Further characterization of the influence of crowding on medication errors.

STUDY
OBJECTIVES: Our prior analysis suggested that error frequency increases disproportionately with Emergency department (ED) crowding. To further characterize, we measured this association while controlling for the number of charts reviewed and the presence of ambulance diversion status. We hypothesized that errors would occur significantly more frequently as crowding increased, even after controlling for higher patient volumes.
MATERIALS AND METHODS: We performed a prospective, observational study in a large, community hospital ED from May to October of 2009. Our ED has full-time pharmacists who review orders of patients to help identify errors prior to their causing harm. Research volunteers shadowed our ED pharmacists over discrete 4- hour time periods during their reviews of orders on patients in the ED. The total numbers of charts reviewed and errors identified were documented along with details for each error type, severity, and category. We then measured the correlation between error rate (number of errors divided by total number of charts reviewed) and ED occupancy rate while controlling for diversion status during the observational period. We estimated a sample size requirement of at least 45 errors identified to allow detection of an effect size of 0.6 based on our historical data.
RESULTS: During 324 hours of surveillance, 1171 charts were reviewed and 87 errors were identified. Median error rate per 4-hour block was 5.8% of charts reviewed (IQR 0-13). No significant change was seen with ED occupancy rate (Spearman's rho = -.08, P = .49). Median error rate during times on ambulance diversion was almost twice as large (11%, IQR 0-17), but this rate did not reach statistical significance in univariate or multivariate analysis.
CONCLUSIONS: Error frequency appears to remain relatively constant across the range of crowding in our ED when controlling for patient volume via the quantity of orders reviewed. Error quantity therefore increases with crowding, but not at a rate greater than the expected baseline error rate that occurs in uncrowded conditions. These findings suggest that crowding will increase error quantity in a linear fashion.

INTRODUCTION

Background

Among the many challenges that emergency departments (EDs) are facing, crowding appears to be a major influence. ED crowding represents an international crisis that may affect the quality of and access to healthcare.[1] Causes of crowding have been categorized as; input factors (i.e., non-urgent visits, so-called frequent-flyer patients, the influenza season, etc.); throughput factors (i.e., lower staffing and mandated medical screening for all patients who present to an ED); outflow factors (i.e., inpatient boarding[2] with 22% of all ED patients boarding at one time in some cases); and hospital bed shortages.[1] Regardless of the aggravating factors, crowding appears to result in increased risk of death and disability,[34567] poor quality of care in patients with severe pain, and medication errors. Medication errors (including dosing protocols or routes that are incorrectly administered[8]) contribute to significant morbidity, mortality, and costs to the health system.[9] One study of a national medication error database found that nearly 11,000 medication errors were reported over a 5-year period by EDs in 484 unique facilities.[10] Despite many reports on the impact of crowding on medication errors[11] few studies of this problem have been published.[121314]

Study objective

Our prior analysis suggested that medication error frequency increases disproportionately with crowding.[15] And another prior study suggested that Emergency Department Work Index (EDWIN) scores may provide some measure to quantify crowding and perhaps discriminate between conditions that may be associated with higher error frequencies.[16] In order to better determine the relation between crowding and medication error occurrence, and control for variables not measured in our earlier study; we measured the association between medication errors and crowding in the ED over discrete time intervals while controlling for the number of charts reviewed and the presence of ambulance diversion status.

Hypothesis

We hypothesized that escalation in medication errors is correlated positively with crowding in the ED regardless of controlling for higher patient volumes.

MATERIALS AND METHODS

Study participants, data source, institutional setting, and patient census.

We performed this prospective, observational study in a large, community hospital ED from May to October of 2009. ED full time pharmacists, who worked weekday morning and evening shifts, reviewed orders of patients to help identify errors prior to their causing harm. Our ED has 50 licensed beds with annual patient visits at the time of the study of approximately 85,000. Our ED supports a 3-year emergency medicine residency program with 33 residents in postgraduate year (PGY) 1-3 format, and at the time of the study employed two full-time ED pharmacists working different 8-h shifts, 5 days a week.

