QJM vol. 97 no. 10 © Association of Physicians 2004; all rights reserved.
GOAL: a simplified mental test for emergency medical admissions
From the Department of Adult Medicine, Royal Gwent Hospital, Newport, Gwent, and 1Department of Epidemiology, Statistics and Public Health, University of Wales College of Medicine, Cardiff, UK
Received 10 February 2004 and in revised form 4 June 2004
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Several scoring systems are used in screening for cognitive impairment, but none are suited to the busy medical assessment environment.
Aim: To construct, validate and assess the reliability of a simple scale (Gwent Orientation and Awareness Listing, GOAL) for this purpose, and to examine its application in consecutive emergency admissions in two general hospitals.
Design: Prospective cohort studies.
Methods: The validity and reliability of GOAL was assessed in three studies of patients aged 
65 years who had been pronounced medically fit for discharge. The evaluation studies were carried out over 4-week medical intake periods in each participating hospital.
Results: Correlation of GOAL with the standard 30-point Mini-Mental State Examination was 0.89, and the inter-observer reliability was 0.90. Based on Receiver Operating Characteristics Curves, patients scoring <8 on GOAL were deemed to be cognitively impaired. Assessment by GOAL took half the time required for the widely used Abbreviated Mental Test Score. Of 1037 consecutive patients admitted to two hospitals' acute medical intakes and remaining for >24 h, 952 were able and willing to be scored by GOAL, and of these 201 (21%) ‘failed’, with a score of <8.
Discussion: Loss of orientation and awareness is common among patients admitted via medical intakes. GOAL is a practical brief screen for identifying and following-up these patients.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
As the population ages, a higher proportion of acute emergency medical admissions are elderly patients with multiple medical problems including cognitive impairment.1,2 This can manifest as delirium secondary to organic illness or its treatment, dysmnesia, dementia or pseudodementia associated with depressive illness. These forms of impairment often coexist.3 Delirium with associated psychosis is usually obvious, but more subtle delirium and subacute confusion often go unrecognized.4 Dementia is thought to be present in nearly 10% of consecutive medical admissions, with one-third of those aged
85 years affected.5 Meagher recently commented that detection of delirium can be improved by routine cognitive testing, but lamented that ‘cognitive assessment is less common in the technological world of modern medicine’ and that ‘knowledge of a patient's prior cognitive state is often minimal’.3 This may be because existing instruments are not suited to the rapid assessment of patients with acute illness.
The Mini-Mental State Examination (MMSE) is used internationally to screen for dementia.6 In addition to testing for orientation, it screens briefly for agraphia and examines short-term memory. Its routine use while admitting acutely ill medical patients is impractical, because the 30-point scale takes around 10 min to complete. Due to its brevity, the Abbreviated Mental Test Score (AMTS) is commonly used in screening for cognitive impairment in the UK.7 This 10-point scale, however, is not a suitable cognitive test in a busy acute assessment environment, because it requires the availability of another person (usually a nurse) whom the patient should be able to distinguish from the admitting doctor. Furthermore, doctors differ in its administration and scoring.8
We perceived the need for a practical and brief scoring system for detecting cognitive impairment among patients admitted as acute emergencies. We chose the term ‘loss of orientation and awareness’ to encompass both acute decompensation and permanent loss of cognitive function.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Development of GOAL
From the existing tests, we took those questions commonly asked in everyday clinical practice, including orientation in time (4 parameters), place (1 parameter), awareness (2 parameters) and personal details (3 parameters). The parameters and suggested leading questions are shown in Box 1. Five of the 10 GOAL questions were taken from the existing ten-point AMTS.7 The five tests omitted and the reasons for their omission are as follows: the date of the First World War (a test of knowledge of history that we found many well-orientated patients unable to answer correctly, especially non-Caucasians); the time to the nearest hour (many acutely unwell patients have no access to timepieces in hospital); the ability to recognize two people, for example a doctor and nurse (a nurse is not always immediately accessible when patients are being seen by a doctor); backward counting from 20 to 1 (which is time-consuming and requires the patient to remain attentive throughout); and later recall of an address (a test of short-term memory requiring the patient to be alert and co-operative throughout). Included instead were the day of week, date of month, month of year, patient's address and name of prime minister. There is evidence that the latter is a highly discriminating parameter.9
| Box 1 Gwent Orientation and Awareness Listing Entry qualification: patients are asked to give their full name (scoring not appropriate if patient unable to give their name) Scoring: One point for each correct response. Maximum score of 10.
