Reliability of Intrabladder Pressure Measurement in Intensive Care

A prospective, nonexperimental, 2-observer repeated measures (between and within) study design was used. With patients supine, a 30° head-of-bed elevation, and instillation of 0 and 25 mL of physiological (normal) saline, IBP was measured in 10 patients, twice by one nurse to assess intraobserver reliability and once by a different nurse to assess interobserver reliability. The 25-mL volume was chosen to represent the recommendation of the World Society of Abdominal Compartment Syndrome (WSACS).¹³ The 0-mL volume was chosen because in studies¹⁴ in humans, this volume was associated with a lower measurement bias and the lowest risk of raising urinary bladder pressure independently of IAP. In addition, not instilling saline would be most convenient for nurses.

The 2 sets of IBP measurements obtained by the first nurse were used to determine intraobserver reliability. The second set obtained by the first nurse were compared with the set of measurements obtained by the second nurse to determine interobserver reliability. Reliability was defined as consistency, stability, and repeatability of results¹⁵ as well as the ability to reproduce and record the same or similar IBP measurements at 2 different times by one or more observers (within a 30-minute period) when a standardized protocol was used. The study was approved by the institutional review boards of the Allegheny Singer Research Institute and Duquesne University, Pittsburgh, Pennsylvania, and was carried out in accordance with the ethical standards set forth in the Helsinki Declaration of 1975. Each patient or a designated surrogate gave written informed consent.

The study was performed in a 700-bed tertiary care hospital designated as a level I trauma center that has 105 adult ICU beds segregated as either surgical units (trauma, neurological, cardiothoracic, and surgical) or medical units (neurological, medical, and cardiac). Collectively these units are staffed by more than 325 ICU nurses; 10% of the nurses have national specialty certifications.

Adult patients admitted to one of the ICUs were eligible to participate in the study if they had a clinical need for a urinary bladder drainage catheter as determined by the attending physician. Patients were excluded from the study if they could not assume a supine position with the head of the bed elevated 30° or if they had a neurogenic bladder, bladder tumor or perforation, or hematuria or anuria, all of which can potentially alter IBP and influence interpretation of IBP measurements. Participation in the study was required only once during the ICU stay.

IBP was measured according to a standardized protocol (Table 1). The method was initially described by Kron et al³ and was modified by Cheatham and Safcsak¹⁶ and Malbrain.¹⁷ The only difference between the modified method of Kron et al and the protocol used in this study was use of the AbViser IAP monitoring kit (Model ABV100, Wolfe-Tory Medical, Inc, Salt Lake City, Utah). Physiological saline was used as the instillation fluid and was kept at room temperature during the study. During the measurements, the urinary catheter (Foley, CR Bard Inc, Gainsville, Virginia) lay on the ICU bed between the patient’s legs with the drainage collection tubing flat until the tubing was beyond the patient’s feet; the drainage collection chamber was suspended from the bed frame. The IBP was recorded from the bedside monitors (HP Model 66, Hewlett Packard, Royal Phillip Electronics, Eindhoven, the Netherlands) at end expiration after 2 full screen sweeps of a stable waveform were observed.

A laser level (Model 9-00085887, Porter-Cable Robotoolz; Toolz, Mountain View, California) was used to level the pressure transducer at the symphysis pubis.¹⁸ The symphysis pubis was chosen as the reference point because it was the reference point recommended by the American Association of Critical-Care Nurses (AACN)⁴ and the WSACS¹⁷ at the time the study was designed, and because it was the reference point used in previous assessments of the reliability of IBP measurements.¹² The only difference between the first set of measurements and the second and third sets was that the AbViser was not removed after the first set. The device was kept in place to maintain a closed system and to reduce the risk of infection. Between the first and second set of measurements, the first nurse releveled and rezeroed the pressure transducer and verified the patient’s position. Before the third set of measurements, the second nurse confirmed the patient’s position, releveled the transducer at the symphysis pubis, performed a manual square wave test to ensure the dynamic properties of the system were maintained and that conduction of pressure from the catheter to the monitor occurred, and then rezeroed the transducer.

