Abstract

Introduction

The goal of patient safety is to reduce the risk of injury or harm to patients by improving both the structure and process of care delivery. Therefore, when unintended risks and hazards associated with the delivery and management of care are eliminated or reduced, patient safety is maximized. To reduce safety risks to patients, health systems expend considerable resources to prevent medication errors in the hospital setting.^1,2,3 Increasingly, inpatient medication risk management efforts focus on preventing errors by improving systems for medication administration and creating a culture of safety.^1,3

Unfortunately, the potential for patient harm and increased medical liability due to medication discrepancies, defined as unexplained differences among documented regimens across different sites of care, continues after hospital discharge. Preventable adverse drug events (ADEs) have been shown to frequently occur in a variety of settings, including the hospital,^4,5 nursing homes,⁶ and the ambulatory/outpatient setting.⁷ In addition to poor clinical outcomes associated with preventable ADEs, several studies have highlighted the significant liability claim costs associated with preventable ADEs occurring in both inpatient and outpatient settings.^8,9 The transition from hospital to community settings, including transfers for those receiving home care services, is an exceptionally risky time, particularly for older adults with multiple chronic morbidities that place them at particular risk for medication discrepancies and associated ADEs.^10,11,12,13

In general, home health care is inclusive of a range of health care services that can be provided within the home environment to manage illness and/or injury. As part of the study described in this paper, participants were receiving skilled services from a certified home care agency. Our previous work indicates that up to 90 percent of patients experience at least one medication discrepancy in the transition from hospital to home care, and discrepancies occur within all therapeutic classes of medications.^14,15 Patients with hospital-to-home-care medication discrepancies are almost twice as likely to be readmitted to the hospital within 30 days as patients with no medication discrepancies.¹⁶ Recent research aimed at improving transitional care and medication safety also demonstrates enhanced patient outcomes and reduced costs.^17,18,19

ADEs occur frequently in both inpatient and outpatient settings and are a major patient safety concern.^{20,21,22,23,24,25,26} ADEs result in increased resource utilization, including additional diagnostic tests, physician visits, medication use, emergency department (ED) visits, and hospitalizations,²⁷ and in 2000, ADE-associated costs were estimated to exceed $117 billion.²⁸ ADEs that are primarily the result of human errors occur when health care team members fail to communicate effectively, do not document information adequately, and/or neglect to “handoff” relevant information to subsequent providers.²⁹ Furthermore, confusion in lines of authority and failure to appropriately consult experts when prescribing and administering medications to selected populations of patients may result in ADEs.²⁹ Such human factors resulting in ADEs are particularly common during transition within or between health systems.²⁹

A critical need exists to develop and implement medication risk management strategies in conjunction with emerging transitional care practices. The purpose of this secondary analysis was to evaluate the potential for medication discrepancies to contribute to ADEs during care transition from the hospital to home care setting. Medication discrepancies were analyzed for their potential to result in an ADE and then classified based on potential health consequences and anticipated health care utilization associated with the potential ADEs. The results of this study are intended to aid in the development of effective risk management strategies to improve medication information transfer and enhance patient safety during care transition from hospital to home care.

Methods

Following institutional review board approval, an analysis was performed of secondary data from a previous study that evaluated and categorized hospital-to-home medication discrepancies via review of medical records and participant interviews. Recruited participants were those discharged from one of two hospitals who received home health care services from a Medicare-certified home health agency. Both recruitment hospitals and the home care agency are part of the same health system. One of the hospitals is the region’s largest tertiary referral center, with nearly 30,000 inpatient admissions annually, while the other is a community medical center with approximately 9,000 annual inpatient admissions. In the larger parent study, medication records on 212 patients discharged from the hospital to home care were examined. Most (89 percent) patients had at least one medication discrepancy, and a total of 1,389 medication discrepancies were identified. Each discrepancy was classified using the Medication Discrepancy Tool (MDT®).³⁰ The MDT® is a tool by which a medication discrepancy can be described and subsequently categorized by causes and/or contributing factors. The “types” of discrepancies that can be categorized utilizing the MDT® are as follows: adverse drug reaction or intolerance; prescription not filled; medication not needed; financial barriers exist; intentional or non- intentional non-adherence; performance deficit in taking medication; conflicting information from care providers; brand/generic confusion; discharge instructions incomplete or inaccurate; incorrect dosage, quantity or label; unrecognized cognitive impairment; or unmet need for caregiver assistance.³⁰

For the current secondary analysis, each discrepancy was further independently evaluated by

two pharmacists with experience identifying and resolving medication discrepancies. For each discrepancy, the reviewing pharmacists were provided with the following information:

1.

