You must be registered and logged in to participate in this eForum.
If you have not already done so, please read the background documents that we previously posted.
Background
The Demographic and Health Surveys are nationally representative household surveys. A Household Questionnaire is used to identify members of the sampled household who are eligible for an individual interview and to collect basic information about the household. Eligible respondents for the Individual Women’s Questionnaire are women of reproductive age (age 15-49 in most surveys.) Data are also collected from men in many countries; the Individual Men’s Questionnaire is used to collect data from men age 15-59.
DHS has a standard Individual Women’s questionnaire that is used as a starting point for data collection, which allows for the establishment of a core set of information to be collected for all countries. However, countries often also examine issues and problems of national interest inserting additional questions into the questionnaire. This results in some differences in final reports, illustrates the variety of country level interests, and highlights the range of data it is possible to collect with the DHS. [Maternal mortality information is collected in the Maternal Mortality Module of the questionnaire, which is an optional addition to the standard questionnaire. This module asks information about the respondent's brothers and sisters, including the number of pregnancies of their sisters and if they are deceased, and whether or not they were pregnant or within two months postpartum at the time of death.]
The standard questionnaires are revised periodically (roughly every 5 years). The recently revised version of the Individual Women’s Questionnaire is the basis for this e-forum discussion.
Section 4, “Pregnancy and Postnatal Care”, of the new questionnaire contains 33 standard questions related to maternal health, including questions on the frequency and content of antenatal, delivery, and postnatal care for births in the three years prior to the survey. This section also includes three questions on desired timing of pregnancies, three questions on the weight of children at birth, four questions on infant feeding practices and eleven questions on child immunizations (the remaining information on child health comes from section 5 of the questionnaire).
The questionnaire contains 18 questions related to antenatal care which are only asked about the last birth. All women are asked about tetanus toxoid injections, iron supplementation, treatment for intestinal worms, utilization of antimalarials, and attendance at antenatal care. Questions on vision trouble during the day and night blindness have been deleted from the revised version. For those who respond “yes” to having attended antenatal care, follow-up questions include the timing of the first visit, number of antenatal visits and physical components of the exam (blood pressure, urine sample, blood sample).
Changes from previous questionnaire:
The new revised questionnaire eliminated four questions related to antenatal care:
- During any of your antenatal care visits, were you told about the signs of pregnancy complications?
- Were you told where to go if you had any of these complications?
- During this pregnancy, did you have difficulty with your vision during daylight?
- During this pregnancy, did you suffer from night blindness?
The questionnaire contains 3 questions on delivery. All women are asked about the location and type of assistance at delivery for each of their previous births in the recent past. Those who delivered in a health facility are asked whether the birth was a caesarian. Follow-up questions on length of stay at place of delivery and whether they received a postnatal check up before discharge have been deleted from the revised questionnaire.
Changes from the previous questionnaire:
The new revised questionnaire eliminated five questions related to delivery:
- How long after delivery did you stay there (in the place of delivery)?
- Before you were discharged after (NAME) was born, did any health care provider check on your health?
- How long after delivery did the first check take place?
- Who checked on your health at that time?
- Why didn’t you deliver in a health facility?
The questionnaire contains 12 questions related to postnatal care and the postpartum period. All women are asked if they had a postnatal visit following their most recent birth. Based on their response, follow-up questions include length of time between delivery and postnatal visit, who performed the postnatal check up, and location of the postnatal visit. Regardless of whether they received postnatal care, women are asked about receiving vitamin A in the first two months after delivery, whether their menstruation has returned since their last birth, and the length of time between delivery and initiating breastfeeding for the last birth. Questions regarding length of time with no menstruation, length of postpartum abstinence, and breastfeeding are asked of all pregnancies.
Changes from the previous questionnaire:
The new revised questionnaire differs from the previous version in the following ways:
- Whether a woman had a postnatal care check up is only asked for the last pregnancy (previously it was asked for all pregnancies)
- A question on the timing of initiation for breastfeeding was added.
- Questions on postnatal check ups for babies were eliminated.
