QJM Advance Access originally published online on January 14, 2008
QJM 2008 101(2):99-109; doi:10.1093/qjmed/hcm136
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Measuring case-mix and outcome for older people in acute hospital care across Europe: the development and potential of the ACMEplus instrument
From the 1The Catalan Agency for Health Technology Assessment and Research, Catalan Health Service, Barcelona, Spain, 2CIBER Epidemiología y Salud Pública (CIBERESP), Spain, 3Sheffield Institute for Studies on Ageing, University of Sheffield and Northern General Hospital, Sheffield, UK, 4The University of Aberdeen, Department of Medicine and Therapeutics, University of Aberdeen, Aberdeen, UK, 5Department of Medicine for the Elderly, Woodend Hospital, Aberdeen, UK, 6Department of Geriatric Medicine, Granollers General Hospital, Barcelona, Spain, 7Department of Geriatric Medicine, University Hospital Birmingham NHS Trust, Birmingham, UK, 8Istituto Nazionale di Riposo e Cura Anziani (INRCA), Ancona, Italy, 9Sextant Co. Athens, Greece, 10The University of Bialystok, Bialystok, Poland, and 11Health office, Turku, Finland
Address correspondence to M. Espallargues, The Catalan Agency for Health Technology Assessment and Research, Catalan Health Service, Barcelona, Spain. email: mespallargues{at}aatrm.catsalut.net
Received 6 July 2007 and in revised form 29 October 2007
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Aim: To identify case-mix variables measured shortly after admission to be included in a patient classification system (ACMEplus) that best explains hospital outcome for older people in different health care systems.
Design: Observational prospective cohort study collecting patient factors (sociodemographics, functional, mental, clinical, administrative and perceived health) at different time assessments.
Methods: Multicentre study involving eight hospitals in six European countries (United Kingdom, Spain, Italy, Finland, Greece and Poland). It included consecutive patients aged 65 years or older admitted to hospital for acute medical problems. Main outcome measures: discharge status, hospital readmission, mortality and length of stay.
Results: Of the 1667 included patients (mean age = 78.1 years; male gender = 43.5%) two-third had at least one ‘Geriatric Giant’ (immobility, confusion, incontinence or falls) on admission or shortly after. The most frequently affected system was cardiovascular (29.2%) and 31% of patients declared poor or very poor health. Mean length of stay was 17.9 days, 79% of patients were discharged to their usual residence; in-hospital and 1-month follow up mortality were 7.4% and 11.6%, respectively. Physical function explained the highest variation (between 8% and 21%), followed by cognitive status and number of Geriatric Giants, for almost all outcomes except readmission.
Conclusions: Factors other than diagnosis (physical function, cognition and presenting problems) are important in predicting key outcomes of acute hospital care for older people and are consistent across countries. Their inclusion in a standardized system of measurement may be a way of improving quality and equity of medical care in older people.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
In acute hospitals, a number of case-mix measurement systems or patient classification systems have been developed to predict resource use (Figure 1). Two of the most widely used instruments, Diagnosis Related Groups (DRGs)1 and the Healthcare Resource Groups (HRGs),2,3 attempt to define homogeneous groups of patients requiring similar amounts of resources (iso-resource groups)4 predominantly using length of stay in hospital as a proxy measure for resource use. However, their application to older people presenting with acute medical problems, with the aim of allowing clinical comparisons of units and the prediction of outcomes of hospital care, has been shown to be inadequate for two main reasons.5 First, they are based on the measurement of often single diagnoses and medical procedures which, although readily available in a computerized form from discharge summaries, have limited capacity to predict hospital outcome in this population and fail to reflect co-morbidity and additional measures of physical and cognitive function.6–8 Secondly, they rely on data collected retrospectively at discharge, whereas it would be desirable to make the assessment soon after admission if we want to improve patient care and outcomes.9
|
Patient classification systems based on direct evaluation of individual patients, while little used to date in acute care, have undergone careful development and refinement in rehabilitation and long-stay care centres. They were first envisaged as tools for the financing of the centres, but later their applications widened to include the management of centres, assessment of quality of care, staff allocation, control of access and formulation of national policies.4 Thus, they have tended to evolve from classifications based on the assessment of functional dependency to classifications progressively including variables related to clinical complexity, and finally to complex systems such as the Resource Utilization Groups (RUGs).10,11
Through the ACMEplus (Admission Case-Mix System for Elderly Patients) project we attempt to develop a new, brief, standardized patient classification system aiming to identify those case-mix variables, measured shortly after admission, which best explain variance in hospital outcome in older people entering hospital for acute medical problems in different European countries.12 The overall goal is to improve quality of care of older people, admitted to hospital as medical emergencies, by producing a robust case-mix instrument that the interested health workers can use to compare the clinical results of their own units, with those of other units elsewhere in Europe and beyond.
