Essay Paper

Implementationofanelectronicsystem.pdf

Home >Nursing homework help >Essay Paper

PRACTICE REPORTS Medication reconciliation

404 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

JOAN S. KRAMER, PHARM.D., is Clinical Research and Hospital Medi- cine Specialist, Department of Pharmacy; PAULA J. HOPKINS, B.S.N., M.S.N., is Adult Medical–Surgical Clinical Nurse Specialist, Depart- ment of Trauma and Surgery; and JAMES C. ROSENDALE is Data Architect, Department of Pharmacy, Wesley Medical Center (WMC), Wichita, KS. JAMES C. GARRELTS, B.S., PHARM.D., is Leader, Clinical Pharmacy Services, Via Christi Regional Medical Center, St. Francis Campus, Wichita; at the time of this study he was Manager, Critical Care Pharmacy Services and Research, Department of Pharmacy, WMC. LADONNA S. HALE, PHARM.D., is Associate Professor, Depart- ment of Physician Assistant, College of Health Professions, Wichita State University, Wichita, and Clinical Pharmacist, Department of Pharmacy, WMC. TINA M. NESTER, PHARM.D., is Internal Medicine Clinical Specialist, Department of Pharmacy; PATTY COCHRAN,

B.S.N., M.S.N., is Clinical Nurse Specialist—Acute Care for the El- derly, Department of Nursing; and LESLIE A. EIDEM, B.S., is Pharmacy Manager, Pediatric/Women’s Services and Medication Safety, Depart- ment of Pharmacy, WMC. ROBERT D. HANEKE, B.S., PHARM.D., is National Clinical Director, PharmaSource Healthcare, Inc., Sylvia, KS; at the time of this study he was Ambulatory and Geriatrics Clini- cal Specialist, Department of Pharmacy, WMC.

Address correspondence to Dr. Kramer at the Department of Pharmacy, Wesley Medical Center, 550 North Hillside, Wichita, KS 67214-4976 (joan.kramer@wesleymc.com).

DOI 10.2146/ajhp060506

Implementation of an electronic system for medication reconciliation

JOAN S. KRAMER, PAULA J. HOPKINS, JAMES C. ROSENDALE, JAMES C. GARRELTS, LADONNA S. HALE, TINA M. NESTER, PATTY COCHRAN, LESLIE A. EIDEM, AND ROBERT D. HANEKE

Purpose. The feasibility of implement- ing an electronic system for targeted pharmacist- and nurse-conducted admis- sion and discharge medication reconcilia- tion and its eff ects on patient safety, cost, and satisfaction among providers and nurses were studied. Methods. This study was conducted in two phases: a preimplementation phase and a postimplementation phase. In the preim- plementation phase, admission medication histories and discharge medication coun- seling followed standard care processes. During postimplementation, pharmacists and nurses collaborated to electronically complete admission and discharge medi- cation reconciliation documentation. Four reports were developed for medication rec- onciliation documentation: (1) home medi- cation profi le report, (2) home medication reconciliation report, (3) discharge medica- tion reconciliation report, and (4) patient discharge medication report. Patients were contacted after discharge to measure their satisfaction with the medication counsel- ing and medication instructions received. Health care providers completed a survey indicating their satisfaction with the elec- tronic medication reconciliation processes. Results. A total of 283 patients were includ-

ed in the study. Patients in the postimple- mentation group took signifi cantly more prescription and nonprescription medica- tions, and their total number of medica- tions signifi cantly exceeded the number taken by the preimplementation group. Pharmacists completed signifi cantly more dosage changes in the postimplementa- tion phase than in the preimplementation phase. In the preimplementation phase, nurses identifi ed more incomplete medica- tion orders, dosage changes, and allergies than they did in the postimplementation phase. Patients in the postimplementation group reported a higher level of agreement on all survey items regarding adequate discharge medication instructions. Conclusion. Patients who had their medi- cations electronically reconciled reported a greater understanding of the medications they were to take after discharge from the hospital, including medication administra- tion instructions and potential adverse eff ects.

Index terms: Computers; Documentation; Dosage; Errors, medication; Hospitals; Nurses; Patient information; Patient infor- mation; Pharmacists; Toxicity Am J Health-Syst Pharm. 2007; 64:404-22

A n estimated 5% of hospitalized patients experience medication errors, 60% of which occur

during transitions of care (i.e., ad- mission, transfer between levels of care, and discharge).1 In the United States, this translates to over 90,000 hospitalized patients who experience medication errors each year.2

Two critical times to prevent med- ication errors are at admission and discharge.1,2 Adverse drug events are responsible for 3.2–9.6% of hospital admissions and are the fi fth leading cause of death in the United States.3-15 Recording an accurate and complete medication history is an important part of the initial patient assessment at admission. Inaccuracies in the medication history result in wasted time and interrupted or inappropri- ate drug therapy and may jeopardize patient safety.1,16-24 Nonprescription medications and herbal preparations are also associated with clinically significant drug interactions and adverse effects.25,26 In a Finnish study of nonprescription and prescription

PRACTICE REPORTS Medication reconciliation

405Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

drug interactions, 4% of adults tak- ing nonprescription products were exposed to potentially clinically sig- nifi cant drug interactions.25

Discharge from the hospital is another crucial time for ensuring medication accuracy and maintain- ing patient safety. Less than half of 43 patients in a New York City hospital remembered medication-related in- formation (i.e., name, purpose, and major adverse effects) at discharge.27

An accurate assessment of a patient’s medications during hospitalization, along with knowledge of medica- tions taken at home, is necessary to write correct discharge medication orders and educate the patient about medications that are to be continued on an outpatient basis. Without a standardized process for medication reconciliation, the reliability of the information recorded at admission is variable and can be infl uenced by the training and background of the personnel involved, the time allotted with the patient, and the patient’s level of familiarity with his or her drug therapies.

Although pharmacist-conducted medication histories and discharge counseling are considered desir- able clinical pharmacy services, only about 5% of U.S. hospitals reported having pharmacists in these roles in 2002.2 More recently, 44.8% of hospitals surveyed reported having a medication reconciliation system in place; however, the report did not reveal whether pharmacists had a role in medication reconciliation.28

This increase in the implementa- tion of medication reconciliation systems likely refl ects a new focus on this service by the Joint Commis- sion on Accreditation of Healthcare Organizations. The American Soci- ety of Health-System Pharmacists (ASHP), in its ASHP Health-System Pharmacy 2015 Initiative, has stated that pharmacists should be involved in managing the acquisition of medication admission histories and provision of discharge counseling

for 75% of hospital inpatients with complex and high-risk medication regimens by 2015.29 A previous pilot study conducted at our hospital16 and other published literature30-36 dem- onstrated that pharmacist-obtained medication histories are efficient and improve patient safety (Table 1). However, many hospitals do not have the funding and support necessary to consistently provide this service.

Several articles explain the impor- tance of medication reconciliation documentation and include paper forms for use in capturing the neces- sary information at the time of hos- pital admission,37-43 but few address completing similar documentation at discharge. Luther Midelfort Hospital (Mayo Health System) has provided one of the few published reports of computer-generated discharge medi- cation lists.39

Development of an electronic process to streamline the fl ow of pa- tients’ information is a relatively new concept for many hospitals, but the Department of Veterans Affairs has been using such systems for many years.44,45 North Mississippi Medical Center, a regional integrated man- aged health system, reported how an electronic-based admission and discharge medication reconciliation process has helped pharmacists with obtaining medication histories and nurses with completing discharge medication processing.46 At Thomas Hospital in Fairhope, Alabama, an automated process to retrieve patient prescription medication histories from insurance carriers or pharmacy benefit management companies is used to begin the medication recon- ciliation process.47

P h a r m a c i s t s a n d nu rs e s co l - laborate daily to provide patient- centered care, particularly in the m e d i c a tion-safety arena. 1 8 Both disciplines are positioned to work to- gether to perform medication recon- ciliation documentation. Our hospi- tal, Wesley Medical Center, located in Wichita, Kansas, and licensed for 760

beds and 102 bassinets, developed an electronic process for medication rec- onciliation using our clinical patient care system (CPCS) (Meditech, Inc., Boston, MA) to enable collaborative, standardized, targeted, pharmacist- and nurse-conducted admission and discharge medication reconciliation documentation. The goal of this study was to evaluate the medication reconciliation system’s feasibility and effects on patient safety.

Methods We conducted a prospective study

in a 48-bed adult general medical unit of our hospital. The study in- vestigators included pharmacists and nurses. Noninvestigators included pharmacists, nurses, and physicians. Pharmacists and nurses collaborated to identify high-risk patients, obtain the admission medication history, and complete the electronic ad- mission and discharge medication reconciliation processes. A repli- cable and generalizable model was created by targeting the service to high-risk patients, identifi ed through a set of trigger questions complet- ed by a nurse, and implementing pharmacist-obtained patient ad- mission medication histories and discharge medication reconciliation documentation. Technology was used to enhance the multidisciplinary fl ow of information.

