Drug and Therapeutics Committees - A Practical Guide
(2003; 155 pages) [French] [Spanish] [Vietnamese] View the PDF document
Table of Contents
View the documentAcronyms and abbreviations
View the documentPreface
Open this folder and view contents1. Introduction
Open this folder and view contents2.Structure and organization of a drug and therapeutics committee
Open this folder and view contents3. Managing the formulary process
Close this folder4.Assessing new medicines
View the document4.1 The need for critical assessment of new medicines
View the document4.2 Sources of information to assess new medicines
View the document4.3 Assessing the efficacy and safety of new medicines from the literature
View the document4.4 Measuring and comparing clinical treatment outcomes
View the document4.5 Measuring and comparing drug costs
View the documentAnnex 4.1 Sources of information
View the documentAnnex 4.2 Checklist to detect common problems encountered in articles
Open this folder and view contents5.Ensuring medicine safety and quality
Open this folder and view contents6.Tools to investigate the use of medicines
Open this folder and view contents7.Promoting the rational use of medicines
Open this folder and view contents8.Antimicrobials and injections
Open this folder and view contents9. Getting started
View the documentGlossary1
View the documentReferences
View the documentFurther reading
View the documentUseful addresses and websites
View the documentBack cover

4.5 Measuring and comparing drug costs

Evaluating a new medicine for the formulary involves not only efficacy, safety and quality, but also cost and cost-effectiveness. A simple determination of price is inadequate for determining the actual cost of a medicine for the health-care system. This section provides basic summary information on how to evaluate the cost of a drug, and compare different drugs, not only in terms of procurement costs, but also in terms of cost impact on the health-care system and patient outcome. Detailed description of pharmacoeconomic methods is beyond the scope of this manual and unnecessary for the average hospital DTC member to know. Nevertheless, it is important for DTC members to understand the basic principles of various pharmacoeconomic methods in order to better understand the literature on drug cost-effectiveness and cost benefit.


4081 asymptomatic men, aged 40-55 years, with dyslipaemia (total cholesterol minus HDL >5.2 mmole/l), were enrolled in a 5-year double-blind randomized study to compare gemfibrozil 600 mg twice daily with matched placebo. The number of events (fatal and non-fatal myocardial infarction and other cardiac death) was measured.




Number of events



Number of subjects



Event rates

56/2051 = 2.73%

84/2030 = 4.13%

relative risk (RR) = 2.73/4.13 = 0.66

i.e. gemfibrozil was associated with less risk of an adverse event

relative risk reduction (RRR) = (4.13 - 2.73)/4.13 = 33.9%

i.e. there was a large relative reduction (33.9%) in risk

absolute risk reduction (ARR) = 4.13 - 2.73 = 1.41%

i.e. only a small number of cases will benefit from the decreased risk

number needed to treat for 5 years to prevent one event (NNT) = (1/1.41%) = 100/1.41 = 70.9

i.e. 71 patients need to be treated with gemfibrozil for 5 years to see an effect in one patient

The researchers concluded that although gemfibrozil was associated with a relatively large reduction in risk (33.9%) of adverse event, the actual numbers of patients normally suffering such events is in fact very small, so the resulting reduction in absolute risk is small (1.41%). Thus, a large number of patients (70.9) must be treated over 5 years in order for one patient to avoid an adverse event. In addition, 2.4% of cases taking gemfibrozil suffered from moderate to severe upper gastrointestinal symptoms, as opposed to 1.2% of cases taking placebo. Taking side-effects and cost into account, many countries and hospitals may decide that the efficacy is not sufficient to justify the cost and increased risk of side-effects.

(Frick et al. 1987)

4.5.1 Price of a drug

The unit acquisition price (for example the cost of a tablet or vial) from a supplier is the easiest and most obvious measure of drug cost that is available to the DTC. Comparison of prices is useful when comparing drugs which are exactly the same chemical entity and dosage form but produced by different manufacturers. Price and other supplier considerations (such as reliability and quality) are compared when choosing which drug product to procure. Usually it is the procurement department that will make such comparisons, but the DTC may have a role to play in deciding whether the different brands are bioequivalent. When comparing medicines of a different chemical entity, even if they have equal therapeutic effect, unit price alone is inadequate for comparison. This is because the unit dosage or treatment duration or mode of administration to achieve the same clinical result will not be the same for the different medicines.

