Do an analysis of the cost of having an unresponsive DT

[sebbacon]

The DT fails because the list prices for generic drugs often significantly lag the possible cost savings that might be achieved by switching to a branded generic.

As prescribing is sticky (GPs are reluctant to switch brands) this is bad because:

• it potentially costs the NHS large sums money when the DT list price is set too high (GPs not switching to cheapest branded generic);
• it creates an incentive for the industry to make branded generics rather than true generics, in an attempt to capture the market (and potentially raise prices again) (GPs who have made the leap to branded generics being reluctant to switch back to generic)

It would be interesting to analyse this problem.

Illustration

Price per pill of different desogesterol products over time:

Prescribing rates of same:

There are only 3 real "players" in the desogestrel market. Cerazette, Cerelle and generic desogestrel. Other brands are available, but volumes are not high (e.g. only 10 pks of Aizea were dispensed in England in February!)

Cerelle was one of the first branded generic, and was first released in January 2013, and cost exactly half of the Cerazette AND desogestrel prescriptions (because desogestrel price was Category C, based on Cerazette). So CCGs started to switch patients, not to the generic but to Cerelle to save money.

But generic desogestrel has been cheaper than Cerelle since April 2015, and the data show that there hasn't been a corresponding reduction in the prescribing of Cerelle, although it does look like switches post-April 2015 may have been to the generic.

[sebbacon]

One way of doing this analysis is model the ideal world for generics:

• the DT would track an average the prices of the cheapest three presentations
• the DT would track the price of the cheapest presentation

This is essentially our current price-per-pill analysis with the baseline of "best decile" swapped for "best product" or similar.

There is a small technical point that our data only includes brands which have ever been prescribed. So if some have been dispensed but not yet picked up by MO teams, they won't be picked up in our analysis, so the possible savings will be underestimated.

Possible savings from fixing the DT could also exclude formulation swaps. In @richiecroker's view those differences are only there to drive the market, not for clinical reasons for patients, but we should probably consider doing versions with and without those swaps.

Alex says:

I think we just need to demonstrate the problem, then offer two solutions:

1. prescribing by brand - horrible, complicated, universally hated by GPs and patients
2. changing drug tariff better - simple, doesn't require the prescribing data, saves more money

[sebbacon]

This relates to #23 which concerns making a list of possibly-gouged items that are already in the DT.

Many of these have gone up considerably in price, and then down again. It would be interesting to see if the decreases in price match them being changed in the DT.