Research paper “Prevention of High-Alert Medication Errors in Hospital Patients”

Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” Research paper based on two articles attached below plus one more original research article Paper includes: 1. Introduction (paragraph of four sentences) see photo with instruction 2. Body of text (must contain three fully developed paragraphs, each discussing a supporting point in depth. Each paragraph must be well thought-out and follow a logical development) 2 pages Paper includes three level 1 headings. Paper is well organized. Paper includes two direct quotes, one personal communication, one in-text paraphrase, and one parenthetical paraphrase cited properly. Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” 3. Conclusion Conclusion offers a good summary of the main points of the paper. 4. Reference page img_5948.jpg high_alert_med_incidents.pdf preventing_high_alert_medication_errors_in_hospital_patients.pdf ORDER NOW FOR CUSTOMIZED AND ORIGINAL ESSAY PAPERS International Journal for Quality in Health Care 2014; Volume 26, Number 3: pp. 308–320 Advance Access Publication: 25 April 2014 10.1093/intqhc/mzu037 Effects of patient-, environmentand medication-related factors on high-alert medication incidents ELIZABETH MANIAS1,2*, ALLISON WILLIAMS3, DANNY LIEW4, SASCHA RIXON5, SANDY BRAAF5 AND SUE FINCH6 1 School of Nursing and Midwifery, Deakin University, Burwood, Victoria, Australia, 2Department of Medicine, Royal Melbourne Hospital, The University of Melbourne, Carlton Victoria, 3School of Nursing and Midwifery, Monash University, Clayton, Victoria, Australia, 4 Collaborative Centre for Clinical Epidemiology, Biostatistics and Health Services Research, Royal Melbourne Hospital, Parkville, Victoria, Australia, 5Melbourne School of Health Sciences, The University of Melbourne, Carlton, Victoria, Australia, and 6Statistical Consulting Centre, The University of Melbourne, Carlton, Victoria, Australia *Address reprint requests to: Elizabeth Manias, School of Nursing and Midwifery, Deakin University, 221 Burwood Highway Burwood 3125 Victoria Australia. Tel: + 61 3 9244 6958; Fax: + 61 3 8344 5391; E mail: [email protected] Accepted for publication 16 March 2014 Abstract Objective. To measure the rate of medication incidents associated with the prescription and administration of high-alert medications and to identify patient-, environment- and medication-related factors associated with these incidents. Design. A retrospective chart audit design was conducted of medical records for patient admissions from 1 January 2010 to 31 December 2010. Setting. Five practice settings (cardiac care, emergency care, intensive care, oncology care and perioperative care) at a public teaching hospital in Melbourne, Australia. Participants. Patients were considered for inclusion if they were prescribed at least one high-alert medication and if they were admitted to one of ?ve practice settings. Main Outcome Measures. High-alert prescribing and administering incidents were measured in each of the ?ve practice settings. Generalized linear mixed modeling was used for data analysis. Results. There were 6984 opportunities for high-alert medication incidents across the ?ve clinical settings. The overall medication incident rate was 1934/6984 (27.69%). There were 1176 prescribing incidents (16.84%) and 758 administering incidents (10.85%). Statistical modeling showed that, in each of the ?ve clinical settings, an increased number of ward transfers was associated with increased odds of prescribing incidents. In addition, statistical modeling demonstrated that an increased number of ward transfers was associated with increased odds of administering incidents in emergency care and perioperative care. Conclusions. Complex relationships were found in managing high-alert medications in specialty clinical settings. Employing measures to address patients’ movements across ward settings can reduce high-alert medication incidents and improve quality of care. Keywords: medication incident, medication therapy management, clinical audit, high-alert medication, hospitals Introduction High-alert medications cause many medication incidents. They