Electronic Prescribing: An Examination of Cost Effectiveness, Clinician Adoption and Limitations

E-prescribing has flourished due to the promise of improving efficiency and decreasing prescription errors caused by its handwritten counterpart, yet only 44% of doctor’s offices use paperless prescriptions. Many studies have assessed the value of e-prescribing in the clinical setting, yet there is no all-inclusive summation of these findings. The aim of this study was to review the literature within the last decade pertaining to the cost effectiveness, clinician adoption and limitations of e-prescribing. Journal articles from January 1, 2003 through May 1, 2013 were compiled through use of the search engines: PubMed, International Pharmaceutical Abstracts (IPA), and Google Scholar. A total of 56 peer-reviewed articles were included in this review. Trends show that most clinicians view e-prescribing positively despite some limitations. Limitations of e-prescribing include cost of implementation and maintenance, insufficient training, and lack of standardization. The cost to implement an integrated system with alerts and decision support is $29,000 per physician for the first year and $4,000 every year thereafter. Furthermore, e-prescribing has little disruption to workflow and no statistically significant increase in processing time. Although limitations exist, an expansion of e-prescribing is expected in the future. Efforts should be increased to promote clinicians adopting e-prescribing.


Introduction
Electronic prescribing (or e-prescribing) is the electronic transmission of prescriptions or prescription-related information between a prescriber, dispenser, pharmacy benefit manager, or health plan [21].In the last 10 years, the use of e-prescribing has flourished due to the promise of improving efficiency and decreasing prescription errors caused by its handwritten counterpart, yet according to 2012 estimates, only 44% of doctor's offices use paperless prescriptions [1,2,3].However, this is a drastic increase from 12% utilization in 2009.While there was initial opposition towards e-prescribing due to perceptions of increased costs, training requirements, and time constraints, many studies have been published to address those concerns.For example, many clinicians felt that the use of e-prescribing would disrupt workflow and take more time.However, it has been shown that e-prescribing actually saves time and can be implemented into the workflow with proper training [4,5].The purpose of this review was to discuss e-prescribing as a preferred form of prescribing and its current impact.With the Medicare Modernization Act of 2003 pushing for e-prescribing and many other insurers following suit, this article summarizes the present state of e-prescribing and its future utilization [6].In addition, the importance of clinician acceptance of electronic prescribing and methods to improve integration are discussed.Clinical pharmacists have an opportunity to detect prescribing errors and improve prescription quality through the use of electronic prescribing.Furthermore, electronic prescribing has been used to monitor and assess patient compliance and prescription discrepancies [23].Several studies show that electronic prescribing has been shown to reduce prescribing errors in United States hospitals [49,50,51,52,53,54]. However, there is no consensus on best practices in the use of electronic prescribing.Further research is needed to explore implementation of electronic prescribing within hospital systems considering that only 67% of eligible users were utilizing electronic health records (EHR) in 2012, and an even lower percentage for electronic prescriptions [7].With more prescribers utilizing electronic prescribing and pharmacists processing electronic prescriptions over the next few years, more analysis of e-prescribing methods can be made and best practices can be developed.These assessments can help patients reach disease state goals and help clinicians reach the overall goal to enhance patient care through better integration and communication between the patient, pharmacist and physician [18,23,24]. Case-series (and poor-quality cohort and case-control studies) 5 Expert opinion without explicit critical appraisal, or based on physiology, bench research or "first principles" Grades of recommendation  Administering and timing the completion of a series of tasks by a physician after a 2 week training program • Physicians were given a training session with the e-prescribing software using mock patients. • They were then evaluated on how quickly and efficiently they were able to redo those tasks on their own. • Physicians were able to quickly complete all tasks but only had difficulty with changing the status of a problem or with renewing multiple prescriptions at once

Results
A total of fifty-six peer-reviewed articles were included in this review.Trends show that most clinicians view e-prescribing positively despite some limitations.
Limitations of e-prescribing include cost of implementation and maintenance, insufficient training, and lack of standardization.As more legal regulations come into effect promoting standardization, there is greater incentive to invest in integrated e-prescribing systems as well as promote clinician adoption.

