This study aimed to evaluate the required test characteristics that a psoriatic arthritis (PsA) biomarker test would need to achieve to be considered cost-effective.
Methods
We adapted an existing Markov model to compare a hypothetical biomarker with current practice. The model followed a patient cohort aged 45 years with moderate psoriasis (PsO) in which PsA was prevalent but unrecognized over a 40-year time horizon. Patients were assumed to be routinely seen at a dermatology clinic. In the current practice arm, patients with PsA were clinically detected. In the biomarker arm, a hypothetical test was assumed to be administered at baseline. Patients who screened positive would accept a combination of conventional disease-modifying antirheumatic drugs and targeted treatment to slow disease progression. Progression was modeled as linear changes in Health Assessment Questionnaire (HAQ) scores. We varied the sensitivity, specificity, and biomarker price based on current development progress. Scenario analyses considered alternative patient cohorts with mild and severe PsO separately.
Results
The base case showed that a biomarker test with 70 percent sensitivity, 80 percent specificity, and a price of US$500 would be cost-effective (incremental cost-effectiveness ratio US$47,566 per quality-adjusted life-year [QALY]). Three-way analyses showed that a test with 80 percent specificity could be cost-effective at a US$50,000 per QALY threshold with a sensitivity as low as 66 percent at US$500. Only a near-perfect test would be cost-effective at a US$1,000 price point. Results were sensitive to HAQ progression under treatment, therapy costs, and the patient population.
Conclusion
This study supports the continued product development of candidate PsA biomarkers.
