Frequency of Delirium in Non-Cardiac Surgical and Medical Intensive-Care Patients-Results from a Comparative, Prospective, Observational Study

Research Article

Austin Crit Care J. 2017; 4(1): 1021.

Frequency of Delirium in Non-Cardiac Surgical and Medical Intensive-Care Patients-Results from a Comparative, Prospective, Observational Study

Zeder M¹*, Müller T², Zeman F³, Schlitt HJ4, Blecha S¹, Graf BM¹ and Bein T¹

1Department of Anesthesiology, Operative Intensive Care, Regensburg University Hospital, Germany

2Department of Internal Medicine II, University Medical Center Regensburg, Germany

3Center for Clinical Studies, University Medical Center Regensburg, Germany

4Department of Surgery, University Medical Center Regensburg, Germany

*Corresponding author: Marius Zeder, Department of Anesthesiology, Operative Intensive Care, Regensburg University Hospital, Franz-Josef-Strauss-Allee 11, 93053, Regensburg, Germany

Received: August 31, 2017; Accepted: December 08, 2017; Published: December 15, 2017

Abstract

Introduction: Delirium has a high incidence in ICUs (Intensive-Care Units) and is associated with adverse outcomes. Amongst other factors, inflammation is predominantly linked to the risk of delirium. We hypothesized that the inflammatory stress of surgery/trauma leads to a higher incidence of delirium in surgical compared to medical ICUs. We performed a prospective comparative study, and determined the risk factors for delirium by multivariable analyses.

Materials and Methods: A prospective single-centre University Hospital study on two ICUs was performed. Patients were screened for delirium with the CAM–ICU (Confusion Assessment Method for the ICU) at several days (1, 3 and 5 after admission/extubation). Demographic data, laboratory values, and administered medication details were gathered. Non-cardiac surgical and medical ICU patients were compared and Odds Ratios (OR) were calculated via univariable and multivariable logistic regression models.

Results: The incidence of delirium in all patients (n=138) was 32.6% and no difference was found between surgical and medical ICU patients. Patients with delirium received significantly more psychoactive medication, benzodiazepines, propofol, and morphine. They had higher Simplified Acute Physiology Score, higher age, and longer ventilation and stay in the ICU. Patients without delirium received significantly more oxycodone/naloxone (15.4mg vs 5.3mg, p=0.041). They had higher levels of serum haemoglobin (10.9g/l vs 9.8g/l, p=0.004) and albumin (27.4g/l vs 23.7g/l, p<0.001) and lower values of urea (47.6mg/dl vs 81.2mg/dl, p<0.001), bilirubin (1.12mg/dl vs 2.3mg/dl, p=0.001), creatinine (1.5mg/dl vs 2.3mg/dl, p<0.001), CRP (C-Reactive Protein) (67.9 mg/l vs 102.6mg/l, p =0.035) and sodium (138.8mmol/l vs 142.0mmol/l, p<0.001). Multivariable logistic regression showed creatinine (p =0.028), CRP (p=0.022), and duration of ventilation (p=0.002), as independent predictors of delirium.

Discussion: Despite higher CRP as a marker of inflammation, patients in the surgical/trauma ICU did not have higher incidence of delirium. A possible explanation could be the higher administration of oxycodone/naloxone, an opioid with a trend towards a “delirium-protecting’ potency. Nevertheless, further studies are needed to prove this hypothesis, as the development of delirium seems to be a multifactorial process, and a “bundle’ for prevention is needed.

Keywords: Delirium; Intensive-Care Medicine; Incidence; Oxycodone; Non-cardiac surgical intensive care; Medical intensive care

Abbreviations

ICU: Intensive-Care Unit; DSM-IV: Diagnostic and Statistical Manual of Psychiatric Diseases IV; SICU: Surgical/trauma Intensive- Care Unit; MICU: Medical Intensive-Care Unit; CRF: Case Report Form; CAM–ICU: Confusion Assessment Method for the ICU; RASS: Richmond Agitation Scale Score; CRP: C-Reactive Protein; SAPS: Simplified Acute Physiology Score; SD: Standard Deviation; IQR: Inter Quartile Range; OR: Odds Ratio; CI: Confidence Interval

Introduction

Delirium is the most common psychiatric disorder in intensivecare patients, with an incidence varying from 10 percent to 92 percent [1-4], and is associated with a higher mortality compared to those patients not suffering from delirium [5,6]. Despite the high incidence, delirium is frequently not detected by the intensive-care unit staff [7] and screening is poorly implemented in German ICUs [8].

Delirium is defined by the Diagnostic and Statistical Manual of Psychiatric Diseases IV (DSM-IV) by impaired consciousness, perception and attention deficit, vigilance disorder and disorganization of thought processes [9]. Furthermore, the International Classification of Diseases also lists disorders of psychomotor activity, emotionality, and the sleep-wake cycle as criteria of delirium [10]. In recent studies, several risk factors for delirium have been examined and it was found that delirium was an independent factor for adverse outcome variables. Thus, prevention or early therapeutic management of delirium is an important goal in the ICU. The use of earplugs, increased input of daylight and early mobilization seem to prevent delirium [11]. Furthermore, the implementation of a so-called “ABCDE bundle’ significantly decreases the prevalence and duration of delirium [12].

