This retrospective economic evaluation followed 197 Thai Buddhist monks aged at least 55 years with Kellgren–Lawrence grade II–IV knee osteoarthritis after failed conservative care and a PRGF injection. WOMAC scores were mapped to utility with the Barton algorithm. Mean utility rose from 0.5957 at baseline to 0.7252 at month 6, then declined to 0.6660 at month 12. PRGF cost THB 6,000 versus literature-derived estimates of THB 3,762.64 for NSAIDs and THB 7,719.78 for celecoxib. The six-month ICER was THB 41,059.01 per utility gain versus NSAIDs and −THB 36,212.45 versus celecoxib. The drug groups were not contemporaneous controls and the outcome was not a QALY, so Thailand's willingness-to-pay threshold cannot be applied.
Key findings
- Mean utility rose from 0.5957 at baseline to 0.7228 at month 1 and peaked at 0.7252 at month 6 before falling to 0.6660 at month 12; utility gain therefore declined from 0.1295 at month 6 to 0.0703 at month 12.
- Six-month direct costs used in the model were THB 6,000 for PRGF, THB 3,762.64 for NSAIDs and THB 7,719.78 for celecoxib.
- PRGF's ICER versus NSAIDs was THB 41,059.01 per six-month utility gain; versus celecoxib it was −THB 36,212.45, indicating dominance only under these externally sourced parameters.
- Nineteen of 197 patients (9.64%) underwent TKA within one year, but no control group means the study cannot show that PRGF prevented or delayed surgery.
- With ±20% parameter variation, the NSAID comparison ranged from THB 27,816–78,289 when six-month PRGF utility varied and THB 19,286–63,163 when PRGF cost varied; dominance over celecoxib persisted across tested ranges.
Why this matters globally
The study demonstrates how real-world outcomes can be linked with costs in a middle-income setting and shows why waning effectiveness after six months matters economically. International transfer requires contemporaneous comparators, locally validated utility measurement, lifetime modelling and a societal perspective, alongside context-specific prices and access.
Thai researcher contribution
Clinicians from Priest Hospital and Police General Hospital combined Thai unit costs with real-patient outcomes to generate a PRGF economic hypothesis and explicitly identified its reliance on external comparators. This provides a basis for a controlled study and a health-economic model better aligned with Thailand's health system.
Limitations to consider
This was a retrospective, nonrandomised single arm with no contemporaneous NSAID or celecoxib control. Some comparator costs and utilities came from different studies or populations, including a baseline of 0.5200 versus 0.5957 for PRGF, creating confounding and non-exchangeability. All participants were monks at one hospital, limiting generalisation to women and other lifestyles. WOMAC utility mapping was developed abroad and not cross-validated in Thailand. One year is short; indirect costs, non-GI harms, downstream TKA in the ICER and lifetime QALYs were excluded. The 9.64% surgical rate cannot be attributed to PRGF. The journal page still labelled the article 'Forthcoming' when checked.