1. Some of the sampled health facilities were not in use. This necessitated either going on to the next facility or surveying one not earlier sampled.

2. Lack of records (stock cards) constituted a problem in the collection of data on stock out duration.

3. Patient records were lacking in some facilities. In these facilities, patients were furnished with notebooks, which served as case files and prescription sheets. Prescriptions were thus not kept in health facilities but recorded in the books each time the patient visited the facility. Consequently, some data on rational use was not available.

4. Generally, clinic attendance in primary health care facilities was very low (about 3 persons per day). Patient interviews were not possible in many facilities.

5. Due to long periods of stock out, cost of medicines was also difficult to obtain. Generally, the cost of the medicines, the last time they were available, was recorded.

6. It was difficult to find single diagnosis of diarrhoea in case notes. This is because diarrhoea is considered a symptom of malaria especially in children resulting in the addition of antimalarials to diarrhoea prescriptions.

7. Access to homes was a problem and this necessitated interviewing patients found in the vicinity of households

8. Respondents could not easily remember their weekly expenditure. Farmers could not, for instance, calculate the cost of farm proceeds, which they obtained free of charge.

9. Respondents could not remember the individual cost of drugs and so the total cost spent on all medicines was recorded

10. The household form did not include the category of drug sellers. In Nigeria, there are patent medicine stores that are licensed by the government to sell certain categories of medicines. These are different from local stores that sell general provisions. The data collectors, therefore, substituted local store with drug seller.