Risk factors for delirium and inpatient mortality with deliriumS Grover, D Ghormode, A Ghosh, A Avasthi, S Chakrabarti, SK Mattoo, S Malhotra
Department of Psychiatry, Postgraduate Institute of Medical Education and Research, Chandigarh, India
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.123147
Source of Support: None, Conflict of Interest: None
Background: There is limited on the risk factors and mortality in patients with delirium from India. Aim: This study aimed to evaluate the risk factors associated with delirium and inpatient mortality rates of patients diagnosed with delirium by psychiatry consultation liaison services. Materials and Methods: Three hundred and thirty-one patients diagnosed as delirium by the psychiatry consultation liaison services were examined on standardized instruments: Delirium Rating Scale Revised 98 version (DRS-R-98), amended Delirium Motor Symptom Scale (DMSS), Delirium Etiology Checklist (DEC), Charlson Comorbidity index, and a checklist for assessment of risk factors. Results: More than three medications as a risk factor and metabolic/endocrine disturbances as cause were observed to play largest role in development of delirium. The inpatient mortality rate was 12.4%. Compared to the survivor group, those who died were more likely to be young (<65 years), had significantly high rate of alcohol dependence and were more frequently restrained prior to development of delirium; of these only age <65 years and use of restraints emerged as the significant predictors of mortality in regression analysis. Conclusion: Age and use of restraints appears to be an important predictor mortality in patients with delirium.
Keywords: Delirium, mortality, predictors
Delirium is the most common psychiatric diagnosis in patient referred from various medical surgical wards to psychiatry consultation-liaison team.  It is associated with a wide range of risk factors and etiologies.  In general delirium is considered to be a reversible condition, but it is also reported to be associated with high mortality. Mortality rates have ranged from 9 to 34.5% during their inpatient stay. ,,,, It is important to note that the studies which have focused on the inpatient and follow-up mortality rates have been limited to elderly population or to those admitted to the intensive care units. ,,,,,, Occasional studies have evaluated the mortality rates of medical-surgical inpatients with delirium referred for psychiatric consultation. Tennen, et al.,  evaluated 1 year mortality rate of 454 patients admitted to various medical surgical inpatients referred for psychiatric consultation and reported a mortality rate of 15.2% within 1 year of referral for psychiatric consultation. Among the various psychiatric diagnosis, delirium was the only psychiatric diagnosis associated with higher mortality at 1 year (52.2 vs 29.9%, P=0.01; hazard ratio=1.7). In a recent study from our center, which included 97 patients with delirium, the inpatient mortality rate was 28% at 6 months post discharge. When an attempt was made to study the relationship of various sociodemographic and clinical variables and mortality in patients of delirium, none of the variables emerged as a significant predictor, possibly because of small sample size.  Although some recent studies from India have studied the phenomenology/symptom profile of patients with delirium seen in consultation-liaison psychiatry, ,, very few studies have evaluated the prevalence of risk factors for development of delirium.
In this background, the current study aimed to evaluate the risk factors and inpatient mortality rates of patients diagnosed with delirium by psychiatry consultation liaison services. An attempt was also made to study the predictors of mortality in patients with delirium.
The study was approved by the Institutional Review Board and written informed consent was obtained from caregivers.
The study was carried out a multispecialty teaching hospital in north India. It was prospective, assessing patients who were admitted to various medicosurgical and emergency wards and diagnosed to have delirium by the Consultation Liaison (CL) team were evaluated to confirm the diagnosis of delirium as per the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) criteria. 
The study involved assessment of referral pattern, phenomenology, motoric subtypes, risk factors, associated etiologies for delirium, outcome of delirium, experience of delirium of patients and the etiological models about symptoms of delirium held by the caregivers. In this paper only the data of risk factors and mortality is presented.
