very poor (scale range: 1-5). A lower score indicates a more positive assessment of participant disease activity.

No additional analyses were performed if the primary (Swollen Joint Count) and major secondary (ACR20) outcome measures resulted in a p-value of >0.05.

Participants were to rate their overall response to the study drug on a 5-point Likert scale with grading as follows: 0=None, 1=Poor, 2=Fair, 3=Good, or 4=Excellent (scale range: 0-4). A higher score indicates a more positive response to study drug.

No additional analyses were performed if the primary (Swollen Joint Count) and major secondary (ACR20) outcome measures resulted in a p-value of >0.05.

The Stanford Health Assessment Questionnaire Disability Index assesses participant functional ability based on 20 questions in 8 categories of functioning: dressing, rising, eating, walking, hygiene, reach, grip, and usual activities. Responses range from 0=No disability to 3=Completely disabled. The score for each category subscale is the single response within the category with the highest score (greatest difficulty). The overall score for the Disability Index is the mean of the 8 category scores and also ranges from 0-3, with a lower score indicating less disability.

No additional analyses were performed if the primary (Swollen Joint Count) and major secondary (ACR20) outcome measures resulted in a p-value of >0.05.

At each clinic visit, participants were to assess their amount of pain due to arthritis during the previous 48 hours on a 100 mm visual analog scale (VAS) that ranged from "No pain" (0) to "Extreme pain" (100). A lower score indicates less pain.

No additional analyses were performed if the primary (Swollen Joint Count) and major secondary (ACR20) outcome measures resulted in a p-value of >0.05.

C-reactive protein levels rise in response to inflammation in the body. The ratio of On-treatment serum C-reative protein:Baseline serum C-reactive protein was calculated to determine a treatment effect. On-treatment C-reactive protein = the mean of serum C-reactive protein levels for Treatment Weeks 8, 10 and 12. A ratio of less than 1.0 is consistent with lower inflammation and was to be considered an improvement.

The APT population consisted of all participants with a baseline and at least one postbaseline observation.

Participants receive matching placebo to MK-0873 twice daily for 12 weeks