This study was conducted in two phases: a preimplemention phase and a postimplementation phase. In the preimplementation phase, admission medication histories and discharge medication counseling followed standard care processes. A nurse obtained each patient’s medi- cation history and called the patient’s physician for admission medication orders. The nurse then handwrote admission medication orders in the physicians’ order section of the medi- cal record. At discharge, the nurse handwrote each patient’s medication list and provided discharge coun- seling. In the postimplementation

PRACTICE REPORTS Medication reconciliation

406 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

phase, pharmacists and nurses col- laborated to electronically complete admission and discharge medication reconciliation documentation. For this study, medication reconciliation was defi ned as an interdisciplinary process involving nurses, pharma- cists, and prescribers who collaborate to decrease medication errors and potential adverse events.48

Staff education. Nurse and phar- macist education. Nurses and phar- macists attended education sessions before study initiation. Nursing edu- cation was provided by investigators at staff meetings and individually. A fl ow chart was created to guide nurses through the admission and discharge medication reconciliation documentation process. All pharma-

cists attended a three-hour, hands-on computer education session. Before pharmacists were scheduled to work on the study unit, they completed an electronic medication order-entry competency evaluation covering admission through discharge using a test patient.

A medication history reference sheet was available to pharmacists;

Pharmacist histories were time-effi cient; pharmacists collected more information regarding prescription and nonprescription medications, herbal supplements, and allergies and made more clinical interventions when allowed to conduct medication history.

Face-to-face pharmacist interview was more accurate; patient-completed form was neither eff ective nor timesaving.

Pharmacist-conducted histories had fewer errors (32% vs. 52%, p < 0.001) and fewer signifi cant errors (1.6% vs. 3.1%, p < 0.01).

Mean ± S.D. medication errors were reduced from 1.45 ± 0.39 to 0.76 ± 0.07; pharmacists and nurses reported that new system was more accurate, safe, and effi cient.

Pharmacists recorded more prescription drugs (mean 5.6 ± 3.1 vs. 2.4 ± 1.3).

Pharmacists recorded at least one piece of additional information in 95% of patients and more prescription medication information for 54% of patients; 11% of pharmacist-conducted histories contained clinically signifi cant information missed by physician.

Pharmacists found discrepancies in 98% of patients (52/53) in the intervention group vs. physicians fi nding discrepancies in 53% of patients (26/50); pharmacists decreased drug cost at discharge by 22.5%, prevented 4 potential major (5%) and 48 moderate (43%) medication events, and decreased drug costs by a predicted $500,000 per yr (25 admissions/day, 5 days/wk, 46 wk/yr); results secured funding for a full-time clinical pharmacist to participate in multidisciplinary rounds.

Table 1. Studies Comparing Pharmacist-Conducted Medication Histories or Discharge Counseling with Those Conducted by Other Disciplines

Comparisona

Ref. Disciplines Di sc

ha rg

Summary of ResultsAd m

iss io

Sig ni

fi c an

Sa tis

fa cti

Fe as

ib ilit

Pharmacists and nurses

Patients and pharmacists

Pharmacists and unspecifi ed “usual process”

Pharmacy technicians and nurses

Pharmacists and physicians

aSignifi cance = some type of clinical signifi cance assigned to the prevented or actual medication error, satisfaction = satisfaction with new process measured in some fashion, feasibility = feasibility, cost-eff ectiveness, or eff ect on personnel time evaluated in some fashion.

PRACTICE REPORTS Medication reconciliation

407Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

the reference gave examples of pa- tient medication history interview techniques (e.g., open-ended ques- tions).49 Pharmacists recorded the patient medication history using a standardized form, which included fi elds for patient name, height, and weight; community pharmacist or pharmacy name and number; al- lergies and reactions; prescription medications, nonprescription prod- ucts and herbal supplements; home medication dosage, route, frequency, indication, and date and time last taken; and reminders to request information from the patient about nonprescription products, herbal supplements, patches, inhalers, eye drops, and vitamins and to record interventions electronically.

Physician education. Posters were placed on the medical unit to edu- cate physicians about the medication reconciliation process, including information on how to view medica- tions in the CPCS, what to complete on the reports, and whom to contact with questions. Individual education was provided for physicians who frequently admitted patients to the unit. In addition, patients enrolled in the study had orange sheets placed in the front of the medical record with written instructions explaining how to view medications in the CPCS, what to complete on the reports, and whom to contact with questions.

Care coordinator education. Care coordinators are either registered nurses or licensed social workers who direct case management activities. Since care coordinators often assist with compilation of discharge or transfer medication lists, investiga- tors provided inservice education to explain the medication recon- ciliation documentation process. Investigators also contacted nursing homes and skilled-nursing facilities to explain the intent of the medica- tion reconciliation discharge and patient medication discharge reports and to obtain feedback for improve- ment. Suggestions were taken into

consideration for future long-term implementation plans.

Pharmacist order entry. Specifi c problems created by study procedures and related to pharmacist order entry were addressed, and solutions were developed before implementing the medication reconciliation process. The computer system did not have a mechanism to inform nurses that the home medications were not to be administered during hospitalization, so a method was devised to prevent confusion. Special order types—HM (home medication) and DM (dis- charge medication)—were developed in the CPCS for home and discharge medications. These special order types prevented home medications from being visible to nurses in the electronic medication administra- tion record but allowed pharmacists to view the medications throughout each patient’s hospitalization. All medications entered by the phar- macist are normally accessible by the nurse from computer-controlled medication-dispensing cabinets in patient care areas. A false (or dum- my) doctor mnemonic, HM, was created and used for all home and discharge medication orders. The automated medication-dispensing cabinets were unable to recognize this false doctor mnemonic and therefore could not visualize home or discharge medications for nurses, preventing access to medications that were not to be administered during hospitalization.

Nonformulary items were entered using the same special order types. All home and discharge medications defaulted to one-minute stop times to allow medications to drop off of the pharmacist’s order-entry screen, leaving only inpatient orders active. Additional lines and special instruc- tions were added for discharge medi- cations with a tapering schedule (e.g., prednisone).

After home medications were en- tered, the home medication profi le was locked to prevent pharmacists

from accidentally modifying the admission medication history. A spe- cial function was created to lock the medication profi le by editing patient data. To lock the profi le, the phar- macist completing the medication history answered yes to the question “Home profi le entered?” in the Edit Patient Data screen.

The home medication profile was accessible to all providers on the Patient Care Inquiry screen. Instructions for accessing the home medication profi le for viewing in the CPCS were placed in the front of the patient medical record. Nurses and pharmacists were provided hand- outs explaining how to access this information.

Report development. Four re- ports were developed for medica- tion reconciliation documentation: (1) home medication profi le report, (2) home medication reconciliation report, (3) discharge medication reconciliation report, and (4) pa- tient discharge medication profi le. All reports contained information required by Health Information Management (HIM) to be retained as a part of the permanent medical re- cord. A meeting with HIM also estab- lished sections of the medical record where reports would be placed.

Home medication reports. Each report was developed with a specifi c purpose in mind. The home medi- cation profi le report and the home medication reconciliation report were generated at the same time and presented the same information in different formats. The home medica- tion profi le report listed the home medication history for reference; this report was placed in the History & Physicial/Education Record section of the medical record. The home medication reconciliation report listed the medication history and in- cluded a physician signature line and date (Figure 1). This report was used to reconcile the home medication history with the inpatient admission medication orders and was placed in

PRACTICE REPORTS Medication reconciliation

408 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

Figure 1. Home medication reconciliation report.

the physicians’ order section of the medical record.

Discharge reports. The discharge medication reconciliation report (Figure 2) was printed by an inves- tigator and placed in the Physicians Order section of the medical record daily before physicians conducted rounds. The discharge medication reconciliation report could also be printed on demand by nurses and pharmacists. This report grouped home medications and active inpa- tient medications by the American Hospital Formulary Service (AHFS) drug classifi cation to prevent medi- cation duplication at discharge.

Medication duplication often oc- curs when medications are changed to hospital formulary products. For example, a patient leaves the hospital with a prescription for famotidine, the hospital’s formulary product, and continues to take ranitidine, a home medication prescribed before hospital admission. Prescribers used the discharge medication reconcilia- tion report to indicate what medica- tions patients should be taking after discharge.

The patient discharge medication profi le was created after the phar- macist entered any new discharge medications. Two copies of the pro-

fi le were printed: one to give to the patient and one to put in the patient medical record (Figure 3). The report printed medication instructions in lay language if drug dictionary ad- ministration frequencies were used at the time of order entry. Pharmacists were instructed to enter lay language instructions for all orders during ini- tial order entry.

