Implementing a Drug Use
Indicators Study
Session Guide
SESSION GUIDE IMPLEMENTING A DRUG USE INDICATORS STUDY
Implementing a Drug Use Indicators Study
SESSION GUIDE
PURPOSE AND CONTENT
The purpose of this session is to train participants how to describe the drug use situation in a particular setting. Experience from many countries has encouraged the development of a standard method of measuring drug use practices, based on drug use indicators. This unit will expose you to the process of implementing a drug use indicator study.
The three activities will help you to gain experience in carrying out the basic steps of a real study through simulation exercises.
OBJECTIVES
This session will develop your ability to—
1. Identify the importance of drug use indicators.
2. Know the steps to implement a drug use indicators study.
3. Understand the strengths and weaknesses of different methods of sampling and data collection.
4. Understand how to analyze drug use practice in a given setting using the drug use indicators.
5. Present the results of an indicators study.
PREPARATION
1. Read the Session Notes.
2. Become familiar with the WHO Manual How to Investigate Drug Use in Health Facilities. (WHO/DAP 93.1).
3. Read Hogerzeil H.V., et al. “Field tests for rational drug use in twelve developing countries.” The Lancet 4 (December 1993): 1408-1410.
SESSION NOTES
OVERVIEW
Researchers, managers, service providers, and practictioners need to understand drug use problems so that they can work to improve performance. (See Introduction How to Investigate Drug Use in Health Facilities).
"Drug use" involves not only the actual prescribing of drugs but also a wide range of behaviors that occur during the drug use encounter, the interaction between the prescriber (provider) and the patient. These behaviors include the processes of making a diagnosis, prescribing, dispensing, and using of drugs by the patient.
Drug use indicators studies can be used for four main purposes. (See Table 4, p. 26 of the WHO Manual). These purposes are:
1. A descriptive cross-sectional study: to measure drug use in a representative group of facilities.
2. A comparative cross-sectional study: to compare facilities, providers, or groups at a single time.
3. Supervision (monitoring): to identify whether a facility is above or below a set norm of practice.
4. Evaluation: to assess the impact of an intervention in an intervention group and a control group, by measuring indicators before and after.
For each purpose, different sampling methods may be required.
In assessing drug use, you would ideally want to know about the whole process of care, but unfortunately this is not always possible. A set of limited indicators, namely the Drug Use Study Indicators, has been developed to assist in such an assessment. These indicators have been selected through a process of discussion, field testing, and revision, involving a wide range of people coordinated by INRUD, with support from WHO/DAP. Other indicators may be used when different needs arise.
Types of Indicators
The WHO/DAP indicators are divided into three groups: Prescribing Indicators; Patient Care Indicators; and Facility Indicators. The list of the indicators follows:
Prescribing Indicators
1. Average number of drugs per encounter
2. Percentage of drugs prescribed by generic name
3. Percentage of encounters with an antibiotic prescribed
4. Percentage of encounters with an injection prescribed
5. Percentage of drugs prescribed from Essential Drugs List or Formulary
Patient Care Indicators
6. Average consultation time
7. Average dispensing time
8. Percentage of drugs actually dispensed
9. Percentage of drugs adequately labeled
10. Patient’s knowledge of correct dosage
Facility Indicators
11. Availability of Copy of Essential Drugs List or Formulary
12. Availability of key drugs
DESIGNING A DRUG USE INDICATORS STUDY
The design of a drug use indicators study will vary from setting to setting. The nature and scope will depend on many factors, such as:
· Information needs of health managers
· Capabilities of the record systems
· Types of providers whose behavior is to be described
· The resources available to carry out the work
The purpose of the study will determine which indicators to use and how the study should be designed. Possible purposes of drug use indicators studies include the following:
· Describe current treatment practices
Useful for problem identification or collection of baseline information at the beginning of a project. Usually more general/basic indicators are used to describe the situation.
· Compare performance among health facilities or regions
More specific indicators usually are used; the choice of which indicators to use is made on the basis of previous findings or knowledge of the practices in an area.
