Availability of priority maternal and newborn health indicators: Cross-sectional analysis of pregnancy, childbirth and postnatal care registers from 21 countries

Data from national health information systems are essential for routinely tracking progress, programmatic decision-making and to improve quality of services. Understanding the data elements captured in patient registers which are building blocks of national HMIS indicators, enables us to standardize data collection and measurement of key indicators for tracking progress towards achieving maternal and newborn health goals. This analysis was done through a review of antenatal care (ANC), childbirth and postnatal care (PNC) registers from 21 countries across five geographic regions. Between July and October 2020, country-based maternal and newborn experts, implementing agencies, program managers, and ministry of health personnel were asked to share the registers in use. Both paper-based and electronic registers were obtained. Twenty ANC registers, eighteen childbirth and thirteen PNC were available and analyzed. Both longitudinal and cross-sectional ANC and PNC registers were obtained, while the childbirth registers included in the analysis were all cross-sectional. Fifty-five percent (11/20) ANC registers and 54% (7/13) PNC registers were longitudinal. In four countries, the registers were electronic, while the rest were paper-based (17 countries). Sub-analysis of registers from four countries (Ghana, Kenya, Nigeria, and Zambia) where the 2017/2018 and 2019/2020 registers were available showed that the latest versions included 21/27 (78%) of data elements that are critical in the computation of key maternal and newborn care indicators. This analysis highlights some areas in where there are data gaps in data on pregnancy and childbirth. Program managers and health workers should use data gathered routinely to monitor the performance of their national health system and to guide the continuous improvement of health care services for women and newborns. The findings can help to inform the standardization of pregnancy and childbirth registers, and provide information for other countries seeking to introduce indicators in their health systems.


Introduction
The global burden of maternal deaths is estimated at 810 women dying each day from preventable conditions related to pregnancy and childbirth [1,2]. Although there was 38% reduction in the maternal mortality ratio (MMR) between 2000 and 2017, the current MMR of 211 deaths per 100,000 live births remains unacceptably high especially, with two-thirds of the deaths occurring in low-to middle-income countries (LMICs) [1]. At the same time, the number of stillbirths is rising especially in Sub-Saharan Africa, increasing from 0.77 million in 2000 to 0.82 million in 2019. In some high-income countries more stillbirths than neonatal deaths occur, and in some cases, even surpass the number of infant deaths [2,3]. Most maternal and newborn deaths are preventable if proper preventive measures are in place. In many settings poor quality of care is a greater contributor to poor health outcomes than care coverage and it is estimated that about half of the maternal deaths and 58% of the newborn deaths could be averted with quality health care [4]. Maternal and newborn mortality is considered to be proxy indicator of limited access to quality healthcare service delivery in a population. Avoidable maternal and newborn health inequalities arise because of the socio-cultural and socioeco-political circumstances in which a woman and her newborn grow, live, work, and age, and the systems put in place to deal with illness [5].
The data needed for routine program implementation monitoring and evaluation are gathered through a variety of modalities, with health facility registers being an important source of information to document events around antenatal care (ANC), childbirth, and postnatal care (PNC), hence the need to periodically assess their content and utility [6]. However, there are challenges related to using current disjointed data collection and reporting systems which in part include focus on meeting the needs of specific funders, poorly developed health information systems (HIS), and inadequate human resources capacity [7]. Furthermore, health data are often recorderd and/or collated by clinicians who have to deal with high patient caseloads, especially those working in the public sector, which might lead to staff burnout, compromised quality of care and poor data quality [8]. Health facility data have been reported to have many shortcomings which include indicators of poor quality, such as inaccuracies, incompleteness and not reporting on time through the health management information system (HMIS) [9][10][11][12][13][14][15][16].
The national HMIS is essential for decision-making, providing regular data on delivery and utilization of services which can be used for program planning and track progress towards national and subnational objectives [17]. Many maternal and newborn health indicators can be reported through the HIS (and are used by countries to track their progress toward local and national goals). One such commonly used indicator which serves as a marker of service utilization is proportion of women attending at least four ANC visits during pregnancy. Data elements used to compute this and other indicators are captured at the facility level and reported routinely, often flowing from facility to subnational and then to national level, with consolidation of data along the way per the existing processes and systems in the country [18][19][20][21].
Routine health information systems (RHIS), which comprise the regular collection, reporting and analysis of health facility data, are often weak in LMICs, with limitations in the representativeness and quality of reported data [17]. In many LMICs countries, routine data are captured through use of printed paper registers, and only available through electronic platforms in summary form, using software such as the district health information system (DHIS2) [22]. On the other hand, high-income countries such as Denmark, Norway, Sweden, and United States of America have robust electronic data capture systems [23][24][25][26][27].
Program and facility managers can use data reported through the RHIS dfor health service delivery planning, implementation and monitoring purposes [28]. Improving the quality and consistency of data collected and reported, requires a thorough understanding of the capacities of the national HIS and how and what information is recorded in facility registers. Analysis of ANC, childbirth and PNC registers is an important step towards improving our understanding the availability of key data elements related to maternal and newborn health that can reported through the national HMIS, with a view towards promoting standardization in data collection, analysis, and use [13,29]. This standardization could help countries better use RHIS data for decision-making by improving the accuracy of what is reported and the ability to analyze trends over time. Additionally, such standardization could help to promote the use of routine data for reporting progress towards global targets, rather than relying on periodic surveys and estimates. This paper reports on an analysis of key data elements in ANC, childbirth and PNC registers from 21 countries describes differences in the way data elements are captured, and highlights opportunities for inclusion of some critical data elements that are not currently collected in registers as well as possibly removing some data elements to reduce redudancy across the different registers and other data collection documents.

