Measuring the Success of Your Reconciling Activities
A set of standard tools for calculating core measures to evaluate your reconciling activities is available to the Reconciling Medications Collaborative. Chart abstraction protocols and electronic spreadsheets are available from the Coalition.
Some Frequently Asked Questions
Should my baseline measurement be institution wide, or just for the unit where I am initiating my first cycle of tests?
We recommend that you pull a random set of charts institution-wide to develop a global baseline measure but that you also review 20 charts on the specific unit before you begin. The institution-wide measure may provide some important insights about where to start, and it will also serve as a benchmark for gauging the penetration of your reconciling activities throughout your organization. The unit-level baseline measurement would be followed by monthly data collection on that unit. The institution-wide evaluation could be less frequent (e.g. six month intervals).
What do you mean when you say “pull a random set of charts”?
When you’re evaluating a unit where you’re testing, you should always pull a random set of charts from that unit. Some hospitals have made the mistake of selecting charts only from patients for whom reconciling has been attempted. You really want to see how well you are doing unit-wide, so you need to be sure your sample covers all patients. That way, your results will reflect both how successful you are at actually resolving variances for the patients for whom you attempted to reconcile medications and what percent of patients you even tried.
When I’m calculating the percent of medications with an error, do I count just the medications on the home list or include any new meds ordered too?
Since the primary initial focus of this reconciling initiative is on how accurately home medications are ordered at admission, we recommend using the number of medications on the home medication list you develop as the denominator. Your review would attempt to identify how many of those were accurately ordered at admission (or annotated to reflect reason to discontinue (or change dose, etc). The review of medications at discharge involves looking at the accuracy of discharge orders for that same list of home medications as well as the last set of medication orders from the patient’s MAR. Hospitals moving to the next phase of implementing reconciling at discharge would then expand the number of medications in the denominator to include both the home med list and any additional medications last set of medication orders from the patient’s MAR.
How have people used their baseline measurement to help guide their process?
Hospitals have used their baseline assessment to identify key breakdown points. For example, one hospital identified failure to restart psychotropic medications at discharge from intensive care units as a particular problem to address. Hospitals have also successfully used their results to build the case for the importance of reconciling. Reading through the examples of what hospitals have learned from their baseline assessment (Collaborative binder, p. 93+) before you begin your assessment may give you some ideas about what to look for.
When looking through the patient’s charts, I’m having a hard time finding complete information on the home medications. What should I do?
For this retrospective analysis, you can expect to encounter instances where there is inadequate information to truly evaluate whether there were medication variances that should have been reconciled. When in doubt, count “unknown” as a reconciling failure.For example:
What really constitutes an unreconciled medication? Often there are clear clinical reasons why a med would be discontinued, so should these really be counted as an error just because it wasn’t documented?Hospitals are using their own judgement in their evaluations. We are encouraging you to error on the side of over-counting errors to help encourage more rigorous documentation in patient charts, a key component of the safety principals embedded in improving communication and information transfer among caregivers. One hospital handled this problem by calculating the baseline measure two ways, with all unknowns counted as errors and under a more lenient definition. Some examples of underlying assumptions:
How will the results be used?
The primary purpose for collecting data is to gauge your own performance over time. The assessment process will remain somewhat subjective as different organizations use different guidelines in determining what constitutes an error, and hospitals are working in a range of different settings with different levels of risk, making cross-hospital comparisons difficult. We will encourage data sharing, however, to offer some rough benchmarking, and we also hope to develop composite measures of improvement.
Why are there different forms for the baseline assessment and ongoing measures?
The forms are essentially similar, with the Ongoing Assessment Form just modified slightly to allow you to record additional data that is only available once you start reconciling. The individual patient collection form used in the baseline assessment will not be needed for any patients who have reconciling forms in their charts. When one is not there, then you can use either Form I or your own reconciling form to collect information on whether there were any omissions or variances between the home meds and the orders.
The Collaborative has presented two different measures; which should I use?
The Collaborative is recommending that all hospitals evaluate their success in reconciling medications with a measure of the “Percent of Medications Unreconciled”. Originally, the Collaborative worked with a measure of “Reconciling Errors per 100 Admissions”, but hospitals found the units for that measure difficult to explain and therefore recommended expressing reconciling failures as a percentage of all medications reviewed. The numerator in both measures is identical, but the current measure divides the number of unreconciled medications by the total number of medications, rather than by the number of patient charts reviewed.
Document last updated 9/8/03.