· Assess the impact of an intervention
In this case, certain key outcome indicators have been determined at the beginning of an intervention project. This study is meant to measure whether the objectives of the intervention have been met by measuring changes in these indicators.
· Periodic monitoring of drug use.
Supervisors can select important performance indicators to monitor over time in a set of health facilities, so that their supervision can be more focused and targeted at specific improvements.
SAMPLING ISSUES
A great deal of work has been done to determine the optimum sampling arrangement to assess a drug use situation. (See the Session Guide entitled Sampling to Study Drug Use and the Annex to the WHO Manual.)
When an assessment of the drug use situation in a region or a set of health facilities is required, a sample of 30 prescriptions from each of 20 sites is the optimal arrangement. If possible, increase the number of sites.
When a survey is undertaken to measure impact of an intervention or to assess practices in individual facilities, at least 100 prescriptions per site should be studied.
The basic design of an indicators study carried out to characterize drug use practices in a region would call for a sample of at least 20 health facilities, with at least 30 encounters being recorded in each facility. Studying 20 facilities, or more if possible, will increase the reliability and generalizability of indicators.
If you have identified fewer that 20 health facilities of interest, clearly all facilities should be included; the number of encounters studied in each facility should be increased to ensure that 600 observations or prescriptions are included. However in this case, because the number of facilities studied is low, conclusions about practices in other facilities that were not studied should be advanced cautiously.
If there are more than 20 facilities in the study areas that will be included in the prescribing survey, it may be necessary to draw a sample of facilities from the larger population. There is danger of introducing error into the results of the prescribing analysis at this step by selecting facilities for the sample that share a certain bias that might influence their prescribing patterns. One example would be only selecting and collecting data from facilities that are easier to reach; because they are more accessible, such facilities might also have more reliable drug supply and because of this, their prescribing practices might be different.
PRESCRIBING INDICATORS
The first five indicators record information about how many and what type of drugs are prescribed:
· Average number of drugs prescribed
· Percent prescribed by generic name
· Percent of patients receiving antibiotics
· Percent of patients receiving injections
· Percent of drugs prescribed on the EDL
Two forms for collecting these indicators (the Simple Prescribing Indictors Form and Detailed Prescribing Indicators Form) are included in Annex 1.
The following four steps describe the process to measure the prescribing indicators.
1. Decide on Source of Data for Prescribing Encounters
One of the basic decisions in designing an indicators study is whether to use retrospective data extracted from historical records or prospective data collected from current patients as they are treated. There are advantages and disadvantages to both types of data collection.
Retrospective treatment records exist in many facilities, although often in different forms. These records are typically kept as part of the normal morbidity or drug consumption recording and reporting systems, or else as part of a facility-based system of medical or pharmacy records.
Retrospective data are usually easier to collect than prospective data and they suffer fewer potential biases. If they are well-maintained, it is often possible to define a retrospective study period of a year or longer, and spread cases throughout this period. This serves to minimize any possible bias due to seasonal variations in health problems. It also means that treatment is observed at many points in the drug supply cycle. If practices tend to change during periods of drug shortage, this will be reflected in the data sampled from these different periods. The major weakness of retrospective data is that they are often incomplete. Individual or entire series of records can be missing, either because they were misplaced or because they were simply not recorded in the first place. The validity of retrospective data is often difficult to verify. Since retrospective records are usually kept for a different purpose than prescription analysis, key data elements such as an exact identification of the pharmaceutical prescribed, health problem, or even whether a drug was dispensed as prescribed can be consistently missing or of uncertain accuracy.
Prospective data have the advantage that they are usually complete. However, since prospective treatment data are often collected over a very short period of time, they can suffer biases due to seasonality, peculiarities in staffing, inconsistencies in the supply cycle, or most importantly, the fact that providers are aware that their behavior is being observed. Of course, in the absence of retrospective sources of data, study planners have little choice but to collect data prospectively and try to guard against these possible sources of bias. Both sources of data have been validated in INRUD studies.
