Printer Friendly

Nurse staffing is an important strategy to prevent medication errors in community hospitals.

THE CHANGING REIMBURSEment system with increased emphasis on quality outcomes tied to payment has elevated the importance of the role of nurses in patient care. The delivery of medications is a highrisk activity involving many individuals--physicians, pharmacists, and nurses--and errors may be the result of failures at a variety of steps in the process. A number of strategies have been implemented to improve the safety of medication prescribing, transcribing, dispensing, and administering. These include computerized physician order entry, medication reconciliation, automated medication dispensing systems, bar code administration systems, and smart pumps (Elias & Moss, 2011; Jayawardena et al., 2007; Richardson, Bromirski, & Hayden, 2012). Yet the incidence of medication errors remains an issue. It is widely believed the occurrence of medication errors far exceeds the reported errors (Carlton & Blegen, 2006; Ulanimo, O'Leary-Kelley, & Connolly, 2007). Barker, Flynn, Pepper, Bates, and Mikeal (2002) found that 19% of the doses administered in 36 institutions were in error with 7% regarded as potentially harmful. Thus, the impact of medication errors on patient quality and safety is an ongoing concern, and research into causes of errors and strategies for improvement is needed.

Although medication safety is a concern of all health care professionals, the nurse is a key clinician in the process and is most likely the final barrier between the patient and an error. Carlton and Blegen (2006) noted medication errors occur due to active failures and latent conditions. Active failures in dosage calculation, following protocols, and lack of pharmacology knowledge are critical for nurse leaders to address, but just as important are latent conditions that contribute to errors.

Latent conditions include inadequate staffing, time pressures, unit environment, and fatigue. For the nurse, administering medications is only one of many duties and is often fraught with interruptions and the possibility of failures at many points. Elganzouri, Standish, and Androwich (2009) found that every medication pass was interrupted because of other staff members, missing medications, or other patient care needs. Kalisch, Landstrom, and Williams (2009) surveyed nearly 500 nurses in three hospitals and found that 85% of missed care was attributed to too few labor resources (nurse staffing). They found when too few nurses are available to meet the needs of patients, nurses may omit steps, take shortcuts, or deviate from approved standards in order to get the work done. Popescu, Currey, Gert, and Botti (2011) observed medication passes and interviewed nurses to identify factors influencing medication safety. Their study had similar findings; they observed nurses who were distracted or who had many competing demands, omitted some of the 5 Rights, failed to follow best practices guidelines, or omitted other care activities to give medications. All of these compensatory behaviors potentially could result in an error, and there should be an examination of the adequacy of staffing as a latent factor influencing medication errors.

The purpose of our study was to examine the relationship between nurse staffing and the occurrence of medication errors on medical-surgical units. Nurse staffing has been studied as an important influence on the occurrence of medication errors, but more research is needed to identify the most effective staffing levels to achieve desired patient outcomes and avoid errors (Schmalenberg & Kramer, 2009).

Literature Review

Several early studies of nurse staffing showed the relationship of staffing on the occurrence of medication errors. Blegen and Vaughn (1998) examined 11 hospitals over a 10 quarter period using multiple regression. They found a significant inverse relationship between the proportion of registered nurses (RNs) in the skill mix and medication errors and falls; as the proportion of RNs increased, the medication errors decreased. The effect ended when the proportion reached 85%. In a study of 95 patient care units in 10 hospitals, Whitman, Kim, Davidson, Wolf, and Wang (2002) found medication errors were higher in cardiac care units and non-cardiac intermediate units when staffing levels were lower. Both Pearson parametric and Spearman's nonparametric correlational methods were used to analyze the relationships between staffing hours and medication errors. Although the data were hierarchical (e.g., hospitals measured over time or units within hospitals), no hierarchical or nested models were used in either of these studies. Hall, Doran, and Pink (2004) examined nurse staffing and patient outcomes with data collected at the unit level. The effects of the unit characteristics, including the mix of nursing staff on medication errors, were examined using multiple regression analysis. Hierarchical linear modeling was used to analyze the effects of patient complexity and age in the first level, and that of the unit variables on the individual patient care hours and costs, but not on medication errors. A lower proportion of RNs and licensed practical nurses (LPNs) in the skill mix on the unit was associated with a higher number of medication errors and wound infections.

