This article is a follow-up report on the National Prescribing Centre's hospital prescribing information project
The National Prescribing Centre's (NPC) hospital prescribing information project has now been running for over two years and is drawing to a conclusion in its current form. It was developed in response to the lack of available prescribing information within secondary care compared with that available for the primary care sector via the Prescription Pricing Authority's (PPA) Prescribing Analyses and Costs (PACT) system. This disparity in availability has recently been brought into even sharper focus by the move to "unified" and cash-limited budgets as outlined in the Government's White Paper "The new NHS: modern, dependable".1
The preliminary results from the project were reported in The Pharmaceutical Journal in 1998.2 That article successfully demonstrated the feasibility of collecting, collating and analysing prescribing expenditure data from a range of hospitals across England to both British National Formulary section level and individual drug usage level.
In an attempt to provide a "representative" sample of data, the hospitals initially selected and then retained during the second project year were large, "general" hospitals, geographically spread throughout the country. The project findings, therefore, continue to reflect a representative national picture for acute general hospital prescribing. The project data now available cover the period from January, 1997, to December, 1998, and include cost information on prescribing trends to BNF section level and drug level, year-on-year cost comparisons, FP10(HP) prescribing and antibiotic prescribing.
During the second year of the project the opportunity was taken to explore different areas of potential national interest related to hospital drug use. For example, FP10(HP) prescribing is an area where medicines are initiated by the hospital but the prescription is dispensed by a community pharmacist. As such, these drug costs and prescribing details cannot be obtained from current hospital pharmacy computer systems. Accordingly, an agreement with the PPA was reached which has now enabled such data to be included within the project information.
The project analysis in the first year highlighted BNF chapter 5 (infections) as the therapeutic area with the greatest overall expenditure nationally. This finding, coupled with the 1998 House of Lords Select Committee report on antibiotic resistance3 and the guidance issued by the Standing Medical Advisory Committee,4 prompted a decision by the project management team to look more closely at the patterns of antibiotic use within the selected hospitals. The decision was further vindicated, in March, 1999, when the NHS Executive published a Health Service Circular5 on antibiotic resistance, setting out specific action for the NHS to take in response to the House of Lords report.
Sixteen hospitals have submitted BNF expenditure data continuously since the start of the project. Figure 1 shows the trend in quarterly total drug expenditure for these hospitals over the two-year period from January, 1997.
Figure 1: Trends in total drug expenditure (January, 1997, to December, 1998) |
The Figure demonstrates a steady rise in outpatient costs throughout the period with an increase of 29 per cent between the first and last quarters. The trend for inpatient costs shows a slight decrease in expenditure during 1997 followed by a gradual rise throughout 1998 resulting in an overall 7 per cent increase over the period. As might be expected, the aggregate trend (inpatients plus outpatients) reflects more closely the fluctuations in inpatient costs. Here, a rise of 12 per cent in overall expenditure is seen between the first and last quarters.
BNF chapter 5 (infections) had already been identified as the highest cost therapeutic area for both inpatients and outpatients.2 This pattern has been confirmed following inclusion of the second year's data. Once again BNF section 5.1 (antibacterial drugs) was identified as the most significant for inpatient costs and BNF section 5.3 (antivirals) for outpatients.
Figure 2 indicates that inpatient costs for BNF section 5.1 decreased throughout 1997 but started to increase slowly during the middle of 1998, with a sharp rise during the last quarter of the year. This sharp rise could be due to a number of factors including the "'flu crisis" experienced across the country last winter and an increase in the use of more specialised antibiotics, eg, for the treatment of MRSA (methicillin-resistant Staphylococcus aureas). However, over the two-year period, inpatient antibiotic costs have been broadly contained, showing a rise of only 0.6 per cent between the first and last quarters.
Figure 2: BNF section level expenditure (chapter 5) (January, 1997, to December, 1998) |
Outpatient costs for BNF section 5.3 are seen to exhibit a different trend pattern by rising steadily over the two years and resulting overall in an 89 per cent increase in expenditure. This large rise in antiviral outpatient costs is the single largest driver for the 29 per cent increase in aggregate outpatient expenditure for all drugs (see Figure 1).
|
|||
| BNF Chapter |
Total 1997(£) |
Total 1998(£) |
Increase (%) |
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 All other |
2,563,134 11,407,923 2,979,191 10,741,409 23,611,784 5,126,795 2,261,299 16,639,228 11,354,336 1,034,864 2,430,437 590,438 6,262,221 3,353,172 8,145,406 13,191,705 |
2,547,506 13,613,491 2,945,752 11,701,659 24,463,138 5,536,568 2,453,893 12,457,382 17,041,156 1,107,583 2,385,685 566,219 6,045,675 3,431,990 8,370,412 13,442,243 |
–1 19 –1 9 4 8 9 10 2 7 –2 –4 –3 2 3 2 |
The rise in total annual drug expenditure, between 1997 and 1998 for the project hospitals was 5 per cent. Inpatient expenditure increased by 3 per cent while that for outpatients rose by 12 per cent between the two years. Table 1 shows the year-on-year variations in annual expenditure in more detail by splitting the information down to BNF chapter level for the same period. As can be seen, chapter 2 (cardiovascular) exhibits the greatest percentage growth (19 per cent) over this period. Chapter 8 (malignant disease and immunosuppression) is next with a 10 per cent rise followed by chapter 4 (central nervous system) on 9 per cent
During 1997, prescribing expenditure across the project hospitals on FP10(HP) prescriptions was £2.96m. This figure represented approximately 2 per cent of the total project expenditure on hospital prescribing and 8 per cent of their total outpatient prescribing.
