Pharmacological treatment of self-injurious behavior in mentally retarded persons.

Singh et al. (1985) hunted every drug paper on self-injury in people with intellectual disability. They looked at studies from the 1960s through 1984. They wanted to know which pills, if any, really stop head-banging, hand-biting, or eye-poking.

The team graded each study for size, design, and follow-up. Most trials were tiny, short, or lacked control groups. The authors called the evidence base 'sparse and methodologically weak.'

Only two drug classes showed even weak promise: antipsychotics like haloperidol and antimanics like lithium. Effect sizes were not reported. Side-effect notes were missing in half the papers.

The review found no good proof that any pill beats placebo for self-injury. The authors warned clinicians to stay cautious and to keep behavioral plans as the first line.

Iwata (1993) updated the same question eight years later and still called the data 'limited.' The later review kept the same bottom line: pills are backup, not frontline.

Coe et al. (1997) gave teams a road map. Their behavioral diagnostic paradigm tells you when to test a med and when to fade it as behavior plans work. The paper turns the 1985 warning into action steps.

van der Geest et al. (2002) showed the next step: mentor frontline staff. Their multiple-baseline study proved that short mentoring lifts team use of combined behavioral and drug plans. The 1985 gap moved from 'we need evidence' to 'we need teamwork.'

If you write behavior plans for adults or kids with ID and self-injury, this paper is your shield against quick-fix pharmacy requests. Use it to show families and physicians why you start with functional assessment, skill building, and reinforcement before trialing any med. When a drug is trialed, set clear behavioral targets and withdrawal criteria just as Coe et al. (1997) suggest. The review also signals you to demand better data: ask for baseline, control phases, and side-effect logs any time a new pill is proposed.