Cost Effectiveness
Illegible prescriptions and drug interactions have been attributed to causing approximately 7,000 deaths per year [59].E-prescribing has been shown to reduce prescribing errors and increase long term benefits [1].One health system in Michigan saved $3 million annually with the implementation of e-prescribing [59].Some of the benefits include: better legibility, less prescription duplications, and improved efficiency of processing in dispensing medication.In 2010, Kaushal and colleagues showed that error rates when converting from paper prescriptions to electronic prescriptions decreased by sevenfold [2].Furthermore, pharmacist, physician, nurse and staff efficiency has been shown to improve with e-prescribing [20].In addition, the use of electronic prescribing has been shown to increase patient medication compliance, which then further decreases the number of subsequent hospital visits [18].An example is a 2011 study that showed for every $10 increase in a patient's initial hyperlipidemia medication, the likelihood for a patient to be at goal decreases by 5% [11].A second study showed that for every dollar invested to increase medication adherence, there is a potential to save $3-$10 in medical costs [19].For some health plans, this could be a total saving of $14 billion/year. While the potential savings seem to be substantial, one of the biggest reasons for the lack of implementation of e-prescribing is the cost investment.The Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 was developed to pave the way towards making a technological approach to healthcare by setting aside $22 billion for EHR utilization [8].Once EHRs have been successfully implemented within a clinical setting, the progression is to focus on electronic prescribing.However, with the HITECH Act only focusing on EHR implementation, institutions are faced with having to invest their own funds in an e-prescribing system.In addition, there are no set standards that are required for an e-prescribing program and thus institutions and facilities have freedom in the type of system they choose. A 2007 study by Scalise and colleagues revealed that the cost to implement a basic e-prescribing program ranges from $1,500 -$4,000 per physician.The price for an advanced system with alerts, reminders and system integration is $29,000 per physician in the first year and $4,000 per physician every year thereafter [9].While there is a huge difference in pricing, multiple studies have shown that physicians will utilize the system more if it is user-friendly, integrated to have access to patient files and provides clinical and formulary decision support options [10.11,12,13,14].With many different commercially systems available and a multitude of different options and features, an issue arises in choosing a system and validating the investment cost.This decision is further complicated by the need to assess costs to train employees, technological maintenance of the program and ensure that the system is reducing error and providing a return on investment (ROI).Studies show that while e-prescribing software has been shown to be cost-effective for practices of all sizes, the time frame for a return on investment varies.Larger practices achieve a quicker ROI [20].Therefore, it is not difficult to imagine that many institutions are awaiting an "all-in-one" package and delaying the need to invest in electronic prescribing [15].
Pharmacies are also hesitant to implement e-prescribing into their workplace.A recent 2013 study by Banks and Galvez assessed that pharmacies did not want to spend initial start-up costs as well as maintenance and transaction fees to insurers [27].Furthermore, the lack of demand for use from physicians or patients promotes a delay in adoption.Some pharmacies commented that they didn't understand the benefits to be gained from using e-prescribing while others reported that the costs didn't outweigh the benefits.Additionally, independent pharmacies or pharmacies in rural areas have identified lack of prescription volume as a reason not to invest in such a costly program.On the other hand, with the HITECH Act of 2009, many states are providing financial assistance to offset those costs.In Tennessee, the state's pharmacists association allocated $675,000 in grants to 124 independent, community pharmacies [28].The state of North Dakota dispersed over $6 million to promote the adoption of information technology and has seen a 42% increase in active pharmacy utilization [28].Other states are following suit and offering state appropriations, grants and revolving loans so that smaller pharmacies can meet their technological needs, especially with the changes in state and federal requirements advocating for e-prescribing.
