Cognitive function has a significant effect on the older person and their admission to hospital. Assessment of cognition is relevant at the time of admission because impairment can lead to an inaccurate history (if the patient is solely relied upon for the history), and because there may be a change to patients' mental status associated with the acute illness.Assessment of cognitive function can also provide prognostic information about risk of delirium, an under-recognised and therefore under-treated condition with significant associated mortality.1,2,6,7,9 Delirium affects both recovery of function, which can lead to threatened independence and need for residential care,3,4 and funding of the episode of patient care.Delirium is a common problem estimated at affecting 10-40% of elderly inpatients. A New Zealand study measured rates between 25-28%.2 Often the acute changes in cognitive function associated with the acute illness are of major concern to a patient's family and friends, sometimes more so than the acute illness itself. Assessment and explanations about this assist them to understand the complications of the illness.Routine assessment of cognitive function, on hospital admission, may lead to earlier diagnosis of cognitive impairment, early dementia or other psychiatric conditions like depression. This may provide the opportunity to discuss this diagnosis with the patient and family or, refer them on to the appropriate out patient or community service for assessment (which may be the more appropriate response). These are important conditions to diagnose as there are treatment options and supports available.It is important for medical staff to not only be aware of cognitive function at admission, but to document it. Previous studies have demonstrated that delirium and cognitive impairment documentation is variable and these conditions are often diagnosed but not documented.1,5 An acute confusional state may be the only symptom of an illness or deterioration.4During a hospital admission, a number of medical staff are involved in a patient's care. While the admitting team will be familiar with the patient, on call and out of hours medical staff may not. It is important that observations about cognitive function, including pre-morbid cognitive function, and associated syndromes, like delirium, are documented. This is especially relevant as the signs and symptoms and behaviours of a delirium may only occur late in the day or overnight, when the patient's medical team have finished for the day.The objective of this study was to measure the documentation of cognitive function and factors that affect cognitive function and delirium risk, within the first 24 hours of admission to an acute general medical service.Methods This was a prospective review of medical records of all patients admitted under acute medical teams over a 2-week period from 26 July 2008 until 8 August 2008 inclusive. Inclusion criteria were any patient aged 65 years and older admitted to an acute medical team. Patients were excluded if they were admitted for elective procedures under the acute medical team or if they transferred to another speciality prior to the post-acute consultant ward round. If a patient was admitted more than once during the audit period, only the first admission was counted. All patient records were reviewed by one medical registrar. Aspects of the admission reviewed included the medical notes made by doctors within the first 24 hours (including the admission note and post-acute consultant ward round), general demographics (taken from the clerical admission sheet) and the formal discharge summary. Information in nursing notes or allied health workers notes or from residential care units was not included. Information gathered included age, sex and living circumstances. Previous history of dementia and previous measurements of this were recorded as well as pre-morbid mental status and whether or not this had changed with the acute illness. The process used to evaluate mental status was also recorded. These measures included general observation and general comment or specific measures (10-point Mental State Examination or 30-point Mini-Mental State Examination or 100-point Modified Mini-Mental State Examination). Functional ability with regards to mobility and sensory disturbance (i.e. visual or hearing impairment) were also recorded. The working diagnosis was recorded as that documented on the consultant ward round. Results In the 2-week period, there were 266 acute medical admissions. Of these, there were 2 patients who were readmitted during the audit period and 7 patients whose notes were not available for screening either during or after the specified timeframe. Therefore there were 257 admissions in total that were audited. Of these, all but one had formal discharge summaries in electronic form. Results were summarised by simple tabulation and as a proportion of the audited notes. Table 1. Demographics Variables N (%) Age 65-69 70-79 80-89 90+ 26 (10.1%) 78 (30.4%) 127 (49.4%) 26 (10.1%) Sex Male Female 126 (49%) 131 (51%) Location of residence Christchurch + suburbs Rural/greater Christchurch area Other 224 (87%) 26 (10%) 7 (3%) Place of residence Home/independent living alone Home/independent living with spouse/partner/carer/family Home/independent living but not stated if alone or with family Rest Home resident Private Hospital resident Place of residence not noted 88 (34.2%) 101 (39.3%) 6 (2.3%) 42 (16.3%) 5 (2%) 15 (5.8%) Primary working diagnosis on admission Respiratory disease Cardiac disease Neurological condition Falls/collapse Confusion/Delirium Cellulitis Renal condition/urinary infection Malignancy Other 77 (30%) 38 (14.8%) 36 (14%) 29 (11.3%) 11 (4.3%) 6 (2.3%) 4 (1.6%) 2 (0.8%) 54 (21%) The primary working diagnoses are summarised as seen in Table 1. "Other" included medication side effects, gastroenteritis, anaemia, arthritis with associated pain, postural hypotension, hypoglycaemia, vertigo, gastrointestinal bleed and musculoskeletal problems. The majority of patients were over the age of 80 years (153, 59.5%) and there was an even number of males and females. Most patients lived within Christchurch (224, 87%) and were in independent living situations (195, 75.8%), either alone (88, 34.2%) or with at least one other (101, 39.3%). The most common working diagnosis on admission was respiratory disorders (77, 30%). Prior functional status in terms of mobility and sensory disturbance (i.e. vision and hearing disturbance) was poorly documented, with only 128 (50%) patients' mobility status recorded and only 63 (24.5%) patients with a comment about presence or absence of sensory disturbance. Past medical history of dementia/cognitive impairment was recorded in 28 (11.3%), see Table 2. Table 2. Premorbid function Mobility Independent Walking frame/stick/mobility scooter Immobile No comment made about mobility 52 (20%) 74 (29%) 2 (0.8%) 129 (50%) Hearing Hearing impairment No comment made 13 (5%) 237 (92.2%) Visual Visual impairment No comment made 31 (12%) 214 (83.3%) Dementia Past medical history of dementia/impaired cognition documented in admission note 29 (11%) Table 3. Cognitive state Mental state Previous mental state noted on admission 36 (14%) Present mental state Usual Worse Improved No comment 15 (6%) 35 (14%) Nil 207 (81%) Measure of previous mental state noted on admission 3MS*/MMSE** 10-point Mental State Exam General comment Nil Nil Nil 34 (14%) 223 (87%) Measure of present mental state noted on admission 3MS*/MMSE** 10 point Mental State Exam General Comment Nil Nil 15 (6%) 138 (53%) 104 (41%) Delirium Delirium included in diagnosis at discharge 8 (3%) * Modified Mini Mental State Examination; **Mini Mental State Examination. Previous mental status and whether the present mental status has changed from baseline was poorly documented, with 207 (81%) patients not having a comment made about this at all. While lack of collateral history may have contributed to this, in some cases there were comments made by other health professionals within the notes that documented premorbid functioning. It is not known whether or not these were noted by the documenting doctor. Present mental state was measured and documented in 153 (59%) patients. The most commonly used measure of mental status was a general comment (e.g. alert and orientated). Only 8 patients had delirium noted within the discharge diagnosis list. However there were another 10 patients whereby a description of delirium was made in the body of the discharge summary but the diagnosis was not listed. Of the admissions reviewed, there were nine inpatient deaths. None of these patients were noted to have altered mental state or confusion on their discharge diagnosis (although discharge summaries for deceased patients are often brief and only include cause of death rather than an extensive problem list for the admission). Discussion In keeping with other observational studies, the documentation by medical staff of cognitive state and impairment or acute delirium is inadequate.5 This was despite other healthcare workers involved in patient care documenting information suggestive of acute or long-standing cognitive decline. It is already acknowledged in the literature that acute delirium on acute medical wards is under-recognised.9 As mentioned previously, delirium is a treatable condition with an associated mortality, hence the importance of detection and intervention. It is also considered to be preventable.6 Delirium, when documented within the body of the clinical notes by a doctor, leads to additional funding for that acute episode of care. Patients suffering from delirium often require increased resources and nursing care, sometimes even one to one supervision, which comes at an added cost in both staffing numbers and time. Non-specific observations like "confusion" or "cognitive impairment" are not sufficient for coding purposes to be classified as delirium. For some patients, there was adequate information in the notes to suggest that they had an acute delirium but such a diagnosis was not specifically documented and therefore could not be coded as such. For the remaining patients, the presence or absence of delirium was poorly documented. There are many reasons for under-recognition and under-reporting of delirium. Acute delirious states can be either hyper or hypoactive and it is the former that is more widely recognised, partly because of the difficult behaviours associated with it. Hypoactive delirium is more likely to be overlooked and has a higher mortality rate associated with it.9 The fluctuating nature of a delirium can also make it difficult to diagnose. While delirium is recognised by medical staff, the documentation of it may be poor as the medical team focus more on the underlying acute illness. Despite the known increased mortality with delirium, it is not treated as a life-threatening illness.9 There are beneficial interventions for both the patient with early dementia and their caregivers.6 While an acute medical admission is may not be an appropriate forum to address previously undiagnosed cognitive impairment, suggestive of a diagnosis of dementia, it is an opportunity to refer the patient and their family for geriatric review as an out patient. Symptom progression can, in some patients, be delayed with cholinesterase inhibitors and both patient and care-givers can benefit from