Date of birth: Can you tell me your date of birth? Age: And how old does that make you? Date of month: Can you tell me today's date? Day of week: Do you know what day of the week this is? Month: Can you tell me what month we’re in? Year: Can you give me the current year? Place: Do you know where you are? (should name hospital if hospitalized) Own address: Please tell me your own address, or where you live? Prime Minister: Who is the current prime minister? (or e.g. President in USA, Chancellor in Germany) Current Monarch: Who is the reigning monarch: king or queen? (or e.g. Vice-President in USA, President in France)
|
To score by GOAL patients are first asked to give their name. An absent response could indicate severe learning disability, delirium with psychosis, coma, profound deafness, dysphasia and/or severe dementia. The ability to give a correct name was deemed an essential requirement for participation. Those able and willing to participate are then asked the ten GOAL questions and scored one point for each correct response.
Validity and practical application of GOAL were tested in three studies of patients aged 65 years who had been pronounced medically fit for discharge from acute medical beds.
Validation of GOAL
To assess concurrent criterion validity, the MMSE was chosen as the most appropriate comparator for GOAL. During a four-week period, 251 consecutive patients were invited to participate in this study. All were recovering while under the care of consultant physicians and had been deemed medically fit for discharge from acute hospital beds. Verbal consent to be tested with GOAL, followed by MMSE, was obtained from 240 of the patients. Of these, 52 (21%) were unable to complete both GOAL and MMSE (14 were partially sighted, 11 withdrew due to physical frailty, 10 due to dysphasia, 8 due to deafness, 6 due to illiteracy and/or dyslexia, and 3 due to a language barrier). Thus both scales were completed for 188 patients. Spearman's rank correlation was used to examine the correlation between results for each patient obtained by each of the two scoring systems. Possible cut-offs for GOAL were derived from calculation of its sensitivity, specificity and predictive value for detecting impairment as defined by MMSE, and by plotting the Receiver Operating Characteristic Curve (ROC).10 Unfortunately, however, there is disagreement on the cut-off point for an abnormal result on MMSE. Although there is agreement that a score of 23 or lower is abnormal, some workers have proposed age-related cut-offs, with a score of <26 indicating impairment in those aged >80.11,12
Inter-observer and test-retest reliability
One hundred and two further medical in-patients, aged 65, whose condition had stabilized sufficiently for discharge or transfer, were tested by GOAL on consecutive days by two independent investigators. One patient withdrew during the second assessment. Each investigator remained blinded to the results of the other until all 101 completed patients had been scored. In contrast to Box 1, so as to reflect everyday practice, we chose not to standardize the leading questions for each parameter. Spearman's rank correlation coefficient was calculated to measure inter-observer reliability and agreement in diagnosing cognitive impairment based on GOAL testing was assessed by calculating the kappa coefficient.10
Use of GOAL in clinical practice
All patients admitted during acute general medical takes, and remaining in hospital for at least 24 h, were assessed over two 28-day periods. One study period took place at the Royal Gwent Hospital, Newport (serving a largely urban and suburban population) and the other at Nevill Hall Hospital, Abergavenny (which serves a more rural catchment area). Patients were identified using the patient information system on the hospital clinical workstation and/or from the admission diaries of the Medical Assessment Units or Accident and Emergency. There is no separate Care of the Elderly intake at either hospital, so all patients aged 16 and above were assessed. After explanation of the purpose of the investigation, patients were invited to give verbal consent to testing by GOAL on the day after their admission. The admitting physicians were unaware of the results obtained by means of GOAL, but were free to perform their own assessment of the patients’ mental state on admission and prior to discharge.
Data was collected on patient demographics, whether they lived alone, whether they had been admitted from home or from an institution, the referral source, history of dementia or other psychiatric illness, alcohol and substance misuse and/or learning disability. Descriptive statistics including means and ranges were used where appropriate and independent sample t tests and multi-layer 
2 tests were analysed. The ‘Statistix’ (version 1, Analytical Software) programme was used for calculation.
