To better understand factors contributing to underutilization of laboratory services for health care delivery in sub-Saharan Africa, we conducted a study in Senegalese Antenatal Care clinics (ANC) and laboratories to determine the extent of underutilization, contributing factors, and bottlenecks in the cascade of care from first ANC visit, test uptake, to availability of test results and appropriate clinical management.
At 16 health facilities, pregnant women attending for their first ANC visit were consecutively recruited and information was prospectively collected on the request, execution, results and clinical management of seven nationally recommended laboratory screening tests for normal pregnancy: hemoglobin concentration (Hb), syphilis serology, HIV serology, determination of proteinuria (PU), determination of blood group and Rhesus factor, Emmel test to detect sickle cell disease, and glycaemia. Health facility staff were interviewed on human resource capacity, management of the ANC and the laboratory, and availability and use of guidelines.
Of 1246 ANC attendants, 400 (32%) had complete results. Completeness varied between facilities from 0–99%. In multilevel logistic regression analysis of women nested in facilities, complete uptake was lower if women started ANC later in pregnancy; very low in rural ANC attendants who ever delivered compared to urban primigravidae (OR 0.064; 95%CI 0.00–0.52); and higher if the facility routinely recommended all seven tests. In the cascade from test request to clinical management, the most frequent bottleneck was non-execution of requested tests, while unavailability of results for executed test was uncommon (<2%). Overall, of 525 abnormal test results 97(18%) had a record of adequate clinical management.
Our study illustrates challenges to test uptake even when laboratory testing capacity is in place, with large differences between facilities, and underscores the importance of management, policy, and the importance of considering local context in order to improve service delivery to expectant mothers.
Citation: van’t Hoog AH, Sarr A, Koster W, Delorme L, Diallo S, Sakande J, et al. (2020) A study to better understand under-utilization of laboratory tests for antenatal care in Senegal. PLoS ONE 15(1): e0225710. https://doi.org/10.1371/journal.pone.0225710
Editor: Charles A. Ameh, Liverpool School of Tropical Medicine, UNITED KINGDOM
Received: April 6, 2019; Accepted: November 11, 2019; Published: January 9, 2020
Copyright: © 2020 van’t Hoog et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the Netherland Organization for Scientific Research, Science for Global Development (NWO/WOTRO), grant W07.4.203.00. to PO. https://www.nwo.nl/en/about-nwo/organisation/nwo-domains/wotro The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
It is increasingly recognized that inadequate laboratory infrastructure in Low- and Middle-Income Countries (LMICs) hampers the contribution of laboratory diagnostics to evidence-based clinical care. This recognition has resulted in laboratory capacity building initiatives[2–4], and in the development of simple, rapid, and affordable point-of-care tests (POCTs), which do not require electricity, a laboratory, or highly trained staff. However, other factors also contribute to underutilization. Antenatal care (ANC) provides a relevant model to study factors contributing to utilization of laboratory testing for health care delivery.
The aims of ANC include prevention and alleviation or management of health conditions that are known to have an unfavorable outcome on pregnancy. ANC involves laboratory screening of infectious diseases and conditions like anemia and pre-eclampsia that are associated with increased maternal and/or newborn morbidity and mortality. The World Health Organization (WHO) has defined a minimum ANC package (focused ANC) for normal pregnancies, which is based on four goal-oriented visits. The increase in ANC coverage is considered a success-story in Africa with 80% of women in sub-Saharan Africa receiving at least one ANC visit, however 52% completing 4 visits. Relevant indicators for the utilization of laboratory screening tests in the context of ANC are scarce, with the exception of syphilis for which 59% of pregnant women in sub-Saharan Africa were tested in 2010 .
The WHO guidelines for focused ANC recommend testing for hemoglobin concentration, syphilis, HIV infection, proteinuria, blood group/Rhesus factors determination and bacteriuria , tests for which testing capacity is widely available in LMICs. The first four are recommended as a minimum for LMICs . Countries are expected to adapt the WHO recommendations to their own epidemiological profile and priorities and implement national ANC guidelines with country specific detailing of the essential components of ANC.
