About one-third of the patients who develop mild, acute confusion (i.e., subsyndromal delirium) will go on to develop a severe acute confusional state (i.e. delirium). Delirium refers to a temporary change in the way a person thinks about things. Delirium occurs in patients admitted to the hospital particularly those patients that are very sick, who are given a number of medications, and who are not able to sleep normally. It affects their behavior, their understanding of the people and things around them, and their ability to make decisions. While ICU doctors do everything possible to eliminate the factors that may cause delirium, delirium may cause a person to become very agitated which if not controlled is dangerous to their safety as well as the safety of those around them. As well, if delirium develops in patients in the ICU, it may increase the risk for death, keep patients in both the ICU and hospital for longer and send patients to a long term care facility rather than home after they are discharged from the hospital. A recent medical report found that patients in the ICU who develop subsyndromal delirium have a mortality rate, a length of stay in both the ICU and the hospital, and a transfer rate to a long term care facility that is nearly as great as patients with delirium and greater than patients who develop neither subsyndromal delirium or delirium. Recent studies in non-ICU patients suggest that if a patient who is at high risk for developing delirium receives a medication called an antipsychotic (e.g. haloperidol) they may not be as likely to develop delirium or if they do develop delirium it will not last as long. No studies have evaluated the effect of administering an antipsychotic in patients in the ICU who have subsyndromal delirium. Another study completed in the ICUs at Tufts Medical Center found that there may be an association between the development of delirium in patients with subsyndromal delirium and the use of haloperidol. However, this small study had many limitations and thus it is not currently known whether using haloperidol in patients with subsyndromal delirium will actually cause fewer of these patients to develop delirium. Haloperidol makes a person sleepy and helps control behavior like agitation. Haloperidol is the drug that is used most often to help control delirium in the ICU. This prospective, double-blind, randomized controlled study will determine if haloperidol administered through the vein four times daily (1mg IV q6h) to patients who have subsyndromal delirium, and who are on a breathing machine and being cared for by the Medical ICU service at Tufts Medical Center, will help prevent patients from developing delirium. A total of 68 participants will be enrolled. Exclusion criteria are extensive and include conditions that could affect the ability to determine if delirium is present or increase the risk for side effects related to the administration of haloperidol. Patients older than 80 will be excluded from the study. Study medication (i.e. haloperidol) will be administered until one the following occurs: 1) delirium develops (that is confirmed by a staff psychiatrist or his designate, 2) the patient is discharged from the ICU at Tufts Medical Center, 3) the patient has received haloperidol or placebo for 10 days or 4) an adverse event potentially attributable to the study drug is experienced by a patient that is deemed, in the opinion of a pulmonologist member of the investigative team to warrant discontinuation of therapy. Haloperidol may cause unwanted side effects such as low blood pressure, twitching, and an unsafe abnormal heart rhythm. Patients with chronic confusion (e.g., a dementia such as Alzheimer's Disease) should not receive haloperidol and will not be included in this study. Patients will be carefully monitored for side effects that are potentially related to haloperidol. Patients who become confusion-free in the ICU before they leave the ICU (i.e., have no subsyndromal delirium) will be asked to provide consent for all research activities that occured in the ICU. If patients where cognition is regained (ie. no subsyndromal delirium or delirium) are not willing to provide consent then any study data collected from them while they were in the ICU will be destroyed and they will not be approached to participate in the post-ICU component of the study. This study also seeks to understand how the use of haloperidol in the ICU in patients with subsyndromal delirium may have affect memory, emotional status, happiness, ability to function, and quality of sleep in patients after they leave the ICU. Patients (that do not have delirium based on CAM screening at the time the 3-10 day and 6 month assessments are attempted) will be approached to participate in this post-ICU component of the study.

Delirium occurs in up to 50% of patients who are not mechanically ventilated and up to 80% of intensive care unit (ICU) patients who are mechanically ventilated. Even after controlling for preexisting morbidities, severity of illness, coma and use of sedatives, patients who develop delirium in the ICU are more than 3-times as likely to be dead at 6 months compared to those patients who do not develop delirium. (1-4) Additionally, ICU delirium is predictive of a threefold-higher reintubation rate and > 10 additional days in the hospital. (5) Each additional day of mechanical ventilation places patients at increased risk for life-threatening sequelae such as ventilator-associated pneumonia and barotrauma. The sequelae of delirium are cumulative; for example, each additional day spent in delirium is associated with a 20% increased risk of prolonged hospitalization and a 10% increased risk of death. Given these negative outcomes, it is not surprising that delirium is independently associated with substantially higher ICU and total hospital costs than for patients who do not develop delirium. (6) Up to a quarter of patients who develop delirium in the ICU develop long-term cognitive impairment that frequently manifests as dementia. (7-10) This post-ICU long-term cognitive impairment affects memory, attention, and executive function. It therefore compromises the ability to return to work, impairs activities of daily living, increases the need for institutionalization, and decreases quality of life. (11-13) More than 5,000,000 patients are admitted to an ICU in the United States each year. Delirium may develop in up 80% of these patients. (1, 2, 14) It is therefore critical that interventions that can prevent delirium from occurring in the ICU be discovered.