In this article, we describe how we identified, in Phase 1 of the ACMEplus project, the patient variables that might contribute to the case-mix system. We report findings about the predictive power of variables measured at 3 days following hospital admission, in terms of key outcome variables of care (death, institutionalization and length of stay), and also examine inter-centre differences. We discuss the potential use of these measures to improve patient care, resource management, audit and comparison of performance.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The ACMEplus project has gathered a large dataset relating to hospital admissions of medical patients aged 65 and over, by means of a 3-year multicentre study involving eight hospitals in six European countries (United Kingdom, Spain, Italy, Finland, Greece and Poland). The study was approved by the local research ethics committees. A ‘long list’ of potential case-mix and outcome variables was chosen as the result of an initial Consensus Conference of participating partners in December 2000, based on information from a systematic review of the literature,6 previous work of the partners on case-mix measurement for elderly patients in acute hospital care13–17 and knowledge of the different healthcare settings in the different centres.
The ‘long list’ was designed to include factors that might have a significant influence on outcome in older people entering hospital for medical reasons. Variables included physical function and activities of daily living (prior to and at the time of admission according to the Barthel Index of Activities of Daily Living, as modified by Collin et al., and administered by a health professional18,19), cognition (according to the six-item Orientation-Memory-Concentration test of Katzman et al. 20), depression,21 socioeconomic status (income, living arrangements, level of education, working status and social support), demographic details (age, gender and marital status), aspects of diagnosis and presence of ‘Geriatric Giants’ (as originally defined by Isaacs22: ‘Immobility, Instability (falls), Intellectual Impairment (confusion) and Incontinence’), perceived health status and feeling of loneliness.
In Phase 1 of the ACMEplus project, from March 2001 to September 2001, data was collected by means of an observational prospective cohort study, where each participating centre aimed to collect 200 consecutive cases. Under close supervision of the local research partners, a structured questionnaire was administered by trained hospital nurses, health services researchers or doctors on a broadly representative sample of patients aged 65 years or older admitted to hospital for acute medical problems. Patients admitted for planned investigations, terminal care or surgical procedures were excluded. In the four centres where local health services were organized so that older people could be admitted either directly to geriatric assessment units or through acute medical receiving units (Aberdeen, Birmingham, Barcelona and Turku), the sample was taken proportionally according to the different admission paths. In the remaining four centres (Sheffield, Ancona, Athens and Bialystok) all acute medical admissions of older people took place through general medicine or medical specialty units.
Patients were assessed on the third day of admission (the admission day being defined as Day 1) and followed up weekly until discharge at the end of their hospital episode, or until they had been in the hospital system for 90 days. They were also contacted 4 weeks after discharge by telephone interview. Data were collected from hospital notes, medical/nursing staff and the patient or his/her caregiver on sociodemographic characteristics, functional and mental status, clinical process and administrative variables and perceived health as identified previously in the ‘long list’.
The statistical analysis of Phase 1 data aimed to identify which of the potential case-mix variables in the ‘long list’ were statistically related to various measures of outcome, the latter including measures of discharge status, readmissions, mortality and length of stay. Chi-square test and logistic regression were used in case of binary outcome variables, with the Nagelkerke R2 statistic23 giving an indication of the amount of outcome variation ‘explained’ by the case-mix variables. Associations with length of stay were examined by linear regression with conventional adjusted R2 statistic (but using the natural logarithm of length of stay as the dependent variable because the untransformed variable was highly skewed). Those variables that ‘explained’ less than 5% (R2 < 0.05) of outcome variation in bivariate analysis were normally excluded from further consideration in multivariable analyses. Known universal ‘confounders’ such as age and sex were also considered and collinearity among variables was checked before their inclusion in the multivariable models. Regression models were evaluated through analysis of residuals and homoscedasticity. All P-values were for two-tailed contrasts with significance at the 5% level. Data were analysed using SPSS version 11.5.