We conducted a transdisciplinary mixed methods study with the overall aim to obtain a better understanding of the underutilization of the laboratory for antenatal care in three West African countries, Senegal, Burkina Faso and Mali (the study was called SociaLab ). Here we report on Senegal and describe the extent of underutilization of laboratory tests, contributing factors, and bottlenecks in the cascade of care from first ANC visit, test uptake, to availability of test results and appropriate clinical management. A qualitative analysis to further understand reasons underlying these patterns, and the activities to translate study findings into practice are reported elsewhere . With 96% of pregnant women in 2014 attended ANC at least once, ANC coverage is high in Senegal compared to sub-Saharan Africa (SSA) in general, and 48% had 4 ANC visits, which is slightly lower . Testing capacity for ANC screening tests is widely available. Maternal mortality was estimated at 315 death /105 live births in 2015, declining from 427 in a decade  and neonatal mortality estimates declined from 30.8 to 21.3 per 1000 life births . Syphilis testing in ANC Senegal was reported to be 5% in 2013, 11% in 2014, and 30% in 2015 .
Setting and laboratory tests
Senegal is a low-income country of West Africa with a gross national income of USD 950 per capita, a population of 15.4 million of whom 47% live below the poverty line. The adult literacy rate is 43% .
The policy of the Senegalese ministry of Health on ANC laboratory tests is in line with WHO recommendations . National guidelines recommend seven mandatory tests for normal pregnancy : hemoglobin concentration (Hb), syphilis serology, human immunodeficiency virus (HIV) serology, determination of proteinuria (PU), determination of blood group and Rhesus factor, Emmel test to detect sickle cell disease, and glycaemia. The national guidelines include normal ranges and recommended clinical management in case of critical results (Table 1). The seven recommended tests are to be requested and completed at first ANC visit. During following visits, repeat testing is on indication only (such as repeated proteinuria or glycemia in case of initially high values). Tests with lifelong result validity such as the blood group/Rhesus factor determination and the Emmel test are not supposed to be repeated once the result is known, but their status should be recorded on the ANC cards. The costs of the ANC consultation and the laboratory tests are covered by the patients, except for the rapid test for HIV serology, which is provided for free. Tests available as rapid format (HIV serology and proteinuria) can be delivered either at the laboratory facility or at the point of ANC, at the discretion of the health facility.
During the first ANC visit, a pregnant woman is usually provided with a test request slip, which she is expected to take to the laboratory herself, once she is ready to pay and have her sample(s) collected. The laboratory hands over the results to the pregnant woman, usually on the same day or the day after, to share with the ANC staff during the next ANC consultation (usually some two months after the first visit). In case of abnormal results, the woman needs to be referred to ANC immediately. Results of rapid tests executed at the ANC are immediately available.
Within the Senegalese health system, the Directorate of Reproductive Health and Infant Survival (DSRSE) is responsible for the clinical component of ANC, while the Directorate of Laboratory services (DL) is responsible for laboratory testing. Screening for HIV infection is part of the Prevention of Mother-to-child Transmission (PMTCT) component of the national HIV program. Laboratory capacity for the seven recommended ANC screening test is in place as part of the tier-specific testing package at all health facilities of levels 2 (Etablissement Public de Santé: EPS) and 3 (Centres de Santé: CS) of the health care pyramid. Rapid testing for HIV and albuminuria are also supported at the level of health posts (Postes de Santé: PS), which do not host a laboratory. Laboratories at health facilities serve ANC women (and other persons) seeking clinical services at the same facility and as well as those referred directly to the laboratory by smaller public or private facilities with no access to laboratory testing.
For the SociaLab study we targeted health facilities at the intermediate level of the Senegalese public health care system (level 2 and 3), providing services to more than 500 ANC clients per year. Of 93 level 2 and 3 facilities, we purposively selected 16, balancing geographic representation of Senegal’s 14 administrative regions and accessibility.
Pregnant women attending for their first ANC visit (ANC-W) were eligible to participate. From February 2014, at each study facility one or two midwives (depending on staff availability) were trained on the study procedures. The study-midwifes recruited consecutively among women assigned to their clinic hours, until the facility close-out visit in July/August 2014 or a maximum of 120 was reached, whatever came first. The sample size assumed on average 500 clients per year per facility (based on records from previous year) and the ability to determine a 50% test uptake (of 7 tests) with a 95% confidence interval of ± 3%, requiring a minimum sample size of 947 women. We allowed the 16 facilities to enroll 50% more women in anticipation of potential loss to follow-up. In addition to ANC-W, we included pregnant women who were referred to the laboratory from health posts and private practitioners for ANC laboratory tests (LAB-W) during the same period, up to a maximum of 50 per facility.