Patient age and sex.

2.

Medication name.

3.

A brief narrative description of the discrepancy (e.g., medication not listed on patient’s discharge list, prescription states 5 mg but the patient takes 10 mg, patient did not want to take any more pills, etc.).

4.

A categorization for the type of discrepancy (as categorized by use of the MDT®).

ADE review and classifications reported in the current study utilized the conceptual model and scheme developed by Weingart and colleagues to assess drug-drug interaction safety alerts.³¹ Weingart’s model (see Figure 1) was tailored to better assess the potential of a medication discrepancy to cause or contribute to an ADE that would then require additional health care utilization. Based on the medications within each patient's regimen and the type of discrepancy identified, the pharmacists projected whether a medication would provoke a potential ADE and, if so, the potential severity of the ADE (serious, significant, or minor). Minor potential ADEs were defined as those that could result in minimal injury or discomfort such as flushing or mild stomach upset, for example. Significant potential ADEs were defined as those that could cause or contribute to symptoms, such as fever or pruritus; laboratory changes, such as hypoglycemia or hyperkalemia; or result in changes in vital signs, such as tachycardia. Lastly, serious ADEs were defined as those that could potentially result in significant organ failure or injury, such as major gastrointestinal bleeding or onset of angina.³¹

Figure 1.

Conceptual Model for Assessing the Potential of Medication Discrepancies To Contribute to an ADE and Increased Health Care Needs. Note: ADE, adverse drug event; ED, emergency department. Adapted from Weingart, Simchowitz, Padolsky, et al. Arch Intern (more...)

Furthermore, the predicted health consequences of the potential ADEs were categorized as follows: death, permanent disability, temporary disability, symptoms present 30 days or greater, symptoms present less than 30 days, abnormal lab results, or ADE but with no practical consequence. Finally, the pharmacists also independently rated the impact of a potential ADE for health care utilization (hospitalization, ED visit, office visit, provider phone call, or ADE but with no practical impact on health care use).

Cross tabulations among the independent pharmacists’ judgments for whether medications would potentially provoke an ADE revealed 84 percent agreement for medications judged to provoke an ADE and 62 percent agreement for medications not considered to provoke an ADE. The overall agreement in judgments across all 1,389 medications was 68 percent, leading to a Cronbach’s alpha for inter-rater reliability of 0.59. We deemed this as an insufficient level of agreement among raters, and therefore, we used an adjudication method for disagreements. For each medication in which there was disagreement on the potential for an ADE or its impact, the pharmacists discussed the medication in the context of the entire regimen until consensus was reached with regard to potential ADE risk. If a disagreement still existed, the physician co- investigator evaluated the case to make the final assignment of risk category. In the event of physician co-investigator review, the physician independently reviewed disputed cases using our predetermined conceptual model (Figure 1). In all cases of disagreement, the physician co- investigator’s assessment agreed with that of one of the pharmacist investigators. Accordingly, the assessment of the physician co-investigator was utilized in such cases.

Results

A total of 1,389 medication discrepancies were identified in 212 patients. This equated to an average of 6.6 discrepancies per patient. Of the total 1,389 discrepancies identified, 566 (40.7%) discrepancies involving 182 individual patients were determined to result in a potential ADE (a mean of 3.1 ADEs per patient). Of these, 6 (1.1%), 386 (68.2%), and 174 (30.7%) of the potential ADEs were rated as serious, significant, or minor, respectively. The 6 discrepancies determined to potentially result in a severe ADE occurred in 6 individuals, while 386 ADEs were rated as significant and 174 rated as minor occurred in 152 and 174 individual patients, respectively.

Medication classes most commonly involved in potential ADEs are displayed in Table 1. Medication discrepancies determined to contribute to serious potential ADEs involved the antineoplastic agent exemestane (1 patient), enoxaparin (1 patient), warfarin (1 patient) and sublingual nitroglycerin (5 patients). Potential ADEs ascribed a significant rating often involved opioids, antihypertensives, anticoagulants, diuretics, antidiabetic agents, antiarrhythmics, antiinfectives, systemic corticosteroids, antipsychotics, immune suppressants, and antiepileptic medications. Minor potential ADEs most often involved OTC agents, antacids, cholesterol lowering drugs, allergic rhinitis agents, nonsteroidal anti-inflammatory drugs (NSAIDs), and thyroid supplements. While 566 of the 1,389 (40.7%) medication discrepancies were judged to contribute to a potential ADE, 833 discrepancies (59.3%) were assessed as unlikely to cause or contribute to a potential ADE. Medications commonly involved in discrepancies assessed to have insignificant consequences involved acetaminophen, laxatives, aspirin, allergic rhinitis agents, and antihistamines.

