Lessons from an eight-country community health data harmonization collaborative

Figures & data

Table 1. Initial indicators chosen for community health data harmonization and pooling

		Metric	Description	Numerator definition	Numerator source	Denominator definition	Denominator source
S P E E D	1	Integrated community case manage- ment (iCCM) Speed	Percentage of children assessed, with a symptom of malaria, diarrhea, or pneumonia, within 24 hours of symptom onset	Number of children assessed, with a symptom of malaria, diarrhea, or pneumonia, within 24 hours of symptom onset	CHW activity data	Number of children assessed with a symptom of malaria, diarrhea, or pneumonia	CHW activity data
C O V E R A G E	2	Pregnancy Speed	Percentage of pregnancies registered in first trimester	Number of pregnancies registered in first trimester	CHW activity data	Number of new pregnancies registered in the month	CHW activity data
	3	Postnatal Care (PNC) Speed	Percentage of women with home delivery receiving 1st PNC visit within 48 hours of delivery	Number of women with home delivery who received 1st PNC visit within 48 hours of delivery this month	CHW activity data	Number of women giving birth at home this month	CHW activity data
	4	Proactive Coverage	Percentage of households visited at least once per month (where family was home)	Number of households visited once or more per month	CHW activity data	Number of households in CHW catchment area	CHW activity data or population survey
	5	U5 Coverage	Number of assessments of children under 5 years of age	Number of assessments of children under 5	CHW activity data	-	-
	6	Contraceptive Coverage	Contraceptive prevalence rate	Number of women 15–49 using modern family planning	Facility data + CHW activity data	Number of women 15–49 years old	CHW activity data
	7	Deliveries Coverage	Percentage of deliveries at a health facility	Number of women giving birth in a health institution under the care and supervision of trained healthcare providers	Facility data or CHW activity data	Number of women giving birth	Facility data + CHW activity data
Q U A L I T Y	8	Treatment Quality	Percentage of correct pre-referral treatment administered by CHW, when recommended	Number of people who received the correct treatment at their doorstep (before being referred, if referral is recommended)	CHW activity data	Total number of consultations	CHW activity data
	9	Referrals Quality	Percentage of referral follow-ups with health facility visit confirmed	Number of referral follow ups with health facility visit confirmed	CHW activity data	Number of referral follow ups	CHW activity data

Table 2. Geographic scope of the collaborative today

Organization	Country	Number of Districts/Counties	Number of CHWs	Number of indicators shared
Integrate Health	Togo	4	139	7
Muso	Mali	1	225	9
Lwala	Kenya	1	402	8
Partners In Health – Malawi	Malawi	1	1,128	5
Living Goods	Uganda	19	4,511	7
Living Goods	Kenya	5	1,656	8
Wuqu’ Kawoq	Guatemala	10	50	3
Possible	Nepal	2	104	4
VITAL	Pakistan	6	100	4

Figure 1. Initial reporting levels across the nine indicators.

Figure 2. Current status of indicators reporting.

Table

Text Box: One organization’s experience with definitional alignment Lwala, based in Rongo, Kenya, took the following actions as a direct result of the Community Health Impact Data Collaborative: Adjusted how existing data were reported to align with peers, leading to new insight. Lwala measured the amount of weeks in a ‘first trimester’ differently for the Pregnancy Speed indicator. We adjusted our data timeframe to capture the same definition and in the process decided to start systematically tracking the first antenatal care (ANC) visit at 12, 14, and 16 weeks, as we found it better informed our program’s understanding of when a woman seeks her first ANC visit, a key focus area for Lwala programming. Viewed data at a more granular level, which offered fresh program trend and data quality opportunities. Coalition data were reported monthly with numerators and denominators. As Lwala generated this data, it promoted an ongoing review of granular trends and changes in the monthly data, between monitoring and evaluation and program teams. Moreover, understanding the data from a different angle also offered new vantage points for checking data quality; for example, if a denominator changes suddenly without explanation, a data quality assessment (DQA) exercise is initiated. Redesigned parts of Lwala’s CHW digital data collection tool to optimize additional data for learning. Lwala found we were not measuring certain indicators such as CHW speed of assessment. The Coalition exercise prompted Lwala to undertake a learning activity that included technical knowledge – that is, sharing meetings with another Coalition partner during which they demonstrate how they successfully collected this data in their digital system. As a result, Lwala is updating parts of CHW digital data collection tool to optimize measures of treatment speed and quality – and to further support CHWs who use the digital tool.

Figure 3. Proactive coverage data for 6 sites, 2018-2020.

Figure 4. U5 coverage data for 12 sites, January-July 2020.