Notes
Technique for Calculation of the True Costs of Antibiotic Therapy
Vol. 11. No. 9
823
depending on the antibiotic and the hospital area in question. We used this technique to assess the cost of antibiotic use in an acute medical ward.
Materials and Methods. After group discussion it J.R. K e r r 1., J.G. B a r r 1, E.T.M. Smyth 1, J. O ' H a r e 2
An in-house method for costing antibiotic therapy is presented which quantifies hidden costs including costs arising from intravenous administration, labour, serum antibiotic assay, monitoring of haematological and biochemical indices and disposal of sharp instruments. A study of various hospital procedures relating directly to antibiotic therapy was undertaken in an acute medical ward, which involved determination of staff members performing various procedures, consumables used and time taken. Results of this study facilitated accurate quantification of hidden costs of i.v. antibiotic therapy in this ward. Using these results, the cost of five-day courses of gentamicin, cefuroxime, penicillin G, fiucloxacillin and erythromycin were calculated. The costing of adverse effects was not attempted. It is recommended that a costing technique of this sort is used in cost-benefit analysis of antibiotic use, as the cost of the drug alone is misleading.
Financial restraints on the National Health Service have increased gradually in recent years. An area in which there is enormous scope for rationalisation is drug prescribing. Antibiotics account for a large part of all hospital pharmacy budgets, however the actual cost of their prescription is unknown. It is a widely held belief among health care staff in general that the cost of antibiotic therapy is equal to the cost of the drug itself. Hidden costs arising from for example intravenous administration, labour, serum antibiotic assay, monitoring of haematological and biochemical indices and waste disposal (1-3), or adverse effects of antibiotics (4-6) are seldom Considered. In order to obtain a more accurate assessment of the cost of antibiotic therapy, we developed a costing technique within which subtotals vary 1Department of Bacteriology and ~Dcpartment of Pharmacy, Royal Group of Hospitals, Belfast BT12 6BA, UK.
was concluded that for any antibiotic the total cost could be divided into eight cost categories (Table 1). The quantification of these cost categories requires, firstly, the raw costs of consumables, labour, transport and incineration of waste and, secondly, knowledge of consumables used and time spent by the various grades of staff on each of the procedures involved. The raw costs within these eight cost categories were obtained as follows: antibiotic costs were obtained from the Monthly Index of Medical Specialties (MIMS), contract costs of other consumables were obtained from the Departments of Purchasing and Pharmacy, costs of standard haematological and biochemical tests were obtained from their respective departments, costs of labour were obtained from the various personnel departments and information on employers' National Insurance and Superannuation contributions were obtained from the Department of Health and Social Services. For the labour categories "staff nurse (S/N)" and "porter" the most commonly occurring levels of each of these staff grades was taken as representative (grade D staff nurse and grade 3 porter). This simplification was not necessary in the case of the "junior house officer (JHO)" labour category as all the participating members of staff in this category were employed at the basic point on the scale. To obtain knowledge of the time taken and consumables involved in the performance of various hospital procedures relating directly to antibiotic therapy, a three-month study was undertaken in an acute medical ward assessing the following procedures: i.v, cannula insertion (JHO); drug delivery (JHO); drawing of blood for laboratory tests (JHO); packing of full sharp instrument boxes into clinical waste bags and erecting new, empty boxes (S/N); and delivery of laboratory samples from the ward to the laboratory (porter). The last task involving the porters could not be recorded due to trade union objection and was therefore simulated with the help of laboratory staff. Costs of collection of clinical waste bags could not be recorded for the same reason and have therefore not been included. The staff member performing a procedure himself recorded his name, grade, the procedure, the time taken and any consumables involved, including disposal of waste and handwashing, if appropriate.
824
Eur. J. Clin. Microbiol. Infect. Dis.
