As early as 1885, Sir Thomas Loch recognized the need for discharge planning and its pivotal role in continuity of care when he said, "There is a need for someone at the hospital to direct the patient, to represent the patient and his interests . . . someone instructed as to ail existing means of preventing and treating illness. He should help to make those who can become self-reliant or obtain help for those who need it or else the medical care may fail of its good purpose" (Sharmansky, 1984). Today, the standards of the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), the regulations of the Health Care Financing Administration, and the legislative mandates of the Omnibus Budget Reconciliation Act of 1986 speak to its importance ("Accreditation Manual," 1984; Birmingham, 1987; Omnibus, 1986).
Persons 65 years of age or older may benefit most from discharge planning because they make up 31% of hospital admissions and account for 42% of patient care days (Fowles, 1988). The literature suggests that the elderly are at higher risk for longer and more frequent lengths of stay than other adults (Anderson, 1984; Andrews, 1986; Cable, 1983; Goodling, 1985; Inue, 1981). Comprehensive discharge planning programs, including early identification of those at risk, can alter these statistics (Cable, 1983; Kennedy, 1987; Lamont, 1983; Naylor, 1990; Wachtel 1987).
Screening inventories to identify patients at risk have been available for many years (Cunningham, 1984; Inue, 1981; Kitto, 1985). Most consist of a list of criteria that, if met, flag patients in need of social service assessment. These criteria include such factors as the patient's age, the patient's residence, whether the patient lives alone, whether the patient has had repeated hospital admissions, and whether the patient is newly disabled. If the patient meets one or more of these criteria, the social worker begins planning for discharge.
Other tools, such as the Continence, Ambulatory, Age, Social Background and Thought Process (Glass, 1976) and the Katz Index of Independence in Activities of Daily Living (Katz, 1970), assess functional abilities and can be used to predict the discharge planning needs of the elderly. The Long-Term Care Information System includes an initial assessment of the patient that translates the discharge needs into services available to meet those needs (Falcone, 1981). Unfortunately, these tools are not specific to the needs of the elderly, nor are they comprehensive or practical for nurses to use at the bedside. We developed the BIaylock Risk Assessment Screen (BRASS) to overcome these problems.
THE BLAYLOCK RISK ASSESSMENT SCREEN
The BRASS, administered as part of the admission assessment, identifies patients at risk for prolonged hospital stay and in need of discharge planning resources. Ten characteristics of the patient are assessed (Figure). To use the screen, the clinician circles the assessment options that best match the patient. For functional status and behavior pattern, the clinician circles each option that matches the patient. For each of the other characteristics, the clinician circles a single option.
Each option has a numerical value that represents the degree to which the characteristic affects the need for discharge planning. Total scores, computed by summing all characteristics, range from 0 to 40. Low scores (less than 10) suggest that the patient has few needs for discharge planning and a low demand for discharge planning resources. The patient has a low risk factor index score (RIS). A moderate RIS (a score of 10 to 19) suggests that the patient's problems are more complicated and require extensive discharge planning resources, probably without institutionalization. A high RIS (a score greater than 19) suggests that the patient's problems are vast, require extensive discharge planning resources, and probably will involve institutionalization or rehabilitation.
The BRASS was developed as part of a discharge planning system with special concerns for patients 65 years old or older. The literature associated with the needs of the elderly served as the primary basis for identification and selection of content for the screen. The review of the literature and the first author's experience as a gerontological clinical nurse specialist resulted in identifying the following factors: age (Blake, 1984; FIoyd, 1987; Wachtel, 1987), functional status (Anderson, 1984; Andrews, 1986; Narain, 1988; Wachtel, 1987; Waters, 1987), cognition (Narain, 1988; Wachtel, 1987), social support (Davis, 1987; Wachtel, 1987), living situation (Roy, 1989), number of previous admissions/ emergency room visits (Anderson, 1984; Andrews, 1986; Cable, 1983; Goodling, 1985), and number of medical problems (Anderson, 1984; Andrews, 1986; Cable, 1983; Goodling, 1985).
We also included behavior patterns, mobility, sensory deficits, and number of medications because they were not part of either functional status or cognition but thought to be relevant to the elderly (Andrews, 1986). The first author's experience suggested that the greater the number of medications that the patient had taken, the greater the likelihood of noncompliance. The first author developed the assessment options and the weights associated with each.
The first author and two master's prepared gerontological clinical nurse specialists reviewed the screen for content validity. By consensus, they made revisions to produce the final screen.
A review of 10 randomly chosen patient records with known discharge planning outcomes aided in identifying cutting points for the three levels of risk. For each record, a total BRASS score was computed. We used the BRASS scores and the information about known discharge planning outcomes to establish the cut points: less than 10 - low risk with few discharge planning resources needed; 10 to 19 = moderate risk with multiple discharge planning resources needed; and greater than 19 = high risk with the patient likely to be discharged to a nursing home, extended care, or rehabilitation facility.
To evaluate the adequacy of the BRASS in identifying patients in need of discharge planning resources, we collected data for 3 months on all patients admitted or transferred to a 23-bed general medical nursing care unit in a 400-bed tertiary care teaching hospital. Within 48 hours of the patient's arrival on the unit, a staff nurse completed the BRASS and identified the risk index score.
