Safety risks with investigational drugs: Pharmacy practices and perceptions in the veterans affairs health system

Abstract

Objectives:

Rigorous practices for safe dispensing of investigational drugs are not standardized. This investigation sought to identify error-prevention processes utilized in the provision of investigational drug services (IDS) and to characterize pharmacists’ perceptions about safety risks posed by investigational drugs.

Methods:

An electronic questionnaire was distributed to an audience of IDS pharmacists within the Veteran Affairs Health System. Multiple facets were examined including demographics, perceptions of medication safety, and standard processes used to support investigational drug protocols.

Results:

Twenty-one respondents (32.8% response rate) from the Northeast, Midwest, South, West, and Non-contiguous United States participated. The mean number of pharmacist full-time equivalents (FTEs) dedicated to the IDS was 0.77 per site with 0.2 technician FTEs. The mean number of active protocols was 22. Seventeen respondents (81%) indicated some level of concern for safety risks. Concerns related to the packaging of medications were expressed, most notably lack of product differentiation, expiration dating, barcodes, and choice of font size or color. Regarding medication safety practices, the majority of sites had specific procedures in place for storing and securing drug supply, temperature monitoring, and prescription labeling. Repackaging bulk items and proactive error-identification strategies were less common. Sixty-seven percent of respondents reported that an independent double check was not routinely performed.

Conclusions:

Medication safety concerns exist among pharmacists in an investigational drug service; however, a variety of measures have been employed to improve medication safety practices. Best practices for the safe dispensing of investigational medications should be developed in order to standardize these error-prevention strategies.

Background

Unlike commercially available prescription products approved by the United States Food and Drug Administration (FDA), rigorous practices for the safe processing and dispensing of investigational medications have not been widely standardized [ISMP, 1997, 2007a, 2007b; Grissinger, 2011]. Examples of medication safety measures implemented to ensure safe dispensing of commercial products include the use of product differentiation strategies such as different colored packaging, tall man lettering, tablet imprinting for easy identification, use of varying shapes and colors, and unit of use packaging. In addition, to combat the risk for human error, often both a pharmacist and technician will be checking the medication. In some instances, this additional check may even include a second pharmacist performing an independent double check [ISMP, 2007a, 2007b; HOPA, 2014].

In contrast to commercial medications, differentiation of investigational drug products is sometimes less desirable. Both active and placebo tablets must appear similar to maintain subject and investigator blinding. In addition, outside packaging is often black and white and contains small print. Also, the outside packaging may at times be overloaded with information in one or more languages or, in contrast, contain incomplete labeling such as omission of the drug name, strength, or quantity. This presents the opportunity for potential errors in selection of an incorrect product in the dispensing to study subjects. Adding to this risk, dispensing of investigational drugs may only involve one pharmacist performing the entire process [ISMP, 2007b; Grissinger, 2011]. According to a survey conducted by Young and colleagues, storage considerations and packaging/labeling were identified by 21% and 20% of 299 respondents, respectively, as needing improvement by study sponsors [Young et al. 1984]. Unfortunately, the specific reasons needing improvement were not evaluated by the authors.

An Institute for Safe Medication Practices (ISMP) medication safety alert identified several potentially dangerous practices of study sponsors, including naming practices, drug labels, packaging, product appearance, and expiration dating [ISMP, 2007b]. Unsafe naming practices of investigational drugs may create the potential for miscommunication and errors in dispensing from indiscriminate use of similar or identical abbreviations [ISMP, 1997]. General recommendations for preventing errors due to these practices include promotion of safe storage (e.g. neat and organized placement on uncrowded, labeled shelves with restricted access to authorized personnel only), enhancement of prescription labeling, increased reporting of potential errors to study sponsors, and inclusion of pharmacists during the Investigational Review Board (IRB) approval process [ISMP, 2007b].

Procedures related to the provision of investigational drugs within a healthcare system are regulated by the FDA [FDA, 2013]. The Veteran Affairs (VA) Health System, specifically, also has additional policy guidance for the handling and dispensing of investigational drugs [VA, 2012]. While the FDA and VA publish regulations on the storage and handling of investigational medications, they do not provide insight on medication safety processes. Other statements do exist on the role of the pharmacist in the use of research medications, as well as best practice standards for hematology/oncology investigational drugs services (IDS) [HOPA, 2014; Wermeling et al. 1998]. The American Society of Health-System Pharmacists recommends an integrated approach to dispensing investigational drugs mirroring the medication-use practices involved in dispensing FDA-approved medications, such as pharmacy profile entry, labeling, and quality assurance typically used [Wermeling et al. 1998].

The Hematology/Oncology Pharmacy Association (HOPA) advises that policies and procedures pertaining to investigational drug management, storage, temperature, and expiration date monitoring should be developed, including standard operating procedures for instances when usual IDS staff are not present [HOPA, 2014]. Specifically, it is suggested to store drug supplies in a designated area, separate from commercial products, with medications separated and arranged according to dosing strength and protocol. Temperature monitoring for storage conditions should be tracked on a continuous basis and alarms should be in place to alert users of excursions. It is also advisable to incorporate the use of ordering templates, additional pharmacy labeling, and independent double checks during the verification process [HOPA, 2014]. Although the aforementioned recommendations exist for the handling and dispensing of investigational medications, it is not clear which, if any, of these practices are actually being implemented by investigational drug pharmacies. This study was conducted in order to assess whether or not these practices have been incorporated into the dispensing activities of investigational drug pharmacists. The objective of this investigation was to describe the site characteristics and processes employed by IDS pharmacists within the VA for safely dispensing study medications and to evaluate pharmacist perceptions about the safety risks posed by investigational drugs.

Methods

A survey-based electronic questionnaire was conducted to assess demographics, perceptions, and practices of VA sites dispensing investigational drugs. The survey was pilot tested among local clinical pharmacy staff, including individuals with prior experience supporting the investigational drug service, and revisions were made based on feedback received.

Once finalized, all VA pharmacists involved in dispensing investigational drugs were invited to participate in the survey and an email requesting participation in a national survey was distributed via listserv to research pharmacists within the VA Health System. Sites not utilizing a pharmacist for dispensing investigational drugs were excluded due to this mechanism for targeting survey participants. The message contained a link to the electronic questionnaire, which was supported by SelectSurvey.net (v4.102.015, ClassApps.com, Kansas City, MO, USA). The surveyors requested that only one response be completed per site in order to avoid duplication. The electronic questionnaire was available from 23 January to 20 February 2014. Reminder emails were sent to the listserv approximately 2 and 4 weeks after the initial communication from the investigator. This protocol was determined to be IRB exempt and was subsequently approved by the local Research and Development Committee.

The questionnaire consisted of 42 questions pertaining to three main areas of interest, including site demographics, VA pharmacist perceptions of medication safety issues, and pharmacy practices related to the dispensing of investigational drugs. Demographic information was collected, including staffing patterns, volume and types of protocols supported, hours of operation, and learner involvement in IDS. VA pharmacists were asked about their perceptions of the safety risks associated with investigational drugs. Finally, practices related to drug storage and security, dispensing, documentation, and error reporting were assessed.

Descriptive statistics were used to report data obtained from this investigation. Continuous data were reported as means with ranges. Nominal and ordinal data were reported as raw values with percentages where appropriate.

Results

There were 114 pharmacists contacted at 64 sites. Twenty-one responses were received within the 30-day period, which resulted in a 32.8% response rate. Demographics for the participating sites are shown in Table 1. In addition to supporting the investigational drug service, other pharmacy roles were fulfilled by 15 out of 21 (71.4%) responders. These included inpatient and outpatient staff pharmacists, anticoagulation, oncology, nutrition, drug information, home-based primary care, quality improvement, and others. Of the 21 sites participating in the survey, all were involved in investigator-initiated studies, 20 (95.2%) were involved in industry studies, 20 (95.2%) in VA cooperative studies, and 4 (19%) supported protocols other than these descriptors. Randomization of study subjects was the responsibility of the IDS pharmacist at 15 facilities (71.4%). Eighteen sites (85.7%) solely permitted pharmacists to dispense study medications, but a minority allowed dispensing by the physician, nuclear medicine staff, or study coordinator.

Table 1.

Site demographics (N = 21).^*

Geographic region
Midwest	5 (23.8%)
Noncontiguous	2 (9.5%)
Northeast	4 (19%)
South	8 (38.1)
West	2 (9.5%)
Mean pharmacist FTEs	0.77 (0.2–1.5)
Mean technician FTEs	0.2 (0–1)
Mean active protocols	22 (3–65)
Study subject type
Inpatient only	0
Outpatient only	3 (14.3%)
Both	18 (85.7%)
Learners
None	10 (47.6%)
IPPE students	4 (19%)
APPE students	5 (23.8%)
PGY1 residents	6 (28.6%)
PGY2 residents	1 (4.8%)
Hours of operation
Monday to Friday	12 (57.1%)
7 days per week	2 (9.5%)
24 h service	6 (28.6%)

^*Demographics represent all sites submitting this information, although one site did not complete the full survey.

APPE, advanced pharmacy practice experience; FTE, full-time equivalent; PGY1, postgraduate year one; PGY2, postgraduate year two; IPPE, introductory pharmacy practice experience.

Pharmacists rated their degree of concern regarding medication safety issues with investigational medications as none (n = 4, 19.05%), mild (n = 12, 57.1%), moderate (n = 4, 19.05%), and serious (n = 1, 4.8%). The significance of being proactive in preventing dispensing errors from occurring (i.e. before initiation of a new study) was rated very important (n = 15, 71.4%), somewhat important (n = 1, 4.8%), neutral (n = 0), somewhat unimportant (n = 0), and very unimportant (n = 5, 23.8%). Specific concerns related to the characteristics of packaging and labeling of study medications were collected and are summarized in Figure 1. These concerns included a lack of product differentiation, expiration dates, and barcodes. In addition, font size and color, amount of information, and use of sequential numbering were also cited. Sequential numbering refers to investigational drug packaging that is numbered consecutively rather than randomly. The risk is that one may be more likely to “mispick” an investigational drug if it is numbered sequentially rather than randomly (i.e. non-sequentially). For example, confirmation bias may be more likely to occur if five out of six numbers are similar and in the same order on a particular package.

Figure 1.

Packaging concerns. Multiple packaging concerns permitted for each responder, therefore total responses exceed total responders.

Pharmacists’ perception of IDS workload was assessed. The workload was characterized as acceptable for the time allotted by 71.4% (n = 15) of respondents. The remaining 28.6% (n = 6) felt the workload was too heavy. The level of agreement that time constraints prevent the pharmacist from dispensing study medications in a safe manner was strongly agree (n = 0), agree (n = 3, 14.3%), neutral (n = 3, 14.3%), disagree (n = 10, 47.6%), and strongly disagree (n = 5, 23.8%). The level of agreement that time constraints prevent thorough documentation of investigational drug dispensing was strongly agree (n = 0), agree (n = 5, 23.8%), neutral (n = 4, 19.05%), disagree (n = 8, 38.1%), and strongly disagree (n = 4, 19.05%).

Respondents were asked if study sponsors were receptive when safety concerns related to a protocol were voiced by the pharmacist. Seven sites had not reported safety concerns. Of the remaining 14, the level of agreement was strongly agree (n = 1), agree (n = 7), neutral (n = 3), disagree (n = 3), and strongly disagree (n = 0).

A series of questions were asked related to processes implemented in the dispensing of investigational medications. The mode for prescribing study medications was found to be a combination of paper and electronic platforms at 42.9% (n = 9) of sites; 19% (n = 4) exclusively used hardcopy prescriptions, and 38.1% (n = 8) solely utilized electronic prescribing.

The use of templates (i.e. with prefilled fields for dose, quantity, etc.) for ordering investigational drugs was assessed. There were five sites (23.8%) that reported use of paper templates for ordering investigational drugs. Each of these sites indicated that their paper templates included a prefilled entry for directions for use, and four of the five sites also incorporated predefined selections for dose and quantity in the template. Electronic templates were used by 11 sites (52.4%) and each included prefilled information for directions and dose in the electronic template. Ten out of the 11 sites also included a predefined quantity to dispense. Two out of the 11 sites incorporated refills in the template.