The pathophysiology of delirium is poorly understood to date. There are seven mostly complementary theories for the development of delirium, which include neuroinflammatory processes, neuronal aging, oxidative stress, neurotransmitter deficiency, neuroendocrine factors, diurnal dysregulation and network disconnectivity hypotheses [13]. It was found that inflammation [14], low albumin level [15], use of analgesics [1] and increased volume load (during surgery)-predominantly in cardiac surgery patients [16]-were associated with increased risk of delirium. Surgery can lead to systemic immunosupression and pro-inflammatory periods [17]. Additionally persistent infections [18] as well as sepsis [19] are common complications in surgical ICUs. Therefore we hypothesized that the incidence of delirium would also be higher in intensive-care patients after trauma or non-cardiac surgery compared to medical intensive-care patients, if higher inflammatory stress due to previous surgery/trauma could be related to delirium [14,20].

The aims of this study were to analyze the frequency of delirium in non-cardiac surgical/trauma compared to “medical’ admissions to the ICU in a prospective single-centre University Hospital study and to determine risk factors for delirium by multivariate analyses.

Material and Methods

A prospective study was performed in two ICUs of a University Hospital: a 26-bed Surgical Intensive-Care Unit (SICU) and a 14-bed Medical Intensive-Care Unit (MICU). The approval of the regional Institutional Review Board (Ethikkommission der Universität Regensburg) was obtained (Approval No.15-101-0101).After a twoweek pilot period, during which the Case Report Form (CRF) and the study protocol were tested, all patients who were admitted to the ICUs between June and November 2015 and who were treated for at least one day and more, were screened consecutively with the Confusion Assessment Method for the ICU (CAM-ICU) at least on days one, three and five after admission to the ICU (non-ventilated patients) or after extubation (ventilated patients). The CAM-ICU is a validated tool with high specificity and good sensitivity for detecting delirium in the ICU [21]. Patients were found to be eligible for delirium assessment when fulfilling the following criteria: a Richmond Agitation Scale Score (RASS) of-3 or higher [22], the absence of mechanical ventilation and no contraindication for the screening with the CAM-ICU. If patients were transferred after day one to another ward, delirium assessment was continued there. If patients stayed in the ICU longer than five days, screening was performed every two days until transfer.

Furthermore, the following variables were recorded: C-Reactive Protein (CRP), urea, albumin and procalcitonin from serum; the arterial pH, haemoglobin, blood glucose, and electrolytes from the 8a.m. blood gas analyses were registered (all at first delirium assessment). In addition, the cumulative doses were recorded of sedative (propofol, lorazepam, midazolam) and analgesic (morphine, hydromorphone, sufentanil, oxycodone/naloxone) drugs, as well as vasoactive medications (noradrenaline, clonidine) and psychoactive drugs (levomepromazine, promethazine, quetiapine, melperone, haloperidol)-24 hours prior to every screening (days one, three and five).Administration of pain medication in the ICUs was standardized using a visual analoge scale. All this information was retrieved from pre-existing data in the patient data management system (Metavision, iMDsoft, Düsseldorf, Germany).

Body temperature, arterial oxygen saturation and mean arterial pressure during screening, as well as age, gender, SAPS of the first 24 hours after admission to ICU, the main reason for admission and the duration of previous ventilation and stay in the ICU were recorded. Additionally, the highest value of serum bilirubin and creatinine during the stay was collected. All CAM-ICU assessments were performed by one investigator (MZ) after an extended supervised training through experienced intensives (TB).

Statistical Methods

Statistical calculations were performed with SPSS Statistics 23 (IBM Corp., Armonk, New York, USA). Statistical significance was defined by p-value<0.05.

Quantitative variables are presented by arithmetic mean ± Standard Deviation (SD), median and Inter Quartile Range (IQR); categorical variables by count and percentage. Admissions to SICU/ MICU were analysed using Student’s T-test for normally distributed values and Mann-Whitney U-test for non-normally distributed values. Results of both tests are shown with mean ± SD in text and tables to improve readability. The Shapiro-Wilk test was used as a test of normality (only temperature showed normal distribution). For comparisons of categorical variables-gender, intubation, tracheotomy and administration of psychoactive drugs-Pearson’s Chi-Square test were used and Mantel-Haenszel odds ratios were calculated.

Univariable logistic regression models to identify risk factors for delirium were calculated. Independent variables were the significant variables of prior tests and the variables described as significant predictors of delirium in the literature.

Further, a multivariable logistic regression model using significant (p<0.05) predictors of the univariable models was calculated. Results from both logistic regression analyses are presented as Odds Ratios (OR) with 95% Confidence Interval (CI).

Results

142 patients were screened for delirium. Four patients were excluded from the study due to impossibility of data gathering, incomplete screenings or for fulfilling the exclusion criteria.138 patients were included in the study (SICU: 71, MICU: 67). More male patients were analyzed (n=88, 63.8%).86(62.3%) patients received mechanical ventilation through endotracheal tube and seven (5.1%) received tracheotomy. Of all the patients, 32.6% developed delirium and 3.6% died during their stay in the ICU.

The mean age was 61.7±16.3 years (median 65 years, IQR 21), with a mean SAPS of 36.5±14.4 (36, IQR 20), patients were mechanically ventilated for a mean of 2.6±4.2 days (1, IQR 3) and they stayed in the ICU for an average of 7.2±6.2 days (5, IQR 7). The characteristics of the cohort are displayed in (Table 1).