Data handling and statistical analysis
Of the participating ANC-W, we prospectively transferred routinely collected information on age, pregnancy status, test requests, execution, results and clinical management (if applicable) to individual paper study cards (S1 Supporting File). For the study participant that were referred to the laboratory from smaller clinical facilities (for LAB-W), the same study cards were used, but only information on test request and execution were collected.
We obtained information about the selected health facilities during one-week visits to each health facility, between July 2013 and August 2014. One researcher (AS) conducted surveys based on 3 structured questionnaires addressing general characteristics, staffing/human resource, availability and use of guidelines, management and organization of the laboratory and the ANC services (S2 Supporting File). The three questionnaires were designed in such a way that they targeted 1) the ANC in charge; 2) the laboratory in-charge and 3) the finance administrator of the health facility.
Two additional research team members observed the ANC and laboratory premises to collect data about infrastructure and cleanliness. Information of the average waiting time of ANC visits was based on observing 10 random women attending the ANC clinic on the first day of the facility visit by the study team. In total each facility was described by one set of responses to the three different questionnaires. Two trained research assistants entered and stored both women and facility data in electronic databases (Open Clinica version 3.1. Copyright © OpenClinica LLC and collaborators, Waltham, MA, USA, www.OpenClinica.com). All entries were checked for completeness and quality against the source documents.
Test uptake: the percentage of women attending for their first ANC visit who obtained complete results for all seven recommended ANC laboratory tests.
Test utilization: the percentage of women attending for their first ANC visit who, from the recordings on the cards, received appropriate clinical care in case of critical (out of range or abnormal) laboratory test results. For this outcome two investigators (PO, AS) reviewed the recorded information on clinical management of all participants possessing test results that were not negative or that fell outside the normal range. They categorized clinical management recordings as ‘appropriate’ when fully in accordance with the guidelines , ‘partly appropriate’ when necessary preventive or complementary measures to the treatment were missing or as ‘not appropriate’ in case no treatment or inadequate treatment was prescribed.
We recoded parity as ‘ever delivered’, implying experience with pregnancy and childbirth, versus ‘never delivered (nulliparous)’, assuming inexperience. We calculated turn-around times of test request, execution, recording of results, and return with results at the ANC for all tests combined, using the dates that were recorded on the cards.
We constructed cascades of care for ANC-W, per test and per facility, reflecting the total numbers of women attending for first ANC visit, with test request, test execution, available result, result abnormal or otherwise indicative of the need for further action, and adequate clinical management.
We summarized the data on health facilities by question and used the WHO health systems building blocks framework  to categorize questions as: Service delivery and quality; Governance and leadership; Information and research; Infrastructure, equipment and supplies; Human resources; Health care financing (Table 2). Two investigators (AH, PO) then, for each building block, assessed the suitability of all questions for inclusion as facility level co-variates in the analysis of factors associated with test uptake. We discarded questions that we considered too remote or otherwise irrelevant, had incomplete responses, lack of any variability in responses based on eyeballing the data. Table 2 shows the resulting selection of variables, their categories and distributions.
As described in the methods, parameters from the health facility assessment questionnaires were assessed for completeness and relevance for inclusion in further analysis. The building blocks were slightly adapted to the ANC-laboratory setting. The columns at the left show the distribution of the characteristics among the 16 facilities.