More recent studies continue to show relationships between staffing and medication errors. In a retrospective study of 10,187 elderly patients, Picone et al. (2008) used generalized estimating equations methods and found that for every 20% decrease in staffing below the lowest RN average hourly time spent with a patient, the odds of a medication error increased by 18%. Patrician et al. (2011) used data from 13 military hospitals to examine associations between nurse staffing and patient outcomes and medication errors. The probability that a medication error occurred in an 8-hour shift was modeled using Bayesian hierarchical logistic regression with the nesting of shifts within days within units. Patient factors were not controlled in this study. Findings indicated a higher number of total nursing care hours per shift was associated significantly with medication adverse events. Breckenridge-Sproat, Johantgen, and Patrician (2012) found an increase in LPNs in the skill mix in intensive care units was associated with a higher number of medication errors in military hospitals. This finding was not present in medical-surgical units. Their study included staffing variables as well as adverse events, including medication errors, which were aggregated from shifts of each unit to the month level for the unit. Thus, a nested model, negative binomial generalized linear mixed regression model, was applied.

Methods

Design and data. We conducted a retrospective, correlational study in collaboration with a system of community hospitals to examine factors associated with medication errors. Medication error records were collected from 24 medical-surgical units in 8 hospitals. Some units were removed from analysis due to low numbers of medication errors reported, and others were removed due to insufficient data for hierarchical linear modeling. As a result, nine units, all from a single hospital, met criteria for this analysis. Although the hospital had a variety of units where medication errors occurred, (pharmacy, emergency department, intensive care units, radiology, surgery, and ambulatory services), only medication errors that occurred in the medical-surgical nursing units were included in the analysis because nurse staffing was the variable of interest and the staffing practices were similar in these areas. Table 1 provides descriptive statistics on the nine units in the final sample. One unit was designated as surgery, one as oncology, four as medical, and three as medical-surgical. The mean number of discharges per week was 39 (SD=9.2). All of the units used Pyxis[R] to assist in medication dispensing; however, none of the units used bar code technology or physician order entry during the study period.

Database administrators extracted all staffing and patient data from July 2008 to June 2010 (2 years of data) associated with the nine units. An incident-reporting database contained information for medication errors using the National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP, 2005) taxonomy (error types A-I) for the time period. Only medication errors that occurred in the units of interest were included. We queried an enterprise-wide data warehouse for patient and unit characteristics. We matched weekly staffing data (for nine units over 2 years there are a total of 936 weekly staffing intervals) from a proprietary payroll system to the reed ication error and patient information. The staffing data included nursing hours per equivalent patient day (HPEqPD included inpatient and observation patients on a unit) for direct care providers during productive hours; all of these were identified by provider type. After matching weekly staffing data with medication errors and patient information, our final sample consisted of 31,080 patient observations within 801 weekly staffing intervals (135 weekly staffing intervals were removed due to missing data).

Statistical analysis. The study consists of data from patients (MSDRG weight, age, gender, and race) and weekly staffing units (HPEqPD). Because patients are assigned typically on nursing units by diagnoses, and the same group of nurses provides care for patients on particular units, the outcomes can be influenced by unit assignment. Thus, the patients' outcomes are not completely independent of each other making the use of ordinary least squares (OLS) regression technique inappropriate because the assumption of independence required for OLS would be violated. Hierarchical linear modeling (HLM) is a statistical analytic technique developed to appropriately deal with such situations where the independence assumption is violated (Bloom & Milkovich, 1998; Raudenbush & Bryk, 2002), and the analytic approach has been used in past nursing research (e.g., Breckenridge-Sproat et al., 2012; Patrician et al., 2011). Based on the nature of the data and the purpose of the study, we used HLM as our statistical method for analysis. Because medication errors are measured dichotomously (where 1 represents the occurrence of a medication error and 0 otherwise), we used the Logit model in the tests for medication errors.

