*3.4. Acceptability of Clinics*

Approximately 96% (n = 43/45) of caregivers interviewed at the market clinic and 87% (n = 54/62) of caregivers interviewed at the fishing community clinic reported that they planned to return for future services. Additionally, 98% (N = 105/107) of exit interviewees stated that the container clinics made it easier to receive child health services. The top reasons mentioned for why services were easier to receive included: "the clinic is easier to get to than other clinics" (98%), "more suitable hours" (27%), and "immediate attention" (22%). Over three-quarters (86%; N = 92/107) of the caregivers across both sites reported that the container clinic services were either better or equal to those received at other clinics (Table 1).

Despite the high levels of acceptability among the exit interviewees, caregivers reported the need for more services. Among the market clinic caregivers, 24% (n = 14/45) reported that the clinics should provide antenatal care (ANC) services. Of the caregivers attending the fishing community clinic, 23% (n = 14/62) also suggested the addition of ANC services and many recommended that the clinic adds prescription services (40%; n = 25/62). Among the nurses, language barriers were also noted, which may affect the acceptability of services. The desire for expanded services and the challenge of language barriers were corroborated in the qualitative findings:

*So since it is a clinic I think if it is expanded a bit, bring in more workers and increase the facilities so that it won't be a small place that only dishes out para [cetamol] for a headache then you have to go again to Polyclinic; but it should be a permanent place where if I have stomach pains I can be treated, admitted and if I need infusion I should be given so, if it is expanded a bit it will help us.*—Caregiver, the market community

*Language barrier because most of them are from these French countries and the north. So they don't speak English and they don't understand the Twi too unless their hometown language or sign language and when you do the sign language they understand or sometimes they come along with other people who understand the Twi and their language.*—Nurse, the market clinic

### **4. Discussion**

Our results suggest container clinics are an acceptable strategy to improve access to routine immunization services for the two urban communities in Ghana, and possibly in other similar urban areas. Following Peters et al.'s (2008) conceptual framework [5], we found that the placement of the container clinics within these communities addressed the availability, geographic and financial accessibility, and acceptability of routine immunization services. During the first 12 months of implementation, immunization sessions increased from monthly (when the clinics were only outreach sites) to daily; all recommended vaccines and supporting supplies and tools were stored on-site, and caregivers reported high satisfaction with both clinics. Improvements in geographical and financial accessibility were observed by minimizing commuting distance, time, and the indirect costs of lost wages for caregivers.

Implementing the container clinics required transforming the locations from outreach posts to fixed routine immunization sites. Transforming the sites into functioning clinics was essential because the outreach sites where the container clinics were built would cancel services due to poor weather that impacted vaccination sessions. Lack of resources, supplies, and tools are well-documented barriers to fully vaccinating children and also have implications for building trust with caregivers to return for future vaccination services, as well as the ability of health workers to effectively trace defaulters [17–20].

The utilization of vaccination services gradually increased as the container clinics became more established over the 12 months. In addition to the three-fold increase in vaccine doses administered during the first 12 months, both clinics exceeded their annual target population for the administration of MR1 and MR2 vaccine doses. Vaccine utilization is driven by a combination of demand and access factors, which this evaluation was not designed to disentangle. However, it is important to note that, in other studies, strategies for increasing demand include improving service frequency, design, and delivery [17–19,21] factors that the container clinics addressed. Thus, we hypothesize that increased community demand for services due to the availability of a geographically proximate, brand new clinic was likely a key factor underlying the observed increase in the utilization of services at these sites.

While most studies have assessed geographical accessibility in the rural poor, recent studies have identified how urban caregivers experience limited access, especially when the distance to the health facility is more than 1–2 km from their home [20,22]. We found that container clinics can provide immunization services to previously hard-to-reach urban populations—nearly every exit interviewee stated that the clinic made receiving care for their child easier compared to accessing previous health service sites. Most caregivers from the exit interviews reported that they walked to the clinics, and data from the caregiver FGDs indicated that the proximity of the clinics was favorable.

Previous studies have found that employed caregivers faced higher opportunity costs when bringing their children to the clinic for vaccinations because of the loss of potential daily earnings [9,20,21,23,24]. In our evaluation, the market community was chosen to provide services to a notably vulnerable group, the kayayei, or head porters. Typically, these head porters are young, rural-urban migrants from the northern part of Ghana [12,25,26] who may forgo accessing health services due to cultural discrimination [26] or because of financial barriers related to lost wages [12]. Our results showed that the majority of the caregivers at the market clinic were indeed head porters or other migrant traders. Interestingly, more market caregivers reported they would have had to miss work if not for the container clinic than mothers in the fishing community did. Thus, placing container

clinics near markets may address unique barriers faced by working caregivers, including head porters and migrant traders.

Finally, our results support the growing literature on how the urban poor are not a monolith and require tailored urban immunization strategies to achieve success [2,4,16,21,27]. The fishing community clinic served a stationary population with higher levels of unemployment and education, while the market clinic served a highly mobile population of working mothers with very little education. These differences influenced not only caregiver expectations of the clinic but also the overall acceptability of the clinic; we observed an almost 10% difference in acceptability when caregivers were asked if they planned to return to the container clinic or seek services elsewhere, with more market caregivers indicating they planned to return.

Acknowledging the diversity of urban contexts is important because although timely, the current focus on urban health is in danger of overlooking the nuances and unique challenges faced by the multitude of urban populations in Africa. A recent paper highlighting the diverse urban contexts in Johannesburg also argues that 'place matters':

*Whilst the challenge of addressing the health of urban populations within developing countries is acknowledged, the diverse urbanisation experiences of different urban groups remain under-explored* ... *for action to improve the health of poor urban populations to be successful, urban policy makers and programmers need to understand the complexity of the urban context* [28]