also carry the highest possibility of devastating adverse outcomes [1, 2]. Examples of high-alert medications include opioids, electrolyte parenteral solutions, insulin, chemotherapy and anticoagulants. Past research has generally involved examining high-alert medications as a subset of all medications. Beckett et al. undertook a retrospective, case–control study of reported medication errors at three US hospitals [3], whereby high-alert medications was an independent predictor of patient harm (OR 4.00; 95% CI 2.38–6.75). Stavroudis et al. conducted a retrospective crosssectional study of 6749 neonatal intensive care medication error reports submitted by 163 US health care facilities [4]. High-alert medications were mentioned in 1487 (22.0%) reports, and were likely to cause harm. In these studies, reported errors were examined, which are a signi?cant underestimation of actual errors [5].Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” Only two studies have speci?cally assessed high-alert medication incidents. Using an observational design, Silva et al. International Journal for Quality in Health Care vol. 26 no. 3 © The Author 2014. Published by Oxford University Press in association with the International Society for Quality in Health Care; all rights reserved Downloaded from https://academic.oup.com/intqhc/article-abstract/26/3/308/2849751 by Nova Southeastern University user on 20 January 2018 308 Medication incidents • Safety examined high-alert medications in pediatric inpatients within a Brazilian hospital [6] and found prescribing errors in 89.6% (632/705) of high-alert medications. In Bahrain, Al Khaja et al. completed a retrospective, nationwide audit of oral hypoglycemic agents (n = 2773) [7], and found medication errors in 4.0 and 3.6% of high-dose metformin and high-dose glibenclamide prescriptions, respectively. This study aimed to measure the rate of prescription and administration incidents of high-alert medications and to identify patient-, environment- and medication-related factors associated with these incidents. Understanding these factors can help in developing strategies for preventing medication incidents. Methods Design This study utilized a retrospective chart audit design, conducted from 1 January 2010 to 31 December 2010 at a public, tertiary teaching hospital in Melbourne, Australia. The research team also accessed all reported medication incidents that occurred during this time. The hospital institutional ethics board approved conduct of the research. Data collection A strati?ed random sampling technique was undertaken. Medical records were considered for inclusion if patients were prescribed at least one high-alert medication and if they were admitted to one of ?ve practice settings (cardiac care (CC), emergency care (EC), intensive care (IC), oncology care (OC) and perioperative care (PC)). These settings were selected because high-alert medications are commonly prescribed in these environments. High-alert medications were those listed by the Institute for Safe Medication Practices [8]. Medical records were excluded if a patient was included from a previous admission. A random-numbers table was used to select medical records to be accessed. Forty-?ve medical records were accessed each month, leading to 108 medical records from each setting to obtain a sample size of 540 records. In some cases, patients were in clinical settings for no more than 1 day, such as in EC. If patients were in an environment for longer than 1 day, data from the ?rst 3 days of their stay in that environment were documented. Records were scrutinized to locate patient-, environment- and medication-related factors that could in?uence the occurrence of high-alert medication incidents. Core patient-related factors were age, gender, number of chronic diseases, usual accommodation, discharge destination, support network, presence of an interpreter, concession card holder and documented allergies. Concession card holders were individuals receiving bene?ts from government insurance programs. Core environment-related factors were number of ward transfers, day of week and public holidays. Core medication-related factors were number of prescribed