Prescription Management

From TAPASWiki

Jump to: navigation, search

The prescription manager allows users to input, manage, and print prescriptions for patients as part of the medical summary. A streamlined prescription management system is seen by the users as a key feature of a summary tool.

This page discusses some of the design decisions for the prescription manager in TAPAS.

Contents

Frequencies, Durations, and Quantities

TAPAS will calculate end date of prescriptions from a combination of start date, duration and frequency. For the most part this will be quite straight forward. For example: starting April 1 take one tablet twice a day for 2 weeks makes the prescription contain 28 tablets and end on April 13th.

Calculating End Dates on Frequency Ranges

One of the challenges comes from medications that can be taken at a variable time: some medications are commonly prescribed “PRN” (as needed) or in a range (every 4-6 hours). How should we calculate these quantities and durations?

For the purposes of calculating an end date, the minimum time will be used (i.e. the shorter end date in a range). That date is the more clinically relevant value - we are more interested if people are taking MORE of a medication than prescribed than less.


Health Canada Drug Database

openTAPAS has permission to use Health Canada Drug Product Database (DPD). This database is a listing of all active drugs approved in Canada. The relational database is available as a series of text files from the link above.

Unfortunately, the DB is focused on drug brands and DIN numbers, making the prescription of generic drugs awkward. Therefore, we propose to provide some additions that will better support prescription of generic medications (see below).

We will not be using aspects of the database that are not relevant to presribing of medications in primary care such as:

  • Veterinary Medications
  • Disinfectants
  • Packaging Information
  • Veterinary Species
  • Pharmaceutical Company information
  • Inactive Product Information

Using Defaults from the DPD to speed data input

The DPD has several fields that will help speed up data entry for the clinician. For example, the selection of a medication provides some additional information:

  • Default Route
  • Strength and Unit of Strength
  • Form (eg tablet, suspension, etc)

NOTE: These do not map completely to the HL7 MedAdministrationRoute table of e-MS. There are routes (such as IM) that are not in the e-MS.

We will map the default routes, where possible to medications and populate those routes. The default can be over-ridden by the user.

Strength cannot be overwritten - the medication needs to be changed by the user. The # of tablets, however, can be changed.

Default Route Mapping to e-MS CDA

To be added.


Generic Medications

It is preferred that medications be prescribed by their generic name (eg Ramipril vs Altace). With the existing DB this is not always possible because the DPD is brand name centric (with good reason - it records all medications that can be prescribed).

Medications are listed with DIN # (Drug Identification Number) and their labeled brand name. For many medications this may often include a drug strength in the title, but this is not consistent. Users, if they are picking a name straight of the list, are, therefore, able to then pick a drug name without knowing its strength, which is not a good option.

With generic drugs, such as hydrochlorothiazide, there are many many generics, all of which get listed in a search (apo-hydrochlorothiazide, dom-hydrochlorothiazide, nu-hydrochlorothiazide, etc etc). Not good usability. The users are getting confused with which drug to pick.

The DB does contain active ingredients (in the ingred.txt table) which is helpful as well as corresponding strength. We are then able to create a list of generic labels that include the strength, unit, and active ingredient.

This would create a unique, generic name that the use can pick from for all medications. It would fail with any compound medications which have multiple active ingredients. We could model them with a list of ingredients, but that would be too long a name. Besides, most clinicians prescribe those medications by their brand names (Septra, Tylenol #3, Diovan HCT, etc). The pharmacists then pick the appropriate generic name.

THE PROPOSAL:

Create a new "Search_Name" in the Drug Table.

  • Combine "Brand_Name" and all "ingredients" associated with each drug.

Create a new "Prescription_Name" field in the Drug Table.

  • Where number of active ingredients = 1 (NOTE this is a field in the drug table) Prescription_Name = Strength & Strength_Unit & Ingredient
  • Where Number_of_AI > 1 use the Brand_Name

Cull the Database of Duplicates:

  • Remove veterinary and disinfectant elements (obviously - this would be silly to keep in and I know they are removed, but for completeness...)
  • Remove all duplicate "Prescription_Name" elements.

Sort "Prescription_Name" ascending so the user knows how to find things.

This should cut the DB down considerably - speeding searching both from the computer's perspective but also (and more significantly) from the user's perspective as the potential number of elements will be dramatically reduced.

Using the search name allows users to type in "Altace" and get ramipril as there would be only one ramipril for each dose, the altace from the search name would not be deleted.

LIMITATION 1: The user could not prescribe a specific generic brand. This should rarely be a problem, however, as in practice a generic is a generic is a generic. For medications like amoxicillin, the user could not pick Trimox or Amoxil, only amoxicillin. It could be written into the free text comments if the user needed to prescribe "APO-amoxicilin" with no substitutions if this was an issue.

