Purpose The results of the 2015 ASHP national survey of pharmacy practice in hospital settings are presented.
Methods A stratified random sample of pharmacy directors at 1432 U.S. general and children’s medical–surgical hospitals were surveyed using a mixed-mode method. IMS Health supplied data on hospital characteristics; the survey sample was drawn from IMS’s hospital database.
Results The survey response rate was 22.7%. Since the 2000 survey, the proportion of hospitals reporting that pharmacists monitor at least 75% of patients has increased from 20.3% to 57.8%. The use of therapeutic drug monitoring has increased from 63.0% to 70.6% since the 2012 survey. The percentage of hospitals reporting that pharmacists have primary responsibility for discharge counseling has increased from 1.2% to 7.3% since the 2012 survey, with 33.8% of hospitals reporting pharmacist counseling of at-risk patients. Virtually all hospitals (97.5%) have partially or completely implemented electronic health records; most have computerized prescriber-order-entry (84.1%) and barcode-assisted medication administration (93.7%) systems. At an increasing percentage of hospitals (86.2% in the 2015 survey), medication orders are reviewed by a pharmacist before a dose is made available and administered to a patient.
Conclusion The role of pharmacists in measuring, monitoring, and managing medication use in health systems continues to be significant, important, and growing. The evolution of electronic health information and technologies that make this information more readily available to patients is transforming healthcare in a positive way and enabling pharmacists to more efficiently contribute to improving medication use.
The ASHP national survey of pharmacy practice in hospital settings focuses on practices and technologies for managing and improving the medication-use system and the role that pharmacists play in this effort. The national surveys are organized according to six components of the medication-use system: prescribing, transcribing, dispensing, administration, monitoring, and patient education. The survey focuses on two components in the medication-use system each year. The 2015 survey evaluated practices and technologies related to monitoring and patient education. The most recent three surveys therefore represent a composite picture of the ways hospitals and health systems are managing and improving the entire medication-use system and current roles of pharmacists in medication-use system management.
In assessing monitoring and patient education practices, the 2015 survey provided data to describe current practices and trends in pharmacists’ medication therapy monitoring activities, inpatient therapeutic drug monitoring activities, drug therapy management activities, methods used to monitor and report adverse drug events (ADEs), antimicrobial stewardship activities, responsibility for medication reconciliation, patient education and counseling activities, implementation of medication-use system technologies, outpatient dispensing pharmacy operations, methods of providing specialty pharmacy products and services to patients, and use of key metrics to track pharmacy department performance.
The data gathered in the 2015 survey also describe human resource commitments, national vacancy rates for hospital pharmacist and pharmacy technician positions, acquisition costs of pharmaceuticals, and the scope of medication order review and entry.
The 2015 questionnaire was developed using procedures suggested by Dillman.15 Questions from previous surveys that pertained to topics of interest in this survey were evaluated for clarity and response. As with past surveys, data about hospital characteristics (i.e., number of beds, U.S. Census Bureau region, ownership, and U.S. Bureau of Census metropolitan statistical area status)16 were obtained from the IMS Health hospital database.17
From the IMS database of 7053 hospitals, a sampling frame of 4893 general and children’s medical–surgical hospitals in the United States was constructed. Specialty, federal, and Veterans Affairs hospitals were excluded from this sampling frame. Hospitals were stratified by size before sampling, and random samples of hospitals within these strata were taken to select the sample of 1439 hospitals. We sampled 300 hospitals with fewer than 50 beds to account for historically lower response rates in hospitals of this size. We sampled all hospitals with 600 or more staffed beds (n = 139) to collect data from enough of these very large hospitals to provide reliable estimates. Two hundred hospitals were sampled in each of the other hospital size categories. After eliminating closed hospitals, the adjusted sample was 1432 hospitals (Table 1).
The percentage of patients being monitored by pharmacists has increased significantly.
Pharmacists routinely provide medication management services in more than half of hospitals.
Pharmacists and pharmacy technicians are becoming more involved in patient education and medication reconciliation.
The rapid adoption of electronic health information is transforming medication-use systems.
A mixed-mode survey method was used. Respondents were offered a choice of completing a paper survey or an online survey (Qualtrics, Provo, UT). Pharmacy directors in the sample were contacted up to six times during the survey period. An announcement letter was sent in July 2015. This first contact letter explained the survey and directed respondents preferring to complete the survey online to the online data collection site. Directors opting for online data collection received e-mail reminders, whereas those not completing the online contact option received mailed paper surveys and reminders. Two weeks after the announcement letter was mailed, the first survey mailing was distributed. Two weeks after the initial survey mailing (in August 2015), respondents received a reminder postcard. The surveys was distributed a second time to nonrespondents one week later. The survey was sent a third time to the remaining nonrespondents at the end of August. A final telephone contact asking nonrespondents to complete the survey was made during September 2015.
Each hospital in the sample was assigned a unique identification number. This number allowed the survey response to be matched with the hospital characteristics in the IMS Health database.