Table 1.

Top Ten Therapeutic Drug Classes Resulting in Potential Adverse Drug Events (ADEs)

The most common potential health consequence assessed was “symptoms lasting less than 30 days,” which equated to 269 (47.5 percent) of the 566 total potential ADEs evaluated (Table 2). The potential health consequence of “abnormal laboratory values or altered vital signs” was also a common consequence accounting for 266 potential ADEs (47 percent). The pharmacist reviewers determined that none of the potential ADEs were likely to have resulted in “death or permanent disability.” “Temporary disability” was assessed as the likely outcome for one (0.2 percent) of the identified potential ADEs, while the health consequence of “symptoms lasting 30 or more days” accounted for seven (1.3 percent) potential ADEs. No likely subsequent health consequences were evaluated for 23 (4 percent) potential ADEs.

Table 2.

Study Identified Potential Health Consequences of Potential Adverse Drug Events by Rated Severity

If an ADE was determined to have the potential to result in a subsequent health consequence, it was then assessed for the type of health care entity or modality most likely utilized to address and manage the health consequence. It was deemed that inpatient hospitalization would likely be required to manage three potential ADEs (0.6 percent), with an ED visit likely required to manage 12 (2.1 percent) of the potential ADEs (Table 3). Discrepancies and resulting potential ADEs involving exemestane, warfarin, and enoxaparin were deemed to potentially require hospitalization in each of three separate patients. ADEs anticipated to potentially require an ED visit involved enoxaparin, warfarin, aspirin, dipyridamole + aspirin, albuterol and/or ipratropium, and levofloxacin in 12 separate patients. While 65 (11.5 percent) of the potential ADEs were deemed unlikely to require any additional contact with health care providers, 71 (12.5 percent) were assessed to potentially necessitate an office visit. The great majority of discrepancies (n=415 or 73.3 percent) were judged to be most appropriately communicated and managed with a phone call to the health care provider’s office.

Table 3.

Study Identified Potential Health Care Utilization of Medication Discrepancies Leading to Potential Adverse Drug Events

Discussion

Our findings in a home care population further support the frequency and potential severity of medication discrepancies during care transitions from hospital to home. These findings further support the need for the development of effective risk management strategies to improve medication information transfer to enhance patient safety during care transitions from hospital to home care. In this study of 1,389 medication discrepancies, 40.7 percent were evaluated to likely result in an ADE; 69.3 percent were assessed to be potentially serious or significant. Six serious potential ADEs were identified in six individual patients, with 386 significant potential ADEs identified in 153 patients. Gandhi and colleagues estimated that ADEs occurred at a rate of 27.4 per 100 ambulatory patients, with 13 percent (3.6 per 100 patients) classified as serious.³² In a hospital study involving 563 patients, 225 (40 percent) of the patients were found to have a discrepancy at admission or discharge.³³ Discrepancies were rated on a 1-3 harm scale (1 being minor, 2 being moderate, and 3 being severe), with 162 patients having a discrepancy rated as 2 or 3, roughly correlating with our scale of significant or serious, respectively. Specifically at discharge, 51 percent of the 167 discrepancies identified were rated as a 2, with 23 percent being rated as a 3.³³ In an ambulatory care population, there were 5.24 discrepancies identified per patient, and medication errors were identified in 88.9 percent of the study visits, with 75 percent, 24 percent, 0.3 percent of the medication errors being rated as minor, significant, and serious, respectively.³⁴

In our study, the top five medication classes involved in a potential ADE were antihypertensives, opioids, anticoagulants, injectable and oral antidiabetic agents, and inhaled medications for chronic obstructive pulmonary disease (COPD) or asthma. Classes of medications most frequently contributing to an ADE in a study involving outpatients from four different primary care offices were selective serotonin reuptake inhibitors (SSRIs), beta-blockers, angiotensin converting enzyme inhibitors (ACEIs), calcium channel blockers (CCBs), and nonsteroidal anti- inflammatory agents (NSAIDs).³² Similar to this study, we identified antihypertensives (which include beta-blockers, ACEIs, and CCBs) as likely to result in an ADE commonly in community-dwelling patients. However, the remaining top medication classes differ; a possible explanation being that the participants involved in Gandhi’s study were able to report actual ADEs, and the associated symptoms directly related to the offending drug (e.g., sexual dysfunction and SSRI use, loop diuretic use associated urinary incontinence). Furthermore, our study involved patients discharged from hospital to home care, while the Gandhi study involved patients receiving primary care services, apart from a recent hospitalization, further distinguishing among differences in ADEs that may be experienced by two different community dwelling populations with potential variability in health acuity.