medications per person and number of high-alert medications per person. Prior clinician interviews identi?ed additional factors that should be examined because of their perceived value. Additional patient-related factors included triage category for EC and cancer type for OC. The Australasian Triage Scale, comprising ?ve categories, classi?es the urgency of emergency department presentations [9]. Additional environment-related factors included time from presentation to treatment for EC and length of stay in IC. Additional medication-related factors included route of administration for CC, route of administration for EC and type of medication for IC. Measures High-alert medication incidents were ‘any preventable event that may cause or lead to inappropriate medication use or patient harm’ [10]. The National Coordinating Council for Medication Error Reporting and Prevention (NCC-MERP) tool [10] guided data collection. Two research assistants with quali?cations in biomedicine and knowledge about pharmacology and medication safety received extensive training and completed data collection. They simultaneously collected data relating to the outcome variable, medication incidents and explanatory variables.Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” One author independently checked data entry accuracy and obtained 100% agreement in 50 randomly selected records. An opportunity for an incident was de?ned as any medication prescribed, any unordered or omitted medication and any dose given and any dose omitted. For each setting, prescribing and administering medication incidents were examined. Objective measurement of medication incidents were obtained by utilizing current, evidence-based guidelines [11–16] and the standard taxonomy provided by the NCC-MERP [10]. The research assistants examined each high-alert medication in terms of predetermined types of incidents identi?ed from the guidelines and taxonomy. Examples of predetermined prescribing incidents included incorrect dose, wrong dose frequency, wrong route of administration, presence of documented allergies or drug–drug interactions, wrong infusion rate or wrong technique. Examples of predetermined administering incidents included omitted or extra dose, wrong time (one hour on either side of the speci?ed time), wrong patient, wrong infusion rate, wrong route, wrong technique, dosageform error, wrong preparation, wrong technique or wrong route or site of administration. Data analysis The prevalence of a medication incident was estimated to be 34% of patients from previous work [17]. A sample size of 539 was required to produce a con?dence interval with ±4% precision (or 0.08) for the speci?ed width of the 95% con?dence level [18]. Thus, a total of 540 medical records were accessed. Statistical analysis was undertaken using GenStat (15th edition, Hemel, UK). Since a single patient could contribute multiple medication incidents in a given setting, statistical modeling needed to account for the potential clustering effect of the patient, as the outcome of interest was binary. Generalized linear mixed modeling was speci?ed as a random effect. The effects of explanatory variables in these models 309 Downloaded from https://academic.oup.com/intqhc/article-abstract/26/3/308/2849751 by Nova Southeastern University user on 20 January 2018 Manias et al. were quanti?ed using odds ratios. Given the complexity of modeling and the number of explanatory variables involved, modeling was undertaken in two stages. In the ?rst stage, separate models for patient-, medication- and environmentrelated factors were ?tted for each setting. In the second stage, for each setting, any explanatory factor in any of the three models ( patient, medication and environment), with P <0.3, was considered for a ?nal combined model. The factors age, gender and number of prescribed medications were included and retained in the combined model regardless of the P-value achieved, due to their perceived importance in health care [1, 17]. A backward step-wise approach was used to re?ne the combined model and remove some explanatory variables. If the proportion of prescribing or administering incidents for a particular setting