The Problem
A 42-year-old male presents to you with a 1-day history of left great toe pain. He was helping his neighbors move when he developed subacute onset of toe pain. He took ibuprofen and applied ice with some improvement. His history is significant for presenting to the clinic with similar pain 5 years previously. At that time, an interphalangeal joint aspiration was performed, which was nondiagnostic. He currently denies fever, chills, night sweats, or antecedent trauma. On exam, he is afebrile with some mild swelling and pain to passive range of motion of his first metatarsophalangeal (MTP1). Mild warmth and redness of the joint is evident. You suspect gout but you are not certain. The patient declines aspiration and, instead, requests a prescription for indomethacin, which worked well for him in the past. You wonder what clinical features increase the likelihood of gout in the absence of joint fluid analysis.
The Question
What clinical features predict gout in patients with monoarticular arthritis?
The Search
You open PubMed and go to “clinical queries.” You enter “acute gouty arthritis” selecting “Category: Diagnosis” and “Scope: Broad.”
Our Critique
This article presents the first available clinical decision rule for the diagnosis of acute gouty arthritis. The clinical variables used for the model are clinically relevant readily available on most patients in most clinical encounters. However, not all clinicians obtain a serum uric concentration. The scale is scored as follows: Male sex (2 points), previous arthritis attack (2 points), onset within 1 day (0.5 points), joint redness (1 point), MTP1 involvement (2.5 points), hypertension or at least 1 sign/symptom of cardiovascular disease (angina pectoris; myocardial infarction; heart failure; cerebrovascular accident; transient ischemic attack; peripheral vascular disease) (1.5 points), and serum uric acid level greater than 5.88 mg/dL (3.5 points). As confirmed by the presence of monosodium urate (MSU) crystals, the prevalences of gout at the three cutoff scores (4 or lower, greater than 4 to less than 8, and at least 8) on the final diagnostic rule were 2.8%, 27.0%, and 80.4%, respectively. A gout calculator has been developed that calculates the risk for a given patient. This tool might benefit from validation in a different sample of patients.
Clinical Decision
Based upon the clinical decision rule, he has a score of 9.5 (highest score possible without the serum uric acid level). You decide to treat him with indomethacin. You contact him a week later, and his symptoms have completely resolved. You decide to place him on a PPI for the duration of the treatment and now have second thoughts that the evidence might not support this approach. Next month, we will explore the answer to this question.
The Evidence
Janssens HJ, et al. “A diagnostic rule for acute gouty arthritis in primary care without joint fluid analysis,” (Arch. Intern. Med. 2010;170:1120-6).
P Study Design and Context: A prospective diagnostic study in which patients with signs and symptoms of monoarthritis were seen by Dutch family physicians.
P Gold Standard Comparison: Joint fluid analysis with the presence of MSU crystals.
P Procedures: Clinicians were invited to enroll their patients presenting with monoarthritis who then referred the patients to a regional hospital research center with concealed information about their diagnosis. When subjects arrived, they completed a standardized interview, physical examination, and laboratory testing within 24 hours of seeing their physician. Synovial fluid was obtained from the affected joint, which was microscopically analyzed. If MSU crystals were not identified, patients were diagnosed with psoriatic, reactive, or unknown arthritis and were followed for 1 year. If MSU crystals were later identified, then the patient was classified as having gout.
P Statistical analyses: FP gout diagnosis (index test) was related to the presence of MSU crystals (reference test) to evaluate diagnostic test characteristics. A model (model 1) was developed based upon statistical significance of clinical and laboratory variables obtained from the subjects. Models were developed for convenient clinical use, one without laboratory testing (model 2) and one with laboratory testing (model 3).
P Results: Ninety-three family physicians (FPs) enrolled 381 patients with monoarthritis with a mean age of 58 years; 75% were male. MSU crystals were identified in 216 patients and 328 (86%) had an FP diagnosis of gout. The reference test (MSU crystals) indicated an index test (FP diagnosis) sensitivity of 0.97, specificity of 0.28, positive predictive value (PPV) 0.64, and negative predictive value (NPV) of 0.87. Model 1 (statistically optimum model) consisted of six clinical and two laboratory values that were independently predictive of gout: (1) male gender; (2) previous patient-reported arthritis attack; (3) MTP1 joint involvement; (4) hypertension or at least one sign/symptom of cardiovascular disease; (5) beer consumption; (6) serum uric acid concentration greater than 5.88 mg/dL; (7) erythrocyte sedimentation rate greater than 20 mm/hr for men or greater than 30 mm/hr for women; and (8) presence of tophus. Based upon the receiver operating characteristic curves, model 3 was chosen as the final model containing seven variables (listed with scoring method in Our Critique).