Abstract
Background
Computerized provider order entry (CPOE) has been developed and implemented within cancer center hospitals nationwide in Japan. To ensure that high-quality services are routinely provided by oncology pharmacists, this study was designed to evaluate the interventions through reviewing the orders that are generated by CPOE.
Methods
The objective of this retrospective chart review was to evaluate how pharmacists contributed to safe cancer treatment using paper-based pharmacy records. Data were collected from a total of 35,062 chemotherapy regimens for 18,515 outpatients between January and December 2013.
Results
Of these 35,062 chemotherapy regimens, the rate of pharmacists’ interventions was 1.1% (n = 408). Among them, 53.1% (217/408) of the chemotherapy prescriptions were modified due to pharmacist interventions. The reasons for interventions included “changes in the chemotherapy regimen were unclear” in 49.5%, “physicians’ prescription errors” (22.0%), “pharmacist suggestions to improve chemotherapy” (15.1%), and “finding differences between physicians’ chemotherapy records and their chemotherapy prescriptions” (13.2%). The top three reasons for the 217 prescription modifications due to pharmacist interventions were “finding prescription errors” (34.5%), “reasons for change in the chemotherapy regimen were unclear” (32.7%), and “finding differences between physicians’ chemotherapy records and their chemotherapy prescriptions” (28.5%).
Conclusion
The computer could not evaluate chemotherapy protocols or adjust doses of anticancer medicines according to patients’ conditions. Therefore, oncology pharmacists should continue to ensure safe and appropriate administration of cancer chemotherapy.
Introduction
A medication error is defined as any error that arises in the process of medication use. 1 Within oncology pharmacy practices, pharmacists play a substantial role in reviewing and checking chemotherapy-containing prescriptions, because chemotherapy agents are inherently toxic, thus highlighting the need to be vigilant to prevent medication errors. With the use of technology in contemporary clinical practice, such as computerized provider order entry (CPOE), there has been a sharp reduction of medication errors. 2 The benefits of CPOE for safe cancer chemotherapy administration have been previously reported in Japan.3,4 However, prescribers may not favor the usage of CPOE for various logistical reasons. 5
In Japan, CPOE has been developed and implemented within cancer center hospitals nationwide. At the National Cancer Center Hospital East (NCCHE), the division of pharmacy has introduced CPOE to ensure safe administration and to validate cancer chemotherapy since 2003. The CPOE system provides template formats which are used for anticancer drugs and any premedication and supportive drugs. The CPOE template is programmed for checking the dose and anticancer medicines and the chemotherapy interval. The drug protocols are developed by oncologists who submitted their chemotherapy regimens to the pharmacy division. The pharmacy division has a role to evaluate the validity of the chemotherapy. After the reviewing process, the pharmacy division builds the protocol into the order template in the CPOE system. After the template is approved, the oncologist only needs to choose the name of a chemotherapy regimen to order the chemotherapy on each patient’s electronic medical record (EMR). The CPOE utilizes weight and height to calculate dose of chemotherapy regimens and the template provides the sequence for anticancer agents, along with supportive medicines and hydration, which oncologists have authorized for the specific protocol. The final templates are approved by a committee that is comprised of oncologists and pharmacists. CPOE also allows oncologists to select their choice of anticancer drugs and infusion bags, calculate drug dosages and administration rates, include appropriate supportive medicines, and schedule the administration of these drugs accordingly. An EMR and an electronic medication administration record (eMAR) are also incorporated in order to enable medical professionals to check critical components. In addition, the computerized system provides print labels and prescriptions that are easier to read and thus safer than handwritten prescriptions.
To ensure that high-quality services are routinely provided by oncology pharmacists at NCCHE, this study was designed to evaluate the interventions that were conducted by Japanese oncology pharmacists by reviewing the orders that were generated by the CPOE.
Methods
Study design
The current study was a single-center, retrospective review of pharmacy records at the NCCHE between January and December 2013. The data do not contain patients’ personal information, since information was recorded by the pharmacists in such a manner that subjects were de-identified. Therefore, this study was exempted from full review by the institutional review board.
Settings
The NCCHE is one of two national center hospitals for cancer treatment in Japan. It has a leadership function and plays an important role in cancer medicine. The NCCHE, a 425-bed cancer-specific hospital, serves about 1000 inpatients and approximately 20,000 outpatients per month. The NCCHE pharmacy division currently registers about 400 chemotherapy regimen templates on the CPOE. Among the 10 pharmacists at NCCHE, four are routinely involved with chemotherapy checking. At the NCCHE, a board-certified oncology pharmacist must perform the final review in the chemotherapy dispensing process.
Workflow
All chemotherapy agents are prescribed through CPOE. The pharmacy division used the IBM MIRACL® (until the beginning of May 2013) and the Fujitsu MISSION® systems (after May 2013) as their CPOE. The CPOE provides chemotherapy templates, which consist of registered anticancer drugs and any premedication and supportive drugs. The CPOE provides checking of the dose and anticancer medicines and the chemotherapy interval. The system also provides printed prescriptions and labels; the pharmacy division does not accept handwritten anticancer prescriptions for risk management purposes.
