Can medication mistakes and patient harm be prevented through the use of a well-developed electronic prescribing system in a hospital?

Over 80 percent of medication mistakes (other than missing doses) were prevented at a teaching hospital by a computerized physician order-entry system once it was fully developed with clinical decision support features. Medication mistakes that caused patient harm or had the potential to cause harm (and were not intercepted before reaching the patient) were reduced by 86 percent.

Why is this important?

Medication mistakes are a common occurrence in hospitals. Although most are not serious, an estimated 380,000 to 450,000 cause some kind of harm (from minor to life-threatening) to hospital patients each year. Many others are "near misses" that might have caused harm.

To reduce this threat to patient safety, the Institute of Medicine has called for greater use of information technologies for prescribing and dispensing medications (IOM 1999, 2006).

Interventions

A large Boston teaching hospital developed a computerized physician order-entry (CPOE) system over a five-year period, with more sophisticated clinical decision support features added over time (Bates et al. 1999).

• At baseline (October–November 1992), all prescriptions were written manually.
• At Period 1 (October–December 1993), basic features had been implemented, including medication name, dose, and frequency selection from standard lists; display of relevant laboratory results; limited drug-allergy checks; redundant medication checking; and notification of a few life-threatening drug-to-drug interactions and drug-laboratory problems.
• At Period 2 (November–December 1995), improved drug-allergy checking had been implemented.
• At Period 3 (March–April 1997), improved drug-to drug interaction checking had been implemented, along with improved potassium ordering rules.

Findings

The overall rate of medication errors (other than missing doses) decreased by 80 percent when the CPOE system was fully developed (Bates 1999).

• Serious medication errors—those that actually caused injury or had the potential to cause injury and were not caught before reaching the patient—fell by 86 percent.
• Preventable adverse drug events—injury due to medication mistakes—declined by 62 percent.
• Potential adverse drug events—"near misses" that reached the patient but didn't cause injury by chance—were reduced 100 percent to zero.
• Error reductions occurred at all stages of the medication use process (drug ordering, transcribing, dispensing, and administering).

A cost analysis found that the CPOE system cost $11.8 million to develop, implement, and operate over 10 years but produced total savings of $28.5 million, for net savings of $16.7 million over 10 years including net savings of $9.5 million to the hospital's operating budget (Kaushal et al. 2006). A few high-impact system elements were responsible for most of the cost-savings.

Implications

Effective use of well-designed computerized physician prescribing systems in every non-rural hospital in the United States could prevent an estimated 522,000 serious medication errors each year. Assuming that only one-tenth of one percent of such errors resulted in patient death, over 500 deaths could be prevented each year as a result of hospital-based computerized prescribing (Birkmeyer and Birkmeyer 2000).

Implementing a CPOE system is a complex undertaking, with success likely to be influenced by a combination of financial, technical, and organizational factors. The Institute of Medicine noted that hospitals must carefully evaluate CPOE systems to ensure good fit, adequately train staff to use systems effectively, and engage in concomitant clinical process design and ongoing evaluation and improvement to ensure that such systems meet their objectives (IOM 2006).

Interviews with hospital executives suggested the following policy levers to speed adoption of CPOE: external pressure to improve patient safety, financial incentives and access to capital, standardization of vocabulary and protocols, and knowledge-building to understand and overcome barriers (Poon et al. 2004).

Measure:

The study included all patients admitted to two general-care medical units and one medical intensive-care unit. Data were collected during four specific study periods ranging from 49 to 68 days over four years. Medication mistakes, adverse events, and potential adverse events were detected by: 1) pharmacists during dispensing and solicited from nurses during daily visits, 2) evaluation by trained reviewers of medication sheets received by the pharmacy, and 3) chart review by a study investigator. Medication errors were defined as "an error in the process of ordering, dispensing, or administering a medication." Serious medication errors were defined as "those associated with a preventable adverse drug event or a potential adverse drug event" (Bates et al. 1999).

Limitations:

Results might not be directly generalizable to other hospitals and systems, although they provide a useful benchmark to inform decision-making. Additional research is needed on the costs and benefits of commercially developed systems that are likely to be implemented in many hospitals.

Source:

The study was conducted by clinician-researchers and informaticists at Brigham and Women's Hospital, a 700-bed academic tertiary-care hospital located in Boston, and at Harvard Medical School and the Harvard School of Public Health (Bates et al. 1999).

References:

* Indicates source of data used in the chart(s).* Bates, D. W., J. M. Teich, J. Lee et al. 1999. The Impact of Computerized Physician Order Entry on Medication Error Prevention. Journal of the American Medical Informatics Association 6 (4): 313–21.

Birkmeyer, J. D., and C. M. Birkmeyer. 2000. Leapfrog Safety Standards: The Potential Benefits of Universal Adoption. Washington, D.C.: The Leapfrog Group.

IOM (Institute of Medicine). 1999. To Err Is Human: Building a Safer Health Care System. Washington, D.C.: National Academy Press.

IOM (Institute of Medicine). 2006. Preventing Medication Errors. Washington: National Academy Press.

Kaushal, R., A. K. Jha, C. Franz et al. 2006. Return on Investment for a Computerized Physician Order Entry System. Journal of the American Medical Informatics Association 13 (3): 261–6.

Poon, E. G., D. Blumenthal, T. Jaggi et al. 2004. Overcoming Barriers to Adopting and Implementing Computerized Physician Order Entry Systems in U.S. Hospitals. Health Affairs (Millwood) 23 (4): 184–90.