Given the negative sequelae of delirium, current practice guidelines recommend routine delirium screening in the ICU with a validated screening tool such as the confusion assessment method for the ICU (CAM-ICU) or the Intensive Care Delirium Screening Checklist (ICDSC). (15-17) The ICDSC (Appendix 1) is an 8-item checklist producing a score between 0 and 8, where each item (corresponding to a value of 1 out of 8) reflects a DSM-IV (or alternative) diagnostic characteristic of delirium and has many advantages over the CAM-ICU. (15, 16, 18) The data are collected in real-time during each nursing shift. An ICDSC score ≥ 4 has been shown to highly correlate with a formal psychiatric diagnosis of delirium using DSM-IV criteria (sensitivity 99%, ROC curve area = 0.901), and has excellent inter-observer reliability between nurses and between nurses and critical care physicians.

Recent investigations, both within and outside of the ICU, have moved beyond simply assigning patients the dichotomous outcome of either being delirious or not, and instead focused on grading the severity of delirium. (19, 20) Patients who have some of the characteristics of delirium, but not full clinical delirium, are deemed to have subsyndromal delirium.[10] The ICDSC, given that its 8 different domains have each been shown to have good item reliability towards the diagnosis of delirium, allows one to identify those ICU patients that fall short of the threshold diagnostic criteria for overt or 'clinical delirium' but who still may have clinically important sub-syndromal delirium (ie. an ICDSC of 1-3). (19)

One recent ICU study compared the outcomes of patients who developed sub-syndromal delirium (ICDSC 1 to 3) over the course of their ICU stay to those who developed delirium (ICDSC ≥ 4) or who developed neither delirium or subsyndromal delirium (ICDSC = 0). In their study of 537 consecutive, non-comatose patients, 31.5% developed neither clinical delirium or subsyndromal delirium, 33.3% developed subsyndromal delirium (but not clinical delirium) and 35.3% developed clinical delirium. Patients who developed subsyndromal delirium had an ICU mortality rate that was more than 4x greater than patients who developed neither clinical delirium or subsyndromal delirium (10.6% vs 2.4%, p=0.002). In addition, the patients who developed subsyndromal delirium, when compared to those patients without clinical or subsyndromal delirium, had both significantly longer ICU and hospital lengths of stays and were more likely to be discharged to a long-term care facility than home.

When correction of the potential causative factors for delirium fails to reverse delirium, clinicians frequently turn to antipsychotic therapy. (21) Haloperidol remains the first-line pharmacologic therapy for the treatment of delirium by critical care clinicians and is advocated for use in this population by practice guidelines. (14, 22) One large retrospective cohort analysis, where potential confounding variables were addressed, found that use of haloperidol within 2 days of mechanical ventilation reduced death. (23)

Two recent studies demonstrate the benefit of antipsychotic therapy in non-ICU patients at high risk for developing delirium. (23, 24) In one randomized study of elderly patients undergoing hip surgery, a population at high risk for developing delirium, administration of oral haloperidol 0.5mg q8h decreased the average Delirium Rating Scale (DRS) Score (14.4 ± 3.4 vs 18.4 ± 4.3, p < 0.001), the duration of delirium (5.4 vs 11.8 days, p < 0.001) and the mean number of days in the hospital (17.1± 11.1 vs 22.6 ± 16.7, p < 0.001). No haloperidol-related side effects were noted. (23) A more recent placebo-controlled study demonstrated that the post-operative administration of a single dose of risperidone in patients undergoing coronary artery bypass surgery decreased the incidence of delirium from 31.7% to 11.1% (p= 0.009). (24) It should be noted that neither of these studies enrolled ICU patients and the incidence of subsyndromal delirium at baseline was not measured.

The safety and efficacy of treating ICU patients with delirium symptoms with antipsychotic therapy has not been evaluated. Recent data highlights the deleterious sequelae of subsyndromal delirium in the ICU. In addition, the results of two studies demonstrate the benefits of administering low-dose antipsychotic therapy in patients at high risk for delirium. We therefore hypothesize that treatment of subsyndromal delirium in the critically ill with haloperidol will prevent progression to clinical delirium. This should be associated with decreased agitation, a shorter duration of mechanical ventilation, and improved cognitive function after ICU discharge. Given the documented safety of low-dose haloperidol therapy in clinical practice, we anticipate the potential risk-benefit ratio for this intervention to be low - particularly in an ICU environment where patients receive continuous cardio- respiratory monitoring.

