Printer Friendly

ASIM: a system for monitoring and evaluating the long-term care of the elderly and disabled.

The purpose of this article is to present the ASIM system for monitoring and evaluating the long-term care of the elderly and disabled, and to discuss its completeness, reliability, and validity using data from an application of ASIM in the Swedish municipality of Solna.

Results from different analyses that have been performed using data collected by the ASIM system and experiences from its application are presented in other articles by the same author (Lagergren 1992a, 1992b, 1993a, 1993b, 1993c). For a description of the Swedish public system of care for the elderly and the disabled, the reader is referred to Sundstrom (1987) or Thorslund (1991).

The purpose of the ASIM monitoring system is to provide long-term care management with basic information concerning the operations under their control. The system is thus not intended for individual care planning. The collected information is used only in the form of statistical aggregates.

The ASIM system can be seen as an application of the systems approach (West Churchman 1968). The system for providing long-term care services for the elderly and disabled can be structured both in terms of the intensity and character of the services and the needs of the clients. One way of structuring the services is by identifying different levels of care. Correspondingly, the clients can be divided into categories of need according to some measure. Each combination of level of care and measure of need constitutes a "cell" in the system of care. One way of describing the state of the system at a point in time is to count the number of persons in each cell.

The combined changes of levels of care and of needs will constitute flows between the different cells in the system. By monitoring these flows and the resulting distribution of the assisted persons it is possible to study the functioning of the care system, to identify and analyze allocation and resource problems, and to suggest solutions for improving the system's performance (Densen 1987).

Several applications of systems science thinking exist in the area of care of the elderly. Many of these are concerned with classifying patients and assessing dependency (cf. Bay, Leatt, and Stinson 1982; Cavaiola and Young 1980; Wenger 1986; Wilkin 1987)- often in connection with the determination of adequate staffing levels (Rhys-Hearn 1983; Green and Rayner 1985).

A vast amount of literature exists concerning measures of dependency and systems of patient classification in general; for a review see Kane and Kane (1981) or Brorsson (1980). One often used and well-documented measure, for example, is the Katz Index of Independence in Activities of Daily Living (ADL) (Katz, Ford, Moskowitz, et al. 1963; Brorsson and Hulter Ahsberg 1984). In the case to be described in this article, however, this index was not considered suitable -- mainly because of poor resolution for persons with minor deficiencies who constitute the majority of cases in the integrated medical and social care system (Spector et al. 1987; Travis and McAuley 1990).

Some studies have attempted to describe and analyze the dynamics of the care system. The different studies by Mackeprang and co-workers bear strong resemblance to the approach that is described in this article in that they are especially geared to the study of flows between levels of care and dependency (Mackeprang and Brauer 1977; Mackeprang and Bentzon 1978a, 1978b).

Most of the forementioned studies are ad hoc research projects without direct connection to policymaking and planning. In the case of routine systems for patient classification and dependency assessment the purpose as a rule is individual care planning. An exception to this is the British Balance-of-Care system (Boldy and Rhys-Hearn 1984; Boldy, Howell, and Smith 1979; Bowen and Forte 1988), which, like the system to be described here, is intended as a planning tool to be used by management on the local level.

MATERIALS AND METHODS

THE ASIM SYSTEM

The development of the ASIM system started in 1984 in Solna municipality near Stockholm, Sweden. Since 1989, the system has been applied on a routine basis in the municipality. Another application of the system started in 1987 in Sigtuna municipality, and the system is now operating routinely there as well.

The ASIM system represents an application of the general principles described at the beginning of this article. The care system is defined in this case as the public organization -- operated partly by the municipality and partly by the county council -- for providing long-term care services for the elderly and the disabled together with the clients using this care who live in the municipality. This organization consists of five different levels of care:

* Ordinary living with home-help services, home nursing, day care, or a home nursing allowance;

* Sheltered housing (service flats);

* Residential home/group unit living;

* Long-term health care at geriatric hospital/psychogeriatric ward/nursing home/institution for senile dementia; and

* Acute hospital care (only in the case where the patient actually does not need acute care but stays at the hospital waiting for some appropriate facility).