GOAL in detecting sensitivity to change
A pilot study was done to examine the potential of GOAL in detecting trends in orientation and awareness associated with treatment of medical illness. Patients scoring <8 on their initial assessment were considered cognitively impaired. Those who had survived their acute illness were reassessed immediately before discharge or transfer from acute care wards.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Validation and cut-offs
Figure 1 shows the results of GOAL in relation to those obtained by MMSE testing for the 188 patients with complete data. There was good correlation between the two scores (r = 0.89, 95%CI 0.85–0.94, p < 0.0001). Data on the accuracy of GOAL in diagnosing cognitive impairment as defined by MMSE are shown in Table 1. To take account of discrepancies in cut-offs, Table 1 and Figure 1 present comparative data using MMSE scores of 25 and 24, and GOAL of 9 and 8 as lower limits of normal.
|
|
While lacking sensitivity (68%) a GOAL result of <8 proved to be a strong predictor of cognitive impairment, as defined by MMSE of <24 (specificity 97%, positive predictive value 91%). Using these cut-offs, agreement between the two scores in defining presence or absence of cognitive impairment was good (
= 0.69). Increasing the GOAL threshold to 8 led to a rise in sensitivity to 90%, while reducing specificity to 84%. The two investigators completely agreed on GOAL scoring of 36 of the 101 patients. There was a one-point difference for 50 patients, a two-point difference in 13 patients and a three-point difference in two patients (one of whom was deemed impaired with GOAL<8 by both investigators). Spearman's rank correlation of the results obtained by each investigator for each patient was significant at 0.90 (p < 0.001).
Speed of administration
Twenty-five further patients agreed to undergo AMTS and GOAL in random order. Mean time to complete AMTS was 176 s, compared with 91 s for GOAL (p < 0.001).
Clinical study
Of 1037 patients who remained in hospital the day after admission to the general medical wards, 10 did not wish to participate and 14 were considered too unwell by their admitting team. Of the remaining 1013, 61 were not studied further because they were unable to give their name correctly; of these, 25 were in coma or sedated, 29 had dysphasia, four could not speak English, two had learning difficulties and one had extreme deafness. This left 952 patients, of whom 490 (52%) were admitted via their General Practitioner; the remainder came directly through Accident and Emergency or had been transferred from smaller hospitals nearby. Of the 952 patients, 496 had been admitted to the Royal Gwent Hospital and 456 to Nevill Hall Hospital.
In total, 201 patients (21%) scored <8 on their GOAL assessment. Of these, only 29 (14%) had a previous diagnosis of dementia. As expected, patients who scored <8 were significantly older (mean 77 years) than those scoring 8+ (mean age 64) (Table 2). Loss of orientation and awareness was significantly more prevalent amongst patients who lived alone (80/284, 29%), compared with those living at home with others (74/592, 12%) (2 = 32.1, p < 0.001). This difference could be explained by the older age of the patients living alone (mean 80 vs. 74 years, p < 0.001). Patients admitted from residential or nursing homes were even more likely to fail (38% scored <8, 2 = 40.6, p < 0.0001). Failure to score 8+ out of 10 was more likely among patients taking psychotropic medication (p < 0.03), those with epilepsy (p < 0.0004) or chronic schizophrenia (p < 0.0001).