As with past surveys, data are presented here by categories of staffed beds to more closely align with data from the American Hospital Association.18
Because of the stratified random sampling procedure, it was necessary to employ a design-based analysis.19 This technique results in population estimates that are more accurate than those derived from unweighted results.
Data were entered into SPSS, version 19 (IBM Corporation, Armonk, NY). Data were converted to an Intercooled Stata, version 8 (Stata Corp LP, College Station, TX)–readable format using DBMS Copy, version 7 (Conceptual Software Inc., Houston, TX). All non–design-based analyses were conducted using SPSS 19. All design-based analyses were conducted with Stata 8 using the set of survey commands. To account for the sampling method, weights were assigned to adjust for each respondent’s contribution to the population estimate. The assigned weights were as follows: 25.57 for hospitals with fewer than 50 staffed beds, 14.54 for hospitals with 50–99 beds, 27.39 for hospitals with 100–199 beds, 14.14 for hospitals with 200–299 beds, 8.14 for hospitals with 300–399 beds, 6.04 for hospitals with 400–599 beds, and 4.09 for hospitals with 600 or more staffed beds. The strata were the categories for number of staffed beds, and the finite population correction was the total number of hospitals in the population (4893).
Descriptive statistics were used extensively. Chi-square analysis and analysis of variance or regression were used to examine how responses differed as a function of hospital characteristics. The a priori level of significance was set at 0.05.
Overall, 228 individuals elected to receive the survey electronically; of those, 154 provided usable responses via online survey submission. Paper surveys were completed by an additional 171 respondents. A total of 325 hospitals submitted usable data for analysis. The overall response rate was 22.7%.
Table 1 presents data on the size, location, and ownership of the respondents’ hospitals, the nonrespondents’ hospitals, the surveyed hospitals, and the 4893 general and children’s medical–surgical hospitals. The characteristics of the surveyed hospitals are presented to highlight the complex sampling design employed in this survey. Respondent and nonrespondent institutions were statistically different in regional location.
Medication therapy monitoring
For the purposes of the survey, medication therapy monitoring was defined as monitoring that occurs after a patient receives a medication. Medication therapy monitoring activities included but were not limited to the following: monitoring therapeutic drug levels, monitoring patient outcomes, monitoring patient laboratory results, ADE monitoring, adjusting medication regimens due to organ function changes or changes in clinical status, and monitoring medication errors. Pharmacist activities performed before medication administration (e.g., identification of medication-selection issues, order review) are part of the prescribing process and were not included in the survey definition of monitoring.
Overall, 57.8% of hospitals had pharmacists monitor 75% or more of patients, 18.9% of hospitals had pharmacists monitor 51–75% of patients, 14.0% of hospitals had pharmacists monitor 26–50% of patients, and 9.3% of hospitals had pharmacists monitor less than 26% of patients (Table 2 and Figure 1). The proportion of hospitals having pharmacists monitor more than 75% of patients has increased over the past 16 years, from 20.3% in 200020 to 57.8% in 2015.
Identifying patients in need of monitoring
Overall, 30.8% of hospitals had pharmacists monitor all patients; 43.4% used computerized data-mining functionality within the hospital electronic health record (EHR) or clinical surveillance software (e.g., TheraDoc [Premier Inc., Charlotte, NC], Sentri7 [Pharmacy OneSource, Inc., Bellevue, WA], Med-Mined [CareFusion, San Diego, CA]) to identify patients in need of pharmacist monitoring, and 16.0% and 9.8% used an informal or formalized paper-based screening process to accomplish that task (Table 3). The survey findings in this area varied by staffed-bed count; smaller hospitals were the most likely to have pharmacists monitor all patients, and the largest hospitals were the most likely to use computerized data-mining functionality as a method to identify patients in need of daily monitoring by pharmacists (p < 0.0001).
Hospitals used many criteria to select patients for daily medication therapy monitoring by pharmacists. Overall, 75.8% of hospitals identified patients for monitoring using abnormal laboratory values that prompted dosage adjustments. The use of this practice varied significantly based on staffed-bed size; smaller hospitals were the least likely to use abnormal laboratory values as a trigger for monitoring (uncorrected χ2 = 30.6582, df = 6, design-based F[4.29, 1337.50] = 6.2347, p < 0.0001). For example, 58.8% of hospitals with fewer than 50 staffed beds used this method, whereas 90.6% of hospitals with 600 or more staffed beds used that method, and 79.1% of hospitals with 300–399 staffed beds used it. The use of abnormal laboratory values that prompt dosage adjustment as a method to identify patients for medication therapy monitoring has remained relatively constant since 2006.3,6,9
Furthermore, 66.9% of hospitals used a formalized list of medications to identify patients who require daily monitoring by pharmacists. The use of this practice varied significantly based on staffed-bed size; smaller hospitals were the least likely to use specific medications as a method to identify patients in need of daily monitoring by pharmacists (uncorrected χ2 = 69.7275, df = 6, design-based F[4.29, 1337.9] = 14.2194, p < 0.0001). For example, 38.2% of hospitals with fewer than 50 beds used such a list, whereas 87.5% of hospitals with 600 or more staffed beds used a list of medications, and 74.5% of hospitals with 50–99 staffed beds used a list of medications.
