I have read many posts over the last year attempting to model sales revenues. There has been discussion about ramp up, employee training, the number of shifts, hours of operation, and plant utilization. Some have mentioned manufacturing constraints but most have worked their models backwards from the company’s statements regarding the maximum capacities of the three plants. DendLONG (351857) and Fordwill1953 (351860) also worked on the idea of how many patients will be treated per month but were mostly starting from the company’s statement that 2000 patients will be treated in the first 12 months. Well, why don’t we work from the bottom up? How many patients will be treated per month?

In my opinion, the biggest bottleneck or manufacturing constraint is at the incubators. In Ocyan’s post 352220, he states that “It takes about 4-5 hours to process the blood before incubation which then takes about 40 hours. Afterward, another 1-2 hours for postprocessing. So the entire process takes about two days.” These timeframes appear to be accurate and allows 12 hours of travel time to and from the plant in order for the patient’s blood to be re-infused in three days. Note that after each sample is received at the plant, it spends 85% of its overall processing time sitting in an incubator. Each workstation has two incubators. Therefore, two patient samples can be processed every two days. I would assume they schedule the samples so that one completes its incubation cycle each day. Each workstation then would process 30 samples per month. Since each patient receives three infusions, 10 patients per month at 100% theoretical capacity will have been served starting the second month.

Ocyan suggested a plant utilization factor of 70%. I believe all of the time each sample spends in the incubator will allow ample time for the staff to prepare their next move. The staff will efficiently process each sample both prior to and post incubation. Therefore, I would guess the plant utilization factor will be higher, say 90%, as a_technical_1 suggested in post 352244. Now each workstation is likely to serve 9 patients per month.

However, what happens if someone misses their apheresis appointment or weather prevents a plane from arriving in New Jersey on time? Another factor must be applied for delays of blood samples reaching the plant as scheduled. Now let’s assume each workstation can really serve patients at a rate of 8 per month (all 3 infusions).

12 workstations x 8 patients x 12 months equal 1,152 patients per year (I know this seems low). The company has stated that they will serve 2,000 patients. This number would be easily achieved if the other 36 workstation were to come online in nine months (March) or earlier. Obviously, the company has worked the overall through put equation in great detail and designed their plant accordingly. I think the company’s guidance will be accurate but I don’t think that they intended to “under promise and over deliver” in this case. I will bet that Deutsche Bank, J P Morgan, and possibly even Fidelity and Citibank have seen detailed breakdowns on Dendreon’s product through put and ability to deliver to patients. I just wanted to add my 2 cents and together, work this out to level the playing field a bit. I welcome any comments about my assumptions. In fact, I would prefer to be shown an error that would cause the production capacity to increase.

Disclosure: I’m long and strong. My first PM will be to Saulk to add to our collective position.