Caplan argued recently  for treating someone with an addiction against her will by administering naltrexone involuntarily. Denying her right to refuse treatment is defended ethically by a seemingly contradictory or paradoxical reason: it is expected to restore her autonomy. In our view, the article's justification of forced treatment is debatable for three reasons.
First, only the biological component of addictions is examined. Psychological and social factors and their complex interactions with the biological are omitted; so naltrexone's actual and sustained ability, regardless of involuntary/voluntary use, to curb addiction may be much less than presumed.
Second, cravings are described as coercive. Coercion is defined typically as involving threats. While telling a client ‘swallow this pill or no afternoon pass’ constitutes a threat and thereby coercion, how do neuropsychological processes threaten someone? Cravings can be difficult, but coercive? Addiction is also described as involving ‘compulsion’ and overriding free will. However, many people manage their cravings and stop using substances on their own, so cravings are not entirely compulsive . The concepts of coercion and compulsion are not as fitting as needed.
Similarly, the portrayal of people with an addiction seems extreme. It is important to remember that tobacco and alcohol are the most harmful and common addictions, not illegal drugs. Phrases such as ‘truly addicted’ and ‘in the throes of addiction’ can imply that people either are or are not addicted; yet addiction is about degrees. Autonomy is presented in extreme ways, too. To say ‘an addict cannot be a fully free, autonomous agent’ presumes that people usually are. This is questionable. Recent debates about autonomy as a ‘hypervalue’[3,4] and the corrective of relational autonomy  warrant more contextualized use of autonomy in discussions about addictions.
Third, we believe that Caplan argues for a provocative conclusion because he knows that successful recovery can be an exhausting series of successes and relapses. However, social justice considerations are overlooked. Many addicted people are multiply disadvantaged: unemployed, poor, inadequately housed, medically untreated, isolated and stigmatized. Social justice demands increased caution if rights (i.e. refusal of medical treatment) may be denied. As per the precautionary principle , unless there is credible, convincing evidence of the effectiveness of ‘denying autonomy to create it’, the proposed practice is unsupportable. Furthermore, what could be needed to administer a medication involuntarily has been overlooked. Oral forms have choking risks. Injectable forms may require physical restraints. Damage to therapeutic alliances demands very cautious use of forced therapies.
The paper accepts that addicted people still have capacity for all types of decisions. Instead of involuntary treatment, some type of advance directive or Ulysses contract is more respectful of autonomy . Also, a harm reduction approach for successful, sustainable treatment and intervention, as demonstrated by opioid agonist maintenance therapy  and supervised injection sites, is more balanced. Reliable health, social and outreach services plus non-coercive drug use expectations typify these programmes. They help to empower a person to be more autonomous in terms of enjoying the freedom of decision-making, accepting attendant responsibilities and pursuing what is personally meaningful.