Patients were also identified for monitoring on the basis of disease state (47.5%), directives from a hospital committee (46.6%), care being provided by specific medical or surgical services (39.0%), and use of high-cost medications (36.2%). The practice of focusing on specific disease states as a method to identify patients for medication therapy monitoring has increased from 30.7% in 2012, whereas use of the other methods has remained relatively constant over the past three years.3 Larger hospitals have been more likely to use these methods.
Data collection for pharmacist monitoring
The most common process used to compile clinical information used for daily monitoring by pharmacists was data-mining functionality within the EHR (58.6%), followed by a manual process in which pharmacists abstracted pertinent data from the patient record (54.6%) (Table 4). Overall, 28.4% of hospitals had clinical surveillance software that assisted pharmacists with compiling clinical information. Relative to larger hospitals, smaller hospitals were more likely to use a manual process (p = 0.0074) and less likely to use data-mining functionality within the EHR (p = 0.0092) or clinical surveillance software (p < 0.0001).
Therapeutic drug monitoring
Overall, 94.6% of hospitals had pharmacists routinely monitor serum medication concentrations or their surrogate markers (Figure 2). In 82.3% of hospitals, pharmacists had authority by protocol to order initial measurement of a serum medication concentration, and 81.7% of hospitals allowed pharmacists to adjust the dosage of a medication being monitored. Over the past three years, therapeutic drug monitoring by pharmacists and their authority to order measurement of serum medication concentrations and to adjust dosages have remained stable.3
Drug therapy management
Drug therapy management was defined in the survey as a multidisciplinary team process for selecting appropriate drug therapies, educating patients, monitoring patients, and continually assessing outcomes of therapy. Pharmacist activities in drug therapy management may include but are not limited to the following: initiating, modifying, and monitoring a patient’s drug therapy; ordering and performing laboratory and related tests; assessing a patient’s response to therapy; counseling and educating a patient about medications; and administering medications.
Overall, 53.2% of hospitals routinely had pharmacists assigned to provide drug therapy management services to a majority of patients in the hospital at least eight hours per day, five days per week (Table 5 and Figure 3). The use of this practice varied significantly by hospital size; larger hospitals were more likely than smaller hospitals to have pharmacists assigned to provide drug therapy management services. The practice of having pharmacists routinely assigned to provide drug therapy management services to a majority of patients in the hospital for at least eight hours per day and five days per week has increased, from 34.2% in 2011 to 53.2% in 2015.1–4
Areas where pharmacists were most frequently assigned were medical–surgical units (43.5%), critical care units (43.5%), oncology units (37.5%), and cardiology units (32.9%). Rates of pharmacist assignment to these types of units varied significantly by hospital size; hospitals with 300 beds or more were more likely than smaller hospitals to have pharmacists assigned to specific areas and services to provide drug therapy management services.
Overall, 47.1% of hospitals had pharmacists actively involved in developing a plan of care with the interdisciplinary team for a majority of inpatients; the use of this practice did not vary by hospital size.
Methods to identify ADEs
Because operational definitions for ADEs vary, the following definition was provided: “An adverse drug event is an injury resulting from the use of, or not using, a needed medication. For the purposes of this survey, consider adverse drug events to include both adverse drug reactions and medication errors, including both errors of commission and omission that result in adverse clinical outcomes.”21
Hospital pharmacy directors were asked which methods pharmacists use to routinely monitor patients for ADEs. The most common methods were notification by nursing staff (86.6%), an ADE incident reporting system (84.1%), alerting orders or trigger medications (73.0%), therapeutic drug monitoring (70.6%), and routine review of laboratory values (67.2%). Less common methods were chart review (41.1%), an ADE hotline (26.5%), pharmacist rounds with physicians to assess ADEs (23.6%), pharmacist rounds without physicians (23.5%), patient counseling (21.7%), and medical record “e-codes” (18.9%). Larger hospitals were more likely than smaller ones to use each of these methods, with the exception of the use of alerting orders or trigger medications, notification by nursing staff, and chart review. Hospitals, regardless of size, used notification by nursing staff consistently as a method to monitor patients for ADEs. The use of all these methods remained stable or increased over the last three years. The use of therapeutic drug monitoring increased from 63.0% in 20123 to 70.6% in 2015 during that period, the use of routine review of laboratory values increased from 52.7% to 67.2%, and the use of the practice of identifying ADEs through patient counseling sessions increased from 11.7% to 21.7%.
Review of ADEs
Most hospitals (84.8%) had a multidisciplinary committee (including physicians, pharmacists, and nurses) responsible for review, analysis, education, policy formulation, and corrective action related to ADEs (Table 6). The smallest hospitals were least likely to have a multidisciplinary committee to review ADEs.