Also noted during the study period was the number of courses of gentamicin therapy given in the ward, the length of each and how many times serum gentamicin and urea and electrolyte (U&E) measurements were performed during each course. Each time a serum gentamicin assay was performed it was noted whether or not dose readjustment occurred in response, and how much time this process took. The number of courses of the other antibiotics used were not noted as serum antibiotic assay of these is not required. The proportion of a sharps disposal box occupied by waste from the various antibiotic course regimens was also calculated. In analysis of results, it has been assumed that for all blood specimens sent to the laboratory in a specimen bag for any routine test, the same time was required by the JHO and the porter. However, consumables for serum gentamicin and U&E measurements were noted specifically. In the case of drug delivery, it has been assumed that all antibiotics require the same time for administration by the JHO. However, consumables for particular antibiotic courses were noted specifically where possible. Times taken for the various procedures were arrived at by calculating average values. Consumables used in the various procedures were determined by noting the minimum number necessary to perform a particular procedure. As the results showed that serum gentamicin and U&E measurements were each performed once for each course of gentamicin, calculations were made on this basis. Also since, no dose readjustments occurred after each serum gentamicin assay, this cost category (number 5) was considered to be equal to zero for the sake of this study. The cost of sharps disposal was based
on the most common sharps box in use, how much waste filled this particular box to the recommended level, how many courses of a particular antibiotic regimen this represented, nursing time, and the cost of transport and incineration. The resulting data were used to calculate the cost of the following five day courses of i.v. antibiotics: gentamicin 80 mg t.i.d. (loading dose of 120 mg), cefuroxime t.5 g t.i.d., penicillin G 1.2 g q.i.d., flucloxacillin 1 g q.i.d, and erythromycin 1 g q.i.d. This was done by itemising and costing each of the cost categories from numbers I to 7 for each of the above regimens and adding the subtotals to obtain totals. The five day regimen chosen was used to facilitate comparison of cost between different antibiotics. The itemisation and costing of the gentamicin regimen is shown in detail; for reasons of confidentiality contract costs of specific items could not be reproduced and costs therefore appear as subtotals. All costs were calculated as contract costs except the antibiotic cost which is taken from MIMS. Value added tax has not been included. R e s u l t s a n d D i s e u s s t o n . Table 2 shows the results
of the study of procedures relating directly to antibiotic therapy in an acute medical ward. Four courses of gentamicin therapy were given in the ward during the three-month study period; the length of the courses ranged from 5 to 9 days and in each case one set of peak and trough serum gentamicin assays and one set of urea and electrolyte measurements was done. In each case of serum gentamicin assay, the dose of gentamicin was not adjusted. Table 3 shows the approximate proportion of a ten-litre plastic sharps disposal box filled by simulated waste from the various
Table 1: The eightcost categoriesmakingup the costof any courseof antibiotictherapy. Cost category
Costs summarized
1. Antibiotic
The drug itself
2. M a i n t e n a n c e of IV Access
Insertion of a eannulaand maintenanceof its pateney
3. Drug Delivery
Administrationof the drug to the patient
4. Drug Monitoring 5. Dose Readjustment
Measurementof serum levels to ensure therapeuticlevelsand avoid toxiclevels Modificationof the dose followingmeasurementof serum levels
6. General Monitoring 7. Sharps Disposal
Haematologicaland biochemicalindicesrelating to the use of the drug Packingin plasticdisposal boxes,then in clinicalwaste bags,deliveryto the place of incinerationand incineration Not assessed in this study
8. Adverse Effects
Vol. 11, 1992
825
Table 2: Time taken, consumables used and staff members involved in hospital procedures relating to antibiotic therapy. Procedure (cost category number)
Staffmember
Number of measurements
Time range (rain)
Insertion ofi.v, cannula (2)
Drug delivery (3)
Mean time Consumables (min)
junior house officer
69
4-9
6.7
i.v. carinula xl anticoagulant xI 10 ml syringe x2 cotton wool xl alcohol swab xl bandage xl
junior house officer
184
-
4.7
for 80 mg gentamicin: needle xl 2 ml syringe xl for 1 g erythromyein: 20 ml sterile water xl 20 ml syringe xl needle xl for 1.5 g cefuroxime 1 g flucloxacillin & 1.2 g penicillin G: 10 ml sterile water xl 10 ml syringe xl needle xl
Drawing of blood (4 + 6)
junior house officer
90
-
5.3
Sharps disposal (7)
staff nurse
41
2-6
4.0
Delivery of blood to laboratory (4 + 6)
porter (simulated by laboratory staff)
41
5-9
6.9
Table 3: Approximate proportion of a ten-litre plastic disposal box filled by sharps waste from various five-day courses of antibiotic therapy. Antibiotic regimen (5 days) Gentamicin 80 mg t.i.d. Penicillin G 1.2 g q.i.d. Cefuroxime 1.5 g t.i.d. Flucloxacillin I g q.i.d. Erythromycin 1 g q.i.d.
