www.midwesterndoctor.com/p/healing-back-pain-and-ending-the

The article changes "during the first year" into "every year", which is a materially different claim.

Recent reviews of glucocorticoid-induced osteoporosis describe a rapid initial phase of bone loss after starting systemic glucocorticoids, followed by a slower continuing decline. For example, the Belgian Bone Club consensus review states that in patients starting glucocorticoids there is a 6–12% decline in lumbar-spine bone mineral density in the first year, followed by a slower but continuous decline thereafter. Another review cited in the literature likewise says bone loss can be 5–15% during the first year of treatment.

So the article's wording is inaccurate because it implies patients lose 5–15% of bone mass each year indefinitely, whereas the evidence it points to describes that magnitude primarily in the first year, with lower rates later on.

The article appears to conflate excess cardiovascular events with deaths, and then attributes an unsupported combined fatality figure to Vioxx and Celebrex.

For Vioxx (rofecoxib), FDA safety scientist David Graham's Senate testimony gave an estimate of 88,000 to 139,000 excess cases attributable to Vioxx, adding that 30–40% probably died. That implies roughly 35,000 to 56,000 deaths, not 195,000.

For Celebrex (celecoxib), the evidence does show cardiovascular risk warnings, especially at some doses, but the published meta-analysis cited below found no significant increase in cardiovascular death (odds ratio 1.02, 95% CI 0.52–1.99). FDA/DailyMed labeling warns of potentially fatal cardiovascular events, but it does not estimate anything close to another ~140,000 deaths attributable to celecoxib.

So the statement "jointly ... killed roughly 195,000 people" is not supported by the established estimates and substantially overstates the documented death toll.