Overall, 53.1% of hospitals had conducted at least one prospective medication safety–related analytical process using a technique such as failure modes and effects analysis (FMEA) during the preceding year (Table 6). The use of this practice varied significantly with hospital size; larger hospitals were more likely than smaller hospitals to have completed an FMEA in the prior year. Overall, 71.2% of hospitals had conducted at least one retrospective medication safety–related root-cause analysis (RCA) in the previous year, with significant variation by hospital size; larger hospitals were more likely than smaller hospitals to have completed an RCA in the preceding year.
Over half (57.4%) of hospitals used a systematic quality-improvement process (e.g., Lean, Six Sigma) to make improvements to the medication-use system, with significant variation by hospital size; larger hospitals were more likely than smaller hospitals to use a systematic quality-improvement process.
The use of prospective FMEA has declined slightly since 20123 while the use of retrospective RCA has remained stable.
External ADE reporting
Overall, 83.6% of hospitals reported ADEs externally. The use of this practice varied significantly by hospital size; larger hospitals were more likely than smaller hospitals to report ADEs externally (uncorrected χ2 = 12.6402, df = 6, design-based F[4.28, 1321.20] = 2.6023, p = 0.0311). For example, 97.0% of hospitals with 600 or more staffed beds reported ADEs externally, as compared with 95.7% of hospitals with 400–599 beds, 86.0% of hospitals with 300–399 beds, 93.0% of hospitals with 200–299 beds, 81.6% of hospitals with 100–199 beds, 89.4% of hospitals with 50–99 beds, and 75.4% of hospitals with fewer than 50 staffed beds. In 2009, 60.7% of hospitals reported ADEs externally, as compared with 80.6% of hospitals in 2012.3,6
Antimicrobial stewardship programs
Overall, 65.5% of hospitals had an antimicrobial stewardship program (Table 7). The use of such programs varied by hospital size.
Pharmacists most commonly had a leadership and accountability role (57.3%) in their institution’s antimicrobial stewardship program; in substantial percentages of institutions, pharmacists had a clinical support role (26.6%) or a data analysis role (14.3%) (Figure 4). It was uncommon for pharmacists to not be actively involved in their institution’s antimicrobial stewardship program if one existed (1.8% of respondents reported lack of active involvement). The pharmacist role varied by hospital size; pharmacists were most commonly in a leadership and accountability role in larger hospitals, whereas in smaller hospitals, pharmacists most commonly provided clinical support to the antimicrobial stewardship program.
Various strategies were used in antimicrobial stewardship programs to influence and improve antimicrobial prescribing. The most frequently used strategy was conducting a daily review of targeted antimicrobials for appropriateness and contacting prescribers with recommendations for alternative therapy (86.9%), followed by education of prescribers using national or local guidelines (80.8%), restricting dispensing of targeted antimicrobials to approved indications (76.5%), and use of clinical decision support systems (disease-specific order sets or clinical surveillance software) that disseminate patient-specific recommendations during order entry (66.7%). Education and guidelines, formulary restrictions, and information technology strategies were more commonly used in larger hospitals.
The most common way hospitals assessed the impact of an antimicrobial stewardship program was by evaluation of resistance trends (79.0%), followed by evaluation of expenditures on antimicrobial agents (77.4%), compliance with evidence-based guidelines (77.0%), and the incidence of Clostridium difficile infection (76.6%); other commonly used strategies were reducing unnecessary antimicrobial therapy (75.3%) and evaluating antimicrobial utilization patterns (75.1%) and the rate of i.v.-to-oral conversions (67.3%). Less common ways included evaluation of infection-related readmissions (40.4%), infection-related length of stay (40.4%), and infection-related mortality (32.3%).
For those hospitals that assessed the impact of the antimicrobial stewardship program, 81.4% had documented improvements in adherence to evidence-based guidelines, 80.7% had improvements in the i.v.-to-oral conversion rate, and 79.6% had reduced unnecessary antibiotic therapy; 79.7% indicated success with evaluation of antibiotic utilization patterns, 76.0% with monitoring of antimicrobial costs, 68.6% with monitoring of resistance trends, 67.8% with monitoring of infection-related length of stay, 61.8% with monitoring of infection-related mortality, 57.3% with monitoring of the rate of C. difficile infection, and 50.7% with monitoring of infection-related readmissions.
