We’ve written numerous columns on prescription of potentially inappropriate medications in the elderly and noted that such medications are often prescribed during an inpatient hospitalization. Many are started during ICU stays (see our March 2011 What’s New in the Patient Safety World column “Inappropriate Medications Often Start in the ICU”).
Now a new study (Bell 2011) focuses on just the opposite problem – unintentional discontinuation of medications taken for chronic diseases in hospitalized patients. The authors looked at records in Ontario for almost 400,000 patients over a 12-year period for discontinuation of drugs in 5 specific categories (statins, levothyroxine, antiplatelet/anticoagulants, respiratory inhalers, gastric acid suppressants) for patients aged 66 and older. The control group was one with no hospitalizations. The other two groups were those hospitalized with an ICU stay and those hospitalized without an ICU stay.
Compared to the control group that had not been hospitalized, those patients who had been hospitalized had much higher rates of unintended medication discontinuation in all 5 drug categories. Hospitalized patients were almost 80% more likely to discontinue an antiplatelet/anticoagulant medication. Those patients having an ICU stay were statistically more likely to have a drug unintentionally discontinued in 4 of the 5 categories categories (statins, levothyroxine, antiplatelet/anticoagulants, gastric acid suppressants). Those having an ICU stay were almost 50% more likely to discontinue a statin and over twice as likely to discontinue an antiplatelet/anticoagulant medication.
Importantly, these discontinuations were not without consequences. For those patients in whom statins or antiplatelet/anticoagulant drugs were unintentionally discontinued, there was an increase in the composite outcome of death, emergency department visit, or hospitalization in the year following discontinuation.
The study did use administrative data and could not completely verify that the medication discontinuations in some cases were not intentional. However, they did take steps to minimize that possibility and a prior chart review study had provided similar findings.
Whether the findings are generalizable to other patient populations and other medication categories is unknown but we don’t doubt that this problem is more widespread. It is a reflection on our system of fragmented care punctuated by numerous missed communication opportunities and problematic handoffs at multiple levels. Though we cannot take out the human element, there clearly is a system problem that is overriding and we need to redesign the system to make it easier for the players to do the right thing and harder to do the wrong thing.
Though most of us expect technological solutions to the nagging problem of medication reconciliation, some of the best systems we’ve seen have been strictly paper-based. But they require dedicated, focused processes where someone (pharmacist, pharmacy tech, nurse, midlevel, physician, etc.) develops a “Best Possible Medication History” using all available sources and that list is scrutinized by all relevant providers at every transition of care, not just on admission and discharge.
But we do need to utilize our information technologies to help. Maybe there is something to the concept in our August 2011 What’s New in the Patient Safety World column “The Amazon.com Approach to Medication Reconciliation”. Hasan and colleagues (Hasan 2011) borrowed the concept of collaborative filtering to help identify medications omitted from patient medication lists at the time of medication reconciliation. They determined, based on large population databases, that patients who take drug “X” also often take drug “Y”. They established multiple different algorithms and applied them to sample patient data. In fact, their algorithms were able to guess correctly an omitted drug within 10 guesses about 50% of the time (they did even better guessing the therapeutic class of a missing drug). They found some of their algorithms might work better in certain settings or with certain populations. Obviously, the principle might be extended to say patients who have condition “X” also often take drug “Y”. Yes, we are concerned that approach could have unintended consequencnes (i.e. erroneously starting some patients on new medications they were not previously taking) but there certainly is merit in further research into applying this concept to medication reconciliation.
The editorial accompanying the Bell article (Kahn 2011) points out another concerning problem in this population – polypharmacy. They note that the hospitalized patients in the Bell study had a median of 12 medications and 75% received 9 or more medications. So hospitalization could represent an opportunity for medication “optimization”, not just “reconciliation”. We’ve noted in multiple columns that the process of a comprehensive review of patients’ medications often uncovers therapeutic duplications or other unnecessary medications or drug combinations that may be problematic. We usually see an average of 2 medications per patient that might be discontinued or have the dosage altered. But such “optimization” requires involvement of the patient and the primary care provider. The latter is often not included in the medication reconciliation process that takes place at all transitions of care – another communication failure within our system.
Some of our prior columns dealing with medication reconciliation:
Bell CM, Brener SS, Gunraj N, et al. Association of ICU or Hospital Admission With Unintentional Discontinuation of Medications for Chronic Diseases. JAMA 2011; 306(8): 840-847
Hasan S, Duncan GT, Neill DB, Padman R. Automatic detection of omissions in medication lists. JAMIA 2011; 18: 449-458 Published Online First: 29 March 2011
Kahn JM, Angus DC. Going Home on the Right Medications: Prescription Errors and Transitions of Care. JAMA 2011; 306(8): 878-879