The effect of provider characteristics on the responses to medication-related decision support alerts

Insook Cho; Sarah P Slight; Karen C Nanji; Diane L Seger; Nivethietha Maniam; Julie M Fiskio; Patricia C Dykes; David W Bates

doi:10.1016/j.ijmedinf.2015.04.006

The effect of provider characteristics on the responses to medication-related decision support alerts

Int J Med Inform. 2015 Sep;84(9):630-9. doi: 10.1016/j.ijmedinf.2015.04.006. Epub 2015 May 9.

Authors

Insook Cho¹, Sarah P Slight², Karen C Nanji³, Diane L Seger⁴, Nivethietha Maniam⁴, Julie M Fiskio⁴, Patricia C Dykes⁵, David W Bates⁶

Affiliations

¹ The Center for Patient Safety Research and Practice, Division of General Internal Medicine, Brigham and Women's Hospital, Boston, MA, USA; Nursing Department, Inha University, Incheon, South Korea; Harvard Medical School, Boston, MA, USA. Electronic address: Insook.cho@inha.ac.kr.
² The Center for Patient Safety Research and Practice, Division of General Internal Medicine, Brigham and Women's Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA; Division of Pharmacy, School of Pharmacy, Medicines and Health, Durham University, Stockton-on-Tees, UK.
³ Harvard Medical School, Boston, MA, USA; Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USA.
⁴ Partners Healthcare Systems, Inc., Wellesley, MA, USA.
⁵ The Center for Patient Safety Research and Practice, Division of General Internal Medicine, Brigham and Women's Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA.
⁶ The Center for Patient Safety Research and Practice, Division of General Internal Medicine, Brigham and Women's Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA; Partners Healthcare Systems, Inc., Wellesley, MA, USA.

PMID: 26004341
DOI: 10.1016/j.ijmedinf.2015.04.006

Abstract

Background: Improving the quality of prescribing and appropriate handling of alerts remains a challenge for design and implementation of clinical decision support (CDS) and comparatively little is known about the effects that provider characteristics have on how providers respond to medication alerts.

Objectives: To investigate the relationship between provider characteristics and their response to medication alerts in the outpatient setting.

Design and participants: Retrospective observational study using a prescription log from the automated electronic outpatient system for each of 478 providers using the system at primary care practices affiliated with 2 teaching hospitals, from 2009 to 2011 for six types of alerts. Provider characteristics were obtained from the hospital credentialing system and the Massachusetts Board of Registration in Medicine.

Main measures: Override rates per 100 prescriptions and 100 alerts.

Results: The providers' mean override rates per 100 prescriptions and per 100 alerts were 0.52 (95% confidence interval (CI), 0.46-0.58) and 0.42 (95% CI, 0.38-0.44) respectively. The physicians (n=422) on average overrode drug alerts with rates of 0.48 per 100 drugs and 0.44 per 100 warnings. Univariate analysis revealed that six physician characteristics (physician type, age, number of encounters, medical school ranking, residency hospital ranking, and acceptance of Medicaid) were significantly related to the override rate. Multiple regression showed that house staff were more likely to override than staff physicians (p<0.001), physicians with fewer than 13 average daily encounters were more likely to override than others with more than 13 encounters (p (range), <0.001-0.05), and graduates of the top 5 medical schools were more likely to override than the others (p=0.04). All six predictors together explained 30% and 50% of the variance in override rates, respectively.

Conclusions: Consideration of six specific physician characteristics may help inform interventions to improve prescriber decision-making.

Keywords: Decision support; Medication safety; Prescribing; Provider characteristics; Quality.

Publication types

Observational Study
Research Support, Non-U.S. Gov't

MeSH terms

Adult
Ambulatory Care
Cross-Sectional Studies
Decision Support Systems, Clinical*
Drug Interactions*
Drug Therapy, Computer-Assisted*
Electronic Prescribing
Female
Hospitals, Teaching
Humans
Male
Massachusetts
Medical Order Entry Systems / statistics & numerical data*
Medication Errors / prevention & control*
Middle Aged
Practice Patterns, Physicians'*
Primary Health Care
Retrospective Studies