Assessment tools delirium rating scale-revised-98
It assesses the symptom frequency by considering as 'present' if any item rated as 1 or more. Of the 16 items, 13 comprise the severity scale and three are diagnostic items. The severity scale scores range from 0 to 39, higher scores indicate more severe delirium, and a cutoff score ≥15 is consistent with a diagnosis of delirium. It has good interrater reliability, validity, sensitivity, and specificity for distinguishing delirium in mixed neuropsychiatric populations including dementia, depression, and schizophrenia. ,
Amended delirium motor symptom scale
The DMSS classifies delirium into various motoric subtypes. Original DMSS consists of 11 items: Increased quantity of motor activity, loss of control of activity, restlessness, wandering (these items pertain to hyperactive subtype), decreased amount of activity, decreased speed of actions, reduced awareness of surroundings, decreased amount of speech, decreased speed of speech, listlessness, reduced alertness/withdrawal (these items pertain to hypoactive subtype). , A later analysis, based on which amended DMSS was developed suggested that DMSS consists of 13 items (five hyperactive and eight hypoactive items), with one item extra for each of the subtyping. The extra hyperactive item was 'increased speed of actions' and extra hypoactive item was 'decreased volume speech'.  Rating is done on the basis of definite evidence for behavior in the previous 24 h, which is a deviation from predelirious baseline. At least two out of five hyperactive items must be present, or at least two out of eight hypoactive items for the hyperactive and hypoactive subtypes to be diagnosed, respectively. The subtype is considered to be 'mixed' when there is concurrent evidence for both these subtypes and 'no motor subtype' when evidence for neither subtype is present. ,
Risk factors for delirium
A list for assessment of risk factors for delirium was developed for the present study and was based on a review of literature and covered common risk factors associated with delirium.
Delirium etiology checklist
It was used to standardize attribution of etiology on the basis of all available clinical information into 12 categories of etiological causation, each rated on a 5-point scale for degree of attribution to the delirium episode, ranging from "ruled out/not present/not relevant" (0) to "definite cause" (4) as follows: Drug intoxication/drug withdrawal; metabolic/endocrine disturbance; traumatic brain injury; seizures; infection (intracranial); infection (systemic); neoplasm (intracranial); neoplasm (systemic); cerebrovascular; organ insufficiency; other CNS; and other systemic causes. This system allows for the documentation of the presence and suspected role for multiple potential causes of delirium and provides more information specifically relevant to delirium than a listing of current medical conditions. 
Charlson comorbidity index
The Charlson Comorbidity index predicts the 10 year mortality for a patient who may have a range of comorbid conditions such as heart disease, acquired immunodeficiency disease or cancer (a total of 22 conditions). Each condition is assigned a score of 1, 2, 3, or 6 depending on the risk of dying associated with this condition. Then the scores are summed up and given a total score which predicts mortality. The clinical conditions and scores are as follow: 1 each for myocardial infarct, congestive heart failure, peripheral vascular disease, dementia, cerebrovascular disease, chronic lung disease, connective tissue disease, ulcer, and chronic liver disease; score of 2 each for hemiplegia, moderate or severe kidney disease, diabetes, diabetes with complication, tumor, leukemia, and lymphoma; 3 each for moderate or severe liver disease, and 6 each for malignant tumor, metastasis, and acquired immunodeficiency syndrome (AIDS). 
From 1 st January 2012 to 15 th July 2012, caregivers of all the patients diagnosed with delirium by the CL team were approached for participation in the study and explained about the nature of the study. The consenting carers were interviewed to confirm the patients' diagnosis of delirium as per DSM-IV TR criteria. Then the required information was obtained from family members, treating team, patients, and treatment records to complete the clinical profile, DRS-R-98, amended DMSS, DEC, risk factors, and Charlson Comorbidity index. Lastly, to study the clinical outcome the patients were followed-up regularly till they expired or were discharged from the hospital.
The data were analyzed using the Statistical Package for Social Sciences, version 14 (SPSS 14). Descriptive analysis was carried out in the form of frequency, percentages, means, and standard deviations. Comparisons were done by using Chi-square, Fisher exact test, and Mann-Whitney U test. Binary logistic regression analysis was done to study the predictors of mortality. All analyses were done at 5% significance.
A total of 331 consecutive patients of delirium were studies. Of these, 41/331 died during the inpatient stay, giving an inpatient mortality rate of 12.4%. In most other patients delirium improved (N=266; 80.36%) by the time of discharge and in few patients (N=24; 16.3%) symptoms of delirium persisted.
The sociodemographic profile of the whole sample, those who discharged from the hospital (after recovery/improvement with respect to delirium or worsening of delirium) and those died during the hospital stay is shown in [Table 1].