In addition to the computer system, the division has also created a “paper-based pharmacy record” (Figure 1) to document a pharmacist’s evaluation of the chemotherapy order. This paper-based pharmacy record is a tool that supports an oncology pharmacist’s chemotherapy evaluation. The paper-based pharmacy record’s evaluation items are: (1) dose of anticancer medicine; (2) interval; (3) indication for the regimen; (4) administration rates; and (5) premedication, hydration, and supportive medicines (Figure 1). In addition to the check using the record, a pharmacist suggests appropriate bags for each anticancer injection along with administration rates and routes for the regimen and enters the information in the eMAR and EMR. The paper record is an independent tool from the EMR/eMAR. The pharmacists check the chemotherapy using the paper tool and information from the EMR/eMAR. The information originates from the CPOE and EMR/eMAR, however, the checking process is not performed digitally. The hand writing paper-based pharmacy check record allows the pharmacists to follow in the treatment progress each chemotherapy cycles. As described in Figure 1, the paper record is a tool to summarize the chemotherapy history. Before chemotherapy is compounded, a pharmacist prepares the medicines and also checks the chemotherapy regimen prescriptions using the paper-based pharmacy record. A second pharmacist checks the paper-based pharmacy record before the chemotherapy is prepared. After the completion of the double check, the prescription review process in the pharmacy division is considered completed (Figure 2).
An example of a paper-based pharmacy record (FOLFOX + bevacizumab). Workflow.
Definitions and endpoints
Interventions by pharmacists
The pharmacy division recorded all pharmacy interventions, which included order clarifications or recommendations made by pharmacists. The record consisted of the prescribing date, name of the clinician, contents of the intervention, and whether or not the clinician changed the prescription. The incidence of pharmacists’ interventions and the incidence of prescriptions that required modification due to the pharmacists’ interventions were evaluated. These endpoints were chosen because the aim was to capture the number of interventions required by pharmacists through reviewing the prescriptions that are generated by CPOE.
Reasons for pharmacists’ interventions
In order to categorize the interventions, the reasons for interventions were classified into categories depending on the purposes of the interventions. These included: “reasons for change in the chemotherapy regimen were unclear”, “prescription errors”, “pharmacist suggestions to improve chemotherapy”, and “differences between physicians’ chemotherapy records and their chemotherapy prescriptions”.
Pharmacists’ interventions
The types of interventions were classified into categories depending on the focus of the interventions. These included: “dose of anticancer drugs”, “adding/stopping anticancer drugs”, “premedication/supportive medicine”, “anticancer drug dose less than 95% calculated”, “interval”, “chemotherapy regimen choice”, “administration routes/rates”, “bag of anticancer drugs”, “dose according to weight change”, “dose rounding”, and “suggesting laboratory test”. “Adding/stopping anticancer drugs” was defined as interventions that a pharmacist found a physician added or stopped anticancer drugs in a chemotherapy regimen without reasonable reasons. “Chemotherapy regimen choice” was defined as interventions that a pharmacist found a wrong chemotherapy regimen prescription for a patient. “Bag of anticancer drugs” was defined as interventions that a pharmacist suggested an appropriate infusion solution for chemotherapy agents in a chemotherapy regimen prescription. “Dose according to weight change” was defined as interventions in which a pharmacist suggested recalculation due to significant weight change from a previous chemotherapy regimen prescription. “Dose rounding” was defined as interventions in which a pharmacist found a rounding error, such as an anticancer drug dose less than 95% calculated.
Data analysis
Descriptive statistics were used to examine the mean number and frequency of prescription modifications due to pharmacy interventions. All calculations were performed using Microsoft® Excel 2010.
Results
Number and frequency of chemotherapy prescription modifications due to pharmacists’ interventions
A total of 35,062 chemotherapy regimens for 18,515 patients were analyzed during the 12-month study period. The average number of chemotherapy administrations was 2922 ± 168 [±S.D.] per month. The average number of patients was 1543 ± 85 [±S.D.] per month. It was noted that the number of pharmacist’s interventions was different each month. In particular, there were a significantly higher number of inquiries in May 2013 compared to other months (Figure 3). This is because the replacement of the IBM MIRACL® system with Fujitsu MISSION® took place that month. During that period, clinicians and pharmacists were troubleshooting the new CPOE system, and the paper-based pharmacy record played a significant and integral role in checking and reviewing chemotherapy administration.
Number of chemotherapy regimens, number of pharmacist inquiries, and modification rates.
Of the 35,062 administered chemotherapy regimens, the proportion of orders requiring clarification was 1.1% (total 408 pharmacists’ interventions). Among the 408 interventions, the proportion of interventions accepted by prescribers was 53.1% (217/408).
Reasons for interventions
Pharmacist’s interventions (n = 408).