A further Consensus Conference in January 2002 subsequently took into account the statistical analyses of Phase 1 data reported in the present publication, together with clinical and practical considerations (including acceptability, feasibility and reliability), to reduce the ‘long list’ of variables to a ‘short list’ for later data collection in Phase 2 of the ACMEplus project.12
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
In Phase 1 of the ACMEplus project the participating centres contributed data on a total on 1667 patients aged 65 years and over, admitted non-electively to hospital for medical reasons.
Table 1 describes the spread of main sociodemographic, hospital admission and health status characteristics of patients across centres. Some of the variables showed considerable inter-centre variation, especially gender, level of education, living alone, route of admission and type of ward (P < 0.001). Thus, while the majority of patients in the overall dataset were admitted via general medical or a medical speciality ward, in some centres, the majority were admitted directly into acute geriatric units (Aberdeen, Birmingham and Barcelona).
|
Two thirds of patients had at least one ‘Geriatric Giant’ on admission or during the first 3 days. Physical and cognitive function24 at Day 3 showed significant variation across participating centres (P < 0.001). As a broad indicator of diagnosis, in each patient the ‘main system affected’ was recorded at Day 3 using the main chapter headings from the 10th revision of the ICD (International Classification of Diseases). Cardiovascular disease (which included myocardial infarction, cardiac ischaemia and heart failure) resulted the most common category in the majority of centres, as shown in Table 2.
|
The main outcome variables in the eight centres are summarized in Table 3. Mean length of stay ranged from 12 days in Athens to 30 days in Birmingham although medians showed less variability due to the positive skewed distribution. At the end of the hospital episode, 79% of patients were discharged to the same place as on admission (i.e. to their usual residence), and the majority of them did not receive extra services (84%). Variability was also observed in the percentage of patients dying during hospitalization ranging from 2% in Athens to 12% in Bialystok. Although Barcelona showed a low rate of readmission 1 month after discharge, it presented the second highest accumulated death rate at 1-month follow-up (15.9%) after Bialystok (16.2%). Nevertheless, differences in case-mix between centres should be taken into account before interpreting these outcomes.
|
Of the list of potential case-mix variables recorded, those statistically associated with any of the six outcomes (death in hospital, death in hospital or in the month after discharge, discharge to the original residence from which admitted, discharge to different residence or still in hospital at 90 days, readmission within 1 month and length of stay) in bivariate analyses (P < 0.001) are shown in Table 4. This table presents the explained variance (R2) and the statistical significance obtained by logistic regression in the first five of these outcomes and by linear regression in the length of stay. Physical function explained a remarkable (and often the by far highest) amount of variation (between 8% and 20%) for almost all outcome variables. Cognitive status and the number of Geriatric Giants also showed high R2 for several outcomes. Other potential case-mix variables consistently explaining more than 5% of variation in outcome were ‘patient status’ at assessment (although status classification dealt with ‘process’ rather than being a true case-mix item) and main system affected. Difficulty in completing the cognitive questionnaire at Day 3 was highly correlated with death (in hospital or within 1 month of discharge) and low rates of discharge to the former residence, probably due to its correlation with poor physical and cognitive function. No variable explained more than 5% of variability in rates of readmission within 1 month. Like the ‘patient status’ variable, ward and centre were not true case-mix items, probably capturing different pathways of hospital care.
|
Variables not reaching our criterion for inclusion in Table 4 were estimates of physical status and Activities of Daily Living18,19 prior to (rather than at the time of) admission, duration of presenting problem prior to admission, the majority of social and socioeconomic conditions (marital status, current living arrangements, help at home, education, working status and occupation), overall health, the Geriatric Depression Scale21 and loneliness. Many of the potential predictor variables shown in Table 4 were intercorrelated. In particular, the Barthel Index (which measures 10 basic Activities of Daily Living and is thus closely affected by physical function), while usually emerging as the best single predictor of outcome in a number of multivariable models that we explored, was moderately correlated with many other variables including cognitive status, the number of Geriatric Giants and difficulty in answering questions.
