Originally Posted by krst234
Yes. And it doesn't appear that the journal article states that there is an 18-fold difference in zero APGARS, only that there is a difference.
If we could read tables and make inferences based on numbers alone, calculations like p-values would be completely unnecessary in determining statistical significance.
I believe the found 'statistically different' values - I don't believe the study demonstrated any magnitude of difference in values. I've only read the abstract. Nothing in the abstract leads me to believe they performed this analysis. It looks like something gleaned by looking at tables..... or by reading some blogger's interpretation of the data.
The study did demonstrate magnitude of differences in the values. It's not "my interpretation of the data." It's right there in the table, in the study, in column 2. If you look at column 2, you'll see at the top that it says RR (95% CI). RR stands for Relative Risk, which is the topic of this discussion (the different chances the infant will have an Apgar score of zero at five minutes). To be more technical, it's the "ratio of the probability occurring." So, in this study, it's specifically addressing the chance that a zero Apgar score at five minutes will occur, compared between the different professionals.
CI stands for Confidence Interval. Confidence intervals consist of a range of values that act as strong estimates that a number will fall between. It's easiest to explain this with the actual numbers, so please look at the table again. First of all, it says that Hospital MDs serve as the reference group. Check. So that RR will be 1, and the other RRs will be compared to that. Then look at homebirth midwives. That RR is 10.55. That means that compared to hospital MDs, the chance that homebirth midwives will have an Apgar of 0 is 10.55. The confidence interval is 8.62-12.93 and the confidence level is 95%, which means the researchers are 95% confident that the number is between 8.62 to 12.93. So home birth midwives almost certainly have 8.62X to 12.93X as many zero Apgar scores as doctors.
Again, this is not just "my interpretation of the data." It's the researcher's interpretation of the data. It's already in the study, right there in column 2.
Originally Posted by krst234
And it doesn't appear that the journal article states that there is an 18-fold difference in zero APGARS, only that there is a difference.
The study uses the Hospital MDs as a reference group, and shows that Homebirth Midwives have a 10.55x greater chance of having Apgar scores of zero than them.
As rnra pointed out, a better comparison would probably be between Hospital Midwives and Homebirth Midwives. Using the same basic calculations that the researchers did (which yielded the 10.55 ratio between hospital mds and homebirth midwives, and which I already demonstrated how to do in a previous answer), you get that Homebirth Midwives have a 18x greater chance of having Apgar scores of zero than Hospital Midwives.
One more time - the study does demonstrate the magnitude of differences between the groups. That's what column 2 is. The researchers did it, and they provided the confidence intervals and confidence level. They used Hospital MDs as the reference group. Using Hospital Midwives as the reference group, the ratio is even higher.