Discussion Topics
We encourage the following types of inputs:
- Comments on existing questions—their utility for monitoring, programming, policy making
- Recommendations for improving the wording of questions
- Recommendations for additions and deletions of questions
- Suggestions for additions or modifications to standard response categories for questions
You must be registered on the site to post a comment. To post a comment, click on the word ‘Comment’ in blue below. The eForum will remain open until June 15th. Comments will be summarized and submitted to the DHS program by June 18th.
Welcome to the e-forum. We look forward to your comments
The Demographic and Health Surveys along with the Multiple Indicator Cluster Surveys are the best source of information for levels of coverage for many reproductive health interventions in developing countries. Recently the Child Health Epidemiology Reference Group, Johns Hopkins and the Futures Institute with Gates Foundation funding through the US Fund for UNICEF have collaborated on a computer projection model that will project the mortality impact of scaling up proven health interventions for maternal and child health. The name of this model is the Lives Saved Tool (LiST) . It can be downloaded at http://www.healthpolicyinitiative.com/index.cfm?id=software&get=Spectrum. LiST is already being used at the country level and internationally to estimate the levels of health intervention coverage that will achieve the MDGs and other health goals. For example the tool was used to estimate the mortality impact for many interventions in a high level task force headed by Gordon Brown. The projections that are made with this model are extremely useful for determining the direction of MCH programming and resource allocation within the programs.
The DHS collects information on coverage for many of the proven interventions included with LiST. Among these are the overall level of antenatal care, the percent of births that are attended or occur in facilities, tetanus toxoid, etc. However, there are many other interventions, also proven to be effective, that are not collected by the DHS. It would be very useful if a systematic inquiry were made to investigate where simple questions could be added to the DHS to populate the Lives Saved Tool. Among the indicators that would be useful for implementing the Lives Saved Tool are: % of pregnant women tested & treated for syphilis, % of pregnant women receiving balanced energy supplementation, % of pregnant women receiving multiple micronutrient supplementation, % of women delivering in facilities that receive antibiotics for PRoM, % of women receiving antenatal corticosteroids, & home deliveries performed in a clean environment. This list is only indicative: i.e., it is not presented as comprehensive or based on a close examination of the current or proposed DHS questionnaire.
A frequent objection to adding questions to the DHS is that there is not broad based demand for the proposed questions or that they will not be used systematically within the public health community. Adding a few key questions that would feed the LiST model would have immediate utility, would be used within the public health community (at the country level and internationally) and would have impact on decision making (again at the country level and internationally).
Bill Winfrey
Futures Institute
In many of the developing countries unsafe abortion has a substantial role in maternal mortality. Continued information is needed on the consequences of unsafe abortion, particularly potential long-term squeal to portray the magnitude and severity of the problem due to unsafe abortion. Abortion is a sensitive issue for women all over the world, secrecy and stigma surrounding abortion are common issues. However, it is greatly important to know about the change in abortion related outcomes which could be linked to program interventions. I like to see this category of questions in the DHS.
As birth preparedness is a topic of increasing interest, but scant data, the DHS survey could provide an important opportunity to explore the advice women receive about birth preparedness during their antenatal care visits, as well as the extent to which they prepare for childbirth. It is noted that the two earlier questions (i.e. Q412 and 413) on counselling on signs of complications and where to go for complications have been removed, which is regrettable.