Percentage of box filled by sharps waste 33 33 33 33 50
five-day antibiotic courses. Table 4 shows the itemised cost categories and resultant subtotals
for serum gentamiein assay; peak & trough: 10 ml syringe x2 needle x2 alcohol swab x2 cotton wool x2 blood bottle x2 specimen bag xl
for a five-day course of i.v. gentamicin, 80 mg t.i.d with a loading dose of 120 mg, b a s e d on the data shown in Tables 1-3; drug cost and hidden costs account for 44 % and 56 % respectively o f the total cost. T h e true ratio of hidden cost to antibiotic cost is actually greater in our hospital as p h a r m a c y contracting a r r a n g e m e n t s significantly reduce the drug cost c o m p o n e n t . H o w e v e r , for reasons of confidentiality our p h a r m a c y contract costs cannot be quoted. Quantification of adverse effects would decrease the drug contribution still further. Table 5 shows drug and hidden costs for the five different antibiotic regimens, including that of gentamicin. Costs were calculated using the s a m e
826
Eur. J. Clin. Microbiol. Infect. Dis.
Table 4: Itemisation of cost categories and corresponding costs for a five-day course of i.v. gentamicin 80 mg t.i.d, with a loading dose of 120 mg based on results of monitored procedures (Table 2). Cost category
Items involved
Cost
1. Antibiotic
80mggentamicin x16
25.28
2. Maintenance of IV Access
i.v. cannula xl anticoagulant xl 10 ml syringe x2 cotton wool xl bandage xl alcohol swab xl 6.7 minutes 9-5 pm JHO
2.03
needle x15 2 ml syringe x15 47 minutes 9-5 pm JHO •23.5 minutes on-call JHO
6.41
3. Drug Delivery
4. Drug Monitoring (peak & trough serum levels, measured xl)
10 ml syringe x2 needle x2 alcohol swab x2 cotton wool x2 blood bottle x2 specimen bag xl gentamicin assay x2 10.6 minutes 9-5 pm JHO 6.9 minutes 9-5 pm Porter
7. WasteDisposal
8. Adverse Effects (not costed)
17.87 0
5. Dose Readjustment 6. General Monitoring (urea & electrolytes)
(£)
10 ml syringe xl needle xl alcohol swab xl cotton wool xl blood bottle specimen bag urea and electrolytes xl 5.3 minutes 9-5 pm JHO 6.9 minutes 9-5 pm Porter
5.27
1/3 cost of 101 plastic sharps box 4/3 minutes 9-5 pm staff nurse contract cost of waste transport and incineration
0.78
nephrotoxicity ototoxicity bacterial resistance superinfection cannula site infection
Total
principles as those used to calculate the cost of gentamicin (Table 4), however labour and consumables vary from regimen to regimen, accounting for variations in hidden costs. Antibiotics which required serum level monitoring, e.g. gentamicin, had larger hidden costs than those which
57.64
did not, e.g. cefuroxime, other factors being equal (Table 5). Antibiotics ad ministered four times per day, e.g. penicillin G, had higher hidden costs than those administered three times per day, e.g. cefuroxime, other factors being equal (Table 5).
Vol. 11, 1992
827
Table 5: Drug cost (MIMS) versus hidden costs for five-day courses of five antibiotics.