Findings from our secondary analysis additionally identified discrepancies involving medications considered to be high risk medications for contributing to ADEs and ED visits, per previous studies and reports. In regard to ADEs leading to acute care utilization, we identified approximately 2 percent of the potential ADEs involving 10 patients that were likely to result in an ED visit. The involved medicines were anticoagulants or antiplatelets (e.g., enoxaparin, warfarin, dipyridamole + aspirin), and sublingual nitroglycerin. Hospitalization was deemed likely necessary in 0.5 percent of potential ADEs, involving three patients, and associated with the antineoplastic drug exemestane and the anticoagulants enoxaparin and warfarin. In a different study examining community-dwelling older adults, insulin, digoxin, and warfarin were identified as medications commonly leading to an ADE requiring an ED visit.³⁴ In another investigation, the top five categories of medicines implicated in ADEs requiring an ED visit included the following broad classes: (1) central nervous system agents (including opioids); (2) systemic antimicrobial agents; (3) hormone modifying agents, including insulin and oral antidiabetic medications; (4) hematologic agents, including anticoagulants and platelet inhibitors; and (5) cardiovascular agents (e.g., ACEIs, digitalis).²³ Thus, insulin and antidiabetic medicines, as well as anticoagulants and platelet inhibitors, were identified in all three studies as contributing to an ADE or potential ADE that may require subsequent acute care utilization. Overall, medications and medication classes identified in our study as high risk for potential ADEs and subsequent health care utilization largely align with other studies reported in the literature.

In addition to patient safety and quality of care implications associated with the potential ADEs identified, our findings have additional implications in terms of medical liability. As discussed previously, preventable ADEs have been shown to occur frequently in a variety of settings, including the hospital,^4,5 nursing homes,⁶ and the outpatient setting.⁷ A retrospective analysis of a New England malpractice insurance company claims records from January 1, 1990 to December 31, 1999, concluded that ADEs represented 6.3 percent of malpractice claims filed.⁹ Of the ADE claims analyzed, 73 percent of them were deemed preventable and were nearly evenly divided between ADEs occurring in the inpatient and outpatient settings. Similar to our findings, the most common medication classes involved were antibiotics, antidepressants, antipsychotics, cardiovascular drugs, and anticoagulants.9 Because the occurrence of medication discrepancies during transition from hospital to home are frequent, interventions designed to identify and resolve medication discrepancies during transitions of care is one potential strategy in preventing ADEs to improve patient care and mitigate medical liability risk.

Our findings concerning medications and medication classes frequently associated with medication discrepancies and potential ADEs are supported by the findings of other studies in other settings and health care environments. The study described herein additionally provides important insight into the frequency and potential severity of medication discrepancies that can help identify patients who may be at risk for experiencing medication-related ADEs during transition from hospital to home. The results of our study should, however, be considered within its limitations. The impacts of actual ADEs were not contained in the data and therefore were unknown. Furthermore, study participants in the present study included patients receiving care from one health system in the Northwestern United States, and our findings may not be generalizable to other health systems or health care that patients receive in other parts of the country.

Conclusion

Medication discrepancies frequently occur during the transition from hospital to home care. In the current study, more than 40 percent of all medication discrepancies were classified as having the potential to result in an ADE. Of these, a majority were classified as potentially serious or significant, with the most common likely resultant health consequences being transient symptoms lasting fewer than 30 days or alterations in laboratory values or vital signs. Many of the potential ADEs were assessed as likely to be resolved by a telephone call to a provider, with a lesser percent requiring a visit to the physician’s office, ED, or hospital.

This project highlights the need to proactively implement strategies to minimize potential ADEs occurring as a result of medication discrepancies surfacing between hospital discharge and home care. Additionally, specific medications and drug classes that were associated with the greatest potential of conveying a negative impact on the patient’s health or resources should be given priority for improvements. Ultimately, this information creates a foundation for developing an effective risk management strategy for improving patient safety. Our findings can be utilized by health care professionals and health systems to develop interventions to improve the safe use of high risk medicines frequently associated with discrepancies that are likely to result in ADEs requiring further health care utilization.

