Impact of maternity waiting homes on facility delivery among remote households in Zambia: protocol for a quasiexperimental, mixed-methods study

Introduction Maternity waiting homes (MWHs) aim to improve access to facility delivery in rural areas. However, there is limited rigorous evidence of their effectiveness. Using formative research, we developed an MWH intervention model with three components: infrastructure, management and linkage to services. This protocol describes a study to measure the impact of the MWH model on facility delivery among women living farthest (≥10 km) from their designated health facility in rural Zambia. This study will generate key new evidence to inform decision-making for MWH policy in Zambia and globally. Methods and analysis We are conducting a mixed-methods quasiexperimental impact evaluation of the MWH model using a controlled before-and-after design in 40 health facility clusters. Clusters were assigned to the intervention or control group using two methods: 20 clusters were randomly assigned using a matched-pair design; the other 20 were assigned without randomisation due to local political constraints. Overall, 20 study clusters receive the MWH model intervention while 20 control clusters continue to implement the ‘standard of care’ for waiting mothers. We recruit a repeated cross section of 2400 randomly sampled recently delivered women at baseline (2016) and endline (2018); all participants are administered a household survey and a 10% subsample also participates in an in-depth interview. We will calculate descriptive statistics and adjusted ORs; qualitative data will be analysed using content analysis. The primary outcome is the probability of delivery at a health facility; secondary outcomes include utilisation of MWHs and maternal and neonatal health outcomes. Ethics and dissemination Ethical approvals were obtained from the Boston University Institutional Review Board (IRB), University of Michigan IRB (deidentified data only) and the ERES Converge IRB in Zambia. Written informed consent is obtained prior to data collection. Results will be disseminated to key stakeholders in Zambia, then through open-access journals, websites and international conferences. Trial registration number NCT02620436; Pre-results.


5
Is MWH a government policy of Zambia? There is no reference while describing it in the introduction. However reference has been used in the following sentence 12 6 If BMJ has no word limitation the intervention needs to be more elaborated, otherwise it would be difficult to relate with the following sections of the protocol. In addition, if MWHs initiative is successful how it would be replicable to other low resource countries cannot be answered 12-15 9 Controversial description between intervention and control groups. The construction of sentences needs to be clearer on which group is consisting of what.

9
The selection process due to political constraints needs to be clarified to understand its rationale for non-randomization of 20 clusters 19 10 Does the MWHs has the facility to store the drugs properly 15 12 Generally the sample size for qualitative methods is flexible, however 10% sample for the qualitative component (240)  The study design does not have the power to estimate and compare outcomes like maternal and neonatal deaths. The outcomes can look at other severe adverse maternal and neonatal outcomes rather than death. Regression models have been mentioned but the data analysis should describe how the maternal and neonatal outcomes will be compared between intervention and control groups 17-30 18 Care should be taken for using the tense in a sentence while the enumerators were already trained and completed the baseline survey. There are some controversial description about the interview time and interval with that of the data collection section. Moreover providing cash of even small amount would bias the interview procedure which is unethical.