The clients are structured in two different ways according to their needs. The first way is to define the need category in terms of that level of care judged by the staff to be the most appropriate with regard to the person's needs. In this way, four categories are defined according to the first four levels of care just described. (The fifth category does not apply, since acute somatic care is not an appropriate level of care for long-term, chronically ill patients).

The other alternative for structuring the clients is to group them into categories of dependency according to a measure based on the assessment of five variables designed to describe different forms of disability (cf. further on).

The method of monitoring the state of the care system is by yearly surveys. Net flows between the different "cells" of the care system are calculated by matching subsequent surveys on the individual level using the patient's national registration number. The yearly survey focuses on the situation on a certain day. The registration and the assessments are made by the regular staff at the care unit where the client is registered.

METHOD OF REGISTRATION

Included in the survey are all persons, registered as living in the municipality, who during the day of the survey were receiving care or assistance on a long-term basis (actual or expected duration in excess of two months) at a care service unit organized by the municipality or the county council. At registration the following items are recorded:

* National registration number (from which age and sex can be determined), marital status, living arrangement (cohabitation or living alone);

* Present care unit (determining level of care) and amount of assistance (hours per week of home help, home nursing, etc.);

* Application (if any) to another level of care and time of application;

* Appropriate level of care and amount of assistance (hours per week) as assessed by the registering person;

* Social environment -- quality of housing, accessibility of housing, social support; and

* Disability -- functional disability (ADL and instrumental ADL), mobility disability, incontinence, insecurity, dementia.

The social environment variables as well as the disability variables are all measured on a scale from 0 to 3, where 0 means no problems and 3 maximum problems. The scale steps are defined in operational terms as much as possible (cf. Appendix). In contrast, no specific criteria have been developed concerning the "appropriate level of care," which shall reflect the overall view of the responsible staff and shall be based on their professional judgment.
The sums of the disability variables are used as an index of
dependency ranging from 0 to 15. From this a grouping into
dependency categories has been constructed -- the dependency
class -- according to the following definition:
Sum of Disability Dependency Class
 0-3 Low
 4-6 Medium
 7-9 High-medium
 10 + High


The choice of the disability variables was determined by the need for the same information system to cover, using a very limited set of data collected by the regular staff, a wide variety of clients, of which the majority belong to the least dependent category and rely mostly on social services.

PROBLEMS OF COMPLETENESS, RELIABILITY, AND VALIDITY

In the survey sequel, the problems of completeness, reliability, and validity are treated in that order. The analysis uses data from the Solna municipality only, where more effort has been put into data quality control.

Completeness. The first measure in achieving completeness of registration consists in checking that the staff is well instructed on the importance of including all clients and that forms are delivered by all units of care. However, since no total system existed for registering all clients in the municipality or the county council, it was possible to check completeness only indirectly.

During the development period 1984-1987, completeness was checked by combining each survey with the previous one and comparing them with the admission/discharge registrations during the intermediate period. (The latter registrations were cancelled in 1987). Using the national registration number for each person, a computer check was run to verify whether or not the "care chain" -- that is, the sequence of registrations starting at the first survey, proceeding with successive registrations of transfers from one unit of care to another, and ending at the second survey -- was complete and interlocking. This means, for example, that if a person was registered in a certain unit for the first survey and not registered at all for the second survey, the intervening period should have shown a discharge registration from the unit to "outside." Conversely, if the person's registration number showed up only at the second survey, there should have been an admission registration from "outside" at some point during the period to the unit where the person was registered.

For each fault that was detected in this way, a message was sent to the units involved requesting clarification regarding the missing link. This resulted in completions or corrections, either of the first or the second survey or of the flow registrations. Of course, it would still have been possible for a person to be missed, namely, if by extreme chance he or she was missed by both surveys and all transfers during the intervening period. Judging from the result of the completion procedure, however, it was estimated that fewer than 1 percent of the clients were still missing after the completion.