While the above statistical analyses were a major consideration when the members of the second Consensus Conference decided which of the ‘long list’ of variables collected in Phase 1 of the ACMEplus project were to be retained in a ‘short-list’ for subsequent data collection, clinical judgement, local knowledge and practical experience gained during Phase 1 data collection were also taken into account. For example, the Rankin Scale of disability25 was excluded because inter-rater agreement depended heavily on the skills and experience of the interviewer, which could not be assumed if the item was later used in routine health care, and also because it was highly correlated with the Barthel Index. Also, the ‘reason for admission as stated by the referring doctor’ was excluded from further consideration, as it was often unavailable in practice, particularly in hospital systems where self-referral or presentation via an Accident and Emergency Department was the norm.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The ACMEplus instrument is intended to allow individual health workers to compare their own local hospital performance, in terms of outcomes of patients aged 65 and over admitted as medical emergencies, with that of other hospital units admitting similar patients elsewhere in Europe. The ultimate aim is, through such comparisons, to improve quality of care, and so the immediate goal has been to develop a broadly applicable set of case-mix indicators with the aim of categorizing patients into ‘iso-outcome’ groups (using the six measures of outcome listed in Table 4), as opposed to ‘iso-resource’ groups (where costs would be the main measure of outcome).
The recording of a range of outcome measures, rather than a single one, is also important because single outcome measures can be heavily dependent on a mixture of administrative and clinical factors, which can paint an unrealistic picture of a unit's performance. For example, a very short duration of stay with a low intra-hospital death rate might be considered an excellent outcome, but doubts would be raised if there were a high death rate in the month after discharge and/or a high readmission rate. The ACMEplus system is intended to be widely applicable to older people hospitalized non-electively for medical (as opposed to surgical, trauma or psychiatric) problems in a wide range of European settings and health services.
Our findings confirm the importance of factors other than diagnosis in predicting key outcomes of acute hospital care for older people, and these findings were broadly consistent across the eight centres included in the study. Other factors are likely to include variation in admission policies and the organization of care in the different centres, both within the hospitals and in post-acute care services in the community, as well as other variations in case-mix that we were unable to detect using the current range of variables. We recognize that additional variables might need to be added in the future for all, or some groups of patients. In particular, a more detailed assessment of the social and functional status of the patients prior to admission might well add value in individual cases, although the accuracy of such data is dependent on there being relatives or carers available to give the information, especially when a patient is cognitively impaired.
Early involvement of specialist teams for patients with old age-related needs is a key principle of geriatric medicine, particularly for those patients at high risk of institutionalization.26 There is debate about whether an age or a needs-based system should be used to identify those needing referral to old age specialist teams.27 Our study suggests that needs defined by physical function, cognitive function and atypical disease presentation (such as through a ‘Geriatric Giant’) are better than age at identifying patients at risk of institutionalization. Systematic measurement of these key variables might be used to identify patients for early referral to specialist teams.
Day 3 of the hospital admission (the admission day being defined as Day 1) was adopted by the Consensus Conference as the optimum day for data collection. It was judged that this was a time in the hospital admission when a degree of clinical stability would have usually been achieved, but where further input by health care professional in regard to medical care and rehabilitation was still likely to have a significant effect on hospital outcome. However, since some of the data (such as ‘Geriatric Giants’, age and gender) were potentially available on Day 1, there might also be the possibility of collecting a set of broad admission data items on Day 1.
Prospective payment systems have operated in hospitals in the USA for decades, creating a strong incentive to reduce length of stay. More recently, they have been introduced into European care systems. Our study suggests that physical function at Day 3 and the presence of one or more of the ‘Geriatric Giants’, are more powerful predictors of length of stay than diagnostic group, and adds weight to the argument that such variables should be included in case-mix grouping for resource management systems that involve prospective payment.28 As well as length of stay, death is obviously an important outcome measure for audit and comparison of performance in acute hospital care. Our results also suggest that physical function and cognitive status of patients measured at Day 3 should be taken into account when comparing mortality rates for older people following acute hospital care.
