During my first week of work, I became acquainted with our online Verbal Autopsy system: a way of noting how many neonatal, under 5 years, and maternal deaths have occurred; it includes the patients’ case histories and why they died.
“See this?” my co-worker pointed to one of the first names in the spreadsheet, “Baby.”
“You will often see Baby and the family name,” he continued. “It is because it frequently happens that parents may not have had the time to name their child before they died.”
I absorbed the weight of this fact and leaned in to read more details of the particular case. The name belonged to a little boy who died two days after he was born.
This year, I am plunging into the world of data during my GHC fellowship. I’m fueled by my belief in the need for more thoughtful, evidence-based interventions: solutions that are founded on an adequate understanding of the problems they try to address. Data can provide extremely powerful and compelling reasons to take action. So I craved more experience in this area of work, with hopes that it would equip me to do more good.
But when I heard this short explanation of why I may see many deaths sharing the first name “Baby”, I was reminded of how much I never want to forget the individuals behind all the data.
My role in UN Millennium Villages Project thankfully gives me the opportunity to make data meaningful by integrating it with individual interactions and translating it into interventions. Whenever a neonatal or maternal death occurs, the sexual and reproductive health facilitator and I conduct a mortality audit. We meet with the deceased’s family members and the health workers who were responsible for the patient and listen to them recount the circumstances surrounding the death. We try to take it beyond the numbers, focusing on not only how many deaths occurred in a month but also why they occurred and how we can prevent similar deaths in the future.
We conduct mortality audits during our quality improvement meetings. Here, health center staff are reviewing a report on the area’s deaths.
My main responsibility is to analyze our 68 Community Health Workers’ (CHWs) performance, looking at health indicators such as the percentage of children visited, the number of Middle-Upper Arm Circumference measurements taken, the number of children discovered to be malnourished, the percentage of children with positive Rapid Diagnostic Tests for malaria who were treated, the percentage of reproductive age women on family planning, and the percentage of pregnant women with the correct number of antenatal care visits. After I pinpoint the worst-performing CHWs, our health facilitators, Community Health Workers Manager, and I meet with them to discuss their performance, challenges they may be facing, and how they can improve.
It doesn’t help to just throw the data at the CHWs: in order for data to have its impact, you have to translate data into an understandable language for others. At these CHW performance review sessions, the poorly performing CHWs take turns one-by-one to sit beside me and listen to me as I go over how they are doing in each area. My first time, after a few minutes of explaining how the CHW was doing in terms of percentages and receiving some slow nods with eyebrow raises, I quickly changed to explaining things in terms of numbers and individuals.
“You are responsible for about 100 children. Now see, you have only visited 20 of them, so there are around 80 more that you did not visit this month.” The rough estimates and conversion from percentages to numbers helped the CHWs better understand their performance. For the next CHW performance review session, I came back with all the percentages converted to specific numbers: “Seebo, you remember you are responsible for 93 children that you need to visit every month. But this month, you have only visited 58 and did not visit 25 children at all. Why is this?”
As CHWs explain to us the reasons and challenges that led to their poor performance, the two-way feedback provides the stories behind the data. Of course, you have a couple of instances in which CHWs excuse themselves by saying they happened to be sick for the whole month and forgot or didn’t know they should have notified the Community Health Workers Manager. But you also have the case in which I was appalled at one Community Health Worker for doing absolutely nothing one month: no child visits, no household visits, nothing. When I met with her, the Community Health Workers Manager, who does her best to keep tabs on all the CHWs, nodded with acknowledgment as the CHW explained that she recently had a Caesarean section to give birth and she had been bedridden and experiencing too much pain to walk around to the households for which she was responsible.
My heart panged at this: based on only the numbers, it had been so easy to dismiss this CHW as truant and uncommitted, deserving of being put on probation or fired. Once I heard her story and the circumstances behind her performance, I could look at the data with more nuance. The CHW promised that she would start moving around to the households closer to her home now that she was feeling a little better, and the Community Health Workers Manager said she would work on having the other CHWs in her parish help the CHW with her caseload until she fully recovers.
Data can be powerful and compelling. But data cannot live up to its potential for impact if it is collected but not used, if it is used but not translated into a manner that can be comprehended, if it is used without seeking a better understanding of the circumstances, stories, and faces behind the data.