Critically ill patients with subsyndromal delirium (ICDSC 1-3) are also at risk for substantial morbidity and mortality and are at high risk for developing delirium (ICDSC ≥ 4). Recent studies suggest that the administration of low-dose antipsychotic therapy to non-ICU patients at risk for delirium prevents transition to delirium, reduces delirium severity and shortens length of hospital stay without causing adverse effects. Given that up 1.5 million Americans may develop subsyndromal delirium each year, an intervention that is found to be safe and effective, particularly if it is of low cost, will have a profound impact on the U.S. healthcare system.

A subsyndromal risk factor study evaluated ICSDC nursing assessment data for 100 consecutive mechanically ventilated patients admitted to the ICU at Tufts MC for ≥ 24 hours without a neurologic reason for admission (e.g., acute stroke) or having acute alcohol withdrawal. Of the 100 patients, 72 developed subsyndromal delirium (ICDSC 1 to 3). Of these 72 patients, 26 (36%) went on to develop delirium (ICDSC ≥ 4) during their ICU admission. The remaining 28 patients developed neither subsyndromal delirium or delirium over the course of their ICU stay. For the 26 patients who developed delirium, we found that only 2 of the patients received ≥ 1 dose of haloperidol therapy in the period prior to when delirium was first identified. In contrast, for the 46 patients with subsyndromal delirium who did not progress to delirium, 14 of these patients received ≥ 1 dose of haloperidol during their ICU stay. Therefore, there was a weak association for a lower incidence of delirium in those patients with subsyndromal delirium who received ≥ 1 dose of haloperidol at any time during their ICU admission than patients who did not receive haloperidol [2/16 (12.5%) vs 24/56 (43%), p < 0.0001]. This retrospective, observational data suggests that patients with subsyndromal delirium who are exposed to haloperidol may remain in subsyndromal delirium and not progress to delirium. However, there are many methodological limitations with this pilot study thus it remains unknown if administering low-dose haloperidol to ICU patients with subsyndromal delirium will prevent the development of delirium.

Over the past 3 years, our delirium research group, a joint collaboration between Northeastern University and Tufts Medical Center has conducted a number of studies surrounding the identification and treatment of delirium in the ICU environment. (18, 26-29) A three-hospital evaluation lead by our group found that an educational intervention consisting of both patient cases using script concordance theory and a didactic presentation improved the ability of the bedside nurse to use the ICDSC correctly and to recognize delirium when compared to a judge. (26) Another study our group conducted found physician use of the ICDSC improves their ability to recognize delirium compared to a validated judge. (27) A large regional survey focusing on critical care nurse preferences and practices regarding delirium assessment compared delirium assessment practices to those for sedation assessment. (28) Finally, our group concluded a 36 patient double-blind randomized controlled trial evaluating the use of quetiapine for the delirium treatment in the ICU that was recently published in Critical Care Med. (29) The track record of our research group is important given the many challenges of implementing an ICU delirium-focused investigation such as the one that we have proposed. We therefore feel that we are qualified to conduct the study we have proposed to evaluate the safety and efficacy of haloperidol therapy in critically ill patients with subsyndromal delirium.

This prior data and our experience conducting delirium research demonstrate:

1. That progression to delirium may be lower in patients with subsyndromal delirium who receive haloperidol.
2. Our ability to conduct high quality research in the field of ICU delirium.
3. A multidisciplinary ICU delirium research group that includes a psychiatrist.
4. An academic environment where delirium assessment with the ICDSC has become a standard of care.

Delirium occurs in up to 80% of mechanically ventilated patients admitted to the intensive care unit (ICU). (1-4) Patients who develop delirium are more likely to die, stay in the ICU longer and cost the healthcare system more. (1-6) Moreover, delirium may cause cognitive dysfunction that persist long after recovery and impact long-term functional ability and quality of life. (7-10) Given that more than 5 million patients are admitted each year to an ICU in the United States, delirium in the critically ill is a major health care issue. Patients who meet some but not all of the Diagnostic and Statistical Manual of Mental Disorders (DSM) IV criteria for delirium are deemed to have subsyndromal delirium. (19) Patients who develop subsyndromal delirium are also more likely to die in the ICU and require a higher level of care after ICU discharge. (19) Current practice guidelines recommend routine delirium screening in the ICU using a validated tool such as the Intensive Care Delirium Screening Checklist (ICDSC). (14, 22) Delirium often persists despite the reversal of causative factors thereby necessitating the use of pharmacologic therapy. Haloperidol remains the pharmacologic treatment of choice for delirium in the ICU and its use has been associated with lower ICU mortality. (23) However, its potential role in preventing the conversion of subsyndromal delirium to delirium is unknown. In this study we plan to determine whether the treatment of subsyndromal delirium in the critically ill with haloperidol prevents progression to delirium without causing adverse effects.