showed an imbalance of occurrence or nonoccurrence of <15%, no statistical modeling was undertaken. Results Demographics and high-alert medication incident characteristics The mean (SD) age of patients was 64.06 (15.68) years. Data from 318 men and 222 women were included. Their median length of hospital stay was 6.3 days, ranging from 22 min in OC to 122 days in IC (Table 1). There were 6984 opportunities for high-alert medication incidents. The overall medication incident rate was 1934/6984 (27.69%), comprising 1176 prescribing incidents (16.84%) and 758 administering incidents (10.85%) (Table 2). Table 3 shows the characteristics of severity ratings. Explanation of predictor effects by logistic regression modeling Statistical modeling was undertaken of prescribing incidents in CC, IC and OC, and of prescribing and administering incidents in EC and PC (Table 4). The ?nal combined general linear mixed model for prescribing incidents in CC had six factors. Prescribing high-alert medications on weekends had 6.40 times greater odds of a prescribing incident compared with weekdays (95% CI 2.62, 15.64; P < 0.001). Increasing the number of ward transfers during a patient’s hospital stay increased the odds of a prescribing incident 4.45 times (95% CI 2.74, 7.22; P < 0.001). Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” Route of administration had reduced odds of 0.45 for the subcutaneous route (95% CI 0.28, 0.71; P < 0.001) compared with the intravenous route. For each additional prescribed medication, the odds of a prescribing incident decreased 0.93 times (95% CI 0.89, 0.97; P = 0.003). The combined model for prescribing incidents in EC had ?ve factors (Table 4). Women had 2.68 times greater odds in experiencing a prescribing incident (95% CI 1.22, 5.89; P = 0.016) compared with men. Increasing the number of ward transfers increased the odds of a prescribing incident 4.35 times (95%CI 1.91, 9.90; P < 0.001). The subcutaneous route had 0.28 times the odds of a prescribing incident compared with the intravenous route (95% CI 0.17, 0.46, P < 0.001). With 310 Downloaded from https://academic.oup.com/intqhc/article-abstract/26/3/308/2849751 by Nova Southeastern University user on 20 January 2018 respect to administering incidents in EC (Table 4), increasing the number of ward transfers increased the odds of an administering incident 3.18 times (95% CI 1.61, 6.27; P = 0.001). Route of administration showed that the subcutaneous route had 5.52 times the odds of an administering incident (95% CI 2.81, 10.85; P < 0.001) compared with the intravenous route. The combined model for prescribing incidents in IC had six explanatory factors (Table 4). Women had 3.05 times the odds of experiencing a prescribing incident (95% CI 1.02, 8.95; P = 0.045) compared with men. The odds of a prescribing incident increased 5.96 times for each additional ward transfer (95% CI 3.52, 9.99; P < 0.001). Lengths of stay between 8 and 21 days and over 21 days had reduced odds of 0.11 (95% CI 0.03, 0.46; P = 0.002) and 0.15 (95% CI 0.03, 0.67; P = 0.014), respectively, of a prescribing incident compared with stays of a week or less. For each additional prescribed medication, the odds of a prescribing incident reduced 0.95 times (95% CI 0.92, 0.99; P = 0.007). Cardiovascular medications showed 2.43 times the odds of a prescribing incident (95% CI 1.53, 3.83; P < 0.001) while endocrine medications had 2.82 times the odds of a prescribing incident (95% CI 1.52, 5.16; P = 0.001) relative to the baseline. The combined model for prescribing incidents in OC had seven explanatory factors (Table 4). The odds of a prescribing incident were 0.22 (95% CI 0.06, 0.75; P = 0.016) for four or more chronic diseases compared with fewer than four. Increasing the number of ward transfers increased the odds of a prescribing incident 87.1 times (95% CI 7.37, 1030; P < 0.001). For each additional prescribed medication, there were 1.67 greater odds of a prescribing incident (95% CI 1.16, 2.39; P = 0.005), while for each additional prescribed high-alert medication, there were 0.48 times decreased odds of a prescribing incident (95% CI 0.24, 0.96; P = 0.037). The combined model for prescribing