Methodology
The approach used in this review was informed by previous similar work that focused on maternal, newborn and child survival [13,29]. Selection of the countries included in the review was guided by the findings of a literature review conducted prior to embarking on the analysis of the registers as well as the availability of the records. Countries included in this review had some literature published related to ANC, childbirth, or PNC registers. A few countries, especially those from regions where published literature was not available, were added to the list to present a more global perspective, taking into account the geographical distribution as well as the access to the registers at the time of the review. In the end, twenty-one countries drawn from five geographical regions (Africa, Europe, South-East Asia, Eastern Mediterranean, and the Americas), were included in the review (Table 1).

Gathering the registers
Between July and October 2020, maternal and newborn experts, implementing agencies, program managers, and ministry of health personnel were contacted and requested to share the current versions of the ANC, childbirth, and PNC registers in use in the selected countries. We received 20 ANC, 18 childbirth, and 13 PNC registers from 21 countries in various formats such Microsoft Excel and Word, portable document format, images, and links to websites. Both paper-based and electronic registers were obtained. Unavailability of some registers for analysis from any of the 21 countries did not mean that the registers do not exist in the country, but that they were not able to be easily shared, especially in countries using electronic registers. For the paper-based registers, the years of their production ranged from between 2015 to 2020. For some of the electronic registries, no specific dates could be assigned to when they were produced. However, majority had some time-periods documented in various ways; for example, the medical birth register of Sweden reviewed spans the period 1967 to 2018 [24,25]. This review was limited to the three types of registers and did not include other records such as the partograph and data summary forms.
Registers in languages other than English were translated prior to analysis. Since registers are revised periodically, we analyzed the most recent versions of the documents available for each country. Eleven of the twenty (55%) of ANC registers, and 7/13 (54%) of PNC registers analyzed were longitudinal ( Table 1). All 18 childbirth registers analyzed were cross-sectional. Registers even from the same country were not linked to each other; they were siloed by the type of care.

Capturing the data elements
A Microsoft Excel worksheet was used to record key data elements manually extracted from each register for each country. These data included: name of the register, year of production, format (cross-sectional or longitudinal), and the data elements contained. For each country file, separate worksheets were population with ANC, childbirth (or labour and delivery), and PNC data elements. Thereafter, all data elements contained in the register were marked under appropriate worksheets, organized in logical sections such as demographics, medical history, laboratory investigations, client outcomes and/or services offered. Abstraction of the information was done in a standardized manner by one person (MK) over a period of three months. Data quality checks were conducted through a second review of select registers from the 21 countries by a different person (FP).