I would suggest retaining question 413 if possible, and adding in a new question after Q411 on birth preparedness counselling during antenatal consultations as follows:
During any of your antenatal care visits, were you advised about: (multiple responses allowed)
Your expected due date
The benefits of delivering in a health facility
Signs of complications during pregnancy and delivery
Setting aside funds for delivery
Making transport arrangements
Identifying a blood donor
Q413 could be revised slightly as follows:
Were you advised where to go if you had any complication during pregnancy or delivery? (yes/no)
I would suggest adding two new questions on household preparation for birth immediately before 432 (referring only to the most recent birth):
1. Did you or your family make any advance preparations for the birth while you were pregnant? (yes/no) [If no, skip to Q432]
2. Which of the following preparations did you make for the birth? (multiple responses allowed)
• Discussed place of delivery
• Set aside funds for delivery
• Made arrangements for transport
• Identified blood donor to accompany to the facility
DHS surveys could also provide an important opportunity to understand household decision-making about childbirth. Toward this end, it would be great to add a question on the decision-maker about place of delivery. The easiest place to add such a question would be immediately after Q435 – something along the lines of:
Who made the decision about where the delivery of (NAME) should take place? (one response allowed)
• Respondent
• Husband/Male partner
• Other family member
• Non family member
Another important issue is whether women are required to pay for the delivery or have to purchase supplies for the delivery. Cost is known to be a major barrier to the use of maternal health services, and particularly delivery care. Even as cost-sharing programs are being phased out, women still incur costs, even for services that are nominally provided free of charge. However, when there is no reliable data on whether or not women are paying for services, it is easy for central-level decision-makers to ignore the issue. To explore this, it would be great to add in a question after Q438 (with skip pattern to be adjusted):
1. Did you have to pay for any supplies or drugs for your delivery? (yes/no)
2. Did you have to pay any service charges or fees for your delivery (yes/no). [Skip to Q445]
It seems unfortunate to have dropped Question 443, which was in the old version of the questionnaire (i.e. Why didn’t you deliver in a health facility?) as this can be illuminating. If possible, this question should be retained.
There appears to be an error/typo in the skip pattern for Q438, as it jumps to 455 when it should probably go to 445. The skip to 455 would mean that all women who delivered in a health facility will not be asked any questions about postpartum care.
Lastly, it would be great to include a question to explore the content of women’s postpartum care check-up(s) immediately after Q447, along the lines of:
During your check, did the provider do any of the following? (multiple response allowed)
• Examine your abdomen?
• Examine your breasts?
• Perform a vaginal examination?
• Ask you about vaginal discharge?
• Ask you about bleeding?
• Discuss family planning?
• Give a family planning method?
• Give advice on breastfeeding?
• Give advice on infant care?
Otherwise, no information is collected on the actual content of postpartum care.
Thanks to the MHTF for getting this underway. I encourage everyone to forward Ann Blanc’s email NOW to all of their colleagues in the maternal health world. NOW is the time to act.
The summary of the questions on pregnancy and pregnancy care is striking. 18 questions on antenatal care, 3 on delivery and 12 on postpartum care. This absolutely does not reflect a focus on the time period where the risk is the greatest and when the large majority of maternal deaths occur. The DHS questionnaire may not change with our suggestions, but it definitely will NOT change without them. Please – strike up conversations now regarding key maternal/newborn health issues that are not currently represented in the questionnaire.
Specific suggestions:
I suggest changing the formulation of the question on caesarean to be: Was {NAME} delivered by caesarean, that is, was there an operation to cut your stomach to deliver the baby? Or something along those lines to define the word caesarean, as this word is sometimes used with/by patients to mean any kind of surgery that occurs at birth. It is also important to differentiate it from episiotomy. I think defining caesarean in the question would be helpful to both the respondent and the interviewer.
More comments coming…
In response to the comment on including questions related to abortion:
Are there one or two specific questions related to abortion or postabortion care that would be particularly helpful? The 2007 Ukraine DHS has a chapter on abortion which might be a good reference to identify questions. For example, they report on:
- Lifetime experience with induced abortion (by background characteristics)
- Use of contraception before abortion
Feedback from Ann Fitzmaurice, Wendy Graham, Jacqueline Bell, David Braunholtz, Emma Pitchforth, Sheetal Sharma (Immpact + LSE)
We agree with several of the previous comments, particularly those pointing out the limitations of the questions around delivery and the desirability for more on birth preparedness
We assume the maternal mortality module is not currently being revised as it hasn’t appeared in any of the available materials. One question we would strongly like to see added to this, if/when the time comes, is on the place of death (facility/home). In fact we feel this is a very useful piece of information to have regarding any death. With population-level data on place of death we could start to make some sense of selection biases in facility-based mortality. An additional question on region of death would also help to localise estimates.
***Suggested additions to core questionnaire***
In view of the strict need to limit the number of questions in the core questionnaires, we will propose minimal additional questions in this round of revisions. However, given the long list of potential questions we have come across with so far, on this forum and in other places, we would like to propose that DHS consider introducing a dedicated maternal health module, to explore the key issues in more depth in countries where the need is greatest. We would also like to see more special maternal and newborn health surveys, along the lines of that in the Philippines, Ghana and Bangladesh.