Data were analyzed by using SPSS statistical software (SPSS-15, SPSS Inc, Chicago, Illinois), and figures were created by using Prism 4 (Version 4.03, GraphPad Software Inc, San Diego, California). Continuous data were expressed as means and standard deviations and were analyzed by using paired t tests, Pearson correlation, and the Bland-Altman method for comparison with the limits-of-agreement approach.¹⁹ The Bland-Altman plot graphs the difference between 2 measurements as a function of the mean of 2 measurements for each participant, which is considered to be the best estimate of the true value of the measurement. Bias is the difference between one measurement and the other. If the first measurement is sometimes higher and the second measurement is sometimes higher, then the mean of the difference should be close to zero. The closer the bias is to 0, the lower the variability and the higher the reliability. If the bias is not close to 0, the 2 measurements are not similar. Sample size was calculated by using PASS (Number Cruncher Statistical Software, Kaysville, Utah, published April 23, 2007). A sample size of 10 had a 97% power to detect a difference of 0.8 between the null hypothesis correlation of 0.9 and the alternative hypothesis correlation of 0.1 with a 2-sided hypothesis test with a significance level of 0.05.

A total of 6 men and 4 women took part in the study (Table 2). Of these 10 patients, 5 were admitted to the ICU for medical reasons and 5 for postoperative care; 3 of the 5 postoperative patients had had abdominal surgery. No patients were paralyzed or receiving mechanical ventilation, but 2 were treated with positive airway pressure during the study. Of the 10 patients, 1 was receiving enteral nutrition, 4 were not permitted anything by mouth, 1 was permitted ice chips, and 4 were permitted regular oral intake.

When no physiological saline was instilled into the bladder, the first IBP measurement ranged from −5 to 12 mm Hg. Of the 10 patients, 8 had positive values, 1 had a value of 0 mm Hg, and 1 had a value of −5 mm Hg. The second IBP measurement ranged from −4 to 14 mm Hg; compared with the first measurement, the second was higher for 6 patients and lower for 4 (Figure 1). The patient with the negative value for the first measurement also had a negative value for the second measurement. The means of the first and second measurements were 6.1 (SD, 5.9) and 7.9 (SD, 6.1) mm Hg, respectively. Pearson correlation of the 10 pairs of measurements was 0.74 (P = .007). The mean of the difference was −1.8 (95% confidence interval, −4.9 to 1.3; P = .22).

When 25 mL of saline was instilled, the first measurements ranged from 7 to 16 mm Hg and the second measurements ranged from 10 to 18 mm Hg. The second measurements were higher than the first in 5 patients, lower in 3 patients, and the same in 2 (Figure 1). The means of the first and second measurements were 12.1 (SD, 2.7) and 12.7 (SD, 2.7) mm Hg, respectively. Pearson correlation between the 10 pairs of measurements was 0.81 (P = .002). The mean of the differences was −0.6 (95% confidence interval, −1.8 to 0.6; P = .28).

When the IBP measurements were obtained with no instillation of saline, the bias of the estimated measurements was −1.8 mm Hg (Figure 2). The limits of agreement or the expected difference between measurements of future samples was between −10.3 and 6.7 mm Hg. When the IBP measurements were obtained after instillation of 25 mL of saline, the bias of the measurements was −0.6 mm Hg, and the limits of agreement were between −3.82 and 2.62 mm Hg; both values (bias and limits of agreement) are lower than those for the set of measurements obtained with no saline.

When no saline was instilled into the bladder, the second nurse obtained IBP measurements that were the same in 2 patients, lower in 4 patients, and higher in 4 patients than the values obtained by the first nurse (Figure 3). The patient with the lowest IBP value obtained by the first nurse also had the lowest value of the 10 patients when IBP was measured by the second nurse. IBP values ranged from −4 to 14 mm Hg for the first nurse and from −3 to 16 mm Hg for the second nurse. The means of the IBP measurements were 7.9 (SD, 6.1) for the first nurse and 6.9 (SD, 7.1) mm Hg for the second nurse. Pearson correlation between the 10 pairs was 0.78 (P = .004). The mean of the differences was 1.0 (95% confidence interval, −2.2 to 4.2; P = .49).

When 25 mL of saline was instilled, compared with the first nurse, the second nurse obtained IBP measurements that were higher in 5 patients, lower in 4 patients, and the same in 1 patient. The measurements ranged from 10 to 18 mm Hg for the first nurse, and 8 to 20 mm Hg for the second nurse. The means were 12.7 (SD, 2.7) mm Hg for the first nurse and 13.4 (SD, 4.1) mm Hg, for the second nurse. The Pearson correlation between the 10 pairs was 0.82 (P = .002). The mean of the differences was −0.7 (95% confidence interval, −2.45 to 1.05; P = .39).

When IBP was measured without instillation of saline, the bias of the measurement was 1.0 and the limits of agreement were between −7.76 and 9.76 mm Hg. When IBP was measured after instillation of with 25 mL of saline, the bias of the measurement was −0.7 mm Hg and the limits of agreement were between −5.5 and 4.0 mm Hg (Figure 4).

Our findings indicate that both intraobserver and interobserver reliabilities of measurements of intrabladder pressure are strong when IBP is measured with patients supine, the head of the bed elevated 30°, and an instillation volume of either 0 or 25 mL is used. However, compared with no instillation of saline, instillation of 25 mL reduced the variability of repeated measurements and increased the reliability.

Our results agree with those of 1 experimental²⁰ and 1 clinical¹³ study on the reliability of IBP measurements. Wolfe and Kimball²⁰ built a laboratory model of the abdomen by using a 210-L container with a urinary catheter exiting from the base of the container. The proximal end of the urinary catheter tip was sealed with a 100-mL bag and was placed at the base of the container, simulating a urinary bladder, and the distal end was connected to the AbViser system and a pressure transducer, which was connected to the monitor. The pressure transducer was leveled and zeroed at the level of the simulated bladder at the base of the container. A column of fluid was placed within the container to simulate IAPs of 5, 10, 15, 20, 25, 30 and 40 mm Hg. A total of 11 observers each obtained 5 measurements for each of the 7 simulated IAPs. Wolfe and Kimball found little interobserver variability of the measurements and the standard deviations of the measurements were small (0.1–0.5). In that study,²⁰ Wolfe and Kimball assessed variability of the monitoring system but did not address more important questions such as variability introduced by patients or by the technique of the ICU nurses.

The intraobserver and interobserver reliabilities of IBP measurements obtained after instillation of 25 mL of saline in our study agree with those in the study by Kimball et al,¹² the only other study in which the reliability of IBP measurements was assessed by using a technique and clinical setting similar to the ones we used. Both studies indicated no difference between the mean IBP values and a high Pearson correlation for intraobserver and interobserver reliabilities. In the study by Kimball et al,¹² the Pearson correlation for paired samples for IBP measurements obtained after instillation of 50 mL of saline was slightly higher than in our study for measurements obtained after instillation of 25 mL (0.93 vs 0.81 for intraobserver reliability and 0.95 vs 0.82 for interobserver reliability, respectively), and their mean difference was lower (0.57 vs −0.6 mm Hg for intraobserver reliability and 0.00 vs −0.7 mm Hg for interobserver reliability, respectively). Nonetheless, both Kimball et al and we found low variability and high intraobserver and interobserver reliabilities.

The slightly higher intraobserver and interobserver correlations in the study by Kimball et al¹² might be due to 3 important differences in the experimental design. First, Kimball et al measured IBP with patients supine, the head of the bed unelevated, and instillation of 50 mL of saline. We obtained IBP measurements with patients supine, the head of the bed elevated 30°, and instillation of 0 and 25 mL of saline. These differences are important because a marked position-volume interaction can affect measurements of IBP. For instance, in a larger study,²¹ when participants were positioned supine, IBP measured after instillation of 50 mL of saline was higher than when 25 mL was instilled, but when the participants were positioned supine with the head of the bed elevated 30°, the opposite occurred. IBP measured after instillation of 50 mL was lower than the IBP measured after instillation of 25 mL.

Second, the technique of establishing the reference point for zeroing and leveling the pressure transducer differed. In the study by Kimball et al,¹² the transducer remained fixed at the symphysis pubis and was not releveled or rezeroed between IBP measurements. In our study, the pressure transducer was releveled and rezeroed, and the entire monitoring system was rechecked and a square wave test was performed between IBP measurements obtained by the first nurse and those obtained by the second nurse. Thus, the paired-sample data obtained in our study not only reflect the variability due to the administration of saline, reading the monitor, and the condition of the participant but also include the variability due to differences in the process used to measure IBP, such as transducer positioning, leveling, and square wave testing between ICU nurses. Last, Kimball et al¹² had a larger study population (18 patients) than we did (10 patients) and a larger sample size (18 vs 10 paired samples).

However, in the study by Kimball et al,¹² from 1 to 39 paired samples per patient were used for analysis, and some of the paired samples were excluded because they exceeded the collection time. Having few subjects contribute a large number of the sample data points and omitting data from the analysis will bias the data toward lower variability and higher correlation. In contrast, we had 10 patients, with 3 pairs of measurements per patient, and no pairs were omitted from the analysis.

The reliability of an IBP measurement is a function of the several factors that may affect the measurement, such as the accuracy of the equipment, the technique of the nurse or observer, and other patient-related clinical factors. Previous investigators^5,22 have shown that when properly calibrated, bedside monitoring equipment, as used in our study, is sensitive and produces a reliable transduction of hydrostatic pressures. Some of the most pertinent patient-related factors are body weight,²³ use of sedatives,²⁴ breathing and ventilatory status with positive airway pressure,²⁵ net fluid balance,²⁶ and body position.²⁷ The patients in our study had a mean body mass index of 28.2 and a mean length of stay of 3 days and were not sedated or treated with mechanical ventilation, characteristics that may partly explain the low variability of our results.

Factors related to a nurse’s technique that can contribute to variability in IBP measurements include the accuracy in positioning the patient, proper assembly of the equipment (ie, ensuring no air bubbles are present in the tubing, placing and leveling the transducer properly, and zeroing the pressure transducer), proper identification of the waveform, and reading the monitor at end expiration.⁵ Because ICU patients are usually extremely ill and are receiving multiple therapies, patient-related factors are difficult to standardize and would be expected to contribute markedly to the variability of paired IBP measurements.

Among the potential factors contributing to variability in IBP measurements, the technique nurses use is an area of focus to limit variability. Our findings indicate that variability in IBP measurements is low when the nurses’ technique is standardized. We standardized the technique by using an ICU bed that allowed accurate elevation of the head of the bed to 30°, using a laser level to level the pressure transducer at the symphysis pubis before zeroing for all measurements, using the AbViser system instead of stopcocks or needles inserted into the sampling port of the urinary drainage system to reduce the risk of infections and technical variability, and keeping the urinary drainage collection tubing on the bed until the tubing was beyond the patient’s feet and then suspending the tubing from the bed frame.

WSACS¹ and AACN⁴ recommend that IBP be measured after instillation of 25 mL and 50 mL, respectively of physiological saline. Recent studies^1,28,29 have indicated that in supine patients increases in the volume of saline instilled result in progressive increases in IBP. The main explanation²⁹ for these findings is that a high volume has a higher probability than a lower volume of activating the bladder detrusor muscle and of affecting bladder compliance. Because of the potential for activation of the detrusor muscle, several investigators have advocated the use of a minimal volume to ensure conduction of the fluid wave.³⁰ Our findings indicate that reproducible IBP measurements can be obtained without instilling any sterile saline into the bladder. However, most patients in the ICU are receiving continuous intravenous fluids, and most likely small amounts of urine would be in the bladder at the time of measurement. In addition, the bladder probably would not be completely empty even when drained.

During our study, the presence of urine became apparent when the drainage collection tubing was clamped for the first IBP measurement and urine filled the AbViser, the drainage collection tubing, and the catheter within seconds of clamping and immediately before the time of measurement of IBP.

Compared with instillation of 25 mL of saline, instillation of no saline resulted in greater variability in IBP measurements and occasional subatmospheric (negative) values, and mean IBP was 6 mm Hg lower (Figures 1 and 3). Subatmospheric IAPs have been measured in previous studies³¹ in animals and cadavers but not in clinical studies,^{12,29,30,32,33} perhaps because volumes from 10 to 250 mL were instilled into the bladder. Our results agree with those of DeWaele et al,³⁰ who reported that up to 76% of the time a fluid wave could be detected when no saline was instilled and that 2 mL was adequate to ensure fluid wave transmission for the remaining cases. On the basis of our results and the results of others,³⁰ measuring IBP without instillation of saline should be avoided.

In our study, the saline instilled into the bladder was at room temperature (22°C). Recently, Chiumello et al³⁴ found that instilling normal saline into the bladder at room temperature results in slightly higher values than does instilling the same amount of normal saline at body temperature (37°C). This effect was more pronounced with high volumes. A possible explanation³⁴ is that room-temperature saline injected into the bladder can activate the detrusor muscle. The validity of the explanation is questionable; the bladder cooling reflex occurs only in infants and young children and in patients with abnormalities in the spinal upper motor neurons, and no detrusor response to cold temperature occurs in patients with abnormalities in lower motor neurons or in patients with no neurological abnormalities.³⁵ On the basis of these results, use of room-temperature saline probably would not alter interpretation of our results.

Our findings support the WSACS recommendation for instilling 25 mL of normal saline into the bladder before measuring IBP because compared with no instillation, this amount reduces variability and improves reliability. We did not investigate the reliability of IBP measurements with volumes larger or smaller than 25 mL, and we cannot comment on whether higher reliability could be obtained with a higher or lower volume. For instance, Kimball et al¹² found higher interobserver and intraobserver rater reliability of IBP measurements when 50 mL of physiological saline was used.

AACN⁴ and WSACS¹ recommend that IBP be measured with patients in the supine position only, but a rationale for this recommendation has not been provided. Studies^21,36,37 have shown that IBP measured with patients in the supine position with no elevation of the head of the bed is lower than IBP measured with patients supine and any degree of elevation. This finding is to be expected, because any elevation of the head of the bed increases abdominal wall tension and increases the gravitational effect of intra-abdominal organs on IBP.²⁷ However, positioning patients supine even for a short period just for the purpose of measuring IBP increases the risk of aspiration pneumonia and is against recent practice guidelines³⁸ that recommend ICU patients have the head of the bed elevated a minimum of 30° at all times. Our results indicate that experienced ICU nurses can obtain highly reproducible IBP measurements by following a standardized protocol with patients supine and the head of the bed elevated 30°. This finding is important for clinical practice because most patients are positioned supine with the head of the bed elevated 30° most of the time in the ICU, and IBP is measured more than once per day and by more than one nurse in a day. Maintaining patients supine with the head of the bed elevated 30° should reduce the risk of equipment dislodgement and decrease the risk of aspiration pneumonia for patients receiving mechanical ventilation and enteral nutrition.³⁸ Furthermore, measuring IBP with patients in this position is convenient for them, especially those who have difficulties achieving the supine position, and also reduces ICU nurses’ workload. Because of these practical advantages, the supine position with the head of the bed elevated 30° should be considered as the recommended position for measuring IBPs.

Our study has several limitations. First, only intraobserver and interobserver reliabilities of IBP measurements in patients who were positioned supine with the head of the bed elevated 30° were assessed, and IBP was only measured with instillation of 0 and 25 mL of normal saline. Studying other positions and smaller and larger instillation volumes would have strengthened the study. Second, only 2 experienced ICU nurses collected data, and we do not know if similar results would be obtained by less experienced or novice nurses. Third, we did not address other potential sources of variability such as the ability of the ICU nurse to correctly identify the anatomic location of the reference point used to level and zero the transducer²² or the ability of the ICU nurse to interpret the pressure wave from the monitor.³⁹ 

Last, we measured IBP in a small and relatively homogeneous number of patients. Despite efforts to recruit more critically ill patients into the study, most of the patients (93%) who participated were not sedated or receiving mechanical ventilation. Patients who were sedated or receiving mechanical ventilation could not give informed consent, and their surrogate decision makers did not give informed consent to participate in this study. For this study, 260 patients were approached, and only 120 gave informed consent. In addition, in some instances, the attending physician and/or the bedside nurse did not refer a patient because of an unstable clinical status or other reasons. Other critical care studies have this same limitation.⁴⁰ 

Measurement of IAP is the new frontier in critical care monitoring. Before this technique becomes more widely used and becomes a standard of care to guide the medical management of patients, nurses should critically assess each aspect of IAP measurement and application. Failure to do so will risk committing the same mistakes that have occurred with hemodynamic measurements.^41,42 

Because abdominal compartment syndrome is defined on the basis of measurements of IBP obtained with patients supine, and if the supine position with the head of the bed elevated 30° is to be used instead, further studies are needed in more heterogeneous ICU populations to identify the levels of IBP in patients supine with the head of the bed elevated 30° that indicate abdominal compartment syndrome and need for treatment. In addition, further studies are needed to delineate additional key elements of IBP measurement, such as the proper anatomic reference for the positioning and leveling of the transducer, the temperature of the physiological saline instilled, and the contribution of bedside nurses’ knowledge and experience related to IBP measurement.

This research was a subordinate reliability study conducted within a larger prospective randomized observational study whose aims were to systematically explore the effects of patient position and volume of saline instilled on measurement of IBP and to identify other clinical factors that influence IBP measurements. Although the larger study is not germane to this article, understanding the design of the larger study is helpful. Briefly, 120 patients who consented to participate were randomized to 1 of 12 groups (10 patients per group). The 12 groups were combinations of 4 body positions and 3 instillation volumes. The 4 positions were supine with no elevation of the head of the bed (flat), supine with the head of the bed elevated 30°, right lateral with the head of the bed elevated 30°, and left lateral the head of the bed elevated 30°. The volumes of saline instilled were 25, 50, and 200 mL. The randomization scheme was prepared by the statistician and delivered to the principal investigator in sealed envelopes that were opened after the patients signed the informed consent forms. The complete description, design, and results of the parent study can be found elsewhere.²¹ 

Melanie Horbal Shuster was a doctoral student at Duquesne University in Pittsburgh, Pennsylvania, when this study was done, and the research was completed in partial fulfillment of the requirements for the degree of doctor of philosophy. L. Kathleen Sekula served as chair of the dissertation committee, and John Kern served as the statistician and was a member of the dissertation committee.