Skip to main content

Table 5 Multivariable multinomial logistic regression showing association between choice of first provider, and respondent’s characteristics

From: Biomedical drugs and traditional treatment in care seeking pathways for adults with epilepsy in Masindi district, Western Uganda: a household survey

VariablesHospitalChurch/ MosqueFriend/ Neighbour
Adj. Coeff95% CIp-valueAdj. Coeff95% CIp-valueAdj. Coeff95% CIp-value
Reason shifted from first provider
 Did not get better, hence sought better treatment1  1  1  
 Nearer2.160.74, 3.590.003**      
 Friend or neighbour’s advice1.16−1.42, 3.740.3792.16−0.45, 4.770.1050.25−2.31, 2.800.851
Age in completed years
 18–251  1  1  
 26–352.51−0.89, 5.900.1482.54−0.30, 5.380.0791.45&−2.64, 5.540.488
 36–452.13−1.16, 5.410.2052.39−0.52, 5.310.1081.38&−3.41, 6.170.573
 46–552.82−0.65, 6.300.1112.80−0.43, 6.030.0905.83&0.67, 10.990.027*
 56–652.56−1.05, 6.170.1640.85−2.84, 4.540.6533.76&−1.33, 8.860.147
  > 652.16−1.83, 6.160.2890.84−3.13, 4.820.6786.56&0.98, 12.130.021*
Sex
 Female1  1  1  
 Male0.33−1.98, 2.640.780−2.01−4.29, 0.280.085−1.60−3.77, 0.570.147
Marital status
 Married/ cohabiting1  1  1  
 Separated/ Divorced0.69−1.75, 3.130.578−2.35−5.50, 0.800.144   
 Single/ Never married0.98−1.71, 3.660.475−1.03−4.10, 2.040.5104.93&0.56, 9.300.027*
 Widow/ Widower0.37−2.43, 3.180.793−1.34−4.18, 1.490.353−0.23−3.38, 2.910.884
Occupation
 Farming1  1  1  
 Government worker         
 Trading2.500.39, 4.610.020*1.29−1.05, 3.630.2812.870.71, 5.040.009**
 Unemployed   1.66−2.66, 5.970.451   
Currently receiving anti-epileptic medicines
 No1  1  1  
 Yes−0.52−1.86, 0.830.451−0.22−1.58, 1.130.7460.57−0.91, 2.050.452
Perceived purpose of treatment
 Control fits/seizures1  1  1  
 Cure epilepsy0.71−0.49, 1.920.2471.910.38, 3.430.014*1.29−0.01, 2.580.051
  1. Coefficients are relative probabilities compared to attending the health centre as first choice provider. *Significant at P < 0.05, **p < 0.01, &Coefficient for bivariate analysis of choice of provider 1 as a traditional healer (not friend /neighbour), and occupation as predictor variable. Hosmer-Lemeshow goodness-of-fit Chi-square test is 30.51, df = 40, p = 0.860