We examined associations between test uptake (i.e. attainment of the complete set of seven recommended ANC test results among ANC-W) as an outcome and all available women and facility-related co-variates in univariate analysis and in a multilevel mixed model using the STATA XTMELOGIT command, to accommodate the hierarchical nature of the data . We first modeled women nested in facilities without co-variates and added women level co-variates as fixed effects (parity trimester of first visit, women’s age) based on the magnitude of the association in crude analysis. We then added urban vs. rural (as our primary level 2 variable of primary interest) as a fixed effect and examined interactions between this variable and level 1 co-variates. Lastly, we added type of facility (hospital vs. health centre) and the facility level co-variates as listed in Table 2 and representing the WHO health system building blocks, one-by-one as fixed effects. Likelihood ratio tests (LRT) were used to compare models, and to explore if any of the fixed effects improved model fit. We obtained intra-class correlation coefficients (ICCs) with the Stata ESTAT command. Records with missing values for age and/or trimester of the pregnancy (n = 63) were excluded from the regression analysis. STATA version 12.1 (StataCorp LP, College Station, Texas, USA) was used for all analyses.
Ethical approval for the SOCIALAB study was granted by the Senegalese Comité National d’Ethique pour la Recherche en Santé (CNERS, protocol number SEN13/09). All study participants provided oral informed consent to preserve the anonymity of data collection and as per approval of the ethical review board. Study participants were provided a study number and all data were anonymized. Test request and results of HIV and syphilis testing were coded in such a way that the disease screened could not be identified.
Facilities and pregnant women
Of the 16 selected facilities three were hospitals and 13 health centers, of which four were in rural areas, as defined by the Senegalese administrative criteria (Table 3). The number of women attending for ANC visit in 2012 ranged between 517 and 2917. The price of the full ANC package ranged between XAF 7,150 and 13,000 (Euro 10.9–19.8). At two facilities all seven ANC tests were conducted in the laboratory. At 14 facilities HIV serology rapid tests were conducted in the ANC and at nine of those proteinuria rapid tests as well. In addition to nurses, midwives and laboratory technicians, one of the four rural facilities had a gynaecologist, and five of the 12 urban facilities had both gyneacologist(s) and specialized laboratory staff. Additional health facility characteristics categorized by health system building block are summarized in Table 2.
In total 1692 women participated in the study, per facility a median of 76.5 ANC-W (range 64–97) and 26 LAB-W (range 24–48) (Table 3). Of the participants, 1246 (74%) were ANC-W of whom 438 (35%) were nulliparous and 754 (61%) were in the first trimester of their pregnancy (Table 4). LAB-W (n = 446) had a similar age distribution, and 191 (42.8%) were nulliparous. Compared to ANC-W they presented at the lab at a more advanced stage of the pregnancy (34% in the first trimester) and were less often attended by specialized ANC staff (i.e mid-wife or gynecologist, 75%) compared to ANC-W (99%).
ANC test uptake and determinants
Of the seven recommended tests, ANC-W on average had 4.2 results available (sd 2.7), and 400 (32%) had complete results for the set, implying optimal test uptake. Completeness varied widely between facilities (Table 3), from 94% and 99% at the highest end, to 0% at two facilities, where women on average had 1.7 and 3.3 test results.
In multilevel logistic regression analysis, ANC-W had half the odds of a complete set of 7 results if they made their first ANC visit in the third trimester (compared to first) or had ever delivered. Women’s age was not associated with availability of complete results. Facility-level variation (as a random effect) explained most of the variation in the data (ICC 72%). The addition of facility-level characteristics to the model as fixed effects explained some of this variation (ICC 58%). The final model (Table 5) suggests that women attending rural facilities were at lower odds of complete results, but this interacted with parity, such that for women who ever delivered and attended ANC at rural health facilities, the odds of complete results were very low compared to urban primigravidae (OR 0.064; 95%CI 0.00–0.52). At urban facilities women who ever delivered had slightly reduced odds of complete results compared to urban primigravidae (OR 0.72; 95%CI 0.49–1.06). Whether a facility was a hospital or health centre had no significant effect. Of the facility-level building block questions listed in Table 2, two improved the model. Routine recommendation of all 7 tests increased the odds, and a larger number of daily opening hours of the lab for sample collection decreased the odds of complete results, however, the 95%CIs of the effect sizes were wide.
