Patient Safety Tip of the Week


December 1, 2009     Patient Safety Doesn’t End at Discharge



Have you ever heard this response when you are discussing DVT prophylaxis with one of your physicians: “He had those same risk factors for DVT before he came into the hospital and we didn’t have him on DVT prophylaxis at home.”? That physician, of course, has a very valid argument.


Admittedly, hospitalization has additional risk factors (eg. relative immobility, factors related to acute infection, etc.) that further increase the risk for DVT.  But many of those patients had been relatively immobile at home. And when we discharge many of those patients, we know they will continue to be relatively immobile at home and their other risk factors for DVT will persist as well. The problem is nobody knows how long we should continue DVT pharmacoprophlaxis and whether it impacts clinically important patient outcomes (like mortality, rehospitalization, etc.). With few exceptions, the evidence base is simply lacking. There have been studies on the continued DVT risk after certain orthopedic injuries or procedures and those have resulted in some guidelines for duration of prophylaxis after discharge for those procedures.


Now, oncologists are beginning to ask the same question. A recent consensus statement (Khorana et al 2009) from multiple oncology organizations on the use of DVT prophylaxis in cancer patients was published in the Journal of Clinical Oncology. While it notes that prophylaxis in ambulatory cancer patients is not currently recommended (with certain specific exceptions such as patients receiving thalidomide or lenalidomide therapy), it issues a call to action encouraging research to determine the role of prophylaxis in ambulatory patients and to study its impact on mortality. However, they also noted that the prophylaxis recommendations for ambulatory patients may change with use of the thrombosis risk assessment tool developed at the University of Rochester School of Medicine and Dentistry (Khorana et al 2008).


That tool was derived from a large database of cancer patients undergoing chemotherapy and then validated on an independent group of cancer patients. It utilized the primary site of cancer, platelet count, leukocyte count, hemoglobin level, use of erythropoiesis-stimulating agents, and body mass index as predictive factors. Patients whom they were able to classify as being in the highest risk group had almost a 7% risk of symptomatic thrombotic events. Most of those thrombotic events occurred during the first 1-2 cycles of chemotherapy. They have suggested use of this tool in selecting patients in studies looking at ambulatory prophylaxis. Obviously, such studies will need to weigh the risks as well as benefits and consider other key questions such as impact on ultimate survival and cost. We fully expect that in the next year or so we will see emerging guidelines on identification of patient populations for whom ambulatory prophylaxis will be recommended.


But let’s extend the same basic question to some other patient safety interventions we do on inpatients and ask “why don’t we continue these after discharge?”. Think of all the interventions you currently do on inpatients. You do assessments and interventions for fall risk, decubitus risk, delirium risk, hypercapnia risk, high-alert medications, and probably several other conditions.


Let’s consider fall risk. Falls, by the way, used to be one of the commonest causes for readmissions after hospitalizations (Mahoney et al 1994). We don’t have any current statistics on that but would not be surprised if they still rank high on the list. Pretty much every hospital has a program for fall risk assessment and then uses interventions based upon the results of that assessment. Some of those interventions are generic ones applied to patients who score at high risk levels (see our comments in our August 4 2009 Patient Safety Tip of the Week “Faulty Fall Risk Assessments?”). Others are more targeted interventions to address the specific risk items for the individual patient.


Since many in-hospital falls occur during toileting activities, one of the key things you likely do is attend to the toileting needs of the fall-prone patient. You probably make sure he/she uses the toilet shortly before he goes to sleep each night and probably try to time doses of any diuretics so they are less likely to cause nocturia. But how many of you really go over those same interventions carefully with the patient and their family/caregiver at the time of discharge. We find that such important interventions are easily overlooked once the patient returns home.


Similarly, medications are usually one of the prime contributing factors to falls. Drugs in the following classes have been linked to increased risk of falls, particularly in the elderly: sedative/hypnotics, antidepressants (particularly tricyclics), benzodiazepines (particularly long-acting ones), anticonvulsants, antipsychotics/neuroleptics, and drugs that cause orthostatic hypotension. Many of these drugs are on Beer’s list (see our January 15, 2008 Patient Safety Tip of the Week  Managing Dangerous Medications in the Elderly ” and June 2008 What’s New in the Patient Safety World “Potentially Inappropriate Medication Use in Elderly Hospitalized Patients”) and should generally be avoided in the elderly.


Discontinuation of these drugs makes sense, though there really are no randomized controlled trials out there documenting that this strategy successfully reduces falls (Ferreri et al 2008). Nevertheless, we often discontinue drugs in these categories while the patient is an inpatient. However, we often fail to convey to the patients’ other physicians the reasons we discontinued them, only to see the drugs restarted once the patient has been discharged. Ironically, a hospital admission offers the ideal opportunity to get patients off these drugs that may increase their fall risk. Firstly, some may need to be tapered while the patient is under observation, because withdrawal syndromes may be seen. More important is a phenomenon we have often mentioned in the past. Physicians, when informed that their patient is on one of the Beer’s List drugs, will almost never discontinue those drugs. They may reduce their new prescriptions for those drugs in other patients but they almost never go back and take patients off the drugs they are already taking. So our ability to discontinue these drugs while the patient is hospitalized often gets the patient back to a state where the outside physicians do not have to take a patient off the drugs. So again, discharge ought to be the ideal opportunity to discuss the reasons for drug discontinuation not only with the patient, but also with the family/caregiver and the primary care physician (or other physicians rendering ambulatory care). But how many of you really take the time to do that?


At any rate, if you considered the patient to be at high risk for falls while hospitalized, you can bet that he/she remains at high risk after discharge. In fact, recent hospitalization is often listed as a risk factor for falls in the community-dwellling elderly. Addressing that risk is an important part of the discharge process.


So our basic premise is simple: if you do a patient safety risk assessment and intervention on the inpatient side, you must ask yourself if you (or someone else) should also be doing it after discharge. Patient safety doesn’t end at discharge.






Khorana AA, Streiff MB, Farge D, et al. Venous Thromboembolism Prophylaxis and Treatment in Cancer: A Consensus Statement of Major Guidelines Panels and Call to Action. Journal of Clinical Oncology 2009; 27: 4919-4926



Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 2008; 111: 4902 – 4907



Mahoney J, Sager M, Dunham NC, Johnson J: Risk of falls after hospital discharge.

J Am Geriatr Soc 1994, 42(3):269-274,f1000m,isrctn



Ferreri S, Roth MT, Casteel C, Demby KB, Blalock SJ. Methodology of an Ongoing, Randomized Controlled Trial to Prevent Falls Through Enhanced Pharmaceutical Care. Am J Geriatr Pharmacother 2008; 6: 61-81





Update: See our December 2009 What’s New in the Patient Safety World column “December 2009 More on Ambulatory DVT Prophylaxis”. A major population study has really added some urgency to the question of DVT prophylaxis in ambulatory patients. Analysis done on the UK’s Million Women Study (Sweetland et al 2009) looked at the duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women. They found that the incidence of DVT and pulmonary embolism after surgery is substantially increased in the first 12 postoperative weeks, though the risk varies by type of surgery. Risk was highest in the first 6 weeks post-op, peaking during the 3rd week, but remained high for at least 7-12 weeks post-op. The risks for pulmonary embolism paralleled that for DVT. Major orthopedic surgery and cancer surgery had the highest risks, but the increased risks spanned almost all types of surgery. Most interestingly, even same day surgery was associated with a prolonged risk for DVT. That certainly pours water on the argument that immobility is the biggest DVT risk factor.


While this study is an eye opener, don’t run out there and start anticoagulating all your surgical patients for months at a time. We clearly need prospective randomized trials that demonstrate the benefits of prolonged DVT prophylaxis would outweigh the risks. Such studies would have to include a very large sample size, particularly since it is likely that certain subgroups would have greater or lesser risks and benfits.


Another new review (Spyropoulos et al 2009) looks at issues surrounding DVT prophylaxis in medical patients. It notes that data from large clinical trials support LMWH prophylaxis for up to 14 days in medical patients but that most patients are discharged without such prophylaxis. That paper has a good analysis of the duration of prophylaxis issue and also has great discussions on LMWH vs. unfractionated heparin, proper dosing of unfractionated heparin, mechanical DVT prophylaxis, and special populations (eg. obese or those with renal impairment).






Sweetland S, Green J, Liu B, et al. Duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women: prospective cohort study. BMJ 2009; 339: b4583 (Published 3 December 2009)



Spyropoulos AC, Mahan C. Venous Thromboembolism Prophylaxis in the Medical Patient: Controversies and Perspectives. The American Journal of Medicine 2009; 122(12): 1077-1084












Patient Safety Tip of the Week Archive


What’s New in the Patient Safety World Archive