A descriptive cross-sectional mixed method study was undertaken among midwives in Kisii and Kajiado Counties in Kenya. According to Kenya National Bureau of Statistics (KNBS) Kisii County has a population of 1,266,860 while Kajiado has 1,157,873. The two counties are largely rural but includes peri-urban areas. The midwives in the two Counties are homogenous professionally given that they undergo the same training. A list of all primary health facilities were stratified by region (Kajiado and Kisii) and by type of facility (public and non-public). The facilities were then purposively selected to ensure representation of views both from the public and non-public primary health care facilities in the two Couties. Having obtained a list of the nurses in the respective primary health facilities, a systematic sampling was used in selecting the nurses mid wives to be interviewed. Although Kajiado county is closer to the National Capital City County Nairobi as compared to Kisii county, both Counties have limited access to basic diagnostic technology for improved perinatal care in Primary Health Care settings.

A questionnaire, focus group discussions (FGD) with nurse mid-wives and mothers attending ante-natal care in Kisii and Kajiado Counties was conducted. Six research assistants, fluent in English, administered the questionnaires, facilitated FGDs and KII, and obtained informed consent from participants. The research assistants who were to have the ability to use computers and mobile phone applications, had a training in health or social sciences and were familiar with the respective regions had a two day training for this study. The questionnaire and consent forms were pre-tested with 27 respondents and revised based on feedback received.

A mobile application tool Open Data Kit (ODK) was used to collect quantitative data from the nurse mid-wives.

All the respondents for the FGDs and KII were purposively selected based on their positions, knowledge and their availability to participate in the study. A total of four FGDs were conducted among midwives. In addition, Key Informants Interviews (KIIs) was carried out among two representatives of the nursing council of Kenya (NCK) and society of radiography in Kenya (SORK). One KII each was conducted to a member of the mid wives association of Kenya, representative of the Obstetrics and Gynecologists society of Kenya and respective Chief County nurses in the two Counties. Information collected through FGD and KII sought to compliment the questionnaire data in the gaps that exist as far as the knowledge and skills of the midwives on Point of Care obstetric ultrasonography screening (POCUS) at the Primary Health Care level. The qualitative data also described the obstetric ultrasonography knowledge and skills gaps among midwives based in primary health care facilities.

Each FGD lasted for 45 - 60 minutes and was audio taped after getting written consent from the participants.

The study protocol, data collection tools, and consent forms were reviewed and approved by Amref Ethics and Scientific Review Committee (ESRC). Further authorization was obtained from the relevant administrative authorities in Kajiado and Kisii Counties in Kenya.

Being a snap survey is possible to have omitted respondents with the most reliable information

As shown in Table 1, Kisii County comprised of 69% of the respondents with 73.4% being female while 86.9% had a degree as their highest education level. More than 80% of the respondents were not satisfied with the income they receive from professional employment with only 27% indicating that they had an income above (USD 500). Public facility had the majority of the respondents (80.3%) with nearly two thirds working in level 3 facilities. ANC clinic and labour ward had the highest deployment areas in terms of the midwives at 48.9% and 32.1% respectively. Access to ultrasound services in the two Counties was low with 20.4% access in Kajiado while Kisii had less than 10%. Only 17.9% of the respondents from public facilities had access to an ultrasound machine as depicted in Table 2 but it was also reported that sometimes they are dormant for lack of technicians to operate them.

A whopping majority comprising of 260 (94.5%) of the midwives responded that they have never been trained on any basic obstetric ultrasound screening skills. The remaining 15 (5.5%) indicated that they had earlier been exposed to ANC ultrasound screening. Six (6) and nine (9) of the fifteen were from Kisii and Kajiado counties respectively; they had been trained in early stages of a midwives entrepreneurship project offered by Amref but they lacked equipment to practice (Table 3). In general midwives from both Kajiado (Peri-urban) and Kisii counties (rural) expressed willingness to learn the basic ultrasound screening skills but on bivariate analysis there was no association between type of county and desire or motivation to be trained.

The midwives were further asked to specify the skills they wish to learn on ultrasound screening and their feedback is as presented in Figure 1. In general the highly ranked skill the midwives wanted to learn is how to operate an ultrasound machine (98%), confirm gestational age of a pregnancy (91%) and fetal viability (91%). A total of 269 (98.2%) of the respondents said they were very motivated to learn the basic obstetric screening skills and they further Table 4 gave various reasons why they are motivated. On perceptions mothers may have, 271 (98.9%) of the respondents believed that mothers will have no problem trusting the basic ultrasonography screening services being offered by a midwife.

More than three quarters (77.74%) of the respondents opined that they wished to provide personalized care services to ANC mothers/clients in their community at a fee with the respondents listing birth preparation (95.8%), nutrition education (94.4%) and Post-natal education (92.5%) among the additional services

they could provide to ANC mothers in a personalized community care arrangement. Only 13.9% of the 274 respondents had knowledge that obstetric screening should be done before 24 weeks gestation Figure 2. Notably, 110 (40%) of the respondents on a scale of 1 to 5 (one being the least), scored themselves a one (1) on the level of confidence they have using a ultrasound machines.

Significant difference were also observed between the Counties and ANC clients screening and monitoring practices at service delivery points; the factors included whether ANC clients are referred for ANC profiling (χ² = 14.05, P = 0.000, 0R = 0.36, 95% CI 0.21 - 0.62), if pregnant women in the facilities have access to ultrasound screening services (χ² = 73.92, P = 0.000, 0R = 0.86, 95% CI 0.05 - 0.16), ability of mother to pay for ultrasound (χ² = 13.91, P = 0.000, 0R = 0.35, 95% CI (0.20 - 0.61), and number of mothers refered (χ² = 18.59, P = 0.000, 0R = 0.31, 95% CI (0.18 - 0.53).