Among ANC-W, HIV serology results were most frequently available (82%), followed by albuminuria (63%), while Hemoglobin results were least available (48%), as shown in Fig 1. Considering the cascade from test request to availability of result, the most frequent bottleneck for ANC-W was non-execution of requested tests, which was observed in 13% of women for HIV serology, in 23% for albuminuria and between 36–41% for the other five tests. The lack of a test request by ANC staff was second, but only for Hemoglobin and albuminuria (13% of women) and was <5% for the other 5 tests. Unavailability of results for executed test occurred for only 1–2% of the ANC-W. LAB-W most commonly had 5 (30%) or 4 (26%) test results (Table 6), in line with the expectation that LAB-W had already received HIV, and/or PU rapid testing at the level of the health posts. The lab requests of LAB-W most commonly included Syphilis (93%), followed by Blood group/Rhesus and Emmel test (83%), and Hemoglobin (65%). Of Hemoglobin requests 11% were not executed, but otherwise non-execution or non-availability of results was rare among LAB-W with a test request. The prior status of test results with lifelong validity (blood group/Rhesus factor and Emmel test) was documented for 7% of ANC-W and LAB-W, almost exclusively women with previous pregnancies.
Adequate clinical management was recorded for 20/37 ANC-W with Rhesus-negative status (54%), and 5/10 for a HIV-positive status (Table 6). For other tests the appropriateness of recorded clinical management was lower. Overall, 97(18%) of 525 abnormal test results had a record of adequate management. The proportion of LAB-W with abnormal results was similar for each test as observed for ANC-W.
Uptake patterns and turn-around times
Complete test uptake among ANC-W was lowest at the rural health facilities, 0–1% in three, and 24% in the fourth facility. When further stratified by test, parity, and routine recommendation all 7 tests, Hemoglobin testing was frequently left out at the rural facilities (Fig 2). Among women with prior deliveries, uptake of most tests was ±20–30%, except for HIV serology–and to some extent syphilis and PU—which had higher uptake. In urban health facilities that reported that all seven tests were routinely recommended, the uptake was highest and quite similar among women who ever vs. never delivered. Uptake was more equal for all tests, although HIV and PU still stood out. In the urban facilities without routine recommendation, uptake across tests and women with and without prior delivery experience was also quite similar, except for PU which was available for a quarter only.
Percentage ANC women with results for each recommended test, separate for those who never (circle) vs. ever delivered (triangle), stratified by facilities in rural vs. urban areas and whether the facility reported to routinely recommended all seven tests or not. Number of rural facilities without routine recommendation of the 7 tests n = 2. Number of rural facilities with routine recommendation of the 7 tests n = 2. Number of urban facilities without routine recommendation of the 7 tests n = 3. Number of urban facilities with routine recommendation of the 7 tests n = 9.
The complexity a woman encounters in getting her seven tests executed and obtain results depends on the number of rapid tests that were conducted in ANC vs. at the laboratory. As explained in Table 3, at facilities where HIV rapid testing is conducted at ANC, PU rapid testing at the laboratory on the same day, and other tests at the laboratory but not same-day, women require three moments at two places, while conducting all tests at the lab requires two moments at one place. For women with incomplete uptake, the number of available results peaked at the number of rapid tests that were available at ANC (Fig 3), but this was not associated with uptake in regression analysis.
For executed tests, test result availability for ANC-W was same-day for 83% and next day for 10%. For LAB-W this was 73% and 15% respectively. Test request, execution and clinical management for ANC-W with incomplete uptake was same day for 49%, and within one week for 75%. This was 21% same-day and 61% within one week for ANC-W with complete uptake of 7 results.
This analysis showed that among women attending for their first ANC visit at health facilities with laboratory capacity, uptake of the seven recommended laboratory tests  was low: only one in three obtained the complete set of test results. However, uptake varied widely between facilities, was lower in rural facilities, and more so among women with earlier deliveries. In urban facilities uptake was higher and depended less on women’s prior delivery status, especially in facilities that routinely recommended all seven tests. Failure to execute a requested test was the most common bottleneck, while, once executed, missing tests results were uncommon (<2%).
Low uptake of ANC tests despite available laboratory infrastructures and commodities, and high (first) ANC attendance indicate missed opportunities for quality ANC in Senegal. Even more so for rural women who, in west-Africa, are already far less likely to receive ANC than their urban counterparts .
