Characterization and Distribution of Medicine Vendors in Nigeria: A study of Lagos and Kaduna States.

In 2014, Nigeria issued the task-shifting/sharing policy for essential health services, which aimed to ll the human resource gap and improve the delivery of health services across the country. This study focuses on the characteristics, spread, and Family planning (FP) stocking practices of medicine vendors in Lagos and Kaduna, assessing the inuence of health training on the provision and stocking of FP services and commodities by vendors. We conducted a census of all patent and proprietary medicine vendor (PPMV) shops and followed up with a facility assessment among 10% of the mapped shops; utilizing an interviewer-administered questionnaire. Bivariate analysis was conducted using the chi-square test, and multiple logistic regression was used to estimate the adjusted odds ratio (OR) and condence intervals (C.I) for the test of signicance in the study.


Introduction
Across several low and middle-income countries (LMICs), national health systems have been functioning suboptimally due to recurring challenges within the health sector, with this particularly affecting the delivery of accessible and affordable healthcare services [1]. The health system, in most cases, is synonymous with publicly owned health facilities, with several important private and non-state actors being downplayed [2]. The functional capacities of the health system in these settings have gradually weakened, having failed to recognize and maximize efforts of all parastatal primarily devoted to improving health (including organizations, institutions, structures, and resources) [1,3].
Health workforce is an integral part of the health system and plays a critical role in achieving effective healthcare delivery. According to the World Health Organization (WHO), they are people "primarily engaged in action with the intent of enhancing health" diagnosing illnesses, healing, caring for people, monitoring health outcomes, supporting treatment adherence, providing medical information, and preventing diseases [4][5]. They consist of physicians, nurses, midwives, dentists, pharmacists, laboratory workers, environment & public health workers, community health workers, other health workers, and health management, and support staff [6][7]. The correlation between the availability of health workforce and positive health outcomes has been observed by several studies [1,8].
In 2007, there was an inadequate gure of 35 doctors and 86 nurses per 100,000 populations of Nigerians comparable to Sub-Saharan average of 15 doctors and 72 nurses per 100,000 population; a region that contributes to a quarter of the global burden of disease [5,9]. These gures have remained inadequate over the years; with Nigeria reporting a health workforce of 195 (doctors and nurses) per 100,000 ten years from the 2006 world health report [10]. This scenario has been linked to inadequate production and inequitable distribution across the country [3,[10][11]. With massive historical brain drain which already has been documented to have caused setbacks in the country's health system [12] and declining recent brain drain still being experienced in Africa (including Nigeria); especially of doctors (25%) [10] and nurses (5%) [13], the number of Community Pharmacists are also grossly inadequate [14] and hence the need for other cadres of staff to ll in the gap and improve healthcare (especially basic public health services).
Although not always recognized as front-line health workers, across many countries in Sub-Saharan Africa (including Nigeria), medicine shops offer an alternative, when there are shortages in the health sector (including human resource shortage) and are usually the rst point of care-seeking in most communities [15][16][17][18]. As reported by the National demographic health survey, nearly half (41%) of contraceptives users in Nigeria obtain their contraceptives from the private sector, with more than half of all mothers/guardians seeking care from this sector for childhood illness such as Diarrhea (54.1%) and Fever (57.7%) [19]. Major reasons noted for this occurrence in several studies include the absence of formalities, reduced waiting time at the Medicine shop, the proximity of the outlets to clients, and the perception of clients that they will pay more at a health facility [20][21]. Patent and Proprietary Medicine Vendors (PPMVs) are increasingly being recognized as important providers of health commodities, as interest continues to grow among policymakers and Program implementers regarding their engagement in Primary Health Care delivery in Nigeria [16]. In some settings, training medicine vendors to provide high-quality basic services, such as family planning (FP) services, treatment of common childhood illnesses and malaria, may offer a cost-effective way of delivering community-based health programs [17,20,22]. The Pharmacists Council of Nigeria (PCN) undertakes the registration, regulation, and licensure of PPMVs, conducting orientation programs for new PPMVs, continuing education programs for existing PPMVs, inspection, and publishing the Approved Patent Medicine List (APML) [23]. Studies have however shown a reluctance of medicine vendors to register with the regulatory body and preferring to register with their trade associations, which also provide drug stocking support, facilitate education and training, and give business and nancial assistance [18,[24][25].
