Context

Nigeria was chosen as the country in which medicines will be collected for this study. Nigeria is a developing country with a large population with previously demonstrated availability of substandard antimalarial medicines.18 The prevalence of hypertension in Nigeria is currently estimated to be 28.9% and this contributes a substantial proportion of the total burden in Africa due to the large population of the country which is currently estimated to be over 170 million.19 This increasing burden is increasing the need for government and other health policy planners to ensure the availability of high-quality evidence-based medicines for hypertension.20 Nigeria has an area of 923 768 km² (about 574 000 square miles).21 It is the most densely populated country in Africa and represents about 20% of the total population of sub-Saharan Africa.21 Although Nigeria has a large number of ethnic groups (∼250), the three most populous are the Hausas in the North, Yorubas in the West and Ibos in the Eastern part of Nigeria. Nigeria has a total of 36 states and 774 local government areas (LGAs) that are subordinate to the relevant states and the Federal Capital Territory. According to the WHO, over half a million Nigerians died from non-communicable diseases, including CVDs, in 2012 and that 35% of Nigerian adults had elevated BP in 2008.22 Other population-based research reports that the prevalence of hypertension in Nigeria is around 23% of the adult population with rates being higher in urban than in rural dwellers (32.7% and 12.9%).23 For this study, we have purposively selected a state from three geopolitical zones.

Identification of antihypertensive classes for collection

To identify which generic brands to include for sampling, we initially identified the most commonly prescribed brand and antihypertensive for each of the five primary classes of antihypertensive medicines available in Nigeria (table 1). The major classes identified were (calcium channel blockers, β blockers, ACE inhibitors, angiotensin receptor blockers and centrally acting drugs) based on information obtained from local opinion leaders and from publically available information via the National Agency Food and Drug Administration and Control (NAFDAC) agency.24 Based on the classes and brands identified (listed in table 1), a detailed and comprehensive list of all the available generic brands was prepared based on data publically available through NAFDAC.24 It was decided that the project would focus on the two most commonly prescribed antihypertensive classes and these were amlodipine and lisinopril. A computerised random sequence was then used to randomly select 10 generic brands of amlodipine and lisinopril. These 20 generic medicines are available in Nigeria and will form the basis of the samples for collection.

Geographical sampling framework

A stepwise and semi-random sampling framework was developed to determine which retail outlets would eventually be included in the study. The purpose and concept behind the sampling framework is to identify a combination of urban and rural locations in northern and southern states of Nigeria via a stratified yet random and systematic approach. The steps involved in the sampling framework are as follows:

1. Identification of three states representing varied geopolitical areas through consultation with local experts (purposive sample).

2. For each of the three states, we will then obtain a list of LGAs from which we will randomly select one rural and one urban LGA. Therefore, across the three states, a total of six LGA will be included for sampling (three rural and three urban).

3. It was not feasible to source a list of all pharmacies in each of the LGA, so we will use geographical mapping software (QGIS Geographic Information System, V.2.8.3-Wien, 2015) to randomly select four geographical coordinates in each of the selected LGAs.

4. The nearest available pharmacy outlet to each of the geographical coordinates will then be identified by the research assistant who will act as a mystery shopper and purchase the predetermined generic medicines from each shop. The total sample will therefore include 24 retail outlets (pharmacy shops) from six LGAs across three states in different geopolitical zones of Nigeria.

Medicine purchasing, processing storage and transport

Once the final list of 24 retail outlets is confirmed, a research assistant will act as a mystery shopper (a hired research assistant who speaks the local language) who will visit the identified outlet and systematically purchased all the available and preselected generic brands for each of the chosen antihypertensive medicines. The covert approach is to minimise the risk that a seller might realise they are to be a part of a research project and therefore raise suspicion and minimise the risk that the seller declines to sell the required samples. Trained research assistants within our existing networks will collect the medicine samples from the pharmacies and will also record the exact geographical coordinates of the pharmacy. Researchers will pose as clients to buy the 20 samples from each of the 24 pharmacy shops (total of 480 samples). A full box (usually 28 or 30 tablets) for each generic medicine will be purchased so the number of dosage units in a box will be the number purchased.

Once purchased, each sample will be retained in their point-of-purchase packaging and placed in an individual zip-lock bag together with a completed standardised data collection form. The data collection form for each sample will include recording of the outlet type, date of purchase, price paid, brand name, formulation, batch number, manufacturer, number of tablets per sample and expiry date. Samples will then be safely and securely stored in an air-conditioned room (∼20°C) in a central location at one of our researcher's sites, pending dispatch. As such, efforts will be made to ensure that the medicines are not degraded by inadequate storage conditions, which could confound the findings. As soon as all samples have been collected, they will be dispatched, via international courier (eg, DHL Express), to the London School of Hygiene and Tropical Medicine (LSHTM) in London, UK for further processing and chemical content analysis.

Sample processing and laboratory analyses

For every sample under investigation, a digital photograph of the packaging and contents of each sample will be taken. Sample information (including name, stated active ingredients, dose form, outlet, district, date of purchase, manufacturer, NAFDAC registration number, batch number, expiry date, number of tablets per packet and price paid) will be compared to the data collected for each sample during the fieldwork collection and logged on an excel sheet. The integrity of the packaging, the features of the blister pack with indented or printed batch number, and dates of manufacture and expiry, as well as the appearance of tablets, form the first steps in determining the quality of the product. Included on the package should be the list of active ingredients, name and address of the manufacturer, storage conditions, batch or lot number, date of manufacture and directions for use. Enclosed with the tablets should be an instruction leaflet in the appropriate language without any spelling errors. All tablets will then be weighed, and their dimensions, colour and physical appearance characteristics (such as any cracks, logos and chips) recorded prior to laboratory analysis.

Chemical content analysis

Quantitative analysis will be conducted at the LSHTM to measure the amount of stated API in the tablets using high-performance liquid chromatography (HPLC). We will use the published methods in the USP monographs for the analysis of tablets of amlodipine besylate and those of lisinopril.25 ,26 Tablets will be dissolved in a solvent, sonicated and centrifuged. The supernatant will then be injected into the HPLC column and the amount of API present in each tablet will be determined and expressed as percentage found. We will also consult USP monographs for the two antihypertensive medicines to see what the tolerance rules stipulate which is generally considered acceptable quality if the API measured is 90–110% for amlodipine besylate tablets and 98–102% for lisinopril tablets.

Sample size and analysis

Laboratory data will be combined with sample data and analysed using STATA (Stata/IC). Descriptive statistics will be used to compare the generic medicines on their physical and chemical characteristics with the reference product (medicine). In addition, we will compare equivalence in rural versus urban areas. Proportions will be evaluated using χ² tests and associations between poor quality and high-quality antihypertensive medicines will be explored. The USP recommends 30 dosage units for a single tablet/capsule medicine of the same lot number from each location. Such a sample size gives enough dosage units to determine identity and accurate estimates of content of active ingredients.

Ethics approvals

The study will seek ethical approval from the LSHTM Ethics Committee and the relevant Nigerian state-based ethics committees. This study does not involve patients and hence written informed consent will not be required.