The study of the supply and distribution of generic drugs in the UK has pointed out a number of perverse incentives in the current reimbursement system, despite the effectiveness of the underlying yardstick benchmarking. OXERA proposes the following three policy options to modify the reimbursement system itself:

• expand the Drug Tariff basket to include more manufacturers and wholesalers, and change the weighting method;
• abolish Category D;

• reform the Discount Inquiry.

These are explained fully in this section. The first and third of these options could, in principle, both be implemented individually. The second option should ideally be implemented in combination with the first, and/or with some of the options outlined in the following sections. Alternatively, a new mode of contracting could be designed, whereby pharmacists play little role in the NHS purchasing strategy. Such centralised purchasing is discussed in section 8 and more fully in sections 11–16; if fully implemented, this could obviate the need for reimbursement.

As noted in section 3.3, there are indications that the Drug Tariff prices are higher than the actual market prices. The Drug Tariff (specifically Category A) is based on a basket of only a limited number suppliers (ie, UniChem and AAH, the two main national full-line wholesalers, and Norton, Cox and APS, the three manufacturers). As such, it is unlikely to be representative of the market price. It excludes the newly established national full-liner, Phoenix, as well as the remaining regional full-liners (although the latter have a small overall market share). The basket also excludes the other three large manufacturers (ie, Generics UK, CP Pharmaceuticals and Lagap), the first having a larger turnover than both Cox and APS. Most importantly, the Drug Tariff does not account for the fact that pharmacists obtain a substantial part of their generic drugs (perhaps up to 40–50%) from the short-liners.

Expanding the Drug Tariff basket would lead to more representative Category A prices. The basket should at least include the third national full-line wholesaler, Phoenix, as well as the other three large manufacturers (ie, Generics UK, CP Pharmaceuticals and Lagap). The PPA has said that it would be relatively straightforward to include these companies. It would have to ask all these suppliers for their price lists, something that already happens for Category D purposes.

Ideally, the Drug Tariff basket should also include short-liners, since these are a major provider of generics to pharmacies. However, there are problems with including the short-liners:

• they only supply a limited range of drugs, and their product offerings may vary regularly in accordance with changing profit opportunities;

• their product availability and supply conditions may vary between regions (depending on where they have depots).

The first is not a serious constraint. Short-liners’ prices would only affect prices where they held stocks of the product. In addition, some larger short-liners do offer a wide range of generic drugs and even delivery once or twice a day. For example, at least one ‘short-liner’ apparently stocks the full-line generics (1,200 products) and offers delivery twice a day. A second short-liner can deliver one or two times a day in the five regions where it has a depot, and uses couriers for overnight delivery in other areas. Including these and other short-liners in the basket is therefore feasible.

The second is more problematic, and is the reason that even regional wholesalers are not currently included. Constructing a different Drug Tariff for each region would not be feasible. Thus, either short-liners should be left out of the Drug Tariff, even though this makes Category A prices less representative, or they should be included, and no compensation should be made for the fact that some pharmacies may not have access to stocks at the reported prices.

Expanding the basket—whether by including only Phoenix and the three manufacturers, or by including some short-liners as well—has the following advantages:

• Drug Tariff prices for Category A drugs would better approximate the actual market prices paid by pharmacists;
• Category D would be triggered less frequently, since the PSNC’s and PPA’s stock inquiries would cover a greater number of Category A suppliers. Of course, Category D should not be triggered in the new system simply because one of the basket suppliers has insufficient stock, as could happen in the current system. Instead, provided that a certain minimum number of basket suppliers have sufficient stock, Category D would not be triggered;
• higher prices resulting from genuine shortages would be reflected more quickly in the Category A prices;

• suppliers in the basket have less incentive to hoard.

The cheaper prices that pharmacists receive from short-liners and the discounts offered by manufacturers and full-liners would still have to be captured through the Discount Inquiry, as under the current system. The stocks of products in shortage held by short-liners, and thus overlooked in current stock inquiries, could be discovered through sales information requirements.

