Integrated health information systems based on the RAI/MDS series of instruments

The Method for Assigning Priority Levels (MAPLe): a new decision-support system for allocating home care resources

BACKGROUND

Home care plays a vital role in many health care systems, but there is evidence that appropriate targeting strategies must be used to allocate limited home care resources effectively. The aim of the present study was to develop and validate a methodology for prioritizing access to community and facility-based services for home care clients.

METHODS

Canadian and international data based on the Resident Assessment Instrument - Home Care (RAI-HC) were analyzed to identify predictors for nursing home placement, caregiver distress and for being rated as requiring alternative placement to improve outlook.

RESULTS

The Method for Assigning Priority Levels (MAPLe) algorithm was a strong predictor of all three outcomes in the derivation sample. The algorithm was validated with additional data from five other countries, three other provinces, and an Ontario sample obtained after the use of the RAI-HC was mandated.

CONCLUSION

The MAPLe algorithm provides a psychometrically sound decision-support tool that may be used to inform choices related to allocation of home care resources and prioritization of clients needing community or facility-based services.

Standardizing assessment of elderly people in acute care: the interRAI Acute Care instrument

OBJECTIVES

To examine the frequency distributions and interrater reliability of individual items of the interRAI Acute Care instrument.

DESIGN

Observational study of a representative sample of older inpatients; duplicate assessments conducted on a subsample by independent assessors to examine interrater reliability.

SETTING

Acute medical, acute geriatric and orthopedic units in 13 hospitals in nine countries.

PARTICIPANTS

Five hundred thirty-three patients aged 70 and older (mean age 82.4, range 70-102) with an anticipated stay of 48 hours or longer of whom 161 received duplicate assessments.

MEASUREMENTS

Sixty-two clinical items across 11 domains. Premorbid (3-day observation period before onset of the acute illness) and admission (the first 24 hours of hospital stay) assessments were conducted.

RESULTS

The frequency of deficits exceeded 30% for most items, ranging from 1% for physically abusive behavior to 86% for the need for support in activities of daily living after discharge. Common deficits were in cognitive skills for daily decision-making (38% premorbid, 54% at admission), personal hygiene (37%, 65%), and walking (39%, 71%). Interrater reliability was substantial in the premorbid period (average kappa=0.61) and admission period (average kappa=0.66). Of the 69 items tested, less than moderate agreement (kappa<0.4) was recorded for six (9%), moderate agreement (kappa=0.41-0.6) for 14 (20%), substantial agreement (kappa=0.61-0.8) for 40 (58%), and almost perfect agreement (kappa>0.8) for nine (13%).

CONCLUSION

Initial assessment of the psychometric properties of the interRAI Acute Care instrument provided evidence that item selection and interrater reliability are appropriate for clinical application. Further studies are required to examine the validity of embedded scales, diagnostic algorithms, and clinical protocols.

The MDS Challenging Behavior Profile for long-term care

The objective was to construct a reliable and valid challenging behavior scale with items from the Minimum Data Set (MDS). Exploratory factor analyses of a sample of 656 nursing home residents yielded a 16-item Behavior Profile containing four internally consistent and valid subscales measuring conflict behavior, withdrawn behavior, agitation and attention seeking behavior (alpha range: 0.69-0.80). On a second dataset of 227 nursing home residents, internal consistency, inter-rater reliability and validity against the Behavior Rating Scale for Psychogeriatric Inpatients (GIP) were established. Internal consistency of the subscales ranged between 0.54 and 0.78. The overall inter-rater reliability of the items was 0.53 (kappa); of the scale it was 0.75 (ICC). The MDS Challenging Behavior Profile could potentially be an important contribution to existing clinical MDS-scales but additional studies on reliability, validity and usefulness are needed.

Using machine learning algorithms to guide rehabilitation planning for home care clients

Background

Targeting older clients for rehabilitation is a clinical challenge and a research priority. We investigate the potential of machine learning algorithms – Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) – to guide rehabilitation planning for home care clients.

