Participants, Interventions, and Outcomes
Study Setting: The study setting is in the home of women enrolled in the “Madres Sanas”community-based nursing program offered by the Fundación para la Salud Integral de los Guatemaltecos (FSIG) 6. FSIG supports a community-based clinic called the Center for Human Development in a region of Guatemala informally referred to as the Southwest Trifinio 6. This region is at the intersection of the boundaries of three departments in Guatemala, and as such no single Department takes responsibility for the health of the migrant workers that reside there6. The impoverished area has a population of around 25,000 people and experiences poor pregnancy outcomes6. This is why the University of Colorado Center for Global Health, in partnership with AgroAmerica, built the Center for Human Development clinic and initiated the community-based maternal and perinatal care programs, called Madres Sanas and Ninos Sanos, respectively6.
Madres Sanas is executed by teams of Guatemalan community nurses. The nurse teams are comprised of two nurses who are responsible for a segment of the communities in the region. There are ten communities in the Madres Sanas program that were combined into eight clusters; our biostastician did this in order to achieve similar cluster sizes, which was determined according to the number of births by community in 2017. Because our study enrolls women at their final Madres Sanas visit, which is a postpartum visit that occurs about forty days out from delivery, the cluster size was based on delivery volume of the communities. The nurse teams are assigned by the nursing supervisor to their respective communities. They drive auto rickshaws provided by the Center for Human Development out to the communities to conduct their home visits, which includes four antenatal visits and two postpartum visits. During the visits the nurses both provide clinical care and collect quality improvement and research data, and as such serve a dual function in their role. As noted, this study takes place at the final Madres Sanas visit, which occurs forty days after delivery. Routine clinical care, including postpartum contraceptive education, culminates at this time, although counseling on postpartum contraception begins at the initial enrollment visit in pregnancy. After routine clinical care is provided, the nurses offer enrollment in our study. Figure 1 is a map of the study communities 6.
Eligibility Criteria: All women capable of providing informed consent who participate in the Madres Sanas program and have a forty-day postpartum visit are eligible to enroll in the study if they are between the ages of 15 – 35 and have not already started a contraceptive method. Those unable or unwilling to provide consent were excluded.
Interventions: All study activities and procedures begin at the forty-day postpartum visit, which is the sixth and final scheduled visit of the Madres Sanas program. At this visit the community nurses perform routine maternal and neonatal clinical care, which includes amongst other activities, final counseling and education about postpartum contraception. In intervention clusters, nurse teams bring a kit with them to the visit. After the visit they describe the study and offer consent. Once a woman has discussed the risks, benefits, and alternatives with her providers and determines she wishes to participate in the study, she signs the consent form and the nurses offer her condoms, contraceptive pills, a medroxyprogesterone contraceptive injection, or the levonorgestrel implant. The kit is stocked with 10 condoms (Vive Amor®), one pack of pills (Segura Plus®), one syringe of medroxyprogesterone (Cyclofem®), and one implant (Jadelle®) for each postpartum visit planned for that day. It also contains all the necessary materials to place the implant or administer the injection under sterile conditions, such as alcohol swabs and sterile gloves, etc.
All contraceptives are purchased using study funds and are sourced from a local provider of contraceptive medications and devices. All contraceptives are routinely available and approved for distribution in Guatemala. Because these are routinely available medications and we are not testing their effectiveness as contraceptives, there are no criteria for discontinuing or modifying allocated interventions for a given trial participant (such as changing the drug dosing). Women are, however, screened for contraindications to the contraceptive methods provided using the Medical Eligibility Criteria 7. The nurses have the eligibility criteria chart in Spanish, laminated, and included in their kit for use during study enrollment. As this is a pragmatic trial, there are no restrictions on care and interventions that are permitted or prohibited during the trial. For example, if a participant did not initiate a contraceptive method at the Madres Sanas nurse visit and wishes to, she can seek the method in the community. Conversely, if she chose a method and opted for the implant but does not like it, she is free to remove it at any time. The nurses will either remove it in the home setting or advise the woman to present to the Center for Human Development where they can remove it and potentially initiate a new method. The initial contraceptive provided in the study setting is free, but any contraceptives sought or utilized after the study enrollment visit is the woman’s responsibility to locate and finance. Study contraceptives are only provided in the home setting by the nurses at the enrollment visit and are not available or offered at any subsequent visit.
Outcomes: Our primary outcome is the proportion of women using the contraceptive implant at three months after enrollment in the intervention clusters as compared to the control clusters. We are planning a difference in differences analysis of proportions. Our secondary outcomes are to also compare overall contraceptive uptake, continuation, satisfaction, and pregnancy rates between study arms. All this data is collected by maternal self-report through enrollment, three-month, and twelve-month surveys.
The clinical relevance of our primary outcome is that we have the potential to reduce unintended and short interval pregnancies by increasing utilization of a highly effective method of postpartum contraception (the implant). The clinical relevance of monitoring continuation, satisfaction, and pregnancy rates is to observe the acceptability of the implants in the community; it is possible that the primary outcome is significant, but that we also find women did not like the implant and took it out before three months, etc. This will be important information to contextualize our primary outcome.