The mean age of the study sample was 46.2 years and 264 (79.7%) patients were aged less than 65 years of age. About two-third of the patients were males and the mean duration of education was 9.2 years at the time of assessment. Except for the fact that those who died had significantly higher proportion of non-elderly patients, there were no significant differences on any of the other sociodemographic variables [Table 1].
The mean duration of delirium at the time of first assessment was 2.27±2.2 days. Most of the patients (77.6%) had hospital emergent delirium, were not treated with any antipsychotic medications or benzodiazepine (89.1%) and the mean number of medications received for all ailments at the time of baseline assessment were 0.9±1.4. There was no significant difference between those died during the hospital stay and those who survived during the hospital stay on any of these variables. Majority of the patients were admitted under the medical services of the hospital. Although mortality was higher in those admitted in the surgical setups, but the difference between the two groups was not significant.
On DRS-R-98 diagnostic items, all the patients fulfilled the diagnostic criteria of 'presence of an underlying physical disorder' and majority of them fulfilled the 'temporal onset of symptoms (92.7%)' and 'fluctuation of symptoms (97%)' criteria. In the whole sample, >80% of the patients had disturbances in sleep-wake cycle, lability of affect, motor agitation, attention, orientation, and short-term memory impairments. The least common symptoms were motor retardation, delusions, and disturbance in long-term memory [Table 2]. The two groups did not differ with respect to frequency of symptoms as assessed on DRS-R-98.
As shown in [Table 2], the two groups were similar for the severity of DRS-R-98 item scores. There was no significant difference with respect to the total DRS-R-98 score, DRS-R-98 severity score, DRS-R-98 noncognitive domain score, and DRS-R-98 cognitive domain score.
Motoric subtypes of delirium as per amended DMSS
As shown in [Table 2], more than half (58.9%) of the patients had hyperactive delirium. This was followed by mixed subtype and hypoactive subtype. As there were very few patients in the nonhyperactive group, all the three nonhyperactive groups were combined together for comparison. On comparison no significant difference was found between those who expired and those who survived (Chi-square test=1.71; P=0.18).
Risk factors for delirium
Factors which were present prior to onset of delirium were considered to be risk factors predisposing a person to develop delirium. Based on the review of literature a specific risk factor list was prepared. As is evident from [Table 3], the most common risk factors which were present in our sample were use of more than three medications (58.3%). It was followed by increased urea level (31.7%), anemia (28.7%), increased creatinine level (24.4%), immobility prior to delirium (27.8%), hyponatremia (23.2%), smoking (23%), acute infection (21.1%), and admitted to intensive care unit (ICU) (20.2%), etc. When the risk factors were compared between those who expired and those who survived, significant difference was found on the variables of patients with age >65 years (more frequent in the survivor group), history of alcohol dependence (more frequent in those who expired), and use of restraints (more frequent in those who expired).
Etiologies associated with delirium
According to DEC each etiology is rated on a 5-point scale for degree of attribution to the delirium episode, ranging from "ruled out/not present/not relevant" (0) to "definite cause" (4). For analysis the ratings of definite cause, likely cause, and cause present and possibly contributory were combined and considered as present and the rating of "present but not contributory" and "ruled out/not present/not relevant" were considered as absent. As is evident from [Table 4], metabolic and endocrine disturbances were most common abnormalities (77%) followed by organ insufficiency (24.8%), systemic infection (20.5%), and drug withdrawal (12%).
Charlson's comorbidity index
Charlson's comorbidity index predicts the 10-year mortality for a patient with a range of medical illnesses (a total of 22 conditions). The weighted index of comorbidity in the study sample was 1.4 and the combined condition and age adjusted score was 2.4 and the estimated 10 year survival of the study sample was 73%. As shown in [Table 5], there was no significant difference between those who expired and those who survived with respect to any of these parameters.
Predictors of mortality
To assess the predictors of mortality, binary logistic regression analysis was done. For the same outcome variable, that is, expired or survived by the time of discharge was used as the dependent variable and all other variables, for which the significance value (P-value) was less than 0.1 were used as independent variables. As is evident from [Table 6], the significant predictors of mortality with odds ratio of more than 1 was age less than 65 years of age and use of restraints.
The present study was an attempt to evaluate the risk factors associated with delirium and the factors associated with mortality in patients admitted to general hospital inpatient unit. Patients were assessed on standardized instruments to assess the phenomenology, motoric subtypes, associated etiologies, and the risk factors for developing delirium. Although previous studies from India have evaluated the phenomenology, the data is limited with respect to the prevalence of various risk factors, associated etiologies, and predictors of mortality from India. This is possibly the largest sample size study from India of patients with delirium. Although sample size calculation was not done, the sample size can be considered to be reasonable enough to draw firm conclusions.
The demography of our sample was similar to the previous studies from our center. ,,, Because the symptoms of delirium tend to be most florid immediately after presentation, assessments were carried out as early as possible. Resultantly, the mean duration of delirium at assessment of 3.27 days as seen in the previous studies from our center. ,, Delirium developing after hospitalization in 78.4% of our patients was higher than in the previous studies from different centers/countries, ,, but similar to studies from our center. ,
The phenomenology of delirium as noted in the present study is very similar to that reported in previous studies. ,, Predominance of patients with hyperactivity suggests that patients with behavioral disturbance, especially, those with agitation are most often referred, as noted in previous studies from our center ,, and other parts of world, which suggests that hyperactivity is a common clinical presentation in CL referral versus elderly medical, palliative care, and ICU populations. 
The most common etiologies associated with delirium in the present study were metabolic/endocrine disturbances, organ insufficiency, and systemic infection is in keeping with the existing literature. ,
Most of the previous studies from India have not evaluated the risk factors for delirium. One study  evaluated the risk factors for delirium in elderly and reported preexisting cognitive deficits, neurological illnesses, urinary tract infections, visual impairment, hearing impairment, current proteinuria, leukocytosis, raised blood ammonia, hyponatremia, and disturbances in the potassium levels to be associated with development of delirium. In contrast, in the present study we found that more than half of the patients with delirium were receiving more than three medications, and another one-fifth to one-third of the patients were having higher urea/creatinine level, had anemia, were immobile prior to onset of delirium, were smoker, were admitted to intensive care unit, had history of alcohol dependence, use of restraints, hyponatremia, acute infection or receiving nonsteroidal anti-inflammatory drugs (NSAIDS), etc., When we compare the findings of the present study with most of the existing literature from developed countries, the risk factors appear to be similar. 
In the present study, the inpatient mortality was 12.4%, which was similar to that in the previous studies from the West (9-34.5%) ,, and from India.  However, it is important to note that most of the mortality seen in the present study can be considered to be limited to patients with hyperactive delirium, as more than 80% of the patients in the present study either had hyperactive or mixed subtype of delirium. Studies do suggest that a large proportion of delirium (32-67%) remains unrecognized in general medical wards  and only a minor proportion of inpatients with delirium (10%) are referred to psychiatry consultation-liaison services. ,, It is important to remember that in general it is suggested that mortality is higher in patients with hypoactive delirium.  If these facts are taken into account, the actual mortality rates in patients with delirium may be higher, as many patients with same would not have been evaluated by psychiatry CL team. Hence, it can be said that the mortality rate in the present study may be an underestimate.
It is important to note that most of the previous studies which have evaluated the mortality rates in patients with delirium have done so in elderly and those admitted to ICUs. In contrast, the present study predominantly had young patients (79.7%) and further it was noted that young patients had higher mortality rates than elderly despite having nearly equal Charlson Comorbidity index. This finding suggests that delirium may be a risk factor for mortality irrespective of the age of the patients. Similar finding was noted in the previous study from our center too. 
Following limitations must be kept in mind while interpreting results of the present study. Studies have shown higher mortality rates in patients with delirium during the first 1-2 years after the diagnosis. However, we did not follow-up the cohort after discharge. We also did not study the influence of treatment on the outcome of delirium.
To conclude, present study suggests that the common risk factors for development of delirium include use of more than three medications, higher urea/creatinine level, anemia, immobility prior to onset of delirium, smoking, being admitted to intensive care unit, history of alcohol dependence, use of restraints, hyponatremia, acute infection and use of NSAIDS. The most common etiologies associated with delirium include metabolic/endocrine disturbances, organ insufficiency and systemic infection. About one-sixth of the patients with delirium die during the initial inpatient stay. Mortality is higher in patients who are young and who are restrainted.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]