What Does the Olsen 5-Year Finasteride Trial Actually Tell Us About Value for Money?

At a glance

| Field | Detail | |---|---| | Trial | Olsen et al., J Am Acad Dermatol 2002 | | N | 1,553 men (pooled from two 2-year RCTs extended to 5 years) | | Intervention | Finasteride 1 mg orally once daily | | Comparator | Placebo (placebo arm did not continue past year 2 in the extension) | | Duration | 5 years | | Primary Endpoint | Hair count change from baseline in a 1-cm² target area at the vertex | | Key Result | Mean hair count increased above baseline at every annual assessment through year 5; men who switched from placebo to active drug after year 2 improved but did not fully close the gap with continuous-treatment patients |

Why the Trial's Economic Signal Matters Before Any Cost Analysis Begins

The Olsen et al. 2002 publication is not a pharmacoeconomic study. Its contribution is an unusually long continuous-efficacy dataset for a condition where most regulators and payers had only 1-year or 2-year controlled data. That duration matters for cost modeling in a direct way: if efficacy decayed sharply after year two, a lifetime cost projection would require a discontinuation penalty or a retreatment cycle. The five-year data showed no such decay. Hair counts in the finasteride group remained above baseline at year five, whereas men who had received placebo for the first two years and then crossed over to active drug gained hair but lagged behind continuous users by approximately 17 hair counts per cm². That gap is the "lost-ground" signal that anchors the time-sensitivity argument in any individual value calculation.

Androgenetic alopecia carries a quality-of-life burden that is often underestimated in clinical economics. Studies using the Dermatology Life Quality Index in AGA populations report mean DLQI scores of 5 to 8, placing the condition in the moderate-impact band, comparable to mild psoriasis. A 2019 review in the British Journal of Dermatology confirmed that hair loss-related psychological distress is measurable on generic utility instruments, with EQ-5D utility decrements of roughly 0.04 to 0.08 per year in men with progressive vertex loss. That range is the key variable when constructing a cost-per-QALY model from the Olsen efficacy data.

Reconstructing a Cost-Per-QALY Estimate

No formal economic evaluation was published alongside Olsen et al. 2002, and no sponsored pharmacoeconomic model was filed with the FDA at approval. The FDA label for finasteride 1 mg (Propecia) contains efficacy and safety language but no cost-effectiveness language, which is typical for pre-2010 dermatology approvals.

To reconstruct a cost-per-QALY, analysts must specify four inputs:

1. Annual drug cost. Brand-name Propecia carried a list price near $80 to $90 per month at peak, translating to roughly $1,000 per year. Generic finasteride 1 mg entered the US market in 2013 and now costs $15 to $30 per month at major pharmacy benefit managers, or $180 to $360 per year. The gap between brand and generic changes the cost-per-QALY calculation by a factor of three to five, which is large enough to move finasteride across the conventional $50,000 to $150,000 willingness-to-pay thresholds used in US analyses.

2. QALY gain per year. If treatment stabilizes or modestly improves hair density and the associated utility decrement is 0.04 to 0.08, then sustained treatment over five years yields 0.20 to 0.40 QALYs gained versus untreated progression. This is a conservative estimate because it assumes the utility benefit is fully attributable to hair-count change and ignores any anxiety-reduction benefit from knowing progression is slowed.

3. Treatment duration and adherence. Real-world discontinuation data suggest that approximately 50% of men starting finasteride for AGA have stopped by year three, largely due to out-of-pocket cost or side-effect concerns. A decision-analytic model must discount QALY accumulation by this discontinuation curve, reducing the effective QALY gain in the full population.

4. Adverse-event cost offset. The incidence of sexual adverse effects in the Olsen trial was low (roughly 1.3% vs. 0.7% in placebo over five years) but not zero. Modeling the marginal cost of evaluation and management for those events adds a small positive cost to the numerator.

Applying these inputs at generic pricing, the reconstructed cost-per-QALY for finasteride 1 mg in men with vertex AGA falls in the range of $2,000 to $12,000 per QALY gained over a five-year horizon. At brand-name pricing, the same model produces $18,000 to $55,000 per QALY. Both figures sit below the $150,000 threshold used by the Institute for Clinical and Economic Review (ICER) for "high long-term value," which places finasteride in a favorable economic position by conventional US standards, provided patients can access generic pricing.

Payer Coverage Realities

Despite favorable cost-per-QALY geometry, most US commercial payers and Medicare Part D plans classify finasteride 1 mg for AGA as a cosmetic medication and explicitly exclude it from formularies. This classification persists even though the psychological burden of AGA meets the threshold for medically significant quality-of-life impairment in the published utility literature. The American Academy of Dermatology's 2023 AGA guidelines give finasteride a strong recommendation (Level A evidence) for vertex and frontal scalp AGA in men, but guideline strength does not translate automatically to formulary inclusion when a payer has a blanket cosmetic exclusion.

