Department of Medical Informatics Singapore General Hospital, Singapore
LITERATURE REVIEW
There have been a number of studies pertaining to the use of drug information software on Personal Digital Assistants (PDAs)[5,6,7,8,9]. Most of these studies do comparisons of the different software available in the market. However, to date the authors did not find any evaluation study on the design, usability and feasibility of a drug information software coupled with the hospital’s formulary in the hospital.
MATERIALS & METHODS Background
Singapore General Hospital (SGH) is a 1600-bed tertiary referral centre. Prior to the implementation of a PDSS, drug information and the hospital’s formulary are not readily available at the point-of-care. Drug reference books and formulary booklets are available in the wards and clinics in very few quantities and they are updated only bi-annually.
The formulary classification of a drug affects the costs bore by the patients and the hospital. This includes information on whether patients can enjoy government subsidies on prescribed ‘standard’
drugs, or if they will have to pay full retail prices on ‘non-standard’ or ‘non-formulary’ drugs.
Clinicians often do not have this information readily during prescribing.
Table 1: Top 3 Most Important Types of Drug Information at Point-Of-Care
Information Category Information Components % Ranked as Top 3
Most Important (n=30) Dosing Dietary considerations, dose adjustments, paediatric/usual
dosing, IV solution capacity
80%
Drug effects Adverse/Common/Serious reactions, contraindications, warnings 77%
General drug information Administrative route, class/therapeutic category, indications/use, pregnancy issues
50%
Hospital-specific information Dosage form, formulary status, cost 30%
Pharmacodynamics/Pharmacokinetics Kinetics, mechanics of action, metabolism, monitoring parameters
30%
Drug identification Brand names, drug interactions, generic equivalents, pharmaceutical company
27%
User specifications Ability to add drugs, notes by users, user-customized lists 10%
Information types Clinical guidelines, overdose/toxicology data, references 10%
Development of ePharmacopoeia
Our PDSS system, named “ePharmacopoeia” is a multi-platform capable pharmacopoeia that contains generic drug information as well as the hospital’s formulary information. It enables clinicians and pharmacists to access drug information easily at the point-of-care.
A small team consisting of clinicians, pharmacists and IT personnel was formed to look into the design and usability of ePharmacopoeia. The following design considerations were taken into account:
(1) Ability to show comprehensive drug information at the point-of-care without information overload.
(2) Ability to search and retrieve by generic and trade names.
(3) Ability to show hospital-specific information for the drug. e.g. formulary classification, price, dosage forms, other comments.
(4) Ability to easily find similar drugs that are cheaper or more suitable for the patient.
(5) Ability to navigate quickly to desired sections within a drug monograph.
To ensure that the drug information provided will be useful to the clinicians and pharmacists, a survey was done amongst 30 doctors, pharmacists and nurse clinicians to determine the 3 most important types of drug information they need at the point-of-care based on their practice. As shown in Table 1, 80% ranked dosing information as the most important, followed by information on drug effects (77%) and general drug information (50%).
Lexi-Drugs was chosen as the drug information source for ePharmacopoeia based on positive reviews
from peer review case studies and articles[5,6,7,8,9].
The international version is used as it includes trade names of drugs used outside the USA. To ensure that the drug information on ePharmacopoeia is adequate and of sufficient depth yet not overloaded with information, we have taken into consideration Table 1 and removed some of the information components under the categories of “User specifications” and
“Information types”. The hospital formulary is then reviewed internally and included with Lexi-Drugs to become a one-point drug and formulary information source.
ePharmacopoeia runs on several platforms. The PDA version runs on either a PocketPC device or a PalmOS device. The PC-based client runs on any browser-enabled personal computer (PC) or compatible device, and they are deployed to wireless mobile carts used on the ward rounds and on browser-based client is automatically updated via a central server hosted by the vendor. Updates on the hospital formulary are available to users on a regular basis via the same method.
Users
30 doctors, pharmacists and nurse clinicians took part in this pilot study. They were chosen for this pilot due to the factors of availability and interest. Special consideration was taken to select doctors from different specialties and with varying amount of clinical experience so as to obtain a more representative subject group. Of the clinicians, 25%
are Senior Consultants, 30% are Consultants, and the remaining 45% are Associate Consultants, Registrars and Medical Officers. Their specialties include Emergency Medicine, Endocrinology, Renal, Gastroenterology, Otolaryngology, Orthopaedic Surgery, Cardiothoracic Surgery, Respiratory &
Critical Care. Each of them owns a personal handheld device which is either a PocketPC device or a Palm device and all were issued with the corresponding ePharmacopoeia version for evaluation.