There were 17 sites (81%) which permitted dispensing of investigational drugs outside of IDS’s normal hours of operation. In order to disseminate study-specific information to off-hours staff, a combination of verbal, written, and electronically posted/email communication was used by all but one of the sites. One site indicated that it solely used verbal communication for relaying study-specific information to backup staff. Orientation of alternate staff responsible for covering IDS in the absence of the designated pharmacist was conducted in a formalized manner at six sites (28.6%), with the remainder participating in ad hoc, informal procedures. Fifteen of the facilities (71.4%) included dispensing guidelines with study protocol documents.

One survey participant did not complete the entire questionnaire, so the remaining results and percentages are based on responses from 20 participants. Methods for storing investigational drugs both prior to and after dispensing were assessed. Multiple responses pertaining to storage were permitted; therefore, a variety of methods could be employed at the 20 sites completing this assessment. Before dispensation, 13 sites (65%) used a separate pharmacy dedicated to investigational drugs. Locked cabinets within the main pharmacy were employed by five sites (25%), and open shelves away from commercial stock but still within the main pharmacy were used at another five sites (25%). A locked refrigerator or freezer was maintained by three of the responders (15%). After filling, but prior to release of the medication to the ultimate user, security measures for storing orders awaiting pickup were in place at 16 sites (80%).

Monitoring procedures for maintaining temperature conditions (e.g. refrigeration or freezing) and expiration dating were tabulated. Twenty sites (100%) stored refrigerated or frozen investigational products with alarmed devices used to track temperature excursions. A formal procedure for managing expiration dating was in place at 13 sites (65%).

Study participants were asked about various storage processes used to minimize the risk of selecting an incorrect product during dispensing. Options included dedicated shelves for individual drug studies, dedicated baskets for individual drug studies, separation of drugs according to look alike sound alike, and placement of additional labels on the packaging of study medications. Nineteen sites (95%) used a dedicated shelf for storing medication of a single study protocol. Drug supply was placed in devoted baskets for individual studies at six sites (30%). Separation of look-alike-sound-alike drugs and placement of additional auxiliary labels on the packaging of study medications for additional emphasis was reported by four sites (20%). Regarding bulk packaging (e.g. 10,000 tablets in one container), pharmacists were asked about measures taken to minimize the risk for errors (e.g. accidental contamination of entire drug supply if placebo tablets were erroneously poured back into active drug container or vice versa), such as repackaging into smaller units or storing overstock in a separate location. A total of nineteen sites (95%) stated that no processes had been incorporated. One site placed overstock in a separate storage area. A formal process for proactively identifying potential dispensing errors (i.e. before initiation of a study at a particular site) was in place at six sites (30%).

During the dispensing process, pharmacy-generated prescription labels were reported to always be affixed to bottles prior to dispensing at 18 sites (90%). One site stated it usually participates in this practice. Another site never participates in this practice.

Pharmacists were asked if an independent double check was regularly performed (defined as more than 75% of the time) by two individuals during the dispensing process. Only seven of the sites (35%) routinely perform this error prevention strategy. Figure 2 summarizes the major practices employed at the participating facilities.

Figure 2.

Medication safety practices.

In the event of an actual dispensing error, the occurrence would be reported to the IRB or study sponsor at 16 sites (80%). Other entities receiving communications include the primary investigator, study coordinator, Research and Development Committee, Safety Committee, Director of Pharmacy, and Pharmacy Safety Officer.

Discussion

The results of this survey attempt to characterize the current landscape of investigational drug service dispensing within the national VA Health System, with an emphasis on pharmacist perceptions of medication safety risks posed by investi-gational drugs and the current risk-mitigation strategies employed during the dispensing process. We found that most sites have fewer than one pharmacist and technician who are tasked with supporting an average of 22 protocols. While more than half operate only during normal business hours, almost one-third provide 24 h services for both inpatient and outpatient study protocols. Over 80% of sites surveyed do in fact permit dispensing outside of the service’s normal hours of operation.

