--------------------------------------------------------------------------------------------------------------------------------- name: log: D:\Consults\Hindum_clean\confound2.log log type: text opened on: 29 Mar 2020, 15:51:41 . *univariable assesssment of health facility treatment* . . xtlogit antibio_any i.hc_med , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(1) = 14.33 Log likelihood = -553.75052 Prob > chi2 = 0.0002 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.799766 .2794165 3.79 0.000 1.327595 2.439869 | _cons | 1.220891 .1852932 1.32 0.188 .9067576 1.643852 -------------+---------------------------------------------------------------- /lnsig2u | -.7736613 .3742065 -1.507093 -.0402299 -------------+---------------------------------------------------------------- sigma_u | .6792061 .1270817 .4706944 .980086 rho | .1229799 .0403604 .063095 .2259929 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 32.59 Prob >= chibar2 = 0.000 . . *adjusting for other variables to assess for potential confounders for effect of health facility treatment on antibiotic use* . . xtlogit antibio_any i.hc_med i.residence , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 20.92 Log likelihood = -550.46252 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.840686 .2839097 3.96 0.000 1.360471 2.490406 | residence | Rural | 1 (base) Peri-urban | 2.500387 .851638 2.69 0.007 1.282584 4.874483 | _cons | 1.043327 .1574993 0.28 0.779 .7761115 1.402545 -------------+---------------------------------------------------------------- /lnsig2u | -1.119778 .4338999 -1.970206 -.2693501 -------------+---------------------------------------------------------------- sigma_u | .5712724 .1239375 .3734007 .8739999 rho | .0902468 .0356242 .0406579 .1884371 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 18.08 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.occup , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(4) = 16.62 Log likelihood = -552.55155 Prob > chi2 = 0.0023 ----------------------------------------------------------------------------------- antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] ------------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.81789 .2830862 3.84 0.000 1.339728 2.466714 | occup | Peasant | 1 (base) Skilled Labour | 1.386893 .4341545 1.04 0.296 .7509 2.561557 Business Owner | 1.219677 .252119 0.96 0.337 .8133795 1.828927 Unskilled Labour | .8802071 .2488335 -0.45 0.652 .5057678 1.531858 | _cons | 1.154027 .1937739 0.85 0.394 .8304044 1.60377 ------------------+---------------------------------------------------------------- /lnsig2u | -.8088301 .3823659 -1.558253 -.0594067 ------------------+---------------------------------------------------------------- sigma_u | .6673671 .1275892 .4588065 .9707334 rho | .1192368 .0401558 .0601374 .2226561 ----------------------------------------------------------------------------------- Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 29.84 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.education , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(5) = 16.55 Log likelihood = -552.5306 Prob > chi2 = 0.0054 ------------------------------------------------------------------------------- antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] --------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.805394 .2816138 3.79 0.000 1.329835 2.451018 | education | No education | 1 (base) Primary | .90305 .1672289 -0.55 0.582 .6281802 1.298193 O Level | 1.164009 .3513944 0.50 0.615 .644159 2.103388 A Level | .4595085 .2717815 -1.31 0.189 .1441587 1.464692 Tertiary | 1.02936 .3433394 0.09 0.931 .5353676 1.979168 | _cons | 1.258629 .2166269 1.34 0.181 .8982443 1.763603 --------------+---------------------------------------------------------------- /lnsig2u | -.6993804 .3729605 -1.43037 .0316088 --------------+---------------------------------------------------------------- sigma_u | .7049064 .1314511 .4891017 1.01593 rho | .1312185 .0425176 .0677853 .2388056 ------------------------------------------------------------------------------- Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 33.82 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.self_med , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 15.33 Log likelihood = -553.19613 Prob > chi2 = 0.0005 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 2.054506 .4132079 3.58 0.000 1.385197 3.047217 | self_med | Yes | 1 (base) No | .8044814 .166708 -1.05 0.294 .5359528 1.207551 | _cons | 1.23177 .1866676 1.38 0.169 .91524 1.65777 -------------+---------------------------------------------------------------- /lnsig2u | -.7859569 .376246 -1.523385 -.0485284 -------------+---------------------------------------------------------------- sigma_u | .6750433 .1269912 .4668755 .9760278 rho | .1216599 .0402052 .0621387 .2245446 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 31.75 Prob >= chibar2 = 0.000 . . *we notice that self medication was a strong negative confounder of the association between health facility med and use of anti > biotics* . . xtlogit antibio_any i.hc_med i.child_age1 , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(3) = 22.33 Log likelihood = -549.43846 Prob > chi2 = 0.0001 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.827117 .2854348 3.86 0.000 1.345211 2.48166 | child_age1 | 1-12 | 1 (base) 13-36 | .9520087 .1726635 -0.27 0.786 .667207 1.35838 37-59 | 1.628389 .3493613 2.27 0.023 1.069394 2.479582 | _cons | 1.106006 .2213341 0.50 0.615 .7471625 1.637194 -------------+---------------------------------------------------------------- /lnsig2u | -.8046647 .3806383 -1.550702 -.0586273 -------------+---------------------------------------------------------------- sigma_u | .6687584 .1272776 .4605421 .9711118 rho | .1196749 .0401013 .0605657 .2227911 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 30.13 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.diarrhea_any , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 20.11 Log likelihood = -550.69565 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.907262 .3006719 4.10 0.000 1.400303 2.597758 | diarrhea_any | 0 | 1 (base) Yes | .6648445 .1096129 -2.48 0.013 .4812631 .9184544 | _cons | 1.338226 .2054979 1.90 0.058 .9904195 1.808173 -------------+---------------------------------------------------------------- /lnsig2u | -.8535129 .3857308 -1.609531 -.0974945 -------------+---------------------------------------------------------------- sigma_u | .6526225 .1258683 .4471927 .9524218 rho | .1146235 .0391459 .0573037 .2161336 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 28.74 Prob >= chibar2 = 0.000 . . *diarrhea also confounds the association* . . xtlogit antibio_any i.hc_med i.cough , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 82.58 Log likelihood = -509.2453 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.744988 .2853543 3.40 0.001 1.266477 2.404295 | cough | 0 | 1 (base) Yes | 7.969414 1.934936 8.55 0.000 4.95175 12.82608 | _cons | .2117448 .0559891 -5.87 0.000 .1261072 .3555376 -------------+---------------------------------------------------------------- /lnsig2u | -.7977566 .3823296 -1.547109 -.0484043 -------------+---------------------------------------------------------------- sigma_u | .6710724 .1282854 .4613702 .9760884 rho | .1204046 .0404915 .0607704 .2245663 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 28.73 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.fever , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 14.65 Log likelihood = -553.57309 Prob > chi2 = 0.0007 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.77405 .2787709 3.65 0.000 1.303796 2.413916 | fever | 0 | 1 (base) Yes | 1.103704 .1827784 0.60 0.551 .7977923 1.526917 | _cons | 1.149932 .209866 0.77 0.444 .8041284 1.644443 -------------+---------------------------------------------------------------- /lnsig2u | -.7561209 .3730994 -1.487382 -.0248595 -------------+---------------------------------------------------------------- sigma_u | .6851891 .1278218 .475356 .9876472 rho | .1248843 .0407754 .0642702 .2286928 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 32.93 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.fast_breath, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 22.30 Log likelihood = -548.99597 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.762216 .2744313 3.64 0.000 1.298676 2.391208 | fast_breath | 0 | 1 (base) Yes | 2.484634 .7727434 2.93 0.003 1.350614 4.570816 | _cons | 1.148162 .1730674 0.92 0.359 .8544724 1.542796 -------------+---------------------------------------------------------------- /lnsig2u | -.8412352 .384018 -1.593897 -.0885736 -------------+---------------------------------------------------------------- sigma_u | .6566412 .126081 .4507023 .9566795 rho | .1158754 .039342 .0581542 .2176488 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 28.93 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.hc_med i.runny_nose, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 22.17 Log likelihood = -549.3846 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- hc_med | 0 | 1 (base) Yes | 1.781438 .2777126 3.70 0.000 1.312426 2.418057 | runny_nose | 0 | 1 (base) Yes | .5289582 .1167061 -2.89 0.004 .3432555 .8151268 | _cons | 2.104767 .5114384 3.06 0.002 1.307285 3.388738 -------------+---------------------------------------------------------------- /lnsig2u | -.8120011 .3771248 -1.551152 -.0728501 -------------+---------------------------------------------------------------- sigma_u | .6663098 .125641 .4604385 .9642304 rho | .1189042 .0395098 .0605401 .220338 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 31.36 Prob >= chibar2 = 0.000 . . *In summary, from above results, area of residence, self medication, having diarrea and occupation(though weak) are negative co > nfounders of the association be tween health facility medication and antibiotic use, while having a cough is a positive confou > nder* . . *below, we assess for potential confounders of the association between cough and getting any antibiotic drugs. First with more > distal factors* . . xtlogit antibio_any i.cough, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(1) = 75.14 Log likelihood = -515.15135 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- cough | 0 | 1 (base) Yes | 7.97843 1.9114 8.67 0.000 4.988779 12.7597 | _cons | .2814766 .06863 -5.20 0.000 .1745425 .4539241 -------------+---------------------------------------------------------------- /lnsig2u | -.9213466 .3935451 -1.692681 -.1500123 -------------+---------------------------------------------------------------- sigma_u | .6308588 .1241357 .428982 .9277378 rho | .1079173 .037887 .0529738 .2073687 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 25.16 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.cough i.fast_breath, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 79.91 Log likelihood = -511.79039 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- cough | 0 | 1 (base) Yes | 7.668753 1.836904 8.50 0.000 4.795519 12.26348 | fast_breath | 0 | 1 (base) Yes | 2.217143 .7113752 2.48 0.013 1.182181 4.158183 | _cons | .2729569 .0662768 -5.35 0.000 .1695952 .4393136 -------------+---------------------------------------------------------------- /lnsig2u | -.9969081 .4074497 -1.795495 -.1983213 -------------+---------------------------------------------------------------- sigma_u | .6074691 .1237565 .4074865 .9055972 rho | .1008554 .036949 .0480467 .1995405 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 21.80 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.cough i.runny_nose, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 77.18 Log likelihood = -513.99621 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- cough | 0 | 1 (base) Yes | 7.556413 1.826573 8.37 0.000 4.704995 12.13591 | runny_nose | 0 | 1 (base) Yes | .7138126 .1601758 -1.50 0.133 .4598106 1.108127 | _cons | .3915749 .1281851 -2.86 0.004 .2061424 .7438106 -------------+---------------------------------------------------------------- /lnsig2u | -.9521508 .3971083 -1.730469 -.1738328 -------------+---------------------------------------------------------------- sigma_u | .6212166 .1233451 .4209529 .9167537 rho | .1049873 .0373143 .0511098 .2034807 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 24.26 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.cough i.diarrhea, re i(vill) or nolog base diarrhea ambiguous abbreviation r(111); . . xtlogit antibio_any i.cough i.diarrhea_any, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 78.12 Log likelihood = -513.50901 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- cough | 0 | 1 (base) Yes | 7.928335 1.895622 8.66 0.000 4.962086 12.66776 | diarrhea_any | 0 | 1 (base) Yes | .7352929 .1243267 -1.82 0.069 .5278801 1.024202 | _cons | .3103015 .0766816 -4.74 0.000 .1911768 .5036544 -------------+---------------------------------------------------------------- /lnsig2u | -1.005841 .4088134 -1.8071 -.2045811 -------------+---------------------------------------------------------------- sigma_u | .604762 .1236174 .4051289 .9027672 rho | .1000483 .036809 .0475187 .1985425 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 21.66 