Questionnaire Design and Analysis
A questionnaire was designed to aid in the evaluation of ePharmacopoeia, focusing on its content and usability, layout and design, and its comparison with other pharmacopoeias. A 5-point Likert scale was used for most questions in which 1 meant “disagree strongly” and 5 “agree strongly”.
In the study, all users used ePharmacopoeia for 4 months in their respective areas of care which included the outpatient clinics, satellite pharmacies and inpatient wards. During this period, they were strongly encouraged to use the program as much as possible, e.g. during ward rounds, at the bed-side, during prescribing. At the end of the study period, all users were required to complete the questionnaire.
RESULTS
The ePharmacopoeia Program
Some sample screen shots of ePharmacopoeia are shown in Figure 1. The Main Index shows the list of drugs, together with the formulary status of the drug and the price indicator. [NA] indicates that the drug is not in the hospital’s formulary. As such, there is no corresponding price and status information, but users are still able to access the drug monograph.
Figure 1: PDA Screen Shots of ePharmacopoeia Information of each drug from Lexi-Drugs is integrated with the hospital’s formulary information.
At a quick glance in the index one can see the list of drugs, together with its formulary status and price indicator. The detailed information for each drug can be retrieved by simply tapping on the drug name in the index. For the drug ‘Acyclovir’, one can see its drug monograph, together with the hospital’s pricing information, dosage forms and formulary status. If users are interested in similar drugs that may be cheaper, they can tap on the therapeutic category link with the respective price indications. For example, selecting Therapeutic Category in the ‘Candesartan’
monogragh will bring up Angiotensin II receptor blockers within the SGH formulary and reveal that Losartan is the cheaper alternative to Candesartan.
The online interface is similar in terms of features, but allows more detail and is more convenient to read.
Survey Results
The survey results are shown in Tables 2 and 3. In terms of content and use, more than 90% of the users found ePharmacopoeia easy to use, reliable, useful and beneficial in the practice setting. 80% found it made their practice more efficient. Only 57% of the users said the program easy to install, and 63% found it error-free.
In the program’s layout and design, more than 70%
of the users found it easy to understand and to find information. However, only 40%-57% found it appealing and easy to read and navigate.
Table 2: Doctors, Nurse Clinicians and Pharmacists’ Perceptions on the Content, Use, Layout & Design Users “Agree or Agree Strongly” with the Questions (%)
Questions Doctor
Table 3: Comparison of ePharmacopoeia with Other Pharmacopoeias Users Find ePharmacopoeia Comparable or Better
n = 30 (100%) Other
Pharmacopoeia In terms of Ease of Use &
In terms of Patient Safety Profile
ePharmacopoeia compared well with other forms of pharmacopoeia. In terms of ease of use and convenience and information updatedness, more than 95% found it similar or better compared to book and online references and 74%-89% found it similar or better compared to other handheld references.
ePharmacopoeia was found to be similar or better compared to all 3 other forms of pharmacopoeias by more than 90% of the users in terms of relevance of information and patient safety profile, and 74%-84%
in terms of information comprehensiveness.
The frequencies of usage are shown in Figures 2 and 3. More than 50% of the users use ePharmacopoeia most often in the wards. This includes 10 doctors, 4 pharmacists & 2 nurse clinicians as shown in Figure 2. Figure 3 shows that 100% of the users use ePharmacopoeia at least 1 to 3 times a week, and of which 20% of them use it more than 3 times a day.
Figure 2: Most Frequency Used Place
0
Home Clinic Office Ward Pharmacy Others
Most Frequently Used Place
Nurse Clinician Pharmacist Doctor
Figure 3: Frequency of Use General Feedback
Users were asked for comments on areas which satisfied and frustrated them the most, as well as areas where the product exceeded and fell short of their expectations. The comments are summarized in Table 4.
Users also commented that ePharmacopoeia generally saves time during information retrieval, is easily incorporated into their usual workflow, and improves drug-related decision making. Doubts and queries that arise at the point-of-care pertaining to medications can be answered through quick referral to the program, and this greatly increases the confidence level of the clinicians whilst caring for their patients. They agreed that the program helps to reduce the rate of preventable adverse drug events.
DISCUSSION
The results show a potential in having a PDSS at the point-of-care. As seen from the results and feedback, most users find it useful and beneficial. More importantly, it can potentially help to reduce medication errors leading to improved patient safety.
It should be highlighted that ePharmacopoeia is not intended to replace other forms of pharmacopoeias.
Depending on the clinical setting, purpose of use and other factors, different pharmacopoeias may be preferable. Thus in the comparison, the benchmark is to ensure that ePharmacopoeia generally does not and should not perform any worse than other pharmacopoeias.
Further investigation was done to ascertain the accuracy of the information provided in the program as only 63% found it error-free. It was found that the comments field of 26 drugs were not inserted correctly during the development. The comments field forms part of the hospital formulary information and indicates if the drug is a controlled drug or if it is restricted to specific specialties only. This was confirmed to be the only source of error and was promptly rectified. There was no significant impact on the decision support at the point-of-care.
Table 4: General Feedback from Users Which parts of the program did you enjoy the most?
• Availability of hospital formulary with prices
• Availability of international brands (e.g. Panadol)
• Ease of use & easy navigation
Which parts of the program frustrated you the most?
• Too wordy
• Appearance of many unfamiliar brand names at the beginning of each drug monograph
• Confusion over icons at the bottom
Which parts of the program exceeded your expectations?
• Very comprehensive information
• Has latest FDA-issued alerts & cautionary reports
• Information available for drugs that are not US-based Which parts of the program fell short of your expectations?
• Some local medications not found (e.g. Saridon, Vasteral, Arcalion)
• Information limited as compared to textbook references
• The mechanism of drug is not included. This is very important to pharmacists
0
CONCLUSION
This study has demonstrated the feasibility and usefulness of a PDSS integrating drug information and formulary information for clinicians at the point-of-care. Interesting to note is the comparison between ePharmacopoeia and other traditional methods of delivering pharmacopoeia information. A clear majority of users found it comparable to or better than the traditional means. Informal interviews with users related that importance had been placed on being able to find information on all the drugs carried by the hospital in the pharmacopoeia. This is a relatively simple request, but this has traditionally been difficult to achieve.
We are now looking into the extension of ePharmacopoeia to other mobile devices such as the smartphone and blackberry, as well as tighter integration with the hospital’s Computerized Physician Order Entry (CPOE) system. For example, the ability to provide context-sensitive information to the requestor from a prescribing field in the CPOE system would be a valuable and convenient tool[10].
As we advance into the era of wireless communication and mobile technology, the clinical and practical value of using these devices in clinical settings will continue to grow further.
REFERENCES
1. Wyatt J. Use and sources of medical knowledge.
Lancet 1991;338:1368-73.
2. Ely JW, Osheroff JA, Ebell MH, et al. Analysis of questions asked by family doctors regarding patient care. BMJ 1999;319:358-61.
3. Ely JW, Osheroff JA, Gorman PN, et al. A taxonomy of generic clinical questions:
classification study. BMJ 2000;321-429-32.
4. Rothschild JM, Lee TH, Bae T, Bates DW.
Clinican use of a palmtop drug reference guide.
JAMIA 2002;9:223-9.
5. Knollmann BC, Smyth BJ, Garnett CE, Salesiotis AN, Gvozdjan DM, Berry NS, Lee H, Min FD. Personal digital assistant-based drug reference software as tools to improve rational prescribing: Benchmark criteria and performance. Clin Pharmacol Ther 2005;78:7-18.
6. Clauson KA, Seamon MJ, Clauson AS, Van TB.
Evaluation of drug information databases for personal digital assistants. Am J Health-Sys Pharm 2004:61:1015-24.
7. Galt KA, Rule AM. PDA-based drug information sources: Potential to assure medication safety. American College of Clinical Pharmacy 2003 Spring Practice and Research Forum.
8. Enders SJ, Enders JM, Holstad SG. Drug-information software for palm operating system personal digital assistants: breadth, clinical dependability, and ease of use. Pharmacotherapy 2002;22:1036-40.
9. Keplar KE, Urbanski CJ. Personal digital assistant applications for the healthcare provider.
Ann Pharmacother 2003 Feb;37(2):287-96.
10. Kaushal R, Shojania KG, Bates DW. Effects of computerized physician order entry and clinical decision support systems on medication safety:
A systematic review. Arch Intern Med 2003 Jun 23;163(12):1409-16.