Medication Errors in United States Hospitals

C. A. Bond, PharmD, FASHP, FCCP, Cynthia L. Raehl, PharmD, FASHP, FCCP, and Todd Franke, PhD

Disclosures

Pharmacotherapy. 2001;21(9) 

In This Article

Results

A total of 1336 hospitals (84%) of the 1597 general-medical, surgical hospitals from the 1992 National Clinical Pharmacy Services database reported having a medication error reporting system in place. The AHA database identified 3444 hospitals that could have been included in the survey; of these, 46% (1597) responded.[36] A total of 1116 hospitals provided specific information on the number of medication errors. Information on the number of medication errors that adversely affected patient care outcomes was reported by 913 hospitals. The 1116 hospitals (70% of the 1597 hospitals) constituted the study population. Hospitals that reported total medication errors and those that reported medication errors adversely affecting patient care outcomes correspond to the useable response rates of 32% and 27%, respectively (1116/3444 and 913/3444 hospitals). Data comparisons are typically lower than the 1116 and 913 hospitals, as not all survey respondents completed all parts of the survey and the variables used to adjust for severity of illness (health care outcomes) were not available for all hospitals. The mean number of total medication errors reported/year/hospital was 385.83 ± 466.96. The mean number of medication errors reported/year/hospital that adversely affected patient outcomes was 18.99 ± 25.30, or 4.9% (18.99/385.83) of total medication errors/hospital. For each medication error that adversely affected patient care outcomes, 20.32 medication errors were actually reported.

The mean number of admissions/year for the study hospitals was 7611 ± 6514 admissions/hospital or 8,493,876 total admissions (36% of all admissions to U.S. hospitals). The average daily census for study hospitals was 170.3 ± 158.0 patients/day. There were 2.26 ± 3.98 medication errors/occupied bed/year and 0.12 ± 0.25 medication errors that adversely affected patient care outcomes/occupied bed/year. The total number of medication errors was 430,586, whereas 17,338 medication errors were documented that adversely affected patient care outcomes. Medication errors occurred in 5.07% of the patients admitted to these hospitals each year. Each hospital experienced a medication error every 22.7 hours (every 19.73 admissions). Medication errors that adversely affected patient care outcomes occurred in 0.25% of all patients admitted to these hospitals/year. Each hospital experienced a medication error that adversely affected patient care outcomes every 19.23 days (or every 401 admissions).

Table 1 shows the association between census regions and the number of total medication errors and number of medication errors that adversely affected patient care outcomes/occupied bed/year. Medication errors were reported more frequently in the West South Central (3.07 ± 5.09), West North Central (3.25 ± 2.73), and Mountain (3.38 ± 2.56) regions. Medication errors were reported less frequently in the East South Central (2.18 ± 2.18), South Atlantic (2.03 ± 2.39), and Mid-Atlantic (1.32 ± 1.87) regions. Medication errors that adversely affected patient care outcomes were more common in the West North Central (0.14 ± 0.29), West South Central (0.15 ± 0.26), and Mountain (0.24 ± 0.37) regions. Medication errors that adversely affected patient care outcomes were less frequent in the Pacific (0.09 ± 0.26), New England (0.09 ± 0.13), and the Mid-Atlantic (0.05 ± 0.10) regions.

Table 2  shows the associations between hospital size and the number of overall medication errors and the number of medication errors that adversely affected patient care outcomes/occupied bed/year. Medication errors were more common in small (2.87 ± 5.36), than in medium (1.55 ± 1.98) and large (1.34 ± 1.70), hospitals. Table 3 shows the association between pharmacy and hospital teaching affiliation and the number of total medication errors and the number of medication errors adversely affecting patient care outcomes/occupied bed/year. Hospitals that were affiliated with a pharmacy teaching program (1.99 ± 2.48) had a much lower number of total medication errors than hospitals that had nonpharmacy teaching affiliations (2.62 ± 3.05) or no teaching affiliations (3.43 ± 9.01). Likewise, hospitals affiliated with a pharmacy teaching program (0.07 ± 0.22) had a lower number of medication errors that adversely affected patient care outcomes than hospitals that had nonpharmacy teaching affiliations (0.12 ± 0.27) or no teaching affiliations (0.19 ± 0.33).

