The research team approached the Pharmacy Network Project Team at the Newfoundland and Labrador Centre for Health Information (the Centre). The Centre is the province’s primary custodian of electronic health data, and is responsible for the development and implementation of the confidential and secure provincial electronic health record, including the provincial Pharmacy Network. The Centre maintains key health databases, prepares and distributes health reports, supports and carries out health analytics and applied health research activities, and undertakes benefits-evaluation projects. Additionally, the Centre provides quality information to health professionals, the public, researchers, and health system decision makers. Given the Pharmacy Network Project Team’s relationship with community pharmacies, they were chosen as a means for recruiting community pharmacies to participate in this study. Four of the five pharmacies identified agreed to participate, while the fifth declined due to work commitments.

Data were collected at the four study pharmacies between July 20 and December 3, 2011, over a period of 17 to 19 weeks, depending on the pharmacy, using a one-page standardized data collection sheet. This data collection sheet was pilot-tested and refined prior to initiation of the study to ensure validity of the tool. Using this tool, pharmacists were asked to document the prescription type received (handwritten or EMR generated), how the prescription arrived at the pharmacy (brought in by the patient or faxed), name of drug(s) prescribed, which drug(s) required clarification, reason for clarification (illegible handwriting, missing information, dose discrepancy from previous prescription, possible drug interaction, allergy, cost contraindication, insurance issue, or other), and time taken for clarification calls to prescribers. Three pharmacies collected data Monday through Saturday. However, to minimize disruption in workflow, the fourth only collected data on alternating days of the week. Only the total number of new prescriptions (no refills) on designated data collection days was recorded. Data were analyzed using SPSS version 17.0.

Four semistructured interviews were completed with participating pharmacists between December 22, 2011 and January 10, 2012. During these interviews, pharmacists were asked about the reasons that prescriptions require intervention, the time spent and methods used for resolving issues, and specific errors frequently associated with handwritten and EMR-generated prescriptions. They were also asked to estimate the percentage of EMR-generated prescriptions received at their practice site. Sessions were audiotaped and transcribed. Interview transcripts underwent thematic/content analysis with the aid of NVivo 9 software. This study was approved by the Health Research Ethics Authority of Memorial University of Newfoundland and Labrador on June 29, 2011 (REB Ref #11.112).

Overall, for 18,042 new prescriptions filled during the study period, there were 200 (1.11%) clarification calls made. The mean length of time to make a clarification call was 9.1 minutes (SD 5.6). Table 1 provides the characteristics of the study pharmacies, while Table 2 shows the number and percentages of new prescriptions requiring clarification.

An estimated 1.33% (161/12,089) of handwritten prescriptions required a clarification call, compared to 0.66% (39/5953) of EMR-generated prescriptions. For three out of four study pharmacies, the estimated proportion of clarification calls for handwritten prescriptions was higher than that of EMR-generated prescriptions, whereas the fourth pharmacy required clarification for a higher proportion of EMR-generated versus handwritten prescriptions. Across all pharmacies, handwritten prescriptions resulted in the majority of clarification calls to prescribers (161/200, 80.5%).

The mean duration of clarification calls for pharmacies A, B, C, and D were 7.2 (SD 4.8), 13.2 (SD 3.1), 7.2 (SD 6.3), and 8.5 (SD 4.6) minutes, respectively. The mean duration of clarification calls for all pharmacies combined was 9.1 (SD 5.6) minutes.

Table 3 shows the reported reasons for clarification calls grouped across four themes: illegibility, missing information (ie, dose, drug, duration, and frequency), appropriateness of the prescription for the patient (ie, dose discrepancy, confirm dosage, known allergy to drug, possible drug interaction, previous adverse reaction, verify directions, and drug), and other (eg, medication not available, insurance, and cost of drug). Some prescriptions had multiple causes for initiation of a clarification call resulting in a total of 236 reported reasons for clarification.

Overall, the most common reason for clarification was to verify the appropriateness of the prescription for the patient (92/236, 39.0%), followed by illegibility (48/236, 20.3%). Dose discrepancy, a reason within the appropriateness category, made up 22.0% (52/236) of all reasons for clarification calls—40 out of 195 (20.5%) for handwritten and 12 out of 41 (29%) for EMR-generated prescriptions. Missing information was the reason for 19.9% (47/236) of calls. In the Other category, the most common reasons were that the medication was not available (15/236, 6.4%)—12 out of 39 (31%) for handwritten and 3 out of 10 (30%) for EMR-generated prescriptions—and insurance issues (12/236, 5.1%)—11 out of 39 (28%) for handwritten and 1 out of 10 (10%) for EMR-generated prescriptions.

Several important themes emerged from the qualitative interviews, with respect to prescriptions requiring intervention and specific errors associated with prescription type. Four themes are reviewed in Textbox 1, including (1) reduction in clarification calls, (2) elimination of illegibility, (3) unique errors with EMR-generated prescriptions, and (4) errors with reprinting prescriptions. When asked about issue resolution, interview subjects cited engaging the patient, using medication profiles, and contacting the prescriber as common interventions. The two pharmacists that were connected to Pharmacy Network noted that while there are limitations due to partial adoption, the Pharmacy Network was beneficial in identifying potential drug abuse and for accessing the patient’s profile. In addition, the medication management program was cited as an enabler for pharmacists to resolve prescription issues independently. Under this standard of practice, pharmacists in Newfoundland and Labrador are authorized to change the form and/or regimen of dosage, change the quantity of medication, complete any missing information, and make nonformulary generic substitutions under specific circumstances, such as having historical information available from a patient’s medication profile [22].

Although illegibility was eliminated, all participants agreed that certain issues with EMR-generated prescriptions still necessitate contacting the prescriber from time to time. For example, participants noted that certain computer-generated defaults, such as quantity calculations and drug catalogs, sometimes lead to error. Interestingly, one pharmacist alluded to incomplete medication reconciliation as a source of missing information: “The doctor will write out a list of medications for somebody and they omit one…and you have to call to get the prescription.” Perpetuation of errors can also be a problem when discrepancies are resolved by the pharmacist at the point of dispensing, but are not documented in the prescriber’s EMR.

In this study, the overall prevalence of new prescriptions requiring clarification was 1.11% (200/18,042). At three sites, the estimated proportion of clarification calls for handwritten prescriptions was higher than for EMR-generated prescriptions. The fourth site required clarification for a higher proportion of EMR-generated versus handwritten prescriptions, however, it also had the lowest estimated percentage of EMR-generated prescriptions. Consistent with the literature, an important benefit of EMR-generated prescriptions was elimination of illegibility, however, EMR-generated prescriptions still required pharmacist intervention, mostly due to omission of information and dosing [9,13,14,23]. While clarification calls to prescribers will continue to be required, the need to do so for missing information and dose discrepancies present important opportunities for future benefits possible with advanced e-prescribing.

Whereas missing information and dose discrepancy accounted for 41.9% (99/236) of all reasons for clarification in this study, these actually accounted for 54% (22/41) of reasons for clarifications for EMR-generated prescriptions. Errors unique to EMR-generated prescriptions, such as incorrect rounding or default quantities, were also highlighted by interview subjects. With the average time for a clarification call being 9.1 minutes in this study, a reduction in the need for clarification calls could have a considerable impact on the day-to-day activities of pharmacists. This may also translate to productivity benefits for both prescribers and pharmacists, as previously articulated in a Pan-Canadian DIS study [18].

In the Canadian context, fully evolved e-prescribing will enable secure electronic transmission of prescriptions from prescribers to a patient’s pharmacy via a provincial DIS which may further decrease prescription issues [15]. Connecting EMRs and pharmacy management systems to a DIS necessitates implementation of interoperability standards which may help mitigate issues frequently encountered for EMR-generated prescriptions in this study. For instance, defining mandatory information for successful transmission of a prescription to the DIS may reduce missing information, thereby reducing the need for clarification at the point of dispensing. This is analogous to a “forcing function” designed to prevent omitted information as described by Nanji et al [14]. Similarly, the use of specific terminology for drugs to create alignment between EMR drug catalogs and drug identifiers defined at the DIS level may help reduce prescriptions for products that are not available in Canada. As part of implementation, jurisdictions may also impose business rules that could reduce the likelihood of incorrect default quantities being prescribed, for example, requiring the EMR to display the final prescription to the prescriber for review and validation prior to signing off.

In this study and others, consulting patient medication profiles was a frequent pharmacist intervention for resolving problems with prescriptions [10]. As noted in the interviews, once adoption of the provincial DIS in Newfoundland and Labrador (ie, Pharmacy Network) is complete, a more comprehensive medication profile will become available. This tool may further support pharmacists in assessing the appropriateness of prescriptions by providing relevant context such as titration of dosages over time. Advanced e-prescribing functionality will also enable prescribers to have access to the same medication profile, potentially mitigating issues like unintentional dose discrepancies at the time of prescribing.

Finally, despite issues being resolved at the point of dispensing, it was observed in this study that with EMR-generated prescriptions, errors tend to be repeated upon reprinting. With advanced e-prescribing functionality, information about the final prescription would be recorded in the DIS, not just the local pharmacy management system, decreasing the chance that the identical error could be perpetuated in the future.

This study used an exploratory prospective research design focusing on a small convenience sample of community-based pharmacies in a small Canadian city. The small number of qualitative interviews conducted may also limit generalizability. In addition to a limited sample size, another constraint of this study is that for the purposes of reducing respondent burden, the total number of EMR-generated and handwritten prescriptions was based on pharmacists’ estimates rather than collected prospectively. That said, insights from this small study are important from a lessons-learned perspective. Finally, this study did not quantitatively capture data about pharmacist interventions other than clarification calls, such as discrepancies resolved by the pharmacist using available medication management resources or in collaboration with the patient themselves. Including these may have been useful in understanding how advanced e-prescribing might offer additional benefit in either augmenting or making these types of interventions more efficient.

This study provides valuable insight around the impact of early e-prescribing on pharmacists’ clarification calls in four community pharmacies located in St. John’s, Newfoundland. While illegibility has been eliminated by computer-generated prescriptions, advanced e-prescribing functionality with connectivity to an electronic provincial DIS provides an opportunity for further realization of benefits related to medication prescribing.