Pharmacists’ interventions
Among the 408 pharmacists’ clarifications or recommendations that were made over the study period, the subjects of pharmacists’ interventions included: “anticancer drug doses, n = 179 (43.9%)”, “premedication/supportive medicine, n = 78 (19.1%)”, “adding/stopping anticancer drugs, n = 57 (14.0%)”, and others, including “dose rounding”, “interval of chemotherapy”, “chemotherapy regimen choice”, “administration routes/rates”, “chemotherapy administration”, “bag of anticancer drugs”, “dose according to weight change”, “dose rounding”, and “suggesting laboratory test”. Of 408 pharmacy clarifications or recommendations, “anticancer drug doses” (43.9%), “adding/stopping anticancer drugs” (14.0%), and “chemotherapy regimen choice” (2.9%) were obviously critical, and their cumulative total was more than half of the pharmacists’ interventions.
Frequency of chemotherapy prescription modifications due to pharmacists’ interventions
Types of prescriptions in which interventions were successful (n = 217).
Discussion
In this retrospective 12-month chart review study, of the 408 documented interventions, the rate of interventions accepted by prescribers was 53.1% (217/408). More than half of the interventions were related to critical issues, such as choice of chemotherapy agents and their doses. Although not all pharmacists’ interventions were accepted by doctors, most of the inquiries were related to usage or dose of anticancer agents. More than half of the interventions for oncologists were due to insufficient descriptions of their cancer chemotherapy prescription. In fact, of the total 408 interventions, 84.8% were for critical issues, “reasons for change in the chemotherapy regimen were unclear”, “physicians’ prescription errors”, and “finding differences between physicians’ chemotherapy records and their chemotherapy prescriptions”. For critical issues, they are generally referred to the ISMP International Medication Safety Self Assessment® for Oncology. Even though modifications occurred in only 51.1%, the pharmacists’ intervention clarified the reasons for the remaining 169 (48.9%) prescriptions. The number of interventions required for each month ranged widely from 6 to 83. The number of interventions was especially high, particularly when the NCCHE introduced a new CPOE system. All pharmacy interventions were for physicians’ chemotherapy prescriptions ordered via a sophisticated CPOE system. The results showed there were certain errors in the prescriptions that were prevented by pharmacists’ review before chemotherapy. Previously, the pharmacy division of NCCHE reported the effects of clinical pharmacist interventions,6–8 but those reports were about patients’ pharmaceutical care after the chemotherapy prescriptions. The results of this study showed that mistakes are inevitable in physicians’ prescriptions, even though CPOE was used to prescribe chemotherapy. Numerous reports concluded that pharmacist interventions are beneficial in clinical settings, but the time to review prescription processes is generally limited and medication errors are potentially fatal.9–16
Of all of the pharmacists’ inquiries, approximately 20% were pharmacist suggestions to improve chemotherapy by referring to patients’ previous chemotherapy, and approximately 80% were inquiries regarding the validity of the oncologists’ chemotherapy prescriptions. Our goal is to have sufficient checks and verifications, so all errors are corrected before chemotherapy is administered to the patient. Chemotherapy regimen prescription errors included the doses of anticancer drugs, intervals, adding or stopping anticancer drugs, chemotherapy regimen choice, and administration routes or rates, and these errors could have fatal consequences. 17 It is important to note that CPOE does not prevent all types of prescription errors, even with simple prescribing mistakes, because the computer system is unable to identify or judge the validity of the chemotherapy regimens. CPOE is only able to perform checks on basic items, such as dose calculation and the interval of chemotherapy. The CPOE system is able to recognize dose calculations and standard administration for each anticancer drug, but it is still impossible for computers to evaluate cancer treatment according to adverse drug reactions, patients’ conditions, and treatment strategies of cancer chemotherapy, because these factors are not countable and measureable. In the present study, even though the prescription modification rate was not high, interventions were still required with prescriptions that were ordered with CPOE. This suggests that CPOE was unable to prevent components that required medical or pharmaceutical judgement, such as the indications for anticancer agents, dose modification of chemotherapy, and the strategic plan for cancer treatment. For example, adding or reducing anticancer agents and changes in regimens require reasonable and logical decisions by oncologists and pharmacists. Moreover, the computer could not check the validity and accuracy of dose reductions, interval changes, and the choice of chemotherapy, even though the computer order entry system regulated the dose and interval of cancer chemotherapy.
There are several limitations in this study. The study showed how pharmacists picked up prescription errors using paper-based pharmacy records and reduced prescription errors, but the study could not show the actual number and percentage of errors prevented by CPOE. In addition, this was a retrospective study, and the data might be under-reported because the pharmacist may fail to recall some interventions they made when they completed their reports. This study did not investigate whether certain characteristics can predict the occurrence of errors. Future studies should investigate whether certain characteristics are able to early identify problems within the dispensing process in order to improve the efficiency of the dispensing process.
Conclusions
This is the first study to evaluate interventions that were conducted by Japanese oncology pharmacists by reviewing the orders that are generated by CPOE. CPOE could not evaluate chemotherapy protocols or adjust doses of anticancer medicines according to patients’ conditions. Therefore, pharmacists who check prescriptions need to be knowledgable of the chemotherapy to conduct safe cancer chemotherapy. Data from this study suggest that prescriptions ordered through CPOE are not foolproof, and oncology pharmacists should continue to ensure safe and adequate administration of cancer chemotherapy.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.