Differences in outcomes (and also process variables) between countries, after adjusting for case-mix, generate challenges for national policy makers.29 Such data raise questions for policy makers to address, challenging underlying assumptions about the efficiency and effectiveness of their care systems. Apparent disparities, however, should be interpreted with caution until case-mix factors have been fully standardized and differences in pathways of hospital care and in the health care systems of participating countries have been adequately addressed.
The introduction of standardized systems for assessing the health and social care needs of older people in primary care30,31 creates a further potential for adding key variables based on a direct observation of previous functioning to the case-mix classification of older people admitted to acute hospitals. Figure 1 shows the possible relationships of case-mix classification systems that could be integrated across the spectrum of care for older people.
Potential limitations of an ACMEplus instrument should be anticipated. Reliability of key variables used for data collection is a major determinant of its applicability in clinical practice. Studies of inter-observer reliability were carried out in a variety of settings representative of the different health care systems studied (Aberdeen and Birmingham, UK; Barcelona, Spain; Bialystok, Poland). As the majority of the questionnaire items were made up of standardized measurement tools already validated, only a part of the questionnaire was tested for inter-observer agreement (hospital admission details, social items and discharge details). The majority of items showed either very good or good agreement between data collectors, and there was an improvement in a subsequent data collection period.12
There has not been a formal process of cross-cultural adaptation of the complete questionnaire we used for data collection in the present study. Nevertheless, several items and scales were already validated in the participating countries (e.g. the Barthel Index, the Katzman's orientation-memory-concentration test, or the International Classification of Diseases version 10) and some others were translated and back-translated followed by a discussion between centres. Consensus within the project investigators and experts, achieved formally at Consensus Conferences, also ensured that the elements of the ACMEplus questionnaire have remained applicable across a wide range of health care systems and practice styles in Europe.
The burden of administration of an instrument might preclude its feasibility in clinical practice. Available results indicate that the ACMEplus questionnaire would not take long to administer, and it would be possible to complete it on Day 3 of the hospital admission in 5–10 min. Data collection would also involve little or no contact with the actual patient, as most of the necessary information should be available from hospital notes or from a health professional caring for the patient. This should also increase its acceptability for acutely ill or confused elderly medical patients.
Finally, the study showed a great variability of structures of geriatric care across countries32 and it is likely that different pathways of hospital care and diverse health care systems could explain some differences between participating centres. While patients from each centre were sampled broadly in line with local admission policies, these samples were not a random selection from all possible modes of hospital admission.
|
Conclusions |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
An instrument of the type that is being produced through the ACMEplus project offers the potential to improve individual patient care, to compare outcomes of acute hospital care within countries as well as differences in outcomes across European countries. Minimally, such an instrument should include the patient's functional and cognitive function and the presenting ‘Geriatric Giants’. In addition, it should meet at least three requirements to be conceptually and operationally appropriate to its main goal. First, it should be statistically acceptable (the instrument should be able to describe differences in patient outcomes measured as the amount of variance explained). Secondly, it should be clinically meaningful (patients from the same group should have similar clinical/functional characteristics). Thirdly, it should have known incentive consequences (its potential application in planning and managing health services requires caution in choosing the variables for building patient groups as the inclusion of any patient characteristic always provides implicitly positive or negative incentives33,34). Other key features to secure effective implementation raised at our Consensus Conferences were that any instrument should be acceptable and simple to administer, have the potential for computerized assessment, be integrated with community patient information systems, and that, once validated, should become a routine, compulsory system.
Following the production of a ‘short list’ of candidate case-mix variables as a result of the analyses and processes described in the present article, Phase 2 data collection has taken place in the ACMEplus project, and basic information about the Phase 2 dataset have been published12. Further work is required, however, in regard to more detailed analyses of both Phase 1 and Phase 2 data, together with wider questions of feasibility and acceptability of the process of data collection. In regard to the latter, it should be remembered that the ACMEplus project to date has used trained research personnel to collect data, and there is need for further work using data collected in a routine clinical setting in a wider range of healthcare systems throughout Europe. On-going international collaboration in the use of ACMEplus would be particularly worthwhile as a way of monitoring trends in case-mix and case-mix adjusted outcomes of acute hospital care amongst participating countries, and thus act as a stimulus for quality improvement.