We would like to ask the forum whether these are things that should be lobbied for, and if so what is the best way to approach DHS and funders?
For us, the question we would most like to add to the revision is on the person attending the delivery. At present women are asked to name all their attendants, and then the birth is generally reported in DHS reports (& used by the Countdown) as ‘attended by’ the most qualified attendant mentioned. This must cause much misclassification, as, for example, the doctor who sees a woman at admission but then has no part in her care is mentioned in the list and in the analysis becomes her attendant. We think it would be an improvement (i.e. reduce misclassification) if women where asked who was their main attendant; and then who else attended.
An adaption to an existing question on place of delivery we also advocate is that all women are asked the name of the place they deliver (at present it is only asked if the usual classification is not clear). This would enable analysts to calculate the distance a woman travelled to deliver (assuming GPS on health facilities were known/collected). It would also be very helpful for DHS analysts to calculate the distance from each household to the nearest facility offering delivery care, as this is an important predictor of care-seeking behaviour and may help to highlight inequity in service provision/use. (At present household coordinates are scrambled up to 5km before release to the public for reasons of confidentiality, which means they are not very useful in analysis of associations between distance to facility and service use.
A single question on birth preparedness, as already raised by others in the Forum, with a focus on person and place of delivery, finance, and transport would also be very welcome. This could be usefully followed-up with a question on where the woman intended to deliver. A delivery planned at home that end up in a hospital can fairly safely be assumed to involve a complication.
***We have a few ideas for questions that could be removed***
Questions 405-406 don’t really disentangle the ‘unplanned’ from the ‘unwanted’ pregnancy so fail to give a clear view on the woman’s feelings towards her pregnancy; and 406 doesn’t add much to concepts of birth spacing – there are already 4 pages on contraception.
Question 411 tries to capture quality of ANC, but the reporting of a one-off measurement gives little idea of this and we would suggest removing it from the core questionnaire, but keeping questions around QOC in a MH module. We would like to emphasise that we feel quality of ANC is an important topic, but not well captured by the currently-proposed questions. This would be a key theme for a dedicated maternal health module.
We don’t understand why there are 3 questions on TT outside the current pregnancy and wonder whether these could be condensed (a total of 5 on TT)? Similarly the purpose of the supplement questions is not clear.
***A few other questions to the forum***
We would like to know from our colleagues in perinatal health what they think of questions 229-236 as a method for the measurement of stillbirths. We also feel that the birth (live birth?) history currently in place should be expanded to a delivery history to better capture stillbirths. We feel this is key since currently most statistics on skilled attendants rates are based on DHS (or MICS) and on live births. In other words, we have no sense of where and with whom stillbirths happen. Perhaps the Forum should link-up with the newborn/stillbirths community to push for this. We are also surprised to see the removal of questions on checking the newborn, and would support their re-inclusion – in as much as they also give an indication of contact with a provider in the postpartum period.
In the series of malaria questions we don’t understand why there is no question on the use of bed nets.
For caesarean section, we agree the wording should be improved to clarify this is delivery involving surgery. We also wonder if it would be useful to distinguish between “planned/booked/scheduled” CS versus emergency, since there clearly are many countries where the former category is on the increase for non-medical reasons (obstetrician, maternal preference).
Questions on postnatal check (Q453) – we think the wording here should be changed to remove the specific mention of a TBA, as this seems to imply TBAs are routine & accepted postnatal care providers i.e. why just specifically mention TBA.
***For a dedicated maternal health module***
We would like to see, among other things, detailed questions around women’s perceptions of delivery care as a measure of quality; birth preparedness; awareness of danger signs and what to do about them. Perceptions of quality of care may be an important determinant of utilisation. Women will have a perception of the quality of ‘clinical’ care (availability & competence of staff, presence / state of ‘infrastructure’) though this may be tricky to elicit, and can be asked to what degree they were treated with dignity and respect.
