• Comparison of Hospital Admission Medication Histories with Actual Outpatient Medication Regimens

      Glover, Jon; Salek, Ferena; Khunkhun, Sanjeev; Krase, Jeff; Rowen, Derek; College of Pharmacy, The University of Arizona (The University of Arizona., 2006)
      Objectives: To assess the accuracy of medication histories taken upon hospital admission. Methods: This study was a prospective chart review comparing the accuracy of hospital admission medication histories with outpatient pharmacy and physician records. Eleven hundred patients admitted to El Dorado Hospital were eligible for participation. Patients were excluded if they were: under 18 years old, non-English speaking, in a room with precautions, transferred from a hospital or skilled nursing facility, unable to give oral consent, or admitted to a geri-psychiatric unit. After verification of medication, dose, and schedule with outpatient pharmacy and physician records, the findings were documented as perfect agreement, error of omission, or error of commission. Results: One hundred and nineteen patient histories were analyzed. Forty three patients (36%) had at least one error in their medication history. Of the 582 medications verified for accuracy, 491 medications were accurate and 91 errors were identified. Of the errors identified, 64 were errors of commission and 27 were errors of omission. The average age of patients who had at least one error type was 74.9 years, and the average age of patients who did not have any errors was 68.3 years (p = 0.004). Medication histories with six or more medications were more commonly associated with errors (p = 0.001). There was no relationship between medication history errors and specific classes of medications such as cardiovascular, lipid-lowering, and antidiabetic agents. Conclusions: Medication histories for patients older than 75 years or those that contain six or more medications are more likely to contain an error. There appears to be no link between medication class and occurrence of error. While this study does not address methods to reduce error rates, verification of medication histories with outpatient pharmacies and physicians may help reduce hospital medication errors and promote positive health outcomes.
    • A Comparison of Pharmacy Student Intern and Nurse Impact on Home Medication List Completeness During Medication Reconciliation

      Calkins, Linda M.; Salek, Ferena; Michael Ivey; Calkins, Linda M.; Salek, Ferena; College of Pharmacy, The University of Arizona (The University of Arizona., 2014)
      Specific Aims: The purpose of this quality improvement project was to compare the completeness of home medication lists generated upon hospital admission between pharmacy student interns and nurses. Methods: This project was a retrospective review of completed home medication lists obtained by pharmacy student interns or nurses in a Southern Arizona community hospital. During August and September 2013, medication lists from the previous day’s admissions were collected and de-identified. Medication lists were included in the evaluation if the patient was admitted directly to the hospital or through the emergency department, stayed for at least 24 hours and had at least one home medication upon admission. The primary outcome was the number of omissions left on home medication lists completed by pharmacy student interns or nurses. An omission was defined as any missing information in the medication list categories of drug name, dose, unit, route or frequency. Main Results: Fifty medication lists that included 519 medications were collected in the pharmacy student intern group and forty-four lists that included 376 medications were collected in the nurse group. Of the total medications, nurses left significantly more omissions in the categories of dose (19% vs. 1.9%), units (20.2% vs. 2.3%), and frequency (11% vs. 0.7%), where the P-value was < 0.05 for each. Lastly, the total number of omissions left by nurses compared to pharmacy student interns was significantly different (201 vs. 35 omissions, P < 0.05). Conclusion: Compared to nurses, these results suggest pharmacy student interns left fewer omissions and created a more complete home medication list for patients being admitted to the hospital.
    • Description of a Pharmacy Technician and Student Intern-Driven Medication Reconciliation Process and Evaluation of Medical Provider Acceptance of Recommendations to Reorder Critical Medications

      Salek, Ferena; Hall, Edina; Glover, Jon; Hall, Scott Thomas; Salek, Ferena; Hall, Edina; Glover, Jon; College of Pharmacy, The University of Arizona (The University of Arizona., 2011)
      OBJECTIVES: To describe a pharmacy technician and student intern-driven medication reconciliation process and to evaluate medical provider acceptance of recommendations to reorder critical medications. METHODS: Patients admitted to Northwest Medical Center had medication histories taken on admission. A specially trained pharmacy technician or student intern reviewed these histories, with emphasis placed on critical medications as defined by the Pharmacy and Therapeutics Committee. Recommendations to re-order these critical medications were made to medical providers. All patients, excluding those under 18 years of age or current enrollment in the prison system, admitted during the months of May-June 2010 were reviewed for acceptance of critical medication recommendations through information recorded in the pharmacy electronic medical record system. RESULTS: One hundred seventy-eight (178) recommendations were made on 132 patients requiring recommendations. All medical providers accepted 102 (57%, p-value=0.008) of the recommendations made. Hospitalists were more likely than physician specialists or surgeons to accept recommendations made (62.5%, p-value<0.001). Recommendations made regarding thyroid products were accepted the greatest majority of the time (82.1%, p-value<0.001); antidepressants (54.8%, p-value=0.321), anticonvulsants (63.2%, p-value=0.194), and medications classified as other (55.6%, p-value=0.480) were also accepted a majority of the time. Vitamin K antagonists did not have recommendations accepted a majority of the time (31.8%, p-value=0.034). CONCLUSION: Medical providers accepted a majority of recommendations to reorder critical medications made by pharmacy technicians or student interns.
    • Effect of Adherence to the GOLD Guidelines on Chronic Obstructive Pulmonary Disease Related Readmissions in a Community Hospital