Gentamicin 80 mg t.i.d. Penicillin G 1.2 g q.i.d. Cefuroxime 1.5 g t.i.d. Erythromycin 1 g q.i.d. Flucloxacillin 1 g q.i.d
Drugcost (£)
Hidden cost (£)
Total cost (£)
Drugcost as a percentageof totalcost
25.28 4.80 79.35 160.60 74.20
32.36 12.37 10.35 15.11 12.38
57.64 17.17 89.70 175.71 86.58
43.9 28.0 88.5 91.4 85.7
Although numbers of measurements for all procedures are reasonably high, a weakness in the study is that calculations for monitoring serum levels of gentamicin are based on data from only four courses of gentamicin. For each of these courses, one urea and electrolyte measurement and one pair of serum gentamicin assays were performed. Within our hospital this may not be representative; although this is the r e c o m m e n d e d procedure it is our impression that only a fraction of courses of aminoglycosides are monitored thus. Furthermore, it has been assumed that all antibiotics require the same time for drug delivery; this is obviously not the case. For example, a water-soluble antibiotic will take a shorter time to prepare and administer than a relatively insoluble one. It is thus difficult to accurately quantify the costs of this. As these calculations are complex and change frequently with variations in costs, this system is at present being installed in the form of computer spreadsheets. The software chosen was Microsoft Excel, version 4.0 (Microsoft, UK), running under Windows 3.1 (Microsoft, UK), on a Compaq Deskpro 386s personal computer (Compaq C o m p u t e r Corporation, USA). The system consists of one master spreadsheet containing all raw costs which is linked to many individual antibiotic spreadsheets. The antibiotic spreadsheets each begin with a questionnaire which when filled out specifies the course regimen exactly. This information and the appropriate costs from the master spreadsheet are then used to calculate the cost for each cost category; these are then added to obtain the total. Both MIMS and pharmacy costs are calculated and each is displayed both numerically and graphically. Computerisation, a s d e s c r i b e d , simplifies firstly cost updating, as the costs are updated only in the master spreadsheet, secondly calculation, as the calculation mechanism is permanently incorporated into the individual antibiotic spreadsheet, and thirdly data extraction, as only the last section of any antibiotic spreadsheet
need be printed. Clearly, the flexibility of this system is enormous as an antibiotic spreadsheet can accommodate any drug regimen; this is achieved by a questionnaire, the answers to which are incorporated directly in the calculations. An assessment of the true costs of antibiotics is vital in cost-benefit analysis of antibiotic use. True antibiotic costs as described are the best estimate we have at present. In future we hope to incorporate also the costs of adverse effects. However, cost is but one part, albeit major, of the costbenefit equation. The other major part is, of course, efficacy of the antibiotics prescribed. With increased access to patient information and constant analysis of bacterial resistance in relation to antibiotic use, we hope that the ultimate goal of rational antibiotic prescription will be attained.
References 1. Davey P, Dodd T, Kerr S, Malek M: Audit of IV anti-
biotic administration. Pharmaceutical Journal 1990, 244: 793-796. 2. Davey P, Malek M: Costs of hospital acquired infection. Pharmaceutical Journal 1991, 246, Supplement: 12-14. 3. Guglielmo J, Brooks GF: Antimicrobial therapy: costbenefit considerations. Drugs 1989, 38: 473--480. 4. Eisenberg JM, Koffer H, Glock HA, Connell ML, Loss LE, Talbot GH, Shuslerman NH, Strom BL: What
is the cost of nephrotoxicity associated with aminoglycosides? Annals of Internal Medicine 1987, 107: 900-909. 5. Davey P, Hernanz C, Lynch W, Malek M, Byrne D"
Human and non-financial costs of hospital-acquired infection. Journal of Hospital Infection 1991, 18, Supplement A: 79--84. 6. Daschner F: Unnecessary and ecological costs of hospital Infection. Journal of Hospital Infection 1991, 18, Supplement A: 73-78.
