In March 2015, DeKalb County, Georgia, police officer Robert Wilson, was dispatched to check on a naked person (Anthony Hill) who was acting bizarrely and wandering around an apartment building.
When Officer Wilson spotted Hill, he stepped out of his patrol car to make contact. Hill ran directly at him at high speed flailing his arms. Officer Wilson stepped back toward the rear of his patrol car to create distance while yelling for Hill to stop. Hill did not stop and Officer Wilson discharged his firearm killing Hill.
Former Officer Wilson was criminally charged with many counts, including murder. The prosecution’s use of force expert admitted that an electrical weapon deployment towards a charging person required a trick shot (since a probe would have to go into the thorax and the other into a thigh). Nevertheless, former officer Wilson was convicted of various charges and sentenced to 20 years in prison.
Study reviews LE response to the naked subject
Two well-recognized researchers on police use of force, Darrell Ross and Michael Brave, recently published a peer-reviewed study of police response to naked subjects. 
They found 397 incidents occurring from 1998–2018, averaging about 20 incidents annually, and then did a more detailed study of 215 incidents covered in Federal Court (42 U.S.C. § 1983 civil rights litigation) decisions from 1997 to 2019. This study shows how extremely dangerous such thankfully rare incidents can be.
Some of the percentages given below are from the larger (n=397) group and some are from the smaller legal-case group. (For details, you can request the actual paper here.)
In about 80% of the incidents, the subject charged, actively fought, or assaulted the responding law enforcement officer (LEO); in about 36% of the incidents, the person attempted to disarm the LEO. The subject was unarmed in 75% of the incidents, armed with a firearm in 15% and possessed an edged weapon in 10%. The subject was either mentally ill or on illicit drugs in 89% of the cases; and died in 86% of cases.
In the 22 deadly-force incidents the officers had limited reaction time, usually less than 10 seconds, from the beginning of the contact.
Overall, de-escalation techniques were attempted in 80% of the incidents but were never successful.
In the deadly-force incidents, the LEO had limited time to use an electrical weapon, an impact weapon, or an aerosol.
In about 20% of the incidents overall, intermediate weapons were ineffective, requiring LEOs to transition to deadly force based on the subject’s assaultive nature.
About 80% of the subjects exhibited clear signs of profound agitation syndrome with 6 to 11 of the classic diagnostic factors such as unexpected strength, pain insensitivity, extreme stamina, sweating, non-compliance, hyperactivity, agitation, and incoherent speech. Because of the danger of these encounters, the authors recommend that a minimum of 2 (ideally 4) officers respond.
This study shows that an encounter with a naked subject presents an unpredictable and dangerous call, although very rare (1 in 10,000,000 citizen encounters). The courts granted summary judgment to LEOs in 70–75% of the cases, primarily based on the threat and danger posed by the subject. Higher levels of force were often used in response to the higher levels of violence and assaultive behaviors exhibited by the subjects who most were either mentally ill, under the influence of an illicit drug, or both.
Police training for naked subject calls
In the Implications section, the authors focused on policy and training. Since these incidents are very rare, agencies are encouraged to develop and use brief study and memory aides to assist LEOs with applying optimal force methods and procedures, and to document such incidents.
Some links for training and memory aids include:
Predictors of a fatal outcome
The Ross and Brave study is consistent with the results of a recently published Dutch study that compared excited-behavior restraint deaths with non-fatal excited-behavior restraint incidents.  Full or partial nudity increased the risk of death by 4x and thus disrobing should be seen as an independent predictor of a fatal outcome. That, in turn, is consistent with recent studies that have shown that hyperthermia is associated with the most severe agitation cases. 
1. Ross DL, Brave MA. Assessing Use-of-Force Liability and Law Enforcement Response to the Naked Subject. Law Enforcement Executive Forum. 2020;20(1):1-21.
2. Dijkhuizen LGM, Kubat B, Duijst W. Sudden death during physical restraint by the Dutch police. J Forensic Leg Med. 2020;72:101966.
3. Baldwin S, Hall C, Bennell C, Blaskovits B, Lawrence C. Distinguishing features of Excited Delirium Syndrome in non-fatal use of force encounters. J Forensic Leg Med. 2016;41:21-27.