A search feature was programmed into the patient discharge medica- tion profi le to scan for new medica- tions continued at discharge. The following alert appeared on the report below each new medication: “Nurse, please print patient instruc-

PRACTICE REPORTS Medication reconciliation

409Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

tion sheet for this medication.” Nurses printed patient instructions for new medications from a com- mercially available online drug information service written in lay language (Lexi-Comp Online, Hud- son, OH). The patient discharge medication profi le and new medica- tion instructions were provided to the patient at discharge.

Figure 2. Discharge medication reconciliation report.

Patient enrollment. Patient en- rollment occurred during the fi rst work shift (7:00 a.m. to 3:30 p.m.), Monday through Friday, September 13, 2004, through February 28, 2005, for the preimplementation phase and May 6, 2005, through October 21, 2005, for the postimplementa- tion phase. The study protocol and patient informed consent were ap-

proved by the local scientifi c review committee and institutional review board (IRB). Initially, all patients provided written informed consent to participate in the study. However, after study initiation, an exemption for written informed consent was ob- tained from the IRB, and subsequent patients provided only oral consent before study participation.

PRACTICE REPORTS Medication reconciliation

410 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

Potential study participants were identifi ed through a set of trigger questions that the nurse asked pa- tients during the admission assess- ment. Trigger questions included the following:

1. Do you take seven or more medica- tions (total number of prescrip- tion, nonprescription, and herbal products)?

2. Do you have asthma? 3. Do you have chronic obstructive pul-

monary disease (COPD)? 4. Do you have diabetes? 5. Do you have any cardiac condition

(e.g., myocardial infarction [MI], congestive heart failure [CHF], ar- rhythmia, hypertension)?

6. Were you admitted with an adverse drug reaction?

7. Do you need to be vaccinated against pneumococcal disease (i.e., never re- ceived the Pneumovax immunization or received it over fi ve years ago)?

8. Do you need to be vaccinated against infl uenza (i.e., not yet vaccinated this year)?

9. Do you have more than three medi- cation allergies?

10. Do you have medications that need to be identifi ed?

A positive response (yes) to one or more trigger questions notifi ed a pharmacist through an electroni- cally generated report. Trigger ques- tions were based on those developed in a previous pilot study at our hospital.16

Once identified, patients were evaluated to determine whether they met study inclusion criteria. To be included in the study, patients had to be general medical patients age 18 years or older, be admitted to the study unit, and provide written or oral informed consent. Patients were excluded if a nursing medica- tion history was obtained more than 2 hours after admission, they were admitted for 23-hour observation, they transferred to or from another hospital unit, their admission was due to intentional drug overdose, or they could not provide consent. The rationale for excluding patients when the nursing medication history was obtained over 2 hours after admis- sion was to avoid confounding fac-

tors affecting medication reconcilia- tion during a prolonged time period compared with the postimplementa- tion phase. Translators were obtained for non-English-speaking patients, so such patients were not excluded. Patients who met study inclusion criteria and consented to participate were followed from admission to discharge. All study data were kept in a locked fi le cabinet in a pharmacy satellite offi ce. Keys to the fi le cabinet were kept in a computer-controlled medication dispensing cabinet in the patient care area.

Preimplementation phase. Ad- mission and discharge. A standardized data collection form was completed for each patient by an investiga- tor. The form included the follow- ing information: inclusion criteria, verifi cation that no exclusion criteria were met, study number assigned, patient name, hospital identifi cation number, length-of-stay information, and the patient’s telephone number. Consent information was placed in the patient medical record. Patients received the usual care provided on this unit from all health care provid-

Figure 3. Patient discharge medication profi le.

PRACTICE REPORTS Medication reconciliation

411Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

ers: Nurses obtained the medication history and performed medication- related interventions (e.g., calls to physicians were documented), and pharmacists performed medication- related interventions, documented the time required to complete the in- terventions, and processed admission medication orders as usual.

Pharmacist and nurse interventions. Pharmacists completed interventions electronically, the standard method for all intervention documentation in the pharmacy department during both phases of the study. To cap- ture nursing interventions, a home medication clarifi cation intervention was created for nurses to add to the intervention documentation screen in the CPCS. Pharmacist and nurse interventions included the following subcategories: (1) incomplete medica- tions, (2) medication duplications, (3) route changes, (4) dosage changes, (5) adverse drug reactions, (6) drug inter- actions, (7) vaccine recommendations, (8) laboratory test recommendations, (9) allergy documentation changed, (10) missing medications, (11) non- prescription medications identified, (12) herbal products identifi ed, (13) medication histories clarifi ed through a patient’s retail pharmacy, and (14) time to enter allergies into the com- puter. Defi nitions used for all inter- vention documentation are provided in the appendix.

Postimplementation phase. Ad- mission. Before study initiation, an electronic process for medication reconciliation documentation was developed and programmed. After the patient assessment was conduct- ed by a nurse and informed consent obtained, consent information was placed in the patient medical record. Patients meeting inclusion criteria in the postimplementation phase had medications electronically reconciled as follows (Figure 4):

1. After enrollment, pharmacists ob- tained the patient medication his- tory in a private area for patient confidentiality.

Figure 4. Admission medication reconciliation process.

Patient admission assessment taken by

nurse.

Yes

Nurse takes home medication history and documents

medications in the admission history.

Nurse prints home medication history and places in patient chart.

Nurse calls physician to obtain patient admission orders and to review home medication history.

Home medication history is marked to refl ect admission

orders.

Inpatient profi le is entered into pharmacy module by the

pharmacist.

Trigger report printed in pharmacy by the Meditech system.

Pharmacist obtains patient medication history and verifi es

using multiple resources.

Pharmacist takes home medication profi le and documents medications

in the Meditech system.

Pharmacist “locks” the patient home medication profi le in the Meditech

pharmacy module.

Pharmacist prints the home medication reconciliation report and places in the physician order section

of the patient chart.

Pharmacist and nurse coordinate contacting the physician to complete

the admission reconciliation.

Physician called to obtain patient admission orders and to reconcile

home medications.

Home medication reconciliation report is marked to refl ect reconciled

medications to be continued on admission. The report is scanned to

pharmacy for order processing.

Does patient meet trigger questionpatient meet trigger questionpatient meet trigger questionpatient meet trigger questionpatient meet trigger question

criteria?

Home medication history report is scanned to pharmacy for order

processing.

PRACTICE REPORTS Medication reconciliation

412 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

2. Pharmacists created the electronic pa- tient home medication profi le using multiple resources to establish an ac- curate medication history and locked the computerized patient profile, verifying the medications were home medications (Figure 5).

3. Once the home medication profi le was locked, the home medication rec- onciliation report was generated for prescribers to use to reconcile medi- cations on admission. Nurses and pharmacists worked with prescribers to complete admission medication reconciliation documentation.

4. Interventions captured during admis- sion medication reconciliation were electronically recorded and catego- rized by nurses and pharmacists.

Discharge. Discharge medica- tions were electronically reconciled through the following process:

1. The discharge medication reconcili- ation report, available in the front of the medical record or printed on demand, was used to reconcile medi- cations. This computer-generated re- port contained the home medications

and the active inpatient medications, listed by AHFS drug classification (Figure 2).

2. The prescriber marked through or wrote “discontinue” for medications not to be continued after discharge. When the prescriber completed this process, the report was electronically scanned and sent to the pharmacist.

3. The pharmacist verified that the medication reconciliation process was completed. If it was not, the nurse and pharmacist worked together to call the physician to obtain any clarifi ca- tion needed.

Figure 5. Pharmacist order-entry screen.

PRACTICE REPORTS Medication reconciliation

413Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

4. The pharmacist then printed the patient discharge medication profi le, which listed all medications the pa- tient should be taking after discharge in lay language. The nurse used the patient discharge medication profi le for discharge instructions, thereby eliminating the need for handwritten medication instructions (Figure 3).

5. The nurse provided information for new medications written in lay language. After nurses provided med- ication information and discharge in- structions, pharmacists were available for discharge counseling if patients had further questions.

6. Interventions completed during dis- charge were electronically recorded and categorized by nurses and pharmacists.

Outcomes measures. This study evaluated the (1) feasibility and ef- fi ciency of nurse-initiated, targeted, pharmacist-conducted medication histories and admission and dis- charge medication reconciliation, (2) the effect of this clinical service on patient safety, and (3) the satisfac- tion of patients, prescribers, nurses, and pharmacists with this clinical service.

