March 4, 2008
Housestaff Awareness of Risks for “Hazards of Hospitalization”
A recent article (Fernandez et al 2008) found that housestaff often had poor awareness of a number of risk factors their elderly inpatients had for the “hazards of hospitalization”. They developed a 24-question Hazards of Hospitalization Questionnaire and had housestaff answer these questions on a sample of their elderly hospitalized patients. Answers were compared to those of a study investigator on the same patients. Most of the patients had multiple risk factors. While there was good agreement on some of the questions, there was poor agreement or frank disagreement on multiple other risks such as patient orientation to place and duration of hospitalization, patient quality of sleep, presence of pain, history of falls, mood, quantity of food intake, use of assist devices (eg. glasses, hearing aids, canes) at home, and name of the primary care physician.
The authors point out that the areas of agreement tended to be related to things for which physicians had to write orders. They speculate as to why they tended to be unaware of the other risks and provide good discussion about both the medical education system and the need for good communication among the interdisciplinary care team.
We look at their results from a slightly different perspective. They likely would find the same level of unawareness for any physician at any level of training in any hospital in which they administered the questionnaire. For years, those of us who are neurologists have jokingly said that the “Neuro – WNL” that appears in so many H&P’s stands for “we never looked” so it comes as no surprise that some of these risk factors were not recognized. But rather than focusing so much on why they might be unaware, the key finding here is that they identified things that physicians are not good at (i.e. they are not good at identifying certain types of risk factors). The focus then should be to develop systems that can capture these risk factors and then make the physician aware of their presence.
There are obviously several ways to accomplish this. Nursing staff already perform many risk assessments on admission (eg. fall risk assessment, decubitus risk assessment, etc.). DVT risk assessments are also being done more frequently, either by the physicians or nursing staff. We’ve pointed out in several of our Tips of the Week that such assessments need to be updated when changes take place during the hospital stay. We have also found that other members of the care team are much more likely to identify many of the risk factors noted in this paper. The use of nurse case managers on several of our academic teaching services provided much improved continuity of care, particularly after the implementation of restricted workhour rules for housestaff. Multiple other nonphysician providers that interact with patients (physical or occupational therapists, respiratory therapists, pharmacists, etc.) could also identify some of these risk factors.
Trigger tools have received attention recently (see our October 30, 2007 Patient Safety Tip of the Week “Using IHI's Global Trigger Tool”) and many of the questions in the Hazards of Hospitalization Questionnaire could be incorporated into a trigger tool. Trigger tools can be manual or computerized. Unfortunately, most of the risk factors here that were not identified are also not readily identified independently via computerized systems.
If the Hazards of Hospitalization Questionnaire tool developed by the authors can be validated in several settings or populations, it has tremendous potential to help us prevent complications commonly seen during and after hospitalization.
Issues such as risk for delirium, that could be identified early by the Hazards of Hospitalization Questionnaire tool, are extremely important. Not only is delirium associated with increased morbidity and mortality, but it is also associated with prolonged lengths of stay and excess costs (Leslie et al. 2008). At least 2 studies have demonstrated that multifactorial interventions targeted at elderly inpatients at risk for delirium may shorten hospital length of stay, reduce duration of delirium, and reduce mortality (Lundstrom et al. 2005; Naughton et al 2005).
We really like this tool and feel it has lots of potential. It should be tested and validated in other settings and organizations should look for ways to best identify the risk factors and integrate them into a system of alerts and reminders that help focus attention early on preventable events.
Fernandez HM, Callahan KE, Likourezos A, Leipzig RM. House Staff Member Awareness of Older Inpatients' Risks for Hazards of Hospitalization. Arch Intern Med. 2008;168(4):390-396 http://archinte.ama-assn.org/cgi/content/abstract/168/4/390
Leslie DL, Marcantonio ER, Zhang Y, Leo-Summers L, Inouye SK. One-Year Health Care Costs Associated With Delirium in the Elderly Population. Arch Intern Med 2008; 168(1): 27-32. http://archinte.ama-assn.org/cgi/content/abstract/168/1/27
Naughton BJ, Saltzman S, Ramadan F, Chadha N, Priore R, Mylotte JM. A multifactorial intervention to reduce prevalence of delirium and shorten hospital length of stay. J Am Geriatr Soc. 2005; 53(1):18–23
Lundström M, Edlund A, Karlsson S, Brännström B, Bucht G, Gustafson Y. A multifactorial intervention program reduces the duration of delirium, length of hospitalization, and mortality in delirious patients. J Am Geriatr Soc. 2005; 53(4): 622–628