supports and interventions to help deal with behavioural problems associated with dementia.5 Cognitive impairment may be because of other treatable conditions like depression. Under-reporting of dementia and cognitive impairment in an acute medical admission is also multifactorial. Often, due to the busy nature of the acute medical wards, the primary medical problem is the main focus rather than chronic underlying problems. Cognitive function is not routinely tested on or during a medical admission and, as shown in this study, the admitting doctor's preferred tool is general observation. Weaknesses in this study include reliance on the patient for a full history as often a collateral history is not available at the time of admission, thus making it difficult to know if information is reliable and whether present mental state is usual for the patient. Even with full access to patients' hospital records, previous relevant medical history is not readily available. An example of this is the variable filing of formal cognitive testing on previous admissions. Included as part of the initial assessment of the patient was the consultant ward round note. This is written by another member of the team (which can vary from a fourth-year student to a senior registrar) and therefore may not include all the observations or comments made by the consultant. The prospective review was performed over a 2-week period in an attempt to give an overview of medical admissions and a reasonable sample size. However there will be a documentation bias as the night medical teams will be over-represented as they work a 7-night roster. The results of this study do not provide new information but helps to highlight the issue at a local level. This study was designed to measure how well cognitive function is measured and documented in the elderly at the time of admission. The results show that this is poorly done. There needs to be increasing awareness of the importance of such documentation, both for treatment and intervention reasons as well as funding purposes. Preferred tools would include the CAM (Confusion Assessment Method) and MMSE, along with collateral history, for acute delirium6 and collateral history and Modified Mini-Mental State Exam for dementia. Admittedly the 3MS can be quite time consuming during the acute admission process for the admitting doctor and more appropriately performed either later in the admission or when the patient is recovered from the acute illness in an outpatient setting. A targeted history and CAM assessment would be achievable during the first 24 hours of admission. Presenting this information locally may improve awareness and documentation at the Christchurch Hospitals. Further studies would be needed to measure both improvements in documentation and to consider if this leads to improvements in patient outcome.

Cognitive function has a significant effect on the older person and their admission to hospital. Assessment of cognition is relevant at the time of admission because impairment can lead to an inaccurate history (if the patient is solely relied upon for the history), and because there may be a change to patients' mental status associated with the acute illness.Assessment of cognitive function can also provide prognostic information about risk of delirium, an under-recognised and therefore under-treated condition with significant associated mortality.1,2,6,7,9 Delirium affects both recovery of function, which can lead to threatened independence and need for residential care,3,4 and funding of the episode of patient care.Delirium is a common problem estimated at affecting 10-40% of elderly inpatients. A New Zealand study measured rates between 25-28%.2 Often the acute changes in cognitive function associated with the acute illness are of major concern to a patient's family and friends, sometimes more so than the acute illness itself. Assessment and explanations about this assist them to understand the complications of the illness.Routine assessment of cognitive function, on hospital admission, may lead to earlier diagnosis of cognitive impairment, early dementia or other psychiatric conditions like depression. This may provide the opportunity to discuss this diagnosis with the patient and family or, refer them on to the appropriate out patient or community service for assessment (which may be the more appropriate response). These are important conditions to diagnose as there are treatment options and supports available.It is important for medical staff to not only be aware of cognitive function at admission, but to document it. Previous studies have demonstrated that delirium and cognitive impairment documentation is variable and these conditions are often diagnosed but not documented.1,5 An acute confusional state may be the only symptom of an illness or deterioration.4During a hospital admission, a number of medical staff are involved in a patient's care. While the admitting team will be familiar with the patient, on call and out of hours medical staff may not. It is important that observations about cognitive function, including pre-morbid cognitive function, and associated syndromes, like delirium, are documented. This is especially relevant as the signs and symptoms and behaviours of a delirium may only occur late in the day or overnight, when the patient's medical team have finished for the day.The objective of this study was to measure the documentation of cognitive function and factors that affect cognitive function and delirium risk, within the first 24 hours of admission to an acute general medical service.Methods This was a prospective review of medical records of all patients admitted under acute medical teams