MN is a 91-year-old Korean male living in a nursing home since 2011 with a recent change of provider to RMSC. He has moderate dementia difficult to accurately evaluate due to a language barrier. Comorbidities include atrial fibrillation, HTN, BPH, DM II, thrombocythemia and he is hard of hearing. Medications include Glucotrol 5 bid, Lantus 12 u qd, Aspirin 81 mg qd, Calcium 500 bid, Pepcid 20 mg qd, Proscar 5 mg qd, MVI qd, Flomax 0.4 mg qd, Xarelto 20 mg qd, and prns. He was switched from Warfarin to Xarelto several months ago after refusing blood draws. Recent labs include WBC 9000, H/H 11/ 36; MCV 85; RDW 19, Plt 802, Albumin 4.1, Lytes WNL, Glucose 153, GFR 73, Hgb A1C 8.9, TSH 1.65 and B12 672.
On May 8, he slid out of his wheelchair and fell in the hall. There were no signs of an injury noted, though one nurse said he was perhaps more confused than normal. Vitals: Pulse 80, BP 148/82, Resp 20 and temp 98.6. This patient is not considered a high fall risk and has not had other falls in the past 30 days, though he has had occasional noninjury falls in the past 6 months, the most recent of which was in late February. He has not had a recent change in condition, altered mental status or significant change in medications. Neuro checks were started every 15 minutes for 72 hours. Through the evening / night, his vitals and oxygenation remained normal and he was afebrile. At about 8 PM that night, he was noted to be restless, confused and nauseated. The on call provider asked to send him to the ER, though the daughter (who turns out not to be the MDPOA) was called and wanted him to stay and watch him through the night. His MOST form showed that he was a DNR wanting limited additional interventions, appropriate antibiotics and no artificial nutrition or hydration. The patient fell asleep without problems. Unfortunately, he was found unresponsive the next morning with a small amount of coffee ground emesis and a blood sugar of 356. He was transferred to the hospital where it is reported that he sustained a large subdural hematoma and he dies the next day. What are the lessons to be learned?
Dr. Gregory Gahm:
From my point of view, the main issue to point out in this case is when is it appropriate to use both an anticoagulant and an antiplatelet agent? According to the most recently published CHEST Supplement dealing with Antithrombotics and Thrombolytics (2012), there are only 3 times when the intended benefits of concurrent use (decreasing embolic events such as CVA, MI and PE) outweigh the likely risks. Those three situations are:
- Mechanical Valve patients w/ low bleeding risk
- Acute Coronary Syndrome
- Recent Coronary Stents or Bypass Surgery
… that’s it! This patient clearly did not fit in any of these 3 categories. Would his subdural have occurred if he had been on only one of the agents? Perhaps, but we know the risk goes up substantially when patients are on both.
Every other takeaway message in this case pales in comparison to this one. Having said that, there are other potential things to think about, such as:
- Why Glucotrol and Lantus? Both are being used at low doses, so are not really doing much, anyway. Since his renal function is still relatively good, consider getting rid of both and using Metformin, the only chronic treatment choice in type 2 diabetes that has actually been shown to reduce long term morbidity and mortality
- He should be on Vitamin D3 in doses able to raise his OH-D3 level above 40-45 in order to realize the 22% decrease in falls and fractures (levels >30 = “normal”, but are not protective in this regard)
- The MVI is not providing benefit and is not a necessary pill
- He probably gets enough Calcium in his diet to suffice. Once he is on the D3, you could consider discontinuing the calcium as unnecessary
- Whoops! We forgot to ask if the family member telling us and the facility not to send him to the hospital was the MDPOA… Multiple parties culpable here, though mainly the daughter.
Steven Atkinson, P.A.
A sad circumstance for sure.
I really think, given the picture Greg paints, this could have been handled differently. I guess I’ll just jump right in.
- glucotrol… now on the BEERS list. If a secretagogue has to be used, consider glipizide but given the age of this patient and the risk of hypoglycemia, lantus would probably suffice. The AGS supports an A1c between 6.5-8% but a recent observational study (mean age 80) suggested that elders may be “better off” with an A1c of 8-9% (relative risk, 0.88).1
- xarelto has an incidence of head bleeds .5% and coumadin has an incidence of .8% (see package insert); this ROCKET-AF trial included 14,000 patients and 2 years of f/u. Pradaxa has similar data… but the risk of bleeding in the head increases with age and this is the consideration which may have been overlooked. Perhaps the right answer would have been to stick with ASA alone which yields, for primary prevention, a 20 to 30% risk reduction whereas warfarin reduces the risk of stroke approximately 68%.2
- There are other meds that are worth discussing but I think the most valuable are mentioned above.
- Yau CK et al. Glycosylated hemoglobin and functional decline in community-dwelling nursing home–eligible elderly adults with diabetes mellitus. J Am Geriatr Soc 2012 Jul; 60:1215.
- Howard PA. Guidelines for stroke prevention in patients with atrial fibrillation. Drugs.1999 Dec;58(6):997-1009.
Dr. Fred Feinsod
Agree with Steve that Lantus would have been better for blood sugar control. Consider metformin rather than glipizide. His medication burden could also have been reduced by eliminating MVI and calcium. Vitamin D3 was a choice for consideration, especially if his level was <40. This may have impacted his fall risk in a positive manner. Did he need Flomax which also may have impacted his fall risk in a negative way.
Agree with the thought of ASA rather than Xarelto, especially with a high RDW. Stool guaiac may have been contributory. Did the Resident/DMPOA know of this choice (ASA versus anticoagulant)? Consider informed consent for anti coagulants (warfarin, Xa inhibitors, and direct thrombin inhibitors) in the elderly since morbidity/mortality associated with these medications is significant.