The health workers in Kisii County were less likely (0.36 times p-value 0.000) to refer ANC clients for ANC profiling compared to the health workers in Kajiado.

Pregnant women in Kisii facilities were less likely (0.86 times p-value 0.000) to have access to ultrasound screening services compared to their counterparts in Kajiado with mothers in visiting Kisii facilities less likely (0.35 times p-value 0.000) to have ability to pay for the ultrasound services as shown in Table 5.

In addition, the study found that 146 (65.5%) of the midwives working in the public health facilities never access an ultrasound for the clients compared to only 10 (4.5%) and 6 (2.7%) working in private for profit and private for non-profit facilities.

Although 18% of the respondents from public facilities indicated that they had access to ultrasound machines in their facilities, the midwives further observed that mothers who seek ultrasound services are mainly in their 2^nd Trimester (72.62%) and 3^rd trimester (26.19%). Mostly mothers undertake an ultrasound screening upon referral from lower levels of the health system especially when they suspect a fetal death or abnormal hemorrhage (Figure 3). Notably, 58.8% of the respondents do refer such mothers to levels 4 and 5 public health facilities (Figure 4) because they are cheaper as compared to private facilities where an ultrasound service costs USD 15 to USD 250. It is only 25.9% of the respondents who believed that obstetric ultrasound screening should be done routinely for pregnant women; 32% of the respondents indicated that ultrasound screening should only be done when a complication is suspected. Nevertheless nearly all respondents (98.18%) thought it is very necessary for ANC mothers to access obstetric ultrasound screening services in their facilities.

The female were the majority in this study and this confirms previous findings that women’s share of employment in the health and social sector is high. However, according to a report done by WHO in 104 countries, there exist systematic differences in gender distribution by occupation across all regions. According to

the report, male workers are significantly represented as physicians, dentists and pharmacists in the workforce, with female comprising the vast majority of the nursing and midwifery workforce [8].

Eight out of ten service providers worked in public sector. Our results are similar to those of African studies [9] [10]. This could be due to the fact that the public sector remains the major employer of health workers and the high number of public facilities both in Kisii and Kajiado.

The lack of trained providers in point of care ultrasound is said to be a barrier to its utility in middle income countries leading to a demand for training courses directed at filling this need [11]. Training of health personnel is a key factor to implementation of quality obstetric ultrasound services since better management of patients relies on the ability of health personnel to use the machine proficiently and interpret findings accurately [12].

According to The International Society of Ultrasound Obstetrics and Gynecology (ISUOG), basic ultrasound examination should be done to investigate the fetal viability and fetal movements, detection of multiple pregnancy, assessment of gestational age and fetal size by recording biometric measurements, evaluation of amount of amniotic fluid, placenta location and fetal position [13].

According to WHO, there should be a minimum of 3 - 6 months of training accompanied with 300 - 500 examinations for the general ultrasonography curriculum [14]. At the same time, multiple studies have concluded that health professional, mainly midwives can be trained to perform point of care ultrasound to identify high risk pregnancies that can lead to allow cost improvement strategy in maternity services [2] [6]. This is consistent with our study in which almost all midwives believed that a nurse-midwife can learn basic obstetric ultrasound screening skills.

The training of the midwives on basic obstetric ultrasound remains very low among the respondents in this study with 9 out 10 confirming that they are yet to receive training on basic obstetric ultrasound. The figures are quite higher compared to a report in which 38.3% of service providers were found to have no training at all [2] [12]. Studies show that insufficient training for healthcare providers is the primary barrier to regular use of obstetric ultrasound [15] with a qualitative study done among Rwandan midwives showing that apart from offering a formal training for physicians, there is need for introduction of ultrasound training for midwives in order to decrease the burden of work in hospitals and to improve access to ultrasound at healthcare centers particularly in rural areas [7]. The motivation to learn basic obstetric ultrasound was high among those yet to learn and this is consistent with a study by Shah [15].

Our study shows a shortfall in the level of knowledge on the ideal time to perform a fetal biometry and this is higher compared to a study in Guinea in which 69.2% had no idea of the ideal time to perform a fetal biometry. The higher figures might be due to the fact that the study was done among midwives unlike the study in Guinea that was done majorly among obstetric gynecologists [16].

Our study found a significant difference in referral for ultrasonography, access to ultrasonography services, and the ability to pay for the services in the two Counties.

On access to ultrasound services for their clients, our study was consistent to a cross sectional study done in Rwanda in which participants from public health facilities were less likely to report access to obstetric ultrasound when needed compared to participants from private health facilities [15]. The study is also consistent with a surveys conducted across Canada and the United States that identified a lack of equipment, training and inability to maintain skills as major barriers to point-of care ultrasound (POCUS) [17].

Kisii being a rural set-up, the midwives reported that their clients are less likely to access ultrasound unlike those in Kajiado that is peri-urban. In the study done in Rwanda, participants employed at health facilities outside Kigali area were less likely to report access to obstetric ultrasound when needed than participants employed at health facilities in the Kigali area [3]. According to a 25 question online survey from 44 countries, some of the barriers to use of ultrasound include lack of training, lack of equipment, ultrasound machine malfunction and lack of maintenance [15].

A majority of the midwives indicated that they would love to learn basic ultrasound screening services and this is consistent with a qualitative study done in Rwanda [18] in which physicians interviewed believed that it would be one important strategy to improve care for pregnant women particularly the poorest in rural areas who were less likely to be seen by physicians during pregnancy, to reduce the burden on physicians and to reduce delay in decision making. In addition, with the shortage of physicians in rural areas, means that midwives or nurses are to manage complicated deliveries, putting a high responsibility on their shoulders.