      Hall, Edina; Salek, Ferena; Glover, Jon; Binder, William; Clark, Scott; Hall, Edina; Salek, Ferena; Glover, Jon; College of Pharmacy, The University of Arizona (The University of Arizona., 2016)
      Objectives: To assess the relationship between adherence to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines for the management of chronic obstructive pulmonary disease (COPD) exacerbations and the corresponding 30-day, all-cause readmissions rate in a community hospital. Methods: A retrospective chart review was conducted on patients admitted with the primary diagnosis of a COPD exacerbation. Medications administration records relevant to the GOLD guidelines were examined as separate independent variables in relation to a readmission within 30 days of discharge. Additional factors examined included: demographic data, resident of a long-term care facility, pre-index hospitalization, pulmonary consult, vaccines, length of stay (LOS), discharge medications, and follow-up appointments. Results: Electronic health records of 120 patients were reviewed and divided into non-readmitted patients (n = 65, mean age 73.4 ± 10.1 years), all-cause readmissions (n = 55, mean age 70.15 ± 9.69 years), and COPD-related readmissions (n = 21, mean age 70.7 ± 11.1 years). Patients with heart failure (p = 0.024), a LOS >5 days (p = 0.045), a pre-index hospitalization (p = 0.001), or who were long-term care residents (p = 0.024) experienced more all-cause readmissions. Females experienced less all-cause readmissions (p = 0.035). Significantly more patients with a pre-index hospitalization had a COPD-related readmission (p = 0.027). Lastly, adherence to the GOLD treatment parameters was not significantly different across all groups. Conclusions: COPD is a complex disease and adherence to the GOLD guidelines during an exacerbation is unlikely to significantly impact 30-day readmission rates.
    • An Evaluation of Outcomes Related to the Use of a VTE Prophylaxis Assessment Tool in Obstetrical Patients at Northwest Medical Center

      Sullivan, Christopher; Salek, Ferena; Calkins, Linda; Smith, Katherine; Smith, Karen; Blanco, Hayde; Kelly, Melissa; Romero, John; Wilkerson, Aaron; College of Pharmacy, The University of Arizona (The University of Arizona., 2018)
      Specific Aims: To evaluate the relationship between standardized venous thromboembolism (VTE) risk assessment, subsequent use of pharmacologic prophylaxis and outcomes related to VTE and bleeding. Subjects: Pregnant women who were admitted to NMC Women’s Center from January 1, 2017 to October 1, 2017 to give birth and who consented to a post-discharge phone call. Methods: An in-house risk assessment instrument was used to measure an individual’s risk of VTE on admission and post-delivery. Six weeks post-discharge, consented eligible patients were contacted via phone to determine VTE occurrence outpatient or inpatient, VTE prophylaxis occurrence, and experienced outpatient adverse drug events (ADE). All data were recorded in a data collection spreadsheet and analyzed. Main Results: A total of 707 women were included in the study. Most patients had a total risk score of zero (antepartum [AP]: 74.4%, postpartum [PP]: 51.9%). Seventeen (2.4%) AP patients and 72 (12.8%) PP patients scored high risk (defined as total score ≥ 4). Of the 17 high risk AP patients, 8 received postpartum prophylaxis. Of the 72 high risk PP patients, 13 received postpartum prophylaxis. There were no reported ADEs associated with any prophylactic drug therapy. One patient experienced a deep vein thrombosis incident in the six weeks postpartum period and zero in the antepartum period. Conclusions: Among the participants who received a VTE risk assessment pre- and postpartum, one participant experienced a VTE during the six week postpartum window. Additionally, there was no notable difference in ADEs in patients receiving pharmacological prophylaxis despite increased use.
    • Glycemic Control in Hospitalized Type 2 Diabetic Patients Receiving Sliding Scale Insulin