4.5.2 Cost of a drug

The acquisition price from a supplier may be the most basic cost of a drug, but is not the complete cost of using the drug. When choosing between different medicines of the same therapeutic class for inclusion in the formulary, the DTC will want to know the cost of using the medicine, not merely the price of an individual tablet or vial. There are three types of cost associated with drug use in a health-care system: direct, indirect and intangible.

direct costs

- acquisition cost of the drug or drug price

- supplies to administer the medicine

equipment for administration, syringes, gauze, IV sets, filters, pumps, etc.

- supply management costs

salaries of supply staff, transport costs and storage facilities (including warehouse, refrigerator, freezer)

- professional services costs

pharmacist salary, preparation and dispensing of medications
clinical pharmacy activities
nursing salaries, physician fees

- other direct costs

treating adverse drug reactions
inpatient and outpatient treatment of poor response to drug therapy
emergency room use
hospital overhead costs, for example electricity
laboratory services

indirect costs

- cost of illness to the patient
- lost time from work

intangible costs

- quality of life.

Although these three costs, taken together, give the most comprehensive assessment of actual drug cost, they will usually only be analysed at national level or for comparative cost-effectiveness studies. Such a comprehensive cost analysis is necessary when deciding what medicines should be on a national EML, but there is no need for every hospital DTC to re-do all such analyses. However, the DTC may wish to evaluate all the direct costs of using a new drug in order to assess whether there is sufficient budget to add it to the formulary list.

4.5.3 Cost minimization analysis

Cost minimization (cost identification) analysis is a method of comparing two or more medicines of equal therapeutic effectiveness and safety to find out which one is the cheapest. This method of cost evaluation is the one used most often by pharmacy departments; it can be used to compare

• different brands of the same drug, or

• therapeutically equivalent drugs, which are not the same chemical entity but belong to the same therapeutic category and can be used interchangeably.

Such comparison can be difficult for many medicines, as there may not be a reliable measure of equivalence between the two products. If therapeutic equivalence cannot be demonstrated then this particular type of cost comparison should not be used. When therapeutic equivalence between a new and an old drug is studied, the sponsor of the new drug should provide proof of superiority and non-inferiority. Cost minimization should also reflect the cost to prepare and administer a dose:

• pharmacist and nursing time for preparation
• laboratory costs
• cost of any ancillary equipment, for example syringes, needles, IV sets, sterile diluents.

Table 4.1 shows a cost minimization analysis of three oral antimicrobial drugs to treat uncomplicated urinary tract infection. The analysis shows that trimethoprim was the cheapest medicine. Although norfloxacin was more expensive than amoxycillin with regard to loose tablets/capsules this was not so with regard to prepackaged courses of treatment. The assumption that there is therapeutic equivalence may not be true in areas with high rates of antimicrobial resistance. Furthermore, the different rates and cost of side-effects have not been taken into account.

Table 4.1 A cost minimization analysis of three antimicrobial drugs to treat uncomplicated urinary tract infection

Cost categories

200 mg tab.

500 mg cap.

400 mg tab.

Recommended treatment regimen for uncomplicated urinary tract infection

200 mg twice daily x 5 days

3 g twice daily x 1 day

400 mg twice daily x 3 days

No. tabs/caps per course of treatment




Acquisition price for 1 loose tab./cap.




Price for course of treatment




Cost to treat 10,000 patients per year





Treatment regiments and prices were taken from the British National Formulary 2002.
a Amoxycillin sachets: price for prepackaged course of treatment was £4.16.
b Norfloxacin tablets: price for prepackaged course of treatment was £2.19.
£1 = US$ 1.5 approximately.

Table 4.2 shows a cost minimization analysis of three injectable narcotic analgesics, one of which (diamorphine) is given by two routes. The analysis shows that pethidine intramuscular (IM) or subcutaneous (SC) injection is the cheapest option. Diamorphine given by slow intravenous (IV) injection is the most expensive option.