After 1987, this method no longer worked, since there was no discharge/admittance registrations. As a consequence, it must be estimated that a small percentage of the clients were lost in each survey.

Double registration is the reverse of failed registration. The reason for double registration is usually ambiguity concerning which of the units the client belonged to. Such cases are easily detected in computer processing depending on the system of national registration numbers. When double registration is detected a message is sent out to the involved units for clarification of the matter.

Reliability. No formal tests were made of reliability, for example, by letting different members of the staff assess the same clients. Such a procedure would not have been entirely appropriate since it would not have reflected the actual situation; assessment should be based on personal knowledge by the person responsible for the care. However, "quasi experiments" could be constructed using the involuntary double assessment that resulted from double registration. This double registration could have involved either different units of care (in which case the reason often would have been uncertainty of placement in connection with semipermanent transfer) or the same unit. The outcome of different tests of reliability according to the methods described will be given in the Results section.

Validity. Validity concerns the question of whether or not the chosen measures truly reflect the factors one wishes to measure. This means that validity must be checked against the purpose of the data collection. In the case of the ASIM system the purpose is to monitor the development of needs and resources in order to see if resources are adequate with regard to the needs of the served population and if they are being distributed in a proper way.

The "appropriate level of care" as assessed by the staff is obviously a measure of need. Whether it is valid or not depends on the professional ability and attitudes that lie behind these assessments. As is shown in the Results section, there is reason to believe that criteria differ among staff serving different levels of care and with different professional backgrounds. One way of testing the validity of appropriate level of care as a measure of need is to compare it with the allowance for home help and home health care in domiciliary care and sheltered housing. This is shown in the Results section.

The other measure of need used in the ASIM system, the class of dependency formed from five disability variables, should be somewhat less subjective since it is based to a greater extent on precisely formulated criteria. It is, however, a less direct measure of need for care since many factors other than disability or dependency -- for instance, access to informal support, housing conditions, and so forth -- may influence the decision to allot some form of care resources to a client. A comparison between the two measures of need is given in the Results section by showing the mean and standard deviation of the allotted weekly hours of home help and home health care in domiciliary care and sheltered housing given the appropriate level of care respectively the class of dependency.

RESULTS

In testing reliability the procedure mentioned in the previous section was used. Table 1A shows the assessment of appropriate level of care after TABULAR DATA OMITTED pooling all involuntary double assessments from all Solna surveys between 1984 and 1990. Uniformity of assessment turns out to be rather low: the frequency of hits is .59, including "near-misses" .78. The kappa-coefficient of agreement (calculated according to Cohen 1960) is .53. However, if the comparison is restricted to double registrations at the same registering unit, the result becomes much better. Frequency of hits increases to .77 (.90 if "near-misses" are included), and the coefficient of agreement increases to .69.

The method of comparing assessments at involuntary double registration has also been used to test the reliability of the five different disability measures and the class of dependency. Table 2 shows the correlation between the first and second observation at involuntary double registration, the frequency of hits and near-misses (|+ or -~1) and the kappa-coefficient of agreement. The comparison is restricted to registrations at the same unit.

As seen from Table 2 the agreement is quite good in this case, with kappa-coefficients around .70 and almost total agreement when near-misses are included. However, as was the case earlier, agreement becomes worse when double registrations on all levels are included. Frequencies of hits then fall to around .60-.70, with the lowest value for insecurity (.59) and the highest for incontinence (.69).

An important question in this context is whether there is any systematic TABULAR DATA OMITTED bias in the observations between different levels of care. Table 3 gives the mean values for the difference between the first and the second of the double registrations in the case when the first registration was made in domiciliary care. The standard deviations of the differences are all around 0.75 for the disability variables and around 1.50 for the "appropriate level of care."

From Table 3 it can be seen that clear systematic differences exist in the assessment of appropriate level of care, depending on which level of care the assessing staff employed. The difference compared to assessment in domiciliary care is significant for all three higher levels (p = .05), and in all cases the appropriate level tends to follow the actual.