As to the feasibility of this option, the PPA has affirmed that expanding the basket would be a straightforward exercise. In addition, it said that, in principle, it would also be possible to perform stock checks with the basket suppliers before they are included in the weighting for any specific drug. Clearly, this would have resource implications. This could prevent the inclusion of basket suppliers which do not have the specific product available for that month but have not taken it off their price list. At present, the PPA only performs stock checks for Category D drugs, and then only on a voluntary basis.

In addition (or as an alternative) to expanding the basket of Category A suppliers, a different weighting system could be introduced. Rather than taking a weighted average of all the basket suppliers that offer the product, the Category A price could be taken as the average of the lowest three prices offered in any month. The name of the three suppliers with the lowest price should also be mentioned explicitly in the Drug Tariff. This introduces an incentive for every basket supplier to belong to the bottom three. It does necessitate each supplier providing accurate price and stock-level information each month. Whether the bottom-three suppliers actually have sufficient stock has to be checked, otherwise the basket price would be too harsh on pharmacists.

Category D has been designed to secure the supply to patients by ensuring that pharmacists are reimbursed fairly when they cannot purchase a drug in shortage at the Drug Tariff price. Pharmacists often manage to get hold of products that are in shortage by trying different suppliers, or they dispense the equivalent branded drug, and Category D allows to them to be reimbursed for higher prices. Another useful function of Category D has been to signal to manufacturers that there are shortages.

However, Category D has not functioned satisfactorily, and it creates a number of adverse incentives, which destabilise prices.

• By design, once a product is in Category D, pharmacists no longer have an incentive to search for the lowest price.
• The financial benefits of trading in a Category D drug encourage speculative behaviour. Those players not part of the basket of suppliers, particularly short-liners but also manufacturers, are free to hold a product if they suspect there will be supply problems, knowing that this stock holding is not part of the Category D assessment.
• There is an incentive to dispense the brand, or to purchase from suppliers that inflate generic list prices and offer discounts.

• By means of the practices (discussed above) of dual price lists and of vertically integrated wholesalers always endorsing the most expensive version of a Category D drug, pharmacies profit from Category D at the expense of the NHS.

Therefore, one of the policy options is to remove Category D completely. This would eliminate the financial benefits throughout the chain of speculative trading in Category D drugs, while keeping the yardstick mechanism for pharmacies intact at all times, including in times of shortage. The effect of supply shocks, such as those that occurred at the end of 1998 and early 1999, would be mitigated.

The major disadvantage is, of course, that patients may not be supplied in times of shortage because pharmacies cannot find the product at a reasonable price. Alternatively, pharmacists who do supply the patient may be penalised for buying the drug at a higher price than the Drug Tariff price. In addition, Category D would no longer function as a mechanism for signalling shortages to manufacturers.

However, these disadvantages could be overcome by removing Category D in combination with some of the other options presented in this report.

• As noted earlier, in times of shortage pharmacists have resorted to several measures that mitigate the effects of the shortage, such as owings, bartering with nearby pharmacies, or discussions with GPs to seek changes in prescribing practice. These have the beneficial effect of further reducing the effectiveness of hoarding.
• If the Drug Tariff basket is expanded, the absence of Category D will be less severe. First, the Drug Tariff price will be more representative of the actual market price, and any price increase caused by shortages will eventually (after a month) be reflected in an increase in the Drug Tariff. Second, stocks of products in apparent shortage will be checked for with more suppliers, meaning that Category D will be triggered less frequently. Third, the number of companies with an incentive to hoard drugs in shortage is reduced.
• For drugs in real shortage, a mechanism could be devised to replace Category D whereby the Department of Health (through the PPA or another agency) would take over responsibility from pharmacists to search for, and acquire, the product. This option is discussed in detail in section 8.

4.3.1 Changes in discounting practice

The aim of the Discount Inquiry is to ‘claw back’ from pharmacies all discounts to the PPRS and Drug Tariff prices at which reimbursement is made. The key problem for the Discount Inquiry is the vertically integrated chains. More radical options for this underlying problem are discussed in section 7. Here, options to improve the Discount Inquiry that would suit the current industry structure are outlined.