In 2014, Nigeria issued the task-shifting and task-sharing (TSTS) policy for essential health services in Nigeria which aimed to ll the human resource gap and improve the delivery of health services across the country [26]. The policy highlighted the need to expand services to community-based personnel including Medicine vendors to provide treatment, counseling, and referral for some reproductive, maternal, and child health services (including FP) [26][27]. Previous studies have shown that different professionals may be operating within the PPMV space (even some with medical training [17,28]. Thus, the integration of this group of private providers into the formal healthcare system could increase access to high quality, primary health care services throughout the country [18].
In an initial study conducted in 2014, Liu and colleagues estimated the number of vendors in Nigeria and offered clues to drug stocking practices and how this group of healthcare workers could be better engaged to improve healthcare across the country [18]. However, it has been four years since the initial study and with the support recently showed to this group of health workers through the TSTS policy, there is the need to conduct further characterization and distribution of medicine vendors. This study focuses on the characteristics, spread, and FP stocking practices of PPMVs in Lagos and Kaduna states of Nigeria. It also assesses the in uence of health training and other factors on the provision and stocking of FP services and commodities respectively by medicine vendors.

Methodology
Between February and September 2018, a census of all Medicine shops in Lagos and Kaduna states was conducted. The census collected information on the basic characteristics of the shop, educational quali cation of the owner, and number/type of support staff working in the shops. Geographical coordinates of each shop were also recorded using the global positioning system (GPS) con gured into the electronic tool used for data collection; adjusted to an accuracy of 5 meters. The states were purposively selected because of a planned FP intervention, regional differences between northern and southern Nigeria and comparability of ndings with the Lui et al study conducted in 2014 (Lagos was one of the states chosen in that study and will show how things have changed whereas Kaduna will represent a new look and give insights about some states not represented in the 2014 study).
Data collectors worked with a list of PPMVs provided by the PCN, directorate of Pharmaceutical Services of Ministry of Health, and trade associations in the States. From the list, data collectors also snow-balled until all PPMVs in the LGAs were covered.
Following the census, a more robust facility assessment was conducted among a sample comprising about 10% of the mapped PPMV Shops; these shops were randomly selected from the 14 implementation LGAs systematically selected for the project in the two states (5 rural and 8 urban). The assessment aimed to elicit additional information not collected during the census. This assessment occurred between October and December 2018 and utilized an interviewer-administered questionnaire programmed using the Census and Survey Processing System (CSPro) into mobile phones through the CSentry Application.
A medicine vendor was regarded as having health (medical) training if he/she reported having quali cation listed as Medical Doctor, Nurse, Midwife or Pharmacist or having completed a training program as a Community Health Extension Worker or a two/three-year clinical training program as a Junior Community Health Extension Worker [18].
Major outcomes in this study included (1) Current stock of FP products, including any brand of oral contraceptives, injectable contraceptives (and DMPA-SC), emergency contraception, intrauterine contraceptive device; (2) registration with a trade association or the PCN; (3) Source of Supply of FP commodity (de ned as whether or not the vendor bought FP commodities from the Open Market). Major predictors in the study included health training of medicine vendor, training attendance (de ned as PPMV shops in which at least one person has ever attended training on FP products and services), and Location of the PPMV shop (de ned as Rural or Urban).
Other variables recorded in this study included the number of staff working in the facility, the estimated number of FP clients per day, and the length of time PPMV has been in business. The number of shops per 100,000 population in each local government area, using the 2006 national census estimates adjusted for population growth to 2017 was calculated. In addition, the percentage of vendors who had health training was calculated. Both parameters were shown graphically using maps culled from both states (showing LGAs).
Data were reported as frequencies and percentages to describe the various variables used in this study. Bivariate analysis was conducted using the chi-square test. Multiple logistic regression was used to estimate the adjusted odds ratio (OR) and con dence intervals (C.I) for the test of signi cance in the study. Statistical and spatial analyses were conducted using Stata v.14.0, and QGIS v.3.8, all tests were 2 sided and p-values < 0.05 were considered as statistically signi cant.
Two models were employed for the logistic regression; Model 1 examined the association between Health Training of medicine vendor and Registration, Stocking and Source of FP products while controlling for State, Location (Urban/Rural), number of staff, number of FP clients and length of time business started. Model 2 examined the association between Length of time business started and Registration, Stocking and Source of FP products while controlling for State, Location, number of FP clients, and health training of the vendor.