Analysis and synthesis of content
Following completion of the data abstraction into the individual country worksheet, a summary file with three worksheets (ANC, childbirth, PNC) was generated. This process was done in a stepwise manner as follows: Step 1 -Five categories were generated for summarizing the data, and corresponding worksheets in the Microsoft Excel file generated: pregnancy, laboratory tests and investigations, birth outcomes and treatment, and PNC services provided.
Step 2-Country names were listed alphabetically, presented as rows in tables.

PLOS GLOBAL PUBLIC HEALTH
Pregnancy, childbirth and post-partum registers review from 21 countries Step 3-The column headings were generated by identifying the common data elements present in most of the registers. Some key data elements that are either denominators or numerators of key maternal and newborn health indicators were also included, whether they were present in many registers or not.
Step 4 -Summary data were captured from each country file by marking a "check" in the appropriate cell if the data element or term listed on the column heading was present in the country file.
Step 5 -The content of the data elements captured was further categorized into logical categories, guided by standard definitions of key maternal and newborn indicators as per WHO's guidance on data recommended for monitoring maternal, newborn and child health services in health facilities [30][31][32]. (Refer to S1 File-List of selected priority indicators in RMNCAH).
Step 6 -As the process of capturing the summary data progressed, some data elements were added and steps 3-5 were reviewed, with corresponding adjustments being made to the data summary categories identified.
Step 7 -Sub-analysis of registers from four countries (Ghana, Kenya, Nigeria, and Zambia) across two time periods (2017/18 and 2019/20) was done to identify changes in format and content or data elements included. These countries were selected because data on their registers were available from a similar review conducted in 2018 13 . This allowed for comparison of data between two time points. We looked for the presence or absence of twentyseven data elements that are critical to monitoring maternal and newborn services across the continuum of care, from pregnancy to the postnatal period.

Results
The findings on the similarities and differences pertaining to how key data elements are captured in different registers from 21 countries are presented. There was great variation in the number of data elements included from each document reviewed ranging from 24 in Nepal's delivery register to 188 in Eswatini's ANC register. These findings are presented in three broad categories (Pregnancy, Childbirth, Postnatal care, and Mortality outcomes), with subheadings in each category, highlighting the similarities and differences in data elements across the countries.

Pregnancy
Of the 20 countries whose ANC registers were analyzed, age, name of the woman, ANC number assigned at the facility or a personal identifier, and residential address were the most common recorded data elements in 15 countries. In two countries, (Afghanistan and Sweden), the husband or name of the baby's father's citizenship was documented while citizenship of the mother is recorded in Guyana and Sweden. Some other vital information recorded, usually during the first ANC visit, include woman's relevant medical history such as hypertension and diabetes in 13/20 (65%) of the countries, and complications from past pregnancies in 7/20 (35%) of the countries. Recording of behavioural or practices such as smoking, use of alcohol or drugs is documented in five countries, all located either in Europe or the Americas. Utilization of antenatal care services. Countries that used longitudinal ANC registers routinely document between five and eight or more contacts, whereas those using electronic formats (Norway, Sweden, USA, and West Bank-Palestine), have capacity to record many more visits, one visit per row. For the countries using cross-sectional registers, they record the first visit as "New visit" and subsequent visits in a separate column as "Revisit". In some cases, the actual visit number was entered, for example 3 rd visit, while in some cases, there was a column to capture 4+ visits in 8/20 countries.
Routine examination during ANC visits. During each ANC visit, a variety of measurements are taken and recorded; longitudinal ANC registers allow for comparison of a current visit to previous visits in order to identify or monitor possible early danger signs. In over half of the countries examined, routine data elements recorded during a given visit included: gestational age, blood pressure, weight, height, and haemoglobin level or pallor/ anaemia. Less commonly recorded parameters included: fetal heart tones/sounds, presentation of the baby, and the mother's nutrition status ( Table 2).
Malaria in pregnancy. In the countries shown in Table 3 where malaria is endemic, documentation of the doses of intermittent preventive treatment in pregnancy (IPTP) for management of malaria in pregnancy was included in 7/20 (35%) of the registers reviewed. Four of the seven countries document at least four doses while the other three record three doses. Provision of bed nets was included in 9/20 (45%) of the ANC registers reviewed (Table 3). Two countries had columns to document malaria treatment provided.

