Understanding why child welfare clinic attendance and growth of children in the nutrition surveillance programme is below target: lessons learnt from a mixed methods study in Ghana

Table 3 Binary logistic regression model showing factors that predicted an increase in weight-for-age z-score between the first and last attendance on the programme

Variables	n	%	Adjusted regression model
Variables			AOR	95% CI	P value
Programme (Facility-based)	116	47.3	Ref
Community-based	104	52.7	1.27	0.70–2.33	0.434
Age at registration (> 4 weeks)	148	67.3	Ref
Neonatal (≤4 weeks)	72	32.7	0.75	0.39–1.44	0.382
Sex (male)	110	50.0	Ref
Female	110	50.0	1.97	1.07–3.62	0.029
Underweight at baseline (No)	188	85.5	Ref
Yes	32	14.5	11.07	3.95–31.03	< 0.0001
Annual attendance (< 6 visits)	117	53.2	Ref
≥ 6 visits	103	46.8	2.15	1.14–4.07	0.018
Attendance (< 9 annual visits)	190	86.4	Ref
≥ 9 annual visits	30	13.6	4.65	1.44–15.06	0.010

Hosmer-Lemeshow goodness-of-fit test, Chi-square = 10.192, p = 0.252. Cox and Snell R² = 0.155, Nagelkerke R² = 0.210

ISSN: 1472-6955