A simple device to increase rates of compliance in maintaining

30-degree head-of-bed elevation in ventilated patients

Zev Williams, MD, PhD; Rodney Chan, MD; Edward Kelly, MD

Objective: To determine whether a highly visible device that Measurements and Main Results: A total of 268 bed measure-
clearly indicates whether the head-of-bed is adequately elevated ments were made. The average head-of-bed elevation was 21.8
would increase rates of compliance with head-of-bed elevation degrees on beds without the device (n ⴝ 166) and 30.9 degrees
guidelines. on beds with the device (n ⴝ 102; p < .005). When compliance is
Design: A prospective, single-center, multi-unit, two-phase study. defined as a bed angle of >28 degrees, 23% of beds without the
Setting: Surgical, thoracic, trauma, and medical intensive care units. device were compliant while 71.5% of the beds with the device
Patients: Cohort of intubated patients. were compliant. The relative risk and odds ratio of having the
Interventions: A 4-wk trial was performed. At the onset of the device on a compliant bed were 2.2 and 9.25, respectively (p <
trial, nurses were reminded to maintain head-of-bed elevation .005). Seventy-two percent of nurses surveyed (n ⴝ 32) found it
>30 degrees. Over the subsequent 2 wks, head-of-bed elevations to be an improvement over existing methods, 88% found it help-
of intubated patient beds were measured. An Angle Indicator, ful, and 84% would like it routinely used.
designed to clearly display whether the head-of-bed was ade- Conclusions: The Angle Indicator improved rates of adherence
quately elevated, was then placed on side rails of beds of venti- to bed-elevation guidelines, and hospital staff found it helpful.
lated patients, and head-of-bed elevation measurements were (Crit Care Med 2008; 36:1155–1157)
taken for an additional 2 wks. A survey was then handed out to KEY WORDS: ventilator-associated pneumonia; head-of-bed ele-
nursing staff to assess satisfaction with the device. vation; nosocomial infections; infection prevention

V entilator-associated pneumo- over study, Torres et al. (5) demonstrated ical ventilation for ⱖ7 days and a Glas-
nia (VAP) is the most common that the semirecumbent position de- gow Coma Scale score ⬍9 were addi-
hospital-acquired infection in creased rates of aspiration of gastric con- tional risk factors. Again, the only
the intensive care unit (ICU) tents four-fold. Kollef (6) used multivar- modifiable risk factor for the develop-
(1). It occurs in 9% to 40% of all ICU iate analysis for risk factors of developing ment of VAP was elevation of HOB.
patients and has an incidence of 5 to 35 aspiration pneumonia and found that Because of the importance of adequate
cases per 1000 ventilator days (2). The head position ⬍30 degrees in the first 24 HOB elevation in preventing VAP, the
consequences of VAP are severe: a three- hrs of intubation was an independent risk 1997 Centers for Disease Control and
fold increased duration of mechanical factor for developing VAP. The other risk Prevention (CDC) Guidelines for Prevent-
ventilation, a two- to six-fold increase in factors were organ system failure, age ing Aspiration Pneumonia and the 2003
ICU stay, a 2- to 3-day increase in hospital ⬎60 yrs, and previous antibiotic use. CDC and the Healthcare Infection Con-
stay (3). Each case of VAP increases hos- Thus, at the time of intubation, the only trol Practices Advisory Committee rec-
pital costs by $40,000 to $50,000 and modifiable risk factor for the develop- ommend elevating the HOB of a patient
results in a 15% to 45% increase in at- ment of aspiration pneumonia risk was at high risk for aspiration at an angle of
tributable mortality (2, 4). head position. In a landmark study, 30 – 45 degrees unless this is contraindi-
Elevation of the head-of-bed (HOB) of Druculovic et al. (7) performed a ran- cated (4). The Institute for Healthcare
intubated patients is an effective method domized trial assessing the frequency of Improvement Safer Systems Saving Lives
for reducing rates of aspiration pneumo- clinically suspected and microbiologically Campaign has made HOB elevation one
nia. In a randomized two-period cross- confirmed nosocomial pneumonia in of four components of the Ventilator
semirecumbent vs. supine position in 86 Bundle for preventing nosocomial infec-
intubated patients. Thirty-four percent of tions (8). Most recently, the 2006 Society
From Brigham and Women’s Hospital, Boston, MA
(ZW, RC, EK); and Massachusetts General Hospital,
patients in the supine position developed for Critical Care Medicine Outcomes Task
Boston, MA (ZW). VAP compared with only 8% of patients Force endorsed HOB elevation as a method
Dr. Williams has filed a patent application for the in the semirecumbent group. Supine to reduce aspiration pneumonia (9).
bed-angle indicator. The remaining authors have not body position (odds ratio 6.8) and enteral Despite the recommendations of lead-
disclosed any potential conflicts of interest.
For information regarding this article, E-mail:
nutrition (odds ratio, 5.7) were indepen- ing societies and agencies, strong evi-
[email protected] dent risk factors for nosocomial pneumo- dence supporting elevation of HOB as a
Copyright © 2008 by the Society of Critical Care nia. Those patients in the supine position means of preventing aspiration pneumo-
Medicine and Lippincott Williams & Wilkins and receiving enteral nutrition had the nia, and its intrinsic cost-effectiveness,
DOI: 10.1097/CCM.0b013e318168fa59 highest frequency of VAP (50%). Mechan- rates of adequate HOB elevation remain