Accomplishing these aims will determine if a pharmacologic intervention, haloperidol, can be routinely used in critical care practice to prevent patients with subsyndromal delirium from developing clinical delirium. If haloperidol is found to be safe and effective in the proposed study, larger studies will be needed to determine the effect of haloperidol therapy on patient mortality and evaluate predictors for drug response. Information from the proposed study is greatly needed to determine if there is a safe and effective pharmacological intervention for the estimated 1.5 million Americans each year who develop subsyndromal delirium in the ICU.

Survivors of critical illness who develop delirium may experience prolonged cognitive impairment and decline, decreased functionality and may suffer significant anxiety and depression. Sleep disturbance is a common factor in patients with delirium and is known to contribute to impaired function. Hence, a strategy that aims to reduce the incidence of delirium in the ICU may improve both short and long-term outcomes.

Neurocognitive deficits persist in up to 75% of ICU patients 6-12 months after ICU discharge and presumably may be higher in ICU patients who develop delirium (30). Cognitive deficits are global (31-33), including deficits in processing speed, memory, and attention. (31) Hopkins concludes that the most consistent deficit across studies is decreased memory. (30) This pattern of deficits, as well as the consistent findings of impaired social and occupational functioning in ICU survivors (32, 34) suggests impaired executive function. Executive function refers to the capacity to plan, initiate, organize, inhibit, and monitor thought and behavior. It is a pattern of impairment which is thought to represent diffuse frontal lobe damage. Executive function is typically poorly assessed in medically ill patients and yet has potentially substantial impact on overall functioning. Functional incapacity includes everyday issues such as difficulty buying groceries, keeping appointments or adhering to medication regimens. (34, 35) For geriatric patients treated in the ICU, only 3% have normal post-ICU functional capacity. (36) It remains unclear if there is a change in anxiety and depressive symptoms among patients who develop delirium and those that do not.

Delirium will worsen functional ability and it is suspected that it will also worsen emotional well being, although it is not well established whether this is the case. Given the deleterious effects that delirium may have on functional status, it is therefore important to measure physical status, social functioning, mental health and general health perception between between patients exposed to an intervention that may decrease the incidence of delirium. Sleep disturbance is a risk factor for neurocognitive impairment and is a prominent component of delirium. (37) Thus disturbed sleep is understood as both a potential cause and a consequence of ICU-related psychological and cognitive morbidity. Since psychological morbidity often persists after ICU discharge, it is likely that disturbed sleep is frequent during post-ICU hospitalization. An intervention focused on reducing delirium incidence will affect post-ICU sleep quality.

Inclusion Criteria:

• Age 18-84 years old
• Subsyndromal delirium (ICDSC of 1-3)
• Mechanically ventilated
• No objection from the MICU, CCU, or SICU service attending MD for enrollment
• Admitted to the MICU, CCU, or SICU service at Tufts Medical Center
• Patients is expected by their ICU attending physician to require admission to the ICU for ≥ 24 hours

Exclusion Criteria:

• Pregnancy
• Delirium (ICDSC ≥ 4)
• History of severe dementia or an Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE)) score ≥ 4
• IQCODE is not able to be completed.
• Current treatment with donepezil (aricept), tacrine (cognex), rivastigmine (exelon), or memantine (namenda)
• Admitted with a neurologic diagnosis (e.g., CVA)
• Past diagnosis of schizophrenia or a formal thought disorder as defined by DSM IV criteria.
• Treatment with an antipsychotic agent in the 30 days prior to ICU admission
• Current treatment with a neuromuscular blocker or dexmedetomidine
• A patient requiring a level of sedation equivalent to a sedation-agitation scale (SAS) score ≤ 2
• Inability to conduct valid ICDSC assessment (e.g., coma, deaf)
• Acute alcohol or drug withdrawal
• Acute drug overdose
• Severe encephalopathy
• History of end stage liver failure (based on presence of ≥ 1 or more of the following: AST/ALT ≥ 2 times ULN, INR ≥ 2, total bilirubin ≥ 1.5
• Patients with a baseline QTc interval >/= 500 msec or an elevation >/= 60mmHg above baseline
• Current drug therapy with a class Ia, Ic or III antiarrhythmic other than amiodarone.
• History of haloperidol allergy
• History of neuroleptic malignant syndrome.
• Patients expected to die within 24 hours
• Inability to obtain informed consent
• Current participation in another research study.
• Lack of permission of the patient's primary MICU service attending physician for participation in the study.
• Current employment at Tufts Medical Center.
• Age ≥ 85 years.