This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

The Clock Drawing Test Is an Independent Predictor of Incident Use of 24-Hour Care in a Retirement Community

Department of Medicine, Division of Geriatric Medicine, University of Pittsburgh, Pennsylvania.

Address correspondence to Laurie Lavery, MD, Division of Geriatric Medicine, University of Pittsburgh, 3471 Fifth Ave., Suite 500, Pittsburgh, PA 15213. E-mail: laveryla{at}upmc.edu

	Abstract

Top Abstract Methods Results Discussion References

 
Background. It is unclear how early cognitive impairment affects future care needs. Furthermore, the Mini-Mental State Examination (MMSE), a commonly used screening tool in the clinical setting, tends to have a ceiling effect for early cognitive decline. One of the earliest changes in cognitive function is executive impairment. We examined the relationship between executive function, measured with a clock drawing protocol (CLOX1) designed to capture executive impairment, and incident need for increased level of care and total mortality.

Methods. Residents (n = 230) in independent living at a continuing care retirement community were followed for incident need for 24-hour care (mean 2.5 years). Baseline assessment included health status and physical and cognitive function. Time to event analysis was performed to determine the association of the CLOX1 score with the outcomes.

Results. Forty percent of residents had a CLOX1 score <12, and 10% had an MMSE score <26. The event rate for a CLOX1 score <12 was 30 per 100 person-years (p-y) and 13 per 100 p-y for a score 12. Similarly, the event rate was 34 per 100 p-y versus 17 per 100 p-y for MMSE <26 and MMSE 26, respectively. A CLOX1 score <12 was associated with a twofold higher risk of incident use of 24-hour care (hazard ratio 2.2; 95% confidence interval: 1.5–3.4) and death (hazard ratio 2.3; 95% confidence interval: 1.1–4.8) even after controlling for age, sex, comorbidity, and MMSE scores. The MMSE score was not an independent predictor of incident use of 24-hour care or mortality.

Conclusion. The clock drawing test, scored for executive impairment, but not the MMSE, predicted incident use of 24-hour care and mortality in this cohort of independent older adults.

The clock drawing test (CDT) has been examined as a screening tool for early global cognitive impairment (2–5). CDT scores have been shown to correlate with those of the MMSE (6), and when combined with the MMSE, the CDT may provide additional discrimination for cognitive decline and Alzheimer's disease (AD) (7). One of the earliest changes in cognition, regardless of underlying pathology, is an impairment of executive function.

Royall and colleagues (8) have developed a protocol for the CDT designed specifically to assess executive function, and it correlates well with other measures of executive impairment.

Given the importance of early changes in executive function in determining care needs, we hypothesized that the CDT, scored to assess executive impairment, would provide additional predictive value to the MMSE, as part of a multidimensional geriatric assessment. Specifically, we examined whether the CDT was an independent predictor of incident use of 24-hour care, including assisted living and nursing home care, in a retirement community. Because executive impairment is associated with common medical illnesses, we also hypothesized that this test would predict total mortality (9,10).

	METHODS

Top Abstract Methods Results Discussion References

 
Participants
Two hundred sixty-nine independent residents in a continuing care retirement community (CCRC) near Pittsburgh, Pennsylvania were followed from January 1, 1996 through December 31, 1999 to determine the date of first admission to long-term care. The CCRC was established in 1992 and includes independent living and an on-site, traditional nursing home. An assisted living facility was added in 1999. In 1996–1997, all residents in independent living underwent a comprehensive baseline assessment that was performed as part of their clinical care by geriatric nurse practitioners working in the medical clinic at the CCRC.

There were 269 residents in independent living at the CCRC at the time of the baseline assessment. On the basis of a retrospective chart review, a total of 39 residents were excluded from analysis. Of these 39 excluded residents, 13 either died during the initial assessment period or were already receiving 24-hour care in their home. Of the remaining 256 residents, 14 were either missing the CDT (n = 9), or testing was not performed due to blindness or deafness (n = 5). Only 12 eligible residents refused to complete the baseline assessment (12/256 = 4.7%). The remaining 230 residents formed the inception cohort for follow-up; none had any difficulty with activities of daily living (ADLs) at baseline.

Follow-up information was obtained by retrospective, anonymous abstraction of records only. Therefore, informed consent was not obtained. The study was approved by the Institutional Review Board of the University of Pittsburgh Medical Center.

Predictor Variable: CDT
Although the CDT was included in the baseline assessment, it was not scored or used clinically. For the purposes of this study and blinded to current level of care, we retrospectively scored the CDT using a modification of the CLOX protocol (8). The CLOX scoring consists of two parts, CLOX1 and CLOX2. CLOX1 is a free-hand drawing of a clock with the time set at 1:45. We assigned all clocks one point if they had a 3, 6, 9, and 12, although the original protocol is based on actual observation of order of placement. The CLOX2 test, a copy task, was not performed. CLOX1 scores range from 0 to 15, with higher scores indicating better performance. Because most scores were above the previously reported cut-point of 10 for dementia, we used the median CLOX1 score of <12 to define an abnormal score. We also examined the CLOX1 score as a continuous variable. Previously reported reliability data concerning the CLOX protocol include high internal consistency (Chronbach's alpha = 0.82) and high interrater reliability for the CLOX1 (r = 0.94). Interrater reliability for our study was high, based on a subset of 20 participants (kappa = 0.90).

Outcomes
Incident use of 24-hour care was the primary outcome and was defined as (a) transfer to assisted living, (b) transfer to the nursing home, or (c) need for 24-hour supervision in the home. Temporary versus permanent nursing home transfers could not be distinguished as this was not defined in the record, and subsequent movements to other levels (including acute care) were common. The need for intermittent, informal caregiving services in the home or intermittent skilled nursing care in the home was not included.

Outcomes were measured from the time of the baseline assessment until the end of the study on December 31, 1999. Time to event was measured in months from the baseline assessment to the first outcome. Average follow-up was 2.5 years. In a second analysis, death was used as the outcome variable.

Other Covariates
Demographics.-- This included age, sex, marital status, and race, and (in a subset of 40), education.

MMSE score (11).-- Scores range from 0 to 30 with higher scores indicative of better performance. As very few had any impairment, we used a cutoff of <26 (10th percentile) as an abnormal score and also examined the score as a continuous variable.

Geriatric Depression Scale score (12).-- The 15-item Geriatric Depression Scale-short version (GDS) was used to assess for depressed mood using the standard cut-point of 6.

Tinetti Gait and Balance score (13).-- The Tinetti Gait and Balance protocol was used as measure of physical function. Few residents had any difficulty, so we defined a score less than 39 of 40 as impaired physical functioning.

Charlson Index of Comorbidity (14).-- The Charlson Index is a validated measure of comorbidity. We defined a score of 0–1 to indicate low comorbidity and a score greater than or equal to 2 to indicate a higher comorbidity.

Medication inventory.-- A medication inventory was obtained as another measure of morbidity. All oral medications listed in the records at the time of the baseline assessment as prescribed for regular use were counted. Medication usage was divided into two groups with a cutoff of <5 medications as the low medication usage group, based on the median of the distribution.

For the purposes of analysis, all variables were also examined as continuous variables. The cutoffs are reported to provide a dichotomous, clinically relevant measure for each variable. Analyses were similar for the continuous and dichotomized variables.

Statistics
Descriptive statistics including mean and range of demographics and predictor variables were examined. Time–to-event analysis was performed using Cox proportional hazards modeling, censoring for death. The proportional hazards assumption was tested and found to be valid. Unadjusted and age- and/or sex-adjusted hazard ratios (HRs) for the CLOX1 score alone and with the MMSE score were obtained, with additional adjustment for all other confounders. Survival curves in those residents with normal and abnormal CLOX1 test scores were compared. All analyses were performed using the SAS statistical package for Windows (version 6.12; SAS Institute, Cary, NC).

	RESULTS

Top Abstract Methods Results Discussion References

 
The mean age at the time of the baseline assessment was 83.4 (±6.1) years, with a range from 69 to 99 years. Sixty-seven percent of the participants were women. Forty-six percent were married, and 47% were widowed. The residents' educational level was quite high. In a subset with educational level recorded, >90% had a college degree or graduate degree.

Overall, the residents had low comorbidity with 80% of residents with less than two points on the Charlson Index. Only 8% had scores on the GDS indicating depressed mood. Sixty-nine percent of residents had at least some impairment in gait and balance function by performance testing. Ninety-three residents (41%) had clock scores <12, whereas only 23 (10%) had MMSE scores <26.

The rates of incident need for 24-hour care for a CLOX1 score <12 was 30 per 100 person-years (p-y), whereas only 13 per 100 p-y for a score 12 with a crude HR of 2.3 and a 95% confidence interval (CI) of 1.6–3.4. Similarly, the event rate was 34 per 100 p-y versus 17 per 100 p-y (HR: 1.9; 95% CI: 1.1–3.2), comparing MMSE scores <26 and 26, respectively.

The CLOX1 and MMSE scores were each examined in adjusted Cox proportional hazards models. With adjustment for age and sex, a CLOX1 score <12 conveyed a higher risk of need for 24-hour care (Table 1). Although an MMSE score <26 alone was associated with a higher risk of needing 24-hour care, the HR was not significant when adjusted for age and sex.

The higher risk of incident use of 24-hour care with a CLOX1 score <12 is illustrated in the survival curve analysis (Figure 1). After 3.8 years, a higher proportion of residents with CLOX1 scores 12 maintained their independence. Fifty-five percent of residents with a CLOX1 score 12, but only 25% with a score <12, remained independent.

View larger version (21K): [in this window] [in a new window]   Figure 1. Survival curve analysis of maintenance of independence in residents with CLOX1 scores 12 and CLOX1 scores <12

	DISCUSSION

Top Abstract Methods Results Discussion References

The unadjusted HR for an MMSE score <26 was significantly associated with subsequent need for 24-hour care, but when controlling for other factors, MMSE score did not remain a significant predictor whereas a CLOX1 score <12 did remain a significant predictor of incident use of 24-hour care, even after controlling for all covariates. This finding suggests that the CDT, scored with a modified CLOX1 protocol, was a more sensitive measure of the risk of incident use of 24-hour care related to cognitive function than was the MMSE.

Clock drawing is dependent not only on intact visual spatial and constructional skills, but also requires intact executive functioning (15). Executive functions are higher level cognitive processes involved in working memory and planning tasks. These tasks include problem-solving, complex attention, strategy formation, and interference control. Deficits in executive function may result in difficulty adapting to a changing environment.

Executive function, measured in a variety of ways, has previously been related cross-sectionally to disability in older adults. Grisby and colleagues (16) showed an association between impairment in executive function and functional status by self-report and observed instrumental and basic ADLs. In a high-functioning cohort of women, impairment in executive function, particularly planning and initiating, was associated with slower performance on instrumental ADLs (17). Royall and colleagues (18) showed that the EXIT25 (a bedside measure of executive function) is also associated with current level of care in a retirement community. A subsequent study by Royall and colleagues (19) showed an association between several executive cognitive measures, including the CLOX, and level of care needs in noninstitutionalized older adults. Again, after adjustment for covariates, the CLOX1 was significantly associated with level of care whereas the MMSE failed to remain an independent contributor in this cross-sectional study.

Executive impairment may precede memory impairment in AD (20). A recent study by Chen and colleagues (21) suggested that impairment in executive function is one of the earliest manifestations of AD. In their study, scores on tests of executive function declined in the 2 years preceding the development of clinical symptoms of AD. In patients with a normal MMSE score, the CDT may differentiate those with and without executive function impairment (22). Taken together, these studies support the use of the CDT as a marker of early disability, presumably because intact executive function is required for independent living. Executive function is also adversely affected by poor health, such as falls, pain, certain medications, and hypoxemia (23–26). This may explain our finding of an association of the CLOX1 score with mortality, however, low cognitive function has also been shown to predict 5-year mortality after considering numerous markers of physical health status (27).

Whether other protocols of scoring the CDT would be independent predictors of functional decline has not been examined. Many clock-scoring protocols were designed to screen for early AD, not to measure executive function. Scoring protocols for the CDT other than the CLOX protocol have been shown to correlate well with the MMSE (6). No studies have been carried out comparing scoring methods for the CDT to predict functional outcomes.

In interpreting these data, it is important to recognize the high socioeconomic status and high level of education of our study cohort. This population was chosen for precisely this reason. This population is the most difficult to assess for early cognitive changes in the primary care setting, because of the ceiling effect of the MMSE. Use of a higher cutoff (26 or 27) for the MMSE score has been suggested for well educated populations (28). Despite using a higher cut-point for the MMSE in our study, the CLOX1 score remained a better predictor of incident use of 24-hour care. The CLOX1 score may or may not be useful in a more general population.

In our study, scores on both the CLOX1 and the MMSE tended to be clustered at the high end. A more extensive neuropsychological test battery could more fully distinguish cognitive function in this near normal range. Our goal, however, was to test the usefulness of brief, commonly used bedside measures that can be used in the primary care setting.

A limitation of our study is that the scoring of the clock drawings was performed retrospectively. This required some modifications in the CLOX scoring protocol as outlined. An additional limitation of this study is that we were not able to distinguish temporary from permanent placement. We used combined incident nursing home and assisted living admission or 24-hour home care which would be analogous to other studies where "incident disability" is the outcome and recovery is not assessed. Predicting who will need increased care services is important to the planning of the cost of care for the community. Future studies may focus on predicting permanent disability and establishing algorithms to determine level-of-care needs in a retirement community, and to evaluate interventions for high-risk individuals. Our study suggests that the CLOX protocol for scoring CDTs is useful in clinical geriatric assessment.

	Acknowledgments

 
Dr. Lavery was supported by the Hartford/American Federation for Aging Research Academic Geriatric Fellowship Program for this work.

	Footnotes

 
Decision Editor: John E. Morley, MB, BCh

Received December 18, 2003

Accepted April 12, 2004

	References

Top Abstract Methods Results Discussion References

 

1. Folstein MF, Anthony JC, Parhad I, et al. The meaning of cognitive impairment in the elderly. J Am Geriatr Soc. 1985;33:228-235.[Medline]
2. Wolf-Klein GP, Silverstone FA, Levy AP, Brod MS. Screening for Alzheimer's disease by clock drawing. J Am Geriatr Soc. 1990;37:730-734.
3. Watson YI, Arfken CL, Birge SJ. Clock completion: an objective screening test for dementia. J Am Geriatr Soc. 1993;41:1235-1240.[Medline]
4. Mendez MF, Ala T, Underwood KL. Development of scoring criteria for the clock drawing task in Alzheimer's disease. J Am Geriatr Soc. 1992;40:1095-1099.[Medline]
5. Sunderland T, Hill JL, Mellow AM, et al. Clock drawing in Alzheimer's disease. A novel measure of dementia severity. J Am Geriatr Soc. 1989;37:725-729.[Medline]
6. Juby A. Correlation between the Folstein Mini-Mental State Examination and three methods of clock drawing scoring. J Geriatr Psychiatry Neurol. 1999;12:87-91.
7. Ferrucci L, Cecchi F, Guralnik JM, et al. Does the clock drawing test predict cognitive decline in older persons independent of the Mini-Mental State Examination? J Am Geriatr Soc. 1996;44:1326-1331.[Medline]
8. Royall DR, Cordes JA, Polk M. CLOX: an executive clock drawing task. J Neurol Neurosurg Psychiatr. 1998;64:588-594.[Abstract/Free Full Text]
9. Degl'Innocenti A, Agren H, Backman L. Executive deficits in major depression. Acta Psychiatr Scand. 1998;97:182-188.[Medline]
10. Ferini-Strambi L, Baietto C, Di Gioia MR, et al. Cognitive dysfunction in patients with obstructive sleep apnea (OSA): partial reversibility after continuous positive airway pressure (CPAP). Brain Res Bull. 2003;61:87-92.[Medline]
11. Folstein M, Folstein S, McHugh PR. Mini-Mental state: a practical method for grading the cognitive state of patients for the clinician. Psychiatry Res. 1975;12:189-198.[Medline]
12. Sheikh JI, Yesavage JA. Geriatric depression scale (GDS): recent evidence and development of a shorter version. Clin Gerontol. 1986;5:165-73.
13. Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatr Soc. 1986;34:119-126.[Medline]
14. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373-383.[Medline]
15. Cahn-Weiner DA, Sullivan EV, Shear PK, et al. Brain structural and cognitive correlates of clock drawing performance in Alzheimer's disease. J Int Neuropsychol Soc. 1999;5:502-509.[Medline]
16. Grisby J, Kaye K, Baxter J, Shetterly SM, Hamman RF. Executive cognitive abilities and functional status among community-dwelling older persons in the San Luis Valley Health and Aging Study. J Am Geriatr Soc. 1998;46:590-596.[Medline]
17. Carlson MC, Fried LP, Xue Q, Bandeen-Roche K, Zeger SL, Brandt J. Association between executive attention and physical performance in community-dwelling older women. J Gerontol B Psychol Sci Soc Sci. 1999;54:S262-S270.[Abstract]
18. Royall DR, Cabello M, Polk MJ. Executive dyscontrol: an important factor affecting the level of care received by older retirees. J Am Geriatr Soc. 1998;46:1519-1524.[Medline]
19. Royall DR, Chiodo LK, Polk MJ. Correlates of disability among elderly retirees with "sub-clinical" cognitive impairment. J Gerontol A Biol Sci Med Sci. 2000;55:M541-546.[Abstract/Free Full Text]
20. Binetti G, Magni E, Padovani A, Bianchetti A, Trabucchi M. Executive dysfunction in early Alzheimer's disease. J Neurol Neurosurg Psychiatry. 1996;60:91-93.[Abstract/Free Full Text]
21. Chen P, Ratcliff G, Bell SH, Cauley JA, DeKosky ST, Ganguli M. Patterns of cognitive decline in presymptomatic Alzheimer disease: a prospective community study. Arch Gen Psychiatry. 2001;58:853-858.[Abstract/Free Full Text]
22. Juby A, Tench S, Baker V. The value of clock drawing in identifying executive cognitive function in people with a normal Mini-Mental State Examination score. CMAJ. 2002;167:859-864.[Abstract/Free Full Text]
23. Rappaport LJ, Hanks RA, Millis SR, Deshpande SA. Executive functioning and predictors of falls in the rehabilitation setting. Arch Phys Med Rehabil. 1998;79:629-633.[Medline]
24. Iezzi T, Archibald Y, Barnett P, Klinck A, Duckworth M. Neurocognitive performance and emotional status in chronic pain patients. J Behav Med. 1999;22:205-216.[Medline]
25. Coull JT, Frith CD, Dolan RJ. Dissociating neuromodulatory effects of diazepam on episodic memory encoding and executive function. Psychopharmacology 1999;145:213-222.[Medline]
26. Gale SD, Hopkins RO. Effects of hypoxia on the brain: neuroimaging and neuropsychological findings following carbon monoxide poisoning and obstructive sleep apnea. J Inter Neuropsychological Soc 2004;10:60-71.
27. Fried LP, Kronmal RA, Newman AB, et al. Risk factors for 5-year mortality in older adults: the Cardiovascular Health Study. JAMA. 1998;279:585-592.[Abstract/Free Full Text]
28. Kukull WA, Larson EB, Teri L, Bowen J, McCormick W, Pfanschmidt ML. The Mini-Mental State Examination score and the clinical diagnosis of dementia. J Clin Epidem 1994;47:1061-1067.[Medline]

This article has been cited by other articles:

				D. R. Royall Double Jeopardy Chest, December 1, 2006; 130(6): 1636 - 1638. [Full Text] [PDF]

				A.-I. Raffaele, C. Andrea, P. Claudio, T. Luigi, A. Domenico, S. Aldo, I. Orsola, and R. Franco Drawing Impairment Predicts Mortality in Severe COPD Chest, December 1, 2006; 130(6): 1687 - 1694. [Abstract] [Full Text] [PDF]

				H. H. Dodge, Y. Du, J. A. Saxton, and M. Ganguli Cognitive Domains and Trajectories of Functional Independence in Nondemented Elderly Persons J. Gerontol. A Biol. Sci. Med. Sci., December 1, 2006; 61(12): 1330 - 1337. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation