Here is how I think Q1 unfolded in the US:
GILD GT1 starts 56,700 (81% of total 70K harvoni starts). Total Rx is 126850 (derived from IMS using conversion for underreporting based on known starts of 56700 vs NRx and applying that to TRx #s - this conversion rate is in line w previous quarters as well). price realized per Rx is 3002M/126850K = 23,665 per Rx (discount of 25% from WAC, note this is likely to go up per GILD as many exclusives that had discounts tied to them didnt start until later in the quarter)
ABBV cost per Rx in US i am assuming at 15k (45K for full course of tx). I think they had to discount more than GILD because they don't have an 8 week offering to bring down total cost per cure for a particular plan (45K is in line with a GILD 8 week cure). total estimated RX filled in Q1 is 138M/15000 = 9200. The conversion from TRx to NRx for ABBV is estimated using IMS data. Every NRx start translated to 1.63 TRx, so the NRx was 9200/1.63 = 5644 new v pak starts in the quarter, or about 10% of total NRx (56700 vs 5644). This is in line with GILD comments that they estimate 90% share for Q1. This is not an end of Q1 or current market share estimate like you assert - from the prepared remarks: "We estimate that at least 90% of genotype 1 HCV patients beginning treatment in the quarter started therapy on Harvoni". ABBV's low teens estimate (call it 13%) was where they think things stand now
GILD states they have access to 80% of lives, also in line with ABBV's assertion they have 21% contracts under preferred:
"greater than 80% of genotype 1 patients have direct access to the product through either preferred or parity formulary status". Moving forward I think ABBV can slowly but surely reallize most of that 20%. Currently I think GILD is likely doing better in ABBV's accounts, which is offsetting any gains ABBV gets in GILD preferred and parity contracts. I think eventually they will offset themselves pretty evenly with GILD getting 20% share in ABBV accounts and ABBV getting 5% of GILD and parity accounts combined. I expect the GILD share of ABBV accounts to go down somewhat over time as populations like PI failures work through the system, as well as sicker patients who can claim intolerance to ribavirin, which are really the 2 main subgroups where GILD is treatment of choice
Even if ABBV struggles to get to 20% share, I think 15% is a reasonable lower bound to plug in given they got 10% in Q1, are 13% current, and still have some contracts that have yet to come online. Even at 15% the sky is not falling given the massive size of the market. 70Kx.81 (percent of treated that are GT1) x .15 or 8500 pt starts. At 45K a pop that is a respectable run rate at steady state of 382.5M per quarter or 1.53B annually.
couple caveats - Q2 they on't quite be at steady state - over the last month NRx translated to 2.06 TRX, and at steady state it should be closer to 3x minus small % that don't complete 12 weeks. The realized price may also go down as a greater proportion of patients come from the public pool, but if that is the case then it means they are driving good use in these newer medicaid contracts and should get closer to the 20% rather than 15% I estimate they capture in Q2