Clinician-computer interaction: automated review of psychotropic drugs five years later

Abstract

Five years after its introduction at a state mental hospital, an automated drug exception review system continues to show a long-term impact on the prescribing practices of hospital physicians. While the overall rate of exceptions has remained low, approximately one quarter of all new exceptions pointed out by the computer result in a change in the order by the physician. The fact that 60% of new exceptions are justified suggests that some forms of polypharmacy may be appropriate. The integration of the exception reporting system into the clinical review process has avoided the danger of the computer being seen as an adversary to the clinician or as exerting a "big-brother"-like control over psychopharmacologic prescription practices.An automated drug exception system which provides a quality assurance review for psychotropic drug orders in exception to predetermined prescribing guidelines has been in place at Rockland Psychiatric Center (RPC) under the auspices of the Multi-1 State Information System (MSIS) since 1975 . The present study was an attempt to examine the longterm impact of this system as regards: (1) the appropriateness of the guidelines; (2) the nature of the clinician-computer interaction and (3) the longterm impact on prescribing practices. MethodThe automated drug review system consists of a drug ordering segment which captures all psychotropic drug prescriptions ordered for RPC inpatients and an exception review module which automatically reviews all new drug prescriptions against a predetermined set of guidelines falling into two broad categories: dosage and polypharmacy. In addition, polypharmacy guidelines include both prescription of more than one drug within the same drug class (e.g., neuroleptics (NL), tricyclic antidepressants (TCA), antiparkinson drugs (AP), antianxiety agents (AA) and hypnotics (HY)), and prescription of potentially hazardous combinations of differing drug classes.When an exception is identified, the computer produces a letter to the prescribing physician with a copy forwarded to the physician's immediate physician supervisor. In consultation, these physicians decide whether the exception is acceptable, based on the patient's clinical needs, or whether a change in medication orders is indicated. When agreement is reached, the decision is referred to a Drug Monitoring Committee (DMC) of physicians for review which also arbitrates in the event a consensus cannot be reached by the prescribing physician and supervisor. Thus, the computer triggers a physician clinical review process.The present study focused on the computer exception reports generated during the period January, 1981 through January, 1982. The authors retrieved three broad categories of consultation outcome from these reports: (1) justification; (2) medication change; (3) other. ResultsDuring an average month, approximately 7% of the total patient population receiving psychotropic drugs had exceptions reported by the automated system. During the study period, there was a total of 263 exceptions. The most com...

Structured Interventions to Optimize Polypharmacy in Psychiatric Treatment and Nursing Homes

“…Fifteen studies were randomized controlled trials [31][32][33][34][35][36][37][38][39][40][41][42][43][44][45] ; 27 were nonrandomized, noncontrolled before-after designs ; 5 were randomized before-after designs without controls [73][74][75][76][77] ; and 9 were retrospective analyses. [78][79][80][81][82][83][84][85][86] A risk of bias assessment can be found in Supplementary Table 3, http:// links.lww.com/JCP/A800.…”

“…Publication time ranged from 1980 to 2021. Fifteen studies were randomized controlled trials 31–45 ; 27 were nonrandomized, noncontrolled before-after designs 46–72 ; 5 were randomized before-after designs without controls 73–77 ; and 9 were retrospective analyses 78–86 . A risk of bias assessment can be found in Supplementary Table 3, http://links.lww.com/JCP/A800.…”

“…[40][41][42]49,52,[54][55][56][58][59][60]63,65,68,71,87 Most studies (36 of 58) performed a comprehensive review, considering the entirety of drugs used and conditions of the participant, whereas others focused on specific aspects of the medication or specific side effects. 31,32,37,40,42,43,46,49,56,58,59,61,65,66,68,71,73,74,[81][82][83]86 Only few studies (12 of 58) defined the inclusion criteria for performing a medication review with varying definitions of polypharmacy or arbitrary thresholds such as "more than four antipsychotics." 86 Several studies in nursing homes used an automated alert prompting review.…”

“…Adoption rate of changes suggested during the intervention was only reported by 3 studies and varied greatly between 25% and 98.6%. 71,83,85 None reported maintenance of changes over time.…”