Understanding the Barriers in Delirium Care in an Intensive Care Unit: A Survey of Knowledge, Attitudes, and Current Practices among Medical Professionals Working in Intensive Care Units in Teaching Hospitals of Central Province, Sri Lanka.

BACKGROUND: Delirium is a common, underdiagnosed, and undertreated condition that increases morbidity and mortality in ICU patients which has an incidence up to 80%. Barriers that hinder optimum care of delirium include inadequate knowledge, poor attitudes, and low perceived importance of delirium care. AIM: To assess attitudes, knowledge, and current practices related to delirium care among medical professionals working in intensive care units (ICUs) in all teaching hospitals in Central Province, Sri Lanka, as there are no Sri Lankan studies on this regard.
METHOD: A descriptive cross-sectional study was carried out among all medical professionals working in nine ICUs in all (n = 5) teaching hospitals in Central Province. Data were collected using a pretested self-administered questionnaire. Responses to questions were compared between postgraduate trainee medical officers (PG-MOs) and non-postgraduate-trainee medical officers (non-PG-MOs).
RESULTS: Eighty-eight questionnaires were analyzed. More than 80% of PGs and non-PG-MOs regarded ICU delirium as significant problem that should be screened and prevented. Forty-one percent stated confidence in diagnosing delirium. However, more than 75% of non-PG-MOs failed to recognize features of hypoactive delirium. Only 30-50% subjects in incorporated preventive methods in usual practice and more than 60% non-PG-MOs had poor knowledge and experience on delirium screening. More than 80% of the participants did not routinely screen their patients. More than 90% non-PG-MOs (p <0.05) had no recent educational exposure.
CONCLUSION: A positive attitude toward the importance of management of delirium was observed. However, there is a discrepancy between the perceived importance and the current practice related to screening and prevention. Participants, especially non-PG-MOs, lacked knowledge on delirium screening, diagnosis, and identification of risk factors, probably related to a lack of educational exposure. HOW TO CITE THIS ARTICLE: Arachchi TMJ, Pinto V. Understanding the Barriers in Delirium Care in an Intensive Care Unit: A Survey of Knowledge, Attitudes, and Current Practices among Medical Professionals Working in Intensive Care Units in Teaching Hospitals of Central Province, Sri Lanka. Indian J Crit Care Med 2021;25(12):1413-1420.

Introduction

Delirium is defined by the American Psychiatric Association, as a disturbance of consciousness, attention, cognition, and perception which develops over a short period of time that tends to fluctuate during the course of the day. It is associated with increased mortality, morbidity, persistent neurocognitive impairment, increased cost, and prolonged hospital stay.[1-14]

The incidence of ICU delirium is estimated to range from 20 to 80% among critically ill patients.[1,4] The only Sri Lankan study that was published by Lochanie and Ranawaka revealed an incidence of 66% among postoperative ventilated patients in a surgical ICU at Colombo National Hospital, Sri Lanka.[7]

However, delirium is widely underdiagnosed and undertreated with 65–75% of cases going undetected.[3] Poor attitudes on delirium care, lack of knowledge on diagnosis and screening tools, communication challenges, time restraints, and workload concerns among professionals working in ICUs are among the identified barriers that are attributed to poor detection and management.[2]

Though multiple studies are conducted worldwide to identify these barriers, to our knowledge, this is the first study that was conducted among Sri Lankan medical professionals working in ICUs. Thus, this study was designed to assess the knowledge, attitudes, and current practices on delirium diagnosis and management in ICU patients among medical professionals working in intensive care units in teaching hospitals in Central Province, Sri Lanka.

Method

A descriptive cross-sectional study was carried out among all the medical professionals (possessing at least an MBBS degree) working in nine intensive care units in all the (n = 5) teaching hospitals in Central Province, Sri Lanka (Teaching Hospital Peradeniya, Dental Hospital Peradeniya, National Hospital Kandy, Sirimavo Bandaranayake Children's Hospital, and Teaching Hospital Gampola).