|
Notes |
|---|
*In addition to the named authors, the ACMEplus project team consists of: Cari Almazan, Cristian Balducci, Ahmed Abdel Hafiz, Anna Hernandez-Cortes, Janusz Lewko, Maria Gabriella Melchiorre, Montse Moharra, Francesca Polverini, Costis Prouskas, Ramon Pujol, Sabrina Quattrini, Tapio Rajala and Judy Triantafillou.
|
Acknowledgements |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
European Commission (QLK6-CT-1999-02070); Departament d’Universitats Recerca i Societat de la Informació, Government of Catalonia (2003BE 00127 to M.E., 2004BE 00219 to M.E.) provided the funding for this study. We are indebted to the following colleagues for help in data collection: Carme Cabot, Ida Carrau, Giancarlo Cadeddu, Damiano Consales, Michelle Deighton, Paul Findlay, Paolo Fumelli, Sue Humphrey, Sue Hyde, Sylvia McDonald, Marta Millaret, Nello Panichi, Paquita Porras, Gabriele Saccomanno, Osvaldo Scarpino, Silvia Tantina, Karen Thompson, Giuseppe De Tommaso, Annamaria Viola and Lynne Walker. We would also like to thank Dr M. Bain, Dr J.A. García Navarro, Dr G. Lera, Prof. J. Levett, Dr A. Marcobelli, Mr P. Paunio, Dr A. Salvà and Ms B. Yandall for their participation in the Final Conference. The first author belongs to the Grup de Recerca en Avaluació de Serveis i Resultats en Salut (RAR) recognized by the Generalitat de Catalunya (2005SGR 00171).
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
1. Fetter RB, Shin Y, Freeman JL, Averill RF, Thompson JD. Case mix definition by diagnosis-related groups. Med Care (1980) 18(2 Suppl):iii, 1–53.
2. Sanderson HF. The use of Healthcare Resource Groups in managing clinical resources. Br J Hosp Med (1995) 54:531–4.[Web of Science][Medline]
3. Benton PL, Evans H, Light SM, Mountney LM, Sanderson HF, Anthony P. The development of Healthcare Resource Groups – Version 3. J Public Health Med (1998) 20:351–8.
4. García-Altés A, Martínez F, Carrillo E, Peiró S. Sistemas de clasificación de pacientes en centros de media y larga estancia: evolución y perspectivas de futuro. Gac Sanit (2000) 14:48–57.[Medline]
5. Sanderson HF, Storey A, Morris D, McNay RA, Robson MP, Loeb J. Evaluation of diagnosis-related groups in the National Health Service. Community Med (1989) 11:269–78.[Web of Science][Medline]
6. Campbell SE, Seymour DG, Primrose WR. A systematic literature review of factors affecting outcome in older medical patients admitted to hospital. Age Ageing (2004) 33:110–5.
7. Covinsky KE, Justice AC, Rosenthal GE, Palmer RM, Landefeld CS. Measuring prognosis and case mix in hospitalized elders. The importance of functional status. J Gen Intern Med (1997) 12:203–8.[CrossRef][Web of Science][Medline]
8. Inouye SK, Peduzzi PN, Robison JT, Hughes JS, Horwitz RI, Concato J. Importance of functional measures in predicting mortality among older hospitalized patients. JAMA (1998) 279:1187–93.
9. Daley J, Shwartz M. Developing risk-adjustment methods. In: Risk Adjustment for Measuring Health Care Outcomes—Iezzoni L, ed. (1994) Chicago, MI: Health Administration Press. 199–238.
10. Morris JN, Hawes V, Murphy K, Nonemaker S, Phillips C, Fries B. Resident Assessment Instrument Training Manual and Resource Guide (1991) Natik, MA: Elliot Press.
11. Morris JN, Hawes C, Fries BE, Phillips CD, Mor V, Katz S, et al. Designing the national resident assessment instrument for nursing homes. Gerontologist (1990) 30:293–307.[Abstract]
12. Campbell SE, Seymour DG, Primrose WR, Lynch JE, Dunstan E, Espallargues M, et al. A multi-centre European study of factors affecting the discharge destination of older people admitted to hospital: analysis of in-hospital data from the ACMEplus project. Age Ageing (2005) 34:467–75.
