Table 3. Predictors of HIV positivity among study participants (N=11534)

CharacteristicNn (%)Simple Logistic RegressionMultiple Logistic Regression
OR (95% CI)Approx. p-valueApprox. global p-valueAdjusted OR (95% CI)Adjusted approx. p-valueAdjusted approx. global p-value
Rural/urban setting0.089<.001
Rural7427373.0 (5.1)refref
Urban4107275.0 (6.6)1.30 (0.96-1.76)0.0891.65 (1.24-2.19)<.001
Total11534648 (5.7)
Region<.001<.001
Eastern229775.0 (3.8)refref
Nairobi131067.0 (4.9)1.32 (0.85-2.04)0.2160.93 (0.58-1.51)0.779
Central141360.0 (3.8)1.02 (0.64-1.62)0.9310.84 (0.52-1.35)0.471
Coast145266.0 (4.4)1.17 (0.72-1.89)0.5241.05 (0.64-1.73)0.833
Nyanza1621242.0 (15.2)4.59 (2.90-7.28)<.0014.69 (2.92-7.51)<.001
RiftValley204979.0 (3.8)1.00 (0.63-1.59)0.9871.03 (0.66-1.62)0.894
Western139259.0 (4.8)1.29 (0.76-2.18)0.3441.36 (0.79-2.34)0.264
Total11534648 (5.7)
Respondent sex<.001<.001
Men4794193.0 (4.4)refref
Women6740455.0 (6.9)1.62 (1.34-1.96)<.0011.44 (1.18-1.76)<.001
Total11534648 (5.7)
Age category (5 year categories)<.001<.001
15-19184117.0 (1.1)refref
20-24189658.0 (3.1)3.01 (1.57-5.78)<.0012.70 (1.39-5.26)0.003
25-291786112.0 (6.3)6.28 (3.39-11.64)<.0015.63 (2.92-10.86)<.001
30-34141897.0 (6.6)6.65 (3.66-12.09)<.0015.70 (2.94-11.05)<.001
35-391220108.0 (8.7)8.99 (5.02-16.10)<.0017.84 (4.03-15.24)<.001
40-4498490.0 (9.3)9.65 (5.13-18.15)<.0018.11 (4.08-16.13)<.001
45-4971566.0 (9.8)10.22 (5.73-18.24)<.0017.45 (3.84-14.44)<.001
50-5473460.0 (8.4)8.62 (4.71-15.80)<.0016.12 (3.14-11.92)<.001
55-5952324.0 (4.4)4.31 (2.09-8.89)<.0012.85 (1.32-6.18)0.008
60-6441716.0 (4.0)3.92 (1.78-8.63)<.0012.47 (1.06-5.77)0.036
Total11534648 (5.7)
Educational level0.037<.001
No primary133244.0 (3.6)refref
Incomplete primary97446.0 (5.0)1.41 (0.88-2.27)0.1532.25 (1.37-3.69)0.001
Complete primary3667223.0 (6.0)1.71 (1.17-2.49)0.0052.52 (1.68-3.79)<.001
Secondary+5561335.0 (5.9)1.67 (1.15-2.42)0.0072.22 (1.50-3.29)<.001
Total11534648 (5.7)
Marital status<.001<.001
Ever widowed339578.0 (2.3)refref
Married/cohabiting6683342.0 (5.3)2.41 (1.83-3.16)<.0011.04 (0.74-1.46)0.807
Never married/never cohabited55863.0 (11.1)5.33 (3.71-7.68)<.0012.59 (1.72-3.92)<.001
Separated/divorced894165.0 (20.0)10.69 (7.66-14.92)<.0014.52 (2.94-6.95)<.001
Unknown40.0 (0.0)0.00 (0.00-0.00)<.0010.00 (0.00-0.00)<.001
Total11534648 (5.7)
Sexual debut0.517
First sex 15+ years7988482.0 (6.1)ref
First sex <15 years134294.0 (6.6)1.09 (0.85-1.40)0.517
Total9330576 (6.1)
Age At Last Brithday11534648 (5.7)1.03 (1.02-1.03)<.001<.001
Number of sex partners (matrix)8128470 (5.8)1.23 (1.00-1.50)0.0450.045

LogisticRegression_Footnotes

Footnote
Simple logistic regression results are displayed alongside multiple logistic regression results.
Outcome event modelled: 1.
Multiple logistic regression includes: qresid, qprov, sex2, agecat5, educ, marital4.
Adjusted columns are intentionally blank for simple-model predictors not selected for the multiple model: debut, age, n_partners.
Weighted models use 'abweight'.
Cluster-robust standard errors use 'qclust'.
Strata metadata supplied via 'strata2'.
Approximation warning: this Python path does not reproduce PROC SURVEYLOGISTIC Taylor-linearized variance exactly; odds-ratio CIs, coefficient p-values, and global p-values are approximate weighted-GLM/robust-covariance results.

LogisticRegression_Metadata

SettingValue
outcomehiv
outcome_event1
simple_categorical_varsqresid, qprov, sex2, agecat5, educ, marital4, debut
simple_continuous_varsage, n_partners
multiple_categorical_varsqresid, qprov, sex2, agecat5, educ, marital4
multiple_continuous_varsNone
reference_categoriesRural/urban setting=Rural, Region=Eastern, Respondent sex=Men, Age category (5 year categories)=15-19, Educational level=No primary, Marital status=Ever widowed, Sexual debut=First sex 15+ years
simple_only_predictorsdebut, age, n_partners
variance_methodTAYLOR
or_decimal2
pvalue_decimal3
weight_variableabweight
cluster_variableqclust
strata_variablestrata2
survey_variance_noteApproximate weighted-GLM/robust covariance; not PROC SURVEYLOGISTIC-equivalent Taylor variance.