Few prior nurse staffing studies on medication errors using both patient and weekly staffing unit data have accounted correctly for the lack of independence described above. Although both Patrician and colleagues (2011) and Breckenridge-Sproat and coauthors (2012) used a HLM method, neither study was conducted using individual patient data. Thus, this research provides a methodological contribution to the nurse staffing and medication error literature through a method appropriate for the data and through the inclusion of separate RN and LPN staffing variables in a single model, allowing study of the effects of patient factors, and that of nurse staffing on the medication errors for individual patients.

Findings

Descriptive findings. The final data from the nine nursing units showed that 335 medication errors occurred among the 31,080 patients; this represents roughly 1% of discharged patients in the 2-year period. The mean age of patients was 66.72 (SD=15). There was a near even split between males (47%) and females (53%). The mean MS-DRG weight was 1.83 (SD=1.42) (see Table 2). The vast majority of patients were White (88%), thus all non-White races were combined for the analysis. The number of RN HPEqPD was consistent among the nine units with a mean of 6.61 (SD=0.42), and the use of LPN HPEqPD was low with a mean of 0.15 (SD=0.21).

There were three major activities that accounted for 97% of errors: administering medications (58%), transcribing orders (22%), and dispensing medications (17%) (see Table 3). Forty-four percent were errors that reached the patient, but no harm was done; 14% of medication errors reached the patient and required monitoring and/or treatment. There were no medication errors that resulted in permanent harm or death (see Table 4). The most common errors were omission of medications, failure to follow protocols, improper dosage, and wrong patient across all of the categories. Table 5 shows the number, percentage, and frequency order of these errors. All other errors were summed because there were many different errors with counts less than 10.

Hierarchical linear modeling results. Table 6 shows the results of the HLM tests. As shown in the table, both models are significant and model fit is reported in the form of a likelihood ratio test comparing models 1 and 2 to the null model, the model with no predictors (Raudenbush & Bryk, 2002). Model 1 includes the control variables in the test of nurse staffing on medication errors and the model shows that age, gender, and MS-DRG weight are significant. The results show patients with higher MS-DRG weights are more likely to face a medication error and males are less likely to encounter a medication error. It is interesting to note older patients face fewer medication errors; however, the effect is very small. Model 2 includes the main effects of RN HPEqPD and LPN HPEqPD, and the results show higher levels of RN HPEqPD result in fewer medication errors (for 1 hour increase in RN HPEqPD from the mean, the probability of medication error decreases by 0.16%); however, higher levels of LPN HPEqPD result in more medication errors (for 1 hour increase in LPN HPEqPD from the mean, the probability of medication error increases by 3% and for one-half hour increase in LPN HPEqPD from the mean the probability of medication error increases by 1.21%; note the maximum LPN hours in data set is 0.75).

Discussion

The patients in our study were most likely to be over 65 and White. There were slightly more females than males in the study. Patients who had a higher MSDRG weight and were female and non-White were the most likely to experience a medication error. This finding was consistent with Thompson-Moore and Liebl (2012) and with Picone and colleagues (2008). However, our finding about the inverse relationship between patient age and medication errors has not been reported in other studies.

The number of HPEqPD in our study was similar to those reported by Kane, Shamliyan, Mueller, Duvall, and Wilt (2007) in a meta-analysis where surgical patients received a mean of 8.1 HPPD and medical patients received a mean of 6.1 HPPD. The mean RN HPEqPD in our study was 6.61 with a minimum to maximum of 5.35 to 8.21 hours for the sample.

