Showing posts with label reconciliation. Show all posts
Showing posts with label reconciliation. Show all posts

Wednesday, 28 December 2011

Scenarios for good use of medication information

A system for keeping track of medication data would eventually see these scenarios:
(scenarios --> refer to available data, and what kind of static initial conditions can exist)

-    Patient’s medication is inside the same system and follows the basic scenario workflows – prescribed, and validated, dispensed and administered exactly as prescribed. CMPD.

-    (Chronic) patient is inside the hospital and all his medication activity is recorded inside the same system and follows the basic scenario workflows – prescribed, and validated, dispensed and administered exactly as prescribed

-    (Chronic) patient is inside the hospital and all his medication is inside the same system and follows the basic scenario workflows. In addition, patien has a contrasted exam and a surgery in the hospital.

-    (Chronic) patient is inside the hospital and all his medication is inside the same system and follows the basic scenario workflows, but medicaton is replaced in the Pharmacy by an equivalent

-    Patient is admitted to new hospital and reports known history to admitting physician

-    Patient is admitted to new hospital and history is fetched from community repository

-    Patient is admitted to new hospital and history is fetched from community repository, but from wrong patient

-    Patient is admitted to new hospital and history is fetched from community repository, patient reports additional medication to admitting physician

-    Patient is admitted to new hospital and history is fetched from community repository, patient reports verbal dose changes to admitting physician

-    Patient is admitted to a country but his medication history is available in another country

-    Patient is admitted to hospital for surgery. Search reveals that there is no current medication, but many years ago, patient took antidepressives that may still affect anesthesia. Complete history for the last X months (for vertain type of medication) is fetched from community repository in another country

-    Patient has a new treatment. In one of the therapies, there is a mismatch between the expected treatment start time and the real dispense time.

-    Patient has a continued treatment. In one of the therapies, there is a mismatch between the expected treatment start time and the real dispense time.

-    Patient is admitted to the hospital. Admitting physician builds the patient’s medication history from all sources and other consumers in the hospital reuse that information.

-    Patient is admitted to the hospital with infection. Patient refers previous antibiotic treatment followed correctly, and prescription and dispense data are according to patient’s statement. Antibiogram indicates treatment was not effective, and patient admits that 2 doses were skipped and taken after time.

-    Phychiatric patient reports adherence. Physician suspects the opposite.

-    Patient is admitted with complications, physician suspects the use of oral contraceptive but patient denies it for cultural reasons.


AND...
A system for keeping track of medication data would eventually be used in these cases:
(use cases --> refer to processing of the data, and what kind of dynamic behaviour can be expected)

- The health authorities decide to investigate the long-term effects of a counterfeit medication. Using unique medication item identification (see ePedigree), the patients that took the counterfeit medication are identified, and this is checked with the condition that was being treated with the medication. The patients are identified and called for a consultation.

- To eradicate H. Pylori, the physician wants to determine the ideal therapy. The physician consults the patient's previous treatments, which shows that a previous protocol was followed by the patient at home. The physician wants to investigate whether the treatment was properly followed, including administration times, to investigate whether there can be some resistance to the previous antibiotics.



I'll later post the features that are required by these use cases.



Real use cases also to follow.

Monday, 5 September 2011

Medication Documentation

Looking at Medication Reconciliation, it seems that before looking at reconciliation (i.e. solving apparently contradictory information) we must look at the data and its meaning.
In medication the problem is very simple (the solution less simple): Except in institutional settings, it is usually not possible to determine what medication is a patient really taking.

Prescribed and dispensed medication may be a good indication of what has been administered, but not 100% accurate:
 - Patient may be taking medication that was not prescribed;
 - Dispense does not indicate the adherence to the therapy
 - Patient's statement may not be accurate or trustworthy.

Other challenges:
- How to consider the personal medication dispensers? How to take into consideration that barriers to reporting (cultural, others) may exist (e.g. contraceptive pill)?
- What is current? Last 6 months? Last 6 years? May vary across sites, or for different medication types.

This started with an effort to query a patient's current medication in IHE. "Current Medication" was not a steady target, which led to this proposal:
https://docs.google.com/leaf?id=0BzWchnBwBx3eN2EwM2M3MTUtMDY1MC00ODQ0LTlmMTUtMTEzNmM4MWViYjNk&hl=en_US

Short version:
  1. We store the data and its clear meaning
  2. The data must be queriable. It may be possible or intended to display this data.
  3. Any view or interpretation of that data that has reduced quality can also be stored, as long as the meaning is also stored with it.
  4. These views or interpretations are also queriable
  5. Original data may be relevant for display. One example of displaying detailed information in a perceivable view:
Other, simpler views are also possible.

A white paper is now being prepared for medication.

This same approach would also apply for other areas where information is distributed, and there is no single consensual interpretation of a view - or when there are many possible and desired views.
- Radiation exposure
-...