Technique for Calculation of the True Costs of Antibiotic Therapy
Vol. 11. No. 9
823
depending on the antibiotic and the hospital area in question. We used this technique to assess the cost of antibiotic use in an acute medical ward.
Materials and Methods. After group discussion it J.R. K e r r 1., J.G. B a r r 1, E.T.M. Smyth 1, J. O ' H a r e 2
An in-house method for costing antibiotic therapy is presented which quantifies hidden costs including costs arising from intravenous administration, labour, serum antibiotic assay, monitoring of haematological and biochemical indices and disposal of sharp instruments. A study of various hospital procedures relating directly to antibiotic therapy was undertaken in an acute medical ward, which involved determination of staff members performing various procedures, consumables used and time taken. Results of this study facilitated accurate quantification of hidden costs of i.v. antibiotic therapy in this ward. Using these results, the cost of five-day courses of gentamicin, cefuroxime, penicillin G, fiucloxacillin and erythromycin were calculated. The costing of adverse effects was not attempted. It is recommended that a costing technique of this sort is used in cost-benefit analysis of antibiotic use, as the cost of the drug alone is misleading.
Financial restraints on the National Health Service have increased gradually in recent years. An area in which there is enormous scope for rationalisation is drug prescribing. Antibiotics account for a large part of all hospital pharmacy budgets, however the actual cost of their prescription is unknown. It is a widely held belief among health care staff in general that the cost of antibiotic therapy is equal to the cost of the drug itself. Hidden costs arising from for example intravenous administration, labour, serum antibiotic assay, monitoring of haematological and biochemical indices and waste disposal (1-3), or adverse effects of antibiotics (4-6) are seldom Considered. In order to obtain a more accurate assessment of the cost of antibiotic therapy, we developed a costing technique within which subtotals vary 1Department of Bacteriology and ~Dcpartment of Pharmacy, Royal Group of Hospitals, Belfast BT12 6BA, UK.
was concluded that for any antibiotic the total cost could be divided into eight cost categories (Table 1). The quantification of these cost categories requires, firstly, the raw costs of consumables, labour, transport and incineration of waste and, secondly, knowledge of consumables used and time spent by the various grades of staff on each of the procedures involved. The raw costs within these eight cost categories were obtained as follows: antibiotic costs were obtained from the Monthly Index of Medical Specialties (MIMS), contract costs of other consumables were obtained from the Departments of Purchasing and Pharmacy, costs of standard haematological and biochemical tests were obtained from their respective departments, costs of labour were obtained from the various personnel departments and information on employers' National Insurance and Superannuation contributions were obtained from the Department of Health and Social Services. For the labour categories "staff nurse (S/N)" and "porter" the most commonly occurring levels of each of these staff grades was taken as representative (grade D staff nurse and grade 3 porter). This simplification was not necessary in the case of the "junior house officer (JHO)" labour category as all the participating members of staff in this category were employed at the basic point on the scale. To obtain knowledge of the time taken and consumables involved in the performance of various hospital procedures relating directly to antibiotic therapy, a three-month study was undertaken in an acute medical ward assessing the following procedures: i.v, cannula insertion (JHO); drug delivery (JHO); drawing of blood for laboratory tests (JHO); packing of full sharp instrument boxes into clinical waste bags and erecting new, empty boxes (S/N); and delivery of laboratory samples from the ward to the laboratory (porter). The last task involving the porters could not be recorded due to trade union objection and was therefore simulated with the help of laboratory staff. Costs of collection of clinical waste bags could not be recorded for the same reason and have therefore not been included. The staff member performing a procedure himself recorded his name, grade, the procedure, the time taken and any consumables involved, including disposal of waste and handwashing, if appropriate.
824
Eur. J. Clin. Microbiol. Infect. Dis.