|
On follow-up of the 106 Royal Gwent Hospital patients who scored < 8 on initial assessment, 11 (10%) had died and 54 (51%), including the 13 with a previous diagnosis of dementia, failed again at reassessment. Four patients were too ill for re-testing and six were lost to follow-up. Only 31 (29%) of those who scored < 8 on initial assessment recovered to pass with a score of 8+ at time of discharge. Failure to reach a score of 8+ within this group was significantly associated with increased age, mean age 82.2 (range 67–91) vs. 73 (range 53–84), p = 0.008.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
This study has demonstrated loss of orientation and awareness in >20% of patients admitted on an acute general medical intake. A pre-admission diagnosis of dementia had only been made in 12% of the patients who scored < 8 on the GOAL assessment. The results show that cognitive impairment is common among general medical intake patients, and suggest that such impairment may often go unrecognized. Failure of orientation and awareness to return with treatment of the acute illness suggests that pre-existing dysfunction is more commonly to blame than transient impairment related to the illness. No formal psychiatric assessment or follow-up was done during this study, and the results must therefore be interpreted with caution. Further studies would be required in order to investigate possible underlying causes and the potential for reversibility of cognitive impairment in this setting. Cognitive impairment was most often found in patients residing in institutions, and among those living alone, a finding which accords with the results of recent prospective follow up studies.13,14
There appears to be an increasing frequency of cognitive impairment in old age even when conditions such as dementia and delirium are excluded. Several large community-based studies have shown a high prevalence of impaired cognitive function in the elderly. The Canadian Study of Health and Ageing estimated the prevalence of cognitive impairment in 10 263 people aged >65 years. Its prevalence approached 25%, but only about one third of this could be accounted for by dementia.15 Other causes were identified as depression, alcohol and drugs, psychiatric illness and learning disability. Other studies have confirmed these findings and have shown that cognitive impairment is associated with increased use of health services and mortality.16,17
There are several potential causes of loss of orientation and awareness in patients presenting as acute medical emergencies. The prevalence of dementia increases exponentially with age, affecting approximately one third of patients aged 85 years.18 While advanced age is the main risk factor, moderate or severe dementia may affect up to 10% of medical in-patients aged >50 years.5 Acute confusional states are also more common in the elderly, and may be precipitated by prior medical intervention and other factors such as visual impairment or urinary tract infection.19 The consequent delirium has been shown to increase hospital stay, institutionalization and mortality.20 Delirium often occurs in those with pre-existing cognitive impairment.21 Cognitive impairment is also associated with increased mortality, institutionalization and use of resources.16
The most widely used scale for assessing cognitive function at the bedside is the MMSE.6 Just over one-fifth of the patients in our comparative study could not complete the MMSE, due to frailty or other incapacity. The use of GOAL appears to provide a simpler, more rapid and easily remembered scale for assessing orientation and awareness in the acute setting. Moreover, GOAL can be readily administered to a wider spectrum of hospital in-patients, including the frail and partially sighted. A potential weakness of GOAL in the acute admitting environment is that it tests the patient's knowledge of their date of birth, age and address. Responses to these parameters might be difficult to verify in circumstances where patients are not carrying any identity.
The results of this study suggest that a score of 8 on GOAL is an appropriate cut-off point. In the comparative study, every patient scoring 
7 scored 25 on MMSE. Only one of the patients known to have dementia scored >8 on GOAL. The reliability study also has a bearing on the cut-off point for GOAL. If a score of 8 is taken as equivocal, indicating the need for later re-testing or more detailed assessment with other instruments, none of the 101 patients would have been classified as definitely impaired (GOAL < 8) by one investigator yet definitely normal (GOAL >8) by the other. The regarding of 8 as an equivocal score also allows for time parameter errors on the first and last days of the month. Unimpaired patients could, for example, drop their score to 8 on 1 August by giving the date as 31 July. We suggest that a score of 8 should be interpreted in its clinical and chronological context. An equivocal score on admission may indicate the need for later reassessment or more formal cognitive testing.
Simple tests such as GOAL may aid the initial assessment of older patients admitted to hospital, because the early recognition of cognitive impairment may alter the doctor's interpretation of a patient's clinical history. Early detection allows for interventions not only to address any underlying treatable cause, but also to facilitate the planning of social interventions such as day hospital care, which may reduce the level of subsequent institutionalisation. Routine cognitive testing at the time a patient is admitted to hospital gives a baseline against which subsequent assessments can be compared. Cognitive testing near the time of discharge may also assist in identifying patients at risk of non-compliance with medication once at home. Furthermore the early recognition of some forms of dementia might become increasingly important as new therapies evolve.22
In conclusion, GOAL is proposed as a practical and clinically useful screen for cognitive impairment in general medical inpatients. Its simplicity and applicability to the acute admitting environment mirrors that of the Glasgow Coma Scale.23 Further work is being done to examine the potential of GOAL in the assessment of surgical and orthopaedic admissions, and in the evaluation of patients recently discharged from hospital.
|
Acknowledgments |
|---|
We thank the consultant physicians under whose care these patients were admitted. We are grateful to Professor John Pathy for his advice during the conception of GOAL. Elin Hughes, Sarah Farmer, Justin Kington, Stefan Bek, Eleri Griffiths and Amanda Lacey performed some of the cognitive assessments for their Special Study Modules during their final year as students of the University of Wales College of Medicine. Dr Jackie McNair helped to oversee the students at Nevill Hall Hospital.
|
Footnotes |
|---|
Address correspondence to Dr M.C. Allison, Consultant Physician, Royal Gwent Hospital, Newport NP20 2UB. e-mail: miles.allison{at}gwent.wales.nhs.uk
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Khaw KT. How many, how old, how soon. Br Med J 1999; 319:1530–2.
2. Capewell S. The continuing rise in emergency admissions. Br Med J 1996; 312:991–2.
3. Meagher DJ. Delirium: optimising management. Br Med J 2001; 322:144–9.
4. Inouye SK. The dilemma of delirium: clinical and research controversies regarding diagnosis and evaluation. Am J Med 1994; 97:278–88.[CrossRef][Web of Science][Medline]
5. Erkinjuntti T, Wikstrom J, Palo J, Lauri A. Dementia among medical inpatients. Evaluation of 2000 consecutive admissions. Arch Intern Med 1986; 146:1923–6.
6. Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method for grading the cognitive state of patients for clinicians. J Psych Res 1975; 12:189–98.
7. Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing 1972; 1:233–8.
8. Holmes J, Gilbody S. Differences in use of abbreviated mental test score by geriatricians and psychiatrists. Br Med J 1996; 313:465.
9. Fuller GN, Meeran K. The Prime Minister, Mrs Thatcher, and dementia. Lancet 1991; 337:1362.[Web of Science][Medline]
10. Streiner DL, Norman GR. Reliability. In: Health Management Scales – A practical guide to their development and use, 3rd edn. Oxford, Oxford Medical Publications, 2003:119–23.
11. Hodges JR. Cognitive assessment for clinicians. Oxford, Oxford Medical Publications, 1994.
12. Tombaugh TN, McIntyre NJ. The Mini-Mental State Examination: a comprehensive review. J Am Geriatr Soc 1992; 40:922–35.[Web of Science][Medline]
13. Fratiglioni L, Wang HX, Ericsson K, Maytan M, Winblad B. Influence of social network on occurrence of dementia: a community-based longitudinal study. Lancet 2000; 355:1315–19.[CrossRef][Web of Science][Medline]
14. Bassuk SS, Glass TA, Berkman LF. Social disengagement and incident cognitive decline in community-dwelling elderly persons. Ann Intern Med 1999; 131:165–7.
15. Graham JE, Rockwood K, Beattie BL, et al. Prevalence and severity of cognitive impairment with and without dementia in an elderly population. Lancet 1997; 349:1793–9.[CrossRef][Web of Science][Medline]
16. Breteler MMB, Claus JJ, Grobbee DE, Hofman A. Cardiovascular disease and distribution of cognitive function in elderly people: the Rotterdam study. Br Med J 1994; 308:1604–8.
17. Brayne C, Gill C, Paykel ES, Huppert F, O’Connor DW. Cognitive decline in an elderly population—a two wave study of change. Psych Med 1995; 25:673–83.
18. Ott A, Breteler MM, van Harskamp F, Claus JJ, van der Cammen TJ, Grobbee DE, et al. Prevalence of Alzheimer's disease and vascular dementia: association with education. The Rotterdam study. Br Med J 1995; 310:970–73.
19. Inouye SK, Charpentier PA. Precipitating factors for delirium in hospitalised elderly persons. JAMA 1996; 275:852–7.
20. Stevens LE, de Moore GM, Simpson JM. Delirium in hospital: does it increase length of stay? Aust NZ J Psych 1998; 32:805–8.
21. Henon H, Lebert F, Durieu I, et al. Confusional state in stroke: relation to pre-existing dementia, patient characteristics and outcome. Stroke 1999; 30:773–9.
22. Jonsson L, Lindgren P, Wimo A, Jonsson B, Winblad B. The cost-effectiveness of donepezil therapy in Swedish patients with Alzheimer's disease: a Markov model. Clin Ther 1999; 21:1230–40.[CrossRef][Web of Science][Medline]
23. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet 1974; ii 81–4.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||