Most pharmacists expressed a mild level of safety concerns related to the dispensing of investigational drugs; over two-thirds felt that the pharmacist workload was acceptable and disagreed that time constraints hinder safe dispensing practices. Most concerns were related to drug packaging and the most commonly cited characteristics were lack of product differentiation, expiration dates, and barcodes, as well as the font size and color selected. These results expand upon another national survey of pharmacy-led investigational drugs services, in which 21% of respondents indicated that packaging and labeling of study medication needed improvement by study sponsors [Young et al. 1984].

Some of the more common practices employed at VA sites include use of alarming systems for tracking temperature for storage conditions such as refrigeration and freezing, placing pharmacy labels on bottles prior to dispensing, and having security measures in place for storing investigational prescriptions prior to pick up. We were pleased to find that 90% of sites participated in the practice of always attaching pharmacy-generated labels, which conventionally include patient’s name, drug name, dose, directions for use, quantity, pharmacy name and phone number, prescriber, fill and discard dates. This high compliance rate is in contrast with the findings of a survey conducted in Michigan, in which date dispensed, name of study drug, directions for use, prescriber’s name, quantity dispensed, and expiration date were only included by 90.7%, 88.4%, 86.0%, 79.1%, 58.1%, and 55.8% of respondents, respectively [Cohen et al. 1989]. Important to note, however, is that the Michigan study was conducted 25 years before the current investigation, at a time when the role of dedicated staff in managing investigational drug dispensing was in its early stages of development. Since that time, processes and procedures have likely been refined.

The use of paper or electronic templates for ordering investigational medications occurred at 80% of the sites surveyed. While this percentage is reasonably high, it would be desirable to achieve an even higher rate of compliance with the use of preprinted order sets. These are important because they may help reduce the potential for medication errors as well ensuring that study protocols are followed [HOPA, 2014; Tamer and Shehab, 2006].

Less commonly employed practices at the sites surveyed included repackaging of bulk product into smaller units of use, a formal policy and procedure for proactively identifying potential errors in dispensing prior to initiation of a study protocol, and performance of an independent double check by two different individuals at the time of dispensing. One potential reason for not performing a double check could be the relatively low number of IDS FTEs at most institutions.

To our knowledge, this is the first survey characterizing specific dispensing practices and perceptions of investigational drugs conducted at a national level, and it is representative of each geographic region of the United States. There are, however, some limitations worthy of note. First, although the authors requested only one response per site, it was not possible to determine if duplicative responses were received since personally identifying information was not collected. Next, since ‘Best Practices’ for dispensing investigational drugs have not specifically been addressed elsewhere, some of the terminology used in the questionnaire may have been open to interpretation. In order to overcome this limitation, we attempted to control for ambiguity in language through local pilot testing prior to release. Use of standardized terminology, defined in a consensus statement, would help to improve communication among investigational drug pharmacists. Finally, the response rate was lower than expected, and the length of the survey could have precluded some from participating. Because we identified participants via a national listserv of IDS pharmacists, nonpharmacist providers engaged in the dispensing of investigational drugs were not represented in the results, nor were non-VA entities. Thus, it is possible that the landscape of IDS practice in nongovernment-funded medical centers may be different from what was found in this study (e.g. larger staff size, higher number of studies different policies and procedures related to dispensing, etc.). Our results may be less generalizable to this group; however, we feel this study is a step toward initiating dialogue of the IDS-specific medication safety practices being employed, and it lays the groundwork for additional research into such practice settings. In addition, this study touches upon some universal safety principles that are foundational to any IDS practice, and these concepts could be considered applicable to both new and established investigational drug services.

Conclusion

It was identified that most research pharmacists had some level of concern regarding safety risks associated with dispensing investigational drugs. A large amount of variation in practices utilized across sites was observed. Consensus recommendations for ‘Best Practices’ in the investigational drug service should be standardized and disseminated in order to improve communication among pharmacists and to improve the safety and quality of services provided.