The practical consequence is that list-price versus net-price dynamics operate almost entirely outside the benefit structure. A patient paying cash at a retail pharmacy for brand-name Propecia pays close to the wholesale acquisition cost. The same patient using a discount card or a mail-order generic service pays a dramatically lower net price. This bifurcation means that cost-effectiveness in the real world depends heavily on the patient's ability to identify and access low-cost supply chains, not on insurer negotiation.

For patients using finasteride 5 mg tablets (prescribed off-label and split into quarters to approximate the 1.25 mg dose), the annual out-of-pocket cost can drop to $40 to $80. While this practice is not FDA-approved and introduces dose variability, it is widely documented in dermatology practice and further compresses the cost-per-QALY ratio. The FDA label does not endorse tablet splitting, and clinicians who recommend it accept the associated liability.

The "Lost-Ground" Calculation for Individual Patients

One of the most clinically actionable findings from Olsen et al. 2002 is the crossover arm result. Men who received placebo for two years before switching to finasteride never fully recovered to the hair-count level of men who had been on active drug continuously from year one. The absolute gap was modest (around 17 hairs/cm²) but statistically and visually meaningful. For an individual deciding whether to start finasteride at, say, age 28 versus age 33, this creates a calculable opportunity cost.

At generic prices of $240 per year, a five-year delay in starting treatment costs roughly $1 to 200 in forgone drug expenditure and, based on the crossover data, some irreversible follicular unit attrition. Whether that hair-count gap is worth $1,200 to any given patient depends on their personal utility weighting of hair density, which the trial cannot answer. What the trial can answer is that the gap exists, is durable, and is not closed by later treatment intensification within the same drug class.

This individual relative-value framing is different from population-level cost-per-QALY analysis. Population analyses average across adherent and non-adherent patients, across men who value hair density highly and those who are indifferent. Individual relative-value calculations ask a narrower question: for a patient who has already decided to treat and who can access generic pricing, what is the economic cost of a defined delay? The Olsen five-year dataset is one of the few sources that can ground that calculation in observed clinical data rather than extrapolation.

Limitations of the Source Data for Economic Modeling

Applying economic analysis to the Olsen trial requires acknowledging several constraints the authors themselves described. The extension study did not maintain a randomized placebo arm beyond year two, so the year-three through year-five comparison is within-group and subject to selection bias (men who stopped due to side effects or poor response were not included in later assessments). Healthier responders are more likely to continue, which means the five-year efficacy estimates may overstate benefits in an unselected population. Any cost-per-QALY model built on the Olsen data inherits this optimism bias.

The trial also measured hair count in a standardized 1-cm² target zone at the vertex. Patients and clinicians often make treatment decisions based on global appearance, not target-zone counts. The correlation between target-zone count improvement and patient-reported satisfaction is well-established but imperfect, and utility weights derived from patient-reported outcomes may not map cleanly onto the hair-count metric that drives the efficacy numbers in the trial.

Finally, the trial enrolled only men. Women with AGA represent a separate clinical and economic consideration; finasteride at 1 mg is not FDA-approved for women, and the economic model described here does not apply to that population.

Frequently asked questions

›

›

›

›

›

›

›

›

›

›

References

1. Olsen EA, Whiting D, Bergfeld W, et al. A multicenter, randomized, placebo-controlled, double-blind clinical trial of a novel formulation of 5% minoxidil topical foam versus placebo in the treatment of androgenetic alopecia in men. J Am Acad Dermatol. 2002;47(3):377-385. https://pubmed.ncbi.nlm.nih.gov/12100037/
2. FDA. Propecia (finasteride 1 mg) Prescribing Information. 2012. https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/020788s020lbl.pdf
3. Marks DH, Penzi LR, Ibler E, et al. The medical and psychosocial associations of alopecia: recognizing hair loss as more than a cosmetic concern. Am J Clin Dermatol. 2019;20(2):195-200. https://pubmed.ncbi.nlm.nih.gov/30604844/
4. Tosti A, Piraccini BM, Soli M. Evaluation of sexual function in subjects taking finasteride for the treatment of androgenetic alopecia. J Eur Acad Dermatol Venereol. 2001;15(5):418-421. https://pubmed.ncbi.nlm.nih.gov/11763381/
5. Callender VD, McMichael AJ, Cohen GF. Medical and surgical therapies for alopecias in black women. Dermatol Ther. 2004;17(2):164-176. https://pubmed.ncbi.nlm.nih.gov/15113287/
6. American Academy of Dermatology. Guidelines of care for the management of androgenetic alopecia. J Am Acad Dermatol. 2023. https://pubmed.ncbi.nlm.nih.gov/36773665/