We obtained scores on 206 patients between 16 and 85 years of age (average age 52; SD = 17); 33 patients were 65 years old or older. BRASS total scores ranged from 0 to 28 and correlated significantly with age (r = .40) and length of hospital stay (r = .5). Patients with low risk scores (less than 10) had an average age of 48 years (SD = 16) with an average hospital stay of 8 days. Those with moderate risk scores (between 10 and 19) had an average age of 58 years (SD = 14) and, on average, were hospitalized for 14 days. Those with the highest risk scores (greater than 19) had an average age of 69 years (SD = 11) with an average hospital stay of 19 days.
Both findings support the validity of the BRASS. We developed the BRASS to reflect the needs of the elderly as well as to identify those patients who might experience extended lengths of stay and, consequently, need discharge planning services and resources. In each risk category, the average number of days of reimbursed care for the patients was lower than the mean length of stay. For those with low risk scores, the average number of days of reimbursed care was 6; for those with moderate and high risk scores, it was 8 days. Because of the current cost of health care, reimbursed lengths of stay need to be considered in discharge planning programs. The data from this study suggest that aggressive discharge planning for patients with a moderate to high RIS may yield the greatest economic savings.
Twenty-six percent of the patients returned to the hospital within 2 months of discharge; most (19%) returned only once. Five percent of the patients had two readmissions; only 1% had three readmissions. Although this readmission rate seemed high, others have reported equivalent rates (Wennberg, 1989). No differences in readmission rates of patients with low, moderate, or high RISs existed.
Using length of stay as the criterion variable and all other variables as predictors, regression analysis yielded five statistically significant predictors: meal preparation and medication administration (options of functional status), confusion and appropriateness of behavior (associated with behavior pattern), and the body system affected by the illness. Because the patients had a great number of diagnoses, the diagnoses were grouped into one of 12 body systems. These five variables accounted for 30% of the variability in the BRASS scores.
We completed similar analyses on the data from those patients who were 65 years or older. These analyses must be regarded only as suggestive, however, because of the small number of patients in this age group (n = 33). For this group of patients, the difference in length of stay associated with RIS approached significance (F = 2.76; df=2,29; p = .07). Seventeen of the patients had RISs of less than 10; 11 had RISs of 10 to 19; and only four had RISs greater than 19. Ability to prepare meals, the only significant predictor of length of hospital stay, accounted for 48% of the variability in length of stay.
The first author conducted an independent assessment of 49 patients whom staff nurses had assessed. These patients were between 43 and 85 years old (mean = 69 years; SD = 8), had a variety of medical diagnoses, and had BRASS total scores between 2 and 28 (average score = 19; SD=I). Fourteen staff nurses completed the BRASS on one to nine patients. The BRASS took about 15 minutes, and nurses had no difficulty completing the tool correctly. The BRASS scores obtained by the staff nurses correlated well (r = .84) with those obtained by the senior investigator. Consistency among raters was also good (intraclass correlation coefficient =.87). These data indicate that the BRASS has good reliability.
Blaylock Discharge Planning Risk Assessment Screen
DISCUSSION AND IMPLICATIONS
These results suggest that the BRASS can identify those patients needing discharge planning resources because it is valid and staff nurses can use it reliably. It is simple, quick, and bedside-friendly. Furthermore, staff nurses completed the assessment without difficulty after only minimal training. Thus, the BRASS overcomes the problem of user-unfriendliness that many other screening inventories have.
Neither dependence in meal preparation nor medication administration has been identified in previous studies as especially important in assessing discharge planning needs. However, studies have associated functional decline in activities of daily living and physical ability with higher incidences of nursing home admission and increased length of stay (Narain, 1988; Wachtel, 1987; Waters, 1987. Both meal preparation and self-medication administration could be considered as activities of daily living. The results of this study suggest that meal preparation and selfmedication administration play a more important role than previously thought and should be included in the assessment for discharge planning.
An informal review of selected patient records suggested that those patients with difficulty in meal preparation or self-medication administration had higher readmission rates. Knowing that a patient has difficulty with meal preparation and selfmedication early in the hospital stay permits the staff to identify ways to meet the patient's needs more effectively following hospitalizauon. Common self-medication problems among these patients included not knowing how to take as-needed medications (patients tended to take them routinely or not take them until their symptoms were very severe), and taking the same medication as if it were two different medications, one prescribed prior to admission under a generic name and the other prescribed on discharge under a brand name. The hospital nurse could evaluate both problems and begin appropriate education. Discharge medications could be identified and prescribed more appropriately with knowledge of preadmission medications. Preparation for home selfmedication could include information on medicating with both prehospital and discharge medications.
Because nurses administer the BRASS shortly after the patient's admission, they can use it to plan both hospital and posthospital care. For example, assessment of the patient's cognition shortly after admission permitted nurses in this study to use this information in making observations about the effects of hospitalization on the patient. The nurses easily identified changes in cognition during the hospital stay and explored contributing factors. They invariably identified and implemented interventions appropriate to the patient's needs.