Feasibility and effi ciency. Timed studies were completed to determine personnel resources required to implement this service. On admis- sion, the time for the nurse to obtain initial medication assessment and the time for the pharmacist to enter and clarify initial home medications, including patient medication assess- ment, were measured. At discharge, the time required for nurses to com- plete medication discharge instruc- tions and paperwork, for pharma- cists to print reconciliation reports and provide discharge counseling, and for prescribers to complete medication-related paperwork and write discharge prescriptions was measured.

Nursing and pharmacy admis- sion and discharge medication rec- onciliation accuracy was measured by the mean number of prescrip-

tion medications, nonprescription medications, herbal products, allergy descriptions, and medication dupli- cations identified. Nursing, phar- macy, and prescriber completeness of medication-related admission and discharge documentation was mea- sured as the percentage of patients with incomplete allergy descriptions (explanation of allergic reaction), medications (dosage, schedule, time of last dose taken, indication), vac- cination documentation, and allergy documentation (allergies identifi ed and documented in the computer system).

Patient safety. The number and type of potential errors prevented at admission and discharge were identi- fi ed by the mean number and type (intervention subcategory) of recon- ciliation interventions or discrepan- cies documented in the computer- ized database. Severity of potential errors prevented were categorized using the hospital’s policy for cat- egorizing medication errors and the 30-day readmission rate.50

Patient and health care profes- sional satisfaction. Patient, prescriber, nurse, and pharmacist surveys were conducted using a fi ve-point Likert scale to assess satisfaction with the medication reconciliation process.

Patient satisfaction survey.Patient satisfaction survey. At the time of admission or shortly thereafter, patients in both groups were offered the opportunity to participate in a telephone satisfac- tion survey to occur after discharge from the hospital. Patients willing to participate provided contact infor- mation to one of the investigators. Researchers attempted to contact all patients willing to participate as soon as possible after discharge. Attempts to contact these patients continued for up to 14 days after discharge. If requested, a family member was al- lowed to answer the survey questions if the patient was unable to come to the telephone (only an option if the family member was present when the patient was discharged).

The same scripted survey was used for all patients. The survey uti- lized a fi ve-point Likert scale, where 1 = strongly disagree, 2 = disagree, 3 = undecided, 4 = agree, and 5 = strongly agree. Patients were asked to respond to the following fi ve state- ments using this scale:

1. When I was discharged from the hospital, I was given clear instruc- tions about which medications I was supposed to continue taking at home.

2. I was given clear directions about how much and how often I am supposed to take my medications.

3. I was given clear directions how and when to take my medicine (e.g., take medications with food or on an empty stomach, in the morning only or at bedtime only).

4. I was given clear information about possible side effects of my medicine.

5. Overall, I feel like I understand my medicines.

After completion of this survey por- tion, patients were asked the follow- ing questions:

1. Did you receive any new prescriptions when you were dismissed from the hospital? If they answered yes, they were asked, Did you have all the new prescriptions fi lled? If they did not, they were asked, Can you tell me why you did not have them fi lled?

2. Do you have any additional questions about your medications?

3. Do you remember talking to a phar- macist about your medications while you were in the hospital?

Telephone surveyor training.Telephone surveyor training. All researchers conducting telephone surveys successfully completed a questionnaire measuring compre- hension of a review of survey design methodology, telephone survey eti- quette, and avoidance of bias in tele- phone surveys modeled from Health Research and Design Methodology.51

Although all researchers completed

PRACTICE REPORTS Medication reconciliation

414 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

the telephone survey competency, primarily three researchers conduct- ed the telephone surveys. For non- English-speaking patients, a trans- lator was utilized. This translator received abbreviated oral training be- fore contacting each patient. Proper etiquette training included impor- tance of confi dentiality, being cour- teous and professional, introduction of self, purpose of the telephone call, and responding to those who declined participation. Avoidance of bias training included avoiding deviation from the scripted survey, appropriately clarify ing unclear questions, utilizing a neutral voice, not rushing the respondent, and avoiding expression of satisfaction or dissatisfaction with the respondent’s answers. If a patient did have addi- tional questions regarding medica- tions, the surveyors were instructed to write down the questions and tell the patient that a letter would be sent to his or her physician for follow-up.

Health care professional survey.Health care professional survey. At the end of both the preimplementa- tion and postimplementation phases, surveys were completed by nurses, pharmacists, physicians, nurse prac- titioners, and physician assistants. The survey utilized a fi ve-point Lik- ert scale, where 1 = strongly disagree, 2 = disagree, 3 = undecided, 4 = agree, and 5 = strongly agree. They were asked to respond to the follow- ing fi ve statements:

1. Patients are provided clear instruc- tions about which medications they are to continue at home (previous home medications and any new dis- charge medications).

2. At the time of hospital discharge, patients have a clear understanding of how much, how often, and when to take their medications at home.

3. Most patients receive written infor- mation about new medications prior to discharge.

4. The patient medication discharge process is effi cient for me as a health care provider.

5. I use the computer system on a regular basis to view patient home medications.

The last two questions allowed the health care professionals to write in comments and suggestions about how they would improve the patient medication discharge process.

Statistical analysis. Power analy- sis indicated that 48 patients were needed in each group to achieve statistical signifi cance (two-sided p < 0.05) using α = 0.05 (two-sided) and β = 0.20. Primary outcome vari- ables were evaluated to determine whether or not they approximated a Gaussian distribution. Variability was assessed using the mean, median, and standard deviation. Finally, 95% confi dence intervals (CIs) were con- structed to assess whether the range of values included the true popula- tion values. For normally distributed data, parametric statistical tests were used. For study results not follow- ing Gaussian distribution, we used nonparametric statistical tests. Data were analyzed with GraphPad InStat (GraphPad Software, San Diego, CA). Patient survey data utilizing the Likert scale were compared using the Mann–Whitney U Test. Categorical U Test. Categorical U data were analyzed using Fisher’s exact test. Continuous variables were compared using the unpaired t test. t test. t Statistical signifi cance was set a priori at p < 0.05.

Results Patient characteristics. A total

of 283 patients were included in the study: 147 patients were enrolled during the preimplementation phase and 136 patients in the postimple- mentation phase. The study popula- tion consisted of patients with mul- tiple medical problems, including diabetes mellitus, asthma, COPD, cardiac conditions, and pneumonia, with an average length of stay of 5.7 days in the preimplementation phase and 5.5 days in the postimplementa- tion phase (Table 2).

Patient responses to most trigger questions did not signifi cantly differ between groups (preimplementation and postimplementation); however, significantly more patients in the postimplementation group were taking seven or more medications (p < 0.0001; 95% CI, 0.5284–0.7604) and had a history of coronary artery disease (CAD) (p < 0.0001; 95% CI, 0.3274–0.5444) (Table 2).

Patients in the postimplementa- tion group took signifi cantly more prescription and nonprescription medications, and their total number of medications signifi cantly exceeded the number taken by the preimple- mentation group (Table 3).

F e a s i b i l i t y a n d e f f i c i e n c y. Investigator-timed studies revealed that, in the postimplementation phase, pharmacists completed the admission medication history in 12.9 ± 9.34 minutes, clarifi ed medi- cations in 1.18 ± 5.84 minutes, and performed interventions in 1.4 ± 2.25 minutes. Pharmacists’ mean ± S.D. self-documented time to obtain patients’ admission history in the postimplementation phase was 16.3 ± 17.5 minutes. Prescribers com- pleted the admission and discharge medication reconciliation process for 78 patients (57.4%). Prescribers did not complete admission medication reconciliation for 10 patients (7.3%), whereas discharge medication rec- onciliation was not completed for 34 patients (25%). A total of 14 patients (10.3%) were discharged from an- other area of the hospital, left against medical advice, or died.

Pharmacist interventions. Types of interventions pharmacists com- pleted during the study included eliminating drug therapy duplication; correcting incomplete, missing, or incorrect dosage or frequency infor- mation; documenting allergies; and providing laboratory test and vac- cine recommendations. Pharmacists completed signifi cantly more dosage changes in the postimplementation phase than in the preimplementation

PRACTICE REPORTS Medication reconciliation

415Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

phase (Table 4). As expected, phar- macists in the preimplementation phase identifi ed no allergies, since they were not directly involved in obtaining the patient history; how- ever, 24 allergies in 17 patients were identified in the postimplementa- tion phase (p < 0.0001). A total of 50 retail pharmacies were called during the postimplementation phase to

provide patient medication infor- mation. Pharmacists completed 24 interventions in the preimplementa- tion phase and 48 total interventions in the postimplementation phase (p = 0.0003).

Nurse interventions. In the pre- implementation phase, nurses iden- tifi ed more incomplete medication orders, dosage changes, and allergies

than they did in the postimple- mentation phase (Table 4). Nurses completed 59 total interventions in the preimplementation phase and 27 total interventions in the postimple- mentation phase (p = 0.0003).

Time variables. Computerized allergy information availability was defi ned as the time from patient ar- rival on the unit until allergies were documented in the computer. The mean ± S.D. time required for nurses to enter allergies in the computer was signifi cantly longer in the post- implementation phase (141.1 ± 238.8 minutes versus 69.1 ± 98 minutes) (p = 0.0315). The time required for pharmacists to enter allergies de- creased in the postimplementation phase (64.1 ± 38.7 minutes versus 112.9 ± 70 minutes) (p < 0.0001). These changes occurred because the nurse was primarily responsible for the medication history in the preim- plementation phase, while the phar- macist was primarily responsible for the medication history in the post- implementation phase. Time from admission to trigger notifi cation did not signifi cantly differ between the groups. The mean ± S.D. time re- quired by pharmacists to initiate the admission medication history after receiving trigger notification was 18.8 ± 20.2 minutes (range, 1–140 minutes).

Mean ± S.D. no. prescription medications 4.9 ± 3.5 6.2 ± 4.3 0.0059 (0.3779–2.2222) Mean ± S.D. no. nonprescription medications 1 ± 1.6 2 ± 1.9 <0.0001 (0.5872–1.413) Mean ± S.D. no. herbal supplements 0.1 ± 0.34 0.1 ± 0.6 NS Mean ± S.D. total no. medications 6 ± 4 8.3 ± 5.2 <0.0001 (1.208–3.392) No. (%) patients receiving vaccinations 142 (97) 136 (100) NS

Table 3. Results of Admission Medication Reconciliation in Preimplementation and Postimplementation Groups

Postimplementation (n = 136)Variable

Preimplementation (n = 147) p (95% CI)a

aCI = confi dence interval, NS = not signifi cant.

Mean ± S.D. age, yr 64.4 ± 16 65.7 ± 17.6 Male, % 50 51 Mean ± S.D. height, cm 169.5 ± 18.3 169.2 ± 10.3 Mean ± S.D. weight, kg 81.8 ± 26.4 81.8 ± 24.3 Mean ± S.D. length of stay, days 5.9 ± 5.6 5.9 ± 6 No. (%) patients with admission

diagnosis Angioplasty 16 (11) 16 (12) Ankle swelling 55 (37) 52 (38) Arrhythmia 36 (24) 35 (26) Asthma 18 (12) 18 (13) Coronary artery disease 47 (32) 103 (76)b

Cardiac catheterization 44 (30) 45 (33) Congestive heart failure 26 (18) 32 (24) COPDc 10 (7) 12 (9) Diabetes mellitus 42 (29) 42 (31) Hypertension 81 (55) 75 (55) Myocardial infarction 27 (18) 27 (20)

Table 2. Characteristics of Patients in the Preimplementation and Postimplementation Groupsa

Characteristic Preimplementation

(n = 147) Postimplementation

(n = 136)

aUnless otherwise specifi ed, comparisons were not signifi cant. bp < 0.0001. cCOPD = chronic obstructive pulmonary disease.

PRACTICE REPORTS Medication reconciliation

416 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

Timed studies of selected study activities were completed for both study groups. Pharmacists completed the admission medication history in 12.9 ± 9.34 minutes, clarifi ed medi- cations in 1.18 ± 5.84 minutes, and performed interventions in 1.4 ± 2.25 minutes.

Readmissions. In the preimple- mentation phase, 17 patients were readmitted 18 times within 30 days after discharge, compared with 8 patients readmitted 8 times in the

postimplementation phase. In the preimplementation phase, 9 patients (10 visits) were seen in the emergency department (ED), and 12 patients in the postimplementation phase had a total of 19 ED visits. Neither compari- son yielded signifi cant differences.

Patient safety. Although attempts were made to determine potential medication errors, the effect of the medication reconciliation process on medication errors could not be determined due to the lack of in-

tervention documentation in both study phases. Category B and C medication errors were discovered during documentation of admission medication history and reconcilia- tion. Three medication errors were reported during the preimplementa- tion phase: two category B errors and one category C error. Four medica- tion errors were reported during the postimplementation phase: three category B errors and one category C error. In the postimplementation

Table 4. Number of Patient Interventions by Discipline

Postimplementation (n = 136)Intervention Type and Discipline

Preimplementation (n = 147) p (95% CI; RR)a

Incomplete medications Pharmacist 8 in 4 pts 4 in 4 pts NSa

Nurse 24 in 8 pts 6 in 4 pts 0.0016 (1.56 to 8.779; 3.701) Duplicate medications Pharmacist 5 in 2 pts 3 in 3 pts NS Nurse 5 in 2 pts 0 NS Route changes Pharmacist 0 0 NS Nurse 0 0 NS Dosage changes Pharmacist 5 in 3 pts 15 in 12 pts 0.0184 (0.1151 to –0.8259; 0.3084) Nurse 11 in 7 pts 0 0.0009 Adverse drug reactions Pharmacist 1 in 1 pt 1 in 1 pt NS Nurse 1 in 1 pt 0 NS Drug interactions Pharmacist 0 0 NS Nurse 0 0 NS Vaccination recommendations Pharmacist 0 0 NS Nurse 1 in 1 pt 0 NS Laboratory recommendations Pharmacist 0 0 NS Nurse 0 0 NS Allergies identifi ed Pharmacist 0 24 in 17 pts <0.0001 Nurse 10 in 5 pts 21 in 14 pts <0.0227 (0.3518 to 1.001; 0.5934) Allergies changed Pharmacist 5 in 2 pts 1 in 1 pt NS Nurse 1 in 1 pt 0 NS Incomplete allergies Pharmacist 0 1 NS Nurse 0 0 NS

aCI = confi dence interval, RR = relative risk, NS = not signifi cant.

PRACTICE REPORTS Medication reconciliation

417Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

phase, completed home medication reconciliation reports not scanned to the pharmacy for order-entry proc- essing contributed to the greatest number of medication errors. The completed home medication recon- ciliation reports were processed but not in a timely manner because of scanning delays.

Patient satisfaction survey re- sults. Results from the patient satis- faction survey are listed in Table 5. In the preimplementation phase, 63% of patients were contacted within 14 days after hospital discharge, compared with 50% in the post- implementation phase (p = 0.002), and agreed to participate in the tele- phone survey. The average length of time between discharge and patient contact was shorter in the postimple- mentation phase (p = 0.031). Patients seen during the postimplementation phase reported a signifi cantly higher level of agreement on all fi ve state- ments regarding adequate discharge medication instructions (Table 5). A total of 35 patients (8%) in the pre- implementation group remembered speaking with a pharmacist about medications while in the hospital, compared with 68 patients (63%) in the postimplementation group (p < 0.001). Interestingly, no additional questions regarding medications were asked during the telephone sur- vey in either group.

Health care provider satisfac- tion survey results. Results from the health care provider survey are provided in Table 6. Physician, nurse, and pharmacist responses for all questions did not signifi cantly differ between the preimplementation and postimplementation phases. Nurse practitioners and physician assistants reported decreased satisfaction with patient understanding of medica- tions at discharge and patient dis- charge instructions and did not feel that the discharge process was more effi cient in the postimplementation phase (n = 5) than in the preimple- mentation phase (n = 15). Ta

b le

5 .

P a

ti e

n t

S a

ti sf

a ct

io n

S u

rv e

y R

e su

lt s

S u

rv e

y It

e m

a n

d G

ro u

p n

M e

an ±

S .D

. R

at in

g a

M e

R an

k S

u m

o f

R an

ks z

S co

re z

S co

re z

W h

en I

w as

d is

ch ar

g ed

fr o

m t

h e

h o

sp it

al , I

w as

g iv

en c

le ar

st ru

ct io

n s

ab o

u t

w h

ic h

m ed

ic at

io n

s I w

as s

u p

p o

se d

t o

c o

n ti

n u

e ta

ki n

g a

t h

o m

re im

p le

m en

ta ti

o n

Po st

im p

le m

en ta

ti o

n I w

as g

iv en

c le

ar d

ir ec

ti o

n s

ab o

u t

h o

w m

u ch

a n

d h

o w

o ft

en I

su p

p o

se d

t o

t ak

e m

y m

ed ic

at io

n s.

P re

im p

le m

en ta

ti o

st im

p le

m en

ta ti

o n

I w as

g iv

en c

le ar

d ir

ec ti

o n

s ab

o u

t h

o w

a n

d w

h en

t o

t ak

e m

y m

ed ic

in e

(e .g

., to

t ak

e it

w it

h fo

o d

o r

o n

a n

e m

p ty

s to

m ac

h , i

n t

h e

m o

rn in

g o

n ly

o r

at b

ed ti

m e

o n

ly ).

P re

im p

le m

en ta

ti o

st im

p le

m en

ta ti

o n

I w as

g iv

en c

le ar

in fo

rm at

io n

a b

o u

t p

o ss

ib le

s id

e eff

e ct

s o

f m y

m ed

ic in

re im

p le

m en

ta ti

o n

Po st

im p

le m

en ta

ti o

n O

ve ra

ll, I

fe el

li ke

I u

n d

er st

an d

m y

m ed

ic in

es .

P re

im p

le m

en ta

ti o

st im

p le

m en

ta ti

o n

9 1

.3 5

± 1

.0 6

3 .1

6 6

5 6

.5 0

–2

.6 7

0 .0

0 7

6 8

.7 8

± 0

.4 5

9 .1

6 0

6 3

.5 0

9 0

.3 6

± 0

.9 8

2 .4

6 5

2 2

.5 0

–2

.6 8

0 .0

0 7

6 8

.7 3

± 0

.5 4

8 .8

6 0

3 8

.5 0

9 0

.8 9

± 1

.2 9

1 .4

6 4

3 2

.5 0

–2

.7 7

0 .0

0 6

6 8

.4 6

± 0

.8 2

0 .1

6 1

2 8

.5 0

9 0

.0 1

± 1

.4 1

9 .3

6 2

3 7

.5 0

–3

.3 1

0 .0

0 1

6 8

.7 5

± 1

.1 1

2 .9

6 3

2 3

.5 0

9 2

.2 2

± 0

.9 9

1 .4

6 5

7 3

.0 0

–3

.2 6

0 .0

0 1

6 8

.6 8

± 0

.5 3

2 .7

6 3

0 7

.0 0

a B as

ed o

n a

fi ve

-p o

in t

Li ke

rt s

ca le

, w h

er e

1 =

s tr

o n

g ly

d is

ag re

e an

d 5

= s

tr o

n g

ly a

g re

PRACTICE REPORTS Medication reconciliation

418 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

Ta b

le 6

. H

e a

lt h

C a

re P

ro v

id e

r S

a ti

sf a

ct io

n S

u rv

e y

R e

su lt

S u

rv e

y It

e m

a n

d D

is ci

p li

n e

p (9

5 %

C I)

Pa ti

en ts

a re

p ro

vi d

ed c

le ar

in st

ru ct

io n

s ab

o u

t w

h ic

h m

ed ic

at io

n s

th ey

a re

t o

c o

n ti

n u

e at

h o

m e

(h o

m e

m ed

ic at

io n

s an

d a

n y

n ew

d is

ch ar

g e

m ed

ic at

io n

s) .

P h

ar m

ac is

u rs

h ys

ic ia

h ys

ic ia

n a

ss is

ta n

t/ n

u rs

e p

ra ct

it io

n er

A t

th e

ti m

e o

f d is

ch ar

g e

fr o

m t

h e

h o

sp it

al , p

at ie

n ts

h av

e a

cl ea

r u

n d

er st

an d

in g

o f h

o w

u ch

, h o

w o

ft en

, a n

d w

h en

t o

t ak

e th

ei r

m ed

ic at

io n

s at

h o

m e.

h ar

m ac

is t

N u

rs e

P h

ys ic

ia n

P h

ys ic

ia n

a ss

is ta

n t/

n u

rs e

p ra

ct it

io n

M o

st p

at ie

n ts

r ec

ei ve

w ri

tt en

in fo

rm at

io n

a b

o u

t n

ew m

ed ic

at io

n s

p ri

o r

to d

is ch

ar g

P h

ar m

ac is

N u

rs e

h ys

ic ia

P h

ys ic

ia n

a ss

is ta

n t/

n u

rs e

p ra

ct it

io n

er Th

e p

at ie

n t

m ed

ic at

io n

d is

ch ar

g e

p ro

ce ss

is e

ffi c

ie n

t fo

r m

e as

a h

ea lt

h c

ar e

p ro

vi d

er .

h ar

m ac

is t

u rs

P h

ys ic

ia n

h ys

ic ia

n a

ss is

ta n

t/ n

u rs

e p

ra ct

it io

n er

I u se

t h

e M

ed it

ec h

c o

m p

u te

r sy

st em

o n

a r

eg u

la r

b as

is t

o v

ie w

p at

ie n

ts ’ h

o m

e m

ed ic

at io

n s.

P h

ar m

ac is

u rs

h ys

ic ia

h ys

ic ia

n a

ss is

ta n

t/ n

u rs

e p

ra ct

it io

n er

P re

im p

le m

e n

ta ti

o n

P o

st im

p le

m e

n ta

ti o

1 5

.6 ±

0 .8

1 2

.5 ±

0 .7

9 7

S 1

3 .6

4 ±

0 .9

1 0

.6 ±

1 .1

N S

1 9

.4 7

± 1

.2 2

3 .6

2 ±

0 .9

N S

1 5

.1 3

± 0

.6 4

.2 ±

0 .8

0 .0

1 7

(– 1

.6 7

8 t

o –

0 .1

8 2

0 )

1 5

.3 3

± 0

.4 9

2 .5

± 0

.7 9

N S

1 1

.2 7

± 0

.6 5

3 .9

± 0

.9 9

S 1

3 .3

2 ±

1 .2

1 3

.2 3

± 0

.9 3

S 1

4 .2

± 0

.6 8

.4 ±

0 .8

0 .0

4 8

(–

1 .5

9 4

t o

0 .0

0 5

9 7

)

1 5

.2 ±

0 .8

1 2

.5 ±

1 .2

N S

1 1

.8 2

± 1

.4 7

2 .7

± 1

.0 6

S 1

3 .4

7 ±

1 .0

1 3

.1 5

± 1

.0 7

S 1

3 .6

7 ±

1 .2

2 .8

± 0

.8 4

1 5

.2 ±

0 .9

1 2

.4 ±

1 .1

N S

1 1

.2 7

± 1

1 0

.2 ±

1 .0

N S

1 9

.8 4

± 1

.1 7

3 .4

6 ±

1 .0

N S

1 5

.9 3

± 0

.5 9

.6 ±

0 .5

0 .0

0 0

(– 1

.9 6

1 t

o –

0 .6

9 9

3 )

1 5

.8 ±

1 .3

1 2

.9 ±

1 .2

N S

1 1

.6 4

± 0

.9 2

3 .3

± 1

.1 6

S 1

2 .5

6 ±

1 .2

1 3

.6 9

± 1

.2 5

S 1

2 .6

4 ±

1 .4

2 .2

± 1

N S

a B as

ed o

n a

fi ve

-p o

in t

Li ke

rt s

ca le

, w h

er e

1 =

s tr

o n

g ly

d is

ag re

e an

d 5

= s

tr o

n g

ly a

g re

b C

I = c

o n

fi d

en ce

in te

rv al

, N S

= n

o t

si g

n ifi

ca n

n M

e an

± S

.D .

R at

in g

a n

M e

an ±

S .D

. R

at in

g a

PRACTICE REPORTS Medication reconciliation

419Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

Discussion The electronic medication rec-

onciliation documentation process developed and implemented at our institution could be adapted and implemented in other facilities. Requirements for successful im- plementation include adequately staffed pharmacist personnel to expand the reconciliation docu- mentation process hospitalwide, the ability of nurses and pharmacists to collaborate with prescribers, and the availability of technical support. Until computerized prescriber order entry is implemented at our institu- tion, the medication reconciliation process will require prescribers to handwrite changes on the printed reports.

Hospital admissions occur dur- ing all hours throughout the week. During our study, the majority of admissions occurred during the sec- ond shift, and we had not planned for study coverage to complete the rec- onciliation process during this time. To continue pharmacist-obtained medication histories and electronic discharge reconciliation, it would take at least two additional pharma- cists solely devoted to medication reconciliation to continue the pro- cess in limited areas of the hospital, such as the ED, for an annual total cost of $180,000 plus benefi ts.

Developing and implementing electronic medication reconciliation documentation was challenging. The greatest diffi culty encountered regarding data collection in the study was the lack of documentation of pharmacist and nurse interven- tions. A special screen to document interventions electronically was cre- ated for our CPCS for nurses, but the nurses had to add the documenta- tion screen for each patient to record interventions. Nurses continued to document interventions in a variety of locations for patients in the post- implementation phase, making it diffi cult for investigators to fi nd and retrieve information.

Pharmacists were instructed to use the usual electronic process to document interventions; however, multiple steps in the electronic medi- cation process often resulted in a lack of intervention documentation. Re- minders were sent, and one-on-one education was frequently provided. The complexity of the process likely detracted from the staff ’s willing- ness and ability to capture interven- tions. The fact that at least 50 retail pharmacies were called to obtain patient medication lists, yet only 48 individual interventions were docu- mented, indicates that the number of home medication clarifications was not accurately captured. In addi- tion, pharmacists interpreted calls to pharmacies as new orders rather than order clarifi cations, since a prescriber was not involved. Therefore, phar- macists did not include these calls in their interventions.

The time required for pharmacists to complete the medication history (12.9 ± 9.34 minutes) was similar to the time spent in a previous pilot study at our hospital (13.4 ± 6.7 min- utes).16 In the present study, the time pharmacists spent clarifying medi- cations (1.18 ± 5.84 minutes) and performing interventions (1.4 ± 2.25 minutes) was less than the 6 ± 6.5 minutes it took pharmacists to rec- oncile subsequent home medication orders in the pilot study.16 The time difference may have been attribut- able to timed evaluations in the cur- rent study versus self-documented time in the pilot study.

One investigator prepared the pharmacist schedule (26 full-time equivalents) for the department; pharmacists were scheduled to work in one-week intervals throughout the study to provide continuity. Pharmacists not routinely working on the study unit found it diffi cult to complete the reconciliation process without referring to fl ow diagrams or calling investigators.

Nurse and pharmacist commu- nication with physicians remained

challenging during the study. Phar- macists and nurses worked together to contact physicians to complete admission and discharge recon- ciliation. Although most physicians returned pages, not all calls were returned immediately. When calls were returned, physicians preferred to address admission medication reconciliation during rounds, so the reconciliation process was completed during the second shift of the day or the next day.

Discharge medication recon- ciliation. Discharge medication reconciliation was also diffi cult to complete. Both nurses and pharma- cists expressed frustration because many prescribers did not complete the discharge medication reconcili- ation report, even when prompted. Considerable time and effort were spent during admission to ensure the accuracy and completeness of the medication list. In addition, time was taken to develop, print, and place the discharge medication reconciliation form in the patient chart. A total of 34 patients (25%) in the postimple- mentation group did not have their discharge medication reconcilia- tion report completed by physicians before discharge, even though the report was fl agged in the chart.

Some physicians indicated that they appreciated the reports and the completeness of the medications, especially the discharge medication reconciliation report. By listing med- ications using AHFS classifi cation on the discharge report, physicians were able to easily visualize medication duplications and prevent medication errors at discharge. When physicians were asked why they did not com- plete the medication reconciliation form, several said they viewed the process as repetitive, since discharge prescriptions still had to be writ- ten. Initially, the investigators had intended to create and incorporate electronic physician-generated pre- scriptions. However, the hospital had already started to explore this op-

PRACTICE REPORTS Medication reconciliation

420 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

tion, so it was decided before study initiation not to include electronic prescription capability during the postimplementation phase.

Technical support. Technical sup- port is required for the electronic medication reconciliation process to work smoothly. A dedicated pharma- cy department computer program- mer spent many hours program- ming, testing, and resolving system problems to make implementation successful. Grant funding, together with hospital-matched funds, sup- ported the reconciliation develop- ment and programming. Each time the computer system was upgraded, computer problems would prevent reports from printing or cause them to print incorrectly. Sometimes after upgrades, pharmacists enter- ing home or discharge medications would be “kicked out” of the com- puter system entirely, losing all of their work. This wasted time, taking from a few minutes to one hour to reenter the lost data, and was frus- trating. Each problem was addressed rapidly during the study, but, for a hospitalwide system, dependability to complete the electronic process and funding to support the program cost are absolutely essential.

Patient participation and satis- faction. It is unknown what infl u- ence, if any, the difference in average time between discharge and contact- ing the patient may have had on the survey outcome since the average dif- ference was only approximately one day. Common reasons for not being able to contact patients during both study phases included patient death during hospitalization or after dis- charge, patients being transferred to other facilities (e.g., skilled-nursing facilities) before returning to their residence, incorrect patient phone numbers obtained at admission, and patients being too ill to communicate or simply not wishing to participate in the survey.

Overall, patients were very will- ing to participate in the study. They

voiced a realization of the impor- tance of health care providers and patients working together to ensure a complete medication review. Many times, patients stated that they wanted to participate because they had experienced medication-related diffi culties in the past or were being admitted with medication-related problems. Other patients had a fam- ily member admitted or discharged with medication-related diffi culties in the past.

Study limitations. Although phar- macists identifi ed more medications in the postimplementation phase, the results may have been skewed by two important triggers: the number of home medications on admission and the diagnosis of CAD. Possible rea- sons for this include the following: (1) patients with CAD tend to take a greater number of medications, (2) nurses preferred for pharmacists to complete medication histories for patients on multiple medications, so the trigger question asking about the number of medications taken was answered more frequently, and (3) pharmacists elected not to complete uncomplicated medication histories if already completed by the nurse or physician.

Our study required noninvestiga- tors, such as physicians, nurses, and pharmacists, to complete different steps of the medication reconcilia- tion process. As a result, documenta- tion was not always complete. More data may have been captured if our study allowed total investigator in- volvement in all steps of the process.

The admission reconciliation process was much easier to complete than the discharge reconciliation process. There was a lack of physician participation in the medication rec- onciliation discharge process. Also, patients anxious to leave the hospital sometimes chose not to wait for a fi nalized list of medications. Com- pleting reconciliation at discharge is a very important, but challenging, goal. Future studies should investi-

gate creative solutions for discharge reconciliation.

Conclusion Patients who had their medica-

tions electronically reconciled re- ported a greater understanding of the medications they were to take after discharge from the hospital, including medication administration instructions and potential adverse effects.

References 1. Rozich JD, Resar RK. Medication safety:

one organization’s approach to the chal- lenge. J Clin Outcomes Manag. 2001; 8:27- 34.

2. Bond CA, Raehl CL, Franke T. Clinical pharmacy services, hospital pharmacy staffi ng, and medication errors in United States hospitals. Pharmacotherapy. 2002; 22:134-47.

3. Pouyanne P, Haramburu F, Imbs JL et al. Admissions to hospital caused by adverse drug reactions: cross sectional incidence study. French Pharmacovigilance Cen- tres. BMJ. 2000; 320:1036.

4. Altman DE, Clancy C, Blendon RJ. Im- proving patient safety—fi ve years after the IOM report. N Engl J Med. 2004; 351:2041-3.

5. Bates DW, Spell N, Cullen DJ et al. The costs of adverse drug events in hospital- ized patients. JAMA. 1997; 277:307-11.

6. Caamano F, Pedone C, Zuccala G et al. Socio-demographic factors related to the prevalence of adverse drug reaction at hospital admission in an elderly popula- tion. Arch Gerontol Geriatr. 2005; 40:45- 52.

7. Classen DC, Pestotnik SL, Evans RS et al. Adverse drug events in hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA. 1997; 277:301-6.

8. Gandhi TK, Weingart SN, Borus J et al. Adverse drug events in ambulatory care. N Engl J Med. 2003; 348:1556-64.

9. Gurwitz JH, Field TS, Harrold LR et al. Incidence and preventability of adverse drug events among older persons in the ambulatory setting. JAMA. 2003; 289:1107-16.

10. Kennedy GA, MacLean CD. Clinical in- ertia: errors of omission in drug therapy. Am J Health-Syst Pharm. 2004; 61:401-4.

11. Malhotra S, Karan RS, Pandhi P et al. Drug related medical emergencies in the elderly: role of adverse drug reactions and non-compliance. Postgrad Med J. 2001; 77: 703-7.

12. Omori DM, Potyk RP, Kroenke K. The adverse effects of hospitalization on drug regimens. Arch Intern Med. 1991; 151: 1562-4.

13. Oren E, Shaffer ER, Guglielmo JB. Impact of emerging technologies on medication

PRACTICE REPORTS Medication reconciliation

421Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

errors and adverse drug events. Am J Health-Syst Pharm. 2003; 60:1447-54.

14. Pirmohamed M, James S, Meakin S et al. Adverse drug reactions as cause of admis- sion to hospital: prospective analysis of 18820 patients. BMJ. 2004; 329:15-9.

15. Peyriere H, Cassan S, Floutard E et al. Adverse drug events associated with hos- pital admission. Ann Pharmacother. 2003; 37:5-11.

16. Nester TM, Hale LS. Effectiveness of a pharmacist-acquired medication history in promoting patient safety. Am J Health- Syst Pharm. 2002; 59:2221-5.

17. Top-priority actions for preventing adverse drug events in hospitals. Recom- mendations of an expert panel. Am J Health-Syst Pharm. 1996; 53:747-51.

18. Pharmacy–nursing shared vision for safe medication use in hospitals: execu- tive summary session. Am J Health-Syst Pharm. 2003; 60:1046-52.

19. Institute for Safe Medication Practices. Building a case for medication recon- ciliation. www.ismp.org/newsletters/ acutecare/articles/20050421.asp (accessed 2006 Nov 15).

