Table 4.
Predicted probabilities of the correct assessment for illnesses of observed sick children seen by HSAs in districts using the iCCM mobile application and paper tools*
| Symptoms |
iCCM mobile application |
Paper tools |
P-value† |
|||
|---|---|---|---|---|---|---|
|
N |
Weighted % |
95% CI |
Weighted % |
95% CI |
||
| Children checked for presence of cough | 987 | 97.9 | 96.6, 99.2 | 90.7 | 85.5, 95.9 | 0.001 |
| Children checked for presence of diarrhoea | 987 | 93.9 | 90.8, 96.9 | 87.4 | 82.1, 92.6 | 0.026 |
| Children checked for presence of fever | 987 | 96.7 | 94.4, 99.0 | 92.6 | 87.6, 97.6 | 0.056 |
| Children with cough assessed for presence of fast breathing through counting of respiratory rates | 716 | 97.1 | 94.3, 99.8 | 95.7 | 92.6, 98.9 | 0.463 |
| Children with cough assessed for the presence of fast breathing in which HSA counted respiratory rate within ± 3 breaths of gold standard (N = 699) | 699 | 84.8 | 81.3, 88.3 | 86.6 | 82.2, 91.0 | 0.488 |
| Children with fever assessed for malaria with rapid diagnostic test | 652 | 83.8 | 73.3, 94.2 | 88.6 | 81.9, 95.3 | 0.507 |
| Children assessed for three general danger signs | 987 | 87.6 | 83.6, 91.6 | 78.6 | 73.3, 84.0 | 0.009 |
| Children checked if able to drink or eat anything | 987 | 94.9 | 92.9, 97.0 | 89.4 | 86.0, 92.9 | <0.001 |
| Children checked if vomit everything | 987 | 94.1 | 90.7, 97.6 | 91.1 | 86.9, 95.4 | 0.270 |
| Children checked if have convulsions | 987 | 92.8 | 90.7, 94.8 | 84.0 | 80.1, 87.9 | <0.001 |
| Children assessed for five physical danger signs | 987 | 79.9 | 75.9, 84.0 | 61.7 | 55.0, 68.4 | <0.001 |
| Children checked for chest indrawing | 987 | 94.6 | 92.8, 96.3 | 78.2 | 73.5, 82.9 | <0.001 |
| Children checked if sleepy or unconscious | 987 | 98.6 | 97.0, 100.1 | 96.5 | 93.6, 99.5 | <0.001 |
| Children checked for palmar pallor | 987 | 99.1 | 98.5, 99.8 | 89.6 | 84.6, 94.6 | <0.001 |
| Children checked for malnutrition with MUAC tape | 987 | 86.3 | 82.9, 89.7 | 82.6 | 77.6, 87.6 | 0.182 |
| Children checked if swelling of both feet | 987 | 96.6 | 95.0, 98.2 | 85.9 | 80.9, 91.0 | <0.001 |
iCCM – integrated community case management, HSA – health surveillance assistant, MUAC – mid-upper arm circumference
*Probabilities adjusted for child characteristics (age and gender), HSA characteristics (gender, highest education level, tenure as an HSA, type and duration since most recent iCCM training, patient case load, and village clinic location), and district characteristics (access to improved water source and median number of years of women’s education) using logistic regression with standard errors clustered at the HSA level.
†Compared against Holm-Bonferroni adjusted significance levels.