Among our numerous columns on potentially inappropriate medication use in the elderly, we’ve done a few specifically on deprescribing (see our Patient Safety Tips of the Week for March 4, 2014 “Evidence-Based Prescribing and Deprescribing in the Elderly” and September 30, 2014 “More on Deprescribing” and our May 2015 What’s New in the Patient Safety World column “Hospitalization: Missed Opportunity to Deprescribe”).
Since our last column on potentially inappropriate medications (PIM’s) in the elderly there has been yet another study showing the problem may be getting worse rather than improving. Lund and colleagues found that despite intervention studies demonstrating up to 80% reduction in PIM use during acute hospitalization, a significant increase in PIM use was observed in a naturalistic setting in Medicare beneficiaries with acute MI (Lund 2015).
But two papers have proposed excellent approaches to minimize PIM use and facilitate deprescribing. As we noted in our May 2015 What’s New in the Patient Safety World column “Hospitalization: Missed Opportunity to Deprescribe” a hospitalization provides a logical time to determine whether a patient is a good candidate for deprescribing. And one group did just that. They implemented a brown bag medication reconciliation process in the hospital setting to decrease medication discrepancies by encouraging evaluation of medication adherence, side effects, and monitoring at posthospitalization follow-up (Becker 2015). After implementation, a 7% decrease in reportable errors was noted.
We had discussed some of the good work of Scott and colleagues on deprescribing in several of our previous columns on deprescribing. Now that group from Australia has put together an excellent 5-step protocol to aid the deprescribing process (Scott 2015):
The Scott paper has an excellent discussion about identifying the reasons for each drug and determining whether the drug was prescribed for symptom or disease control or for prevention. It stresses looking at whether the condition for which the drug was originally prescribed is still present and merits continuation. And it stresses looking at the risk:benefit ratio of the drug in the context of expected lifespan. The Scott paper notes that predicting lifespan is very difficult. However, it notes that the “surprise question” (i.e. “would you be surprised if this patient were to die within the next 12 months?”) is reasonably predictive. Note that this might also be one of the few potential uses of a tool recently developed that accurately predicts death in recently hospitalized patients (van Walraven 2015). While that tool seems to be quite accurate, we did not think that either physicians or patients were likely to want to use it. The Hospital patient One-year Mortality Risk (HOMR) model, using easily available administrative data and originally derived and internally validated to predict the risk of death within 1 year after admission, was recently validated externally in three medical centers in Canada and the US. The HOMR score was strongly and significantly associated with risk of death in all populations and was highly discriminative. The authors felt the HOMR model might be useful for risk adjustment in analyses of health administrative data to predict long-term survival among hospital patients. While the time window for the HOMR score is one year, one might identify patients with a high likelihood of death within one year who would not likely benefit from continued use of drugs with a longer time horizon (eg. bisphosphonates).
The Scott paper also notes that patients may no longer derive any benefits from some drugs. They note patients who no longer need antihypertensives because they are now normotensive in response to lifestyle modifications. Or they note patients who were originally prescribed nitrates for chest pain that did not turn out to be of cardiac origin.
The Scott paper has a nice algorithm for deciding the order and mode in which each drug could be discontinued. Their algorithm also asks whether it is likely the patient would have withdrawal symptoms or a disease recurrence if the drug were discontinued. If so, tapering the drug and monitoring for adverse effects is indicated.
The paper also has a good discussion about patient-level and physician-level barriers to deprescribing and system-level strategies that might help promote deprescribing.
Particularly as we move forward with healthcare models like accountable care organizations and population health management, it makes both clinical and financial sense to take a hard look at what we might do better to both avoid potentially inappropriate medications in the first place and to consider deprescribing in those already on them. Such programs can reduce adverse drug events and improve patient satisfaction at the same time they save money (from direct drug costs and costs associated with adverse drug events).
Some of our past columns on Deprescribing in the Elderly:
Some of our past columns on Beers’ List and Inappropriate Prescribing in the Elderly:
Lund BC, Schroeder MC, Middendorff G, Brooks JM. Effect of Hospitalization on Inappropriate Prescribing in Elderly Medicare Beneficiaries. J Amer Geriat Soc 2015; 63(4): 699-707
Becker D.Implementation of a Bag Medication Reconciliation Initiative to Decrease Posthospitalization Medication Discrepancies. Journal of Nursing Care Quality 2015; 30(3): 220-225
Scott IA, Hilmer SN, Reeve E, et al. Reducing Inappropriate Polypharmacy. The Process of Deprescribing. JAMA Intern Med 2015; 175(5): 827-834
van Walraven C, McAlister FA, Bakal JA, et al. External validation of the Hospital-patient One-year Mortality Risk (HOMR) model for predicting death within 1 year after hospital admission. CMAJ 2015; First published online June 8, 2015
Arguably the most important interventions in reducing CLABSI’s (central line-associated bloodstream infections) are avoiding central venous catheters in the first place and minimizing their duration when there is a legitimate initial indication. All too often central lines (which also include PICC catheters) are placed for reasons that may not be optimal (see our December 2014 What’s New in the Patient Safety World column “Surprise Central Lines”).
A recent study showed that use of an online physician audit tool led to a substantial reduction in central venous catheter use (McDonald 2015). The tool listed several potential indications for central venous catheters or PICC’s and also allowed for input of “other” reasons. Its use resulted in a 46.6% reduction in use of CVC’s or PICC’s compared to the pre-intervention period. But the investigators still found substantial room for improvement in that a third of the CVC’s/PICC’s were used for “ease of drawing blood” or “just-in-case the patient deteriorates”.
The success of this intervention should not be surprising. Way back in the early 1990’s one of our first patient safety interventions was placing a brightly colored sticker on the chart of patients with indwelling urinary catheters requiring the physician to specify the indication and intended duration of the catheter (see our May 8, 2007 Patient Safety Tip of the Week “Doctor, when do I get this red rubber hose removed?”). That simple intervention led to almost a 50% reduction in Foley catheter use. Later we used alerts delivered during CPOE and saw even further reductions in Foley catheter use. So merely making the physician think about the indications for any catheters can help reduce unnecessary use.
We’ve discussed unnecessary central venous catheter or PICC line use in our previous columns of January 21, 2014 “The PICC Myth” and December 2014 “Surprise Central Lines”. In the latter we noted that oversight is needed to avoid what we consider an example of a predictable error and put systems in place to help us avoid the problem. One of the items on our checklist for daily rounds on patients in all locations should be “Does this patient have any catheters or lines in place and, if so, are they still necessary?” Use of such lines should be evidence-based where possible. Alert fatigue aside, we also recommend that flags be set in the electronic medical record (EMR) to highlight for the clinician that such catheters are in place and need to be reviewed for continuation on a daily basis.
But the best interventions are those that can be delivered before a central venous catheter or PICC line are placed. There are two ways to do this. One is to require an order in CPOE that requests the indication before the catheter insertion kit is dispensed. The other is a paper-based checklist with indications that is filled out before the catheter insertion kit can be opened. We actually prefer the latter since your checklist can and should also include all the elements for proper catheter insertion, i.e. the classic Pronovost checklist (see our March 2011 What’s New in the Patient Safety World column “Michigan ICU Collaborative Wins Big”), and correct patient and procedure verification.
As a related aside, in our experience we have seen more “other” indications pop up when a prompt appears for justification of a central line or catheter already in place than in those checklists that must be filled out prior to catheter placement. That does not take away from the need to assess the need for continued use of a central line on a daily basis but does again emphasize that delivering the prompt before catheter placement is more likely to result in catheter placement only for appropriate indications.
McDonald EG, Lee TC. Reduction of Central Venous Catheter Use in Medical Inpatients Through Regular Physician Audits Using an Online Tool. JAMA Intern Med 2015; Published online May 04, 2015
One of the most egregious errors in healthcare is failure to follow up on significant test results. These may be results of bloodwork, other laboratory tests, pathology studies, imaging studies, and other diagnostic tests.
In our October 13, 2009 Patient Safety Tip of the Week “Slipping Through the Cracks” we highlighted a study that showed almost 8% of patients with critical imaging results had not received appropriate follow-up within 4 weeks (Singh 2009).
Researchers from that same group have now developed and validated a “trigger” tool to detect such delays in treatment for patients having abnormal results of lung imaging studies (Murphy 2015). The researchers developed an electronic health record (EHR) algorithm based upon reports of chest CT images and/or conventional chest radiography that had been read as “suspicious for malignancy”. The algorithm excluded patients in whom appropriate and timely follow up had taken place (such as a pulmonary visit) or cases in which no follow up was necessary (eg. patients with terminal illness). The trigger algorithm was retrospectively applied to the records of over 89,000 patients. Of 538 records with an imaging report that was flagged as suspicious for malignancy, 131 were identified by the trigger as being high risk for delayed diagnostic evaluation. Manual chart reviews were then done and confirmed a true absence of follow-up in 75 cases, of which four received a diagnosis of primary lung cancer within the subsequent 2 years. The positive predictive value (PPV) for the trigger was thus 57.3% for detecting evaluation delays. The researchers now plan to apply the trigger tool prospectively.
We clearly need ways to ensure that patients with significant imaging (or lab) findings have had appropriate follow-up. Phone calls and emails may be cumbersome if they need to be made on all patients with such findings. However, using trigger tools such as the one described here can filter down the list of cases needing such phone calls or email follow-ups can go a long way to ensuring no one “falls through the cracks”.
See also our other columns on communicating significant results:
Singh H, Thomas EJ, Mani S, et al. Timely Follow-up of Abnormal Diagnostic Imaging Test Results in an Outpatient Setting. Arch Intern Med. 2009; 169(17): 1578-1586
Murphy DR, Thomas EJ, Meyer AND, Singh H. Development and Validation of Electronic Health Record–based Triggers to Detect Delays in Follow-up of Abnormal Lung Imaging Findings. Radiology 2015; published ahead of print May 11, 2015
The Canadian Patient Safety Institute has developed a patient safety and incident management toolkit (CPSI 2015) that has some resources we think you will find very valuable. The toolkit is quite comprehensive and breaks down into 3 sections: (1) system factors (2) patient safety management and (3) incident management.
In the “system factors” section you find, in addition to those organizational issues you’d expect, factors “outside the boundaries” of your organization. These include engaging the public, addressing legislative and regulatory issues, and others.
The patient safety management section has excellent tools and recommendations regarding patient safety culture, promoting teamwork and communication, identifying and prioritizing patient safety goals, monitoring methods, reporting systems and learning systems.
The incident management section obviously deals with incident investigation and root cause analysis but also has good sections on immediate response, disclosure, and follow through plus sections on closing the loop and learning lessons.
The toolkit also has a great list of links to patient safety resources.
And, of course, we’d be remiss if we did not note another great resource just released: NPSF has published a comprehensive resource on doing root cause analyses (NPSF 2015). We’ll likely be doing a full column on that next month.
CPSI (Canadian Patient Safety Institute). Patient Safety and Incident Management Toolkit. CPSI 2015
NPSF (National Patient Safety Foundation) RCA2. Improving Root Cause Analyses and Actions to Prevent Harm. NPSF 2015
Print “PDF version”