Ethical clearance was obtained from the Ethical Review Committee of the Faculty of Medicine, University of Peradeniya (ERC No: 2020-EC-14).

Data were collected using a pretested self-administered questionnaire from July 1, 2020 to September 1, 2020 after obtaining informed consent from the subjects. A thorough literature survey was undertaken to identify the key issues related to ICU delirium (knowledge, attitudes, and current practices related to diagnosis, screening, impact on the patient and the society, prevention, and education) for the development of the questionnaire. The questionnaire was consisted of several close-ended questions, and it covered the following areas: demographic data, knowledge on clinical features of hypoactive, hyperactive and mixed-type delirium, risk factors and risk categories, adverse effects, commonly practiced preventive methods, recommended screening tools (Confusion Assessment Method for ICU [CAM-ICU] tool) for delirium, diagnosis of delirium (Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) criteria], attitudes toward management, screening, and current practices on diagnosis and prevention.

Data were analyzed with IBM SPSS version 25. Results were presented as frequencies and percentages. Significance between the responses of PGs and non-PG-MOs was analyzed using Pearson's Chi-squared test and Fisher's exact test when indicated. p values less than 0.05 were taken as significant.

Results

A total of 123 questionnaires were distributed. Ninety-four have responded to the study. Six questionnaires were discarded due to >10% incomplete answers. A total of 88 questionnaires were analyzed, and the response rate for the questionnaire was 76%.

The age of the population ranged between 29 and 60 years, with a mean age of 38.7 years, and 45 (51%) subjects were females.

Of the total, 29 (33%) were PGs with experience in anesthesiology, critical care, and emergency medicine [senior registrars n = 10 (11%), registrars n = 19 (22%)]. Non-PG-MOs comprised 67% (n = 59) of the population.

Sixty-nine percent of the population had more than 1 year of working experience in an ICU setting, while 31% revealed that they had <1 year experience.

Attitudes Related to Delirium Care

Overall, the participants had a positive attitude toward delirium care in critically ill patients. It is encouraging that the majority (80%) regarded ICU delirium as a significant problem that should be screened and prevented (Table 1).

Table 1

Attitude toward ICU delirium management in ICU

	Response (n = 88)		PGs (n =)		non-PG-MOs (n =)
Question	Yes	No	Yes	No	Yes	No	p value
Recognition of delirium as a common occurrence	65 (74%)	13 (15%)	27	2	38	23	0.002
Incidence estimate of ICU delirium
• <10%	24 (27%)		3		13
• >10%	64 (73%)		26		46
Self-reported confidence in diagnosis of delirium	36 (41%)	52 (59%)	13	16	23	36	0.769
Delirium screening is important	83 (94%)	5 (6%)	28	1	55	4	0.999
ICU delirium is a major problem	70 (80%)	18 (20%)	27	2	43	16	0.046
ICU delirium is preventable	72 (82%)	16 (18%)	27	2	45	14	0.046
It is important to prevent ICU delirium	87 (99%)	1 (1%)	28	1	59	0	–
Recognize delirium as a predictor of increased mortality	52 (59%)	36 (41%)	23	6	29	30	0.01
Recognize delirium as a cause for increased morbidity	72 (82%)	16 (18%)	27	2	51	8	0.486
Recognize delirium as a cause for persistent cognitive impairment	49 (56%)	39 (44%)	23	6	26	33	0.002
Recognize delirium as an unpleasant occurrence to the patient and family	87 (99%)	1 (1%)	29	0	58	1	–
Recognize that delirium increases length of hospital stay and health care cost	86 (98%)	2 (2%)	29	0	57	2	–

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered as significant

Knowledge on ICU Delirium Diagnosis and Screening

Components of DSM-5 criteria that included cardinal features of delirium (inattention, disordered cognition, acute onset of symptoms with fluctuating course) were recognized as diagnostic criteria by a majority (>65%) of participants, with a significantly high proportion of PGs identifying the criteria. The most striking feature noted is that only two (2.2%) PGs were able to correctly identify all the diagnostic criteria (Table 2).