Results
We visited a total of 8,318 PPMVs in the study; the majority of these were in the urban areas (76.2%). Tables 1 and  2  There were 2,345 shops (28.2%) in which the medicine vendor (person in charge) had some form of health training. Figure 3 shows the proportion of PPMVs with health training in each of the LGAs across both states. The average percentage was signi cantly higher in the rural areas (32.2%) and in Kaduna State (37.9%) than urban areas (19.7%) and Lagos State (11.3%). About half (50.9%) of the shops were solely manned by the medicine vendor without assistance, about one-third (31.9%) had staff ranging from 1-3 persons and only a quarter of the shops had more than four (4) staff.  The assessment data also showed similar patterns with the census; with Kaduna (65.2%) and urban areas (84.4%) contributing largely to the sample.  Signi cant associations were also found between the length of time the PPMV business started and outcomes (except stocking of oral and emergency contraceptives). We also found a signi cant association between length of time business opened and health training of vendor; ndings show that a signi cantly larger number (56.3%) of newly opened PPMV shops are manned by those who reported to have health training; those newly opened shops are 1.8 times more likely to be health trained compared to those opened > 5 years before the study. After controlling for confounders in the multivariate analysis (shown in Table 5), many of the associations remained signi cant (especially those relating to health training of vendors). Consequently, there was a signi cant association between health training and the likelihood of vendor stocking oral and injectable contraceptives. Also, there is a signi cantly reduced likelihood of health trained vendors stocking from the open market compared to those not health trained.
Although there was an observed increased likelihood of newly opened PPMV shops to utilize the open market as the major means of stocking products; this relationship was not shown to be signi cant after adjusting for confounders.
Health training of the vendor did not prove to be a signi cant predictor of registration status although some differences exist. Those who are health trained seem to be more likely to register with the PCN than with national/local medicine dealers' associations compared to those not health trained. The same can be said for those who reported opening within 5 years of the study compared to PPMV shops which have operated for longer periods.

Discussion
The ndings from this study, like the Liu et al study among others, revealed that PPMV shops were numerous in the two states. With fewer health facilities and health posts opening across the country [29] and fewer pharmacies opening up (especially in rural and hard to reach areas) [14], PPMVs have remained the most widespread health structure across the two (2) states; and across Northern and Southern Nigeria [18]. However, the study found a greater concentration of shops in the Northern state of Kaduna and especially in the urban areas; this is a slight deviation from the norm in many studies that PPMVs are mainly located in the rural and hard to reach areas [16,22,30].
As Nigeria strives to achieve Universal Health Coverage (UHC) and improve access to quality health services, the role of PPMVs has become more important. One of the domains of access listed in the UHC includes physical access to health services; this physical accessibility is dependent on the proximity of health facilities to potential clients [31]. In a recent study where the distribution of health facilities across Nigeria was documented, primary health facilities, more than other facilities were most prominent across the country (ranging 15.5 and 18.4 per 100,000 population in Southern and Northern states respectively) [29]. The ndings from our study showed that the distribution of PPMV shops in these regions more than doubled that of the most populous health facilities; ranking 32 and 46 per 100,000 respectively in Southern state of Lagos and Northern state of Kaduna; this corroborates the ndings of Liu and colleagues that PPMVs may be more accessible than health care facilities in many places across Nigeria. The gures in this study revealed a considerable increase from the gures shown in 2014 where there were 32 versus 24 per 100,000 in the Southern states and a similar 17 per 100,000 in the Northern States between PPMV shops and health care facilities [18]. The ndings remained the same even when state comparison was done between this study and the study by Makinde and colleagues. In addition, the study found that PPMV shops have continued to grow progressively in the last ve (5) years; further increasing the margins between the numbers of PPMV shops and health care facilities and thus, becoming the most proximal health facility for potential clients for different health care services.
Also, there is an emergence of PPMVs reporting formal health training. These ndings have also been observed by a few studies in the past [15,18,28] and are still relevant today. This suggests a deviation from the general de nition of medicine vendors, who are characterized as being able to read and write [23], having completed mainly primary education and not having formal training in medicine and pharmacy [32][33]. Many of them were also thought to learn the trade through an apprenticeship program before opening their shops [16]. Our ndings show that more than half of the vendors who reported health training are relatively new in the business, opening within the last ve (5) years. This suggests the rise of vendors who may be able to deliver high-quality health services and complement the existing healthcare infrastructure [17] due to their formal training and previous experience (some being retired providers who worked at public/private health facilities across the country). The ndings from this study show a signi cant relationship between health training of vendors and stocking of some FP commodities (especially Injectable contraceptives) with clients that patronize them for FP services every month (this supports ndings elsewhere in Nigeria and Africa that PPMVs already offer injectable contraception) and other services [27,34]. This supports the evidence that PPMVs can support the formal health systems as these are services that they can provide, because of their training and the recent task-shifting and task-sharing policy of the country. Unfortunately, some services being provided are currently prohibited within the scope of PPMVs [17,26,30,32]. As many other studies have suggested, this study found that concerns may exist in the delivery of quality services by PPMVs as many of those interviewed stocked illegal commodities (beyond their legal scope of practice) and refused to register with the pharmacy practice regulatory body (PCN) [16,18,30]. Despite the ndings, studies have suggested that demand (especially women and adolescents) could be driving the stocking and sale of those categories of drugs especially when the drugs are not available in the formal health facilities [30,35]. This, in addition to the geographic spread of the medicine vendors, has implications for the healthcare system in Nigeria and necessitates the expansion for more services (including FP) to PPMVs beyond their current scope of practice especially for those reporting formal health training, as have been demonstrated in many pilot studies across Africa (including Nigeria) [22,[36][37][38][39][40]. This will be in line with the TSTS policy of the country which calls for an increase in the capacity of community people to provide some reproductive, maternal, and child health services [18]. Strengthening the collaboration between this informal sector and the formal health sector may go a long way in improving the con dence of medicine vendors to legitimately stock drugs that they are capable of providing according to their quali cation and improving referral for higher levels of care sought by the vast population that patronize them.