20. Institute for Safe Medication Practices. 2004 ISMP medication safety self as- sessment for hospitals. www.ismp.org/ selfassessments/Hospital/Login.asp (ac- cessed 2006 Nov 15).

21. Joint Commission on Accreditation of Healthcare Organizations. Using medication reconciliation to prevent e r ro r s . w w w. j o i n tco m m i s s i o n . o r g / SentinelEvents/SentinelEventAlert/sea_ 35.htm (accessed 2006 Nov 15).

22. Institute for Healthcare Improvement. 100k lives campaign: prevent adverse drug events. www.ihi.org/IHI/Programs/ Campaign/Campaign.htm?TabID=2# PreventAdverseDrugEvents (accessed 2006 Nov 15).

23. Beers MH, Munekata M, Storrie M. The accuracy of medication histories in the hospital medical records of elderly per- sons. J Am Geriatr Soc. 1990; 38:1183-7.

24. Kohn LT, Corrigan JM, Donaldson MS, eds. To err is human: building a safer health system. Washington, DC: National Academy Press; 1999.

25. Sihvo S, Klaukka T, Martikainen J et al. Frequency of daily over-the-counter drug use and potential clinically signifi cant over-the-counter-prescription drug in- teractions in the Finnish adult popula- tion. Eur J Clin Pharmacol. 2000; 56:495- 9.

26. Flaherty JH. Psychotherapeutic agents in older adults. Commonly prescribed and over-the-counter remedies: causes of confusion. Clin Geriatr Med. 1998; 14: 101-27.

27. Makaryus AN, Friedman EA. Patients’ understanding of their treatment plans and diagnosis at discharge. Mayo Clin Proc. 2005; 80:991-4.

28. Pedersen CA, Schneider PJ, Scheckelhoff DJ. ASHP national survey of pharmacy practice in hospital settings: dispensing and administration—2005. Am J Health- Syst Pharm. 2006; 63:327-45.

29. American Society of Health-System Pharmacists. Goal 1 of the 2015 Initiative. www.ashp.org/2015/Obj1.cfm (accessed 2006 Nov 16).

30. Montpetit LM, Roy MT. Evaluation of a patient-completed versus health profes- sional-conducted medication history. Drug Intell Clin Pharm. 1988; 22:964-9.

31. Duggan C, Feldman R, Hough J et al. Re- ducing adverse prescribing discrepancies following hospital discharge. Int J Pharm Pract. 1998; 6:77-82.

32. Michels RD, Meisel SB. Program using pharmacy technicians to obtain medica- tion histories. Am J Health-Syst Pharm. 2003; 60:1982-6.

33. Gurwich EL. Comparison of medica- tion histories acquired by pharmacists and physicians. Am J Hosp Pharm. 1983; 40:1541-2.

34. Badowski SA, Rosenbloom D, Dawson PH. Clinical importance of pharmacist- obtained medication histories using a validated questionnaire. Am J Hosp Pharm. 1984; 41:731-2.

35. Fertleman M, Barnett N, Patel T. Improv- ing medication management for patients: the effect of a pharmacist on post- admission ward rounds. Qual Saf Health Care. 2005; 14:207-11.

36. Vira T, Colquhoun M, Etchells E. Rec- oncilable differences: correcting medica- tion errors at hospital admission and discharge. Qual Saf Health Care. 2006; 15:122-6.

37. Gleason KM, Groszek JM, Sullivan C et al. Reconciliation of discrepancies in medi- cation histories and admission orders of newly hospitalized patients. Am J Health- Syst Pharm. 2004; 61:1689-95.

38. Rogers G, Alper E, Brunelle D et al. Rec- onciling medications at admission: safe practice recommendations and imple- mentation strategies. Jt Comm J Qual Saf. 2006; 32:37-50.

39. Rozich JD, Howard RJ, Justeson JM et al. Standardization as a mechanism to improve safety in health care: impact of sliding scale insulin protocol and rec- onciliation of medications initiatives. Jt Comm J Qual Saf. 2004; 30:5-14.

40. Rodehaver C, Fearing D. Medication reconciliation in acute care: ensuring an accurate drug regimen on admission and discharge. Jt Comm J Qual Saf. 2005; 31:406-13.

41. Form reconciles meds, but doctor buy-in diffi cult. ED Manag. 2006; 18(2):3-4.

42. Pronovost P, Weast B, Schwarz M et al. Medication reconciliation: a practical tool to reduce the risk of medication er- rors. J Crit Care. 2003; 18:201-5.

43. Ketchum K, Grass CA, Padwojski A. Med- ication reconciliation: verifying medica- tion orders and clarifying discrepancies should be standard practice. Am J Nurs. 2005; 105:78-9,81-2,84-5.

44. Chantelois EP, Norman TS. A pilot program comparing physician- and pharmacist-ordered discharge medica- tions at a Veterans Affairs medical center. Am J Health-Syst Pharm. 2003; 60:1652- 6.

45. Rolland P. Occurrence of dispensing er- rors and efforts to reduce medication errors at the Central Arkansas Veteran’s Healthcare System. Drug Saf. 2004; 27: 271-82.

46. Crosswhite R, Beckham SH, Gray P et al. Using a multidisciplinary automated discharge summary process to improve information management across the sys- tem. Am J Manag Care. 1997; 3:473-9.

47. Weeks G, Stanley L, Vinson MC. Automa- tion of the medication history process: a case report. Hosp Pharm. 2005; 40:1057- 61.

48. Bond CA, Raehl CL, Franke T. Clinical pharmacy services and hospital mortality rates. Pharmacotherapy. 1999; 19:556-64.

49. Foss MT, Panning CA, Broomfi eld JF. Medication history taking: techniques for a more productive patient–physician interaction. Prev Med Manag Care. 2002; 3:19-25.

50. Nickerson A, MacKinnon NJ, Roberts N et al. Drug-therapy problems, inconsis- tencies and omissions identifi ed during a medication reconciliation and seamless care service. Healthc Q. 2005; Spec No 8:65-72.

51. Okolo EN. Health research and design methodology. 1st ed. Boca Raton, FL: CRC; 1990.

PRACTICE REPORTS Medication reconciliation

422 Am J Health-Syst Pharm—Vol 64 Feb 15, 2007

Appendix—Defi nitions of data points for disease, intervention, and time documentation

Adverse drug reactions Asthma Congestive heart failure Dosage changes

Drug interaction

Duplicate medications Hypertension Incomplete medications Laboratory recommendation

Medication error

Number of allergies identifi ed Number of incomplete descriptions of allergies Number of medication histories clarifi ed with

pharmacist Number of medications Number of missing medications

Nurse’s time to enter allergies into computer

Pharmacist’s time to enter allergies into computer

Route changes Time from patient admission until trigger

notifi cation Time from trigger notifi cation until medication

history initiated Vaccine recommendation

Mild, moderate, or severe using institution’s current policy regarding adverse drug reactions Yes or no (obtained from nurse’s screening history) Yes or no (obtained from nurse’s screening history) Drug dosage increases or decreases based on laboratory test results (e.g., renal function, hepatic

function) or maximum and minimum recommended doses per the drug manufacturer An identifi ed problematic interaction that could possibly cause an adverse drug reaction if continued

(not morphine causes itching or a purposeful drug interaction) Duplications of therapeutic drug class Yes or no (obtained from nurse’s screening history) Missing information (drug name, strength, dosage schedule, time of last dose taken, indication) Drug therapy recommended based on laboratory values, laboratory values recommended as a result

of drug therapy, laboratory values recommended based on previous laboratory results Categories A through C classifi ed by National Coalition Council Medication Error Reporting

Program taxonomy; category A = circumstances or events that have the capacity to cause error, category B = an error occurred but did not reach the patient, category C = an error occurred and reached the patient but did not cause harm

Any additional allergies added to the patient profi le on admission; true allergies (not nausea) Any allergy listed without an explanation of what the allergic reaction was to the drug implicated Number of medications requiring a phone call to the patient’s retail pharmacist to verify drug,

strength, dosage, or reasons for use Include vitamins as nonprescription medications Additional medications the patient should be taking or was taking at home before admission identi-

fi ed by the nurse or pharmacist Time difference between when the nurse documented the information in the computer system and

when the patient was admitted to the unit Time difference between when the pharmacist documented the information in the computer system

and when the patient was admitted to the unit Intravenous to oral or oral to intravenous (or other routes) Time documented from the patient arriving on the unit to the time reported on the computer-

generated trigger notifi cation report Time from computer-generated trigger notifi cation to the time the clinical staff pharmacist begins

the patient medication history (postimplementation phase only) Any vaccine recommendation provided during the hospital stay

Data Point Defi nition