Methods

This study is a secondary analysis of data on 24,724 longer-term clients from eight home care programs in Ontario. Data were collected with the RAI-HC assessment system, in which the Activities of Daily Living Clinical Assessment Protocol (ADLCAP) is used to identify clients with rehabilitation potential. For study purposes, a client is defined as having rehabilitation potential if there was: i) improvement in ADL functioning, or ii) discharge home. SVM and KNN results are compared with those obtained using the ADLCAP. For comparison, the machine learning algorithms use the same functional and health status indicators as the ADLCAP.

Results

The KNN and SVM algorithms achieved similar substantially improved performance over the ADLCAP, although false positive and false negative rates were still fairly high (FP > .18, FN > .34 versus FP > .29, FN. > .58 for ADLCAP). Results are used to suggest potential revisions to the ADLCAP.

Conclusion

Machine learning algorithms achieved superior predictions than the current protocol. Machine learning results are less readily interpretable, but can also be used to guide development of improved clinical protocols.

The K-nearest neighbor algorithm predicted rehabilitation potential better than current Clinical Assessment Protocol

OBJECTIVE

There may be great potential for using computer-modeling techniques and machine-learning algorithms in clinical decision making, if these can be shown to produce results superior to clinical protocols currently in use. We aim to explore the potential to use an automatic, data-driven, machine-learning algorithm in clinical decision making.

STUDY DESIGN AND SETTING

Using a database containing comprehensive health assessment information (the interRAI-HC) on home care clients (N=24,724) from eight community-care regions in Ontario, Canada, we compare the performance of the K-nearest neighbor (KNN) algorithm and a Clinical Assessment Protocol (the "ADLCAP") currently used to predict rehabilitation potential. For our purposes, we define a patient as having rehabilitation potential if the patient had functional improvement or remained at home over a follow-up period of approximately 1 year.

RESULTS

The KNN algorithm has a lower false positive rate in all but one of the eight regions in the sample, and lower false negative rates in all regions. Compared using likelihood ratio statistics, KNN is uniformly more informative than the ADLCAP.

CONCLUSION

This article illustrates the potential for a machine-learning algorithm to enhance clinical decision making.

Prevalence and predictors of sexual dysfunction in psychiatric inpatients

The authors examined the prevalence and predictors of sexual dysfunction in a sample of 3,717 psychiatric inpatients assessed with the Minimum Data Set-Mental Health Version 1 (MDS-MH 1.0). Sexual dysfunction was found to be less prevalent in inpatient psychiatry (17%) than is typically reported in community settings. Severe depression symptoms, use of antidepressants, and cardiopulmonary conditions emerged as powerful predictors of sexual dysfunction. More research is needed on the assessment and treatment of sexual dysfunction in psychiatric inpatients, particularly focusing on attitudes of assessors, patients, and interactions between medical, psychiatric, and medication characteristics.

Premature termination of treatment in an inpatient eating disorder programme

This retrospective study was conducted to explore rates, timing and predictors of two forms of premature termination of treatment (PTT) in an inpatient eating disorders programme: patient dropout (DO) and administrative discharge (AD). A chart review was conducted to obtain demographic, Eating Disorder Inventory-2 (EDI-2), and Resident Assessment Instrument-Mental Health (RAI-MH) data for 186 patients being treated for bulimia nervosa (BN), anorexia nervosa (AN), or eating disorder not otherwise specified (EDNOS). Overall, of the 37.6% of patients who terminated treatment prematurely, 22.1% of patients dropped out, and 15.5% of patients were administratively discharged. Time at which discharge occurred was found to be associated with the type of premature termination. The presence of DSM-IV Axis-I comorbidity was found to be the only factor associated with an increased risk of being administratively discharged. No factors were predictive of patients dropping out of treatment. The findings support the notion that AD and patient DO are different events that may have different factors influencing their rates and timing. Implications for future research and programme planning are discussed.

Stroke care in long-term care facilities in southwestern Ontario

PURPOSE

This study describes the current state of stroke care and rehabilitation in facility-based long-term care (LTC).

METHOD

LTC representatives, community partners (including physiotherapists and occupational therapists), stroke survivors, and family members were interviewed about stroke care provided in LTC facilities.