The main proposal is to accept the different ownership structures in the industry and to design different claw-back solutions to match them. In addition, it is important to be clear about the source of discounts and to ensure that the Discount Inquiry questionnaire is designed to identify them. There are three different ownership structures:

• independent pharmacies, perhaps with a large number of stores or participating in a buyer group;
• wholesalers, which supply their own retail pharmacies and other pharmacies;

• Boots, which supplies only its own retail outlets.

Discounts can arise from two basic sources:

• purchasing at a better price than is reflected in the list prices of the basket suppliers. This ‘better price’ could be offered by an alternative supplier or as a discount from the basket suppliers’ list price;
• negotiating a better deal on the wholesaling/distribution part of the service.

It is likely that the first of these will result in greater discounts than the second. As consolidation continues at the pharmacy level, more centralised purchasing occurs. The guaranteed dispensing volume significantly enhances the bargaining power of the pharmacy group. The current Discount Inquiry seeks information at a pharmacy level, to compensate appropriately for the fact that pharmacies with smaller turnover will incur higher costs in drug supply. It is this that enables integrated chains (and even the supermarkets) to argue that the prices they would report for their smaller stores would not be comparable with those of small independents. By doing this, the additional discounts arising from the buyer power of these larger groups are not captured by the NHS. The proposed changes focus on identifying this ‘discount’ explicitly.

The particular discount practices that are prevalent in the industry are:

• PIs for branded drugs (supplied by full- and short-line wholesalers);
• volume discounts offered by main full-line wholesalers, to encourage consolidation of purchasing. These discounts apply to all drugs and other products supplied by full-liners;
• lower prices for generics from suppliers not included in the Drug Tariff;
• additional discounts and rebate schemes operated by manufacturers, smaller wholesalers and buying groups to encourage pharmacists to consolidate purchasing. Rebates can include cash-back or free product;

• as a pharmacist is supposed to be reimbursed exactly for any Category D drug, no product-specific discounting should be possible, although these purchases would contribute towards any volume discount. However, there is evidence of (sometimes significant) differentials between list prices and market prices for Category D drugs from some suppliers. This suggests a potentially new source of discount for the Discount Inquiry.

The main structure and workings of the Discount Inquiry would not change for pharmacies remaining independent of wholesalers. The change would be in focusing on the key margin that the Discount Inquiry is trying to identify—the difference between a pharmacy’s actual expenditure on all drugs purchased and the amount reimbursed by the PPA for drugs.

During interviews for this study, most independent pharmacists said they could not, or did not, undertake this calculation. This is because of the difficulty in tracking the price at which a drug was purchased (possibly some time before dispensing) and that at which it was reimbursed (up to three months after dispensing). All claimed not to look at this metric of purchasing skill, even on an annual basis. While the complexities of the Drug Tariff may make it difficult for pharmacies to keep track of purchases, it is not surprising that pharmacists would protect this information, as it is a clear signal for the appropriate level of claw-back.

A move towards an annual Discount Inquiry on the total actually paid for drugs dispensed would undoubtedly require different systems to be put in place by pharmacists to track stock value more closely. However, it does not seem too much to expect a pharmacy chain to be able to answer the question, ‘How much did you pay for your drugs last year?’ Any IT system (discussed in section 5) which tracked information through the chain would enhance this process. More detailed information on suppliers, volume-related discounts, proportion of ‘true’ generics dispensed, and use of PIs could all be ascertained. However, to achieve this using the current system of looking through invoices is extremely time-consuming and leads to long lags in the process. Until computerised systems are in place, it may be better to use the Discount Inquiry to look at the overall total margin and not the details. Other ways of obtaining such information are discussed in section 5.