Nurses had primary responsibility in about two thirds of hospitals for collecting medication regimen information at admission (Table 8). Physicians, pharmacists, and pharmacy technicians were not commonly responsible for collecting medication regimen information at admission. Responsibility for collecting medication regimen information on admission varied by hospital size; nurses in the largest hospitals were less likely than those at smaller hospitals to have responsibility for collecting medication regimen information at admission. In recent years, the reported rate of assignment of primary responsibility for collecting medication regimen information at admission decreased for nurses (from 79.4% in 2012 to 67.1% in 2015) and increased for pharmacists and pharmacy technicians (from 2.0% and 2.9%, respectively, in 2012 to 5.5% and 10.4%, respectively, in 2015).3
Nurses had primary responsibility in about four fifths of hospitals for disseminating medication regimen information at discharge (Table 8). Physicians and pharmacists were not commonly responsible for disseminating medication regimen information at discharge. Responsibility for disseminating medication regimen information at discharge varied by hospital size; nurses in the largest hospitals were less likely than those at smaller hospitals to have responsibility for disseminating medication regimen information at discharge. In recent years, the percentage of hospitals reporting that pharmacists had primary responsibility for disseminating medication regimen information at discharge has increased, from 1.6% in 2012 to 7.3% in 2015.3
Patient medication education and counseling
Overall, 33.8% of hospitals had pharmacists routinely conduct discharge medication counseling for at-risk patient groups (Table 9), with significant variation by hospital size; larger hospitals were more likely than smaller hospitals to have pharmacists provide discharge medication counseling to at-risk groups. Furthermore, 10.6% of hospitals routinely had pharmacists follow up with selected high-risk patient groups on appropriate medication use after discharge (e.g., through phone calls); use of this practice also varied by hospital size, with the largest hospitals most likely to conduct postdischarge follow-up.
In 35.5% of hospitals, pharmacists routinely provided either discharge medication counseling or postdischarge follow-up through telephone calls to at-risk patient groups; this figure was substantially higher than that documented in the 2011 survey (20.3%).4
Health information technologies
Overall, 97.5% of hospitals had partially or completely implemented an EHR (Table 10 and Figure 5). Fully 37.5% of hospitals had a complete EHR system, with no paper records; 60.0% of hospitals had a partial EHR, with some components still using paper; and 2.5% had an all-paper system and no EHR. From 2003 to 2015, the percentage of hospitals having paper-only health records declined from 69.4% to 2.5%, with a significant decline seen since 2012, when 18.5% of hospitals still reported the use of paper-only records.1–6,8,9,11,12 There has been a corresponding increase in the use of complete EHR systems over the last four years.
Overall, 84.1% of hospitals had computerized prescriber-order-entry (CPOE) systems (Table 10), a major increase from 2.7% of hospitals in 2003, with substantial increases reported in each of the last five surveys.1–12
Overall, 93.7% of hospitals had barcode-assisted medication administration (BCMA) systems to verify patient identity and electronically check doses administered by nurses (Table 10); that marks a dramatic increase from the 1.5% rate of BCMA use reported in 2002, with significant increases seen in each of the last five surveys.1–13
In 40.3% of hospitals responding to the 2015 survey, pharmacists routinely used a mobile device (a tablet computer or smartphone) while providing patient care (Table 11). The use of this practice varied significantly by hospital size; larger hospitals were more likely than smaller hospitals to have pharmacists using mobile devices. Activities that pharmacists using mobile devices most commonly performed included accessing drug information (93.6%), communicating with other healthcare providers (51.0%), accessing laboratory data (47.9%), reviewing or entering orders (43.6%), and documenting interventions (37.4%). Less commonly performed activities included medication reconciliation and transitions-of-care activities (24.4%), ADE reporting (23.6%), drug shortage monitoring (13.9%), receiving notifications regarding patients in need of pharmacist assessment (10.4%), and remote prescription counseling at discharge (3.1%). These findings varied significantly by hospital size; larger hospitals were more likely than smaller hospitals to have pharmacists using mobile devices for collection of laboratory data, documentation of interventions (including medication reconciliation and transitions-of-care activities), receiving notifications regarding patients in need of pharmacist assessment, ADE reporting, and remote discharge counseling.
Outpatient dispensing pharmacy
Overall, 27.7% of surveyed hospitals and health systems had an outpatient dispensing pharmacy, with significant variation by hospital size; larger hospitals were more likely than smaller hospitals to have an outpatient dispensing pharmacy (uncorrected χ2 = 67.6880, df = 6, design-based F[4.36, 1339.74] = 13.1373, p < 0.0001). For example, 87.9% of hospitals with 600 or more staffed beds had an outpatient dispensing pharmacy, as compared with 68.1% of hospitals with 400–599 beds, 63.6% of hospitals with 300–399 beds, 38.1% of hospitals with 200–299 beds, 24.3% of hospitals with 100–199 beds, 17.4% of hospitals with 50–99 beds, and 10.8% of hospitals with fewer than 50 staffed beds. The proportion of hospitals indicating that they have an outpatient dispensing pharmacy has remained stable since 2009.3,6
Overall, 70.2% of hospitals with outpatient dispensing pharmacies had the inpatient hospital pharmacy department operate the outpatient pharmacy, 19.9% had a separate outpatient division in the organization operate the outpatient pharmacy, and 11.9% outsourced operations to another company.
Among hospitals that had an outpatient dispensing pharmacy, 51.1% offered a discharge prescription service (e.g., a “meds-to-bed” service) through their outpatient pharmacy, 11.6% offered this service through a contracted community pharmacy, and 37.2% did not offer a discharge prescription service. Larger hospitals were more likely than smaller hospitals to offer a discharge prescription service (uncorrected χ2 = 33.7083, df = 12, design-based F[8.17, 988.43] = 2.0891, p = 0.0331). For example, 86.2% of hospitals with 600 or more staffed beds offered a discharge prescription service through the outpatient pharmacy, as compared with 71.0% of hospitals with 400–599 beds, 60.7% of hospitals with 300–399 beds, 43.8% of hospitals with 200–299 beds, 44.4% of hospitals with 100–199 beds, 50.0% of hospitals with 50–99 beds, and 14.3% of hospitals with fewer than 50 staffed beds.
The most common approach to providing specialty pharmacy products and services was to serve patients on a case-by-case basis, with no formal specialty pharmacy business model (83.7%), followed by outsourcing to an external specialty pharmacy (8.6%) and operating a health system–owned specialty pharmacy located within health-system pharmacy space (4.2%) or a standalone health system–owned specialty pharmacy (3.6%) (Table 12). These findings varied significantly by hospital size; compared with smaller hospitals, larger hospitals were more likely to provide specialty pharmacy products and services to patients through a health system–owned specialty pharmacy.
Overall, 69.7% of hospitals used key metrics (sometimes referred to as a dashboard or a balanced scorecard) to measure performance (Table 13). The use of metrics varied by hospital size. For example, 50.0% of hospitals with fewer than 50 staffed beds were using key metrics, compared with 76.1% of hospitals with 50–99 beds, 75.7% of hospitals with 100–199 beds, 85.7% of hospitals with 200–299 beds, 84.1% of hospitals with 300–399 beds, 87.2% of hospitals with 400–599 beds, and 84.8% of hospitals with 600 or more staffed beds.
At hospitals reporting the use of administrative and operational metrics, pharmacy departments tracked and monitored trends in financial metrics (84.2%); pharmacists’ clinical interventions (73.7%); workload and productivity measures (72.5%); dispensing and distribution measures, including turnaround times, missing doses, and dispensing errors (54.5%); inventory measures, including shortages, “stockouts,” and turnover (52.7%); and metrics recommended by the ASHP Practice Advancement Initiative (formerly called the Pharmacy Practice Model Initiative) (11.0%).
With regard to quality and outcomes metrics, pharmacy departments reported tracking and monitoring trends in Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) scores (64.5%), national quality goals (e.g., National Patient Safety Goals, Partnership for Patients Goals) (58.5%), patient satisfaction (50.3%), and 30-day readmissions (48.2%).
Hospitals reporting the use of medication safety metrics indicated that pharmacy departments tracked and monitored trends in ADEs or medication errors reported by staff or patients (91.9%), technology utilization (e.g., bedside bar-coding reports (86.4%), and data from the Institute for Safe Medication Practices Medication Safety Self-Assessment (51.0%).
Medication order review
Overall, 86.2% of hospitals responding to the 2015 national survey had a method for pharmacists to review and enter medication orders on demand (Table 14 and Figure 6). Having the pharmacy department open and staffed 24 hours a day, seven days a week was the most common method (41.6%), followed by using an affiliated hospital with 24-hour services (16.8%), having a telepharmacy company provide afterhours medication order review and entry (16.6%); 11.2% of hospitals had an employee pharmacist on call or at a remote location to provide these services, and 13.8% of hospitals reported that orders were not reviewed by a pharmacist when the pharmacy department was closed. Smaller hospitals were more likely than larger ones to not provide for order review when the pharmacy department was closed (Table 14). Regardless of the method used, the percentage of hospitals in which medication orders are not reviewed by a pharmacist has declined. The percentage of hospitals not reviewing orders afterhours has declined annually since 2005, when 59.6% of hospitals did not provide afterhours order review.1–5,7–10
Overall, 19.7% of hospitals reported providing remote order review to an affiliated hospital. This statistic differed significantly by hospital size, with 34.5% of the largest hospitals (600 or more staffed beds) providing the service, as compared with 41.3% of hospitals with 400–599 beds, 35.7% of those with 300–399 beds, 26.8% of those with 200–299 beds, 17.6% of those with 100– 199 beds, 22.7% of those with 50–99 beds, and 8.3% of those with fewer than 50 staffed beds (uncorrected χ2 = 20.3457, df = 6, design-based F[4.39, 1268.91] = 3.8348, p = 0.0031). The percentage of hospitals providing remote order review to an affiliated hospital increased from 14.6% in 2014 to 19.7% in 2015.1
Information about pharmacy operations collected through the survey provided useful data for pharmacy managers. However, caution should be exercised when reviewing and interpreting these data. The numbers presented are averages of available reported data. Every hospital offered unique services, and these data are not benchmarks or best practices.