incidents in PC included six explanatory factors (Table 4). The presence of a documented allergy showed 9.28 times the odds of a prescribing incident (95% CI 1.61, 53.5; P = 0.013). Increasing the number of ward transfers increased the odds of prescribing incidents 125.8 times (95% CI 11.2, 1412; P < 0.001). For each additional prescribed medication, the odds of a prescribing incident decreased 0.88 times (95% CI 0.79, 0.99; P = 0.026), while for each additional high-alert prescribed medication, the odds increased 1.28 times (95% CI 1.01, 1.61; P = 0.040). In terms of administering incidents in PC (Table 4), there were six factors in the model. The presence of an interpreter reduced the odds of an administering incident 0.13 times (95% CI 0.02, 0.75; P = 0.024). By increasing the number of ward transfers, the odds of an administering incident increased 6.15 times (95% CI 2.23, 16.62; P < 0.001). Administration of medications during weekends had 1.72 times greater odds of an administering incident (95% CI 1.15, 2.54; P = 0.007) relative to weekdays. Research paper “Prevention of High-Alert Medication Errors in Hospital Patients” Discussion This audit reveals the importance of systematically examining patient-, environment- and medication-related factors to Table 1 Patient demographic characteristics for whole sample and for each clinical setting (N = 540 for whole sample, n = 108 for each clinical setting) Variable a Total n b(%) Cardiac care n c(%) Emergency care n c(%) Intensive care n c(%) Oncology care n c(%) Perioperative care n c(%) ……………………………………………………………………………………………………………………………………………………………………………………………………………….. 318 (58.9) 36 (33.3) 57 (52.8) 70 (64.8) 244 (45.2) 121 (22.4) 43 (8.0) 25 (4.6) 25 (4.6) 25 (4.6) 21 (3.9) 18 (3.3) 8 (1.5) 7 (1.3) 2 (0.4) 1 (0.2) 106 (98.1) 68 (63.0) 1 (0.9) 9 (8.3) 31 (28.7) 5 (4.6) 13 (12.0) 14 (13.0) 10 (9.3) 12 (11.1) 8 (7.4) 7 (6.5) 3 (2.8) 2 (1.9) 2 (1.9) 1 (0.9) 2 (1.9) 5 (4.6) 3 (2.8) 7 (6.5) 5 (4.6) 5 (4.6) 5 (4.6) 53 (49.1) 66 (61.1) 40 (37.0) 7 (6.5) 20 (18.5) 11 (10.2) 10 (9.3) 8 (7.4) 6 (5.6) 6 (5.6) 24 (22.2) 23 (21.3) 18 (16.7) 8 (7.4) 8 (7.4) 8 (7.4) 4 (3.7) 4 (3.7) 4 (3.7) 2 (1.9) 1 (0.9) 1 (0.9) 1 (0.9) 1 (0.9) 519 (96.1) 11 (2.0) 9 (1.7) 106 (98.1) 1 (0.9) 1 (0.9) 104 (96.3) 4 (3.7) 97 (89.8) 8 (7.4) 2 (1.9) 107 (99.1) 1 (0.9) 105 (97.2) 2 (1.9) 1 (0.9) (continued ) 311 Downloaded from https://academic.oup.com/intqhc/article-abstract/26/3/308/2849751 by Nova Southeastern University user on 20 January 2018 Medication incidents • Safety Gender Male Admitting diagnosis Cardiovascular Neoplastic Gastrointestinal Hepatobiliary Neurological Respiratory Musculoskeletal Renal and genitourinary Allergy and infection Metabolic and endocrine Dermatological Psychiatric Breast cancer Colorectal cancer Leukemia Bladder cancer Hodgkin’s and non-Hodgkin’s lymphoma Lung cancer Multiple myeloma Esophageal cancer Pancreatic cancer Prostate cancer Gastric cancer Melanoma Mesothelioma of pleura Waldenstrom macroglobulinemia Usual accommodation Home Supportive accommodation Nursing home Manias et al. 312 Table 1 Continued Variable a Total n b(%) Cardiac care n c(%) Emergency care n c(%) Intensive care n c(%) Oncology care n c(%) Perioperative care n c(%) ……………………………………………………………………………………………………………………………………………………………………………………………………………….. Hostel Concession card holder No Yes Documented allergies present No Yes Australasian Triage Scale Category 1 Category 2 Category 3 Category 4 Category 5 1 (0.2) 1 (0.9) 217 (40.2) 323 (59.8) 36 (33.3) 72 (66.7) 37 (34.3) 71 (65.7) 48 (44.4) 60 (55.6) 50 (46.3) 58 (53.7) 46 (42.6) 62 (57.4) 360 (66.7) 180 (33.3) 49 (45.4) 59 (54.6) 72 (66.7) 36 (33.3) 76 (70.4) 32 (29.6) 80 (74.1) 28 (25.9) 83 (76.8) 25 (23.2) 10 (9.3) 50 (46.3) 34 (31.5) 10 (9.3) 4 (3.7) Age M = 64.1 years SD = 15.7 years M = 67.5 years SD = 13.6 years M = 65.2 years SD = 16.2 years M = 62.1 year … Purchase answer to see full attachment Student has agreed that all tutoring, explanations, and answers provided by the tutor will be used to help in the learning process and in accordance with Studypool’s honor code & terms of service . Get a 10 % discount on an order above $ 100 Use the following coupon code : NURSING10