US Pharm. 2012;37(7):31-33.

In an earlier TechRx column, we defined electronic prescribing (e-prescribing) as consisting of three components: 1) creating a prescription using an electronic tool such as an electronic health record (EHR); 2) signing the prescription with an electronic signature; and 3) transmitting the prescription to the pharmacy computer using a computer-to-computer electronic data interchange protocol with standardized data fields.1-3 We noted that the most commonly employed industry standard for transmission of the e-prescription is the National Council for Prescription Drug Programs Prescriber/Pharmacist SCRIPT Standard. In another TechRx column, we focused on the role of computerized clinical decision support and drug-interaction databases.2 As noted, this holds great promise for decreasing medication errors that originate on the prescriber’s end of the equation. Using a pediatric patient example, this column will take us through the entire computerized prescriber order entry–e-prescribing process and will examine the impact at the pharmacy.3

At the Pediatrician’s Office

The pediatrician’s e-prescribing software/EHR can confirm the patient’s insurance coverage either through a batch query that runs overnight for all patients on the following day’s schedule or through a real-time query when the child is being seen. The result of the query can confirm active insurance coverage, pharmacy benefit information, patient-specific formulary information, and copay structure.

The pediatrician can also download medication history. As previously discussed, the medication history is derived from national pharmacy benefits manager (PBM) and retail databases.2 One advantage of the medication history download is that the pediatrician will be able to view prescriptions that were written by other prescribers. The information needs to be reviewed with the patient and parents/guardians, as the patient may not actually be taking all medications that were prescribed. In addition, the patient may be taking medications that were paid for in cash and not reported to a database. This could include some prescription items as well as OTC products and herbal preparations. PBM databases typically hold information on medications that were paid for by insurance coverage, and may lack information on prescriptions that were paid for in cash; however, a national retail pharmacy database does collect information on cash prescription purchases, but a few pharmacy chains that sell very inexpensive generics do not report these cash sales.4 In many pediatric practices, a nurse will review the medication list and update the EHR record to reconcile all medication entries.

When the pediatrician actually enters the prescription electronically, a number of edits can then be performed by the software. As previously discussed, this functionality is known as computerized clinical decision support (CDS). The e-prescribing formulary check determines whether the prescribed pharmaceutical is on the patient’s specific formulary. If not, the software can indicate which alternative items are on the formulary. The pediatrician can either select an alternative or can file a prior authorization request for coverage of a nonformulary item. This entire process can occur before the prescription is transmitted to the pharmacy, thereby avoiding the need for the pharmacist to call the prescriber because of a nonformulary medication.

After the medication is selected, the software will perform various clinical edits. The most common edits are for allergies, adverse drug reactions, and drug–drug interactions. In a pediatrics practice, the availability of a neonate/pediatric medication dosing module is very important.5 The most common method for calculating pediatric dosing is based on the child’s weight. Hence, when a recent weight measurement is available on the system, the EHR software should be able to draw from standard references and suggest appropriate dose ranges for the individual child.  

The system should display the current weight measurement and should be able to determine whether this measurement is current enough to be useful. This determination is age dependent. For example, a rapidly growing neonate would need a more recent weight than an adolescent. When the available weight measurement is outdated, the system should prompt for a current measurement. Entries of weight and height should also be checked by the system against national age-based norms to flag possible errors. The system should also display the formula for the recommended dosage per unit of body weight. Obviously, the system should check for proper units (lb. vs. kg). In instances where medications are supplied in liquid from, the system should specify the appropriate drug concentration and the appropriate volume for each dose. The volume should be appropriately rounded and expressed in convenient units that the child’s parents can understand.

Once a dose is finalized on the prescription, the system should perform an independent dosage check to determine whether the selected dose falls within the appropriate range based on accepted published standards. If no standards exist, the system should inform the prescriber that a final dose check was not performed.

An example will serve to illustrate these useful features. A 6-year-old child weighs 42 lbs. and is being prescribed amoxicillin. The appropriate dose of amoxicillin for the clinical condition is 20 to 50 mg/kg/day in divided doses every 8 to 12 hours. The prescriber selects a dose of 30 mg/kg/day given in divided doses every 8 hours. The oral suspension at 200 mg/5 mL is selected. The system then calculates that the child weighs 19.1 kg. The dosage calculates out to 191 mg every 8 hours. This equates to 4.8 mL of suspension every 8 hours. For the parents’ convenience, this can be rounded to 5 mL or 1 teaspoonful every 8 hours. The software can then double check this final dose to ensure that it adheres to appropriate prescribing standards.

In some instances, pediatric dosing will be based on body surface area or ideal body weight. In these cases, the appropriate pediatric formulas (which are different from the adult formulas) should be calculated by the system. In other instances, the pediatric dosing is based on the child’s age. In these instances, the system should apply the appropriate age-based criteria.

The EHR can provide the most useful edits when the child’s weight, height, medication history, dosage and frequency information, and disease state information are all codified as discrete structured data within the system. More detailed information results in clinical edits that are more specific and more useful to the clinician. Clinicians should be engaged to adjust thresholds so that they are only presented with clinically relevant warnings.

Refinements to the pediatric dosing module can include adjustments for abnormal renal or liver function. These adjustments can be applied by the system if appropriate software is available and if the required laboratory values are available on the system.