Intervention

The morning shift pharmacist reviewed and processed medication orders placed from early morning to late afternoon and the evening shift pharmacist reviewed and processed orders from later that afternoon until 11:00 pm. These orders comprised of medications ordered for newly arrived patients, patients in the critical care area of our ED, and boarded patients. There was no pharmacist coverage during overnight hours and weekends. In addition to satisfying responsibilities pertaining to every patient's medication order, our pharmacists reviewed orders to help identify errors prior to producing harmful results. We organized a group of research volunteers who shadowed our ED pharmacists over a variety of discrete 4-h time periods during their reviews of orders on patients in the ED. We documented total numbers of charts reviewed by pharmacists during their shifts along with the errors identified in those charts. We categorized these errors according to the National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP) Index for Categorizing Medication Errors Guidelines: Category A: Circumstances or events that have the capacity to cause error, category B: An error occurred but the error did not reach the patient (An “error of omission” does reach the patient), category C: An error occurred that reached the patient but did not cause patient harm, category D: An error occurred that reached the patient and required monitoring to confirm that it resulted in no harm to the patient and/or required intervention to preclude harm, category E: An error occurred that may have contributed to or resulted in temporary harm to the patient and required intervention, category F: An error occurred that may have contributed to or resulted in temporary harm to the patient and required initial or prolonged hospitalization, category G: An error occurred that may have contributed to or resulted in permanent patient harm, category H: An error occurred that required intervention necessary to sustain life, category I: An error occurred that may have contributed to or resulted in the patient's death [Table 1: NCC MERP error categories and Table 2: Error categories].

Table 1

National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP) error categories

Table 2

Error category

(Source: http://www.nccmerp.org/medErrorCatIndex.html)

We then measured the correlation between error rate (number of errors divided by the total number of charts reviewed) and ED occupancy rate while controlling for diversion status during the 6-month observation period from May 2009 to October 2009. We estimated a sample size requirement of at least 45 errors identified to allow detection of an effect size of 0.6 based on our historical data.

RESULTS

Pharmacists and research volunteers reviewed a total of 1,171 charts. Total documented error surveillance was comprised of 81 discrete observation periods, each 4 h in duration (324 h total). We identified a total of 87 errors in these 81 observation periods. For each 4-h observation period, average numbers of charts reviewed were 14.4 [Figure 1, Table 3: Percentage of errors per charts reviewed]. Median error rate per 4-h block was 5.8% of charts reviewed (interquartile range (IQR) 0-13) [Figure 2]. Errors occurred throughout the spectrum of care. Errors included; (a) incorrect dosage, (b) incorrect administration routes (e.g., intravenous (IV) medications ordered for intramuscular (IM)), (c) medication duplications, and (d) delays in therapy. No significant change was seen with the ED occupancy rate (Spearman's rho (ρ) = – 0.08, P = 0.49) [Figure 3], whereas previously error frequency showed a positive correlation with daily average EDWIN score (Spearman's ρ = 0.33; P = 0.001).[15] The Median error rate during times on ambulance diversion was almost twice as large (11%, IQR 0-17) [Figure 4], but this rate did not reach statistical significance in univariate or multivariate analysis.

Figure 1

Histogram showing the occurrence of errors during each observation period. Most observations periods yielded no errors, whereas the largest number found in one observation period was eight

Table 3

Percentage of errors per charts reviewed

Figure 2

Showing median error rate per 4-h block was 5.8% of charts reviewed (IQR: 0-13)

Figure 3

Showing no significant change was seen with the emergency department (ED) occupancy rate (Spearman's rho = –0.08, P = 0.49)

Figure 4

Showing median error rate during times on ambulance diversion was almost twice as large (11%, IQR: 0-17)

Limitations

Since we collected data from only one ED, our results may not be generalizable to all EDs. Our pharmacists are experienced and have specific skills and knowledge to perform their function in a busy ED environment. Changes in daily census may impact findings, with an increased inflow of patients possibly resulting in a higher number of medication and therapy orders, which may hinder pharmacists’ availability to review charts for errors and accurately document the results. Our ED has employed pharmacists for two full-time shifts, morning (7:00 am-3:00 pm) and evening (3:00 pm-11:00 pm), but the ED at the time of this study was devoid of pharmacists’ coverage from 11:00 pm to 7:00 am. This lack of coverage limits extrapolation of our data beyond hours when pharmacists were present.

DISCUSSION

Previously we determined a positive correlation between medication error frequency and ED occupancy as we used the EDWIN score as a point of reference for ED occupancy and medication error frequency measurement. We grouped EDWIN score into low, medium, and high crowding days; and determined that the error frequency was significantly increased in the high crowding group.[15] To further characterize this crowding influence on medication error, we reviewed patients’ charts for medication orders. After reviewing 1,171 charts and 324 h of error surveillance; we found that there was not a disproportionate increase in medication errors as ED occupancy increased [Figure 3], rather there is a linear association between crowding and medication error. This may be due to at least two factors: First, our department may be getting better at reducing the number of errors made when crowding conditions increase, which in turn would result in a reduction in the number of errors available to be identified by the pharmacists. Second, increases in errors seen with increased crowding may be occurring in a nonlinear fashion that is not detected by our sample size. Nevertheless, the fact that errors demonstrably increase as crowding increases raises significant concerns, and suggests that greater vigilance is warranted as crowding conditions occur.

The differences in the findings of the current study from our previous work are likely a result of the more granular measurement process employed, combined with the fact that in this study, we controlled for error number by factoring in the number of charts analyzed; whereas in earlier work, we did not. Performance improvement at our institution may additionally be a factor; however temporal changes in these processes are more difficult to quantify.

In 2007, there were about 117 million ED visits in the United States, about one-fifth of ED visits by children younger than 15 years of age were to pediatric EDs, there were 121 ED visits for asthma per 10,000 children under 5 years of age.[17] From 1997 and 2007, total annual visits to US EDs increased from an estimated 94.9 million to an estimated 116.8 million, an increase of 23.1%, with this increase being almost double what would be expected from population growth during this period.[18] Regardless of the causes of crowding; the negative results include delays in treatment,[19] decreased quality of care for patients,[20] and medication errors[21] among others.[2223242526]

Medication error includes components of administrative mistakes, physician focus disruption, and perhaps miscommunication; but it does not necessarily mean direct harm to patients and their care. MEDMARX (the anonymous national database for reporting medication errors) shows 105,603 medication errors documented, with 2,063 (2%) of total errors occurring in the ED. Although most of these were corrected before causing harm to the patient, 147 (7%) resulted in patient harm. Of those cases; 123 resulted in temporary harm to the patient and required intervention, 21 resulted in admission to hospital, one may have contributed to or resulted in permanent harm, another required lifesaving intervention, and one resulted in a patient's death.[27]

In our institution, most of these errors are identified with the help of pharmacists as they review orders before dispensing. A study to determine the frequency of medication errors in one facility's ED before and after an ED pharmacist was assigned to check medication orders found that the rate of errors decreased significantly (66.6%) when pharmacists prospectively reviewed the orders.[28] Pharmacists have assisted ED staff with drug selection, drug administration, and patient monitoring; as well as with emergency and trauma-related codes.[2930] Nurses and physician assistance staff can also provide another filter before uneventful effects of medication errors. With the help of computer technology and medication databases, that may eliminate most obvious errors with the indications of harmful medication effects.[31]

CONCLUSIONS

The error frequency in our ED appears to remain relatively constant across the range of crowding when controlling for patient volume via the quantity of orders reviewed. Error quantity therefore increases with crowding, but not at a rate greater than the expected baseline error rate that occurs in uncrowded conditions. These findings nevertheless suggest that crowding may result in an increase in the error quantity in a linear fashion.