Converting from PDF to Word, Excel or any other format is one of the most complex things you can try to do with a PDF file. It works very well in some cases, in other cases the output has very little to do with the original file. The key for success is that the PDF file needs to be "tagged" - which means that it contains information about the information that is displayed in the file. The best way to make sure that a PDF file is tagged correctly is by using the PDFMaker in Acrobat to create the PDF file from Word or Excel (that's the Acrobat ribbon or toolbar).
Unfortunately there is not much you can do to improve the output without spending a lot of time (e.g. by manually tagging the file). Also, if you are using Adobe's ExportPDF service and don't have access to Acrobat, that is not even an option.
In most cases, the only thing you can do is complain to the original author of the file and tell them that they used a bad PDF generator to create the PDF file.
Sometimes you get better results when you export the whole file to individual TIFF images (e.g. 600dpi), then import these TIFF images again as a PDF file, and finally run OCR and export to Word.
I tried converting it to a TIFF and then saving as a pdf and THEN exporting it to word, and it worked!