Table 4.2 Hypothetical example of a cost minimization analysis of three injectable narcotic analgesics

Cost categories

5 mg vial

50 mg vial

30 mg vial

Recommended treatment regimen for severe pain requiring injectable analgesia

5 mg 4 hourly

5 mg 4 hourly
IM or SC

50 mg 4 hourly
IM or SC

30 mg 4 hourly
IM or SC

Acquisition price for one vial (US$)





No. doses needed per day

6 doses/day

6 doses/day

6 doses/day

6 doses/day

Price for one day’s treatment (US$)





Nursing salary @ US$2.00 per IM or SC injection





Specialist nursing salary @ US$4.00 per slow IV injection





Equipment: syringe + needle US$2.00 per set





Total drug costs per day (US$)





Anticipated no. days treatment per year

3000 days

3000 days

3000 days

3000 days

Total drug costs for 3000 days treatment (US$)






Treatment regimens and prices were taken from the British National Formulary 2002 and converted into US$; equipment prices were from the Drug Tariff November 2002, UK Department of Health, and salary estimated for 3 minutes per IM or SC injection and 6 minutes per slow IV injection.

Sensitivity analyses are very important in any kind of economic analysis. Such an analysis tests how sensitive the conclusions are to the different assumptions made. For example, in table 4.2, if we change the assumption that IV injections take twice as much nursing time as SC or IM injections, and assume instead that IV, IM and SC injections take equal nursing time, then IV diamorphine would cost less than IM or SC pentazocine.

4.5.4 Cost-effectiveness analysis

Cost-effectiveness analysis is used to compare two or more medicines which are not exactly equivalent in terms of dose or therapeutic effect, but which are used to treat the same clinical condition. This form of analysis is difficult and is often only done at the national level. It requires measuring the cost per defined measurable clinical outcome (effect) for each of the drugs. The cost of the drug should include indirect as well as direct costs and some examples of measures for clinical outcomes include:

• hypertension - blood pressure measurements

• diabetes - glycosylated hemoglobin, blood glucose results

• coronary heart disease -frequency of angina attacks

• urinary tract infections - incidence of infections

• obesity - weight measurement

• seizures disorders - frequency of seizures

• HIV/AIDS - CD4 counts

• heart failure (and any other disease) - years of life saved or quality-adjusted life years (QALYs) or disability-adjusted life years (DALYs).

Cost-effectiveness measurement can be presented in many different ways. Some examples include:

• for acute illness: cost per course of treatment or cost per cure
• for chronic illness: cost per month of satisfactory control
• for disease prevention: cost per case prevented
• for health promotion: cost per month of desired outcome.

Table 4.3 shows an example of a cost-effectiveness analysis to compare two types of antibiotic ear drop. Ear drop A costs US$6.50 and has been found to be 80% effective; ear drop B costs US$7.90 and has been found to be 90% effective.

Table 4.3 Hypothetical example of a cost-effectiveness analysis for two antibiotic ear drops


Ear drop A

Ear drop B

Cost (US$)







US$6.50 to treat 0.8 of one case successfully

US$7.90 to treat 0.9 of one case successfully

Amount (US$) needed to treat one case successfully

6.50/0.8 = 8.125

7.90/0.9 = 8.778

Thus, although ear drop A was less effective than ear drop B, it was found to be more cost-effective in terms of the amount needed to treat one case successfully. The additional cost of B was not worth the small extra benefit. The additional cost for extra benefit, known as incremental cost-effectiveness, can be calculated as follows for this example:

(7.90 - 6.50)/(0.9 - 0.8) = 1.4/0.1 = US$14.00

Is it reasonable to pay an extra US$14 per additional case successfully treated? This judgement will need to be made by the DTC.

Steps for conducting a cost-effectiveness analysis

1 Define the objective of the analysis, for example which drug regimen should be the treatment of choice?

2 Identify the different ways to achieve the objective, for example should we use a cheaper slightly less efficacious medicine or a more expensive and slightly more efficacious one?

3 Identify and measure the drug costs of each option.

4 Identify and measure the benefits (clinical outcomes) of each option.

5 Calculate and interpret the benefits of each option. The cost-effectiveness ratio is the total drug cost divided by the number of units of outcome.

6 Perform sensitivity analysis on the conclusions. This is where some of the assumptions in the analysis, for example costs of staff salaries and hospital overheads, are varied to see if changing these assumptions also changes which medicine is found to be most cost-effective. If the conclusion about which medicine is most cost-effective does not change with varying the assumptions, then the conclusion is likely to be valid. If however, the conclusion is very sensitive to changing the assumptions, then the study result is likely to be subject to error and no firm conclusion can be drawn.