Table 2

Responses to knowledge-based questions on diagnostic criteria according to DSM-5

	Total responses (n = 88)			PGs (n = 29)		Non-PG-MOs (n = 59)
Diagnostic criterion	Yes	No	Do not know	Yes	No/do not know	Yes	No/do not know	p value
Onset is acute in nature	59 (67%)	15 (17%)	14 (16%)	26	3	33	26	0.001
Having a fluctuating course of severity of symptoms during a day	61 (69%)	2 (2%)	25 (28%)	25	4	36	23	0.015
Presence of inattention	65 (74%)	3 (3%)	20 (23%)	28	1	37	22	0.001
Disordered cognition	60 (68%)	5 (6%)	23 (26%)	25	4	35	24	0.014
Absence of premorbid neurocognitive disorder	44 (50%)	17 (19%)	27 (31%)	14	15	30	29	0.999
Absence of evidence of other systemic or other underlying cerebral disease or substance that explains the disturbance	55 (62%)	13 (15%)	20 (23%)	21	8	34	25	0.242

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered as significant

Knowledge on Subtypes of Delirium and Clinical Features

Majority (>85%) had recognized features of hyperactive delirium. However, 55–60% of the population, especially a significant proportion of non-PG-MOs (76–82%), failed to recognize features of hypoactive delirium (Table 3).

Table 3

Awareness on clinical features and subtypes of delirium

	Total responses (n = 88)			PGs		Non-PGs
Clinical feature	Yes	No	Do not know	Yes	No/do not know	Yes	No/do not know	p value
Altered conscious level	64 (73%)	10 (11%)	14 (16%)	25	4	39	20	0.073
Inattention	79 (90%)	1 (1%)	8 (9%)	28	1	51	8	0.261
Disorientation	80 (91%)	–	8 (9%)	28	1	52	7	0.262
Agitation	84 (94%)	2 (2%)	3 (3%)	28	1	55	4	0.999
Restlessness	80 (91%)	2 (2%)	6 (7%)	26	3	54	5	0.999
Reduced alertness	41 (46%)	17 (19%)	30 (34%)	28	1	13	46	0.0001
Reduced speech	36 (43%)	17 (19%)	35 (40%)	22	7	14	45	0.0001
Reduced motor activity	35 (40%)	22 (25%)	31 (35%)	24	5	11	48	0.0001
Hallucinations	72 (82%)	8 (9%)	8 (9%)	24	5	48	11	0.0001
Altered sleep/wake cycles	48 (55%)	15 (17%)	25 (28%)	18	11	30	29	0.320

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered as significant

Knowledge on Risk Factors

Interestingly, majority of common risk factors were identified by a greater proportion of the population (Table 4).

Table 4

Awareness on risk factors of ICU delirium

	Total responses (n = 88) (%)			PGs (n = 29)		Non-PGs (n = 59)
Risk factor	Yes	No	Do not know	Yes	No/do not know	Yes	No/do not know	p value
Severe illness	74 (84%)	4 (5%)	10 (11%)	27	2	47	12	0.130
Mechanical ventilation	38 (43%)	18 (20%)	32 (37%)	22	7	28	30	0.020
Increased age	78 (89%)	5 (6%)	5 (6%)	26	3	52	7	0.999
Preexisting cognitive impairment	62 (71%)	7 (8%)	19 (23%)	24	5	38	21	0.087
Metabolic acidosis and electrolyte imbalances	69 (78%)	4 (5%)	15 (17%)	24	5	45	14	0.588
Sedatives—benzodiazepines	73 (83%)	5 (6)%	10 (11%)	27	2	46	13	0.129
Opioids	73 (83%)	4 (5%)	11 (12%)	26	3	47	12	0.367
Preexisting hypertension	25 (28%)	26 (30%)	36 (42%)	13	16	12	37	0.000
Dehydration	67 (76%)	2 (2)%	19 (22%)	22	7	45	14	0.966
Alcohol abuse	78 (89%)	1 (1%)	9 (10%)	27	2	51	8	0.486
Increased noise in ICU	62 (70%)	5 (6%)	21 (24%)	25	4	37	22	0.026
Infections/sepsis	78 (89%)	3 (3%)	7 (8%)	27	2	51	8	0.486
Untreated pain	72 (82%)	7 8%)	9 (10%)	25	4	47	12	0.071

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered as significant

Knowledge and Current Practice-related Screening of ICU Delirium

Significant proportion of non-PG-MOs had inadequate awareness on availability and experience in using CAM-ICU screening tool to screen ICU patients for delirium. Majority of the population did not adhere to the current recommendations on routine delirium screening (Table 5).

Table 5

Current practices related to screening of ICU delirium

	Total responses (n = 88)		PGs (n = 29)		Non-PG-MOs (n = 59)
	Yes	No	Yes	No	Yes	No	p value
Aware about the availability of screening tools	34 (39%)	54 (61%)	20 (69%)	9 (31%)	14 (24%)	45 (76%)	0.0005
Had ever screened a patient for delirium	24 (27%)	64 (73%)	13 (45%)	16 (55%)	11 (19%)	48 (81%)	0.009
Has experience in using CAM-ICU tool	23 (26%)	65 (74%)	18 (62%)	11 (38%)	5 (8%)	54 (92%)	0.0001
Routinely screen patients for delirium	11 (13%)	77 (87%)	5 (17%)	24 (83%)	6 (11%)	53 (89%)	0.493

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered as significant

Knowledge and Current Practice-related Preventive Methods (Tables 6 and 7)

It is remarkable that more than 80% were aware, had good knowledge, about preventive methods. However, the routine practice on incorporation of preventive methods in patient management is inadequate.

Education Related to ICU Delirium (Table 8)

There is a significant unmet need for education among non-PG-MOs (p <0.05) compared to PGs.

Discussion

Remarkably, vast majority of the study population had a positive attitude toward delirium care in ICU, in terms of recognizing delirium as an important factor that increases morbidity, mortality, and health care cost in ICU patients, that needs to be diagnosed and prevented. These findings are in line with the research findings of Davis and Maclullich, and Troglic et al.[2,13]

However, only a quarter of the population had ever screened a patient for delirium while 88% admitted that they do not routinely screen patients for delirium, revealing a disparity between the perceived importance and the current practice. Similar results were observed in studies conducted by Davis and Maclullich, Elliott, Xing et al., and Bellelli et al.[13,15-17]

Responses to the knowledge-based questions did indeed suggest that medical professionals who participated in the study had poor knowledge of the DSM-5 diagnostic criteria, CAM-ICU screening tools, and clinical features of hypoactive delirium, especially non-PG-MOs.

Although only 26% of medical professionals had experience in using CAM-ICU, 41% reported that they are confident in diagnosing delirium in ICU patients. Thus, it may be assumed that these professionals may have screened patients based on their clinical judgment. A study conducted by Devlin et al. suggests that respondents who use a nonstructured means rely primarily on features of hyperactive delirium, which is the least common subtype of delirium, thus missing the cases of hypoactive delirium, the common type, and missing around two-thirds of cases if a validated screening tool is not used.[15] Consequently, it may be hypothesized that many cases in our ICU setup are also undetected and undertreated. This highlights the need for new strategies to improve screening, such as incorporation of daily delirium screening into the ICU monitoring chart and motivation of medical staff to adhere into standard practice.

More than two-thirds of the participants recognized the cardinal features of delirium, inattention (74%), disordered cognition (68%), and acute onset of symptoms with a fluctuating course of severity during a day as essential criteria to diagnose (69%). These findings are in contrast to the findings of Davis and Maclullich, which revealed only 32% recognized inattention as a core diagnostic feature, while 67% regarded acute onset of symptoms as essential.[13] The results observed in our study may suggest that the participants possess adequate knowledge. However, the most striking feature observed was that only two (2%) respondents were able to correctly identify all the essential criteria of DSM-5. This suggests that although the respondents have some idea about delirium, they lack precise knowledge regarding the diagnosis of delirium, with results paralleling the findings of multiple studies.[13,16,17] Furthermore, half (n = 40) of the population regarded that premorbid neurocognitive disorders should be absent to diagnose delirium, thus increasing the possibility of missing cases of delirium superimposed on other neurocognitive disorders. Only 15% correctly responded that the presence of evidence from history, physical examination, or laboratory findings that the disturbance is a direct consequence of another medical condition, substance intoxication, or withdrawal is a criterion to diagnose delirium. This again heightens the possibility of missing delirium diagnosis, rather it would be diagnosed in the context of the somatic disorder; for instance, a delirium triggered by hyponatremia will be diagnosed as a case of hyponatremia, with a reluctance to diagnose delirium. Diagnosis of delirium as delirium is important, as it is preventable and treatable with multiple, simple, and yet-effective nonpharmacological strategies.[4]

The highest gap in knowledge (between PGs and non-PG-MOs p <0.05) was observed concerning the recognition of clinical features of hypoactive delirium. Nearly 75% of non-PG-MOs failed to identify key features of hypoactive subtype,[4] though 66% were aware that there are different subtypes of delirium. This may be related to the fact that compared to non-PG-MOs, a significant proportion of PGs have an educational exposure to delirium in the form of lectures or discussions during their training. This lack of continuous education in non-PG-MOs needs to be recognized, and measures should be taken to minimize this issue.

Generally, the mixed subtype accounts for 53%, hypoactive for 35%, and hyperactive for 11%. However, it is notable that the only study conducted in Sri Lanka regarding the incidence of delirium revealed that the majority (50%) belonged to the hypoactive subtype.[7] Therefore, it is highly likely that these cases of hypoactive and mixed subtypes, that comprise the majority, will go unnoticed and may be diagnosed as depression if this knowledge gap is not addressed accordingly. Underdiagnosis or failure to recognize features of hypoactive delirium is a common problem identified in many circumstances.[4] In such situations, it is commonly mistaken for depression.[4]

Interestingly, majority of participants (>75%) were aware of the common risk factors that trigger delirium in ICU patients. This is encouraging as the awareness of risk factors improves the prediction of high-risk cases in turn aiding early detection and treatment and motivating the treating team to implement preventive strategies early in the course.

In contrast to a study conducted by Zoran et al. where only 20% regarded delirium as preventable, 82% of our participants believed delirium is preventable. Astoundingly, more than 75% of our participants were aware of most of the preventive methods of ICU delirium. Almost all (99%) agreed that it is important to prevent delirium. Regrettably, this recognized importance was not reflected in their practice, without a remarkable difference between PGs and non-PG-MOs. The exact reason or the barriers for the inadequate incorporation of preventive strategies in managing critically ill patients are unclear, but the possibility of lack of knowledge on screening and diagnosis should not be disregarded as a cause for this occurrence.

Despite recent guidelines on delirium management and the recognition ICU delirium has received, aboult three-quarters of the population (significantly a higher proportion of non-PG-MOs) claimed that they have not received any training or lectures delirium management during the recent years, and this was in line with the results of a study conducted by Devlin et al. where 50% of the respondents had never received any education.[15,18] However, it is noteworthy that about 43% of our subjects, majority of PGs, attempted to refresh or gain knowledge through self-learning. This may be an important finding that indicates their positive attitudes toward delirium care and acquisition of knowledge. This gap in knowledge and unmet need for education needs to be recognized by the trainers and policymakers to improve and expand the scope of education and training related to delirium care such that it occupies a position proportionate to its clinical impact in Sri Lanka.

In summary, it is evident that the lack of appreciation of the significance of delirium and poor attitudes does not appear to be fundamental barriers to delirium management in relation to underdetection and undertreatment, among medical professionals participated in this study. Rather, lack of knowledge and inadequate current practice pertaining to screening modalities, diagnostic criteria, and preventive modalities appear to be more important, especially among non-PG-MOs.

One of the strengths in our study was the high response rate (76.4%), probably because the questionnaires were distributed and collected via paper and pen on the spot. However, this may have interfered with the in-depth thinking of participants which may have caused a response bias.

The other limitations are that this survey had included the response bias caused by the poor recollection of clinical experiences, or misunderstanding of questions as these were self-reporting responses. Besides, this survey only has included ICUs in teaching hospitals in Central Province (five hospitals) in Sri Lanka, due to time restraints and limited human resources for data collection. Therefore, it may affect the generalizability of the results of this study.

Conclusion

There is an encouraging positive attitude toward the importance, prevention, and management of delirium among all the doctors in the study population.

There is a lack of knowledge on delirium screening and diagnosis, probably related to an unmet need of education related to delirium care, especially among non-PG-MOs.

A discrepancy was observed between perceived importance, awareness on risk factors and preventive methods, and current practice related to screening and prevention among both PGs and non-PG-MOs.

Orcid

Tilani M Jayasinghe Arachchi https://orcid.org/0000-0002-4543-1918

Vasanthi Pinto https://orcid.org/0000-0001-6938-248X

Table 6

Knowledge on preventive methods

	Total responses (n = 88)			PGs		Non-PGs
Preventive method	Yes [n (%)]	No [n (%)]	Do not know [n (%)]	Yes	No/do not know	Yes	No/do not know	p value
Early mobilization	84 (96%)	–	4 (4%)	28	1	56	3	0.999
Reorientation	78 (89%)	2 (2%)	7 (8%)	27	2	50	7	0.711
Cognitive stimulation	74 (84%)	3 (3%)	10 (11%)	27	2	48	11	0.205
Reduction of nighttime light	73 (83%)	5 (6%)	10 (11%)	27	2	46	13	0.129
Reduction of nighttime noise	69 (78%)	5 (6%)	14 (16%)	25	4	42	17	0.183
Use of hearing aids and glasses when required	56 (64%)	7 (8%)	25 (28%)	22	7	34	25	0.105
Allowing frequent family visits	71 (81%)	8 (9%)	9 (10%)	25	4	46	13	0.406
Reduction of sedation to promote daytime wakefulness	79 (90%)	–	9 (10%)	28	1	51	8	0.261
Prevention of sleep deprivation	76 (86%)	2 (2%)	10 (11%)	28	1	48	11	0.094

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered as significant

Table 7

Current practices related to preventive methods

	Total responses (n = 88)		PGs (n = 29)		Non-PG-MOs (n = 59)
Practice	Routinely done	Not routinely done	Routinely done	Not routinely done	Routinely done	Not routinely done	p value
Early mobilization	56 (63%)	32 (37%)	20	9	36	23	0.491
Reorientation, day/night orientation	43 (49%)	45 (51%)	17	12	24	35	0.172
Cognitive stimulation, use of music and a clock	27 (31%)	61 (69%)	11	18	16	43	0.332
Reduction of nighttime light	29 (33%)	59 (67%)	10	19	19	40	0.815
Reduction of nighttime noise	17 (19%)	71 (81%)	7	22	10	49	0.566
Use of hearing aids and glasses if required	18 (21%)	70 (79%)	7	22	11	48	0.581
Allowing frequent family visits	30 (34%)	58 (66%)	14	15	16	43	0.058
Reduction of sedation to promote daytime wakefulness	44 (50%)	44 (50%)	16	13	31	28	0.999
Prevention of sleep deprivation	42 (48%)	46 (52%)	15	14	27	32	0.653

Table 8

Educational exposure related to ICU delirium

	Total responses (n = 88)		PGs (n = 29)		Non-PG-MOs (n = 59)
	Yes	No	Yes	No	Yes	No	p value
Had lectures/training programs in recent years	23	65	17	12	6	53	0.000
Read journals/books on delirium	38	50	24	5	14	45	0.000

p value was calculated using Pearson's Chi-squared test and Fisher's exact test when appropriate. p <0.05 was considered significant