RESULTS

Limitations of the current system were identified including inadequate provision of therapy, unequal access to specialized rehabilitation, lack of staff training to care for the specialized needs of residents with stroke, and the lack of coordinated and consistent care.

CONCLUSION

This study identified challenges and barriers to providing optimal stroke care in LTC facilities. Recommendations for enhancing stroke care are suggested.

Relationship between nursing interventions and outcome achievement in acute care settings

The extent to which nursing interventions provided during hospitalization are associated with patients' therapeutic self-care and functional health outcomes was explored with a voluntary sample of 574 patients. Nurses collected data on patient outcomes at admission and discharge using the minimum data set (MDS) and the therapeutic self-care scale (TSCS). Research assistants audited charts for documentation of nursing interventions. The results indicated that nursing interventions aimed at exercise promotion, positioning, and self-care assistance predicted functional status outcome. Higher functional status outcome predicted therapeutic self-care ability at hospital discharge. The results demonstrate that nurses can use MDS and TSCS data on patient outcomes to gain insight into the effectiveness of their interventions.

Physiological Outcomes of an Internet Disease Management Program vs. In-person Counselling:A Randomized, Controlled Trial

OBJECTIVE

To compare physiological outcomes and satisfaction for followup care between an interactive diabetes internet program and Diabetes Education Centres.

METHOD

A randomized, controlled trial with outcomes of glycosylated hemoglobin (A1C), fasting blood glucose, total cholesterol, triglycerides (TG), high-density lipoprotein cholesterol, lowdensity lipoprotein cholesterol and patient satisfaction. Enrollment was staggered, with individuals assessed at baseline, 3, 6 and 12 months.

RESULTS

Fifty-seven participants completed the study (20 control, 37 internet). Physiological outcomes were not statistically different between the 2 groups. However, within-group comparisons demonstrated a significant improvement in the internet group's A1C, TG and satisfaction levels from baseline to 3 and 6 months (p<0.0452, p<0.0428 and p<0.0138, respectively).

CONCLUSION

Although the trial was limited in sample size, it yielded significant findings for diabetes management, within group. Further research in using the internet program in other applications for diabetes management is needed.

Measuring healthcare outcomes to improve quality of care across post--acute care provider settings

Post-acute care (PAC) occurs in a variety of settings-skilled nursing facilities (nursing homes), rehabilitation facilities, and home health agencies. To evaluate the impact of care processes on clinical outcomes and implement changes designed to improve outcomes, one must begin by measuring outcomes in a valid, reliable manner that allows for comparisons to reference or benchmarking data. Currently, several data sets exist in PAC settings for the purpose of outcome measurement. However, there is a need for comparable information across settings to ensure the quality and continuity of care. This article reviews various existing data sets used in PAC settings, examines ongoing projects to create a single set of measures, and suggests some directions for future research.

Long-term care funding in Canada: a policy mosaic

When Canada was founded, health care was delegated as a provincial responsibility. Although the federal government shares a portion of health care costs, it is not directly responsible for the planning, delivery, and governance of health services. The 1984 Canada Health Act set national standards for the provision of physician and hospital services, but it does not apply to home care and long-term care facilities. Consequently, each province has established a unique approach to long-term care, resulting in a health policy mosaic. This paper examines different approaches to funding long-term care with a particular emphasis on the impacts of regionalization and of the implementation of case-mix-based funding systems.

The Resident Assessment Instrument-Mental Health (RAI-MH): inter-rater reliability and convergent validity

An important challenge facing behavioral health services is the lack of good quality, clinically relevant data at the individual level. The article describes a multinational research effort to develop a comprehensive, multidisciplinary mental health assessment system for use with adults in facilities providing acute, long-stay, forensic, and geriatric services. The Resident Assessment Instrument-Mental Health (RAI-MH) comprehensively assesses psychiatric, social, environmental, and medical issues at intake, emphasizing patient functioning. Data from the RAI-MH are intended to support care planning, quality improvement, outcome measurement, and case mix-based payment systems. The article provides the first set of evidence on the reliability and validity of the RAI-MH.