LIMITATION 2: The TAPAS system would not be able to reflect all DIN numbers. This is not a significant limitation, however, as the prescription of a medication is rarely DIN specific - it is the dispensing that is DIN specific. That is the MD writes the prescription. The pharmacist selects the specific medication.


FUTURE THOUGHT: It may be worthwhile considering if we can expand "Search_Names" to include all names of medications that are being deleted / "compressed" in this model.


Drug Schedules

At this time we will leave all human drug schedules in our look up tables, this will include homeopathic, over the counter, and narcotic prescriptions. We may, based on user feedback, change this.

NOTE: Schedule is different that frequency (# of times / day) which is not in the DB. (See below)


Table of Dosage Frequencies

Below is a table showing the frequency elements described in the e-MS CDA implementation guide. Comments are provided to discuss reasons and challenges for each.

The order of the table corresponds to how the list should be shown in this version of TAPAS - it makes sense to one clinician. Feedback is appreciated.

NOTE: With the help of Mowafa Househ from the University of Victoria, we have developed a revised list based on commonly prescribed medications in the US as reported by RxList in their Top 200 List. All medications that were prescribed more than 10 million times on that list were summed into frequency categories to create the following list of most common frequencies:

Frequency Number of Prescriptions
od 793898613
tid151271119
bid133101299
q4-6h120651846
qam99606471
qhs54031811
qpm49254483
q12h41393538
weekly31352761
q4h30050056
q6h23665172
q6-8h18873635

From this list of common prescribing frequencies, the TAPAS Rx frequency list has been modified. The list does not match the above list completely as we have clustered and ordered the progression of clusters (eg Q1h, Q2h,... Q12h etc). Currently we included all elements in the e-MS list except "STAT" as that is not a frequency. Future modification will come from user feedback during testing.

Frequency DescriptionComments
QD Once a day This should be the default for a new medication unless we can extract the default from the Health Canada database, which would be the better choice to reduce errors and improve usability.
BID Two times a day
TID 3 times a day
QID Four times a day
QAM Every morning
QNOON Once a day, at noon
QPM Once a day, in the evening
QHS Once a day, at night
Q1H Every Hour Q1H and the more frequent dosage rates are used less than Q8h, Q12h, especially in primary care. However, for consistency, they are placed here in order. BID, TID, QID cover the most frequent ones through their overlap (Q12h, Q8h, Q6h respectively). Usability testing will show us if we need to change it.
Q1-2H Every 1-2 hours
Q2H Every 2 hours
Q2-3H Every 2-3 hours
Q3H Every 3 hours
Q3-4H Every 3-4 hours
Q4H Every 4 hours
Q4-6H Every 4 to 6 hours
Q6H Every 6 hours
Q6-8H Every 6 to 8 hours
Q8H Every 8 hours
Q8-12H Every 8 to 12 hours
Q12H Every 12 hours
Q2D Every 2 days used less frequently
1/weekOnce a week
2/week Twice a week
1/5day Once every 5 days
1/6day Once every 6 days
21/28D 21 out of 28 days Used for the oral contraceptive pill, primarily. For calculations of duration, this will be considered a once a day medication. Doses will not be allowed in the calculation of duration.
1/month Once a month
ONCE A single dose Useful for some antibiotics and acute medications (eg for pain). This is more used in the hospital and will be used less often in the clinic setting.
OTH OtherUsed for anything not considered until now. The end date would be undefined if a duration was calculated in doses. The user must be able to define use in free text.
STAT Now Stat is removed from TAPAS

Some of the rarer items on the list are not going to be used too often. They are here for completeness and may, based on user feedback, be pulled from the UI.

Comments about the e-MS

Part of our project is to provide feedback to the e-MS project to see where the developing standard can be enhanced. This section is here to add pieces as they come up in the prescription module.

In the rare case when the user selects “OTH” the duration will need to be defined in time quantities (days, weeks, months) and, therefore the “doses” option in durations will not be available. This may have to be clarified in the eMS specification.

“Stat” has been removed from TAPAS as it is not a frequency, but an urgency value. If requested by the users, this can be added back in as a qualifier.

For the calculation of a duration of X doses and a frequency of “OTH” it will be impossible to calculate the end date. In this rare case, we will define the end date as nil - as the system cannot define it.

For “21 out of 28 days” the unit of dosing is a package that lasts for 21 days (or 28 days, depending on the packet). Unfortunately, there isn’t a good way, using the e-MS structure to properly quantify the duration of the medication in a completely foolproof way. For now, we are going to simply calculate the duration as a once a day medication and disallow doses as we do for “other”.

Routes

Medication routes in the e-MS are based on HL7 data. However, they appear not to be comprehensive. A particular miss is "intramuscular".

Personal tools