Acknowledgments

This project was funded by the Agency of Healthcare Research and Quality (AHRQ grant HS19552). In memoriam, we acknowledge the leadership and vast contributions to this work from Dr. Stephen M. Setter.

Author Affiliations

Joshua J. Neumiller, PharmD; Associate Professor and Vice-Chair of Pharmacotherapy, College of Pharmacy, Washington State University. Stephen M. Setter, PharmD, DVM; formerly Associate Professor of Pharmacotherapy, College of Pharmacy, Washington State University. Allison M. White, PharmD; Clinical Pharmacist, Providence Sacred Heart Medical Center. Cynthia F. Corbett, PhD, RN, FAAN; Professor and Associate Dean of Research, College of Nursing, Washington State University. Douglas L. Weeks, PhD; Director of Research, St. Luke’s Rehabilitation Institute. Kenn B. Daratha, PhD, MBA; Associate Professor, College of Nursing, Washington State University. Jeffrey B. Collins, MD; Chief Medical Officer, Providence Sacred Heart Medical Center.

Address correspondence to: Joshua J. Neumiller, PharmD, Washington State University, College of Pharmacy, P.O. Box 1495, Spokane, WA 99210; e-mail: ude.usw@rellimuenj.

References

1.

Goeschel CA, Bourgault A, Palleschi M, et al. Nursing lessons from the MHA Keystone ICU Project: developing and implementing an innovative approach to patient safety. Crit Care Nurs Clin North Am. 2006;18:481–92. [PubMed: 17118302]

2.

Hsia DC. Medicare quality improvement: bad apples or bad systems? JAMA. 2003;289:354–6. [PubMed: 12525237]

3.

Pronovost PJ, Berenholtz SM, Goeschel CA, et al. Creating high reliability in healthcare organizations. Health Serv Res. 2006;41(4):1599–617. [PMC free article: PMC1955341] [PubMed: 16898981]

4.

Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. N Engl J Med. 1991;324:370–6. [PubMed: 1987460]

5.

Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients: results of the Harvard Medical Practice Study II. N Engl J Med. 1991;324:377–84. [PubMed: 1824793]

6.

Gurwitz JH, Field TS, Avorn J, et al. Incidence and preventability of adverse drug events in nursing homes. Am J Med. 2000;109:87–94. [PubMed: 10967148]

7.

Gandhi TK, Burstin HR, Cook EF, et al. Drug complications in outpatients. J Gen Intern Med. 2000;15:149–54. [PMC free article: PMC1495358] [PubMed: 10718894]

8.

Lahue BJ, Pyenson B, Iwasaki K, et al. National burden of preventable adverse drug events associated with inpatient injectable medications: healthcare and medical professional liability costs. Am Health Drug Benefits. 2012;5(7):1–10. [PMC free article: PMC4031698] [PubMed: 24991335]

9.

Rothschild JM, Federico FA, Gandhi TK, et al. Analysis of medication-related malpractice claims. Arch Intern Med. 2002;162:2414–20. [PubMed: 12437399]

10.

Foust JB, Naylor MD, Boling PA, et al. Opportunities for improving post-hospital medication management among older adults. Home Health Care Serv Q. 2005;24:101–22. [PubMed: 16236662]

11.

Greenwald JL, Denham CR, Jack BW. The hospital discharge: a review of high risk care transition with highlights of a reengineered discharge process. J Patient Saf. 2007;3:97–106.

12.

Kripalani S, LeFevre F, Phillips CO, et al. Deficits in communication and information transfer between hospital-based and primary care physicians: implications for patient safety and continuity of care. JAMA. 2007;297:831–41. [PubMed: 17327525]

13.

Marcotte L., Kirtane J, Lynn J., et al. Integrating health information technology to achieve seamless care transitions. J Patient Saf. 2015;11(4):185–90. [PubMed: 24522208]

14.

Corbett CL, Setter SM, Neumiller JJ, et al. Nurse identified hospital to home medication discrepancies: implications for improving transitional care. Geriatr Nurs. 2011;31(3):188–96. [PubMed: 20525523]

15.

Setter SM, Corbett CF, Neumiller JJ, et al. Resolving medication discrepancies in older patients transitioning from hospital to home care: impact of a pharmacy/nursing intervention. Am J Health Syst Pharm. 2009;66:2027–31. [PubMed: 19890086]

16.