Also noted during the study period was the number of courses of gentamicin therapy given in the ward, the length of each and how many times serum gentamicin and urea and electrolyte (U&E) measurements were performed during each course. Each time a serum gentamicin assay was performed it was noted whether or not dose readjustment occurred in response, and how much time this process took. The number of courses of the other antibiotics used were not noted as serum antibiotic assay of these is not required. The proportion of a sharps disposal box occupied by waste from the various antibiotic course regimens was also calculated. In analysis of results, it has been assumed that for all blood specimens sent to the laboratory in a specimen bag for any routine test, the same time was required by the JHO and the porter. However, consumables for serum gentamicin and U&E measurements were noted specifically. In the case of drug delivery, it has been assumed that all antibiotics require the same time for administration by the JHO. However, consumables for particular antibiotic courses were noted specifically where possible. Times taken for the various procedures were arrived at by calculating average values. Consumables used in the various procedures were determined by noting the minimum number necessary to perform a particular procedure. As the results showed that serum gentamicin and U&E measurements were each performed once for each course of gentamicin, calculations were made on this basis. Also since, no dose readjustments occurred after each serum gentamicin assay, this cost category (number 5) was considered to be equal to zero for the sake of this study. The cost of sharps disposal was based
on the most common sharps box in use, how much waste filled this particular box to the recommended level, how many courses of a particular antibiotic regimen this represented, nursing time, and the cost of transport and incineration. The resulting data were used to calculate the cost of the following five day courses of i.v. antibiotics: gentamicin 80 mg t.i.d. (loading dose of 120 mg), cefuroxime t.5 g t.i.d., penicillin G 1.2 g q.i.d., flucloxacillin 1 g q.i.d, and erythromycin 1 g q.i.d. This was done by itemising and costing each of the cost categories from numbers I to 7 for each of the above regimens and adding the subtotals to obtain totals. The five day regimen chosen was used to facilitate comparison of cost between different antibiotics. The itemisation and costing of the gentamicin regimen is shown in detail; for reasons of confidentiality contract costs of specific items could not be reproduced and costs therefore appear as subtotals. All costs were calculated as contract costs except the antibiotic cost which is taken from MIMS. Value added tax has not been included. R e s u l t s a n d D i s e u s s t o n . Table 2 shows the results
of the study of procedures relating directly to antibiotic therapy in an acute medical ward. Four courses of gentamicin therapy were given in the ward during the three-month study period; the length of the courses ranged from 5 to 9 days and in each case one set of peak and trough serum gentamicin assays and one set of urea and electrolyte measurements was done. In each case of serum gentamicin assay, the dose of gentamicin was not adjusted. Table 3 shows the approximate proportion of a ten-litre plastic sharps disposal box filled by simulated waste from the various
Table 1: The eightcost categoriesmakingup the costof any courseof antibiotictherapy. Cost category
Costs summarized
1. Antibiotic
The drug itself
2. M a i n t e n a n c e of IV Access
Insertion of a eannulaand maintenanceof its pateney
3. Drug Delivery
Administrationof the drug to the patient
4. Drug Monitoring 5. Dose Readjustment
Measurementof serum levels to ensure therapeuticlevelsand avoid toxiclevels Modificationof the dose followingmeasurementof serum levels
6. General Monitoring 7. Sharps Disposal
Haematologicaland biochemicalindicesrelating to the use of the drug Packingin plasticdisposal boxes,then in clinicalwaste bags,deliveryto the place of incinerationand incineration Not assessed in this study
8. Adverse Effects
Vol. 11, 1992
825
Table 2: Time taken, consumables used and staff members involved in hospital procedures relating to antibiotic therapy. Procedure (cost category number)
Staffmember
Number of measurements
Time range (rain)
Insertion ofi.v, cannula (2)
Drug delivery (3)
Mean time Consumables (min)
junior house officer
69
4-9
6.7
i.v. carinula xl anticoagulant xI 10 ml syringe x2 cotton wool xl alcohol swab xl bandage xl
junior house officer
184
-
4.7
for 80 mg gentamicin: needle xl 2 ml syringe xl for 1 g erythromyein: 20 ml sterile water xl 20 ml syringe xl needle xl for 1.5 g cefuroxime 1 g flucloxacillin & 1.2 g penicillin G: 10 ml sterile water xl 10 ml syringe xl needle xl
Drawing of blood (4 + 6)
junior house officer
90
-
5.3
Sharps disposal (7)
staff nurse
41
2-6
4.0
Delivery of blood to laboratory (4 + 6)
porter (simulated by laboratory staff)
41
5-9
6.9
Table 3: Approximate proportion of a ten-litre plastic disposal box filled by sharps waste from various five-day courses of antibiotic therapy. Antibiotic regimen (5 days) Gentamicin 80 mg t.i.d. Penicillin G 1.2 g q.i.d. Cefuroxime 1.5 g t.i.d. Flucloxacillin I g q.i.d. Erythromycin 1 g q.i.d.