The principal question to answer in deciding to use retrospective or prospective data is whether adequate sources of retrospective data exist. The useful elements for designing a retrospective sample that these historical data provide include:
· The ability to link specific health problems diagnosed with specific drugs prescribed for individual patient encounters
· A method of selecting a random sample of patient encounters that took place within a defined period of time.
Whenever possible, use retrospective data spread over the past year. Draw a sample of records using one of the techniques described in the module entitled Sampling to Study Drug Use (usually systematic or multistage sampling).
2. Define Types of Encounters to Be Included
There are many different types of prescribing encounters that take place in an outpatient health facility. There are sometimes separate clinics for adult and pediatric cases. Even when these cases are mixed in the same clinic, prescribing practices for them will be quite different. New patients are frequently separated from re-attendances for an existing health problem, both in a record-keeping system and also in the type of care they receive. Similarly, cases presenting for curative care are managed differently from those who attend to receive preventive services such as vaccinations, prenatal or postnatal care, or child health services. The usual indicator study includes new patients attending for curative complaints. In defining the types of encounters to include, exclude repeat attendances and preventive or promotive cases.
3. Define Drugs to Be Included as Antibiotics
There are a number of different ways in which antimicrobial agents are classified in different settings. Sometimes drugs such as antiprotozoals, antihelminthics, or antituberculosis agents are placed in a separate category than other antibiotics, while other systems organize all of these products under a single category of antiinfectives or antimicrobials. The indicators of antibiotic use are quite sensitive to whether or not certain groups of drugs are included as antibiotics, especially in environments where problems such as parasitic infestation or tuberculosis are common.
Another issue in the definition of antibiotics for an indicators study is whether to include antibiotic topical preparations, such as skin creams, ophthalmic ointments, or ophthalmic drops. In areas where trachoma, bacterial conjunctivitis, or bacterial skin infections are common, these products may be widely used.
Metronidazole is another problem to classify. At primary level it is used as an antiparasitic agent, though it can be effective against anaerobic bacteria. For purposes of indicators studies, we recommend that metronidazole NOT be classified as an antibiotic. See page 15 of the WHO Manual for further explanation and for the recommended list of drug groups to include as antibiotics.
4. List Drugs to Be Classed as Generic
Data collectors and coders will need to have specific lists of drugs to class as generic in order to compute the proportion of drugs prescribed as generic products. Often an Essential Drug List or national Formulary will be written in generic terms and can be used as the basis for such a list of generic products supplied in a health system.
Certain drugs are difficult to class as generic or branded. For some products, such as aspirin, it makes sense to include all generic and branded forms as generic, since they are used interchangeably and tend to have similar costs. Although most combination drugs would not be classed as generic, some combinations in common use would qualify, such as co-trimoxazole (trimethoprim-sulfamethoxazole). Panadol is often thought to be a generic name though it is not. Abbreviations (e.g., CQ as chloroquin or HCT as hydrochlorthiazide) may be counted as generics.
PATIENT CARE INDICATORS
The second category of INRUD indicators measures the adequacy of the patient care process according to minimum standards of performance. These indicators include:
· Average consultation time
· Average dispensing time
· Percentage of drugs actually dispensed
· Percentage of drugs adequately labeled
· Patient’s knowledge of correct dosage
The length of time that a health worker spends with a patient and the type of examination performed set important limits to the adequacy of diagnosis and treatment from both a clinical and social perspective. Patients who emerge from the clinical process with prescribed pharmaceuticals should, at a minimum, understand the timing and amount of each oral and topical medication.
The following four steps are recommended to plan data collection for patient care indicators.
1. Observe the Logistics of the Patient Care Process
To determine how and when the data required to measure the indicators of patient care will be collected, it is necessary to become familiar with the organization of services at sample health facilities. In particular, data must be collected about both the clinical examination and drug dispensing procedures. In addition, a nondisruptive way must be found to intercept and interview patients at some convenient point after their drugs have been dispensed.