Coleman EA, Parry C, Chalmers S, et al. The Care Transitions Intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166:1822–8. [PubMed: 17000937]

17.

Balaban RB, Weissman JS, Samuel PA, et al. Redefining and redesigning hospital discharge to enhance patient care: a randomized controlled study. J Gen Intern Med. 2008;23(8):1228–33. [PMC free article: PMC2517968] [PubMed: 18452048]

18.

Jack BW, Chetty VK, Anthony D, et al. A reengineered hospital discharge program to decrease rehospitalization: a randomized trial. Ann Intern Med. 2009;150:178–87. [PMC free article: PMC2738592] [PubMed: 19189907]

19.

Phatak A, Prusi R, Postelnick M, et al. Impact of pharmacist involvement in the transitional care of high-risk patients through medication reconciliation, medication education, and postdischarge call-backs (IPITCH Study). J Hosp Med. 2016;11(1):39–44. [PubMed: 26434752]

20.

Evans RS, Floyd JF, Stoddard GJ, et al. Risk factors for adverse drug events: a 10-year analysis. Ann Pharmacother. 2005;39:1161–8. [PubMed: 15897265]

21.

Bates DW, Cullen DJ, Laird N, et al. Incidence of adverse drug events and potential adverse drug events; implications for prevention: ADE Prevention Study Group. JAMA. 1995;274:29–34. [PubMed: 7791255]

22.

Baker GR, Norton PG, Flintoft V, et al. The Canadian adverse events study: the incidence of adverse events among hospital patients in Canada. CMAJ. 2004;170:1678–86. [PMC free article: PMC408508] [PubMed: 15159366]

23.

Budnitz DS, Pollock DA, Weidenback KN, et al. National surveillance of emergency department visits for outpatient adverse drug events. JAMA. 2006;296:1858–66. [PubMed: 17047216]

24.

Zed PJ, Abu-Laban RB, Balen RM, et al. Incidence, severity and preventability of medication-related visits to the emergency department: a prospective study. CMAJ. 2008;178:1563–9. [PMC free article: PMC2396352] [PubMed: 18519904]

25.

Bedell SE, Jabbour S, Goldberg R, et al. Discrepancies in the use of medications. Arch Intern Med. 2000;160:2129–34. [PubMed: 10904455]

26.

Conklin JR, Togami JC, Burnett A, et al. Care transitions service: A pharmacy-driven program for medication reconciliation through the continuum of care. Am J Health Syst Pharm. 2014;71(10):802–10. [PubMed: 24780489]

27.

Al-Tajir GK, Kelly WN. Epidemiology, comparative methods of detection, and preventability of adverse drug events. Ann Pharmacother. 2005;39:1169–1174. [PubMed: 15928265]

28.

Ernst FR, Grizzle AJ. Drug-related morbidity and mortality. A cost-of-illness model. J Am Pharm Assoc. 2001;41:192–9. [PubMed: 11297331]

29.

Institute of Medicine, Committee on Identifying and Preventing Medication Errors, Aspden P, Wolcott J, et al. (Eds). Medication Errors: Quality Chasm Series (pp. 1-25), Washington DC: National Academies Press; 2007. 30 Smith J, Coleman E, Min S. A new tool for identifying discrepancies in postacute medications for community-dwelling older adults. Am J Geriatr Pharmacother 2004; 2(2):141-7.

30.

Smith J, Coleman E, Min S. A new tool for identifying discrepancies in postacute medications for community-dwelling older adults. Am J Geriatr Pharmacother. 2004;2(2):141–7. [PubMed: 15555490]

31.

Weingart SN, Simchowitz B, Padolsky H, et al. An empirical model to estimate the potential impact of medication safety alerts on patient safety, health care utilization, and cost in ambulatory care. Arch Intern Med. 2009;169(16):1465–73. [PubMed: 19752403]

32.

Gandhi TK, Weingart SN, Borus J, et al. Adverse drug events in ambulatory care. N Engl J Med. 2003;348:1556–64. [PubMed: 12700376]

33.

Feldman LS, Costa LL, Feroli ER, et al. Nurse-pharmacist collaboration on medication reconciliation prevents potential harm. J Hosp Med. 2012;7(5):396–401. [PubMed: 22371379]

34.

Budnitz DA, Shehab N, Kegler SR, et al. Medication use leading to emergency department visits for adverse drug events in older adults. Ann Intern Med. 2007;147:755–65. [PubMed: 18056659]