The interplay of homicide death investigations conducted by medical personnel employed by the same agency investigating a suspect is fodder for doubts concerning undo influence. Its the subject of this The Intercept article just out from Jordan Smith.
It is no fallacy that forensic experts not happy with their less than “she/he did the crime” results (like some bitemark people I know) will cheat towards stronger opinions when other evidence comes out from the police. I have had a couple judges actually agree with this tactic. Both times, while in court, my explaining what the “scientific method” says about such activities got me into dispute with the “gatekeeper.”