over a 2-week period from 26 July 2008 until 8 August 2008 inclusive. Inclusion criteria were any patient aged 65 years and older admitted to an acute medical team. Patients were excluded if they were admitted for elective procedures under the acute medical team or if they transferred to another speciality prior to the post-acute consultant ward round. If a patient was admitted more than once during the audit period, only the first admission was counted. All patient records were reviewed by one medical registrar. Aspects of the admission reviewed included the medical notes made by doctors within the first 24 hours (including the admission note and post-acute consultant ward round), general demographics (taken from the clerical admission sheet) and the formal discharge summary. Information in nursing notes or allied health workers notes or from residential care units was not included. Information gathered included age, sex and living circumstances. Previous history of dementia and previous measurements of this were recorded as well as pre-morbid mental status and whether or not this had changed with the acute illness. The process used to evaluate mental status was also recorded. These measures included general observation and general comment or specific measures (10-point Mental State Examination or 30-point Mini-Mental State Examination or 100-point Modified Mini-Mental State Examination). Functional ability with regards to mobility and sensory disturbance (i.e. visual or hearing impairment) were also recorded. The working diagnosis was recorded as that documented on the consultant ward round. Results In the 2-week period, there were 266 acute medical admissions. Of these, there were 2 patients who were readmitted during the audit period and 7 patients whose notes were not available for screening either during or after the specified timeframe. Therefore there were 257 admissions in total that were audited. Of these, all but one had formal discharge summaries in electronic form. Results were summarised by simple tabulation and as a proportion of the audited notes. Table 1. Demographics Variables N (%) Age 65-69 70-79 80-89 90+ 26 (10.1%) 78 (30.4%) 127 (49.4%) 26 (10.1%) Sex Male Female 126 (49%) 131 (51%) Location of residence Christchurch + suburbs Rural/greater Christchurch area Other 224 (87%) 26 (10%) 7 (3%) Place of residence Home/independent living alone Home/independent living with spouse/partner/carer/family Home/independent living but not stated if alone or with family Rest Home resident Private Hospital resident Place of residence not noted 88 (34.2%) 101 (39.3%) 6 (2.3%) 42 (16.3%) 5 (2%) 15 (5.8%) Primary working diagnosis on admission Respiratory disease Cardiac disease Neurological condition Falls/collapse Confusion/Delirium Cellulitis Renal condition/urinary infection Malignancy Other 77 (30%) 38 (14.8%) 36 (14%) 29 (11.3%) 11 (4.3%) 6 (2.3%) 4 (1.6%) 2 (0.8%) 54 (21%) The primary working diagnoses are summarised as seen in Table 1. "Other" included medication side effects, gastroenteritis, anaemia, arthritis with associated pain, postural hypotension, hypoglycaemia, vertigo, gastrointestinal bleed and musculoskeletal problems. The majority of patients were over the age of 80 years (153, 59.5%) and there was an even number of males and females. Most patients lived within Christchurch (224, 87%) and were in independent living situations (195, 75.8%), either alone (88, 34.2%) or with at least one other (101, 39.3%). The most common working diagnosis on admission was respiratory disorders (77, 30%). Prior functional status in terms of mobility and sensory disturbance (i.e. vision and hearing disturbance) was poorly documented, with only 128 (50%) patients' mobility status recorded and only 63 (24.5%) patients with a comment about presence or absence of sensory disturbance. Past medical history of dementia/cognitive impairment was recorded in 28 (11.3%), see Table 2. Table 2. Premorbid function Mobility Independent Walking frame/stick/mobility scooter Immobile No comment made about mobility 52 (20%) 74 (29%) 2 (0.8%) 129 (50%) Hearing Hearing impairment No comment made 13 (5%) 237 (92.2%) Visual Visual impairment No comment made 31 (12%) 214 (83.3%) Dementia Past medical history of dementia/impaired cognition documented in admission note 29 (11%) Table 3. Cognitive state Mental state Previous mental state noted on admission 36 (14%) Present mental state Usual Worse Improved No comment 15 (6%) 35 (14%) Nil 207 (81%) Measure of previous mental state noted on admission 3MS*/MMSE** 10-point Mental State Exam General comment Nil Nil Nil 34 (14%) 223 (87%) Measure of present mental state noted on admission 3MS*/MMSE** 10 point Mental State Exam General Comment Nil Nil 15 (6%) 138 (53%) 104 (41%) Delirium Delirium included in diagnosis at discharge 8 (3%) * Modified Mini Mental State Examination; **Mini Mental State Examination. Previous mental status and whether the present mental status has changed from baseline was poorly documented, with 207 (81%) patients not having a comment made about this at all. While lack of collateral history may have contributed to this, in some cases there were comments made by other health professionals within the notes that documented premorbid functioning. It is not known whether or not these were noted by the documenting doctor. Present mental state was measured and documented in 153 (59%) patients. The most commonly used measure of mental status was a general comment (e.g. alert and orientated). Only 8 patients had delirium noted within the discharge diagnosis list. However there were another 10 patients whereby a description of delirium was made in the body of the discharge summary but the diagnosis