A few ideas on quality of care questions (*not* for the core questionnaire):
To assess quality of care, in addition to women’s perceptions, rates of facility births, and attendants present, it is important to also have an objective assessment of quality of clinical care delivered. It may be possible to obtain valid information from women on whether known good practices were carried out (e.g. support during labour, continuous monitoring, wearing gloves for examinations) and whether any harmful or unnecessary procedures were carried out (e.g. pressing on abdomen, shaving) – though this would need field-testing.
We are not sure whether anything of value about the women’s perception of QOC can be gathered using just one or two questions, ie in the space / time available in DHS core module. We could suggest something like:
“If there were no costs or problems with transport, would you prefer to deliver a baby at home with TBA / at home with community midwife / local health centre / local hospital?”
or : “Would you feel safer delivering a baby at home with TBA / at home with community midwife / local health centre / local hospital ?”
and: “In general (TBAs) are (respectful / somewhat respectful / not at all respectful towards patients)”
“In general (auxiliary midwives) are (respectful / somewhat respectful / not at all respectful towards patients)”
“In general (midwives) are (respectful / somewhat respectful / not at all respectful towards patients)”
“In general (Doctors) are (respectful / somewhat respectful / not at all respectful towards patients)”
We thought such questions could perhaps be usefully asked of all women of reproductive age – ie not just with respect to a specific recent delivery (obviously in analysis could separate out those who had a recent delivery, and where delivered).
There is an interesting section (was 490 in children’s health section in Indonesia 2003 DHS, with clear relevance to delays 1 & 2 in the ‘3 delays’ model – with (in Indo DHS) questions about barriers to getting care for the respondent – were the following a big problem or not: permission / money / distance / transport / lack of companion / lack of female-health provider.
We would like to see this section retained / if possible improved – and asked of all WRAs. e.g. could add or substitute “the safety of care in the facility” “the lack of respect towards patients” to the list.
This is a very brief comment – there is no way that a well informed questionnaire on maternal health can cut down the questions on delivery – this is the period thatis most prone to risk…
I support the suggestion on the wording of the c-section question
I also support this but am not sure this is easy as some of the key maternal interventions only apply to a small subset with complications. However among the ones for all pregnant women AMTSL (or oxytocic injection) would be good if we though women could report on this. Perhaps also use of clean birth kit or providers that washed hands? And again fomr monitring labour, this is hard to thik of good questions. Perhaps if listened to baby’s heart ?
I have responede to some of the comments with my ideas. I would like DHS to consider having a dedicated matenral child health survey. I can see the need to reduce questions and the latest DHSs have been too long. I can see the advantages of having many topics as this allows multiple varables to be examined togther but if some questions needed to be left off perhaps the survey could be split into two: a) Family Planning and HIV; and b) Maternal, Perinatal and Child Health. THen we could have more MCH questions
On a seperate point I would like to see routine tabulations of where women delivered by who delivered them
Completely agree
Strongly endorse need for questions on the place of death but with options for (hospital/ health centre/ private clinic/on way/home) or better.
ALso in maternal and newborn health surveys in Ghana, groupings of ICD classifications of causes of death were strange I thought.
If the suggestion to have the main attendant is taken up, I would still like to see the data tabulated in teh old way too as this also gives a measure of the “highest” level of care accessed. For example if midwife did delivery but doctor these, the doctore presumably cold be called on.
I don’t think DHS should scramble GIS coordinated irrevocably as this loses potentially useful information. If confidentiality is a concern, then perhaps a special “trusted third party” can do the linkage, as is done for example with NHS data. In our case, we have been linking Health Facility Censuses to DHS to look at the role of distance but have to deal with miscalsification due to scrambling.
I think there is merit in Q411 (assuming this is the one that asks about blood pressure blood and urine tests) and do not agree it should be removed though I accept that one measure is not useful as it is a first stab at quality
I agree with comments on TT overkill and am also unsure of the supplemenation questions
I wold endorse aksing stillbirth questions (ie pregnancy history on calendar) not jsut live births if research suggests this is reliable enough and also support re-inclusion of checking the newborn,
I agree with the need for field testing of questions on both beneficial and harmful practices linked to delivery care
Pingback: Baby-Parenting.co.uk - Pregnancy Problems & Pregnancy Complications
I strongly support J. Bell and co’s comment re: adding a follow-up question on booked versus emergency caesareans.