For the individual disability variables the result indicates a significant difference of assessment compared to when assessment was done in domiciliary care in the case of incontinence for sheltered housing, insecurity for residential home, and dementia for long-term hospital care (p = .05). In all of these cases, staff at the higher level of care tends to reduce the problem as compared to staff at the domiciliary care level. For the sum of the disability variables none of the differences are significant (p = .05). However, the number of observations is very small in the cases of sheltered housing and residential home.

As mentioned in the previous section, the validity of the appropriate level of care as a measure of need can be studied by comparing it with the allotted amount of home help and home health care. Table 4A shows mean and standard deviation of the weekly hours of allotted care in domiciliary care and sheltered housing per appropriate level of care. The corresponding result for the class of dependency formed from the sum of the disability variables is shown in Table 4B. In both cases there is a very TABULAR DATA OMITTED clear, monotonous increase of the allotted weekly hours of care and assistance for increasing values of the appropriate level, respectively for the class of dependency. The differences are, of course, strongly significant.

An important question, when it comes to the choice of measures, involves how much particular measures independently contribute to information. Table 5 shows cross-correlations between the five disability variables, the index formed by their sum -- the class of dependency -- and the appropriate level of care.

As seen from Table 5 all these measures are correlated. The highest correlations are found between the two overall measures and between these measures and functional disability, incontinence, and dementia. In contrast, mobility disability and insecurity show much lower levels of correlation with the other variables.

It must also be decided whether all of the disability variables contribute to a combined measure of the need for care. Table 6 shows the result of stepwise multiple regression analysis using the appropriate level of care as dependent variable and the five disability variables as independent.

Table 6 shows that all five variables contribute significantly to the assessment of the appropriate level of care, with functional disability as most important followed by mobility disability and incontinence.
Table 4: Mean and Standard Deviation of Allotted Home Help and
Home Nursing (Hours per Week)
 Allotted Hours per Week
 Mean s.d. N
4A. Given the "Appropriate
Level of Care"
Domiciliary care 4.57 4.83 6470
Sheltered housing 8.01 10.0 1317
Residential home 11.59 8.99 662
Long-term hospital care 17.98 16.1 205
Pearson correlation coefficient: .39
4B. Given the Class of
Dependency
Low 3.69 3.14 5055
Medium 7.39 6.89 2706
High medium 12.93 14.2 730
High 16.85 16.4 179
Pearson correlation coefficient: .44
Source: Pool of six Solna surveys 1985-1990; domiciliary care
and sheltered housing.


TABULAR DATA OMITTED
Table 6: Stepwise Multiple Regression -- Appropriate Level of
Care versus Functional Disability, Mobility Disability,
Incontinence, Insecurity, and Dementia
 Variable Parameter Partial Model
 Entered(*) Estimate |R.sub.2~ |R.sub.2~
Intercept .67 - -
Functional disability .33 .41 .41
Mobility disability .15 .10 .50
Incontinence .25 .03 .53
Insecurity .17 .01 .55
Dementia .37 .01 .56
Source: Solna survey, November 1987, 2,127 observations.
* All variables contribute significantly at the .0001 level.


DISCUSSION

As mentioned in the introductory section, the purpose of the ASIM system is to provide information in support of the management of the care system -- not the management of individual cases. The collected information is used only in the form of statistical aggregates.

Ease of handling, minimum use of extra resources, and simple interpretation of results have been leading objectives in the development of the system. The information provided is intended to act as a signal light for the detection of problems and developments rather than as a complete tool for evaluating efficiency. To achieve the latter purpose more specific information collection, directed at the efficiency problem in question, is needed (Goldberg and Connelly 1982).

The basic problem when it comes to assessing the completeness of the data collection is that no absolute reference is available. Some loss is inevitable but in practice the problem is negligible; the low percentage that might be lost does not affect the main results. A more serious problem from a completeness point of view is that there may be persons in need for care, who for some reason are not receiving any services. Ideally, a monitoring system should also contain, on a sample basis, need assessments of the old or disabled who are not included in the system of care. Resources, however, have not permitted their inclusion in the ASIM system as applied in Solna and Sigtuna.