Prevention of mother to child transmission (PMTCT) of HIV, sexually transmitted diseases (STIs) and routine laboratory tests and investigations.
Pregnancy is a period when specific measures can be taken to prevent a baby born to a HIV-infected woman from being infected. In addition, screening and treating mothers with STIs like syphilis can protect the baby from such infections both in-utero or after birth. The HIV status of the mother was documented in 10/20 (50%) of countries where ANC registers were reviewed-primarily in countries where HIV is endemic. ANC registers from nine countries (45%) document whether the mother is on antiretroviral therapy (ART). Three countries (Eswatini, Kenya, and Uganda) document the WHO stage of disease. The same three countries also have space to document whether the pregnant woman is on cotrimoxazole (CTX) to prevent opportunistic infections.
Regarding syphilis, 17/20 (85%) of the countries record testing for syphilis (Table 4). Most countries capture data on laboratory tests and investigations conducted routinely ANC, with the most commonly recorded tests in the registers being hemoglobin level; Syphilis testing; urinalysis for sugar or proteins; HIV testing; and blood grouping (Table 4). Nutrition and micronutrients supplementation. Nutrition interventions are key to ensuring the good health of the mother and normal growth of the fetus. Provision of health promotion and prevention messages to the mother through nutrition counseling is documented in 9/20 (45%) of the countries. Regarding micronutrients, six countries document provision of iron and folate supplementation separately (Guyana, Indonesia, Kenya, Norway, Tanzania, Uganda, and Zambia), while eight countries record them combined in one column. Provision of Vitamin A and calcium are recorded in one country each, Norway and Zimbabwe, respectively. Related to nutrition, deworming is recorded in only six countries (Fig 1).
Documentation of family planning counselling and screening for gender-based violence. Family planning counseling provided during pregnancy was documented in five countries. At the same time, discussion around the birth plan is recorded as having been performed or not in five countries (Afghanistan, Argentina, Sweden, USA, and Zimbabwe). Screening for occurrence of gender-based violence is documented in Afghanistan ("yes/no" field), while breast cancer screening though physical examination is recorded in Zambia and documented as Yes/No/Not done.

Childbirth
The care received by the mother from the moment of admission to discharge after delivery is recorded mostly in a maternity register. However, some other information such as the personal identifying information of the mother's medical history is recorded in supplemental documents such as ANC cards or other records retained by the client. Gestation age at delivery is documented in only four countries (Afghanistan, Kenya, South Sudan, and Uganda) out of the 18 whose childbirth registers were reviewed. The estimated date of delivery and last known menstrual period dates are not consistently documented in the maternity register.
Monitoring of mother and the fetus during labor. Eighteen country childbirth registers or source documents were reviewed. The information captured in the registers distinguished between vaginal deliveries, assisted deliveries, and cesarean section deliveries in the register. Information in the register on whether a partograph was used is recorded in four countries, as shown in Table 5. Important information might have been lost for those countries not recording the use of the partograph, which allows a close monitoring of vital signs and procedures during childbirth.
Use of a uterotonic (Oxytocin, Ergometrine or Misoprostol) immediately after delivery is documented in 11/18 (61%) countries, while use of magnesium sulphate in the management of pre-eclampsia/ Eclampsia is recorded in three countries. Complications of childbirth are documented in 14/18 (78%) countries. The complications documented include vaginal tears, blood loss or need for transfusion, retained placenta, obstructed, or prolonged labor, ruptured uterus, and sepsis (Fig 2).
PMTCT during childbirth. Childbirth presents another window of opportunity to implement PMTCT. However, documentation of HIV status of the mother is not commonly included in many maternity or childbirth registers. In this review, five countries (Afghanistan, Eswatini, Guyana, Kenya, and Malawi) capture data on either HIV testing at delivery or the HIV status at admission. Couple HIV testing and counseling (HTC) and partner HIV status

PLOS GLOBAL PUBLIC HEALTH
Pregnancy, childbirth and post-partum registers review from 21 countries monitoring of the mother in the childbirth registers (only Kenya and Uganda). Guyana and Kenya are the only countries that record syphilis testing in childbirth registers. Best practices in newborn care. The newborn's first hour of life requires key lifesaving interventions to implement immediately postpartum to ensure the survival of the baby. For newborns experiencing difficulty in breathing at birth, resuscitation is documented in registers from eight countries while breastfeeding within the first hour is recorded in six countries out of the 18 childbirth registers reviewed. The practice of skin-to-skin is documented in Kenya, and immediate cord care through application of either chlorohexidine or chloramphenicol is recorded in Ghana, Kenya, and Nigeria.
In 15 out of the 18 countries, live births are clearly recorded, distinguishing them from fetal and newborn death. The sex and weight of the baby are recorded in 10 and 12 countries, respectively. The Apgar score is recorded in 10 countries, with most of them categorizing by the time the assessment was done (one, five, or 10 minutes after birth). Recording of birth deformities is done in five countries as shown in Table 6.
Services provided to mother prior to discharge. Health promotion interventions provided to the mother prior to discharge include counseling on infant and young child feeding in Afghanistan, Nigeria, Norway, and Tanzania. Distinction is made in Zambia between exclusive breastfeeding, replacement feeding, and mixed feeding. Regarding family planning counseling, only three countries capture these data (Bolivia, Nigeria, and Tanzania). In Tanzania, the type of postpartum family planning method selected by the woman is documented. In some registers, the level of detail includes whether the method is provided pre-discharge.

Mortality and newborn outcomes
Childbirth or labor and delivery registers provide data on maternal and newborn outcomes. Results from the analysis of maternal and neonatal deaths is presented for the same 18 countries.

Maternal death
In 14 out of 18 countries (78%), data on the status of mother at discharge (alive or died) are captured in separate columns, except for Nigeria, Sweden, and Uganda, where "maternal death" is not included as a separate column. Data on mothers who die are sometimes captured through an open text field, such as in Ghana, Kenya, South Sudan, and Zambia, where the cause of death is recorded in this manner. In some other countries, such as Tanzania, maternal deaths are recorded under "discharge" summary, the status of the mother being designated as either alive or dead.

Fetal and newborn death
Where the death of the baby occurs before birth, this is recorded as a fetal death. Registers in Nigeria and Norway have codes to record abortion. Regarding stillbirths, 11/18 (61%) of the countries distinguish between fresh and macerated stillbirths. Bolivia and Malawi do not make such distinctions, although stillbirths are reported ( Table 7). The timing of newborn death is recorded in different ways across the registers. Half the countries document early death occurring less than 7 days after delivery. In Norway, where an electronic register is used, there are separate codes for recording early neonatal death versus death occurring after 7 days, but before 28 days. South Sudan and Zambia have spaces to record newborn deaths whose timing is unspecified (Table 7).

Utilization and timing of postnatal care visits post-discharge
Some countries record the date of the visit and the date of birth therefore making it possible to calculate the timing of the visit. In countries using cross-sectional registers, service providers count the number of visits from a mother-held record, the number of the visits a woman has attended, then enter the visit number in the postnatal care register. The timing of the PNC visit was categorized as day of birth or immediate care after birth, within 48 hours or day 2, day3-6, days 7-14, and at six weeks. The schedule of reporting varied widely across countries, Immunization (e.g. BCG) ART/ Niverapine

Breastfed within an hour
with countries like Zambia using incredibly detailed schedule as follows: 1 st hour, 6 th hour, 12 th hour, 18 th hour, 24 th hour, 48 th hour, 6 th day, and 6 th week. In most countries, the timing of the visit was unspecified.

Routine checks during postnatal period
Routine clinical checks, counseling, laboratory tests are done on the mother while the child receives immunization and routine checks. Blood pressure is recorded in PNC registers of five of the countries reviewed, while complications related to childbirth are recorded in seven countries. Some of the complications documented are postpartum haemorrhage, lochia (excessive or foul smelling), and presence of a fistula. Examination of the site of cesarean section is recorded in Zambia and Kenya). The mother's hemoglobin level is recorded as such or anaemia/ pallor. Some of the less commonly documented practices include examination of the vagina, breast, and the uterus.

Screening and counseling services
The most common type of counseling service during postnatal visits is on family planning and nutrition counseling for both the mother and the baby. Nutrition counseling includes information on the three feeding methods described earlier. Micronutrient supplementation provided, mainly iron and folate, is recorded in 9/13 (70%) of the countries whose PNC registers were analyzed. Postpartum family planning is recorded in 8/13 (62%) of the countries. Screening for TB, cervical cancer, and STIs is recorded in 3, 2, and 1, countries, respectively.

PMTCT during the postnatal period
The postnatal period provides another opportunity for PMTCT so that infants born to HIVinfected women remain uninfected. HIV testing is recorded in the PNC registers of six of the Table 7. Abortion, fetal and early death of the newborn.

Stillbirth-Macerated
Neonatal death (unspecified) X Zambia X X X X countries analyzed while information on whether the mother is on ART is recorded in five countries. Couple testing and the HIV status of the mother's partner is recorded in Kenya and Zambia only (Table 8).

Postnatal care for the newborn
In addition to documenting the scheduled immunizations provided to the newborn, routine checks primarily focus on newborn danger signs or complications such as jaundice and failure to gain weight as expected. PNC registers in five of the 13 countries included in the analysis record newborn danger signs. Regarding preventing HIV infection, ART prophylaxis for the newborn is recorded in four countries (Afghanistan, Eswatini, Tanzania, and Zambia), and it is only in Eswatini where provision of early infant male circumcision is recorded.

Sub-analysis of registers from Ghana, Kenya, Nigeria, and Zambia
Results of analysis of registers from four countries (Ghana, Kenya, Nigeria, and Zambia) where the 2017/2018 and 2019/2020 versions of the registers were available showed that there is an apparent change towards inclusion of data elements that are critical in the computation of key maternal and newborn care indicators. Overall, 21/27(78%) key data elements for tracking maternal and newborn care are present in the latest versions of the registers reviewed ( Table 9 and S1 File of key priority indicators). The ones that are present in to varying degrees are: ANC visits <12 weeks, fetal heart rate, partograph use, newborn resuscitation, breast/ Cervical cancer screening and contraception counseling.

Discussion
This analysis reviews the registers for ANC, childbirth, and PNC for 21 countries in the world's five regions conducted through abstracting data elements from ANC, childbirth, and PNC registers with a focus on evidence-based interventions as well as outcomes for the woman and newborn. Overall, there is great variation in the number of data elements ranging from 24 in Nepal's childbirth register to 188 in Eswatini's ANC register, as well as the types of elements that are captured. Countries' registers also vary in terms of longitudinal versus cross-sectional format. This review has demonstrated the lack of harmonization and consistency in the types of data collected, how they are collected and organized or documented. Given that these registers are a critical source of maternal and newborn data, information gathered plays a key role in program planning and tracking progress towards achieving national and global goals and targets.
Most of the ANC registers reviewed capture routine information about antenatal care, such as gestation age, weight, height, blood pressure, and hemoglobin level. However, there were either gaps in capturing key data as per the 2016 WHO recommendations [32] or lack of . Majority of the countries recording eight or more ANC visits are the ones using electronic registries, indicating that use of electronic medical records provides a more adaptable system for introducing new or updated indicators for alignment with evolving clinical guidelines compared to paper registers. It's noteworthy that most ANC registers included relevant medical history for the pregnant woman, although only one third included complications from previous pregnancies. Most countries captured recommended screening practices such as blood pressure measurement, hemoglobin levels and syphilis and HIV screening and treatment. However, only one third of registers captured whether an ultrasound was performed. In addition, the registers capture screening but not if any action was taken to address risk factors.
For childbirth, the majority of countries capture data on obstetric complications, but few do preventative or management interventions for postpartum hemorrhage or pre-eclampsia. Although there is evidence of documentation of key data on childbirth, it is possible that some other materials such as the partograph, which is recommended by WHO for capturing data during active labor to inform care, should supplement the information captured through registers [33]. Additionally, few countries tracked whether the partograph was used or fetal heart monitoring during childbirth. Furthermore, data on breastfeeding within one hour of birth, newborn resuscitation and BCG immunization were also not routinely captured in most countries.
The PNC registers had the fewest data elements available and the least standardization in terms of what was captured for the woman or newborn. In general, there was more information captured for the woman than for the newborn, especially around counseling on nutrition and family planning, with information for newborns focused on immunization and routine checks. Given that information on newborns, especially immunization, is likely to be captured in other registers, it's not surprising that not much of newborn care is recorded in PNC registers. In sub-analysis of four African countries, there was some improvement over time in terms of the types of data elements collected. These countries track the same data elements in ANC registries because of epidemiological reasons (e.g. HIV testing, STIs screening or malaria treatment). Several newer data elements included newborn care indicators recommended by WHO, UNICEF and other stakeholders [7,[30][31][32].
There were limitations to this review. First, although every effort was made to obtain the most recent registers, these documents were not always available for all selected countries. In addition, many of the data elements reviewed were relevant to only a sub-set of countries (e.g. malaria prevention and HIV prevention and management). It was often difficult to interpret how the registers were supposed to be completed especially where instructions for completing the register were not available. Finally, we were only able to conduct a sub-analysis of four countries to assess change over time, due to limitations in obtaining the relevant registers from more countries. It would be helpful to expand this analysis to additional countries and to include the data summary forms.
Although there are global recommendations for maternal and newborn indicators to be collected within routine data systems [7,[30][31][32], this analysis highlights the great variation among and between countries. There is a need for additional guidance, standardization, and reduction of data elements to better focus on capturing the most important evidence-based practices for improving the quality of care as well as tracking progress toward national and global targets. This analysis also did include the burden on health care workers to collect these data elements in addition to providing quality care, but it is an essential consideration when selecting and updating registers and forms for routine data systems.
Improving the quality and consistency of data reported through a national HIS, requires a thorough understanding of facility registers with respect to how data recorded in facilities can be most easily translated into aggregate indicators. Taking these factors into consideration when designing a register allows for more accurate reporting and for inclusion of indicators that provide information on quality of care, such as number of service contacts and timing of interventions received.

Conclusion
The ANC, childbirth and PNC registers are a major source of essential maternal and newborn data, but standardization in recording of data elements varies across countries. Managers and frontline health workers use these data to monitor the performance of their national health system and to guide the continuous improvement of health care services for women and newborns. The results of this analysis will complement additional work and help to inform the standardization of global indicators and provide information for other countries seeking to introduce indicators in their health systems. There is need for further analysis of the HMIS indicators reported from data captured in the registers to understand what data elements/indicators help more to suggest policy changes and inform programme decisions.
Supporting information S1 File. Priority indicators for maternal health. (DOCX)

MCHIP. A Review of the Maternal and Newborn Health Content of National Health Management Information Systems in 13 Countries in Sub-Saharan Africa and South
Asia. MCHIP, MCSP: 2014. Available from: https://www.mchip.net/technical-resource/a-review-of-the-maternal-and-newborn-healthcontent-of-national-health-management-information-systems-in-13-countries-in-sub-saharan-africaand-south-asia/.