The results of the anthropological components of the mixed-methods SociaLab project (published elsewhere ) give us insights that helped us understand the results of the quantitative analyses presented in the current manuscript. For pregnant women, reasons for non-execution of requested ANC laboratory tests mostly related to difficulties in obtaining money to pay for the tests and for the multiple travels for clinic and laboratory visits. In Senegal women do not have to pay for delivery care in public facilities , but payment is required for ANC consultation, registration, medicines, laboratory diagnostic tests -which are expensive when considering poverty-, and echography . Illiteracy, which is common among women and men , aggravated the financial barrier as not being able to read the test request and thus not knowing how much to request or precisely what to request money for made it hard for women to raise sufficient money, usually from their husbands, for tests and for transportation to collect results.
The anthropological study also revealed the role of midwives as gatekeepers. At the health facilities midwives reported anticipating the financial hardship of women. They tried to limit the prescription of medicines and requests for test that women could not afford, by prioritizing tests that they judged most important. Or they skipped the most expensive ones (e.g. Hb) and instead relied on clinical symptoms, a common aspect of clinical medicine in SSA [1, 4, 20]. In addition, midwives faced heavy workload, shortage of furniture, equipment and consumables in most facilities, and poor pay. A variety of reasons given by midwives for not explaining to women about the tests were: lack of time, forgetting (due to high workload), illiterate women not being able to understand, anticipation that women might refuse the test (for HIV), a language barrier, the lack of clear national guidelines for ANC screening test creating inconsistency in which tests were recommended, not filling a request form for tests that were faulty at the facility’s own lab, or in anticipation of stock-outs. At the laboratory level reported barriers for test execution were the cost of the test, stock-outs of reagents, and broken equipment.
Lack of confidence in laboratory results for clinical management due to lack of equipment, inadequate supply-chain management for consumables and reagents, poor equipment maintenance, lack of clear policies, and insufficient leadership, points out the importance of laboratory capacity strengthening efforts . And the need for human resource management interventions, which can improve health workers' performance, taking into account that different contexts produce different outcomes . During a result dissemination workshop the study findings were shared, barriers were discussed among different stakeholders , and an implementation plan was developed.
Whereas a more in-depth understanding through triangulation of data from different research methods was a strength of our study, it also pointed to limitations. Among pregnant women approached for anthropological interviews in the community, verification of ANC cards suggested that ANC test utilization was even lower than among the women we report on in this manuscript, especially among women attending ANC at health posts, and tests were more often not requested . The differences may be due to sampling or incompleteness of ANC cards, but study procedures likely contributed: in each facility a midwife was trained to recruit and attend to study participants and record clinical data on specially designed study ANC cards to facilitate data entry. This because in routine care standard antenatal cards were often not completed fully, a problem that is more common in this type of studies . The study midwives probably became more mindful of requesting the test, which were then not followed on by the women because of financial constraint, thus suggesting a different bottleneck in the cascade, and possibly a ‘study effect’ towards more complete test request. Equally, higher availability of test results may have occurred from laboratory staff more mindful of delivering better quality of test is results in the context of this study. Collecting data in 16 different facilities was intended to capture variation. Even though the results reported may represent a more optimistic overview of the ANC cascade compared to facilities not participating to the study, the gaps we described remain substantial.
Other limitations apply to the multilevel regression analysis, which was explorative. The available women characteristics were limited to information routinely collected during ANC, and did not include additional assessment of women’s individual socioeconomic status, literacy level or distance to the facility. With 16 facilities power was low and failure to demonstrate an effect of facility level characteristics should not be interpreted as if these aspects are irrelevant. Also, facility assessments were cross-sectionally, so e.g. stock-outs or breakdowns may not have coincided with the full enrolment period. The analysis however suggests that most variation in uptake was explained at the facility level, which invites for interventions targeting the health facility and management level. Moreover, it tones down the blame that health workers tend to put on the women for low uptake of tests. Facilities that routinely recommended all tests had higher uptake of the complete set underscoring the importance of guideline implementation. Longer opening hours of the lab was inversely associated with uptake, which is counterintuitive, as short opening hours of the lab had been suggested as a reason for low uptake by ANC staff. It may be an indication of another unmeasured facility characteristic, but may also indicate that intuitive solutions (i.e. increasing opening hours) may improve client-friendliness, but on its own may not necessarily solve the uptake problem. The national recommendation changed from 6 to 7 tests (glycaemia was added) during the study period , but this had little impact since all facilities already considered glycaemia testing as part of the package at study onset and glycaemia uptake was quite similar to e.g. syphilis testing. A strength of our study is the detail in uptake of different tests, compared to the Senegalese Demographic and Household Surveys (DHS) that measured uptake of urine and blood sampling (90% for urine- and 84% for blood sampling among women attending ANC in 2016) without specifying whether and which tests were carried out on the specimens .
Low uptake of ANC tests is not unique to the setting of our study, and was even lower in a study in Mozambique  that aimed at improving uptake of antenatal syphilis screening through integration of HIV and syphilis rapid tests in ANC. In our study HIV rapid testing, which is managed vertically by the Senegalese HIV program, had the highest uptake, and confirms the perception that point-of-care (POCT) technologies increase access to health services. However, when rapid tests for HIV (free) and/or proteinuria (not free) were conducted at ANC, a subgroup of women had results for these tests only. Similarly, experiences with POCT for syphilis  and CD4+ T-cell enumeration  showed wide variability in uptake, indicating the potential for detrimental effects in some settings or sub-populations. If importance of a full ANC laboratory tests packages could be promoted through a health systems approach, then goals and funding of vertical programs could be utilized to improve lab test uptake and quality and levels of service delivery. A similar approach is used in other laboratory capacity building initiatives, were support for vertical disease programs is combined with horizontal health system approaches [2, 4], and vertical program monitoring could be a monitoring strategy .
In conclusion, our study showed that uptake of recommended laboratory tests among first ANC attendants was low, and lowest in rural facilities and among women with earlier deliveries, illustrating challenges to test uptake even when laboratory testing capacity is in place. Large differences between facilities underscore the importance of management, policy, and the importance of considering local context in order to improve service delivery to expectant mothers by identification of specific issues at each facility that require intervention.
S1 Supporting File. Fiche de suivi / ANC card.
We are grateful to the healthcare providers and the clients attending the 16 health facilities that participated in the study. We are grateful to Dr Frank van Leth, for his input to the statistical analysis plan. We thank the Knowledge Translation Network (KTNET) project from the Makerere University, School of Public Health for assisting in the interpretation of the research results.
- 1. Petti CA, Polage CR, Quinn TC, Ronald AR, Sande MA. Laboratory medicine in Africa: a barrier to effective health care. Clin Infect Dis. 2006 Feb 1;42(3):377–82. pmid:16392084
- 2. Nkengasong JN, Nsubuga P, Nwanyanwu O, Gershy-Damet GM, Roscigno G, Bulterys M, et al. Laboratory systems and services are critical in global health: time to endci the neglect? Am J Clin Pathol. 2010 Sep;134(3):368–73. 10.1309 pmid:20716791
- 3. Gray I.P. and Carter J.Y., An evaluation of clinical laboratory services in sub-Saharan Africa. Ex africa semper aliquid novi? Clin Chim Acta, 1997. 267(1): p. 103–28. pmid:9469247
- 4. Ndihokubwayo J.B.K., F.; Yahaya A.A.; Mwenda J. Strengthening public health laboratories in the WHO African Region: A critical need for disease control. The African Health Monitor, 2010. 47–52.
- 5. Peeling R.W. and Mabey D., Point-of-care tests for diagnosing infections in the developing world. Clin Microbiol Infect, 2010. 16(8): p. 1062–9. pmid:20670288
- 6. World Health Organization, Standards for Maternal and Neonatal Care, in 1.6 Provision of effective antenatal care. 2007, World Health Organization: Geneva.
- 7. Partnership for Maternal Newborn and Health Child, Opportunities for Africa’s Newborns. Practical data, policy and programmatic support for newborn care in Africa. 2006: Geneva.
- 8. UNICEF, UNICEF Data: Monitoring the Situation of Children and Women. https://www.unicef.org/.
- 9. World Health Organization, Joint United Nations Programme on HIV/AIDS (UNAIDS)., United Nations Children's Fund., Global HIV/AIDS response: epidemic update and health sector progress towards universal access. 2011, World Health Organization: Geneva.
