Limitations to the identification of HIV-associated neurocognitive disorders in clinical practice

Authors


Correspondence: Dr Deirdre Catherine Morley, Department of Infectious Disease, Cork University Hospital, Wilton Road, Cork, Cork, Cork, Ireland. Tel: 00353860639574; fax: 00353214921343; e-mail: deemorley@gmail.com

Abstract

Objectives

The objective of this study was to establish the level of awareness of HAND among healthcare providers, the screening tools that are currently used in its detection and factors that limit cognitive assessments.

Methods

We distributed a 12-item questionnaire to doctors and nurses who work in the Department of Genitourinary Medicine and Infectious Disease (GUIDE) service and also to doctors who work in the emergency department (ED) at St James Hospital.

Results

35 surveys were collected, 54% (n = 19) from the GUIDE service and 46% (n = 16) from the ED. 82% (n = 29) of participants were doctors from interns to consultants. There was reasonable appreciation among participants with regards the prevalence of neurocognitive impairment (estimated at 29.1% among patients on HAART, and 39.3% among patients not on HAART). Screening tools were rarely used by GUIDE and ED clinicians (25% vs. 15% of the time). The Mini Mental State Examination (MMSE) was previously used by 37% (n = 13) of the group. Very few people had used the HIV Dementia Scale (HIVDS) 6% (n = 2). 34% of respondents felt that ‘Orientation in Person, Place and Time was a sufficient screening tool for cognitive assessment’. Lack of time, exposed environment and lack of availability of screening tool were cited as limitations to cognitive screening in the ED environment.

Conclusions

This study examines awareness of HAND among healthcare providers and also reasons for inadequate assessment. There is a need for consensus on screening guidelines. A quick, easy to use and readily available screening tool may have a role in the acute setting in identifying high-risk patients.

Introduction

Neurocognitive impairment has a prevalence rate of between 20 and 50% in the HIV-positive population despite highly active antiretroviral therapy (HAART) [1-3]. Rates are high in certain population groups, including patients over 50 years of age and those with opioid addiction [4, 5].

Classification of HIV-associated neurocognitive disorders (HANDs) in the post-HAART era was updated in 2007 to include three clinical syndromes: asymptomatic neurocognitive impairment (ANI), mild neurocognitive disorder (MNCD) and HIV-associated dementia (HAD) [1]. The CNS HIV Anti-Retroviral Therapy Effects Research (CHARTER) group have demonstrated rates of 33% for ANI, 12% for MNCD and 2% for HAD among the HIV-positive population [3]. Risk factors for the development of HAND include older age, female gender, more advanced HIV disease (including CD4 count < 100 cells/μL), high plasma HIV RNA (viral load), comorbid conditions (especially hepatitis C), and a history of injecting drug abuse [1, 6, 7]. The presence of neurocognitive impairment at HAART initiation has been shown to be a strong predictor of persistent neuropsychological deficit [8, 9].HIV-positive patients with neurocognitive impairment are at risk of a shorter life expectancy, a decreased adherence to HAART and poor employability [3, 10, 11].

The objective of our study was to establish the level of awareness of HAND among health care providers, the screening tools that are currently used in its detection and factors that limit cognitive assessment in the acute setting.

Methods

Study site

Our study site was St James Hospital, Dublin. St James Hospital is one of Ireland's largest teaching hospitals, with approximately 1020 beds [12]. The Department of Genitourinary Medicine and Infectious Disease (GUIDE) has a dedicated HIV in-patient team and the largest HIV out-patient service in Ireland. Approximately 46 000 people attend the Emergency Department (ED) at St James Hospital annually [12]. Of these, a significant number are HIV-positive patients requiring acute medical attention. Over a 1-year period (2009−2010), 1745 HIV-positive patients attended St James Hospital (ED or other out-patient departments) at least once [13].

Training for doctors in an Irish hospital includes a 1-year medical/surgical internship, 2−3 years as a senior house officer (SHO) within various specialties, 5−8 years as a registrar/specialist registrar within a chosen specialty and finally consultancy. On-call duties in the ED commence at SHO level.

Questionnaire

We distributed a 12-item questionnaire to doctors and nurses who work in the GUIDE service and also to doctors who work in the ED. SHOs, registrars and consultants working in the GUIDE department at the time of distribution took part in this study. They would typically encounter HIV-positive patients through out-patient clinics and in-patient ward rounds, and occasionally when patients present acutely to the ED. Interns, SHOs, registrars and consultants who worked in the ED also took part in the survey. They would assess HIV-positive patients who present acutely to the ED only. Nursing staff who took part in the study worked specifically within the GUIDE service and would see patients in the out-patient and/or in-patient setting.

The responses were unaided. Data establishing the specialty, clinical experience, and level of exposure to HIV-infected patients were collected. The aim of the questionnaire was to assess the knowledge, skills and attitudes of the group to HANDs.

Knowledge

Participants were asked to predict the prevalence of HAND in patients who are on HAART, and in those who are not on HAART. They were then required to highlight how various factors impact on mortality, morbidity and chance of readmission through the use of numbered Likert scales with scores from 0 to 10, where 10 indicates the highest level of impact. These factors included cognitive impairment, opportunistic infections, lower CD4 count, substance misuse, medication adherence, female gender, time elapsed since diagnosis of HIV infection and ethnicity.

Skills

Participants were asked if an assessment of Orientation in Person, Place and Time (OPPT) was an adequate screening tool for HAND. They were presented with the names of other cognitive screening tools and were asked which they had used previously with HIV-positive patients. The list included the Mini Mental State Examination (MMSE), the Montreal Cognitive Assessment (MoCA) screen, the Abbreviated Mental Test Score (AMTS), the International HIV Dementia Scale (HIVDS) and the Brief Neurocognitive Screen (BNCS). Participants were asked to identify factors that limit them in performing cognitive assessments in the acute setting, including lack of expertise, lack of availability of an appropriate screening tool, exposed environment/lack of privacy, noisy environment, lack of time, poor patient cooperation and communication difficulties.

Attitudes

They were asked whose primary responsibility it is, in their opinion, to perform a cognitive assessment on patients who present acutely to the ED.

Analysis

Descriptive data were presented as proportions and percentages. Ethical approval was acquired prior to the commencement of the study from the hospital ethics committee.

Results

A total of 35 surveys were collected. Characteristics of the group were determined, including gender, hospital grade, length of time qualified and how frequently they would encounter HIV-positive patients in practice (Table 1).

Table 1. Group characteristics (n = 35)
 n (% of total)
  1. GUIDE, Department of Genitourinary Medicine and Infectious Disease; SHO, senior house officer.
Male8 (23)
GUIDE personnel (total), of whom:19 (54)
GUIDE doctors13 (37)
GUIDE nurses6 (17)
Emergency Department personnel, of whom:16 (46)
Doctors16 (46)
Doctors (total), of whom:29 (82)
Consultants4 (11)
Registrars11 (31)
SHOs12 (34)
Interns2 (6)
Length of time qualified 
0–2 years4 (11)
2–4 years9 (26)
4–6 years6 (17)
>6 years16 (46)
The frequency with which participants encounter HIV-positive patients 
GUIDE personnel (n = 19)n (% of GUIDE participants)
At least once per day15 (79)
A few times per week2 (11)
A few times per month2 (11)
Less frequently than stated above0
Emergency Department personnel (n = 16)n (% of ED participants)
At least once per day1 (6)
A few times per week8 (50)
A few times per month5 (31)
Less frequently than stated above2 (12)

Participants' mean estimates of the prevalence of cognitive impairment among patients taking and not taking HAART were 29.1 and 39.3%, respectively (Table 2).

Table 2. Estimated prevalence of cognitive impairment
 Prevalence in those not on treatment (%)Prevalence in those on treatment (%)
  1. ED, Emergency Department; GUIDE, Department of Genitourinary Medicine and Infectious Disease.
  2. *One ED doctor estimated a prevalence of cognitive impairment of 100% among those who are not on highly active antiretroviral therapy (HAART). Excluding this result, ED clinicians estimated prevalence at 41.3% among patients who are not on treatment.
Group (total)39.329.1
GUIDE doctors37.728.1
ED doctors45*31.6
GUIDE nurses22.522.5

The group's perceptions of the impact that various factors have on mortality, morbidity and chance of readmission were determined with the use of Likert scales. Cognitive impairment ranked fifth in order of impact (from highest to lowest) following poor adherence to medication, opportunistic infection, low CD4 count and substance abuse (Table 3).

Table 3. Factors that negatively impact on morbidity, mortality and chance of readmission
 Mean value on Likert scale
Medication non-compliance7.2
Opportunistic infection7.0
Low CD47.0
Substance misuse6.9
Cognitive impairment6.4
Time elapsed since diagnosis4.3
Ethnicity4.2
Female gender4.0

According to our survey, GUIDE doctors used cognitive screening tools on average 25% of the time when reviewing HIV-positive patients, GUIDE nurses 5% of the time and ED doctors 15%. The most commonly used screening tools were the MMSE, used by 37% (n = 12) of participants, followed by the AMTS, used by 20% (n = 7) of participants. Six per cent of participants (one GUIDE clinician and one GUIDE nurse) had previously used the HIVDS. Nine per cent (two GUIDE clinicians and one ED physician) had used the BNCS (Table 4).

Table 4. Screening tools
 n (% of respective group*)
  1. AMTS, Abbreviated Mental Test Score; BNCS, Brief Neurocognitive Screen; ED, Emergency Department; GUIDE, Department of Genitourinary Medicine and Infectious Disease; HIVDS, International HIV Dementia Scale; MMSE, Mini Mental State Examination; MoCA, Montreal Cognitive Assessment.
  2. *GUIDE clinicians, ED clinicians or GUIDE nurse.
MMSE 
GUIDE clinicians7 (54)
ED clinicians5 (31)
GUIDE nurse1 (17)
AMTS 
GUIDE clinicians5 (38)
ED clinicians2 (13)
GUIDE nurse0
MoCA 
GUIDE clinicians4 (30)
ED clinicians0
GUIDE nurse0
HIVDS 
GUIDE clinicians1 (8)
ED clinicians0
GUIDE nurse1 (17)
BNCS 
GUIDE clinicians2 (15)
ED clinicians1 (6)
GUIDE nurse0

Thirty-four per cent (n = 12) of respondents felt that OPPT was a sufficient screening tool for cognitive assessment: 23% of GUIDE clinicians (n = 3) and 56% of ED clinicians (n = 9).

Respondents found lack of time, exposed environment and lack of privacy the most limiting factors when performing cognitive assessment on patients who present acutely to the ED (Table 5).

Table 5. Limiting factors when performing cognitive assessment in the Emergency Department
 Mean value on Likert scale
Lack of time6.1
Exposed environment/lack of privacy5.1
Lack of availability of a screening tool4.9
Lack of patient cooperation4.8
Noise4.6
Unnecessary examination3.4
Lack of expertise3.4

Regarding patients who present acutely to the ED, 44% (n = 7) of ED clinicians felt that it was the responsibility of the GUIDE clinicians to perform cognitive assessments on all HIV-positive patients, while 19% (n = 3) felt that it was the responsibility of the initial reviewer. Thirty-one per cent (n = 4) of GUIDE clinicians felt that it was the responsibility of the ED doctors to perform the assessment, while 31% felt that it was their own responsibility. The majority of nurses interviewed (n = 5; 85%) felt that it was the doctor's responsibility to perform cognitive screens.

Discussion

This study examined awareness of HIV-associated neurocognitive decline among health care providers and also reasons for inadequate assessment. It emphasizes recommending a quick easy to use and readily available screening tool.

Identification of neurocognitive impairment supports earlier initiation or modification of HAART [14]. A reduction in plasma viral load has been associated with a reduction in neurocognitive decline [15, 16]. In addition, detection of neurocognitive impairment should prompt further investigation into other possible causes and management of coexisting conditions (e.g. hepatitis C, depression, and cardiovascular disease) [14, 17]. Patients with HAND have lower antiretroviral adherence, more mental health issues and an overall reduced quality of life [18]. It is therefore important to establish a neurocognitive baseline, identify those at risk and plan for frequent monitoring. It is beneficial to involve other services when managing these patients, including psychiatry, adherence support, social work and occupational therapy [19, 20].

The majority of our respondents were doctors from the GUIDE department and the ED, with representation of all hospital grades from interns to consultants. Many of the participants had a significant amount of clinical experience (46% greater than 6 years of experience). Participants commonly come into contact with HIV-positive patients, with only two ED physicians who encountered patients less than once per month.

There was reasonable awareness among participants with regard to the prevalence of neurocognitive impairment despite HAART treatment (mean estimates were 29.1% among patients on HAART and 39.3% among patients not on HAART). Respondents were also aware of the negative impact of neurocognitive impairment on morbidity and mortality, estimating it as the fifth leading cause.

Respondents rarely used screening tools, emphasizing the poor appreciation or lack of guidelines. GUIDE clinicians used screening tools more frequently than ED physicians (25 vs. 15%, respectively). The majority of ED clinicians in this study felt that it was not their responsibility to do so (44% felt that it was the responsibility of the GUIDE clinician). GUIDE nurses rarely used tools. A significant proportion of participants felt that it was sufficient to perform an OPPT assessment (34%). The most commonly used screening tool was the MMSE (37%). Very few people had used the HIVDS and BNCS (6% and 9%, respectively). The MoCA was rarely used by the group (11%). Lack of time, lack of privacy and lack of availability of a screening tool were cited as limitations to cognitive screening in the ED environment.

It has been shown that the overall prevalence of neurocognitive impairment has not changed significantly across the pre-HAART and HAART eras (the current estimated prevalence is 40 − 50%) [2, 3]. However, the pattern has changed, with an increase in ANI and MCND and a decline in HAD (from 10 − 15 to 2%) [3, 16]. The detection of asymptomatic neurocognitive impairment is challenging, frequently requiring third-party report or prolonged observation [1]. The European AIDS Clinical Society (EACS) guidelines recommend biennial screening involving three questions that focus on self-reports of cognitive difficulty. A positive screen would then mean a further activity of daily living assessment and a formal neuropsychiatric assessment [21].These guidelines have been criticized by the CHARTER group, who concluded that the EACS guidelines were poor at detecting asymptomatic HAND when compared with HIVDS [22].

In addition, there is difficulty when adequately assessing HAD (thought to be associated with subcortical deterioration) [2, 23, 24]. Short screening tools (for example AMTS) do not assess frontal or executive function, and therefore poorly detect subcortical decline [25]. The MMSE has also been shown to have a very limited role in the detection of subcortical processes [26]. The HIVDS and BNCS have 70 and 65% sensitivity, respectively, in the detection of HAND [27, 28].The MoCA may be of benefit as it is designed to detect cortical and subcortical processes, but further studies are needed [27, 29].

The nature of a busy ED department and the fact that many patients have other confounding acute medical conditions discourage against performing formal neurocognitive assessments on these patients. A basic neurocognitive assessment may have a role in identifying patients with a delirium or other neurological disease (e.g. opportunistic infection, lymphoma, stroke or seizure). These conditions may exacerbate or unmask an underlying unidentified HAND diagnosis [30]. A more formal and detailed neurocognitive assessment can then take place during the admission or after the acute illness. This is particularly important for elderly patients who present with an indicant diagnosis of HIV infection, as it is important to establish if HAND contributed to delayed presentation.

In our study, lack of time, lack of privacy and lack of availability of a screening tool were emphasized as limitations to cognitive screening in the ED environment. These findings are similar to those of another study that focused on older patients who presented to the ED. It demonstrated that only one in four older patients were being screened for cognitive impairment (despite a prevalence of up to 42%). Cognitive assessments of older patients were also restricted by environmental factors (lack of privacy and noise, time constraints, and the absence of a standardized brief screening tool) [31].

Routine screening for neurocognitive impairment is recommended for all HIV-positive patients. However, the efficacy of cognitive screening tools varies considerably and there is a lack of world-wide consensus on existing guidelines. There is a clear need for integrated care pathways for patients who are diagnosed with HAND involving relevant neuropsychiatric services, social work, and close infectious disease follow-up (especially on commencement of HAART). Thorough investigation for other causes of neurocognitive impairment is also essential. Cognitive screens should be considered in high-risk patients who present acutely. A brief, sensitive and easy to use screening tool should be available to clinicians in this setting. It is important to follow up these patients with a formal neurocognitive assessment to establish an underlying diagnosis of HAND.

Ancillary