As mentioned in the Materials and Methods section, the survey procedure made it impossible to perform any formal reliability test. The relevant question then is whether the method of comparing double registrations as related earlier serves to simulate such tests in a satisfactory way.

When it comes to comparisons between different levels of care the result may be confounded by an intervening episode of acute hospital care (cf. earlier), which perhaps may change the status of the client. Since this would augment differences in assessment the analysis of double registrations should tend to underrate the reliability of the actual assessment procedure.

Double registrations at the same unit have sometimes been the result of a lack of communication between the district nurse and the home help assistant belonging to the same unit. Differences in assessment in this case would also reflect differences in professional bias. Another possibility is involuntary double registration by the same person simulating a test-retest situation. It is not possible to distinguish between these two situations from the data, because staff members were expected to do the assessment together and the unit often delivered the data without identifying who had assessed which of the clients. Comparing double registrations from the same unit would give a lower limit, however, to the test-retest reliability. Thus, in both cases the shown estimates of reliability ought to underrate the actual test-retest reliability as measured by a formal test.

Another aspect of reliability is whether some members of the staff may feel tempted to bias the assessment in order to achieve some objective, such as an increase of the staffing level at the care unit. The risk of this should not be underestimated. However, linking subsequent surveys makes it easy to detect abnormal developments in dependency, and if the staff is well aware of this possibility the temptation may be resisted.

Summing up, the reliability tests appear to show acceptable reliability taking into account the actual practical situation. No doubt, however, it would be possible to increase reliability by more intense supervising of the staff. The meaningfulness of this must be judged against the effort and the value of the improvement.

The validity of the chosen measures of need was analyzed by comparing them with the allotted weekly hours of home help and home health care in domiciliary care and sheltered housing. It was shown that increased values of both the appropriate level of care and the class of dependency -- based on the sum of the disability variables -- were associated with strongly significant increases in the average amount of allotted weekly care and assistance. Thus, it seems clear that the measures carry some validity when it comes to assessing need for care in a population.

The strength of the "appropriate level of care" concept is that it gives an overall assessment taking all aspects of the client's situation into account. Taken together for all clients it represents the staff's view of the need for resources on the different levels of care. Its weakness is that it is a subjective measure that varies among observers on different levels of care and also possibly over time.

The disability-based class of dependency measure has a more objective foundation (see Appendix) even if it was shown to have some observer variation. Its weakness is that many factors other than disability, such as access to informal support, housing conditions, and so on, may influence the need for care and assistance. By using the two measures together it is possible to achieve a more substantiated total picture of the need situation than each measure gives by itself. It was shown, however, that the two measures were rather strongly correlated (.73, Table 5), which ought to be expected. Also, each of the five separate disability variables contributed in a multiple regression analysis to the assessment of the appropriate level of care, which is an indication that they all carried an independent informational value in describing the client.

The ASIM system was developed in the context of the Swedish public system for long-term care and assistance for the elderly. A relevant question is whether the approach is restricted to that context or if it has wider applicability. No doubt the concrete form of the ASIM survey is directly developed to suit the needs of the management responsible for the care of the elderly and disabled in a Swedish municipality. However, the general idea of the ASIM system, which is to survey a population of clients on a regular basis using ordinary staff and to structure the clients and their needs in terms of the intensity and character of the allotted, respectively required resources of care, could be applied in any context of formal long-term care and assistance. A national registration number that makes it possible to link clients from one survey to the next, is of course a very important tool for more deeply analyzing the functioning of the care system. This is illustrated in further articles from the same project (Lagergren 1992b, 1993b). But even without such identification much can be learned just from observing changes between the regular surveys (cf. Lagergren 1992a).