Reports from staff nurses participating in this study suggested that they delivered more effective care when BRASS data were available on patients. After using it in planning for discharge, they found that they started ambulation or mobility programs earlier, made referrals to other providers earlier in the hospitalization, and prepared for home management earlier.
The BRASS does identify those patients in need of discharge planning resources. The nurse can use the data it provides to improve the care outcomes for patients while in the hospital and in their transition to home care. Thus, it shows promise for use as the first phase of a discharge planning program. Its implementation as part of an admission assessment meets JCAHO's requirement for a mechanism to identify those at risk and in need of discharge planning resources. Given the nature of the items, its results can provide useful initial information to those on the hospital health-care team as well as those who plan for or provide posthospital care.
Research is needed to evaluate the effectiveness of the information that the BRASS provides and discharge planning teams use. Especially for those patients at moderate to high risk, the effects of early assessment and subsequent planning for care on such outcome variables as ease of transition from hospital to posthospital care, ability to manage posthospital care (especially medication and nutrition), and hospital readmission rates need to be identified.
- Accreditation manual for hospitals. Chicago: Joint Commission on Accreditation of Healthcare Organizations, 1984.
- Anderson, G.E, Steinberg, E.P. Hospital readmissions in the Medicare population. N Engl J Med 1984; 311:1349-1353.
- Andrews, K. Relevancy of readmissions of elderly patients discharged from a geriatric unit. J Am Geriatr Soc 1986; 34:511.
- Birmingham, J. Quality discharge planning requires timely documentation of all phases. HPR 1987; Spring:3-5.
- Blake, R. What disables the American elderly. Generations 1984; 8(4):14-17.
- Cable, E.P., Mayers, S.P. Discharge planning effect on length of hospital stay. Arch Phys Med Rehabil 1983; 64:57-60.
- Cunningham, L.S. Early assessment for discharge planning: Adopting a high-risk screening program. QRB 1984; 10:561-565.
- Davis, C., Moore, D. Interdisciplinary discharge planning in the elderly. Little Rock, AR: University of Arkansas for Medical Sciences, 1987.
- Falcone, A.R. Using the Long-Term Care Information System for effective community care. Home Health Care Services (Quarterly 1981; 2:83-97.
- Floyd, J., Buckle, J. Nursing care of the elderly: The DRG influence. Journal of Gerontological Nursing 1987; 13(2):20-25.
- Fowles, D.G. A profile of older Americans. Washington, DC: Program Resources Department of the American Association of Retired Persons and Administration on Aging, US Department of Health and Human Services, 1988.
- Glass, R.I., Weiner, M.S. Seeking a social disposition for the medical patient: CAAST, a simple objective clinical index. Med Care 1976; 14:637-641.
- Goodling, J., Jette, A. Hospital réadmissions among the elderly. J Am Geriatr Soc 1985; 33:595-601.
- Inue, T.S., Stevenson, K.M., PIorde, D., Murphy, I. Identifying hospital patients who need early discharge planning for special dispositions: A comparison of alternative techniques. Med Care 1981; 19:922-929.
- Katz, S., Downs, T.D., Cash, H.R., Grotz, R.C. Progress in development of the index of ADL. Gerontologist 1970; 10:20-30.
- Kennedy, L., Neidlinger, S., Scroggins, K. Effective comprehensive discharge planning for hospitalized elderly. Gerontologist 1987; 27:577-580.
- Kitto, J., Dale, B. Designing a brief discharge planning screen. Nursing Management 1985; 169):28-30.
- Lamont, C.T., Sampson, S., Matthieas, R., Kane, R. The outcome of hospitalization for acute illness in the elderly. J Am Geriatr Soc 1983; 3:282-288.
- Narain, R., Rubenstein, L.Z., Wieland, D., Rosbrook, B., Stiome, S., Pietmszka, F., et al. Predictors of immediate and 6-month outcomes in hospitalized elderly patients: The importance of functional status. J Am Geriatr Soc 1988; 36:775-783.
- Naylor, M.D. Comprehensive discharge planning for hospitalized elderly: A pilot study. Nurs Res 1990; 39:156-161.
- Omnibus Budget Reconciliation Act. Washington, DC: US Government Printing Office, 1986.
- Roy, A.W., Folman, S.J., Ford, A.D. The elderly and risk factors for institutianalization: Evidence from the Cleveland GAO study. Cleveland: Case Western Reserve University, 1989.
- Sharmansky, S., Boase, J., Horn, B. Discharge planning: Yesterday, today, and tomorrow. Home Health Care Nurse 1984; 2(3):14-18.
- Wachtel, T.J., Fulton, J.P., Goldfarb, J. Early prediction of discharge disposition after hospitalization. Gerontologist 1987; 27:98-103.
- Waters, K.R. Discharge planning: An exploratory study of the process of discharge planning on geriatric wards. J Adv Nurs 1987; 12:783.
- Wennberg, J.E., Freeman, J.L., Shelton, R.M., Bubola, T.A. Hospital use and mortality among Medicare beneficiaries in Boston and New Haven. N Engl J Med 1989; 321:1168-1173.