The findings demonstrated the most common errors occurred in administering medications. This is logical because we included errors that occurred only in nursing units of interest, not pharmacy, surgery, or other areas. Because nearly 75% of the errors were either near misses or errors that reached the patient with no harm, little additional treatment was necessary. Neither permanent harm nor deaths resulted from the medication errors in this data set. In a review of literature of observed medication errors, Kiekkas, Karga, Lemonidou, Aretha, and Karanikolas (2011) found the most common consequence of medication errors was the need to monitor the patient. This finding was not true for our study; only 14% of medication errors required additional monitoring or treatment following the error. We found the most common errors were due to dose omission, failure to follow protocols, and improper dose. This finding is consistent with Barker and co-authors (2002) and Kiekkas and colleagues (2011). Kiekkas et al. (2011) postulated dose omissions were more likely to be system issues such as work overload, and wrong doses were more likely to be related to knowledge deficits of health care providers. The concept of missed care by Kalisch and colleagues (2009) does support medication omissions as being related to human resource allocation (nurse staffing).

Our findings indicate nurse staffing is an important human resource to keep patients safe from medication errors. As the RN HPEqPD increased, the medication errors decreased; conversely, as the LPN HPEqPD increased, the medication errors increased. Administering medications to hospitalized patients is not a simple task; it requires thorough knowledge of every medication that is administered to a patient. This includes pharmacodynamics, therapeutic use, side effects, adverse events, and appropriateness of administering the medication given the patient's current response to treatment. The patient's history, allergies, primary diagnoses and comorbidities, and treatment plan are also relevant to medication administration (NCC MERP, 2005). Because of the advanced knowledge of pharmacology, patient condition, and treatment plan required for safe administration of medications, the work is more appropriately assigned to the most highly educated nurses--RNs rather than LPNs. Research bears this out. For example, Elganzouri and colleagues (2009) conducted a time and motion study of medication administration and found LPNs were noted to frequently ask RNs medication questions during the medication pass. While it is desirable for LPNs to seek information, the cognitive work of medication administration may not be appropriate for the educational preparation of LPNs, and limitations in scope of practice for LPNs may require RNs to make calls to physicians to clarify orders. In a fastpaced hospital environment, having uninterrupted time for these cognitive processes is almost impossible. As demonstrated by Elganzouri and colleagues (2009), the information seeking behavior of LPNs also caused disruptions during the medication pass of the RNs.

Implications for Nursing Administration

The costs associated with medication errors are difficult to estimate, but a few sources cite statistics that show the magnitude of unnecessary expenses. For example, the 1999 Institute of Medicine study To Err is Human estimated the total medical costs for treating inpatient preventable adverse medication events at $2 billion annually. Bates and colleagues (1997) estimated the costs per adverse medication event for hospitalized patients as $5,857. After inflation is calculated, this cost would now be $9,580 (Bureau of Labor Statistics, n.d.). For the hospital included in the study, there were 335 errors with 14% (47) of the errors requiring additional monitoring and treatment. The cost of the errors multiplied by the number of errors ($9,580 x 47) produces a cost estimate of $450,260. The estimate only includes costs for errors that occurred on the nursing unit and not in other areas of the hospital, so costs for the entire hospital would be expected to be much higher. Any one error has the potential to cause permanent harm or death, which could result in litigation and high costs associated with defending a claim.

Several human resource options can be used to reduce the incidence of medication errors such as increasing the number of RN hours per patient day in all nursing units, increasing RN hours for nursing units in the lower 50%, or substituting RN hours for LPN hours. Across the range of staffing in this study (5.34 to 8.21 RN HPEqPD), a change in staffing from the lowest to highest resulted in reduction of errors of 0.5% for a typical patient, and for LPN hours across the range, decreasing LPN hours from highest to lowest would reduce errors by 1.7%.

Nurses intuitively know higher levels of RN staffing result in better patient care, and current research is supportive (Chang & Mark, 2009; Hall et el., 2004; Picone et el., 2008). However, research has not established a definitive relationship between RN staffing and medication errors, and the optimum level of RN staffing has not been identified. This study, although limited to one hospital, provides several strategies that can be used to increase medication safety in medical-surgical nursing units. When faced with decisions about reducing costs, nurse administrators must evaluate the potential impact of reducing the RN hours of care on patient outcomes including medication errors. The increased cost of higher RN hours in the staffing can be compared to the savings of preventing a medication error.

Conclusion

The incidence and cost of medication errors continues to be a problem requiring solutions. A number of technology strategies have been implemented to decrease the number of medication errors including computerized physician order entry, automated medication administration records, and bar coding administration; but even with the use of these technologies, errors continue to occur (Ulanimo et el., 2007). Health care leaders need to consider not only technology capital investments but also human capital as a strategy to keep patients safe. This study examined the relationship between nurse staffing and medication errors. Findings indicate even a small number of LPNs in staffing can contribute to medication errors. Even though using LPNs reduces payroll expenses, the safety of patients could be affected. This study adds to the body of evidence that patient care is most safely delivered when there are enough RN care hours and when LPN hours are reduced or eliminated. The cost associated with RN hours must be balanced against the cost of an error.

REFERENCES

Barker, K.N., Flynn, E.A., Pepper, G.A., Bates, D.W., & Mikeal, R.L. (2002). Medication errors observed in 36 health care facilities. Archives of Internal Medicine, 162(16), 1897-1903.

Bates, D.W., Spell, N., Cullen, D.J., Burdick, E., Larid, N., Peterson, L.A. ... Leape, L.L. (1997). The costs of adverse drug events in hospitalized patients. Adverse Drug Events Prevention Study Group. JAMA, 277(4), 307.

Blegen, M., & Vaughn, T. (1998). A multisite study of nurse staffing and patient occurrences. Nursing EconomicS, 16(4), 196-203.

Bloom, M., & Milkovich, G. (1998). Relationships among risk, incentive pay, and organizational performance. Academy of Management Journal, 41, 283-297.

Breckenridge-Sproat, S., Johantgen, M., & Patrician, P. (2012). Influence of unitlevel staffing on medication errors and falls in military hospitals. Western Journal of Nursing Research, 34(4), 455-474. doi:10.1177/0193945911407090.

Bureau of Labor Statistics (n.d.). CPI inflaction calculator. Retrieved from http://www.bls.gov/data/inflation_calculator.htm

Carlton, G., & Blegen, M. (2006). Medicationrelated errors: A literature review of incidence and antecedents. Annual Review of Nursing Research, 24, 19-38.

Chang, Y., & Mark, B. (2009). Antecedents of severe and nonsevere medication errors. Journal of Nursing Scholarship, 41(1), 70-78.

Elganzouri, E.S., Standish, C.A., & Androwich, I. (2009). Medication Administration Time Study (MATS) nursing staff performance of medication administration. Journal of Nursing Administration, 39(5), 204-210.

Elias, B.L., & Moss, J.A. (2011). Smart pump technology: What we have learned. CIN: Computers, Informatics, Nursing, 29(3), 184-190. doi:10.1097/NCN.0b013e3181fcbe6d

Hall, L. M., Doran, D., & Pink, G. H. (2004). Nurse staffing models, nursing hours, and patient safety outcomes. Journal of Nursing Administration, 34(1), 41-45.

Institute of Medicine. (1999). To err is human: Building a safer health system. Washington, DC: National Academies Press.

Jayawardena, S., Eisdorfer, J., Indulkar, S., Pal, S.A., Sooriabalan, D., & Cucco, R. (2007). Prescription errors and the impact of computerized prescription order entry system in a community-based hospital. American Journal of Therapeutics, 14(4), 336-340.

Kalisch, B., Landstrom, G., & Williams, R. (2009). Missed nursing care: Errors of omission. Nursing Outlook, 57(1), 3-9.

Kane, R., Shamliyan, T., Mueller, C., Duval, S., & Wilt, T. (2007). The Association of Registered Nurse staffing levels and patient outcomes: Systematic review and meta-analysis. Medical Care, 45(12), 1195-1204.

Kiekkas, P., Karga, M., Lemonidou, C., Aretha, D., & Karanikolas, M. (2011). Medication errors in critically ill adults: A review of direct observation evidence. American Journal of Critical Care, 20(1), 36-44. doi:10.4037/ajcc2011331

National Coordinating Councilfor Medication Error Reporting and Prevention [NCC MERP]. (2005). About NCC MERP. Retrieved from http://www.nccmerp.org/aboutNCCMERP.html

Patrician, P.A., Loan, L., McCarthy, M., Fridman, M., Donaldson, N., Bingham, M., & Brosch, L.R. (2011). The association of shift-level nurse staffing with adverse patient events. Journal of Nursing Administration, 41, 64-70. doi:10.1097/NNA.0b013e31820594bf

Picone, D.M., Titler, M.G., Dochterman, J., Shever, L., Kim, T., Abramowitz, P., ... Rui, Q. (2008). Predictors of medication errors among elderly hospitalized patients. American Journal of Medical Quality, 23(2), 115-127.

Popescu, A., Currey, J., Gert, G., & Botti, M. (2011, First Quarter). Multifactorial influences on and deviations from medication administration safety and quality in the acute medical/surgical context. Worldviews on EvidenceBased Nursing: Linking Evidence to Action, pp. 15-24. doi:10.1111/j.1741-6787.2010.00212.x

Raudenbush, S., & Bryk, A. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc.

Richardson, B., Bromirski, B., & Hayden, A. (2012). Implementing a safe and reliable process for medication administration. Clinical Nurse Specialist: The Journal for Advanced Nursing Practice, 26(3), 169-176.

Schmalenberg, C., & Kramer, M. (2009). Perception of adequacy of staffing. Critical Care Nurse, 29(5), 65-71. doi:10.4037/ccn2009324

Thompson-Moore, N., & Liebl, M.G. (2012). Health care system vulnerabilities: Understanding the root causes of patient harm. American Journal of Health-System Pharmacy, 69(5), 431-436. doi:10.2146/ajhp110299

Ulanimo, V., O'Leary-Kelley, C., & Connolly, P. (2007). Nurses' perceptions of causes of medication errors and barriers to reporting. Journal of Nursing Care Quality, 22(1), 28-33.

Whitman, G.R., Kim, Y., Davidson, L.J., Wolf, G.A., & Wang, S. (2002). The impact of staffing on patient outcomes across specialty units. Journal of Nursing Administration, 32(12), 633-639.

Executive Summary

* The prevalence of medication administration errors continues to be a problem requiring the attention of nurse leaders.

* In this study the relationship between nurse staffing and the occurrence of medication errors was examined.

* Using a retrospective design, researchers analyzed secondary data from administrative databases of one hospital containing 801 weekly staffing intervals and 31,080 patient observations.

* The current study shows that increasing the number of RN hours and decreasing or eliminating LPN hours can be a strategy to reduce medication errors.

KAREN H. FRITH, PhD, RN, NEA-BC, is a Professor of Nursing, College of Nursing, University of Alabama in Huntsville, Huntsville, AL.

E. FAYE ANDERSON, DNS, RN, NEA-BC, is an Associate Professor of Nursing, College of Nursing, University of Alabama in Huntsville, Huntsville, AL.

FAN TSENG, PhD, is a Professor of Management Science, College of Business Administration, University of Alabama in Huntsville, Huntsville, AL.

ERIC A. FONG, PhD, is an Associate Professor of Management, College of Business Administration, University of Alabama in Huntsville, Huntsville, AL.
Table 1.