Crit Care Med 2008 Vol. 36, No. 4 1155

low (10). In one observational study, HOB elevation ⬎30 degrees were written by elevation ⱖ28 degrees (allowing for error
mean HOB elevations were only 16 –23 the physician and transcribed by the nursing in reading the device), 38 (23%) of the
degrees with the vast majority (70% to staff. Over the subsequent 2 wks, HOB eleva- beds without the device were compliant
86%) of patients being supine (11). In tion measurements were taken on beds of all while 73 (72%) of the beds with the de-
Kollef’s (6) previously cited study, only intubated patients in the surgical, thoracic, vice were compliant (p ⬍ .005). The rate
32% of the patients who were in the su- trauma, and medical ICUs who a) did not have ratio was 3.12 (95% confidence interval,
pine position had an indication for that a specific indication or order to be kept in a 2.3– 4.2) and the odds ratio was 9.35
position. Most recently, van Nieuwenhoven less elevated position; b) were not in the pro- (95% confidence interval of 4.3–20.1).
cess of having a bedside procedure done; or c)
and colleagues (12) assessed the feasibility of There were no instances where the red/
were not in the process of being cleaned, fed,
semirecumbent position and found an aver- green indicator did not correlate with the
or groomed by the nursing staff. The existing
age HOB elevation of 22.6 degrees after 1 wk Johnson Level 700 Magnetic Angle Locator,
ICU beds had manufacturer-installed angle in-
in their study population that had a targeted dicators located along a track at the base of the
allowing for a 2-degree error rate in the
HOB elevation of 45 degrees. bed. All measurements on a bed were made by positioning of the bed-angle indicator.
The purpose of this study was to deter- one of the study investigators ⱖ18 hrs apart to Thirty-two nurses completed the
mine whether a simple, easy-to-view, and allow for change of shift. Nursing staff were anonymous survey—a 100% response
easy-to-interpret device that displayed not aware of when measurements would be rate. Seventy-two percent of nurses sur-
whether the HOB was adequately elevated made. A waiver was obtained from the institu- veyed found the device to be an improve-
could increase rates of compliance with HOB tional review board. ment over existing methods, 88% found
elevation guidelines. Additionally, we sought Over the following 2 wks, the angle indi- it to be helpful in monitoring HOB ele-
to determine whether the device was accept- cator was placed on the side rails of all eligible vation, and 84% wanted the device to be
able to nursing staff. beds. HOB elevation measurements were permanently used. While 94% of nurses
taken starting 24 hrs after placement of the surveyed were previously aware of the
METHODS angle indicator using the same inclusion and HOB elevation guidelines in the ICU,
exclusion criteria used during the control por- 66% found that the device increased
The Angle Indicator (Fig. 1) consists of a piece tion of the study. All HOB and calibration awareness of the importance of ensuring
of glossy photographic printer paper cut into a measurements were taken using the Johnson the correct elevation.
pie-slice wedge. The base of the paper has a stripe Level 700 Magnetic Angle Locator placed onto
of green and then red. A silk suture hangs from the the side rail.
apex of the wedge to its base and a steel nut is tied DISCUSSION
Following the completion of the study, the
at the distal end of the suture. The angle indicator nurses who participated were given a five- This study suggests that compliance
is placed on the side rail of the hospital bed such question survey that they completed and re- with HOB elevation guidelines can be im-
that when the bed is elevated to 30 degrees, the turned anonymously. A section for comments proved using a simple device that clearly
weight hangs at the junction between red and was also provided. Approximately 70 nurses indicates whether adequate HOB eleva-
green. When the bed is elevated ⬎30 degrees, were involved in the care of patients with the tion is achieved. In the control phase of
the weight hangs in the green zone, and when device on their beds. the study, HOB elevation was measured
the elevation is ⬍30 degrees, the weight hangs The study was independently funded at a
in the red zone. This device was designed to
on beds without the device. Despite stan-
material cost of $42.
show in the clearest and most easily seen and dardized orders to maintain HOB ⬎30
interpreted way whether the HOB was ade- RESULTS degrees, traditional HOB elevation indi-
quately elevated. cators, and regular in-service teaching to
A 4-wk trial was conducted. At the initia- A total of 268 bed-angle measure- the ICU healthcare team, the vast major-
tion of the trial, an E-mail message was sent ments were made, 166 without the device ity of beds that should have their HOB
from the president of the hospital to all ICU and 102 with the device (Table 1). The elevated to ⬎30 degrees did not. This is
nursing staff reminding them to maintain average HOB elevation without the device in agreement with previously published
HOB elevation ⬎30 degrees, where appropri- was 21.9 ⫾ 9.1 degrees, while it was 30.9 ⫾ observational studies (11).
ate. As part of the routine care within the 7.5 degrees with the device (p ⬍ .005). Several factors contribute to the low
institution, specific orders for maintaining When compliance was defined as an HOB rates of compliance. McMullin et al. (13)
showed that healthcare providers, includ-
ing attendings, residents, and nurses,
overestimate the angle of HOB elevation.
Several bed manufacturers now place in-