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.cough i.child_age1 , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(3) = 78.22 Log likelihood = -513.29781 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- cough | 0 | 1 (base) Yes | 7.6618 1.839208 8.48 0.000 4.786308 12.26481 | child_age1 | 1-12 | 1 (base) 13-36 | .9167071 .17328 -0.46 0.645 .6328967 1.327787 37-59 | 1.337571 .2986756 1.30 0.193 .8634674 2.071989 | _cons | .2843628 .0780669 -4.58 0.000 .1660314 .4870296 -------------+---------------------------------------------------------------- /lnsig2u | -.9390757 .3987926 -1.720695 -.1574565 -------------+---------------------------------------------------------------- sigma_u | .6252912 .1246808 .4230151 .9242911 rho | .1062223 .037861 .0515859 .2061478 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 23.81 Prob >= chibar2 = 0.000 . . *we see fast breath, runny nose and child age moderately counfound the association between cough and use of antibiotics* . . xtlogit antibio_any i.cough i.residence , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 79.72 Log likelihood = -512.43174 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- cough | 0 | 1 (base) Yes | 7.953249 1.901655 8.67 0.000 4.977586 12.7078 | residence | Rural | 1 (base) Peri-urban | 2.266959 .76579 2.42 0.015 1.16924 4.395252 | _cons | .2479001 .0605278 -5.71 0.000 .1536192 .4000442 -------------+---------------------------------------------------------------- /lnsig2u | -1.226445 .4529099 -2.114132 -.3387578 -------------+---------------------------------------------------------------- sigma_u | .5416028 .1226486 .3474738 .844189 rho | .0818635 .0340416 .0354008 .1780514 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 14.97 Prob >= chibar2 = 0.000 . . *diarrhea and residence are not confounders* . . *Below, we assess for potential confounders of the asssociation between area of residence and use of antibiotic use* . . xtlogit antibio_any i.residence, re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(1) = 5.69 Log likelihood = -558.44647 Prob > chi2 = 0.0171 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.202453 .7291418 2.38 0.017 1.151088 4.214097 | _cons | 1.452789 .1805346 3.01 0.003 1.138743 1.853444 -------------+---------------------------------------------------------------- /lnsig2u | -1.189527 .43727 -2.046561 -.3324939 -------------+---------------------------------------------------------------- sigma_u | .5516929 .1206194 .359414 .8468371 rho | .0846815 .033893 .037782 .17897 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 17.43 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.residence i.hc_med , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 20.92 Log likelihood = -550.46252 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.500387 .851638 2.69 0.007 1.282584 4.874483 | hc_med | 0 | 1 (base) Yes | 1.840686 .2839097 3.96 0.000 1.360471 2.490406 | _cons | 1.043327 .1574993 0.28 0.779 .7761115 1.402545 -------------+---------------------------------------------------------------- /lnsig2u | -1.119778 .4338999 -1.970206 -.2693501 -------------+---------------------------------------------------------------- sigma_u | .5712724 .1239375 .3734007 .8739999 rho | .0902468 .0356242 .0406579 .1884371 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 18.08 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.residence i.education , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(5) = 8.46 Log likelihood = -556.88077 Prob > chi2 = 0.1328 ------------------------------------------------------------------------------- antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] --------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.427369 .849023 2.54 0.011 1.222954 4.817939 | education | No education | 1 (base) Primary | .8567889 .1551497 -0.85 0.393 .6008083 1.221833 O Level | 1.089273 .3245523 0.29 0.774 .6074613 1.953237 A Level | .4044435 .2393671 -1.53 0.126 .1267886 1.290137 Tertiary | .852205 .2857631 -0.48 0.633 .4416899 1.644261 | _cons | 1.528387 .2161064 3.00 0.003 1.158451 2.016457 --------------+---------------------------------------------------------------- /lnsig2u | -1.124236 .432088 -1.971113 -.2773587 --------------+---------------------------------------------------------------- sigma_u | .5700006 .1231452 .3732315 .8705071 rho | .0898815 .035346 .0406226 .1872154 ------------------------------------------------------------------------------- Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 18.17 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.residence i.occup , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(4) = 7.55 Log likelihood = -557.47061 Prob > chi2 = 0.1094 ----------------------------------------------------------------------------------- antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] ------------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.281965 .7963409 2.36 0.018 1.151499 4.522245 | occup | Peasant | 1 (base) Skilled Labour | 1.149981 .3602479 0.45 0.656 .6223575 2.124916 Business Owner | 1.037073 .2130799 0.18 0.859 .6932967 1.551314 Unskilled Labour | .7223204 .2050212 -1.15 0.252 .4141212 1.259889 | _cons | 1.467728 .197144 2.86 0.004 1.12801 1.909758 ------------------+---------------------------------------------------------------- /lnsig2u | -1.189992 .4380966 -2.048645 -.331338 ------------------+---------------------------------------------------------------- sigma_u | .5515649 .1208194 .3590396 .8473267 rho | .0846455 .033944 .0377063 .1791399 ----------------------------------------------------------------------------------- Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 17.27 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.residence i.diarrhea_any , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 9.78 Log likelihood = -556.54794 Prob > chi2 = 0.0075 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.201916 .7092432 2.45 0.014 1.171187 4.139759 | diarrhea_any | 0 | 1 (base) Yes | .7309931 .1171476 -1.96 0.051 .5339509 1.000749 | _cons | 1.593164 .2071816 3.58 0.000 1.234715 2.055674 -------------+---------------------------------------------------------------- /lnsig2u | -1.291957 .45851 -2.19062 -.3932936 -------------+---------------------------------------------------------------- sigma_u | .5241495 .1201639 .3344359 .8214807 rho | .0770725 .0326149 .0328797 .1702098 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 14.73 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.residence i.cough , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(2) = 79.72 Log likelihood = -512.43174 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.266959 .76579 2.42 0.015 1.16924 4.395252 | cough | 0 | 1 (base) Yes | 7.953249 1.901655 8.67 0.000 4.977586 12.7078 | _cons | .2479001 .0605278 -5.71 0.000 .1536192 .4000442 -------------+---------------------------------------------------------------- /lnsig2u | -1.226445 .4529099 -2.114132 -.3387578 -------------+---------------------------------------------------------------- sigma_u | .5416028 .1226486 .3474738 .844189 rho | .0818635 .0340416 .0354008 .1780514 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 14.97 Prob >= chibar2 = 0.000 . . xtlogit antibio_any i.residence i.child_age1 , re i(vill) or nolog base Random-effects logistic regression Number of obs = 856 Group variable: vill Number of groups = 44 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 19.5 max = 34 Integration method: mvaghermite Integration pts. = 12 Wald chi2(3) = 13.24 Log likelihood = -554.55026 Prob > chi2 = 0.0042 ------------------------------------------------------------------------------ antibio_any | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- residence | Rural | 1 (base) Peri-urban | 2.152824 .7100151 2.32 0.020 1.127916 4.109039 | child_age1 | 1-12 | 1 (base) 13-36 | .9272264 .1651198 -0.42 0.671 .6540402 1.31452 37-59 | 1.546032 .326802 2.06 0.039 1.021623 2.339626 | _cons | 1.366736 .2392463 1.78 0.074 .9698002 1.926136 -------------+---------------------------------------------------------------- /lnsig2u | -1.211263 .4454888 -2.084405 -.3381206 -------------+---------------------------------------------------------------- sigma_u | .5457298 .1215583 .3526771 .844458 rho | .0830119 .033911 .03643 .1781447 ------------------------------------------------------------------------------ Note: Estimates are transformed only in the first equation. Note: _cons estimates baseline odds (conditional on zero random effects). LR test of rho=0: chibar2(01) = 16.22 Prob >= chibar2 = 0.000 . . *we notice that heath facility medication confounds the association between area of residence and * . . *Although caregiver education confounds the association as above, education was not found to improve the final model. There for > e was dropped.