Table 4 shows the association between type of hospital and the number of total medication errors and the number of medication errors that adversely affected patient care outcomes/occupied bed/year. Federal hospitals had much lower total medication error rates (0.89 ± 0.91) and medication error rates that adversely affected patient care outcomes (0.02 ± 0.07) than other types of hospitals. Table 5 shows the association between the academic degree of the pharmacy director and the number of total medication errors and the number of medication errors that adversely affected patient care outcomes/occupied bed/year. Hospitals in which the pharmacy director had earned a master of science in pharmacy or a doctor of pharmacy had lower rates of errors than other hospitals (total medication errors, 1.85 ± 2.69; and medication errors that adversely affected patient care outcomes, 0.09 ± 0.14). However, differences between the error rates of these and other hospitals were not statistically significant.

Table 6 shows the association between the physical location of pharmacists within the hospital and the number of total medication errors and the number of medication errors that adversely affected patient care outcomes/occupied bed/year. The lowest numbers of both overall medication errors (1.74 ± 2.51) and medication errors that adversely affected patient care outcomes (0.09 ± 0.19) were seen in hospitals in which pharmacists served in decentralized settings. Hospitals with decentralized pharma-cists had a 45% decrease in medication errors (1.81-fold decrease) when compared with hospitals that only had centralized pharmacists. Moreover, hospitals with decentralized pharma-cists had a 94% decrease in medication errors that adversely affected patient care outcomes (16.88-fold decrease) when compared with hospitals that had only centralized pharmacists.

Table 7 shows a severity of illness-adjusted multiple regression analysis between medication errors/occupied bed/year and four health care outcome variables: hospital mortality rate, drug costs/occupied bed/year, total cost of care/occupied bed/year, and hospital length of stay. As mortality rate (slope = 3.143, p=0.0043) and total cost of care (slope = 0.027987, p=0.0062) increased, the number of medication errors also increased. None of these health care outcome variables had statistically significant associations with the number of medication errors that adversely affected patient outcomes.

Table 8 shows a simple regression analysis between medication errors/occupied bed/year and hospital staffing/occupied bed. Several health care staffing ratios are included in this analysis. As the number of hospital adminis-trators (slope = 2.9931, p=0.04), registered nurses (slope = 0.6908, p=0.0008), ratio of registered nurses to licensed practical/vocational nurses (slope = 2.5563, p=0.314), registered pharmacists (slope = 9.996, p=0.002), physical therapists (slope = 21.4581, p=0.001), and total hospital personnel (slope = 0.1128, p=0.0082) increased, the number of medication errors also increased. Conversely, as the number of medical residents (slope = -3.1541, p=0.0032) and pharmacy technicians (-0.0529, p=0.0029) increased, the number of medication errors decreased. These were the only statisti-cally significant associations with this regression model. The only variable that had a statistically significant association with the number of medication errors that adversely affected patient outcomes was the number of medical residents/ occupied bed (slope = -0.1571, standard error [SE] = 0.07423, p=0.0346, confidence interval [CI] = -0.3028 to -0.0114).

Table 9 shows a multiple regression analysis between all of the variables presented in tables 1-8 and medication errors/occupied bed/year. Only statistically significant associations are presented. Factors associated with increased total medication errors/occupied bed/year were lack of pharmacy teaching affiliation (slope = 0.8875, p=0.0416), centralized pharmacists (slope = 1.0942, p=0.0001), number of registered nurses/occupied bed (slope = 1.624, p=0.032), number of registered pharmacists/occupied bed (slope = 25.0573, p=0.0001), hospital mortality rate (slope = 2.8017, p=0.0192), and total cost of care/occupied bed/year (slope = 0.01432, p=0.0091). Factors associated with decreased medication errors were location in the Mid-Atlantic region (slope = -1.5182, p=0.03), affiliation with a pharmacy teaching program (slope = -1.0252, p=0.0349), decentralized pharmacists (slope = -0.9843, p=0.0037), and number of medical residents/occupied bed (slope = -1.478, p=0.0014). The actual R2 for this analysis was 21.03%, and the adjusted R2 was 15.94%. A similar multiple regression analysis was performed with medication errors that adversely affected patient outcomes. This analysis was not statistically significant.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....