Table 1. Results of bed-angle measurements

Without Device With Device

Total, n 166 102

Average angle, 22 31
degree
Compliant, n 38 73

For the purposes of this study, compliance was

defined as a bed angle ⱖ28 degrees. All measure-
Figure 1. A, bed-angle indicator on side rail of intensive care unit bed. B, close-up. ments were rounded to the nearest degree.

1156 Crit Care Med 2008 Vol. 36, No. 4

dicators that attempt to show the HOB the novelty of using a new device wore off red/green indicators; and improving ac-
elevation. Unfortunately, as shown in the or it began to blend in to the ICU envi- cess to the information by making the
control phase of our study, these have ronment and be forgotten by staff. One indicator readily visible. By these changes
proven unsuccessful in promoting ade- solution for this possible problem would in design, we were able to improve clin-
quate HOB elevation (Table 1). Often the be to have the device visible only when ical care with minimal costs and excellent
angle indicators are small and are located needed. For example, when HOB eleva- overall acceptance by nursing staff. This
under the bed, making them difficult to tion to ⬎30 degrees is either unnecessary bed-angle indicator may be a simple and
read and easily forgotten. Some models or contraindicated, the device could be cost-effective method for improving com-
provide angle measurements in incre- removed or hidden using a cover. Simi- pliance with HOB guidelines and decreas-
ments as small as 5 degrees, which pro- larly, the indicator can be readily digi- ing rates of VAP.
vide more detail than is necessary and tized to allow for monitoring and record-
may be overwhelming in the ICU, which ing of compliance rates or could be fitted REFERENCES
already has an extraordinarily high rate of with an alarm to help indicate when beds
data generation. Patient beds are often are not sufficiently elevated. Further 1. Vincent JL, Bihari D, Suter PM, et al: The preva-
kept in a slight Trendelenburg position to studies over longer periods of time are lence of nosocomial infection in intensive care
units in Europe: Results of the European Preva-
prevent patients from sliding off the beds, necessary.
lence of Infection in Intensive Care (EPIC) Study.
and this makes some of the existing bed- The control and test periods were not
EPIC International Advisory Committee. JAMA
angle indicators inaccurate as they as- run concurrently as there was concern that 1995; 274:639–644
sume a level bed axis. The current meth- the presence of this device in the ICU would 2. Ibrahim EH, Tracy L, Hill C, et al: The
ods do not clearly demonstrate whether improve awareness of HOB elevation guide- occurrence of ventilator-associated pneumo-
the bed is adequately elevated, thereby lines and thus falsely improve the rates of nia in a community hospital: Risk factors and
requiring another level of information adequate HOB elevation in those beds with- clinical outcomes. Chest 2001; 120:555–561
processing for the healthcare providers. out the device. Results of the survey con- 3. Rello J, Ollendorf DA, Oster G, et al: Epide-
The results of the survey suggest that firm that the presence of the device did miology and outcomes of ventilator-associ-
nursing education was not one of the improve awareness of the HOB elevation ated pneumonia in a large US database.
Chest 2002; 122:2115–2121
contributing factors, since 94% of re- guidelines. While this improvement in
4. Tablan OC, Anderson LJ, Besser R, et al:
spondents said they were aware of the awareness would be a confounding element
Guidelines for preventing health-care-
guidelines and the importance of main- in simultaneously run control and test associated pneumonia, 2003: Recommenda-
taining adequate elevation. trial, in practice, it would be an overall tions of CDC and the Healthcare Infection
Placing the bed-angle indicator on the value within the ICU. Control Practices Advisory Committee.
side of beds increased the average HOB Since it was not technically possible to MMWR Recomm Rep 2004; 53:1–36
elevation from 21.9 to 30.9 degrees, and continuously monitor the HOB elevation, 5. Torres A, Serra-Batlles J, Jos E, et al: Pulmonary
beds with the device were almost four intermittent measurements were made. aspiration of gastric contents in patients receiving
times more likely to be compliant with It is therefore possible that sampling er- mechanical ventilation: The effect of body posi-
the hospital, CDC, and Institute for rors could have been made. To attempt to tion. Ann Intern Med 1992; 116:540–543
6. Kollef MH: Ventilator-associated pneumonia:
Healthcare Improvement guidelines. The limit this possibility, measurements in
A multivariate analysis. JAMA 1993; 270:
results of the survey suggest this was due both the control and intervention phases
1965–1970
to several factors. The bed-angle indica- of the study were made at roughly the 7. Drakulovic MB, Torres A, Bauer TT, et al:
tor described here was readily visible same times of day. In addition, all mea- Supine body position as a risk factor for
from the patient’s doorway, displayed surements were made ⱖ18 hrs apart to nosocomial pneumonia in mechanically ven-
clearly whether the bed was adequately prevent oversampling a single bed and to tilated patients: A randomised trial. Lancet
elevated, and was accurate even with the give time for nursing changes. As another 1999; 354:1851–1858
bed in Trendelenburg position. Conse- safeguard against sampling errors, bed 8. Critical Care. http://www.ihi.org/IHI/Topics/
quently, the vast majority of nurses using measurements were not made if a patient CriticalCare/. Accessed October 26, 2007
the device found it helpful and an im- was to have a procedure, test, cleaning, or 9. Curtis JR, Cook DJ, Wall RJ, et al: Intensive
care unit quality improvement: A “how-to”
provement in monitoring bed angles. feeding or if for any other reason was not
guide for the interdisciplinary team. Crit
There are limitations to our study. at his or her “baseline” position.
Care Med 2006; 34:211–218
First, the study required investigators to The study did not evaluate rates of aspira- 10. Evans D: The use of position during critical
enter the patient rooms and measure bed tion pneumonia. Nonetheless, as others have illness: Current practice and review of the
angles, introducing a potential for the shown that a semirecumbent position de- literature. Aust Crit Care 1994; 7:16 –21
Hawthorne effect or observer influence. creases rates of aspiration pneumonia, the 11. Grap MJ, Munro CL, Bryant S, et al: Predic-
To minimize this possible confounding goal of improving HOB elevation was felt to tors of backrest elevation in critical care.
effect, we started with the control phase be a clinically relevant end point. Intensive Crit Care Nurs 2003; 19:68 –74
of our study by measuring beds without Within the ICU setting, large amounts 12. van Nieuwenhoven CA, Vandenbroucke-
the device for the first 2 wks. Therefore, of patient data are constantly being gen- Grauls C, van Tiel FH, et al: Feasibility and
effects of the semirecumbent position to pre-
by the time we started the interventional erated. The bed-angle indicator described
vent ventilator-associated pneumonia: A ran-
portion of the study, the nursing staff here was an attempt to focus the data and
domized study. Crit Care Med 2006; 34:
should have become somewhat accli- facilitate its translation into improved 396 – 402
mated to the daily bed-angle measure- clinical care by eliminating unnecessary 13. McMullin JP, Cook DJ, Meade MO, et al:
ments. Nonetheless, it is possible that information, such as the specific degrees Clinical estimation of trunk position among
over a more prolonged time, compliance of elevation; minimizing the need for in- mechanically ventilated patients. Intensive
rates with the device could decrease once terpretation of the data by providing clear Care Med 2002; 28:304 –309

Crit Care Med 2008 Vol. 36, No. 4 1157