13. Findlay PF, Primrose WR, Gibbons Y, Seymour DG. A prospective study of factors associated with duration of stay in a department of medicine for the elderly. Age Ageing (2005) 27(Suppl. 2):60.
14. Dunstan E. Modified admission case-mix for the elderly. Age Ageing (2001) 30:531–2.
15. Dunstan EJ, Amar K, Watt A, Seymour DG. First steps in building ACME – an admission case-mix system for the elderly. Age Ageing (1996) 25:102–8.
16. Findlay PF, Seymour DG, Primrose WR, Gibbons Y. Pictorial outcome measures for the hospital care of older patients – a suggested toolkit. Age Ageing (2001) 30:27–32.
17. Perlado F, Midon J, Mesa P. ACME: an admission case-mix system for the elderly. Age Ageing (1999) 28:581–2.
18. Collin C, Wade DT, Davies S, Horne V. The Barthel ADL Index: a reliability study. Int Disabil Stud (1988) 10:61–3.[Medline]
19. Fillenbaum GG, Smyer MA. The development, validity, and reliability of the OARS multidimensional functional assessment questionnaire. J Gerontol (1981) 36:428–34.
20. Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a short Orientation-Memory-Concentration Test of cognitive impairment. Am J Psychiatry (1983) 140:734–9.
21. Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res (1982) 17:37–49.[CrossRef][Web of Science][Medline]
22. Isaacs B. Is geriatrics a speciality? In: Health Care of the Elderly—Arie T, ed. (1981) London: Croom Helm. 220–35.
23. Nagelkerke NJ. A note on a general definition of the coefficient of determination. Biometrika (1991) 78:691–2.
24. Royal College of Physicians and the British Geriatric Society. Standardised Assessment Scales for Elderly People (1992) London: Royal College of Physicians and the British Geriatric Society.
25. Rankin J. Cerebral vascular accidents in patients over the age of 60. I. General considerations. Scott Med J (1957) 2:127–36.[Medline]
26. Challis D, Clarkson P, Williamson J, Hughes J, Venables D, Burns A, et al. The value of specialist clinical assessment of older people prior to entry to care homes. Age Ageing (2004) 33:25–34.
27. Department of Health. The National Service Framework for Older People (2001) London: Department of Health.
28. Carpenter I, Bobby J, Kulinskaya E, Seymour G. People admitted to hospital with physical disability have increased length of stay: implications for diagnosis related group re-imbursement in England. Age Ageing (2007) 36:73–8.
29. Ham C, York N, Sutch S, Shaw R. Hospital bed utilisation in the NHS, Kaiser Permanente, and the US Medicare programme: analysis of routine data. Br Med J (2003) 327:1257.
30. Philp I. EASY-Care: A systematic approach to the assessment of older people. Geriatr Med (2000) 30:15–9.
31. EASY-Care [homepage on the internet]. Sheffield, The Centre for Healthy Ageing Sheffield Institute for Studies on Ageing. University of Sheffield [updated 2004 March; access 16 February 2005]. Available from: http://www.shef.ac.uk/sisa/easycare.
32. Primrose W, Campbell S, Seymour D. for the ACMEplus project. Pathways of hospital care for older people in six European countries. In: Proceedings of the 1st Congress of the European Union Geriatric Medicine Society (2001) Paris (France). 29 August to 1 September.
33. Portella E, Peiró S. La financiación de las instituciones de cuidados de media y larga estancia: una visión crítica. Todo Hosp (1994) 106:45–9.
34. Schneider D, Foley W. Case mix payment for long-term care. Bus Health (1987) 4:28–31.[Medline]
35. Schneider DP, Fries BE, Foley WJ, Desmond M, Gormley WJ. Case mix for nursing home payment: resource utilization groups, version II. Health Care Financ Rev (1988) ((Annual Supplement)):39–52.