Incentives for pharmacies and institutions to implement adoption can result in direct cost savings.A 2013 study by Radley et al. showed that the decrease in prescription error from e-prescribing caused a 12.5% reduction in medical errors in over 4,700 hospitals in the United States [41].This translates to cost savings for over 17 million errors per year.Utilization of e-prescribing with formulary decision support can further reduce hospital visits and pharmacy spending on medication.Findings from recent articles revealed that physicians are more likely to prescribe tier 1 (generic) drugs [10,25,26] and patients are more likely to purchase the tier 1 medication prescribed when it requires lower copayment.Furthermore, the resulting decrease in utilization of tier 2 (brand name) or other specialty drugs can result in savings of $845 per patient based on a 20% 20 Electronic Prescribing: An Examination of Cost Effectiveness, Clinician Adoption and Limitations e-prescribing rate [13].Therefore, savings can substantially increase with a higher e-prescribing adoption percentage. Pharmacies may also achieve savings by not having to purchase specialized and expensive inventory that is either a rarely dispensed item or one that a patient rarely will purchase due to its high copayment.Stocking such unpopular items results in loss of income due to items not being able to be returned to the manufacturer or becoming expired.

Clinician Adoption
The 2013 Medicare eRx Payment Adjustment provides an incentive for physicians who serve Medicare patients to switch to e-prescribing.Physicians who do not switch to e-prescribing receive a 1.5% decrease in payment on all Medicare Part B services [5].Furthermore, this adjustment increases to 2% in 2014 [5].A major component for the continuing success and development of e-prescribing is its adoption by physicians, pharmacists, and any other healthcare providers.Many studies have discussed the mixed positive and negative perceptions on adoption and how negative attitudes are preventing the utilization of e-prescribing.Recent studies have shown that physicians' perception of e-prescribing after efficient implementation is viewed as very efficient and easy to use [31,32,39].A 2007 study analyzed the time to track both physician and staff members when using an integrated system.The results showed e-prescribing had no disruption to workflow and had no statistically significant increase in time when compared to handwritten prescriptions [4,40].The use of computerized physician order entry (CPOE) in hospitals to write inpatient prescriptions was found to decrease the likelihood of a prescription error in an acute setting by 48% [41].While cost was one of the main factors contributing to negative attitudes, there were other factors as well.Lack of adequate training, clinical support, formulary support, or software issues were identified as major deterrents [29,32,36].With the exception of lack of training, most barriers were related to the technology.Physicians reported issues of limited connectivity with pharmacies requiring multiple transmissions of the same prescription and pharmacies reported issues in receiving the same prescription more than once [10,32].This transmission issue created an increased workload for both physicians and pharmacists.If resolution of these issues is possible in the future, most physicians will have a more positive mindset towards electronic prescription adoption.For example, physicians with an integrated e-prescribing system who had access to their patient's medication profile were more likely to review it before prescribing a new medication.The use of this feature gave the physician more awareness of potential duplication in therapy, potential allergies and drug-drug interactions [2,36].Increased awareness in this manner gives physicians a better ability to evaluate and diagnose their patients.
Studies on pharmacist's perceptions revealed that they felt e-prescribing may decrease errors and increase efficiency for physicians, but required as much pharmacist intervention as would normally be provided due to omitted information or dosing errors on prescriptions sent electronically [30,32].Additionally, some pharmacists negatively viewed e-prescribing systems due to the over utilization of alerts.On the other hand, several studies showed that pharmacists perceived e-prescribing positively noting better legibility, improved processing time and efficiency as benefits [1].
Although clinician views on the adoption of e-prescribing are mixed, there is a consensus from all groups that proper training is required to improve the efficiency of the dispensing process [29,30,32]. Clinicians may receive incomplete information from time to time which results in distrust in the e-prescribing system.For example, a physician may find incomplete documentation of a patient's medical history [26].Findings from a 2011 study by Nanji and colleagues showed that for the 45 pharmacy interventions required for 3,850 computer generated prescriptions, thirty five percent of those interventions resulted from omitted information or potential adverse drug events [17].Findings from another study in 2012 revealed that 153 pharmacy interventions were needed for 1,678 new prescriptions to address omitted information, excessive quantity, etc. [40].Greater data accuracy and completeness is needed to increase clinician confidence in e-prescribing.
Clinicians also feel that some prescriptions required a faxed or phoned in prescription, citing system limitations especially with controlled substances preventing electronic transmission [32]. As of 2010, the DEA has authorized the use of electronic prescribing of controlled substances (EPCS) [34,35].However, most pharmacies have not yet adopted the utilization and will not allow physicians to electronically prescribe them and therefore, must submit a fax or phoned in prescription [35].While controlled substances only make up 11% of all prescriptions issued, they are prescribed by 90% of all physicians [34].This creates an increase in workflow for both physicians and pharmacies.However, the adoption of EPCS is due to the extra technology regulations to ensure secure transmission by the DEA.Many systems are still in the process of being updated to include these operational security issues.Hopefully, once in effect, the issue of controlled substances will be one less barrier for the proper utilization of e-prescribing.
In addition to studies on controlled substances, studies are being conducted to examine whether e-prescribing can help to efficiently manage patients and their disease states more so than paper prescriptions.A 2011 study by Michelis and colleagues analyzed the effects of using electronic prescriptions to better manage hyperlipidemia patients [11].Patients that were given an electronic prescription were 59% more likely to reach their LDL goals.The study revealed that the utilization of formulary decision support helped patients to obtain their low cost prescriptions.Similarly, a second 2009 study by Smith et al. discussed how diabetes patients benefited from e-prescribing through the reduction of errors such as correct product selection through more legible prescriptions (i.e. Novolog versus Novolin or glipizide versus glimeperide) [23].Furthermore, results revealed that clinical decision support helped physicians and pharmacists better predict drug interactions and assisted with updating dosage regimens.Other studies have also analyzed the use of electronic prescriptions for oral chemotherapy regimens [42].Authors developed a built in field for weight and body surface area to allow for more accurate dosing.Findings revealed that over six hundred clinicians readily accepted and utilized e-prescribing to better enhance oral chemotherapy safety.As illustrated, several studies have highlighted positive benefits in the use of e-prescribing, however, an equal number of studies point to its limitations.

Barriers to Implementation (Limitations)
A major concern that has prohibited the widespread adoption of e-prescribing is the lack of standardization or requirements related to it [43,57].This makes it difficult to assess the value of the different systems available. There are also inconsistencies in the perception of e-prescribing.Some systems are stand-alone and used only to write prescriptions while others are integrated and allow access to patient profiles.Some systems have a combination of varying clinical decision support and formulary decision support while others have none.In 2013, Marceglia and colleagues discussed potential modeling methods to better standardize e-prescribing systems by breaking down the six phases of the e-prescribing process [57].However, they recognized that it is a difficult task due to the large heterogeneity among functionalities and performances of existing systems.There is even variation in the sig codes designated for each system on how to use short codes to type in directions for electronic prescriptions [44].The National Council for Prescription Drug Programs (NCPDP) developed a structured and codified sig format that has been shown to be 95% effective.However, not all systems and institutions utilize this format because it is merely a recommendation and not a guideline that is enforced.
The second concern about e-prescribing relates to sociotechnical error, or an error caused only by human interaction when using the system [45]. It has been suggested that institutions and pharmacies develop an auditing system to ensure staff members are using the system correctly [28].Examples of sociotechnical errors are omitting information in a certain field, inappropriate use or selection of default doses, or a system alert that causes a distraction and interruption of workflow that ultimately causes a prescription error.For example, over-alerts from e-prescribing systems have been shown to be distractions hindering pharmacists' ability to discern true drug interactions.In 2011, Saverno and colleagues analyzed pharmacy e-prescribing systems in hospital and community settings and found that only 28% were able to correctly identify eligible interactions [37].A second 2011 European study showed that 93.7% of 2,729 electronic prescriptions generated an interaction alert.Of those alerts, only 10 were shown to have a true contraindication and 551 average danger alerts with some risk of adverse outcomes [38].Therefore, pharmacy systems usability needs improvement so that pharmacists can more accurately prevent adverse drug events.These changes can drastically improve pharmacist outlook on e-prescribing.Although sociotechnical errors are a concern, findings from Redwood's 2011 study showed that out of 485 medication incidents, only 15% were attributed to e-prescribing [45].While this may be viewed as a small amount, future efforts should aim to resolve these new issues and ensure that other issues do not arise from its use.
The third major concern related to e-prescribing is the lack of adequate training for clinicians.Many studies provide support for the implementation of a pharmacy informatics team to develop training education programs as well as mediate other technological issues, such as system maintenance and upgrades [26,37].With clinical guidelines constantly changing, systems need to be continually updated to reflect those new recommendations.In addition, the informatics team may develop upgrades to make the system more user-friendly or serve as the system liaison to prescribers.For example, some prescribers have indicated that a limitation to e-prescribing was physical access to a computer, which decreased time for patient interactions [47].However, with technology also constantly changing, concerns related to lack of training may be addressed through the development of application upgrades to use on hand-held portable devices [46].The upgrades could be developed by a pharmacy informatics team to help minimize the physical dependence on a desktop computer and allow prescribers to access patient EHRs and write e-prescriptions in places that have an Internet connection.In 2007, Donyai and colleagues provided a discussion on an upgrade developed to allow indication of an "as needed" (PRN) prescription in a hospital system to minimize administration and prescription errors by 48.9% and improve communication between prescribers and nurses [48].The upgrade also reduced the need for pharmacy interventions by 38.3%.
Lastly, not all e-prescribing systems have the capability to adjust dosages for special populations such as pediatric, elderly, or those with hepatic or renal impairments [38,55].While studies have shown the potential benefits for such features, Frolich and colleagues showed that recommended dosing was not altered for renally impaired patients.Furthermore, in 2013, Johnson and colleagues reported that there was limited data to show a significant role for e-prescribing in the pediatrics community due to the lack of pediatric functionality [18].However, the study also highlighted the adoption of e-prescribing emphasizing that changes to current e-prescribing systems are needed.

Conclusion
An expansion of e-prescribing is expected in the future.Despite limitations related mainly to cost investment and inconsistency of technological systems, many studies are emphasizing the benefits of e-prescribing, which include a decrease in prescription errors, improving efficiency and prospective cost savings as a result of utilization.Although studies revealing the cost effectiveness of e-prescribing are substantial, those effects will be seen over time after efficient training has been achieved.In one study, Bartlett and colleagues had 28 physicians undergo a two week training session using an integrated system with clinical and formulary decision support on mock patients.After timed assessments, the study showed that user ability would not be a barrier for adoption as long as there was a user-friendly system and a thorough one-on-one training with troubleshooting support for implementation issues [9,60].However, it is realistically difficult to require clinicians to spend long periods of time training and to expect them to overcome the barrier of investing in such a highly integrated system.The Center for Improving Medication Management (CIMM), along with the cooperation of several physician groups such as the American Medical Association, provides a clinical guide to e-prescribing with annual updates [62].The guide helps to stress the clinical importance of e-prescribing as well as any updates to its use.While it is a good resource, it is not highly well-known and is limited to prescribers.Future efforts should include a target towards other groups involved in the process, such as nurses, physician assistants and pharmacists.
Adoption of e-prescribing can be stimulated by organizational commitment from all members of the healthcare team and technological support teams [9,56].This includes support from payers that may clearly communicate the benefits of e-prescribing, such as initiating a pay-for-performance program to provide a strong incentive.E-prescribing vendors can and should also proactively reach out to clinicians to ensure increased use over time.Informatics teams should develop new applications to keep the technology up-to-date and more user-friendly.Quality management programs are needed to promote the standardization of e-prescribing.Currently, Surescripts is one company aiming to help provide information on this process with the Quality Program.Their goal is to improve the quality of the e-prescribing process by assessing all aspects and work on standards, certification, network monitoring and providing customer support [61].Surescripts works with physician groups, system vendors and pharmacies to collect data to reduce the amount of time a problem is identified and solved.However, the Surescripts' program does not include every provider group, vendor and pharmacy because there are no clear laws as to what is required for the e-prescribing process.The program needs legislative support to ensure that those involved are required to be regulated by a third party.
Positive perception of e-prescribing will be the key to increasing adoption by healthcare professionals.Many physicians and pharmacists seem to focus on the barriers as opposed to the benefits.It is important to facilitate e-prescribing adoption by targeting factors that are salient to each group of clinicians.Further investment in a pharmacy informatics team, especially within institutions, can address these issues of implementation, training, and development of the continuous technological maintenance that e-prescribing systems require.Although a valuable resource, pharmacy informatics, specifically e-prescribing is still relatively a new concept and further assessments are warranted.