The ndings of the study also show that health training of a PPMV and age of business existence may in uence registration status with the PCN, although this was not found to be signi cant after adjusting for the state, location, and the number of FP clients. There is a possibility that newly opened shops with health trained vendors are more likely than others to register with the PCN, suggesting a possible improvement in the relationship between PCN and the association of patent medicine vendors in the last 3 years. This nding shows that health trained vendors may more likely adhere to regulatory and quality assurance guidelines; thus, corroborating the ndings found by other investigators [18,30]. This also suggests that newly opened vendors may be willing to register with the PCN at the start, but the in uence of older PPMVs and the majority of who are without health training may further drag new PPMVs into not desiring to register annually with the regulator but rather with the trade associations, or not registering at all, hence operating illegally. As noted by several studies, the trade associations are perceived to be more bene cial to the vendors than PCN for several reasons, ranging from providing them with protection against law enforcement agents to oversight functions (especially with preventing illegal drug sellers and monitoring of the sale of drugs) to members through a taskforce and peer-to-peer mentoring; these are generally more acceptable by the medicine vendors [24,30]. There is thus a need for a collaborative framework between the associations and PCN to improve registration of PPMVs and their perception about the roles of PCN against the current antagonistic standpoint. PCN needs to be seen to perform more supportive roles to foster trust and respect; which could lead to a suitable accreditation system, an opportunity for professionals to expand their scope through continuous learning, and ultimately have the drug list expanded to t the realities of the sector. This may result in more health trained vendors joining the business and offering a great opportunity to further improve the quality of health services provided to the populace, leading to better health outcomes.
Lastly, our ndings support the claim of many studies that many vendors utilize the open market as the major source of supplying health products [17][18][41][42]. Even though the National Drug Distribution Guideline has as part of its core deliverables, the elimination of the dominance of unregulated drug markets in major cities of Nigeria [43][44], however, the ndings from our study show that these markets still supply the majority of the vendors (especially in the northern part of the country). In the southern part of the country, reliance on open markets is common among those without health quali cations and in the rural areas (where mega dealers/wholesalers are few). Drug distribution requires e cient supply chain systems and appropriate regulation to ensure that the medicines that reach the consumer are in their intended qualitative state, supported with the required infrastructure to ensure rational use; this, however, is not the case across both states as vendors still source for drugs in these markets where drugs are peddled without caution. For the most part, only the public and few private health facilities are catered for by the national drug distribution centers (MDDC and SDDC) [45].
This study has several limitations; chie y, by combining the results of two studies conducted at different times, there is a possibility of missing some key characteristics from PPMV shops not included in both studies, especially as the second study (facility assessment) focused on PPMVs in proposed implementation LGAs. Also, recall bias is a possibility with this study type, as some of the variables self-reported by PPMVs could not be veri ed in both studies, only GPS coordinates, licensing status with the PCN, and product stocking could be independently veri ed.

Conclusion
PPMVs have continued to grow progressively in the last ve years, becoming the most proximal health facility for potential clients for different health care services (especially FP services) across both northern and southern Nigeria. they are now comprised of a considerable mass of health trained personnel, able to deliver high-quality health services and complement existing health care infrastructure if trained. However, restrictions on what is possible within the PPMV shop and lack of access to quality health commodities have resulted in some poor practices among PPMVs. There is therefore a need to identify, train, and provide innovative means of improving access to quality-assured products for this group of health workers. This may signi cantly reduce the stocking of inadequate or substandard medicines, reduce the stock-out of key health commodities, and further improve the quality of services provided by these providers and ultimately, the health outcome of the populace.