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

Discussion

In the 1116 hospitals studied, 5.07% of patients experienced a medication error. This figure is consistent with the medication error rates of 3-6.9% reported in smaller studies.[2,7,8,9,10] We found that 0.25% of patients experienced a medication error that adversely affected their care; this figure is somewhat lower than the rate of comparable errors identified in previous reports (0.7%).[2] In addition, the finding that 4.9% of all medication errors adversely affected patient care outcomes (18.99 ± 25.30 medication errors that adversely affected patient care outcomes/hospital ÷ 385.83 ± 466.96 medication errors/hospital) is somewhat higher than the comparable figure (3%) reported by the U.S. Pharmacopeia.[14] Since all previous studies of medication errors involved few hospitals or a limited number of patients, our results are probably a more accurate reflection of the actual health care system. Our study, which examined 430,586 medication errors, was 69 times larger than the analysis conducted by the U.S. Pharmacopeia, which relied on data voluntarily provided to that institution for 1999 (6224 medication errors).[14] Whereas we can not assess how many medication errors go unreported, it is likely that a substantial number of medication errors were not detected or reported to the hospital medication error reporting system.

The exact reasons for the differences in total medication errors and medication errors that adversely affected patient care outcomes based on census regions are unknown. Pharmacy services and drug costs have been shown to vary significantly by geographic region.[16,30,32,34,45,46] The Mid-Atlantic region generally has lower levels of clinical pharmacy services when compared with other regions.[16] In addition, fewer hospitals in the Mid-Atlantic region have decentralized pharmacists.[16] The Mid-Atlantic region, which had a low number of medication errors, had the lowest drug costs/occupied bed/year in 1992.[30] The West South Central, Mountain, and Pacific regions, which had higher numbers of medication errors, tended to have higher drug costs/occupied bed/year. Since drug costs probably represent an indirect measure of the number of drugs dispensed/patient, these findings simply may be related to number of drugs used/patient rather than any specific hospital or regional factors. This finding also could be explained by pharmacist workload, in that more drugs/patient (in regions with higher drug costs) may indicate a higher workload for the pharmacy (medication orders/hour). It previously was shown that the risk of medication errors increases substantially when the number of medication orders/hour increases.[47,48,49,50] Whereas the risk of medication errors increases as pharmacist workload increases, there appears to be a critical point (20-25 prescription orders/hour) at which errors dramatically increase.[47,48,49,50]

The exact reasons for the differences between smaller and larger hospitals with regard to total medication errors and medication errors that adversely affected patient care outcomes are unknown. Smaller hospitals reported almost twice the frequency of medication errors compared with medium and large hospitals. In our previous publications, we reported that smaller hospitals provided fewer clinical pharmacy services and had fewer clinical pharmacists when compared with medium and large hospitals.[16,31,33,34] This finding may suggest that smaller hospitals lack the level of patient care services and/or training necessary to reduce medication errors. Clearly, further study is needed to determine the specific reasons why medication errors are affected by hospital size.

One of the more striking findings of this study is the association between pharmacy teaching affiliation and medication errors, including medication errors that adversely affect patient care outcomes. Hospitals that had no teaching affiliations reported a medication error rate 72% greater than hospitals affiliated with pharmacy teaching programs (3.43 ± 9.01 vs 1.99 ± 2.48). Likewise, hospitals that had no teaching affiliation reported 171% more medication errors that adversely affected patient care outcomes than hospitals affiliated with pharmacy teaching programs (0.19 ± 0.33 vs 0.07 ± 0.22). The exact reasons for these findings are unknown.

We previously reported that hospitals affiliated with pharmacy teaching programs had significantly more clinical pharmacy services and employed more clinical pharmacists than hospitals without such affiliations.[16,31,32,34] Whether the increased provision of clinical pharmacy services to patients in these hospitals is related to lower medication errors is unknown. Perhaps pharmacy teaching hospitals have patient care services and/or training that helps reduce medication errors. Student training and supervision may lead to greater scrutiny of medication orders and the dispensing process. In addition, it is also likely that colleges of pharmacy affiliate with better hospitals that provide more clinical services than other hospitals. It appears that pharmacy teaching affiliation is a more important factor in the reduction of medication errors than hospital teaching affiliation (i.e., membership in the Council of Teaching Hospitals or the American Osteopathic Teaching Hospital Association).

The reasons for lower rates of total medication errors and medication errors that adversely affect patient care outcomes in federal government hospitals are unknown. When compared with the mean medication error rate for all study hospitals, federal government hospitals had a 65% lower rate of medication errors (0.89 ± 0.91 vs 2.53 ± 3.98) and a 77% lower rate of medication errors that adversely affect patient care outcomes (0.3 ± 0.07 vs 0.13 ± 0.25). This may be due to fewer drugs being used in federal government hospitals, since drug costs/occupied bed/year in federal government hospitals were 24% lower when compared with all hospitals in the study population ($2291 ± $1279 vs $2997 ± $1267).[30] It is also worth noting that federal government hospitals generally had higher levels of clinical pharmacy services compared with other types of hospitals.[16]

Whereas the associations between a pharmacy director's academic degree and total medication errors and medication errors adversely affecting patient care outcomes were not statistically significant, the trend of this analysis suggests that directors with advanced degrees were in hospitals that had fewer medication errors. It is interesting to note that directors with advanced degrees generally served in hospitals offering higher levels of clinical pharmacy services when compared with directors who had only a bachelor's degree.[16,31,32,34]

Although the fundamental reasons for lower rates of total medication errors and medication errors that adversely affected patient care outcomes in hospitals with decentralized pharmacists are unknown, this finding is not surprising. Previous studies documented that a clinical pharmacist providing decentralized services on work rounds in an ICU decreases preventable adverse drug events by 66%.[51] In addition, in a study of two teaching hospitals, ward-based clinical pharmacists reduced potential adverse drug events (medication errors judged to have significant potential for injuring a patient) by 94%.[52] This variable accounted for some of the most dramatic differences in medication error rates. We found a 45% (1.81-fold) decrease in total medication errors in hospitals that had decentralized pharmacists when compared with hospitals that exclusively had centralized pharmacists. In addition, there was a 94% (16.88-fold) decrease in medication errors that adversely affected patient care outcomes in hospitals with decentralized pharmacists when compared with hospitals that had only centralized pharmacists. This finding is identical to the results of a field study of adverse drug events in children in two Boston hospitals.[52] It appears that one of the most effective ways to prevent medication errors that actually harm patients is to decentralize pharmacists to patient care areas.

Hospitals with decentralized pharmacists had more clinical pharmacists and much higher levels of clinical pharmacy services than hospitals in which all pharmacists were centralized.[16,31,32,34] Substantial documentation indicates that clinical pharmacy services can improve therapeutic outcomes and reduce drug costs.[17,18,19,20,21,53,54,55,56,57,58,59,60,61,62,63,64,65] Providing clinical pharmacy services in a decentralized location may reduce medication errors by reducing inappropriate prescribing and improving the monitoring of patients. Decentralized pharmacists often serve as a final check on drugs provided from the central pharmacy. Previous field studies involving decentralized clinical pharmacists in three hospitals,[51,52] in addition to our findings in 1116 hospitals, strongly suggest that decentralizing pharmacists may be one of the most effective ways to decrease medication errors. Additional study will be needed to elucidate the specific services of decentralized pharmacists that reduce medication error rates.

We found that as total medication errors and medication errors that adversely affect patient care outcomes increased, hospital mortality rates, drug costs/occupied bed, total cost of care/occupied bed, and length of stay also increased. However, only two of these measures were statistically significant: mortality rate (slope = 3.143, p=0.0043, R2 = 10.34%) and total cost of care (slope = 0.027987, p=0.008, R2 = 11.12%). These associations seem logical, since medication errors can lead to increased deaths, morbidity, and hospital costs.[1,2,3,4,5,6,7,8,9,10] Since mortality rates are very good indicators of the quality of patient care,[25,26,38,39,40] lower medication error rates are likely to indicate better quality of care. This is the first study to show that increased medication errors predict increased mortality rates and increased total cost of care in a large number of hospitals. This finding demonstrates that medication errors are not isolated events that only affect individual patients; they have a negative impact on health care and system financials.

The multiple regression analysis is the most important of our statistical analyses. Factors associated with increased 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), the number of registered nurses/ occupied bed (slope = 1.624, p=0.032), the 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 the number of medical residents/occupied bed (slope = -1.478, p=0.0014). Whether the lower mortality rates observed in hospitals with larger numbers of medical residents were due to better care, the involvement of more physicians in the care of patients, or simply to decreased resident caseload is unknown. Although we did not specifically measure medical resident workload (number patients/resident), our data suggest that this variable was probably associated with the rate of medication errors. The data show that hospitals with larger medical teaching programs were associated with lower medication error rates. This finding is consistent with studies of pharmacist workloads, which demonstrate that increased workloads are associated with increased rates of medication errors.[47,48,49,50]

As the number of registered nurses and pharmacists/occupied bed increased, more medication errors were detected. In the simple regression analysis, as the number of licensed practical/vocational nurses and pharmacy technicians increased, fewer medication errors were detected. A study published in 1997 found that 11.4% of 2103 medication errors were due to the wrong drug or dosage (dispensing errors).[8] Similarly, a 1995 study found that 11% of medication errors were pharmacy dispensing errors related to the wrong drug or incorrect strength.[9] Most of the medication errors addressed in the 1997 study involved factors such as insufficient knowledge of drug therapy (30%), lack of awareness of patient's condition (e.g., kidney function) (29.3%), and inaccurate calculations and unit or rate expression factors (17.5%). Other investigators have shown similar results.[1,2,3] It appears that only a small minority of medication errors are due to the wrong drug or dosage. Higher cognitive skills are required to detect and report the vast majority of medication errors. Our findings with registered nurses, registered pharmacists, licensed practical/vocational nurses, and pharmacy technicians suggest that better educated and trained personnel (registered nurses and registered pharmacists) may be necessary to detect medication errors caused by lack of drug and patient care information. This conclusion is also supported by previous studies showing that pharmacists and technicians are about equally accurate when simply checking drugs for correct drug and dosage form.[66,67,68,69,70] However, pharmacists performed much better than licensed vocational nurses in detecting medication errors when these errors involved interpreting patient care data.[71] Unfortunately, there are no studies evaluating the abilities of pharmacists and pharmacy technicians to detect this type of medication errors.

The findings regarding hospital staffing suggest that physicians (medical residents) may be more involved with producing medication errors (workload-prescribing), whereas registered nurses and registered pharmacists may be more involved in detecting and reporting medication errors. The data on licensed practical/vocational nurses and pharmacy technicians suggest that the presence of highly trained nurses and pharmacists in patient care areas is essential to detection and reporting of medication errors. It also appears the location of pharmacists is very important for the reduction of medication errors. For pharmacists, it appears that increased staffing is desirable and that staff should be decentralized to improve the quality of patient care.

Clearly, the results of this study provide some direction for reducing medication errors in U.S. hospitals. Census region, hospital size, pharmacy teaching affiliation, type of hospital, location of pharmacists, mortality rates, total cost of care, as well as staffing for medical residents, registered nurses, and registered pharmacists, play important roles in medication error rates.

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....