36. Phillips CD, Hawes C. Nursing home case-mix classification and residents suffering from cognitive impairment: RUG-II and cognition in the Texas case-mix data base. Med Care (1992) 30:105–16.[CrossRef][Web of Science][Medline]
37. Schultz BM, Ward D, Knickman JR. RUG-II (Resource Utilization Group, Version II) impacts on long-term care facilities in New York. Health Care Financ Rev (1994) 16:85–99.[Web of Science][Medline]
38. Fries BE, Schneider DP, Foley WJ, Dowling M. Case-mix classification of Medicare residents in skilled nursing facilities: resource utilization groups (RUG-T18). Med Care (1989) 27:843–58.[CrossRef][Web of Science][Medline]
39. Cooney LM Jr, Fries BE. Validation and use of resource utilization groups as a case-mix measure for long-term care. Med Care (1985) 23:123–32.[CrossRef][Web of Science][Medline]
40. Fries BE, Cooney LM Jr. Resource utilization groups. A patient classification system for long-term care. Med Care (1985) 23:110–22.[CrossRef][Web of Science][Medline]
41. Mitty EL. Resource utilization groups. DRGs move to long-term care. Nurs Clin North Am (1988) 23:539–57.[Web of Science][Medline]
42. Fries BE, Schneider DP, Foley WJ, Gavazzi M, Burke R, Cornelius E. Refining a case-mix measure for nursing homes: Resource Utilization Groups (RUG-III). Med Care (1994) 32:668–85.[CrossRef][Web of Science][Medline]
43. Carpenter GI, Main A, Turner GF. Casemix for the elderly inpatient: Resource Utilization Groups (RUGs) validation project. Casemix for the Elderly Inpatient Working Group. Age Ageing (1995) 24:5–13.
44. Bjorkgren MA, Fries, Shugarman L. Testing a RUG-III based case-mix system for home care. Can J Ageing (2000) 19(Suppl. 2):106–25.
45. Eilertsen TB, Kramer AM, Schlenker RE, Hrincevich CA. Application of functional independence measure-function related groups and resource utilization groups-version III systems across post acute settings. Med Care (1998) 36:695–705.[CrossRef][Web of Science][Medline]
46. Stineman MG, Tassoni CJ, Escarce JJ, Goin JE, Granger CV, Fiedler RC, et al. Development of function-related groups version 2.0: a classification system for medical rehabilitation. Health Serv Res (1997) 32:529–48.[Web of Science][Medline]
47. Stineman MG, Shea JA, Jette A, Tassoni CJ, Ottenbacher KJ, Fiedler R, et al. The Functional Independence Measure: tests of scaling assumptions, structure, and reliability across 20 diverse impairment categories. Arch Phys Med Rehabil (1996) 77:1101–8.[CrossRef][Web of Science][Medline]
48. Tilquin C. PLAISIR: Planifications Informatisée des soins infirmiers requis-Guide de l’usager. In: Montreal, Quebec, Canada, Equipe de Recherche Operationelle en Santé (1993).
49. Fries BE. Comparing case-mix systems for nursing home payment. Health Care Finance Rev (1990) 11:103–19.
50. Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged. The index of ADL: a standardized measures of biological and psychosocial function. JAMA (1963) 185:914–9.
51. Katz S, Downs TD, Cash HR, Grotz RC. Progress in development of the index of ADL. Gerontologist (1970) 10:20–30.[Web of Science][Medline]
52. Morris JN, Sherwood S, May MI, Bernstein E. FRED: an innovative approach to nursing home level-of-care assignments. Health Serv Res (1987) 22:117–38.[Web of Science][Medline]
53. Armstrong-Esther CA. Long-term care reform in Alberta, Canada: the role of the resident classification system. J Adv Nurs (1994) 19:105–13.[CrossRef][Web of Science][Medline]
54. Semradek J, Hornbrook MC, McKenzie D, Giovannetti P, Charles C, Schalm C. Long-term care reform in Alberta, Canada. Alberta's resident classification system: fact, fiction and future prospects. J Adv Nurs (1994) 20:1182–5.[CrossRef][Web of Science][Medline]
55. Tilquin C, Michelon P, D’Hoore W, Sicotte C, Carrillo E, Leonard G. Using the handicap code of the ICIDH for classifying patients by intensity of nursing care requirements. Disabil Rehabil (1995) 17:176–83.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. M. Valderas, B. Starfield, B. Sibbald, C. Salisbury, and M. Roland Defining Comorbidity: Implications for Understanding Health and Health Services Ann. Fam. Med, July 1, 2009; 7(4): 357 - 363. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||