Please help interpret
died alive
TB/HIV+ 16 45
TB/HIV- 17 152
Odds ratio 0.3146
95 % CI: 0.1472 to 0.6722
z statistic 2.985
Significance level P = 0.0028
Please If I`m correct do you mean you treated TB with HIV + and TBHIV -ve
Then what is the efficacy in each?
Regards
I think there was sort of confounding and other variable affects your statistical results, thus difficulties in interpretation. Therefore I suggest multivariate analyses.
The raw interpretation of your findings is that there is no significant relationship between infection with TB & HIV. In other words, if an individual is infected with HIV, it does not significantly increase the risk of same individual being infected with TB bacterium. That explains why your Odds Ratio (OR) is far less than 1.0 (0.3146). This result is, however, practically incorrect. As Dr Amar Miohamed Ismail suggested, I suspect that some confounding factors actually affected your findings.
I calculated Crude OR from your data by manual, it is 3.17 . It is make sense to have OR around this, please observe these number below.
TB/HIV+ died =16/61 (26.23%)
TB/HIV- Died= 17/169 (10.06%)
We can interpret that TB/HIV+ more likely to died 3.17 times comparing to TB/HIV-.
Please revise your analysis again. It should not be the protective factor for death in TB/HIV+.
By the way as comments above, you should control confounding factors by using multivariate analysis.
Died. Alive
TB/HIV+ 16. 45
Tb/hiv- 17. 152
Please I want to know the odd ration for death if They client is coming infected with HIV
Yes TB/HIV more likely to died according to his data but the question is about treatment out come. Therefore I said there was confounding factor.
You can use the link I provided to get answer for your question.
http://vassarstats.net/odds2x2.html
The first coluMN are death second column is alive after treatment for tb
The first coluMN are death second column is alive after treatment for tb
The first coluMN are death second column is alive after treatment for tb
I totally agree with Amornrat's comments. I think the problem George had with interpreting his result was how he constructed his table. Based on the data he presented, bivariate analysis with chi-squared will give the same result as what Amornrat gave. What George did was to place HIV -ve TB patients on top on the row[left side] to get his result. However, the interpretation will be just the reverse to what Amornrat said: That HIV -ve TB patients are x3 LESS likely to die than HIV +ve TB patients.
I want to know the how likely would death be the outcome for HIV co infection in TB clients .Please MRS Amornrat Anuwatnonthakate and MR George NYADORO can i have your email so that i send a copy of the data for some analysis and some hypothesis i hope to decide from the analysis .I shall be grateful.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Infectious Diseases