It is possible via the absorbance shift of the spectra comparing it with the original or undegraded drug. You can also determine the concentration of the drug by using known concentrations of the drug to find the unknown concentration or concentration of the drug after degradation.
Yes, if you established that your absorbance wavelength is away from the other degradants maximum absorbance. It needs HPLC separation first to establish that.
The answer is a bit complex. It is possible based on some assumptions:
1) UV is molecular spectroscopy; drug degradation changes need to generate molecular evolution directly related to specific functions inside the molecule.
2) The changes may not be related to the degradation products with the same chemical functions; otherwise, the chemical shifts can be very tiny, not associated with any degradation pathway, in strong overlapping.
3) It needs to be easily related to a secondary product, increasing the probability of signal shift recognizance.
4) It may change the original signal of the probed molecule significantly, so the identification is straight forward.
5) The changes may be in overlapping and may not be identified as it is. The changes need to be undoubtedly related to the decomposition product.
6) By other means, the change may be monitored by the original signal variances. However, it would be unsure if the signal change is due to decomposition only unless proven by another technique.
As UV spectra are a set of overlapping signals, in general, it is possible, with extra care, not to falsely relate to a simple blocking of chemical function inside the molecule. Doble bonding suppression, salt formation, or even breaking the original molecule leaving the same chemical function available for an extra signal like carboxyl in a degraded molecule may generate a false negative for degradation product.
Finally, the HPLC is a traditional and acceptable method for decomposition product identification in any drug-related quality control method.
The regulatory agencies need to recognize the applied methodology; otherwise, it would only be acceptable as single laboratory identification, not industrially or officially commercial quality control procedure.