Elderly people are being hospitalized in acute-care settings in increasing numbers. A recent survey in a Rhode Island hospital revealed that 43.6% of all admissions were over the age of 65.
Thus, in order to meet the special needs of their patient population, general hospital staff nurses must develop knowledge and skills related to the care of the elder.
The results of the following project* suggest an approach for administrative, educational, and consultation staff to take when assessing elder patients' cognitive capacity in an acute-care setting.
Assessing patient cognitive capacity is a crucial factor in caring for the acutely ill elderly person for at least three reasons.
First, full cognitive capacity (the ability to process, store, and recall information about one's surroundings and life situation) is one of the major cornerstones of coping. Elderly patients who may be experiencing "transfer trauma," which disrupts cognition and increases mortality, or who may be in various stages of dementia, may be handicapped as they attempt to cope with the stress of acute hospitalization.1
Secondly, it may be speculated that without preventive interventions, the stress of hospitalization will overwhelm patients with mild cognitive deficits, possibly leading to further cognitive decompensation or behavioral difficulties.
Thirdly, nurses and case workers need to be aware of the patient's cognitive capacity when formulating discharge planning. Identifying patients' cognitive impairments is not an exercise in useless labeling, but rather the first step in identifying those patients who may need further assistance in coping with an acute hospitalization. The purpose of this study was to identify causes for staff avoidance in cognitive assessment.
It has been estimated that up to 20% of demented patients have dementia secondary to eminently treatable disorders.2 It has been shown that the level of an elderly patient's cognitive functioning can be maintained and even improved through the development of a facilitating and reorienting relationship with a nurse.3 Thus, both medical and nursing staff have complementary roles in this area of care.
Unfortunately, neither physicians nor nurses assess cognition in a regular and consistent fashion. Gehi, Strain, et al have documented that, while 33% of all patients on a general medical ward have cognitive deficits, only 17% are identified as having deficits.4 The identification of cognitively impaired elderly may be worse than the findings of Gehi, Strain, et al, which are not age specific. As the earliest manifestation of dementia is mild, recent memory loss, the patient who is in the early stages of dementia may not appear grossly confused.5 The patient may have compensated for the memory loss with established habits of daily living supported by a network of relatives and friends, and, thus, may not be perceived as having overt cognitive deficits upon entering an acute-care setting. As the patient appears cognitively intact during superficial interpersonal interactions, medical and nursing staff may not pursue formal assessment of cognitive capacity. Without formal assessment, the patient's need for preventive and remedial intervention related to cognitive capacity may not be identified.
At our facility, current informal nursing assessments performed on admission failed to identify a significant number of cognitively impaired patients over 65 years of age, when compared with the number of patients identified by a formal screening tool. One reason for this failure was that if patients appeared cognitively intact during superficial interpersonal interactions, then formal cognitive testing was not performed.
The study consisted of three phases: a comparison of the number of cognitively impaired elderly identified by a standardized mental status exam with the number identified by present nursing assessment techniques; an in-depth review of the nurses' notes in order to gather descriptive data that would clarify the process by which nurses either identified or failed to identify those patients identified by the screening test to have cognitive impairments; and a self-report questionnaire to determine the nurses' perceptions of their assessment techniques.
The first phase consisted of a sixweek period during which a standardized screening test for cognitive impairments, the Cognitive Capacity Screening Examination (CCSE), was administered by one of three examiners to all patients over 65 admitted to four general medical/surgical units within 24 hours of admission.4,6,7 Patients were told that we were studying admission assessment procedures and that we wanted to ask them questions related to concentration and memory. They were given the option to refuse participation in our project.
The CCSE has inter-rater reliability and is a valid measure of cognitive impairment when compared with a standard psychiatric evaluation performed at the same time by a psychiatrist proficient in consultation in the medical setting.6 In a review of the CCSE published in 1979, Kaufman et al found that while it has a high yield in detecting cognitive deficits resulting from metabolic disturbances and degenerative diseases such as Alzheimer's Disease, it is less sensitive in identifying cognitive deficits associated with structural deficits.7
The major limitation of the CCSE is that it is simply a screening tool that uses an arbitrary cutoff score to identify patients with cognitive deficits. Thus, in spite of the fact that a score of 20 has been empirically shown to generate the least total false-positive and false-negative results, false results may still be obtained. A CCSE score can only be considered in light of other aspects of patient evaluation. Likewise, as noted previously, a high CCSE score does not necessarily rule out significant focal lesions. Although the CCSE was not specifically designed for use in a geriatric population, there is no contraindication for its use in this population.