Per definitions used in the national survey, inpatient pharmaceutical acquisition costs included all drug products, blood products, and diagnostic agents but excluded i.v. fluids and administration sets. Larger hospitals had higher inpatient and outpatient expenditures. The mean ± S.E. inpatient acquisition cost per patient-day was $155.79 ± $8.26. The mean ± S.E. inpatient acquisition cost per admission was $663.61 ± $36.11.
The number of pharmacist full-time equivalents (FTEs) per hospital averaged 11.4 and varied by hospital size (Table 15); the larger the hospital, the higher the number of pharmacist FTEs. The number of pharmacy technician FTEs averaged 11.0 and also varied by hospital size, with larger hospitals having a higher mean number of technician FTEs.
The average number of pharmacist FTEs per 100 occupied beds (average daily census) among all hospitals was 15.6 (Table 15); the average number of pharmacy technician FTEs per 100 occupied beds among all hospitals was 14.8. These statistics varied by hospital size, with the smallest hospitals having more pharmacist and pharmacy technician FTEs per 100 occupied beds.
A given hospital’s staffing levels can be compared with the national average for comparably sized hospitals by finding the average number of FTEs per 100 occupied beds for the corresponding size category, multiplying that number by the hospital’s average daily census, and dividing by 100.
The survey data indicate that at the time of the survey, 2.3% of pharmacist FTE positions and 5.1% of pharmacy technician FTE positions were vacant (Table 15); these percentages were calculated as the ratio of vacant to budgeted FTE positions overall and within each hospital size category.
Respondents to the 2015 survey indicated that 24.0% of inpatient hospital pharmacists had completed a postgraduate year 1 (PGY1) pharmacy residency, 5.9% had completed a postgraduate year 2 (PGY2) pharmacy residency, and 16.6% of inpatient hospital pharmacists had earned board certification through the Board of Pharmacy Specialties (BPS) (Table 16). Furthermore, 77.5% of pharmacy technicians were certified through the Pharmacy Technician Certification Board, and 17.5% had completed an ASHP-accredited technician training program. Relative to pharmacists working in smaller hospitals, greater proportions of pharmacists working in larger hospitals had completed PGY1 and PGY2 residencies and held BPS certification.
Staff time allocation
Pharmacy directors participating in the survey were asked to estimate current and future pharmacist and technician time allocations. It was estimated that, on average, 43.8% of pharmacists’ time was devoted to order review and verification, 18.2% to drug distribution, 23.6% to clinical functions, 8.5% to administrative management tasks, 5.6% to training residents and students, and 0.2% to other activities (Table 17). The reported allocation of pharmacists’ time has not substantially changed since 2010, when pharmacy directors reported that pharmacists spent 44.4% of their time on order review and verification, 17.9% on drug distribution, 24.4% on clinical functions, 7.7% on administrative management, 5.1% on the training of residents and students, and 0.5% on other activities.5 When asked about a desired future time allocation, directors indicated that they would like to see 33.7% of pharmacists’ time devoted to order review and verification, 13.6% to drug distribution, 36.0% to clinical activities, 8.4% to administrative management, 7.9% to training residents and students, and 0.5% to other activities; such a reallocation would represent an increased emphasis on clinical activities, with relative decreases in amounts of time spent on order review and verification and drug distribution.
The survey data indicated that, on average, pharmacy technicians devoted 3.2% of their time to order processing and entry, 77.5% to traditional drug preparation and distribution, 10.1% to nontraditional activities, 8.8% to administrative activities, and 0.5% to other activities (Table 17). The reported allocation of pharmacy technicians’ time has not changed substantially since 2010, when surveyed pharmacy directors reported that technicians spent 6.9% of their time on processing and entry, 76.6% on traditional drug preparation and distribution, 7.7% on nontraditional activities, 8.1% on administrative activities, and 0.6 on other activities.5 However, there has been a small increase in technician time devoted to nontraditional activities, with an associated decrease in time spent on order-entry activities. When asked about a desired future time allocation, directors reported that pharmacy technicians would ideally spend 5.5% of their time on order processing and entry, 64.9% on traditional drug preparation and distribution, 20.2% on nontraditional activities, 9.0% on administrative activities, and 0.4% on other activities; that reallocation would represent an increase in nontraditional activities and a decrease in time spent on traditional drug preparation and distribution.
Additional staffing-related data are available online in the ASHP resource centers (www.ashp.org/menu/PracticePolicy/ResourceCenters/Pharmacy-Practice-Managers/Business-Management).
The survey response rate was lower than in past years and has been declining in recent years despite changes made to make the survey easier for respondents. For example, the number of questions has been reduced by more than half since the 2012–14 surveys. An option to respond to the survey electronically was made available. Accepted methods for survey research continued to be employed, including contacting prospective respondents before, at the time of, and after distribution of the survey instrument. Despite the lower response rate, we feel the 2015 survey results are a valid reflection of the current state of medication-use management and the role that pharmacists play in that area.
The monitoring of drug therapy is important after a treatment decision is made. While this is a shared responsibility, pharmacists have an important role, and the results of the survey show that the percentage of patients monitored by pharmacists is increasing. During the past 12 years, the proportion of surveyed hospitals in which pharmacists were monitoring medication therapy in the majority of patients increased from 20.3% to 57.8%. The growth of electronic health information has likely contributed to this increase, as more than 40% of hospitals in the 2015 survey reported using computerized data-mining functionality to identify patients in need of pharmacist monitoring. Abnormal laboratory values that should prompt dosage adjustment were the most common trigger used.
A more comprehensive approach to improving medication use is drug therapy management. In addition to monitoring, this approach includes selecting appropriate drug therapies, educating patients, and continually assessing outcomes of therapy. It is impressive that more than half (53.2%) of hospitals reported having pharmacists routinely assigned to provide drug therapy management services at least eight hours per day, five days per week—a major increase from 34.2% in 2011.
Direct interaction between pharmacists and patients, including patient education and (more recently) medication reconciliation, has always been a challenge; nurses have often performed these activities because they are greater in number and their workplace is routinely closer to patients. Despite that, pharmacists and pharmacy technicians are becoming more involved in patient education and medication reconciliation. For example, while two thirds of hospitals participating in the 2015 survey reported that nurses were primarily responsible for collecting medication regimen information at admission, that figure had declined from 79.4% in 2012; in contrast, this is a growing area of responsibility for pharmacy staff, with the percentages of hospitals reporting primary roles for pharmacists and pharmacy technicians increasing from 2.0% to 5.5% and from 2.9% to 10.4%, respectively, during the same time frame. The proportion of hospitals indicating primary pharmacist responsibility for disseminating medication regimen information at discharge increased from 1.6% to 7.3% from 2012 to 2015. In the 2015 survey, a third of hospitals reported that pharmacists routinely conducted discharge medication counseling for at-risk patients.
The Health Information Technology for Economic and Clinical Health (HITECH) Act promoted the adoption and meaningful use of health information technology.22 The federal investment in fostering the implementation of electronic health information systems through the HITECH Act appears to have accelerated adoption of tools that can improve medication use. Almost all hospitals represented in the current survey had partially or completely implemented EHRs. Integral parts of electronic health information systems that have the potential to improve medication use are CPOE and BCMA systems. These two medication safety systems were in use in the majority of U.S. hospitals in the 2015 survey regardless of hospital size.
Another positive change resulting from electronic health information systems and related technologies is the increase in the percentage of medication orders that are reviewed by a pharmacist before a dose is available for administration to a patient. In the past, this required the presence of a pharmacist in the hospital and, ideally, a 24-hour pharmacy service. Prescriptions and clinical information about the patient are now available electronically so that medication order review can take place remotely. With the use of automated dispensing cabinets, access to a medication can be restricted until after a pharmacist has reviewed the prescription. As a result, the percentage of medication orders that are not reviewed by a pharmacist has declined dramatically, from almost 60% to less than 14%, in the past 10 years.
Mobile devices have the potential to facilitate access to information, improve communication among caregivers and with patients, enhance medication order entry and review, and streamline documentation of patient care activities. Pharmacists were using tablet computers and smartphones to provide patient care in 40% of hospitals represented in the current survey.
Specialty pharmacies constitute an emerging channel of the supply chain that grew out of the unique needs of transplant patients for expensive medications not widely available from pharmacies. The rapid growth in the use of costly medications, including biologicals, has resulted in the rapid growth of this new type of pharmacy. Hospitals are just getting started developing their strategies for distribution of specialty pharmaceuticals. More than 80% of hospitals in the 2015 survey did not have a formal specialty pharmacy business model. The most commonly used model was outsourcing of specialty pharmacy services to an outside vendor, but that practice was reported by less than 10% of hospitals. An even smaller proportion of hospitals had their own specialty pharmacy operating either within existing pharmacy space or as a standalone, health system–owned pharmacy. Those percentages could be expected to increase as the number of expensive medications in use increases and as hospitals and health systems develop strategies for managing these drugs.
The role of pharmacists in measuring, monitoring, and managing medication use in health systems continues to be significant, important, and growing. The evolution of electronic health information and technologies that make this information more readily available to patients is transforming healthcare in a positive way and enabling pharmacists to more efficiently contribute to improving medication use.
The assistance of Moyo Myers, M.S., Renee Barnes, the staff of ASHP, and the pharmacy directors who participated in the 2015 national survey is acknowledged.
The survey was partially supported by a grant from Merck & Co., Inc. The authors have declared no potential conflicts of interest.
An audio interview that supplements the information in this article is available on AJHP’s website at www.ajhpvoices.org. Readers can also access this interview through AJHP’s augmented reality (AR) feature by launching the Layar app and scanning this page with their mobile device.
- Copyright © 2016 by the American Society of Health-System Pharmacists, Inc. All rights reserved.