“The system in place is almost stabilized and then they change everything again. It’s discouraging. You’ve already encouraged [the providers] and then tomorrow you do something else and then the day after something else.”
- Edina, Monitoring and Evaluation officer for an NGO working on maternal and child health
I was recently with Edina* and a team of others from the same organization in the Rukwa district of Kalambo while they were working on various parts of a maternal and child health improvement project. Edina was talking with one of the project’s program officers after we’d spent several days visiting small dispensaries in remote areas of the district. These dispensaries are part of the project and therefore are the recipients of what they call “supportive supervision” visits. I asked to tag along to see what supportive supervision is all about and if it might actually have a chance of helping improve services, given that those two words make it onto many action plans created during reviews of maternal deaths. We had two doctors, who are also basic emergency obstetric and neonatal care (BEmONC) teachers, a regional nursing officer, one program officer who is also a physician, and two district health administrators, including one district reproductive and child health coordinator, and me, the local anthropologist. We broke into two teams of three to four people and spent two days at a facility. The idea is to evaluate the facility in terms of equipment, infrastructure, support systems (pharmacy, lab, ambulance if available, etc.), and personnel knowledge. We had an evaluation “tool” or book, which had a number of standards, each with a number of sub-standards, broken down by topics such as normal labor and delivery, management of complications, post-natal care, and infrastructure. The facility then gets a score based on the number of “yes” answers they get when they meet the sub-standards. Missing just one sub-standard earns you a zero for the entire standard.
These dispensaries are the first line of care for hundreds of thousands of people throughout Tanzania. Each village dispensary that we visited serves between about 2,000 and 6,000 people, depending on if it serves more than one individual village. Many of these facilities are staffed almost exclusively by medical attendants, who are not generally considered skilled personnel. To give you an idea, the medical attendant on the maternity ward at the hospital largely assists in picking up supplies from the main pharmacy, preparing equipment for sterilization, and doing things like folding gauze to be used in delivery packs. She does not administer drugs, help deliver babies, or generally have much direct contact with the care of the women who come to give birth. In the village dispensaries, these same people are responsible for evaluating patient conditions, prescribing medications, ordering supplies, keeping a clean and organized facility, making decisions about referring cases to higher levels of care, and helping women give birth.
We did not visit one dispensary that scored higher than 10% on their evaluation and most scores were around 5%. The challenges in these locations are many and some staff members at the regional hospital like to blame maternal deaths on these providers. While in some cases this may be true, the situation is far more complex. The death of a woman who lives in town and dies in the hospital after giving birth surely cannot be blamed on these front-line care providers in the most remote settings. However, many women in the villages these dispensaries serve may see the poor quality of care and choose to give birth at home with the assistance of a local midwife or relatives and neighbors.
I spent two days at one facility that only has one staff member working although there are supposed to be three. One of the others has been away from her post for nearly a year claiming she has numbness in her hands that causes her to not be able to work. She has not yet been replaced. The other absent provider seems to spend his time in town waiting for the next opportunity to attend a training or seminar at which he will be paid a handsome per diem and have nearly all his meals provided for free. The only working staff person is a medical attendant. While she clearly tries extremely hard, running this type of facility would be a nearly impossible task for one person. The morning we arrived she hadn’t yet shown up for work and it was after 9am. Government health providers are supposed to start work at 7:30 am everyday. The one evaluator immediately noted the woman’s absent and the line of waiting clients. Soon the provider, Salome*, arrived explaining that a woman had given birth at 10pm the night before and she had spent most of the night attending her. That was why she was late arriving at the dispensary to start work for the day. It seemed reasonable to me. The evaluator was not impressed and urged Salome to be sure to get to work by 7:30am. We started with the evaluation process. The organization is particularly interested to see how much knowledge providers have retained after attending BEmONC training (Basic emergency obstetric and neonatal care), which is very comprehensive and generally is conducted over the course of 2 intensive weeks, 6 days per week. Initially, the organization had not included medical attendants in these trainings but then decided to open the trainings up to this lowest cadre in recognition of the fact that these are often the only type of provider at dispensaries.
When we were done evaluating Salome, we sat down with her to create an “action plan.” This action plan is meant to be created by her for her so she can set goals for herself to improve the care she provides. Instead, the evaluators were primarily responsible for generating the action points and the entire plan was written in English, a language Salome does not know. We spent more time debating the dates to put down on the “time line” for bureaucratic purposes than we spent on actually constructing a useful plan and enabling Salome to carry it out. Each short action point took nearly an hour to accomplish as we debated the English wording, the numbering, whether or not I should include the page numbers the action points referred to, what a viable “intervention” was, and what date we should write. I was writing because, I thought, given English is my first language it would be fastest if I took down the notes. I was sorely mistaken. In actuality, we were constructing a document for use by bureaucrats, meant to adhere to a cookie cutter form not sufficiently flexible to be adapted to the needs of each facility and its personnel. The one program officer refused to use Salome’s own words in the plan and insisted on re-wording her statements. One intervention was listed as reviewing support materials in order to better remember the treatment and signs of severe pre-eclampsia and eclampsia. We had a lengthy debate about the date to be indicated on the “timeline.” I argued we ought to say “immediate and continuous review” while the program officer ended up calling some other person in order to confirm that no, we MUST put down one single date. So we chose an arbitrary date two weeks later. It made no sense. Clearly this time line was not for Salome, as we told her, “you know what we really mean, you have to do this every day but we have to write a date, you understand.” This was a time line for the people in the office.
At the very end, Salome thanked us and said she just had two things she wanted to say. First, she asked that the leaders (meaning the district health administrators) receive her requests for support and her efforts to follow-up requests when she makes them. She said often she is met with people who do not care and act as though they don’t have time for her. Her second request was that the BEmONC training materials be made available in Swahili. All of the books, the checklists, the evaluation tool, all given to her in order to help support her and improve the care she provides, are in English, a language she does not understand. I told one of the doctors on the team I was surprised the materials weren’t available in Swahili. He and the program officer both said, “Swahili is hard! How would you say ‘complications of labor and delivery’ in Swahili? It takes up too much room!” So the shorter, more concise English has been used for these complex and crucial clinical and patient care techniques and guidelines. I said, “Yes but wouldn’t it be better if they were in Swahili, so everyone could understand them?” The doctor said, “So you think now they should spend the money to translate these materials all into Swahili and reprint them?” In my mind, that money would be well spent because it’s a one-time cost. Paying for BEmONC trainings for hundreds of people, per diems for the participants and the instructors, accommodations, food, transportation, stationary, all of these costs are repeated again and again. Then, many of these providers go back to their posts and 6 months or a year later score 3% or 5% on the assessment criteria, indicating they’ve retained virtually none of the information they were taught. Perhaps they were simply staring out the window or counting their per diem cash, letting the English float past their eyes and in one ear and out the other. There are deep flaws in the ways these programs are constructed. Despite excellent intentions, solid clinical grounding, and relatively good follow-up, these training programs don’t appear to be doing much good, at least in the facilities we visited over the last two weeks. It’s almost impossible to know if these programs have reduced maternal and neonatal mortality, as they were intended, because the data on these deaths, complications, and outcomes, are also deeply flawed and likely corrupted to the point of uselessness, but that’s a story for another time. The techniques for conducting best-practices and even accounting for and documenting deaths are constantly being changed by the national and local governments, as well as non-governmental organizations. Providers get used to one system and the next day a new person shows up to tell them about another way to do everything. These shifting targets are made more difficult to deal with due to poor organization, leadership, and communication at the level of district health administration and between the districts and higher levels.
*Names have been changed to protect the privacy of the research participants.