Primary Outcomes: The primary outcome of this study is use of the contraceptive implant at three months postpartum.
Outcome Name: Postpartum implant use
Metric/Method of Measurement: Proportion of women in each arm of the trial who self-report using a contraceptive implant when they complete their three-month survey
Timepoint: Three months following enrollment in the study
Key Secondary Outcomes:
Outcome Name: Postpartum contraceptive use
Metric/Method of Measurement: Proportion of women in each arm of the trial who self-report using a contraceptive method when they complete their three-month survey
Timepoint: Three months following enrollment in the study
Outcome Name: Postpartum contraceptive continuation
Metric/Method of Measurement: Proportion of women in each arm of the trial who self-report using a contraceptive method when they complete their twelve-month survey
Timepoint: Three- and twelve-months following enrollment in the study
Outcome Name: Postpartum contraceptive satisfaction
Metric/Method of Measurement: Proportion of women in each arm of the trial who self-report satisfaction (if they are using a contraceptive method) with that method when they complete their three-month and twelve-month survey
Timepoint: Three- and twelve-months following enrollment in the study
Outcome Name: Short-Interval Pregnancy
Metric/Method of Measurement: Proportion of women in each arm of the trial who self-report repeat pregnancy when they complete their enrollment, three-month, and twelve-month surveys
Timepoint: At enrollment and three- and twelve-months following enrollment in the study
Participant Timeline: Women are enrolled as part of their routine postpartum visit in the Madres Sanas program that occurs about forty days after a delivery. Three months after study enrollment, women are either called or visited in their home to complete a survey regarding their current contraceptive use, potential contraceptive continuation, satisfaction with any contraceptive they might be using, and whether or not they have been pregnant since the delivery of their last child. Similar questions are asked twelve months after study enrollment to observe our secondary outcomes, and after the twelve-month survey is conducted the study activities are complete. Figure 2 describes study activities.
Sample Size: Based on previously collected data, we expect that during our study timeframe around 260 women will meet eligibility criteria over the course of one year. This study, with 200 women enrolled (100 from intervention clusters and 100 from control clusters) will be powered to detect a change in Jadelle® uptake rates from 3% to 15% at 85% power and 5% significance, with an intra class correlation of 2%.
Recruitment: As part of our educational efforts within the Madres Sanas program we use “flipcharts.” The flipcharts are a spiral-bound, laminated compilation of images that represent healthcare concepts. As some women in our community are not literate and because many people are visual learners the nurses bring flipcharts that are specific the visit they are executing to help guide them and their patients through the educational sessions that are a component of every care session. These flipcharts have been a great way to both structure the visits and to ensure that important concepts are covered, and they have worked well to impart information. As such, our main strategy for achieving adequate participant enrollment was to create a study and informed consent flipchart to illustrate the main concepts of the study and help the nurses standardize and ensure a proper informed consent process.
The nurses also bring contraceptive educational materials with them to antepartum visits. They introduce the idea of optimizing child health by optimizing pregnancy spacing. The nurses have a model arm with an implant in it that women can palpate, contraceptive pill packs, sample intrauterine devices, and condoms so that over the course of their antenatal care women are able to touch and visualize different contraceptive methods. This allows them to become more comfortable with the concept of postpartum contraception/pregnancy spacing and the various options so that when the study is offered at the end of their Madres Sanas experience, they have some familiarity and health literacy with respect to the options offered as part of the study. Our final strategy for achieving adequate participant enrollment to reach our target sample size is to offer enrollment to all women at their Madres Sanas postpartum visit who meet our inclusion and exclusion criteria.
Assignment of Interventions
Sequence Generation: The first step, as noted previously, was to divide our Madres Sanas communities into study clusters. We were able to obtain historical data about the number of enrollees/births by community to get a sense of the volume of births in each area. We combined some smaller communities into larger study clusters in order to get an expected birth rate of about 100 births per nurse team (nurse teams have varying numbers of clusters), per year for a total of about 300 expected births in one year. Once the clusters were assigned by expected birth volume, which we expect to translate to eventual postpartum visits, the allocation sequence was generated.
Allocation and Concealment Mechanism: The initial allocation sequence was generated by our data analyst using SAS to assign the clusters to either the intervention or the control arm of the trial. Once the nurses were educated about the study and understood all study procedures and activities, they were informed about the cluster assignment. One of the nurse teams was not assigned an intervention group and the nursing supervisor requested that each nurse team have an intervention and a control group. As such, the allocation sequence was rerun to accommodate the “real world” constraints of the study to appease the study staff in order to proceed with study activities; this will be addressed as a limitation of our study when we publish our results. The communities, assigned to clusters, are described below with their nurse team (as indicated by team color) in Table 1.