Descriptive Statistics of the Units

Unit                      Discharges/Week   Medication Error/Patient

Medical Unit 1                 43.7                  0.010
Medical Unit 2                 30.9                  0.0142
Medical-Surgical Unit 1        32.3                  0.0236
Medical Unit 3                 47.8                  0.0007
Medical Unit 4                 43.0                  0.0065
Surgery Unit                   53.2                  0.0135
Oncology Unit                  23.3                  0.0106
Medical-Surgical Unit 2        38.2                  0.0109
Medical-Surgical Unit 3        37.2                  0.0112

Table 2.
Descriptive Statistics on Patient and Staffing Variables

Variable        Number   Mean    SD      Minimum   Maximum

Patient Age     31,080   66.72   15.61     12.00    106.00
MS-DRG Weight   31,080    1.83    1.42      0.39     18.37
RN HPEgPD          801    6.61    0.42      5.34      8.21
LPN HPEgPD         801    0.15    0.21      0.00      0.75

Table 3.
Percentage of Modes of Committing an Error

Mode of Error                 Percentage

Administering                   58.0%
Documenting or Transcribing    722.0%
Dispensing                      17.0%
Monitoring                       1.5%
Patient Compliance               0
Prescribing                      1.6%
Total                          100.0%

Table 4.
Percentage of Types of Medication Errors

Type of Error                                           Percent

A--No error, capacity to cause error                     14%
B--Error, no harm, did not reach patient                 28%
C--Error, no harm, reached the patient                   44%
D--Error, no harm, reached the patient and required      10%
monitoring/intervention
E--Error, harm temporary, intervention required           4%
F--Error, harm, temporary required hospitalization         0
G--Error harm, contributed to permanent harm               0
H--Error, harm required to intervene to sustain life       0
I--Error, death                                            0
Total                                                   100%

Table 5.
Error Typology

                               Protocol Not
              Dose Omission      Followed       Improper Dose
Typology     Rank/N/Percent   Rank/N/Percent   Rank/N/Percent

Category A   1   51 (38%)     2   26 (20%)     3    15 (11%)
Category B   1   95 (36%)     2   46 (18%)     3    28 (11%)
Category C   2   84 (35%)     3   34 (14%)     1   102 (42%)
Category D   1   31 (33%)     4    6 (6%)      2    28 (29%)
Category E   1   15 (42%)     4    0 (0)       2    11 (31%)

             Wrong Patient    All others     Total
Typology     Rank/N/Percent      N/%          N/%

Category A    4    6 (5%)      35 (26%)    133 (100%)
Category B    4   11 (4%)      82 (31%)    262 (100%)
Category C    4   21 (9%)     158 (66%)    241 (100%)
Category D    3    8 (8%)      23 (24%)     96 (100%)
Category E    3    2 (6%)       8 (21%)     36 (100%)

Table 6.
HLM Results for the Effects of Nurse Staffing on Medication Errorse (a)

Variables      Model 1           Model 2

Constant       -2.46 ** (0.04)   -1.70 ** (0.19)
Age            -0.02 ** (0.00)   -0.02 ** (0.00)
Gender         -0.15 ** (0.02)   -0.17 ** (0.02)
Race           -0.01 (0.03)      -0.07 ** (0.03)
MSDRG Weight    0.13 ** (0.00)    0.10 ** (0.01)
RN HPEgPD                        -0.07 * (0.03)
LPN HPEgPD                        0.85 ** (0.07)
[chi square]   60.98 **          59.59 **

(a) Robust standard errors are shown. * p<0.05; ** p<0.01, two-tailed.
COPYRIGHT 2012 Jannetti Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Author:Frith, Karen H.; Anderson, E. Faye; Tseng, Fan; Fong, Eric A.
Publication:Nursing Economics
Geographic Code:1USA
Date:Sep 1, 2012
Words:4684
Previous Article:Attending registered nurse: an innovative role to manage between the spaces.
Next Article:Various factors influence the adoption of ergonomics in health care organizations.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters