Incidence and prognosis of intra-abdominal hypertension in critically ill medical patients: a prospective epidemiological study

Introduction The aim of this study was to determine the incidence of intra-abdominal hypertension (IAH) in patients with two or more categorized risk factors (CRF) for IAH, and their morbidity and mortality during their intensive care unit (ICU) stay. Methods Prospective cohort study carried out at a medical ICU. A total of 151 medical patients were enrolled during a period of 3 months. After ICU whole staff training, we conducted daily screening of the four CRF for IAH based on the World Society of Abdominal Compartment Syndrome (WSACS) guidelines (namely, diminished abdominal wall compliance, increased intraluminal content, increased abdominal content, and capillary leak syndrome or fluid resuscitation). In those patients with risk factors of at least two different categories (≥2 CRF), intra-abdominal pressure (IAP) was measured every 8 h during ICU stay. Data included demographics, main diagnosis on admission, severity scores, cumulative fluid balance, daily mean IAP, resolution of IAH, days of ICU and hospital stay, and mortality. Results Eighty-seven patients (57.6%) had ≥2 CRF for IAH, 59 (67.8%) out of whom developed IAH. Patients with ≥2 CRF had a significantly higher mortality rate (41.4 vs. 14.3%, p < 0.001). Patients with IAH had higher body mass index, severity scores, organ dysfunctions/failures, number of CRF for IAH, days of ICU/hospital stay and hospital mortality rate (45.8 vs. 32.1%, p = 0.22). Non-resolution of IAH was associated with a higher mortality rate (64.7 vs. 35.3%, p = 0.001). None of the cohort patients developed abdominal compartment syndrome. The multivariate analysis showed that IAH development (odds ratio (OR) 4.09; 95% confidence interval (CI) 0.83-20.12) was a non-independent risk factor for mortality, and its non-resolution (OR 13.15; 95% CI 22.13-81.92) was an independent risk factor for mortality. Conclusions Critically ill medical patients admitted to ICU with ≥2 CRF have high morbidity, mortality rate, and incidence of IAH, so IAP should be measured and monitored as recommended by the WSACS. Our study highlights the importance of implementing screening and assessment protocols for an early diagnosis of IAH.

IAH has traditionally been associated with patients undergoing abdominal surgery and trauma patients who require aggressive volume resuscitation [12][13][14][15][16]. Several studies have demonstrated that IAH is also frequently present in critically ill nonsurgical patients, and is now considered to be associated with the general process of inflammation and resuscitation [17][18][19][20][21][22]. However, establishing comparisons for incidence, morbidity and mortality rates in medical patients from data of published clinical trials is difficult, mainly due to three reasons: firstly, the absence of uniform definitions before consensus guidelines, which allow us to reproduce and interpret the results; secondly, the different methods used for IAP measurement; and lastly, the fact that most clinical studies have been performed on mixed populations (both medical and surgical).
The World Society of Abdominal Compartment Syndrome (WSACS) has published a consensus statement including definitions and recommendations for the screening and management of IAH and abdominal compartment syndrome (ACS) [23,24]. Following this consensus, the aim of this study was to determine the incidence of IAH in exclusively critically ill medical patients with high risk of IAH development, and their morbidity and mortality during their intensive care unit (ICU) stay.

Materials and methods
This study was conducted in a university hospital medical ICU in accordance with a protocol that was approved by the local ethics committee. The study fulfilled the requirements of the Declaration of Helsinki, and included an informed consent from the patients on admission (or their family if necessary) to participate and publish the results.
All medical patients admitted to the ICU during a three-month period and expected to stay > 24 h were prospectively enrolled, provided they needed a bladder catheter. Exclusion criteria were ICU stay < 24 h, age < 18 years, pregnancy, contraindication for intravesical pressure measurement (pelvic fracture, hematuria, or neurogenic bladder), and bladder surgery. WSACS consensus for assessment of IAH was implemented as a new protocol [23,24].

Data collection Demographic data
Age, gender, body mass index (BMI) expressed as kg/m 2 [25], main diagnosis on admission, cumulative fluid balance, length of ICU and hospital stay, and hospital mortality were considered in collecting the patients' data. Patients were followed until death or hospital discharge, whatever came first.

Organ dysfunction
On admission, APACHE II (Acute Physiology and Chronic Health Evaluation) and SOFA (Sequential Organ Failure Assessment) scores were recorded using the worst values of the day [26,27]. Organ dysfunction/ failure was evaluated using the SOFA score (dysfunction = subscore ≤2, failure = subscore ≥3). The SOFA maximum score was based on the worst value of each organ system during the ICU stay.

Risk factors for IAH
All patients were prospectively screened for risk factors (RF) of IAH on admission and daily during their ICU stay until death or discharge, based on four categories from the WSACS algorithm (Table 1). Those patients who presented any RF of at least two different categories (≥2 categorized risk factors (CRF)), at any time on admission or during the ICU stay, were considered to have a high risk of developing IAH, and IAP was monitored (starting the day they were found to have ≥2 CRF). Those patients with < 2 CRF on admission and daily during the ICU stay were considered to have a low risk of developing IAH according to WSACS recommendations and their IAP measurement was not recorded. For the main analysis, we calculated the number of RF as categories (one minimum, four maximum) on admission (first 24 h) and during the ICU stay (mean RF categories present per day).

IAP measurement
IAP was measured intravesically using a closed-system Foley bladder catheter according to the modified Kron technique [28,29] with an instillation volume of 25 ml of saline. Abdominal perfusion pressure (APP) was calculated as mean arterial pressure minus IAP [30,31]. Measurements of IAP and APP were recorded every 8 h in patients with ≥2 CRF until resolution of IAH (in case it developed), death, or discharge from the ICU. IAH was defined as a repeated (at least three consecutive values separated by 8 h) elevation of IAP min (the lowest value of each of the three IAP measures) ≥12 mmHg [23]. For the main analysis, IAP mean and APP mean (mean of the three daily values) were calculated per day.

IAH characteristics
We recorded the day of onset, days of sustained IAH, surgical abdominal decompression, and resolution. Resolution of IAH was defined as a sustained IAP min < 12 mmHg in all measurements recorded every 8 h for a minimum of 48 h; and non-resolution was defined as a sustained IAP min ≥ 12 mmHg in all measurements until death or discharge from the ICU. IAH was classified according to the time of onset (on admission or during ICU stay) and the development of symptoms (acute, over a period of hours; subacute, over a period of days; or chronic, over a period of months/years). We also recorded the grade of maximal IAP reached during the ICU stay (I, 12-15; II, 16-20; III, 21-25; IV, ≥25 mmHg) [23].

Statistical analysis Univariate analysis
Continuous non-normally distributed variables were expressed as the median and interquartile range and compared using the nonparametric Mann-Whitney test. Continuous normally distributed variables were expressed as mean ± standard deviation, and compared using the student's t test. Categorical variables were expressed as numbers and percentages and compared using the chi-square test or Fisher exact test. All p values were two-tailed, and p < .05 was considered as statistically significant.

Multivariate analysis
The prognostic relevance of two different parameters was analyzed: variables associated with development of IAH and the prognostic value of IAH and its non-resolution. Backward multiple logistic models were used, including all the variables yielding p < .2 by univariate analysis and those considered clinically relevant. Possible interactions were tested. The results were summarized as odds ratios their respective 95% confidence intervals. The area under the receiver operating characteristic (ROC) curve was used to assess discrimination; an area under the ROC curve of 1.0 denotes perfect discrimination, whereas a value close to 0.50 indicates no apparent accuracy. The Hosmer-Lemeshow goodness-of-fit test was used to evaluate agreement between the observed and expected results across all the probability strata of the outcome of interest (calibration); p > .05 indicates a good fit for the model. Kaplan-Meier curves for survival at 28-day hospital stay were constructed for subgroups with IAH.

Criteria for IAP monitoring and incidence of IAH (Figures 1 and 2)
During the study period, 204 patients were admitted. Fifty-three were excluded: 5 refused to participate, 22 had ICU stay < 24 hours, 12 had contraindications for intravesical pressure measurement or bladder surgery, 11 did not require urinary catheterization, 2 were < 18 years and 1 was pregnant. We finally enrolled 151 medical patients, and the entire protocol was completed in all of them. Main diagnosis on admission was • Hypotension (systolic blood pressure < 90 mmHg or mean arterial pressure < 70 mmHg or a systolic blood pressure decrease > 40 mmHg or > 2 standard deviation below normal for age in the absence of other causes of hypotension) • Hypothermia (core temperature < 33°C).
•  Figure 2 shows the differences in the evolution of the number of CRF during the ICU stay across different subgroups of patients. Comparisons between survivors and non-survivors in the study cohort and high risk group are shown in Tables 3 and 4. There were no statistical differences in IAP levels between survivors and non-survivors (11.8 vs. 12.0 mmHg, p = 0.78).
General characteristics for the comparison between high risk and low risk groups (Table 5) We compared patients with ≥2 CRF (n = 87) with those with < 2 CRF for IAH (n = 64). Severity of illness, ICU and hospital length of stay, and mortality rate (41.4 vs. 14.3%, p < 0.001) were significantly higher in the high risk group.
General characteristics for the comparison between IAH and non-IAH groups (Tables 6 and 7) We compared patients who developed IAH with ≥2 CRF (n = 59) with those who did not despite having ≥2 CRF (n = 28  and IAP mean (13 + 1 vs. 8 + 1 mmHg, p < 0.001) were also significantly higher.
General characteristics of patients with IAH (Table 8, Figure 3) Non-survivors had significantly higher severity scores, number of organ failures, and CRF for IAH, both on admission and during the ICU stay. The more commonly found subtypes of IAH were subacute (51 patients, 86.4%) and grade II (34 patients, 57.6%). IAH was resolved in a high proportion of patients (42.4%) and none presented ACS or required surgical abdominal decompression. Patients in whom IAH was not resolved (57.6%) had a significantly higher mortality rate than patients whose condition was resolved (64.7 vs. 35.2%, p = 0.001). Kaplan-Meier survival analysis for patients with IAH on admission or during the ICU stay depending on resolution is shown in Figure 3.
Multivariate analysis: predictive and prognostic models (Tables 9, 10 and 11) Predictive model for IAH development in the high risk group (87 patients with ≥2 CRF) SOFA on admission, obesity, and number of CRF during ICU stay (odds ratio (OR), 2.31; 95% confidence interval (CI), 1.09 -4.88) were independent predictors of IAH development.
Predictive model of mortality in the high risk group (87 patients with ≥2 CRF) IAH development was a non-independent predictor of mortality (OR, 4.09; 95% CI, 0.83 -20.12).

Predictive model of mortality in the IAH group (59 patients)
Non-resolution of IAH (OR, 13.15; 95% CI, 22.13 -81.92), number of CRF during ICU stay (OR, 9.94; 95% CI, 1.59-62.00), and age ≥65 years were independent predictors of mortality. Even though the ROC curve showed good discrimination, the Hosmer-Lemeshow test indicated inadequate calibration across all the probability strata for mortality in the model.

Discussion
This is a prospective epidemiologic study performed in a population of exclusive critically ill medical patients following the WSACS guidelines for the assessment of IAH (screening, IAP measurement, definitions, and classification recommendations). Other published studies had been performed in mixed populations (medical and surgical), before the WSACS consensus statement, based on different definitions of IAH or limited by incomplete assessment in terms of screening or classification of the IAH [17][18][19][20][21].
Our study revealed several interesting findings. Firstly, there was a large number of patients with high risk (≥2 CRF) for IAH development, and the incidence of IAH in this group was high; in addition, the mortality in this subgroup was significantly higher as compared with the low risk subgroup. Secondly, there was an association of IAH with illness severity and the number/type of RF reported in the literature. Thirdly, IAH was a predictor of mortality in association with other clinical factors. And finally, nonresolution of IAH was an independent predictive factor of mortality. Our results support the importance of implementing screening and assessment protocols, according to WSACS recommendation, in order to identify a subgroup of high risk patients with high mortality and high incidence of IAH whom should be monitored for IAP.

Screening, risk factors and incidence of IAH
A large number of patients presented ≥2 CRF for IAH development, and the incidence of IAH in this group was high (67.8%) on admission or during the ICU stay. In addition, several high risk patients without IAH might have shown underestimated IAP levels because of the different factors usually present in the critically ill population, such as sedation or gastric/colonic decompression. A high incidence of IAH had also been reported in the literature [17][18][19][20][21]; however, when trying to combine our results with previously described findings, three main methodology issues should be accounted for: a) incidence is recorded on admission or during the first week; b) most reports refer to mixed populations (both medical and surgical); c) definitions of IAH differed across the studies (incidence rates can vary depending on the threshold IAP level used as a diagnostic criterion). Unlike previous studies, our study analyzes the incidence of IAH in a high risk population on admission and throughout the entire ICU stay, following WSACS consensus guidelines, and exclusively in critically ill medical patients. Reintam et al. [17] and Vidal et al. [18] have also used WSACS definitions for IAH, and they report incidence rates of 37% in a mixed population, and 43% in the medical subgroup of a mixed population, respectively. Numerous conditions predispose to IAH development, and our results are consistent with those from other studies [20,21]. In the high risk group, we found an  association between IAH and the CRF proposed by the WSACS, specially the increased abdominal content. Obesity and the number of CRF for IAH during ICU stay were independent predictors of IAH development. Data on fluid balance were only qualitative (positive or negative), and the association between IAH and massive fluid loading in the presence of sepsis and/or capillary leak described in the literature could not be studied.

Severity scores and organ failure in patients with IAH
Increases in IAP have deleterious effects on end-organ system function by compromising perfusion, thus leading to multiorganic dysfunction (MOD) and poor outcome. In our population, IAH was significantly associated with higher severity of illness scores and incidence of organ failure. In addition, the study design allowed us to compare organ impairment during ICU stay between patients with and without IAH (Table 5). Previous studies analyzed organ impairment only during the first days of ICU stay, thus making it difficult to study the association between MOD and IAH throughout the entire ICU stay.
Our results in medical patients are consistent with this observation. The IAH classification varies depending on the study; however, no studies include all the classification types currently recommended by WSACS. In summary, a complete classification is important for planning future studies and comparing results, and the differentiation between medical or surgical patients is, in our opinion, necessary for a correct analysis based on the differences with regard to the causes and evolution of IAH in those two types of patient's profile. Fourteen patients reached grade III of IAH, and five grade IV, but none of them developed ACS because a sustained elevation of IAP > 20 mmHg associated with organ deterioration was not detected. In concordance with this, no patient required surgical abdominal decompression.

Prognostic implications of IAH
Development of IAH has been described as an independent predictor of mortality in mixed populations [20]. In our study IAH was a non-independent predictor of mortality, and this supports the fact that IAH was a marker of mortality in association with other clinical factors. In accordance with this, patients with IAH had higher severity scores and number of CRF for IAH that can independently increase mortality risk. In fact, the number of CRF during ICU stay was an independent predictor of mortality in the group with IAH.  We found a high rate of resolution of IAH in our medical patients, but non-resolution was an independent predictor of mortality, resulting in a considerably high OR (non-resolution meant that death was 13 times more likely). Treatment of IAH was nonsurgical in all cases and none developed ACS, although no conclusions can be drawn in the absence of clinical intervention data. Management guidelines for IAH have been proposed by the WSACS [24] although they are based on studies that analyze therapeutic strategies in isolation and in different patient profiles [36][37][38][39][40].
Limitations of the study IAP was not measured in all cases (only in those with ≥2 CRF), and patients with < 2 CRF and no IAP recorded might have developed IAH during their ICU stay (e.g., patients with mechanical ventilation and therefore, diminished abdominal wall compliance but with no RF in other categories and no IAP measured). Therefore, the study followed the WSACS screening recommendations but was not designed to validate them in order to identify a subgroup of high risk patients for IAH development, as compared with a low risk subgroup. The study was underpowered to analyze each RF of the four categories as predictors of IAH (small sample size) or the IAP level as a predictor of mortality (the IAP mean in the patients with IAH were relatively low to perform comparisons in the outcome). Not all patients who had their IAP measured were sedated, which could have caused a falsely elevated IAP. Although resolution of IAH was a prognostic factor, our study was not designed to focus on therapeutic management of IAH, and no conclusions can be drawn or ascertainments can be made about the strategies that were effective.

Conclusions
Critically ill medical patients admitted to the ICU with high risk of IAH development (≥2 CRF) have high morbidity, mortality, and incidence of IAH, so IAP should be measured as recommended by the WSACS. In this subgroup of patients, the presence of IAH is associated with mortality as a non-independent risk factor, and non-resolution of this process is an independent predictor of mortality. Our study supports the implementation of protocols based on WSACS guidelines for IAH screening and assessment to allow an early approach to this syndrome. Large, multicenter randomized controlled clinical trials are needed to confirm our results and to further promote the WSACS recommendations regarding IAH screening and management.

Consent
Written informed consent was obtained from the patients or their relatives for publication of this case report. A copy of the written consent is available for review by the Editor-in-Chief of this journal.