After the patient was administered the CCSE, the admission assessment sheet and the nurse's notes were reviewed for statements recording cognitive impairments. Only definitive statements were counted as documentation of cognitive deficits.
From the comparison of the CCSE scores and the nurses' assessment, four categories of patients were identified:
1. Undetected: Patients with significant CCSE scores but unidentified by nurses to have cognitive deficits;
2. Agree-deficit: Patients with significant CCSE scores and identified by nurses to have cognitive deficits;
3. Agree-no-deficit: Patients with insignificant CCSE scores and identified by nurses to have no cognitive deficits;
4. RN detected: Patients with insignificant CCSE scores but identified by nurses to have cognitive deficits.
The second phase of the study focused on gathering descriptive data that would clarify the process of nursing assessment. The type, frequency, and specificity of observations used to assess and describe a patient's mental status were recorded on a standardized chart review form. In addition, collaborative statements that are not part of a cognitive assessment but that may influence it were also recorded. Collaborative statements were defined as adjectives or observations that have either positive or negative connotations about a patient's:
1. self-care: (for example: ambulatory vs. bedfast; continent vs. incontinent);
2. social interaction: (for example: pleasant vs. irritable; trusting vs. suspicious);
3. emotional status: (for example, happy vs. sad; calm vs. labile); and
4. appearance: (for example: wellgroomed vs. unkempt; clean vs. dirty).
The descriptive data and collaborative statements were correlated with the categories of undetected; agree-deficit; agree-no deficit; and RN detected.
The final phase of the study consisted of developing a questionnaire to elicit a self-report of nursing assessment behaviors. This questionnaire was distributed to the nursing staff of the four medical/surgical units where the patient testing and chart review occurred. All nurses were encouraged to complete the questionnaire anonymously. Although there was no way to ensure that the nurses who would return the questionnaire would be those nurses whose notes we reviewed, we felt that the questionnaire would reflect how nurses perceive themselves performing cognitive assessments on those units from which our patient sample was drawn.
The self-report questionnaire consisted of three parts:
1. a survey of nurses' opinions regarding performing a cognitive assessment;
2. a test of nurses' knowledge regarding dementia and cognitive assessment; and
3. self-report of assessment behaviors.
Of 435 patients over 65 years of age admitted to the hospital during the sixweek period, 182 were administered the CCSE. The remaining 253 patients were unable to be examined for a variety of reasons, ranging from non-compliance to unavailability related to severity of illness or location at testing time. Of the 182 patients who were administered the CCSE, 117 or 64% received a score that did not suggest the presence of reduced cognitive capacity. The remaining 65 or 36% received a score that suggested the presence of reduced cognitive capacity. Only 18 or 28% of these 65 patients with significant CCSE scores were identified by the nurses to have cognitive deficits. Thus, 47 or 72% of those patients with a significant CCSE score were undetected by nursing assessment techniques.
There was no documentation of a formal mental status examination performed by any nurse within the 182 charts reviewed. Thus, documentation of cognitive deficits consisted simply of a check on the "confused" or "disoriented" line on the admission assessment sheet, or of descriptive phrases in the nursing notes. A review of these descriptive phrases revealed a number of weaknesses in the process of nursing assessment.
A review of the charts in the agreedeficit category revealed that the word "confused" was used by nurses to document cognitive deficits 83% of the time. However, the word "confused" appeared without behavioral description. In seven of the 15 charts using the word "confused," the nurses made a point of noting that the patient was confused, yet oriented, so it would seem that the word "confused" was used to refer to cognitive deficits other than disorientation. It is interesting to note that five of these seven patients were shown on the CCSE to be oriented, but demonstrated other cognitive deficits in concentration and memory.
The second most frequently used (44% of the time) word to document cognitive deficits was "disoriented."
However, the nurses only clarified in which sphere (person, place, time) disorientation occurred in half of the patients noted to be disoriented. Finally, although 17 of the 18 patients in the agree-deficit category were unable to remember more than two out of four objects in three minutes during the CCSE exam, the nurses only commented on memory in three patients' charts by using the word "forgetful" without documentation of how this forgetfulness affected the patients' behavior.
A review of the charts in the undetected category revealed that 87% of the time, the nurses' only comment regarding cognitive capacity was the word "oriented." In many instances (67%), these patients were indeed "oriented" during the CCSE exam, but demonstrated other cognitive deficits in concentration, memory or abstraction. Thus, over half the time, the nurses missed documenting the cognitive deficits identified by the CCSE because they did not assess cognition beyond noting orientation.
In 115 charts in the agree-no deficit category, the nurses ruled out cognitive deficits 90% of the time by. simply using the word "oriented" without further comment regarding other cognitive functions.
The two charts in the RN-detected category revealed that nurses documented cognitive deficits in one chart by noting an acute confusional episode with disorientation during the first night of admission and in the other chart by noting a memory problem as "forgetful." It is interesting to note that the items missed on the CCSE exam in both of these patients could have alerted the staff to the potential of cognitive deficits (for example, the forgetful patient received a CCSE score of 26, but missed all four points on a recent memory task).
The distribution of positive ( + ), negative ( - ), or absence (0) collaborative statements within the undetected, agree-deficit, and agree-no deficit categories can be seen in Figures 1 and 2. Only the distribution of collaborative statements regarding social interaction and self-care were statistically significant. The similarity of the higher percentages of positive collaborative statements regarding interaction and self-care in the undetected and agree-no deficit categories suggests that the nature of the statements influenced the nurses' assessments. The Figures also show a statistically significant association between the nature of the statements regarding interaction and self-care, and the correlation, or lack of it, of the nurses' assessments with the CCSE scores.
COLLABORATIVE STATEMENTS REGARDING SOCIAL INTERACTION
COLLABORATIVE STATEMENTS REGARDING SELF-CARE
While the majority of nurses (80%) completing the questionnaire responded positively to opinion statements focusing on assumption of responsibility for and attitudes toward performing a mental status exam, only 64% felt that there was enough time to perform such an assessment during the admission procedure. Only 55% of the nurses felt that physicians would believe a nurse's assessment. While some nurses did well on the knowledge survey, 67% incorrectly answered the true-false question regarding the earliest manifestation of dementia, and 47% incorrectly defined disorientation.
The self-report of assessment behaviors revealed inconsistent results so that it was unclear whether the nurses give or do not give cognitive assessment priority during the admission procedure.
Significance and Implications
Present nursing assessment techniques failed to identify a significant number of cognitively impaired elderly patients upon admission when compared with those identified by a standardized screening tool for cognitive deficits. We propose four factors as possible explanations for the nurses' failures to identify patients with cognitive deficits.
First, nursing assessment did not include formal cognitive testing. While indirect observations often provide data regarding a patient's mental status, subtle deficits may not be identified without formal testing. We had originally speculated that if a patient appeared cognitively intact during superficial interpersonal interactions, formal testing would not be performed. However, formal testing did not occur with any patient, even if the patient displayed behavior that suggested underlying deficits.
Lack of knowledge regarding the specifics of a formal mental status exam may have prevented nurses from performing a formal exam. While a majority of nurses felt that they had the appropriate knowledge and skills to perform a formal mental status exam, only 36% scored 85% or higher on the knowledge survey. In addition to this potential lack of knowledge, another factor that may prevent nurses from performing a formal mental status may be the low priority assigned to such an exam. Although a majority of nurses (over 80%) denied that low priority was a factor preventing cognitive assessment, they also said that they do not give cognitive assessment priority over other aspects of the admission procedure. Finally, in the opinion survey and in the self-report survey, a majority of nurses (over 80%) identified lack of time as a factor preventing formal cognitive assessment.
A second factor, which may explain the nurses' failure to identify patients with cognitive deficits, is the incompleteness of the assessment. In our study, the nurses failed to assess all functions of the patient's cognitive capacity. Fifty-five percent of those patients with deficits undetected by the nurses were missed, not because the nurses "incorrectly" assessed orientation, but because the nurses apparently did not assess other functions of cognitive capacity beyond orientation. In the self-report survey of assessment techniques, nurses acknowledge this failure to assess all functions of cognitive capacity. Even in those patients identified by the nurses and the CCSE to have cognitive deficits, the nurses commented very infrequently on a patient's memory.
The fact that nurses use indirect observations instead of a formal mental status exam does not explain the incomplete assessments. Behavioral observations gathered carefully over time may be valid indicators of cognitive deficits. According to the chart review, nurses were not gathering this data. One reason for the infrequency of comments regarding memory function may be a lack of appreciation for the importance of memory assessment in this age group. Sixty-seven percent of the nurses thought that disorientation rather than recent memory loss was the most prominent early manifestation of dementia.
A third factor associated with identification of cognitive deficits is that nurses may be less skilled at detecting less overt deficits. If one assumes that the lower the CCSE score, the greater the possibility that the cognitive deficit will be overt, then it would seem that those patients who were identified by the nurses had more overt deficits than those patients undetected by the nurses (the average CCSE score of the patients identified by nurses was 10; the average CCSE score of the patients undetected by nurses was 15). In addition, the broad global manner in which deficits were documented (that is, the use of the words "confused" and "disoriented" alone without behavioral descriptions) suggest that nurses may only assess global functioning in relation to cognitive capacity. Obviously, this is a less sensitive measure of cognitive capacity than specific formal testing.
Finally, a fourth factor that may explain the nurses' failure to identify patients with cognitive deficits is the observation that collaborative statements regarding social interaction and self-care may have biased the nurses' assessment. Patients in the undetected and agree-no deficit categories tended to be similarly described in terms of self-care and social interaction; that is, more positively. On the other hand, patients in the agree-deficit category were described as having more difficulty with self-care and fewer positive social interactions. This may support our speculation that if a patient appears to be superficially intact, nurses may assume mat the patient possesses full cognitive capacity and will not pursue formal testing.
As a result of this study, we feel that nursing assessment of cognitive capacity could improve with the routine use of a standardized assessment tool (such as the CCSE) during the admission procedure for the following reasons:
1 . the priority given cognitive assessment would increase if nurses are expected to use an assessment tool during the admission procedure;
2. the meaning of indirect behavioral observations would be clarified as quantitative and objective data are gathered;
3. bias based on relevant but inconclusive observations of selfcare and social interaction would be avoided;
4. assessment would be expanded beyond total global functioning to include specific data related to a variety of cognitive functions; and
5 . baseline data could be gathered in a quantitative manner so that comparisons over time would document any changes in the patient's cognitive capacity during the course of hospitalization.
When introducing the use of a standardized mental status exam to nursing staff, the use of clinical judgment in administering and interpreting the exam must be emphasized in order to avoid indiscriminate labeling of patients. In addition, an inservice program to improve assessment of cognitive capacity should include:
1. a discussion clarifying potentially conflicting values regarding cognitive assessment (target the gap between what staff say they do and what they actually do);
2. a discussion focusing on attitudes toward the elderly and cognition (target any therapeutic nihilism);
3. a lecture and discussion focusing oh dementia and delerium in the elderly (target misinformation and bridge knowledge gaps); and
4. a multidisciplinary planning session to promote multidisciplinary communication among nurses, physicians, and other health care providers within the system.
While our primary suggestion is to improve nursing assessment of cognition in the elderly through the use of a formal mental status exam, we do not mean to imply that indirect observation of a patient's behavior is not an equally important measurement of cognitive deficits as a formal mental status exam. On the contrary, such observations can often complement the quantitative results of a formal exam by describing in behavioral terms the impact of cognitive deficits on the patient's daily functioning. Thus, beyond challenging the neglect to perform a formal mental status exam on admission, and beyond challenging the assumption that a patient possesses full cognitive capacity because he demonstrates good social skills and self-care, another major implication of this study is a challenge regarding the use of such words as "confused" and "disoriented" without a quantitative documentation through a formal mental status exam and without a description of the patient in behavioral terms.
POSSIBLE BEHAVIORAL INDICATORS OF COGNITIVE DEFICITS
In Figure 3, we have a suggested list of behavioral descriptors that may be used in conjunction with a more formal mental status exam. This list is not meant to be exhaustive and is offered only as a partial list of potential behavioral descriptions.
While it is beyond the scope of this paper to discuss the process of a comprehensive assessment, it must be emphasized that a formal assessment tool such as the CCSE and indirect behavioral observations cannot be interpreted in isolation. To be accurate, cognitive assessment of the elderly must be performed within the context of a complete physical assessment; a review of the patient's cultural, social, educational, psychological, and intellectual makeup; an assessment of the patient's visual and auditory capacities; and a consideration of the current environment within which the patient is functioning.
With improved assessment through the use of a formal mental status exam and more precise behavioral descriptors, nursing interventions can be geared more specifically to a patient's needs. Once patients with mild deficits are identified, preventive nursing strategies to minimize stress via reorientation, repetitive and clear explanations, and behavioral memory aids can be developed. Potential difficulties with compliance and discharge planning can be targeted. Finally, the use of a standardized mental status exam such as the CCSE and exploration for causes of behavioral difficulties can reveal important information about the patient other than the potential existence of a cognitive deficit.
Limitations of The Study
The descriptive nature of this study imposes clear limitations on the significance of the discussion. First, the CCSE score was accepted as the only criterion for determining the presence or absence of cognitive deficits in our patient sample. As noted, the CCSE is only a screening tool that may have false-negative and false-positive results. In addition, an organic brain syndrome may fluctuate so that the patient may demonstrate different deficits at different times. Thus, it is possible that, in some instances, the nurse's assessment might have been correct. However, as the CCSE is a screening tool with an accepted level of inter-rater reliability and validity, we felt it was a reasonable standard with which to compare assessments of cognitive capacity performed by nurses.
Secondly, the discussion regarding the process by which nurses fail to identify a significant number of cognitively impaired elderly patients is only speculative. The four factors identified were only proposed as potential contributing factors. There was no attempt to prove causation. Other factors, such as therapeutic nihilism, may be as significant in contributing to the failure to identify cognitive deficits.
Finally, the discussion regarding the process of nursing assessment may only represent the process of assessment within our own institution. However, the problem of incomplete, unclear, and perhaps incorrect assessment of cognition has been documented elsewhere.4,5,8 As our institution is a teaching facility affiliated with a medical school and three nursing schools, and within which primary nursing is practiced, it may be argued that the results of the project are relevant to other institutions.
- 1. Aldrich CK, Mensakoff F: Relocation of the aged and disabled: a mortality study. J Am GeriatrSoc 1963; 2:185-194.
- 2. Katzman R: Early detection of senile dementia. Hosp Pract 1981; 16(6):61-64, 69-71, 75-76.
- 3. Robinson KD: Therapeutic interaction: a means of crisis intervention with newly institutionalized elderly persons. Nurs CUn North Am 1974; 9(l):89-96.
- 4. Gehi M, Strain J, Wiltz N, Jacobs J: Is there a need for admission and discharge cognitive screening for die medically ill? Gen Hosp Psychiatry 1980; 3:186-191.
- 5. Wang AJ: Organic brain syndrome in the elderly: a révaluation of concepts. Postgrad Med 1972; 5 1(2): 237-241.
- 6. Jacobs J, Bernhard M, Delgado A, Strain J: Screening for organic mental syndromes in die medically ill. Ann Intern Med 1977; 86:40-46.
- 7. Kaufman D, Weinberger M, Strain J, Jacobs J: Detection of cognitive deficits by a brief mental status examination: the Cognitive Capacity Screening Examination, a reappraisal and a review. Gen Hosp Psychiatry 1979; 2:247-253.
- 8. Ludwick R: Assessing confusion: a tool to improve nursing care. J Gerontol Nurs 1981; 7(8):474-477.
POSSIBLE BEHAVIORAL INDICATORS OF COGNITIVE DEFICITS