Data innovation in response to COVID-19 in Somalia: application of a syndromic case definition and rapid mortality assessment method

Figures & data

Table 1. Syndromic scoring system

Symptom or test	Scores
Loss or change in taste or smell	0.9
Cough	0.7
Fever	0.7
Shortness of breath or difficulty breathing	0.6
Fatigue	0.4
Sore throat	0
Headache	0
Diarrhoea or stomach pains	0
Body aches	0
Other	0
Positive laboratory test for COVID-19	2
Maximum possible score	5.3

A cut-off score of ≥ 2.0 was used to define a suspected case.

Figure 1. The distribution of participant households by district¹.

¹ District locations are shown with a blue dot

Table 2. Incidence of suspected COVID-19 infections in safety net beneficiaries

Measurement period	Round 1^1 Interview respondents only	Round 2	Round 3	Round 4	Round 5
		All household members
Data Collection	22 Jun–15 Jul	10 Aug–3 Sep	6 Oct–22 Oct	30 Nov–20 Dec	27 Mar–18 Apr
Households included in sample	1,117	1,046	1,115	1,565	1,550
Households interviewed	952	942	947	1,430	1,441
Follow up (%)	85%	90%	85%	91%	93%
Household members included in symptom assessment^1	952	7,381	6,916	7,418	11,541
Sex (% female)	61.0%	51.4%	51.5%	51.8%	52.2%
HH members with symptomatic COVID-19 in recall period^2	9	64	36	12	81
Period prevalence of symptomatic COVID-19	0.9%	0.9%	0.5%	0.2%	0.7%
Symptomatic COVID-19 infection rate (cases/1,000,000/day)	311	285	103	34	78

¹In round 1 only the household respondent was asked about symptoms; in round 2 onwards all household members were included.

²In rounds 1 and 2 a one month recall period was used. In round 3 and 4 the recall period was the number of days since the previous interview. In round 5, it was the first 3 months of 2021.

Figure 2. Comparison of laboratory confirmed cases published by WHO for Somalia and suspected symptomatic cases detected by the BRCiS monitoring system¹.

¹The horizontal bars indicate the recall periods for each round of data collection. In the first two rounds, a one-month recall period was used, contiguous recall periods were used in rounds 3 and 4, and a 3 month recall period (Jan–Mar) was used in round 5.

Table 3. Characteristics of participants with suspected COVID-19

Measurement period	Round 2		Round 3		Round 4		Round 5
	Cases (N)	% Affected	Cases (N)	% Affected	Cases (N)	% Affected	Cases (N)	% Affected
Total cases	64	0.9	36	0.5	12	0.2	81	0.7
Age (years)
< 5	10	0.7	8	0.6	2	0.1	26	1.2
5–17	22	0.7	16	0.5	3	0.1	29	0.6
18–49	25	1.3	9	0.5	4	0.2	12	0.4
50–69	4	0.9	1	0.2	3	0.7	8	1.1
≥ 70	3	1.3	2	0.9	0	0	6	1.4
Sex
Male	27	0.7	18	0.5	2	0.1	28	0.5
Female	37	1.0	18	0.5	10	0.3	53	0.9

Table 4. Crude, under-5 death, and cause specific death rates

	Round 1	Round 2	Round 3	Round 4	Round 5^1
Persons under observation	7,618	7,395	6,917	7,418	11,541
Average recall period (days)	70	52	50	56	118
Person days of observation	528,664	349,942	348,870	350,982	1,281,809
Total deaths reported	16	12	17	10	23
Deaths in children <5 years	4	4	4	3	6
Deaths due to suspected COVID-19	5	4	5	4	8
Cause specific COVID-19 death rate (deaths/million/day)	9.5	11.4	14.3	11.4	6.2
Crude Death Rate^2 (CDR) deaths/10,000/day	0.30	0.34	0.49	0.28	0.18
Under Five Death Rate^3 (U5DR) deaths/10,000/day	_	0.61	0.61	0.46	0.25

¹In round 5 the families of 2 of the deceased could not be traced and no verbal autopsy was done.