I would also like to request an addition to the tables available on STATCOMPILER: a cross tab of where one delivers and birth attendant.
RE: the questions to identify stillbirth. I find the current series of questions extremely difficult. I can’t really imagine training interviewers to go through all of these gymnastics, and jumping forward and backward in time to pick up one kind of outcome and then another. Plus, I think it must be difficult for a woman to respond. I too agree that a complete pregnancy history makes more sense. Thus said, I have only been able to identify one validation study that shows the benefits of a pregnancy history over a live birth history (Espeut D. A validation of birth and pregnancy histories in Matlab, Bangladesh. [ Johns Hopkins University School of Hygiene and Public Health; 2002.) Additional methodological work is needed on this subject.
In light of the strong effort on the part of DHS to reduce the size of the questionnaire and the fact that the current plan is to keep the contraceptive calendar, myself and a group of colleagues have decided that – under these circumstances – perhaps trying to improve calendar data – is a compromise we can live with.
I do agree that a subset of follow up questions on recent stillbirths would be tremendously important for the MNH community -in particular, to know the place of birth and birth attendant for pregnancies ending in stillbirth. Advocates will say publicly that “most stillbirths occur at home”, but there are no global data to support this statement, and I think there is a very good chance that the opposite is true. The inclusion of 2 questions on recent stillbirths (place of birth, attendant) would make a very important contribution to the field.
It would be a very helpful service if DHS could prepare datafiles for the contraceptive calendar that have already been transformed so that each record is a month of observation, with the woman’s characteristics attached to each record. I suspect that this would increase use of the calendar data.
I strongly agree with all comments made related to the inclusion of questions concerning birth preparedness. I am adding here only a few additional reasons such questions (412 and 413) should not be removed from the core DHS questionnaire. Capturing information on birth preparedness can shed insight into the factors that facilitate women’s ability to access skilled delivery care. Also, collecting data on the counseling/educational content of ANC visits is important for assessing the quality of care received, and progress in the implementation of the new WHO antenatal care model (educating women about planning for safe birth, recognition of warning signs and how to deal with them is one of three topical areas to be covered in each ANC visit according to the basic component of this model).
I appreciate DHS’ effort to maintain balance between asking every important question and keeping the questionnaires from being too long. However, I think it is important to restore the questions on whether women were told about the signs of pregnancy complications and whether they were told where to go if they had any of those complications. At least the questions should be asked about the last pregnancy. Information is key for prevention and these questions are important for determining how much women know about those danger signs.
I also think the question about why women do not deliver in a health facility should be retained. Having this information overtime helps check whether and how obstacles that women see for not delivering in health facilities are changing or not and to examine strategies for promoting deliveries in facilities.
I support the suggestion that DHS should help to improve users access to the calendar. Although it has important information, the calendar is underutilized because of the difficult many people have with accessing the information. It is good to see that the calendar is now being used in some of the Sub-Saharan African countries.
I want to emphasize the problem with the scrambling of geographic coordinates of the clusters. DHS policy is now to add up to 5km error to locations in rural areas and to destroy the correct/unscrambled data. In some countries, there is geographic information on health facilities. Linking these to the DHS clusters would allow to study service availability and distance alongside the usual individual and household determinants. While 5km more or less may not matter for things like malaria maps, it matters very much for pregnant women walking to health facilities and the resulting misclassification prevents good analyses of distance effects. Given that much effort is put into getting the geographic data in the first place (modern GPS equipment etc), this seems such a waste of potential. As with other sensitive data, there must be ways of allowing the research community to study it while still preserving confidentiality. It would be nice to see some efforts in this direction.
And another note. I also think it would be useful to have a question to find out what the intended place of delivery was and whether the woman delivered in the indended place, on the way or elsewhere. This could tell apart women who planned to go to a facility but did not manage to get there and women who only went to a facility when complications developed at home (one could also ask when the decision was made to go). Determinants for planned facility delivery are likely to differ from those for emergency facility delivery, and it would be useful to be able to separate those.
Pingback: Pregnancy Problems & Pregnancy Complications