Best Evidence Topics


Alan E. Jones, Vasilios Yiannibas, Charles Johnson, Jeffrey A. Kline
Emergency Department Hypotension Predicts Sudden Unexpected In-hospital Mortality
October, 2006; 941-948
  • Submitted by:Michael Peacock - Medical Student
  • Institution:Michigan State University College of Human Medicine
  • Date submitted:21st August 2009
Before CA, i rated this paper: 6/10
1 Objectives and hypotheses
1.1 Are the objectives of the study clearly stated?
  Yes. The objectives are: What is the prevalence of non traumatic hypotension in the Emergency Department and whether or not this can be a prognostic indicator of in-hospital mortality.
2 Design
2.1 Is the study design suitable for the objectives?
  Yes. This was a prospective cohort analysis. The sample was chosen from the Emergency Department at Carolinas Medical Center which sees greater than 100,000 ED visits per year. There were 182 random 24 hour days during 2004-2005 that were used in this study. Hospital charts were reviewed which placed patients in the exposures and non exposures group.
2.2 Who / what was studied?
  All non traumatic patients older than 17 years who had any episode of hypotension (systolic blood pressure less than 100 mm Hg), while in the Emergency Department, and were admitted to the hospital on a given random 24 hour block during the time span of the study. Those with any episode of hypotension with admission to the hospital were placed in the exposure group while all others admitted to the hospital from the ED were placed in the non exposures group.
2.3 Was a control group used if appropriate?
  Yes. There was a group which was termed "non exposures" which were used to compare in-hospital mortality and sudden unexpected death to the exposures group.
2.4 Were outcomes defined at the start of the study?
  Yes. Primary outcome measures were in-hospital mortality. Secondary outcomes included incidence of sudden and unexpected deaths between the groups, the relationship between he decile of the lowest ED systolic blood pressure and in-hospital mortality, the relationship between duration of hypotension and in-hospital mortality in the exposure group, and the sensitivity and specificity of ED hypotension for in-hospital mortality.
2.5 Was this the right sample to answer the objectives?
  Yes. Patients younger than 17 are more likely to have episodes of hypotension by definition at baseline. Trauma patients are also more likely to experience hypotensive episodes due to the acuity of their injuries.
2.6 Is the study large enough to achieve its objectives? Have sample size estimates been performed?
  Yes. They calculated a sample size based on preliminary data suggestive 15% in the exposures group and 8% in the non exposures group would suffer in-hospital mortality. They planned an enrollment of 325 for the exposures group and 3900 for the nonexposures group to get an alpha = 0.05 and a beta = 0.20.
2.7 Were all subjects accounted for?
2.8 Were all appropriate outcomes considered?
2.9 Has ethical approval been obtained if appropriate?
  The study was approved bt the IRB and Privacy Board at Carolinas Healthcare System prior to the study.
3 Measurement and observation
3.1 Is it clear what was measured, how it was measured and what the outcomes were?
  Yes the outcomes as stated earlier are clearly addressed. Sudden and unexpected death was measured by agreement of two physician observers independent of each other's decision using the defined criteria: age<85 years absence of a defined end stage disease that was to be treated with comfort care only, absence of hospice care, and absence of a do not resuscitate order. ED hypotension was broken into three groups. 1)Transient, which was only one documented systolic blood pressure less than 100 mm Hg; 2)Sustained, which was systolic blood pressure less than 100 mm Hg for greater than or equal to 60 consecutive minutes; 3)Episodic, which was ED hypotesion not categorized as transient or sustained. Continuous data were analyzed using unpaired t-tests, categorical data were reported as proportions, relative risks were calculated, and Kaplan-Meier survival estimates and log rank test were used for time to primary outcome analysis. p<0.05 was considered significant
3.2 Was the assessment of outcomes blinded?
3.3 Was follow up sufficiently long and complete?
  Yes. Follow up was the extent of the hospital stay given the fact that there was not in hospital mortality.
3.4 Are the measurements valid?
  The study never defines methods of blood pressure measurement. There is a high amount of user variability with manual blood pressure technique as well as automatic blood pressure machine measurement. There are a lot of factors that could have falsely elevated or lowered the blood pressure such as cuff size, patient movement, patient size, blood pressure cuff location, and medications. The amount of time spent in the ED was also never addressed. Patients could spend minutes to several hours in the ED and have their vital signs checked as frequent as every 5-10 minutes. The longer the stay and more frequent the vital signs are checked could falsely add patients to the hypotensive group given they only need one documented episode. It is difficult to determine validity without these considerations.
3.5 Are the measurements reliable?
  Until all of these issues are addressed I dont think the use of one episode of hypotension can be an reliable predictor of mortality. I think it is already known that sustained hypotension usually correlates with worse outcomes.
3.6 Are the measurements reproducible?
  I think they would be reproducible in a similar ED setting. I dont think they would be as generalizable to a smaller community ED where the acuity level may be less. The data measured, systolic blood pressure, is one that is simple and common practice. Therefore, it would not be difficult to reproduce the methods of the study but the results may vary based on location of the ED.
4 Presentation of results
4.1 Are the basic data adequately described?
4.2 Are the results presented clearly, objectively and in sufficient detail to enable readers to make their own judgement?
4.3 How large are the effects within a specified time?
  -In-hospital mortality: 8.3% exposures vs 2.8% non-exposures (95% CI 4-8%)
-Sudden and unexpected death: 0.5%(95% CI 0.3-0.7%) 71% in exposures group.
-Lowest systolic blood pressure <80 mm Hg: 18% in hospital-mortality with lowest systolic blood pressure <80 mm Hg and 3% in-hospital mortality for all other systolic blood pressures between groups (95% CI 9-25%).
-Sustained vs transient hypotension: 14% in-hospital mortality in sustained group (95% CI 9-19%)and 5% in-hospital mortality in transient group (95% CI 3.3-8.1%).
-Odds ratio in-hospital mortality between sustained, episodic, and transient hypotension: 5.4(95% CI 3.4-8.5), 3.2(95% CI 1.9-5.1), 2.0 (95% CI 1.2-3.2).

19% non traumatic adults had hypotension in ED. Those exposed to hypotension had a 3-fold increased risk of in-hospital mortality and a 10-fold increased risk of sudden and unexpected in hospital death.
4.4 Are the results internally consistent, i.e. do the numbers add up properly?
5 Analysis
5.1 Are the data suitable for analysis?
  Yes. The data was properly collected and the study design allowed for appropriate power.
5.2 Are the methods appropriate to the data?
5.3 Are any statistics correctly performed and interpreted?
6 Discussion
6.1 Are the results discussed in relation to existing knowledge on the subject and study objectives?
  Yes. On page 945, there is discussion about hypotension being a poor prognostic indicator for certain known diagnoses and also hypotension measured outside the hospital being a predictor of in-hospital mortality.
6.2 Is the discussion biased?
  The discussion recognizes its limitations that it does not talk about intervention to correct the hypotension, lack of physiologic measurements to establish whether or not the hypotension is causing circulatory insufficiency and possible lack of generalizability. I don't believe they provide a biased discussion. They attempt to catch all nontraumatic patients with hypotension to determine a relationship with in-hospital mortality rather than looking at previously known diagnoses that have poor outcomes with hypotension or lead to hypotension with poor outcomes.
7 Interpretation
7.1 Are the author's conclusions justified by the data?
7.2 What level of evidence has this paper presented? (using CEBM levels)
  1b: Differential diagnosis / symptom prevalence study; Prospective cohort study with good follow-up
7.3 Does this paper help me to answer my problem?
  I think this paper tells us what we already know. It may have limited usefulness in making people more aware of sustained hypotension and increased risk of mortality. However it seems less convincing with the transient episodes due to the factors discussed earlier that were not included within the study. Also, there are no current recommendations for intervention based on this study. Until there is a study that shows an intervention that decreases in-hospital mortality for those who have hypotensive episodes in the ED, the clinical usefullness is minimal.
After CA, i rated this paper: 5/10
8 Implementation
8.1 Can any necessary change be implemented in practice?
  No. Hypotension is already recongnized as a possible indicator for disease.
8.2 What aids to implementation exist?
  Ease, common practice, and low cost of blood pressure measurement
8.3 What barriers to implementation exist?
  Incorrect blood pressure measurement technique or human measurement error
8.4 Are the study patients similar to your own?
8.5 Does the paper give any conclusions that will affect what you will offer or tell your patient?