From a routine monitoring system one can never expect the kind of completeness and reliability that would be required in an ad hoc research project. Nevertheless, it seems clear that for the stated purpose -- general monitoring of the system of care -- the ASIM monitoring system does give an acceptable result. The regular description of the balance between needs for care and resources that the ASIM system provides will, in any event, far exceed in quality the usual information available to management in the care of elderly.
APPENDIX
Definition of Scale Steps for Environmental and Disability
Variables: all variables are measured on a scale from 0 to 3
(0 = no problems; 3 = very grave problems)
Quality of Housing
3 = Very bad housing * Lacks at least four
of
 the qualities
mentioned
 below
2 = Bad housing * Lacks at least two of
 the qualities
mentioned
 below
1 = Good housing * Lacks none of the
 qualities mentioned
 below
0 = Housing adapted to personal needs * Qualities: cold
water,
 hot water, central
 heating, modern
cooker
 (electric/gas),toilet
 inside,shower and/or
 bath


Accessibility of Housing

3 = House accessible only to persons without mobility disability

2 = House accessible even to persons with a slight mobility disability (lift)

1 = House accessible even to persons with a relatively severe mobility disability, but not to persons confined to a wheelchair (lift, only a small number of steps)

0 = House accessible to persons confined to wheelchairs (lift, ramp)

Social Support

3 = Person entirely or almost bereft of social support

2 = Person receiving help from family or other social contacts at least once a week

1 = Person receiving help every day from family or other social contacts

0 = Person receiving help from a fit person living in the same household

Functional Disability

Assessment is based on the ability to manage:

* Weekly cleaning

* Shopping, managing mail and bank

* Buying food

* Small amount of washing

* Making the bed

* Taking a bath or shower

* Cooking

* Going to the toilet

* Dressing

* Shaving, washing face and hands

* Getting up/going to bed

* Eating/drinking

1 = Person entirely dependent on help for these activities

2 = Person needs help for the most part

1 = Person able to manage most parts without help

0 = Person able to manage all parts without help

Mobility Disability

3 = Person bound to a wheelchair or confined to bed

2 = Person able to move inside with the proper facilities (including wheelchair if the person can manage to get in or out of it and also can maneuver it), but not to walk outside

1 = Person can walk inside and outside the home with means of help or with the help of another person

0 = Person can walk without difficulty

Incontinence

3 = Person urine or feces incontinent

2 = Person severely urine incontinent

1 = Person slightly urine incontinent

0 = Person entirely continent Note: Catheter is indicated as 2.

Insecurity

3 = Extremely anxious or insecure

2 = Anxious or insecure

1 = Slightly anxious or insecure

0 = Not at all anxious or insecure

Dementia

3 = Suffers from extreme senility, entirely confused

2 = Moderately senile, often confused

1 = Somewhat unclear, sometimes confused

0 = Lucid

REFERENCES

Bay, K. S., P. Leatt, and S. Stinson. "A Patient-Classification System for Long-term Care." Medical Care 20, no. 5 (1982): 468-88.

Boldy, D., N. Howell, and B. Smith. Allocating the Home Help Services. London: Bedford Square Press, 1979.

Boldy, D. P., and C. Rhys-Hearn. Strategies for Assessing and Monitoring Appropriate Long-Term Care for Elderly People. Proceedings of the 3rd International Conference on System Science in Health Care. Berlin, Heidelberg: Springer Verlag, 1984.

Bowen, T., and P. Forte. "A Microcomputer System for Planning the Balance of Care for Elderly People." In Systems Science in Health Care. Proceedings of the 4th International Conference of the International Society for Systems Science in Health Care, Lyon, France, 4-8 July 1988. Paris: Masson, 1988

Brorsson, B. ADL-index: Sammanfattande matt pa individens formaga att klara det dagliga livets aktiviteter. Stockholm, Sweden: Medicinska Forskningsradet, 1980.

Brorsson, B., and K. Hulter Asberg. "Katz Index of Independence in ADL." Scandinavian Journal of Rehabilitative Medicine 16, no. 3 (1984): 125-32.

Cavaiola, L., and J. Young. "An Integrated System for Practical Assessment and Classification and Nurse Staff Allocation for Long-term Care Facilities." Health Services Research 15, no. 3 (1980): 281-306.

Cohen, J. "A Coefficient of Agreement for Nominal Scales." Educational and Psychological Measurement 20, no. 1 (1960): 37-46.