Once the e-prescription has been written and all alerts have been addressed, the prescriber electronically signs the prescription. In many EHR systems, this involves the entry of a password. However, for controlled substances, the Drug Enforcement Administration requires a two-factor authentication protocol to electronically sign the prescription.6  This requirement was previously discussed in the November 2011 column.7

Transmission to the Pharmacy

Once the e-prescription is signed, the software routes the prescription to the patient’s pharmacy of choice. In many cases, the e-prescription is securely routed via an intermediary or e-prescribing network, such as Surescripts. If the intended pharmacy is a network participant and is capable of receiving e-prescriptions, then the intermediary will route the prescription directly to the pharmacy’s computer system. Surescripts requires that participating e-prescribing software systems and pharmacy systems demonstrate compliance with applicable security and data standards. This ensures that prescribing and receiving computer systems are operating with the same national standard, i.e., the NCPDP SCRIPT. In addition, some states have specific rules that apply to the transmission of e-prescriptions. For example, New York State requires that the transmission be encrypted. In the event that the intermediary cannot complete the transaction electronically, the intermediary may convert the e-prescription into a facsimile and transmit the prescription over the phone lines to the pharmacy. However, in the case of an e-prescription for a controlled substance, the intermediary is prohibited from converting the script to a facsimile and must inform the prescriber that the transmission failed.6

At the Pharmacy

Modern pharmacy computer systems can accept the e-prescription directly. Some older pharmacy software required that the prescription first be printed and then manually reentered. Pharmacies utilizing such older software are encouraged to upgrade. Direct transmission of the prescription into the pharmacy system has the potential to decrease the number of prescriptions that are telephoned in and to streamline dispensing by reducing the time spent entering prescription information into the pharmacy computer system. The elimination of pharmacy transcribing and data entry might possibly reduce the frequency of pharmacy dispensing errors.

The impact of electronic transmission of e-prescriptions to the pharmacy with respect to dispensing errors has recently been investigated by Moniz and colleagues.8 In this study, direct electronic transmission of prescriptions entered via computerized physician (prescriber) order entry (CPOE) was compared against prescriptions entered by CPOE and faxed or provided to the patient as a paper script. Prescriptions were analyzed from two control clinics and one e-prescribing clinic. All clinics used a common CPOE system. During a baseline period, prescriptions from all clinics were generated with CPOE and either faxed or given to the patient on paper. During the intervention period, prescriptions from the control clinics continued to be printed and handed to the patient. At the e-prescribing clinic, prescriptions were electronically transmitted to the pharmacy unless the medication was not approved for electronic transmission (e.g., a controlled substance). Controlled-substance prescriptions were printed and handed to the patient.

A total of 41,022 prescriptions were generated during the study, and 72% were from the e-prescribing clinic. Of these, 7,233 prescriptions were available for analysis. These prescriptions were the set in which all required data points were available for review. Dispensing errors were determined by analyzing dispensing histories against prescribing histories. In the e-prescribing clinic, the baseline rate of dispensing errors was 3.4%. This rate dropped to 1.8% during the intervention period when electronic transmission was employed (P = .034). Hence, dispensing errors were reduced by about half when electronic transmission was added to CPOE. These data demonstrate that electronic transmission conveys additional benefits beyond CPOE alone. While the CPOE has error-preventing benefits such as decision support and the elimination of handwriting ambiguities, electronic transmission appears to eliminate additional error-producing factors. The most frequent types of errors identified were with respect to product strength, dose, and frequency. For example, in the e-prescribing clinic there were 4,558 prescriptions analyzed in the baseline period. Of these, 2.3% were dispensed with the wrong dose, wrong frequency, or wrong strength. During the intervention period with electronic transmission, the rate of these specific types of errors dropped to 1.0%.

Conclusions

Computerized prescriber order entry and direct electronic transmission of e-prescriptions to the pharmacy computer have the potential to reduce both prescriber and pharmacy sources of medication errors. In the prescriber’s office, CPOE using decision support has the potential to eliminate many errors on the prescriber’s side of the equation. In pediatric applications, the use of a pediatric dosing module can add significant value and functionality. At the pharmacy, it is important that software upgrades be installed to allow the pharmacy computer to accept the e-prescription directly. Recent data demonstrate that electronic transmission directly to the pharmacy computer reduces dispensing errors by about half.

REFERENCES

1. Figge H. What really is electronic prescribing? U.S. Pharm. 2011;36(9):HS-35-HS-38.
2. Figge H. Computerized clinical decision support and drug interaction databases. US Pharm. 2012;37(3):47-49.
3. Robinson GA, Figge H, Stein RL, Russell J. The life and times of an e-prescription. Hosp Pharm. 2011;46(12):956-959.
4. Choudhry NK, Shrank WH. Four-dollar generics–increased accessibility, impaired quality assurance. N Engl J Med. 2010;363:1885-1887.
5. Spooner SA. Special requirements of electronic health records in pediatrics. Pediatrics. 2007;119:631-637.
6. U.S. Department of Justice. Drug Enforcement Administration. 21 CFR Parts 1300, 1304, 1306, and 1311. Electronic prescriptions for controlled substances. Interim final rule. Federal Register. 2010;75(61):16236-16319.
7. Figge H. HIPAA: privacy, security, and pharmacy information technology. US Pharm. 2011;36(11):78-81.
8. Moniz T, Seger A, Keohane C, et al. Addition of electronic prescription transmission to computerized prescriber order entry: effect on dispensing errors in community pharmacies. Am J Health-Syst Pharm. 2011;68:158-163.

To comment on this article, contact rdavidson@uspharmacist.com.