Box 4.2 shows a real example of how two different kinds of thrombolytic agent for the treatment of myocardial infarction were compared from the point of view of efficacy and cost-effectiveness in Australia. The treatment of myocardial infarction in the usual way was compared with usual treatment plus the use of either streptokinase or plasminogen activator. Comparison was done in terms of (1) total treatment costs, (2) death rates, and (3) cost per life saved (or death averted). The treatment costs included all the direct and indirect costs mentioned in section 4.5.2.

4.5.5 Cost utility analysis

Cost utility analysis is a cost-effectiveness analysis, where a composite measure of effectiveness is used to reflect both the quantity and the quality of health outcome. Examples of utility measures are quality adjusted life years (QALYs) or disability life years (DALYs). As well as measuring outcome of cases cured, deaths averted, or lives saved, these measures take account of the fact that impairment, discomfort and handicap mean that a ‘life-year’ is sometimes too crude a measure of effectiveness to capture all that is clinically important. A quality adjustment factor is normally obtained through surveys where people are asked to indicate their preferences between different states of health. Because of the difficulty of assessing quality of life, this method is controversial for comparing medicines and is likely to be beyond the scope of a DTC.

4.5.6 Cost benefit analysis

In cost benefit analysis, there is calculation of (1) the cost of the medicine, and (2) the monetary value of the benefits or change in outcome. Such benefits should measure the total gain in economic welfare associated with the intervention. This is sometimes broken down into the value of healthy time gained, savings in treatment costs, and other savings or benefits. The selection and valuation of benefits is often controversial and incomplete. Cost benefit analysis is very controversial because of placing a monetary value on clinical outcomes such as life years saved.

The cost benefit ratio is the total drug cost divided by the monetary benefit (in terms of money saved by using the drug, for example less future illness, less hospitalization, etc.). Unlike cost-effectiveness analysis, where comparable medicines are analysed for the same outcome, cost benefit analysis can be used to compare different treatments with different outcomes. However, cost benefit analysis is very difficult to do, requires major assumptions that may be incorrect, and is unlikely to be useful for most DTCs.


A review of the literature concerning the cost-effectiveness of different thrombolytics in the treatment of myocardial infarction was conducted in Australia. The cost of the various treatments and the mortality rate following myocardial infarction were evaluated and the results are shown below. Prices are given in Australian dollars (AUD).

Cost of treatment and mortality rates

Usual care of myocardial infarction (MI):

AUD 3.5 million/1000 cases, 120 die

Usual care of MI + streptokinase (SK):

AUD 3.7 million/1000 cases, 90 die

Usual care of MI + plasminogen activator (TPA):

AUD 5.5 million/1000 cases, 80 die

Comparison of the different treatments

Difference between SK and usual care of MI:

Cost of treatment = AUD3.7-3.5 million/1000 cases = $0.2 million/1000 cases = AUD200/case
No. of deaths that will be prevented = 120-90 = 30 deaths/1000 cases treated
Cost-effectiveness of SK = AUD0.2 million/30 lives = $6700 per life saved

Difference between TPA and usual care of MI:

Cost of treatment = AUD5.5-3.5 million/1000 cases = 2.0 million/1000 cases = AUD2000/case
No. of deaths that will be prevented = 120-80 = 40 deaths/1000 cases treated
Cost-effectiveness of TPA = AUD2.0 million/40 lives = $50 000 per life saved

Difference between TPA and SK treatments for MI:

Cost of treatment = AUD2.0-0.2 million/1000 cases = 1.8 million/1000 cases = AUD1800/case
No. of deaths that will be prevented = 90-80 = 10 deaths/1000 cases treated
Marginal cost of TPA over SK = AUD1.8 million/10 lives = $180 000 per life saved

If one has a budget of only AUD 500 000, which drug should one use?

For SK:

No. cases that can be treated = 500 000/200 = 2500
No. lives that can be saved = (30/1000) x 2500 = 75

For TPA:

No. cases that can be treated = 500 000/2000 = 250
No. lives that can be saved = (40/1000) x 250 = 10


Although TPA is slightly more efficacious and saved marginally more lives, when cost was taken into account, more patients could be treated and more lives saved using SK. In other words, the extra cost of TPA over SK was so high ($180 000 per life saved) that with the limited budget available fewer people could be treated and lives saved, using TPA as compared to SK.

Sources: Fibrinolytic Therapy Trialists’ Collaborative Group (1994); Aylward (1996)

to previous section
to next section
The WHO Essential Medicines and Health Products Information Portal was designed and is maintained by Human Info NGO. Last updated: October 29, 2018