Statistics from Altmetric.com
Medline (1996–2003), CINAHL (1982–2003), HealthSTAR (1975–2003), bibliographies of relevant reviews, and content experts.
Study selection and assessment:
randomised controlled trials (RCTs) that assessed the effects of hospital based case management on length of stay or readmission rate in inpatients ⩾18 years of age. Studies of patients with mental illness, outpatient services, or hospital-to-community based or community based case management were excluded. Methodological quality of individual studies was assessed using the 5 point Jadad scale; intervention quality was assessed based on the presence of 6 components: comprehensive assessment, education/consultation, collaboration, discharge planning, linkage with community, and monitoring/follow up.
length of stay and readmission.
12 RCTs (2876 patients, mean age 72 y) met the selection criteria. 11 studies assessed a nurse led intervention, and 1 assessed a physician led intervention; in all studies, the control group received usual care. 6 of the 12 studies had high quality scores (⩾8 out of 11), and 6 had moderate quality scores (4–7 out of 11). Meta-analysis using a random effects model showed no difference between case management and usual care for length of stay (10 studies, n = 2666; pooled effect size 0.094, 95% CI −0.032 to 0.220) or readmissions (10 studies, n = 2603; odds ratio 0.87, CI 0.69 to 1.04). Sensitivity analyses showed that the effects of case management on length of stay did not differ based on study quality or number of intervention components. However, effectiveness did vary by patient diagnosis: case management reduced length of stay in patients with heart failure (pooled effect size 0.241, CI 0.012 to 0.470) but not in patients with stroke (pooled effect size –0.226, CI –0.542 to 0.089) or frail elderly people (pooled effect size 0.126, CI –0.073 to 0.324). Sensitivity analyses showed that the effects of case management on readmissions did not differ by study quality, number of intervention components, or patient diagnosis.
Hospital based case management did not reduce length of hospital stay or readmissions in adult inpatients.
The systematic review by Kim and Soeken highlights the weakness of the evidence base on the effectiveness of hospital based case management and raises some important questions for practitioners and researchers.
The definition of case management lacks clarity, and this is a major weakness in this area of research. Features of the operational definition of case management in this review are shared to some degree with other complex patient management interventions, and some of the studies have also been included in systematic reviews of protocol based care,1 discharge planning,2 and post-acute intermediate care.3 All of these reviews conclude that there is a lack of supporting evidence for effectiveness or non-effectiveness. The interventions are complex, comprising several components that may act both independently and interdependently.4 As discussed by Kim and Soeken, existing studies have not systematically explored the components or “active ingredients” of case management, which may include practitioner skills and behaviours, and service organisation and location. This limits our understanding of the implications of research findings as well as the selection of appropriate outcome measures; length of stay and readmission, although important, may not be the most appropriate outcomes to measure the effects of case management. Furthermore, the absence of a definition of the “usual care” with which case management is compared hinders interpretation.
Despite this lack of supporting evidence, models of case management continue to be developed and implemented. The review by Kim and Soeken offers little to practitioners and managers in terms of recommendations for practice. However, it does emphasise the need for further research, a crucial element of which would be determining the critical components of case management and how they relate to and affect each other.
For correspondence: MsY J Kim, School of Nursing, University of Maryland, Baltimore, MD, USA.
Source of funding: no external funding.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.