Risk factors for medication errors in the electronic and manual prescription 1

ABSTRACT Objective: to compare electronic and manual prescriptions of a public hospital of Brasilia, identifying risk factors for the occurrence of medication errors. Method: descriptive-exploratory, comparative and retrospective study. Data collection occurred from July 2012 to January 2013, using an instrument for the review of the information contained in medical records related to the medication process. A total of 190 manual and 199 electronic records composed the sample, with 2027 prescriptions each. Results: compared to the manual prescription, a significant reduction was observed in the risk factors after implantation of the electronic prescription, in items such as "lack of the form of dilution" (71.1% to 22.3%) and "prescription with brand name" (99.5% to 31.5%). Conversely, the risk factors "no check" and "lack of CRM of the prescriber" increased. The lack of the allergy registration and the occurrences related to medication were the same for both groups. Conclusion: generally, the use of the electronic prescription system was associated with a significant reduction in risk factors for medication errors, concerning the following aspects: illegibility, prescription with brand name and presence of essential items that provide a safe and effective prescription.


Introduction
The identification of risk factors for errors related to medication administration has proved important to ensure greater safety for patients and health professionals. They can be identified, for example, by analyzing medical prescriptions, which allows preventive actions to reduce the occurrence of adverse events.
In the literature it can be verified that the electronic prescription (EP) system for medications has enabled higher quality care to hospitalized patients and others involved, demonstrating that the choice of this model can help to reduce errors related to medications by up to 50% (1)(2) . Other studies also refer to improvements in antibiotic prescriptions and reduced time and cost of hospitalization (3)(4) . However, there are publications that indicate increased mortality after its implementation (5)(6)(7) , medical team resistance in the use of electronic prescription due to time constraints, impairments in the interaction between patients and nurses, and lack of integration with the flow work (8) .
These data are worrying, because events of this nature are frequent and constitute a concern for healthcare professionals, patients, and government agencies. Accordingly, the Ministry of Health and the National Health Surveillance Agency (Anvisa) launched, in April 2013, the National Patient Safety Program, with the focus on preventing and reducing the incidence of situations that result in harm to patients.
Internationally, the issue has also been the subject of various investigations. Studies have shown that, after the implementation of the electronic prescription system, there has been a reduction in the frequency of medication errors (1)(2)7,9) . In Brazil, however, this strategy has still been little investigated. Among the few studies, one investigated the presence of prescriptions smudged after printing (18%), suspended medications (17%) and lack of information about presentation, time (9%) and route of administration (82%) (10) .
International studies that comparatively investigated manual and electronic prescriptions (11)(12)(13) , showed a reduction in prescribing error rates and improved outcomes for patients following the implementation of an electronic system. Given the above, knowing the prescription system and its functionality is essential for a safety proposal for both patients and health professionals.
With regard to the administration of medication, to talk about safety necessarily refers to reducing the risk of errors that generally occur in the prescription, dispensing and administration stages. Studies show that: 72% of them started with the prescription and 15% during the administration (14). In a study conducted recently in Brasilia, Brazil, with a total of 484 doses observed, errors occurred during the administration of the drug in 69.5% of them: 69.6% during the preparation phase, 48.6% were time errors, 1.7% dosage errors and 9.5% derived from omission (15) .
Despite some advantages, one of the difficulties in the adoption of EP may be the high cost of the system, as well as a high risk that its implementation is ineffective and may produce unintended consequences and harm (4,16) . In Brazil, various institutions have already adopted this system, however, it is still necessary to know how its implementation is going and how professionals interact with it. It is also fundamental to investigate whether it is, in fact, providing greater safety and quality for the care.
Considering the potential benefits of the electronic prescription and how the computerized system contributes to reducing medication errors, reducing costs and assuring the quality of care provided, it was proposed to investigate these aspects in a hospital of the city of Brasilia.
Accordingly, this study aimed to compare electronic (EP) and manual prescriptions (MP) of a public hospital of Brasilia, identifying risk factors for the occurrence of medication errors.

Method
This was an exploratory, descriptive and comparative study, performed in the internal medicine department of a public hospital of Brasilia, the Federal District, Central Region of Brazil. The clinic has 31 beds reserved for patients with chronic and degenerative diseases with long periods of hospitalization, which, because of the treatment, make use of various medications.
It should be noted that in this unit the electronic prescription system was implemented in August 2011.  previously tested and adapted from earlier studies (17) .
The dependent variables analyzed at this stage were: manual prescriptions (handwritten); electronic prescriptions, performed on the computer and typed into the electronic prescription system. In order to achieve the purpose of assessing the presence or absence of information related to medications in the medical developments and outcomes and nursing notes, these were also analyzed in the medical records included in the sample.
The independent variables were diverse, namely:    With regard to incomplete scheduling, there was a 3.43 times greater chance of this being present in the MP than in the EP. Therefore, there was a significant increase of these variables in the electronic prescription, demonstrating another of its benefits. Medications administered without checking, also presented in Table   1, is another worrying factor, as this increased from 31.1% in the MP to 79.2% in the EP, constituting an 8.41 times higher risk.

