Just by chance I came across the following simple substitution: a phenol--->indole change, where the NH of the indole functions like the phenolic OH in hydrogen bonding situations but avoids the unfavourable glucuronidation. Certainly not a general solution, but a useful one that may work in some cases

Reference:
5-, 6-, 7- and 8-amino-2-(N,N-di-n-propylamino)-1,2,3,4-tetrahydronaphthalenes: centrally acting DA and 5-HT1A agonists
European Journal of Medicinal Chemistry, Volume 28, Issue 9, 1993, Pages 693-701
P Stjernlöf, T Elebring, B Andersson, A Svensson, K Svensson, A Ekman, A Carlsson and H Wikström