      Salek, Ferena; Glover, Jon; Bates, Amy; Collier, Kathleen; College of Pharmacy, The University of Arizona (The University of Arizona., 2005)
      Objectives: The purpose was to describe the glycemic control of hospitalized patients treated with sliding scale insulin (SSI) and correlate glycemic episodes to other factors. Methods: The records of 315 patients admitted to a private, 166-bed hospital between August 25 and November 30, 2004 were identified from orders for antidiabetic medications then screened for inclusion and exclusion criteria. 135 of these patients met the criteria for inclusion in this study and their charts were reviewed for demographic data, medical history, medication regimens, blood glucose measurements and SSI use during hospitalization. Hypoglycemia was defined as less than or equal to 50 mg/dL and hyperglycemia greater than 250 mg/dL. Results: Orders for SSI were written for 93% of diabetic patients admitted. 2,904 blood glucose fingerstick measurements were recorded: 15.9% were greater than 250 mg/dL and 0.689% were under 50 mg/dL. The only statistically significant result was the correlation between increased numbers of hyperglycemic episodes and the consumption of the hospital’s “diabetic diet,” p<0.001. The small group of patients admitted for cellulitis (N=6) also experienced more hyperglycemic episodes. There was a trend approaching significance, p=0.055, for an increased number of hyperglycemic episodes in patients with admission blood glucose value over 200 mg/dL. Results based upon the hospital’s standard SSI regimen were not significantly different from other variations of SSI. Implications: SSI was almost always prescribed for hospitalized patients with type 2 diabetes and this regimen resulted in poor glucose control in approximately 17% of fingerstick measurements.
    • An Inpatient Multidisciplinary Educational Approach to Reduce 30-day Heart Failure Readmissions

      Salek, Ferena; Malhotra, Kyle; Salek, Ferena; College of Pharmacy, The University of Arizona (The University of Arizona., 2016)
      Objectives: An estimated 5.7 million Americans had heart failure (HF) in 2012 with an economic cost of $30.7 billion. By 2030 the prevalence of the disease is expected to increase by 46%. Centers for Medicare and Medicaid Services penalizes hospitals for 30-day readmissions. This study evaluated the effect of our multidisciplinary HF intervention on readmissions. Methods: This is a retrospective cohort study. Patients were identified from electronic inpatient admission records from January 1 to December 31, 2014. Patients who received any component of intervention were compared to patients who did not receive any intervention. Intervention included student pharmacist medication counselling, HF education, and post-discharge phone calls with Modified Morisky questionnaire. Age, sex, admission/discharge dates, readmission diagnosis, smoking status, ejection fraction, medications, and Charlson Comorbidity Index (CCI) conditions were collected. Results: A total of 221 patients with 249 discrete admissions were identified. No difference in age (p=0.42), sex (p=0.48), smoking status (p=0.10) existed between the groups. No difference in readmissions was found between patients receiving complete intervention and control (p=0.41) or patients receiving 1 or 2 intervention components and control (p=0.41). Patients with CCI score≥ 8 had greater risk of readmission compared to CCI scores 0-2 (OR 7.7, 95% CI 1.6-36.3, p=0.01). Conclusions: This analysis did not identify an intervention impact on 30-day readmissions in patients with HF; high CCI scores were associated with increased readmission risk. The intervention may be best targeted towards patients with high CCI scores as they have the highest readmission rate.
    • Predictive Value of a Medication Adherence Screening Tool on Hospital Readmission Rates in Patients with Congestive Heart Failure

      Hall, Edina; Salek, Ferena; Glover, Jon J.; Felix, Serena; McGowan, Veronica; Hall, Edina; Salek, Ferena; Glover, Jon J.; College of Pharmacy, The University of Arizona (The University of Arizona., 2013)
      Specific Aims: To examine the relationship between hospital readmission rates and responses to a medication adherence questionnaire (Morisky) in patients with congestive heart failure (CHF). Methods: The Morisky questionnaire, assessing medication adherence, was administered to all CHF patients admitted from September 15, 2012 to March 7, 2013. Information collected from the electronic medical record (EMR) for all patients with complete Morisky questionnaires included: age, sex, ethnicity, insurance, height, weight, marital status, tobacco use, alcohol use, number of home medications, all-cause and CHF admission in the previous 365 days from when the questionnaire was administered as well as the following events/disease states: myocardial infarction, hypertension, atrial fibrillation, stroke, diabetes mellitus, peripheral vascular disease, chronic obstructive pulmonary disease, congestive heart disease and chronic kidney disease. Main Results: Of the 120 patients enrolled, 52% scored 1-5 on the Morisky questionnaire indicating some problem with medication adherence while 48% scored 0 (no problems). There was no correlation between the Morisky score and age (95% CI: -3.3-5.7), number of medications (95% CI: -0.26, 2.85), or number of comorbidities (95% CI: -1.02,0.03). The Morisky questionnaire was not predictive of all cause readmissions (95% CI: 0.35, 2.01) p = 0.691). For CHF readmissions the Morisky score was not significant (95% CI: 0.6, 4.11, p=0.358) but the confidence interval suggests a trend. Conclusion: There is no correlation between Morisky scores, age, comorbidities, and medication number. Readmission rates were not predicted by Morisky scores; with more participants a trend may be detected for CHF readmissions.