The survey was conducted using the WHO Operational Package for Monitoring and Assessing the Pharmaceutical Situation in Countries (April 2003 version). The package contains survey tools for two levels of core indicators and a household survey tool:
Level I - Structural and process indicators were used to assess the existing structures and processes in the national pharmaceutical system
Level II - Outcome indicators supported level I indicators by providing specific data about the important pharmaceutical outcomes
The household survey form complemented the levels I and II indicators by examining issues on access and use of medicines in the community. Levels I and II indicators are almost entirely focussed on health structures and people visiting health facilities. This survey tool ensures data are collected about treatment-seeking behaviour and medicines consumption.
The list of indicators is presented below. A description of the purpose of each indicator, together with instructions on how to collect, record and process the data can be found within the operational package.
Level I indicators
The level I core indicators were collected according to a questionnaire that was completed as part of a desktop exercise (Appendix 6).
Level II indicators
The level II indicators measure the degree to which Kenya is achieving the strategic pharmaceutical objectives of improved access, quality and rational use of medicines. Survey forms for these indicators can be found in Appendix 4.
Access
• Availability of key medicines in public health facility pharmacies, private pharmacies and regional warehouses
• Stock-out duration in public health facilities and regional warehouses
• Affordability of treatment at public health facilities and private pharmacies
• Percentage of prescribed medicines dispensed/administered to patients at public health facility pharmacies
• Access to medicines by households
• Average cost of medicines and related fees in public health facilities
• Price of key medicines in public health facilities and private pharmacies
• Adequate stock record keeping at public health facilities and regional warehouses
Quality and safety indicators
• Presence of expired medicines in public health facility pharmacies, private pharmacies and regional warehouses
• Adequacy of conservation conditions and handling of medicines in public health facility pharmacies and regional warehouses.
Rational use of medicines
• Percentage of adequately labeled medicines dispensed in public health pharmacies
• Percentage of patients who know how to take medicines
• Average number of medicines prescribed in public health facilities
• Percentage of patients receiving antibiotics in public health facilities
• Percentage of patients receiving injections in public health facilities
• Percentage of prescribed medicines on the essential medicines list at public health facilities
• Percentage of prescribed medicines prescribed by INN/generic name at public health facilities
• Availability of standard treatment guidelines (STGs) in public health facilities
• Availability of essential medicines lists in public health facilities
• Percentage of tracer cases treated with medicines recommended or discouraged in STGs
• Use of medicines by households
Sampling and survey population
Five provinces were selected from a possible eight. Nairobi and Eastern Provinces were chosen as the highest and lowest income-generating areas respectively. The other three provinces were chosen randomly, taking into account reasonable accessibility by the data collectors. The three randomly selected provinces were Nyanza, Rift Valley and Coast Provinces. A list of provinces and health facilities can be found in Appendix 3. In each province, the following units were surveyed:
• Six public health facilities treating outpatients and with pharmacy or dispensary units
• Six private pharmacies
• One regional medicines warehouse
• 150 households, divided equitably into those within 5km, those between 5-10km away and those more than 10km away from a surveyed health facility.
Training of the survey team
Adaptation of the survey forms for Kenya (see Appendix 5 for the list of key medicines selected for survey), training of 15 data collectors from both the ministry staff and civil society, location and facility selection and field testing were carried out during a training workshop held in Nairobi from 31st March through 4th April 2003.
Data collection
The survey of public health facilities, public warehouses, private pharmacy outlets and households was conducted between April 8 -17, 2003. Data were collected by five teams each of three data collectors, working concurrently in each of the provinces selected for the study, and all followed the procedures and approaches learned from the WHO operational package during the training.
Scope and limitations of the data
Sampling was done only in public health facilities leaving out private, NGOs and faith-based health facilities which provide a significant part of pharmaceutical services in Kenya.
Preventative aspects of medicine use was not given prominence in this survey and given the importance of this issue in Kenya, there is need for a study to capture access to essential vaccines and preventative pharmaceuticals.
To measure rational medicines use indicators, exit interviews with outpatients were used. The patients to be sampled are restricted to general illness encounters, representing a mix of health problems and ages. These indicators have limitations when applied to pre- and post-natal visits, specialist consultations, or even separate clinics for adults and paediatric cases because treatment practices are different.
Cases of non-bacterial diarrhoea in children under age 5 years were very few. In most instances, it was impossible to identify 10 cases at a facility. The reason for scarcity could not be established. There were also inadequate tracer cases obtained for mild/moderate pneumonia in children under age 5 years and non-pneumonia ARI in patients of any age.
The total number of cases reviewed in some primary health facilities was relatively low because of low patient numbers.
Since this was a cross-sectional study, retrospective and prospective sampling was combined. Retrospective sampling had limitations because some health facilities did not have sufficient records for random selection.
Lack of co-operation, particularly of non-pharmaceutical personnel charged with medicines stores management, caused delays in data collection in some health facilities.
Although in general the household interviews carried out were reliable, some difficulties were encountered in the Nairobi Region as would be expected in an urban centre (security, household members away at work, misunderstanding of symptoms, medicines use and expenditure estimates). Therefore the household results may not be representative of the national household situation.