Implementation: It is the nurse’s responsibility to enroll patients in the study and to only offer home-based contraceptives to women living in intervention clusters. If the nurses incorrectly offer the intervention in control clusters this will bias our study towards the null hypothesis of there being no difference in the uptake of the contraceptive implant.
Blinding: Our study did not involve any blinding procedures; neither the nurses nor the participants were blinded to their assignment, so there was no unblinding necessary.
Data collection, management, and analysis
Data Collection Methods: We have quality improvement data from prior to study initiation that was collected from June 2017 – September 2018 by the Madres Sanas community nurses. This database includes antepartum, intrapartum, and postpartum quality improvement data collected by our community nurses during routine antenatal and postnatal care visits. The data is collected on tablets and transmitted through the application REDCap 8. REDCap (Research Electronic Data Capture) is a secure, HIPAA-compliant web-based application designed for data collection for research studies 8. It provides an easy-to-use data entry system with data validation, the ability to import data from external sources (Guatemala), automated exports to statistical software, audit trails, branching logic and calculations, and sophisticated tool for building and managing online surveys 8. As the community nurses have been using this software for years, our study links to the Madres Sanas dataset, but involves separate forms in a separate REDCap database. Just as the nurses prompt women during routine visits and collect general pregnancy outcomes data for our quality improvement database, so will they use REDCap and the same methods to collect enrollment and follow-up on participants and transmit the data to password protected servers at the University of Colorado. There are cluster-specific REDCap forms that are collected on enrollment, at three months, and at twelve months following enrollment. These surveys are included in Appendix A. The schedule of enrolment, interventions, and assessments are shown in Figure 4, below.
Data Management: Our study supports a data manager who reviews the study data on a biweekly basis for inconsistencies and provides a study dashboard of study progress in terms of enrollment, retention, and follow-up. Our plans to promote participant retention and to complete follow-up include having the nurse teams conduct the three- and twelve-month surveys on their own enrollees from their own communities. In this way, the personal relationship that was built during the antepartum and postpartum care programming is continued after enrollment in an effort to provide continuity of the relationship through study activities.
Statistical Methods: To analyze our primary outcome we are planning a difference in differences analysis of the proportion of women actively using a contraceptive implant three months after enrollment in intervention clusters as compared to control clusters. We plan to provide an unadjusted analysis and an adjusted analysis for any cluster characteristics that are imbalanced between the two study arms. Once our primary outcome is assessed, we will likely use multivariate modeling to determine characteristics of women, in each study arm, who used any postpartum contraception within three months of delivery as compared to those who didn’t, in both adjusted and unadjusted analysis. We plan to use descriptive statistics to compare our other secondary outcomes between study arms (continuation and satisfaction). Finally, we will likely perform survival analysis of time to repeat pregnancy by study arm to observe if there is any statistically significant difference in the time to repeat pregnancy of women who became pregnant by twelve months in the study groups. We plan to use intention-to-treat in our analysis.
Data Monitoring: This study does not have a data monitoring committee because we are not testing the safety or efficacy of a new intervention but rather the association of increasing access to the postpartum contraceptive implant with uptake of the device. We are not planning any interim analyses and do not have any predefined x guidelines. No one will have access to any interim results that might influence a decision to terminate the trial early.
Harms: The Center for Human Development has a close relationship with the community it serves. It meets regularly with community leaders from each community that comprise the Community Advisory Board, which approved the study before its execution. The community leaders permitted the Center for Human Development to develop the community-based nursing programs and help set the priorities for care in the programs. They meet monthly with Center for Human Development leadership to discuss issues related to the provision of care in the community. In addition to the Community Advisory Board as a mechanism for communication between these clinical and research partners, the community nurses themselves serve as genuine link between the community and FSIG/the Center for Human Development. The nurses are in the communities five days a week providing preventative care and managing pregnancy and neonatal/early childhood complications. As such, our main plans for collecting, assessing, reporting, and managing solicited and spontaneous reported adverse events or other unintended effects of our trial intervention depend on the nurses, the Community Advisory Board, and the women themselves. We will not be monitoring specifically for adverse effects of study medications as they are not themselves under study. Therefore, we depend on women to contact their community nurse, and if they do not feel comfortable, their community leadership, regarding any adverse events they are experiencing. The nurses share their work cell phone numbers and patients call them and the nursing supervisor, frequently. Patients also know that the Center for Human Development clinic is open to provide care to them at any time. In this way, we hope that the community safety net established through trust and partnership between the Center for Human Development and the community will help with finding and managing adverse events during the study. There are also standard operating procedures for the study staff to follow if any of these adverse events occur.
Auditing: The Principal Investigator (PI) and the Senior Foreign Investigator (SFI) are responsible for auditing trial conduct in person, while the data manager audits the actual data being entered on a biweekly basis. Our team meets as a group to discuss study activities every other week and data issues and inconsistencies are often addressed at that time. Between the PI, SFI, and other co-investigators on the study, site visits are made every one to three months to observe study activities and to provide audit and feedback on the consent process and data entry. There are no planned procedures for an independent audit of trial conduct.