Percentage of box filled by sharps waste 33 33 33 33 50
five-day antibiotic courses. Table 4 shows the itemised cost categories and resultant subtotals
for serum gentamiein assay; peak & trough: 10 ml syringe x2 needle x2 alcohol swab x2 cotton wool x2 blood bottle x2 specimen bag xl
for a five-day course of i.v. gentamicin, 80 mg t.i.d with a loading dose of 120 mg, b a s e d on the data shown in Tables 1-3; drug cost and hidden costs account for 44 % and 56 % respectively o f the total cost. T h e true ratio of hidden cost to antibiotic cost is actually greater in our hospital as p h a r m a c y contracting a r r a n g e m e n t s significantly reduce the drug cost c o m p o n e n t . H o w e v e r , for reasons of confidentiality our p h a r m a c y contract costs cannot be quoted. Quantification of adverse effects would decrease the drug contribution still further. Table 5 shows drug and hidden costs for the five different antibiotic regimens, including that of gentamicin. Costs were calculated using the s a m e
826
Eur. J. Clin. Microbiol. Infect. Dis.
Table 4: Itemisation of cost categories and corresponding costs for a five-day course of i.v. gentamicin 80 mg t.i.d, with a loading dose of 120 mg based on results of monitored procedures (Table 2). Cost category
Items involved
Cost
1. Antibiotic
80mggentamicin x16
25.28
2. Maintenance of IV Access
i.v. cannula xl anticoagulant xl 10 ml syringe x2 cotton wool xl bandage xl alcohol swab xl 6.7 minutes 9-5 pm JHO
2.03
needle x15 2 ml syringe x15 47 minutes 9-5 pm JHO •23.5 minutes on-call JHO
6.41
3. Drug Delivery
4. Drug Monitoring (peak & trough serum levels, measured xl)
10 ml syringe x2 needle x2 alcohol swab x2 cotton wool x2 blood bottle x2 specimen bag xl gentamicin assay x2 10.6 minutes 9-5 pm JHO 6.9 minutes 9-5 pm Porter
7. WasteDisposal
8. Adverse Effects (not costed)
17.87 0
5. Dose Readjustment 6. General Monitoring (urea & electrolytes)
(£)
10 ml syringe xl needle xl alcohol swab xl cotton wool xl blood bottle specimen bag urea and electrolytes xl 5.3 minutes 9-5 pm JHO 6.9 minutes 9-5 pm Porter
5.27
1/3 cost of 101 plastic sharps box 4/3 minutes 9-5 pm staff nurse contract cost of waste transport and incineration
0.78
nephrotoxicity ototoxicity bacterial resistance superinfection cannula site infection
Total
principles as those used to calculate the cost of gentamicin (Table 4), however labour and consumables vary from regimen to regimen, accounting for variations in hidden costs. Antibiotics which required serum level monitoring, e.g. gentamicin, had larger hidden costs than those which
57.64
did not, e.g. cefuroxime, other factors being equal (Table 5). Antibiotics ad ministered four times per day, e.g. penicillin G, had higher hidden costs than those administered three times per day, e.g. cefuroxime, other factors being equal (Table 5).
Vol. 11, 1992
827
Table 5: Drug cost (MIMS) versus hidden costs for five-day courses of five antibiotics.