- 10. Koster W, Ondoa P, Sarr AM, Sow AI, Schultsz C, et al. Barriers to uptake of antenatal maternal screening tests in Senegal. SSM Popul Health. 2016 Oct 13;2:784–792. eCollection 2016 Dec. pmid:29349190
- 11. World Health Organization, World Health Statistics data visualizations dashboard, Monitoring health for the SDGs. 2016, World Health Organization. p. SDG health and health related targets, 3.1 Maternal mortality; 3.2 Newborn and child mortality.
- 12. WHO, UNICEF, UNFPA, World Bank Group, and United Nations Population Division Maternal Mortality Estimation Inter-Agency Group. Maternal mortality in 1990–2015. Senegal. https://www.who.int/gho/maternal_health/countries/sen.pdf
- 13. World Health Organization, Global Health Observatory indicator views. Global Health Observatory indicator views. 2016, World Health Organization. p. By indicator: Antenatal care (ANC) attendees tested for syphilis at first ANC visit (%) (Global strategy for women's, children's and adolescents' health).
- 14. The World Bank, Data by country. 2018, The World Bank Group.
- 15. Ministère de la Santé de l’Hygiène Publique et de la Prévention, Protocoles de services de SR au Sénégal. 2014, Ministère de la Santé de l’Hygiène Publique et de la Prévention.: Dakar. https://advancefamilyplanning.org/sites/default/files/resources/PMP%202nd%20Version-1_Part1_pp1-95.pdf
- 16. World Health Organization, Everybody business: strengthening health systems to improve health outcomes: WHO’s framework for action., in Framework for Action. 2007, World Health Organization: Geneva.
- 17. Hox JJ.Multilevel analysis: techniques and applications. Quantitative Methodology Series. 2nd ed. 2010. Routledge, New York.
- 18. Stata Corporation, STATAtm version 12.0. Texas, USA.
- 19. Witter S, Armar-Klemesu M, Dieng T, National fee exemption schemes for deliveries: Comparing the recent experiences of Ghana and Senegal, in Financing obstetric care, Richard F., Witter S., De Brouwere V., Editor. 2008, ITM: Antwerp. p. 167–198.
- 20. Elbireer AM, Opio AA, Brough RL, Brooks Jackson J, Manabe YC, and al. Strengthening Public Laboratory Service in Sub-Saharan Africa: Uganda Case Study. Laboratory Medicine, 2011. 42(12): p. 719–725.
- 21. Dieleman M, Gerretsen B, van der Wilt GJ. Human resource management interventions to improve health workers' performance in low and middle income countries: a realist review. Health Research Policy and Systems 2009 7:7. pmid:19374734
- 22. Barriers to the use of prenatal screening test for maternal health in Senegal. Video. https://www.youtube.com/watch?v=JPX2qTSmxOI. Publication date: 28 April 2016.
- 23. Gloyd S, Montoya P, Floriano F, Chadreque MC, Pfeiffer J, Gimbel-Sherr K. Scaling up antenatal syphilis screening in Mozambique: transforming policy to action. Sex Transm Dis, 2007. 34(7 Suppl): p. S31–6.
- 24. Agence Nationale de la Statistique et de la Démographie (ANSD) [Sénégal], et ICF. 2017. Sénégal: Enquête Démographique et de Santé Continue (EDS-Continue 2016). Rockville, Maryland, USA: ANSD et ICF. Available at: https://dhsprogram.com/pubs/pdf/FR331/FR331.pdf
- 25. Taylor M., Newman L., Ishikawa N., Laverty M., Hayashi C., Ghidinelli, et al. Elimination of mother-to-child transmission of HIV and Syphilis (EMTCT): Process, progress, and program integration. PLoS medicine, 14(6), e1002329. pmid:28654643
- 26. De Schacht C, Lucas C, Sitoe N, Machekano R, Chongo P, Temmerman M, et al., Implementation of Point-of-Care Diagnostics Leads to Variable Uptake of Syphilis, Anemia and CD4+ T-Cell Count Testing in Rural Maternal and Child Health Clinics. PLoS One, 2015. 10(8): p. e0135744. pmid:26308345
- 27. World Health Organization, Global Guidance on criteria and Processes for validation: elimination of mother-to-child transmission of HIV and Syphilis,. 2014, World Health Organization: Geneva.