was not listed. Of the admissions reviewed, there were nine inpatient deaths. None of these patients were noted to have altered mental state or confusion on their discharge diagnosis (although discharge summaries for deceased patients are often brief and only include cause of death rather than an extensive problem list for the admission). Discussion In keeping with other observational studies, the documentation by medical staff of cognitive state and impairment or acute delirium is inadequate.5 This was despite other healthcare workers involved in patient care documenting information suggestive of acute or long-standing cognitive decline. It is already acknowledged in the literature that acute delirium on acute medical wards is under-recognised.9 As mentioned previously, delirium is a treatable condition with an associated mortality, hence the importance of detection and intervention. It is also considered to be preventable.6 Delirium, when documented within the body of the clinical notes by a doctor, leads to additional funding for that acute episode of care. Patients suffering from delirium often require increased resources and nursing care, sometimes even one to one supervision, which comes at an added cost in both staffing numbers and time. Non-specific observations like "confusion" or "cognitive impairment" are not sufficient for coding purposes to be classified as delirium. For some patients, there was adequate information in the notes to suggest that they had an acute delirium but such a diagnosis was not specifically documented and therefore could not be coded as such. For the remaining patients, the presence or absence of delirium was poorly documented. There are many reasons for under-recognition and under-reporting of delirium. Acute delirious states can be either hyper or hypoactive and it is the former that is more widely recognised, partly because of the difficult behaviours associated with it. Hypoactive delirium is more likely to be overlooked and has a higher mortality rate associated with it.9 The fluctuating nature of a delirium can also make it difficult to diagnose. While delirium is recognised by medical staff, the documentation of it may be poor as the medical team focus more on the underlying acute illness. Despite the known increased mortality with delirium, it is not treated as a life-threatening illness.9 There are beneficial interventions for both the patient with early dementia and their caregivers.6 While an acute medical admission is may not be an appropriate forum to address previously undiagnosed cognitive impairment, suggestive of a diagnosis of dementia, it is an opportunity to refer the patient and their family for geriatric review as an out patient. Symptom progression can, in some patients, be delayed with cholinesterase inhibitors and both patient and care-givers can benefit from supports and interventions to help deal with behavioural problems associated with dementia.5 Cognitive impairment may be because of other treatable conditions like depression. Under-reporting of dementia and cognitive impairment in an acute medical admission is also multifactorial. Often, due to the busy nature of the acute medical wards, the primary medical problem is the main focus rather than chronic underlying problems. Cognitive function is not routinely tested on or during a medical admission and, as shown in this study, the admitting doctor's preferred tool is general observation. Weaknesses in this study include reliance on the patient for a full history as often a collateral history is not available at the time of admission, thus making it difficult to know if information is reliable and whether present mental state is usual for the patient. Even with full access to patients' hospital records, previous relevant medical history is not readily available. An example of this is the variable filing of formal cognitive testing on previous admissions. Included as part of the initial assessment of the patient was the consultant ward round note. This is written by another member of the team (which can vary from a fourth-year student to a senior registrar) and therefore may not include all the observations or comments made by the consultant. The prospective review was performed over a 2-week period in an attempt to give an overview of medical admissions and a reasonable sample size. However there will be a documentation bias as the night medical teams will be over-represented as they work a 7-night roster. The results of this study do not provide new information but helps to highlight the issue at a local level. This study was designed to measure how well cognitive function is measured and documented in the elderly at the time of admission. The results show that this is poorly done. There needs to be increasing awareness of the importance of such documentation, both for treatment and intervention reasons as well as funding purposes. Preferred tools would include the CAM (Confusion Assessment Method) and MMSE, along with collateral history, for acute delirium6 and collateral history and Modified Mini-Mental State Exam for dementia. Admittedly the 3MS can be quite time consuming during the acute admission process for the admitting doctor and more appropriately performed either later in the admission or when the patient is recovered from the acute illness in an outpatient setting. A targeted history and CAM assessment would be achievable during the first 24 hours of admission. Presenting this information locally may improve awareness and documentation at the Christchurch Hospitals. Further studies would be needed to measure both improvements in documentation and to consider if this leads to improvements in patient outcome.