Densen, P. M. "The Elderly and the Health Care System: Another Perspective." The Milbank Quarterly 65, no. 4 (1987): 614-23.

Goldberg, E. M., and N. Connelly. The Effectiveness of Social Care for the Elderly. London: Heinemann Educational Books, 1982.

Green, T., and T. Rayner. "Hospital Care of the Elderly -- Modelling the Relationship between Direct Nursing Care and Patient Disability." International Journal of Nursing Studies 22, no. 3 (1985): 183-200.

Kane, R. A., and R. L. Kane. Assessing the Elderly. Lexington, MA: Lexington Books, 1981.

Katz. S., A. B. Ford, R. W. Moskowitz, B. A. Jackson, and M. W. Jaffe. "Studies of Illness in the Aged." Journal of American Medical Association 185 (21 September 1963): 914-19.

Lagergren, M. "Allocation of Care in an Area-based System for Long-term Care of the Elderly and Disabled." Submitted, 1993a.

-----. "Dependency Transitions in an Area-Based System of Long-Term Care for the Elderly and Disabled." Submitted, 1993b.

-----." De-institutionalization and Aging: Monitoring the Effects in an Area-Based System of Long-Term Care of the Elderly and Disabled." Submitted, 1992a.

-----. "The Residential Home -- What Difference Does It Make?" Scandinavian Journal of Social Welfare 1 (1993c):

-----. "Transfers between Levels of Care in an Area-Based System of Long-Term Care for the Elderly and Disabled." Submitted, 1992b.

Mackeprang, B., and E. Brauer. "Changes in Age and Need for Care among Patients in a Geriatric Institution during a Two-Year Period." Scandinavian Journal of Social Medicine 5, no. 2 (1977): 67-72.

Mackeprang, B., and M. W. Bentzon. "Minimum Need for Care in Geriatric Institutions." Scandinavian Journal of Social Medicine 6, no. 1 (1978a): 25-29.

-----. "Some Aspects Concerning Admittance to and Accommodation in Geriatric Institutions Housing Persons with Varying Degrees of Need for Care." Scandinavian Journal of Social Medicine 6, no. 1 (1978b): 59-62.

Rhys-Hearn, C. Staffing Levels and Standards of Nursing Care in Australian Nursing Homes. Proceedings of the International Conference SYSTED 1983. Montreal, Quebec, Canada: Systems Science Press Offices, 1983.

Spector, W. D., S. Katz, J. B. Murphy, and J. P. Fulton. "The Hierarchial Relationship between Activities of Daily Living and Instrumental Activities of Daily Living." Journal of Chronic Diseases 40, no. 6 (1987): 481-89.

Sundstrom, G. Old Care in Sweden -- Yesterday, Today and Tomorrow. Stockholm: The Swedish Institute, 1987.

Thorslund, M. "The Increasing Number of Very Old People Will Change the Swedish Model of the Welfare State ." Social Science and Medicine 32, no. 4 (1991): 455-64.

Travis, S. S., and W. J. McAuley. "Simple Counts of the Number of Basic ADL Dependencies for Long-Term Care Research and Practice." Health Services Research 25, no. 2 (1990): 349-60.

Wenger, G. C. "What Do Dependency Measures Measure? Challenging Assumptions." In Dependency and Interdependency in Old Age-Theoretical Perspectives and Policy Alternatives. Edited by C. Phillipsson, M. Bernard, and P. Strang. London: Croom Helm, 1986.

West Churchman, C. The Systems Approach. New York: Dell Publishing Co., 1968.

Wilkin, D. "Conceptual Problems in Dependency Research." Social Science in Medicine 24, no. 10 (1987): 867-73.
COPYRIGHT 1993 Health Research and Educational Trust
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lagergren, Marten
Publication:Health Services Research
Date:Apr 1, 1993
Words:5481
Previous Article:Factors affecting conversion rates to Medicaid among new admissions to nursing homes.
Next Article:Effect of prospective reimbursement on nursing home costs.
Topics:

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters