I was wondering if there is a way to calculate a p value with only the following information:

Group A: n=64; Oswestry Disability Index mean = 30
group B: n=62; Oswestry Disability Index mean = 20.

I was thinking you could use the Excel difference between 2 means program; however, I do not have the raw data, so I can't get the variance for each group.

Any thoughts or ideas would be greatly appreciated!

unfortunately for this you would need also the standard deviation or variance of each group, since the p-value strongly depends on how much the values scatter around their means

It is allowed to make an educated guess. I have never heard of this index so I just made a google search and found this paper. In table 4 they show some values of mean and standard deviations. When the mean is 20 the standard deviation seems to be about 10, and when the mean is 30 the standard deviation seems to be about 20. Also, have a look at figure 4.

So there seems to be a significant difference. In R code:

Just like when you design a new study you will need to guess on the standard deviation, often from previous published work.

So if you want others to cite your work you should help the next statistician by saying: "In a new study we would base the simple size needed on the standard deviation of 10"[for example].