Failure to rescue is shorthand for failure to rescue (i.e., prevent a clinically important deterioration, such as death or permanent disability) from a complication of an underlying illness (e.g., cardiac arrest in a patient with acute myocardial infarction) or a complication of medical care. Failure to rescue rates used for both research purposes and as quality indicators are typically derived from hospital administrative databases. However, it is not clear how identify it, so what are the best indicators to measure it?
Rosario - that's a real tricky one. Apart from retrospective documentation (as you have already highlighted), I can't really see how else it could be done. The main issue is that of ethics. Imagine trying to do something like a clinical trial where one group received an intervention that saved them from failure to rescue and the other group did not and died as a result!! The closest thing that I could think of is something like a retrospective cohort study to measure those who didn't end up in a failure to rescue situation versus those that did - but tun how do you measure those that don't - other than as some sort of 'close call' situation
Dear Dean, I agree with you, but I think if we could be able to "intercept" the main predictors and/or antecedents of "failure to rescue" maybe we could study the phenomenon with study designs different from retrospective (eg longitudinal), focusing on them and not on the event (failure to rescue). I believe that another issue, as well as ethics, concerns the lack of clarity about the theoretical construct of "failure to rescue": the conseguences of "failure to rescue" are often obvious, but are we sure that we know the antecedents of this phenomenon and how they could predict a faiulure to rescue?
Glad we are in agreement Rosario. Your point about a lack of conceptual/contextual clarity immediately draws me to two qualitative approaches. Grounded Theory would be the most robust way of seeking theoretical clarity and perhaps provide a framework - while a concept analysis could at least clarify the concept itself. From a mixed methods perspective, a Delphi technique of experts in the field might do similar. Of course, these approaches would not measure the extent of the problem but, instead, allow a clearer theoretical/philosophical platform to move forward.
Thanks Dean, I think focusing on clarify theory about the "failure to rescue" is crucial to best understand that phenomenon, and as you said that could give us a clearer basis to go deepen in future research endeavor. Maybe we need a reverse gear at the beginning aimed to provide more theoretical clarity, before trying to "measure".
These can only be identified retrospectively. That leaves any research in this topic demographic, descriptive, or at best root cause analysis. As far as identification, claims can be one screen, but a good incident reporting environment, near misses can be identified with future events being prevented.
I also think that retrospective analysis is the only option. Is it accurate to limit failure to rescue to cardiac arrest? Perhaps one of the early warning system frameworks (EWS; http://www.ihi.org/resources/Pages/ImprovementStories/EarlyWarningSystemsScorecardsThatSaveLives.aspx) could be used to expand the conceptualization of "failure to rescue" . It seems that failure to respond, or a delay in response, at any point when the evidence (i.e., vital signs, O2 sat) suggests an intervention could be on a continuum of failure to rescue.
Amelia (Amy) Joseph did her dissertation at the University of South Carolina on this topic. If you go to http://scholarcommons.sc.edu/etd/2487/ you can download her open access dissertation.
She's still at the University of South Carolina's College of Nursing doing a post-doc in Health Systems redesign.
I have developed an instrument on patient perspectives of Nurse Vigilance. I have often thought that using this - and then comparing scores to incidents would be interesting. If there is a correlation, it would be a good indicator that patients are aware of nurse behaviors that indicate safe care or error potential. I am currently working on a publication - but here is the link to the file that an honors research and I worked on-
http://digitalcommons.iwu.edu/cgi/viewcontent.cgi?article=1045&context=nursing_honproj
I'd refer you back to the literature from 2007 and following on Rapid response teams.Most organizations have developed criteria for calling their internal rapid response team, with various professionals responding based on their own institutional policy. The IHI might be a source for you. In my past, we had defined clinical criteria that mandated a call for the response team, and when staff failed to execute that call, we usually found three prominent failures: lack of critical thinking and putting all the empirical data together (usually a novice nurse), pressure of time to closely monitor a patient due to staffing load, and pressure from physicians to NOT call the rapid response team because they felt they had the situation under control. (Often, not.) To mediate these factors, there were three effective strategies; if a nurse felt she would be criticized by peer or physician for not "better managing", we talked about horizontal violence, the power of teams, coping with power differentiation, etc. We correlated the presence of rapid response calls with the absence of "codes outside the ICU", which helped staff feel more secure with their early interventions, and we held physicians responsible for supporting, rather than undermining, institutional policy.
Another useful correlation is the number, and/or successful outcomes, of "codes outside the ICU" compared from daytime to nighttime. Some institutions find that their resources are diminished at night and the nurses cannot "rescue" everyone. Night staff who had access to response teams reported feeling more confident in their assessment skills, and more likely to call physicians with critical findings.
- Badges
- Science topic
- Similar topics
- Psychology
- Social Psychology