Results
The absence of justification for not carrying out the checks also increased in the EP, which is attributed to the ease of noting the reason for not checking in the MP.
Furthermore, the fact can also be associated with the distance of the computers from the bed of the patient, which increases the chance of forgetting to record this information, or with the lack of ability of professionals with the electronic system, so that the check is often performed incorrectly.
The following constitute factors that did not contribute to the prevention of medication errors: "presence of medication without checking" and "lack of justification for not administering the medication".  Regarding the risk factors "lack of form of dilution", "lack of route", "lack of frequency" and "lack of presentation", there was a reduction in all. The possibility of data not being present on frequency, form of dilution, presentation of the medication and route of administration in the MP was significantly higher than in EP. Regarding this aspect, the implementation of EP enabled a safer prescription (p<0.000) ( Table 2).
The risk of prescriptions being issued with the brandwas higher name in the MP than in the EP (p<0.000).
Inappropriate acronyms and abbreviations were found in 100% of the prescriptions, and in the medical and nursing outcomes. Often, abbreviations or acronyms are used in order to save time, however, this is a risk factor, since they may be erroneously interpreted by health professionals.
When used, abbreviations should follow standardization to facilitate their understanding.
However, at the study site, an absence of standard abbreviations and acronyms was observed. The risk of using abbreviations was higher in the MP than in the EP.
According to Table 2, all the risk factors analyzed were enhanced with the electronic system, especially concerning acronyms and abbreviations.

Discussion
Medication errors can cause major health problems, with relevant economic and social repercussions that, in a certain way, directly affect the lives of patients and the health professionals and institution, as well as prolong the hospitalization period and affect the treatment (1)(2)4) .
The EP is one of the main measures to prevent medication errors (18) . Studies have shown the possibility of a significant reduction of serious errors by implementing this system, with advanced support for the clinical decision. Accordingly, the results of this study converge with those of other investigations with regard to improving patient safety and reducing risk factors for such situations (19)(20)(21) .
The systems that support clinical decision are more complete and offer suggestions as to the route of administration and any correction in the drug dose and frequency values. In addition, other more complex systems cover checks for allergies, laboratory test results, drug interactions and provide clinical protocols to support the prescriber (21)(22) .
These tools can also improve the process for safer medication administration, using technology, such as smart infusion pumps, computers at the bedside and medication administration system with barcodes.
In particular, computers at the bedside, in association with the computerization of the system, make the registration faster, reducing the time spent completing documentation by 30% and reducing potential faults (22) . observed in the adjusted probability of errors that did not cause harm (20) .
As in other studies, the benefits of the electronic prescription observed in this study are linked to the fact that it is a technology used to facilitate and ensure the safest prescription of the medications (21) . However, if it is not used properly, it can not achieve these goals. Thus, it is understood that the electronic prescription, by itself, does not eliminate the possibility of medication errors.
Among its disadvantages, it is worth mentioning that this is a complex project that is still expensive, which restricts the number institutions that can adopt it, even in the USA (3) . Some faults observed in the system can configure another disadvantage: Repetition of prescriptions from previous days, without review, and information entered incorrectly (10). These practices can negatively affect the safety of the medication process and therefore require intervention (21) .
Finally, the risk factor of erasures should be highlighted, which is reduced with the EP, thus providing more security to the medication process. In agreement with other studies (10,21) , the lack of registration of the nursing team regarding administration of the medication was another problem identified in this study, which, even with the implementation of EP, has not been solved. Failure to check medications is something usual, however, it is believed that with adjustments to the system its occurrence can be reduced.
Limitations in the present study are related to the use of secondary data and difficulty of access to the patient record manuscripts. The study analyzed the prescriptions generated by the electronic system and did not include the analysis of its structure and functionality or of its acceptance and the interaction of professionals with the system.

Conclusion
The implementation of a electronic prescription system is associated with a reduction of risk factors for medication errors. Elimination of illegibility is an inherent aspect of the electronic prescription process, which also minimizes the use of inappropriate abbreviations,  (23) .
The results of this study demonstrate that the illegibility risk factor is virtually eliminated with EP, similar to that found in other national and international studies, which showed (24) 4% of illegible prescriptions (9,21,25) . It The EP also contributed to the presence of essential information for the prescription (route, dilution, frequency), in that the absence of such data can cause problems at the time of preparation, dispensing and administration of the medication for the patient. To provide all the elements of information is, therefore, essential for a safe prescription.
Other studies show incomplete or inaccurate information as one of the main causes of medication errors, as well as the absence of data on the date and route (10) .
To enter the name of the active ingredient is a procedure considered obligatory by law in the public services of Brazil (Law No. 9. 787, of February 10, 1999), determining that, in these spaces, it is obligatory to prescribe the medications by the generic name. This is a procedure for reducing the exchange of names of similar drugs, as the brand names change from one geographic region to another. Traditionally, the use of the brand name was used in the MP, which can lead to errors. Now, with the introduction of the electronic prescription, this possibility has been reduced, since the name of the active ingredient is used (10,21) .
A study comparing manual and electronic prescriptions showed that the frequency of errors decreased from 18.2% with the MP to 8.2% with the EP. The greatest reductions were seen in adjusted probabilities of errors regarding illegibility (97%), the use of inappropriate abbreviations (94%) and the lack of information (85%). In this study, a 57% reduction was www.eerp.usp.br/rlae