The margin would be determined at company, not individual pharmacy, level. Hence, for example, all Tesco stores would be assumed to have gained the average margin for the company as a whole. Since most supermarkets use a full-line wholesaler to supply them, they are no different from any chain of independent pharmacies. Claw-back would be set for different turnover bands as at present, but would be based on total turnover for pharmacies that are commonly owned—the same claw-back would apply whether the turnover occurred in one store, or across ten small ones, if they were owned by one company. This would ensure that the buyer power of larger groups would be captured by the Discount Inquiry.

An additional reform might be that discounts are calculated at an individual pharmacy or company level. The entirety of the discovered discount is not clawed back, but 30%, say, remains with the pharmacy to ensure that the incentive for good purchasing remains. This option, which is similar to the system in the Netherlands, requires detailed information to be examined for each pharmacy (or chain) in the country. Under a complete revamp of the IT support for drug purchasing, such a scheme might be feasible.

If the Discount Inquiry is to tackle integrated businesses, it is very difficult to avoid addressing the issue of an appropriate return (or margin or fee) for the wholesale business. The issue of the costs and benefits of vertical integration and how to deal with it are discussed in section 7. Here, two basic ways of capturing the ‘true’ price of drugs for pharmacies integrated with wholesalers are explored.

The first option is to use the discounts discovered from analysing the independent sector as a proxy for the integrated chains. This assumes that wholesalers would offer similar deals to their own pharmacies, whether or not they are integrated; what determines the discount is the turnover of the chain as a whole. This is basically the current approach. It is not robust to further integration since it requires a dynamic representative independent sector to produce a sensible benchmark. It may also not be robust to new ways of delivering discounts.

Second, the integrated chains can be asked for information on actual purchase prices of drugs. Subsequently, some estimate of an appropriate margin on this price for wholesale and distribution is added to form an estimate of the transfer price from wholesale to retail business.

• Prices at which drugs were supplied to wholesalers, weighted by volumes, should be provided on a monthly or annual basis. This is the key piece of information that is currently not used by the Discount Inquiry.
• To this price needs to be added the costs of distribution. Information on ‘allowed’ margins from contracts between manufacturers and wholesalers could be used. Currently there are a number of schemes where the wholesaler acts as an agent, including the PPRS. Generally the margin is of the order of 15% for generics. It would be important to ensure that such clauses were not inflated for the Inquiry, with discounts offered subsequently.
• Alternatively, a full formal assessment of the wholesaler’s costs could be undertaken to give a view on the appropriate margin. It could be similar to a five-year regulatory price control, setting the margin for some period of time. A yardstick control could be employed (using the costs of the others as a benchmark for one firm), as there are three large integrated companies.

Combinations of the above could be used, depending on the size of the wholesaler. Short-liners with significant pharmacy interest could be included in the process.

Many of the same issues arise with Boots as with the integrated wholesalers; however, there is a difference. Boots is not a full-line drugs wholesaler. It only carries 4,000 lines, and uses a full-line supplier as its secondary supplier. It has its own distribution system, which is used for a wide range of retail products. Dispensing brings customers in and establishes a trustworthy reputation. This leads to different economies of scale and scope from the full-liners. It is this that has led to difficulties in including Boots in the Discount Inquiry.

As with the integrated chain, Boots could be asked to disclose information on actual purchase prices of drugs, and the distribution margin identified as appropriate for the integrated chains could be applied to Boots’ purchase costs.

The costs of a Discount Inquiry designed to give a better estimate of the costs of distribution could be large. The benefits depend on the value of the hidden information of the price at which wholesalers buy directly from manufacturers. If wholesalers were structuring transfer prices so that profits were taken upstream, this should be apparent in operating profit margins. For the full-liners, these are not very high, although generics are a very small part of turnover.

Such an Inquiry could yield significant information about prices and the value chain in the industry. As such, it would have numerous beneficial effects, including enhancing price transparency, minimising distribution costs, improving the reimbursement system, and increasing the competitiveness of the market. However, many of these objectives can be achieved by other options; a more modest change to the Discount Inquiry may be sensible at the first stage.