Gentamicin 80 mg t.i.d. Penicillin G 1.2 g q.i.d. Cefuroxime 1.5 g t.i.d. Erythromycin 1 g q.i.d. Flucloxacillin 1 g q.i.d
Drugcost (£)
Hidden cost (£)
Total cost (£)
Drugcost as a percentageof totalcost
25.28 4.80 79.35 160.60 74.20
32.36 12.37 10.35 15.11 12.38
57.64 17.17 89.70 175.71 86.58
43.9 28.0 88.5 91.4 85.7
Although numbers of measurements for all procedures are reasonably high, a weakness in the study is that calculations for monitoring serum levels of gentamicin are based on data from only four courses of gentamicin. For each of these courses, one urea and electrolyte measurement and one pair of serum gentamicin assays were performed. Within our hospital this may not be representative; although this is the r e c o m m e n d e d procedure it is our impression that only a fraction of courses of aminoglycosides are monitored thus. Furthermore, it has been assumed that all antibiotics require the same time for drug delivery; this is obviously not the case. For example, a water-soluble antibiotic will take a shorter time to prepare and administer than a relatively insoluble one. It is thus difficult to accurately quantify the costs of this. As these calculations are complex and change frequently with variations in costs, this system is at present being installed in the form of computer spreadsheets. The software chosen was Microsoft Excel, version 4.0 (Microsoft, UK), running under Windows 3.1 (Microsoft, UK), on a Compaq Deskpro 386s personal computer (Compaq C o m p u t e r Corporation, USA). The system consists of one master spreadsheet containing all raw costs which is linked to many individual antibiotic spreadsheets. The antibiotic spreadsheets each begin with a questionnaire which when filled out specifies the course regimen exactly. This information and the appropriate costs from the master spreadsheet are then used to calculate the cost for each cost category; these are then added to obtain the total. Both MIMS and pharmacy costs are calculated and each is displayed both numerically and graphically. Computerisation, a s d e s c r i b e d , simplifies firstly cost updating, as the costs are updated only in the master spreadsheet, secondly calculation, as the calculation mechanism is permanently incorporated into the individual antibiotic spreadsheet, and thirdly data extraction, as only the last section of any antibiotic spreadsheet
need be printed. Clearly, the flexibility of this system is enormous as an antibiotic spreadsheet can accommodate any drug regimen; this is achieved by a questionnaire, the answers to which are incorporated directly in the calculations. An assessment of the true costs of antibiotics is vital in cost-benefit analysis of antibiotic use. True antibiotic costs as described are the best estimate we have at present. In future we hope to incorporate also the costs of adverse effects. However, cost is but one part, albeit major, of the costbenefit equation. The other major part is, of course, efficacy of the antibiotics prescribed. With increased access to patient information and constant analysis of bacterial resistance in relation to antibiotic use, we hope that the ultimate goal of rational antibiotic prescription will be attained.
References 1. Davey P, Dodd T, Kerr S, Malek M: Audit of IV anti-
biotic administration. Pharmaceutical Journal 1990, 244: 793-796. 2. Davey P, Malek M: Costs of hospital acquired infection. Pharmaceutical Journal 1991, 246, Supplement: 12-14. 3. Guglielmo J, Brooks GF: Antimicrobial therapy: costbenefit considerations. Drugs 1989, 38: 473--480. 4. Eisenberg JM, Koffer H, Glock HA, Connell ML, Loss LE, Talbot GH, Shuslerman NH, Strom BL: What
is the cost of nephrotoxicity associated with aminoglycosides? Annals of Internal Medicine 1987, 107: 900-909. 5. Davey P, Hernanz C, Lynch W, Malek M, Byrne D"
Human and non-financial costs of hospital-acquired infection. Journal of Hospital Infection 1991, 18, Supplement A: 79--84. 6. Daschner F: Unnecessary and ecological costs of hospital Infection. Journal of Hospital Infection 1991, 18, Supplement A: 73-78.