|
|||
| Drug | Expenditure (£) |
% of total FP10(HP expenditure |
|
| Alglucerase Epoetin Alfa Somatropin (Rbe) Epoetin Beta Tacrolimus Imiglucerase Methadone hydrochloride Fluticasone propionate Cyclosporin Dornase Alfa Omeprazole X Co-Amoxiclav Dexamphetamine sulphate Ciprofloxacin Aciclovir |
785,390 376,704 341,897 325,454 188,909 139,115 63,293 34,406 34,092 33,702 29,133 23,811 23,290 22,782 21,745 |
24.2 11.6 10.5 10.0 5.8 4.3 2.0 1.1 1.1 1.0 0.9 0.7 0.7 0.7 0.7 | |
The use of these prescriptions varied significantly between hospitals. In fact FP10(HP) expenditure ranged from a minimum of 0.3 per cent of one hospital's outpatient costs to a maximum of 28 per cent in another. Some sites used these prescriptions to prescribe drugs from all BNF chapters, while others "targeted" their use. For example, in one hospital, over 99 per cent of its FP10(HP) expenditure was in BNF chapter 9 (nutrition and blood).
Closer examination of the FP10(HP) data identifies those drugs, which together accounted for over 75 per cent of this expenditure during 1997 (Table 2).
Eighteen antibiotics were selected for further study by the project management team as part of the second year's work. These selected antibiotics accounted for around 70 per cent of the total project antibiotic expenditure during 1997 and 1998. Table 3 shows the relative contributions made by each of these antibiotics to their aggregate spend. As can be seen, ciprofloxacin has the highest expenditure (21 per cent), followed by teicoplanin (17 per cent). It is interesting to note that the top six selected antibiotics together accounted for over three-quarters of the analysed spend.
The project database allowed trend analysis of a number of these antibiotics and was able to highlight areas where clinical choice had changed significantly over time. Figure 3 illustrates an example of this by demonstrating a notable rise in expenditure on intravenous clarithromycin coupled with a corresponding fall in expenditure on intravenous erythromycin
Figure 3: Selected antibiotics - inpatient expenditure trends (January, 1997, to December, 1998) |
| Table 3: Spend on individual antibiotics as a percentage of total selected antibiotics spend between January, 1997, and December, 1998 | |
| Antibiotic | Percentage spent |
| Ciprofloxacin Teicoplanin Cefuroxime Cefotaxime Ceftazidime Co-amoxiclav Flucloxacillin Clarithromycin Erythromycin Amoxycillin Doxycycline Imipenem with cilastatin Sodium fusidate Azithromycin Ampicillin Co-trimoxazole Trimethoprim Cefaclor |
21 17 12 9 9 8 6 4 3 2 2 1 1 1 1 <1 <1 <1 |
The introduction of primary care groups and the prospect of primary care trusts in the NHS has highlighted a potential need for effective and comprehensive prescribing information to be made available across both primary and secondary care. The evolution of clinical governance and the necessity to monitor performance against standards will further increase the demand for comparative aggregated data.
This article provides further examples of the types of information that it is feasible to produce by collecting and collating information from existing hospital pharmacy systems. The examples given are by no means exhaustive and considerably more analysis has been undertaken and provided by the project team to key NHS professionals in a recently published report.6
The problems associated with collecting and processing hospital prescribing data have already been discussed in some detail in our previous article.2 The difficulty these problems create has been reinforced during the second year of the project. For example, we continue to be unable to provide volume data (ie, number of items) in addition to the expenditure data. This is due to the sheer scale of the manual manipulation of each hospital's data required to produce volume information - a direct result of the absence of a common drug identifier across all NHS trusts.
The project team has now outlined a wide range of issues needing careful consideration and more investigation, should the concept of creating a national database of hospital prescribing information be taken further.
In the light of experience gained during the past two years, the project team considers that there is no quick-fix solution to the creation of a comprehensive NHS-based national database of prescribing in secondary care. However, feedback received on the information provided by the project to key NHS professionals and managers has indicated the significant interest in, and potential value of, this new type of data.
The long-term vision for the provision of such data ultimately rests with the development and use of electronic prescribing systems within hospitals. Widespread implementation of such systems will take at least five years. The question now requiring consideration is how we can provide useful management information on hospital drug use to the new NHS in the intervening period.
ACKNOWLEDGMENTS The authors would like to thank the following for their hard work and support: Mr David Evans (LPCS) for his ongoing technical input into the project, the staff of the pharmacy departments within the hospitals taking part in the project for producing the data reports and providing feedback on the preliminary results, and Miss Annette Ireland (NPC) for her advice and support in the presentation of this report.
Debra Walker is project manager and head of Liverpool Pharmacy Computer Services. Clive Jackson is director of the National Prescribing Centre. Correspondence to Mr Jackson at the National Prescribing Centre, The Infirmary, 70 Pembroke Place, Liverpool L69 3GF (tel 0151 794 8137)