20
Cost and payment section is contradicting with the above section of consent procedure, where 1-2 USD is mentioned from past, present and in some instances to future tense.This is confusing to the reader. There is need for clarity and consistency.
3) Dates for the study: Not clear; I am not sure the authors included them 4)Abstract: a) Analysis....it is not clear how the data will be analysed. b)Conclusion: Contrary to guidelines on reporting study protocols, the authors included the conclusion sections in both the abstract and main document. This should be removed. Reading through the conclusion in both sections, I noticed that the content ("To the best of our knowledge"......"This study will generate....") is actually a justification of the study. Let the authors remove the conclusion and take this content to the relevant section/under study justification. What assumptions did they make? 9) Typo and grammatical errors: There are a number of typo and grammatical errors in the document such as "comprised of" instead of "consisted of" or "comprised"(page 8 line 8); "antenatal instead of antenatal care";fathest rather than farthest (page 6 line 31). 10) Page 14: Line 42: "Quality and completeness" should probably read as "accuracy and completeness" as these two are both part of quality! 11) Limitations: Page 20 line 6-7: "...half of study clusters could not be randomly assigned to either the intervention or control group due to political constraints". It is not clear what these political constraints are/were. Let the authors clarify this.
has not been done before.
I have few comments that may help to improve the design and the study: 1. How are selected the 20 clusters that are randomly assigned to treatment and control group (10 to each)? Which is the total sample (how many clusters) from where you choose these 20 and how do you choose them? 2. Regarding the other 20 clusters that are assigned to treatment and control group without randomization: how were they selected in the first place? Was randomization used to select them? The paper says that these 20 clusters were assigned to control and treatment groups without randomization due to political constraints: it would be relevant to know what are the criteria followed to select those clusters that go to the treatment group in order to understand better what is the nature of the bias incurred. Is it based on poverty levels? Is based on number of inhabitants? It is important to make this transparent.
3. I understand that the sample is conformed by women who have delivered a baby in the last 12 months. However, it is not clear to me if these women have delivered in a health care facility, after using Maternity waiting homes or not. If the study is measuring the probability of using maternity waiting homes (and probability of facility delivery), it is difficult to know what is the intention to use them among women that have just delivered a baby if they have not used the Maternity Waiting homes. Women who have used maternity waiting homes and had an institutional birth would be an interesting sample to explore since they may decide, based on their experience, if they want to repeat or not. All these questions should be clarified. 4. I wonder if there is going to be an advertising strategy about the new Maternity waiting homes, so in case the women interviewed have not used them, at least, have heard of them and can say whether is their intention to use them or not. This would be useful not only for the research, but also in operational terms to increase the use of the Maternity Waiting Homes. 5. Finally, impact on health outcomes is going to be measured. In page 17 you talk about primary and secondary outcomes. a. I wonder why you include as a secondary outcome delivery by csection. Explaning the choice of secondary outcomes would be convenient. b. Maternal death and neonatal death can be included as outcomes ( but not maternal mortality rate nor neonatal mortality rate because of the sample size and the short period of analysis: 18 months

GENERAL COMMENTS
This is well written study protocol on a topic of interest to those involved in maternal health in low-and middle income countries. The authors are correct that rigorous evidence on maternity waiting homes is needed and it is of great value that such a study is being implemented. I recommend to accept the protocol with minor revisions. I have added some comments to the attached PDF document. Some minor comments: 1) The authors speak of possible confounders, but do not provide much detail. They could consider reporting using tROBINS-I tool (Risk Of Bias In Non-randomized Studies -of Interventions). This will also allow them to provide arguments on why they call it a rigorous controlled before and after study.
2) The MWH model does not seem to include promotion of the intervention in the community, but their secondary evaluation questions include whether awareness and perceptions have changed over time in the MWH model sites. If the model does not include promotion/communication to the target group, how are women supposed to know about them?
3) It is not clear to me whether the MWH sites all had the model implemented at the same time. Otherwise, this will have an affect on the outcome measures.
4) The reason for having two sets of eligibility criteria for the study sites is unclear for me.

5)
In the introduction, not all evidence on MWH effectiveness has been included.

Response to reviewers: Manuscript ID bmjopen-2018-022224
Title: "Impact of maternity waiting homes on facility delivery among remote households in Zambia: protocol for a quasi-experimental, mixed-methods study"

Corresponding: Nancy Scott
Please find in the table below, a point by point response to the thoughtful comments from reviewers. applicable to the process evaluation protocol.

1
The authors speak of possible Thank you for this suggestion as we were not aware of this confounders, but do not provide tool. We will use the suggested tool to assess risk of bias as much detail. They could we report. We have adjusted the manuscript accordingly consider reporting using in the analysis section (p. 20).

tROBINS-I tool (Risk Of Bias In
Non-randomized Studies -of Interventions). This will also allow them to provide arguments on why they call it a rigorous controlled before and after study. 2.
The MWH model does not seem